Proposal Authors: Abdullah Umar (Michigan Blockchain) & Cameron O'Donnell (Consensys)
For clear graphs and visuals, visit the original forum post at https://gov.uniswap.org/t/rfc-lower-onchain-proposal-threshold/22429
NOTE: PGov is proposing as one address can only submit one vote at a time
We are proposing to lower the current proposal threshold (PT) from 2.5M UNI to 1M UNI with the intention of broadening the delegate base capable of submitting onchain proposals, thereby increasing delegate engagement. This request aligns with the increasing maturity and expanding influence of the Uniswap protocol.
Proposal threshold and voting quorum are two vital parameters for ensuring that a DAO is functioning properly.
It’s important to put quorum and PT in the context of the governance token’s circulating and total supply, allowing us to normalize parameters across multiple DAOs.
At inception, Uniswap DAO’s governance framework required that “1% of all UNI must be cast in favor to submit a proposal, and 4% in order to pass a vote.” In terms of tokens, this meant that the proposal threshold was 10M UNI, and the quorum was 40M UNI. These percentages use the total supply of UNI (1B), not the circulating supply, as the denominator. Acknowledging the restrictive nature of this parameter, the proposal threshold was quickly reduced from 10M UNI to 2.5M UNI. This alteration took place over two years ago.
Since then, we have seen Uniswap governance mature and onboard new delegates. And of the current delegate cohort, ~33 have the ability to post onchain proposals. Unfortunately, many of the entities with enough voting power are scarcely involved in DAO forums and votes. Lowering the PT will help reduce the proposal creation barrier for delegates who are actually active in the DAO. This, in concert with recent initiatives to increase active delegates’ voting power, will aid in further decentralizing Uniswap decision making.
Most DAOs set decentralization as a north star metric that’s achieved over time. Initially, protocols tend to start out more centralized, and their governance parameters are indicative of their degree of maturity. A less mature protocol should start off with a high PT, for example. Core contributors and a circle of trusted entities should primarily be involved in proposal creation to ensure that the protocol grows legs. At this time, the token value tends to be lower as well, along with a relatively insignificant treasury. Therefore, the respective PT and quorum should be high enough to prevent nefarious actors from depleting reserves and collapsing the protocol before it reaches any sort of prominence. Responsibilities and workloads should, however, become dispersed across more and more contributors as the protocol grows.
A growing protocol typically correlates with an increasing token price. So, the economic cost of creating proposals and reaching quorum simultaneously increases. This means that the DAO must over time reduce such parameters. When Uniswap lowered its PT to 2.5M, the economic cost of putting a vote onchain was still upwards of $50M. At that time, the UNI token was less than a year old, and governance was extremely nascent, so caution remained the best policy.
In the past year, the UNI token went as low as ~$3.50. With a PT of 2.5M, the economic cost of putting a vote onchain hit ~$8.75M. And yet, the DAO wasn’t spammed with votes. It’s therefore clear that we should not be worried about an influx of useless proposals upon a reduction of the PT. Plus, as Uniswap v4, X, and perhaps a fee switch are introduced, along with a turn in economic conditions, the token value will likely increase. Hence, a reduction of the PT seems to be inevitable. It’s a matter of when–not if.
We are proposing to reduce the PT to 1M UNI. That is in and of itself not an insignificant amount. If an entity is able to purchase that much UNI or attain an equivalent amount via delegation, it’s likely that they are a trusted and competent party. Okay, perhaps that’s too optimistic. So let’s ask–if the entity is nefarious, what’s the worst that can happen? Answer: you get a bad proposal that never reaches quorum. If a nefarious proposal does reach quorum, we’ve got other problems. But it’s unlikely due to the massive capital requirement.
At the end of the day, Uniswap’s 40M quorum is enough to prevent malice, regardless of how low the PT is. Again, the quorum is a gauge for how difficult it is to pass a proposal, and the PT is a gauge for determining how freely people are allowed to create proposals. Suppressing the latter is counterproductive to growing and decentralizing the DAO.
As of today, ~33 delegates are able to cross the 2.5M PT. Decreasing the PT to 1M increases that number to ~47 (again, with a small margin of error). This is effectively a centralization choke point, where proposers must turn to a select few delegates who may or may not be in support of sponsoring a proposal.
| Wallet | Delegated UNI | Entity | Active? | % Vote | Proposals Posted | | ------------------------------------------ | ------------- | -------- | --------- | ------ | ---------------- | | 0x683a4f9915d6216f73d6df50151725036bd26c02 | 7,501,052.07 | Gauntlet | Yes | 90% | 1 | | 0x553f674dd7d102ad79c644103974a1cc53b62ac2 | 5,000,184.21 | Teemu | Yes | 80% | 1 | | 0xa2bf1b0a7e079767b4701b5a1d9d5700eb42d1d1 | 5,000,021.65 | Ken (UF) | No | 0% | 1 | | 0x13bdae8c5f0fc40231f0e6a4ad70196f59138548 | 3,851,337.13 | Michigan | Yes | 100% | 6 | | 0x070341aa5ed571f0fb2c4a5641409b1a46b4961b | 3,000,601.96 | Penn | Yes | 80% | 2 | | 0xa6e8772af29b29b9202a073f8e36f447689beef6 | 2,502,935.57 | GFX Labs | Yes | 90% | 5 | | 0xed11e5ea95a5a3440fbaadc4cc404c56d0a5bb04 | 2,500,839.08 | she256 | Yes | 80% | 2 | | 0x5d8908afee1df9f7f0830105f8be828f97ce9e68 | 2,500,667.32 | Argent | No | 10% | 1 | | 0x7ae109a63ff4dc852e063a673b40bed85d22e585 | 2,500,030.00 | Berkeley | No | 20% | 1 | | 0xdc1f98682f4f8a5c6d54f345f448437b83f5e432 | 2,500,010.00 | Columbia | Sometimes | 60% | 1 | | 0xb7771f70633c7e54e61dd38d01c26da0e86be1a5 | 2,500,000.00 | Chicago | No | 10% | 1 |
The above table shows the number of proposals created by delegates with large amounts of voting power today. This does not include proposals posted by wallets who received temporary delegation. “% vote” is participation in the past 10 onchain votes. Proposals posted is the number of proposals created onchain since the Governor Bravo upgrade (i.e. since Q3 2021).
If the few delegates who have enough voting power are disinterested in a proposal, it may never go onchain, even if the proposal passes the temperature check phase. One option that proposers currently have is the ability to turn to the Uniswap Foundation, who can then delegate the proposer 2.5M UNI–however, that creates a reliance on the Foundation.
Comparing governance parameters and token ratios between DAOs is a tough exercise since most DAOs are not mature and have different objectives. The reality is that Uniswap sits in a position to set standards for other DAOs, and we certainly don’t have all of the answers, so the best that we can do is experiment. Regardless, below are some data illustrating where the Uniswap DAO sits compared to the industry.
The forum post contains graphs comparing the Proposal Threshold (PT), Total Supply (TS), Circulating Supply (CS), and Quorum (Q) ratios of 7 DAOs on Ethereum.
Uniswap currently has the highest Q/TS ratio, along with Compound, from the given sample. It’s also on the higher end when looking at Q/CS. This means that relative to other DAOs, it takes more votes to pass a proposal. Since we are not proposing to lower the quorum, we can rest assured that the barrier for a vote to pass remains high enough to prevent security issues.
The PT/TS and PT/CS ratios, however, are on average lower for Uniswap relative to the sample. So, relatively speaking, proposal creation is already easier for Uniswap, and the PT reduction to 1M will make it even easier. This may simply indicate that DAOs overall are too constrictive. But a relative comparison here doesn’t tell the whole story. Even though the PT/TS and PT/CS are lower for Uni DAO, the absolute dollar amount that it takes to post a proposal is by far still the highest for Uniswap. The below table shows that the PT in dollar terms is $15.6M, with Aave coming in second at $7.2M and Compound third with only $1.3M. The dollar cost for creating a proposal on Uniswap is ~4.5 times higher when compared to the sample.
| Protocol | Economic Cost of PT | FDV | Mcap | Quorum | Treasury | | ----------------- | ------------------- | ------------------ | ----------------- | --------------- | ----------------- | | Uniswap | $15,575,000.00 | $6,230,000,000.00 | $4,695,966,335.41 | $249,200,000.00 | $2,100,000,000.00 | | Aave | $7,200,000.00 | $1,440,000,000.00 | $1,317,033,360.00 | $28,800,000.00 | $134,000,000.00 | | Compound | $1,275,000.00 | $510,000,000.00 | $349,660,335.00 | $20,400,000.00 | $114,000,000.00 | | ENS | $890,000.00 | $890,000,000.00 | $229,470,773.70 | $8,900,000.00 | $597,000,000.00 | | Hop | $50,000.00 | $50,000,000.00 | $3,761,124.15 | $150,000.00 | $25,800,000.00 | | Arbitrum Core | $1,070,000.00 | $10,700,000,000.00 | $1,364,250,000.00 | $130,465,100.00 | $- | | Arbitrum Treasury | $1,070,000.00 | $10,700,000,000.00 | $1,364,250,000.00 | $78,281,200.00 | $3,900,000,000.00 | | Gitcoin | $1,090,000.00 | $109,000,000.00 | $66,341,074.39 | $2,725,000.00 | $64,300,000.00 |
Above are the PT, TS, CS, and Quorum in dollar terms (with prices as of December 11, 2023)
Today, the role of a delegate is pretty passive. All that’s required is voting yes, no, or abstain on proposals–unfortunately, many delegates don’t even fulfill that responsibility. Some delegates may take it a step further and contribute to discussions. And a select few with enough voting power may also create onchain proposals. The ability to post a vote onchain on behalf of another entity is currently a job that’s not given enough weight. Let’s categorize this relationship further.
Delegate: an entity with delegated voting power
Proposer: the entity that authors/initiates a proposal
Sponsor: the facilitator of the governance process, taking a proposal from early-stage to onchain execution. Delegates with enough voting power to meet the PT are considered sponsors
An ideal DAO ecosystem is one where we have a handful of entities with the ability to sponsor proposals. A rolodex of sponsors should enable proposers to pick and choose which sponsors to approach for facilitating the governance process. Ideally, the proposer seeks to team up with a sponsor who holds similar values so that both parties are motivated to see the proposal through. Sponsors can also help co-author proposals, depending on how involved they’d like to get. A variety of sponsors, each with their own biases and areas of focus, allows for healthy debate and diversity of opinion. The sponsors should therefore be stewards of the DAO and not merely wallets who happen to have a bunch of voting power. An interconnected system like this is also superior to simply asking the Uniswap Foundation for temporary delegation as it’s less communal than that of sponsor–proposer partnerships.
Beyond the collectivistic aspect, sponsors, if they so choose, can take their proposals from RFC to onchain vote independently. This autonomy is something that’s currently scarce, which may be a contributing factor behind a lack of development in DAOs.
Reducing the PT certainly helps lower friction in the mentioned areas, ideally leading to a more democratized and active Uniswap DAO.
| Protocol | Proposal Threshold | Total Supply | Circulating Supply | Quorum | PT/TS | PT/CS | PT/Q | Q/TS | Q/CS | | -------- | ------------------ | ---------------- | ------------------ | ------------- | ------ | ------ | ------ | ------ | ------ | | Uniswap | 1,000,000.00 | 1,000,000,000.00 | 753,766,667.00 | 40,000,000.00 | 0.100% | 0.133% | 2.500% | 4.000% | 5.307% |
Above would be the new ratios upon lowering the PT to 1M