Analysis of the governance of different DAOs.
Uniswap DAO
Uniswap governance is token-based: UNI holders (often via delegates) discuss proposals on the forum, run off-chain “temperature checks” (Snapshot), then submit executable on-chain proposals. On-chain votes pass if “for” votes exceed “against” and reach quorum (commonly cited as 40M UNI in favor), followed by a timelock before execution. (Uniswap Docs)
On decentralization: UNI’s initial allocation gave 60% to the community but ~40% to insiders (team/investors/advisors), which can concentrate voting power. (Uniswap Labs) Programs like treasury delegation can further centralize influence by amplifying select delegates. (The Block)
One broader limitation: many DAOs govern parameters and treasuries, but depend on off-chain teams, legal entities, and privileged actors (front-ends, dev roadmaps, “delegate classes”), limiting true autonomy even when voting is on-chain.
Aave DAO
Aave governance runs in stages: ideas start on the Aave Governance Forum, move to a Temp Check on Snapshot (non-binding), then an ARFC (final comments) with another Snapshot vote, and finally an on-chain AIP vote that—if passed—executes after a timelock (typically via the “Short Executor”; governance-changes use a slower “Long Executor”). (aave.com)
On decentralization: while voting is open to AAVE holders/delegates, voting power is meaningfully concentrated—top 100 holders control ~12.45M AAVE of 16M supply (~78%), suggesting oligarchic influence risks. (Ethereum (ETH) Blockchain Explorer)
One broader limitation: many DAOs still rely on off-chain governance plumbing (forums, delegate politics, and emergency/guardian roles), so “autonomous” protocols can revert to trusted actors during crises. (vote.onaave.com)
Yearn Finance DAO
Yearn governance is largely off-chain, token-weighted: proposals (YIPs) are discussed in the forum, then veYFI holders vote on Snapshot; proposals passing with >50% support are executed in practice by a 9-signer multisig (6/9 threshold). (Yearn Docs)
On decentralization: Yearn’s launch was unusually “fair” (no premine/VC/team allocation; early supply distributed via liquidity mining). (Yearn Docs) Still, current ownership is concentrated: one recent rich-list snapshot shows the top 10 addresses ~38% and top 100 ~84% of YFI. (CoinLore) That concentration plus multisig execution means it’s partly decentralized, but not fully.
One broader limitation: many DAOs can vote, but execution and operations still depend on multisigs, keyholders, and off-chain coordination—so autonomy is social/institutional, not purely on-chain. (Yearn Docs)
MakerDAO
MakerDAO governance blends off-chain deliberation with on-chain control. MKR holders vote on Governance Polls (signaling/parameter preferences) and binding Executive Votes that, when passed, enact “spells” to change protocol parameters (e.g., collateral rules, rates) via the Maker Governance Portal. (community-portal-staging.makerfoundation.com)
On decentralization: decision power is meaningfully concentrated—one recent rich-list snapshot shows the top 10 MKR addresses hold ~33.66% and the top 100 ~83.43% of supply, implying whale/delegate dominance risk despite open participation. (CoinLore)
One limitation of DAO frameworks: even with on-chain votes, agenda-setting, delegation, and operational execution often centralize around a small set of actors, so autonomy remains partly social rather than purely programmatic.
SushiSwap DAO
SushiSwap governance is token-weighted: proposals are discussed by the community, then voted on via Snapshot using SUSHI. Sushi’s own docs note that only proposals posted to Snapshot by the CORE team are considered binding if they pass quorum, and votes typically cover tokenomics, major structural changes, and use of DAO/dev-fund wallets. (docs.sushi.com)
On decentralization: this process has clear centralization chokepoints (proposal “gating” by the core team, plus reliance on multisig-style execution in practice). Recent reporting around treasury-control proposals and governance controversies reinforces that power can concentrate around a small operational group. (The Block)
One broader limitation: many DAOs can vote, but credible, autonomous execution still depends on trusted operators (multisigs/teams) and off-chain coordination—so “autonomy” remains partly social, not purely programmatic.
Aragon DAO
Aragon’s governance has historically been ANT token–weighted: proposals (AGPs) were put to ANT holders, with 1 ANT = 1 vote (often via common DAO tooling such as on-chain voting modules / Snapshot-style signaling, depending on the era and product stack). (Aragon Legacy Documentation)
On decentralization: Aragon is a cautionary tale. In Nov 2023, the Aragon Association announced it would dissolve and make ~86,343 ETH (~$155M then) redeemable by ANT holders—after governance tensions including a described attempted “51% attack” to capture the treasury. That episode suggests meaningful central points of control remained around the Association and treasury custody, despite token voting. (Blockworks)
One broader limitation: many DAOs can vote, but legal wrappers, treasuries, and emergency powers often sit off-chain (or in multisigs), so “autonomy” breaks exactly when high-stakes decisions hit.
Decentraland DAO
Decentraland governance is hybrid: proposals are created and voted on off-chain in Decentraland’s Snapshot space (surfaced via governance.decentraland.org), using Voting Power (VP) derived from holdings of MANA, NAMES, and LAND (e.g., 1 MANA = 1 VP; 1 NAME = 100 VP; 1 LAND = 2000 VP). Passed binding votes are then enacted on-chain through Aragon by a DAO Committee multisig, with oversight from a Security Advisory Board and delays/timelocks. (Decentraland Documentation)
On decentralization: token/asset-weighted VP plus custody/execution by committees means decentralization is partial. Holder concentration is high—one recent rich-list snapshot shows top 10 MANA addresses ~43.66% and top 100 ~73.17% of supply—so a small set can dominate outcomes. (CoinLore)
One broader limitation: many DAOs still rely on multisigs/boards and off-chain processes for execution and security, so “autonomy” breaks down into trusted governance when stakes are high. (Decentraland Documentation)
Compound DAO
Compound governance is on-chain and token-weighted: COMP holders (or delegates) create proposals through Governor Bravo, vote during a set voting period, and successful proposals are queued into a Timelock and then executed on-chain (anyone can execute after the delay). Proposals can change risk parameters, add markets, or upgrade contracts. (Compound)
On decentralization: while open in theory, COMP ownership is concentrated. A recent rich-list snapshot shows the top 10 addresses hold ~36.53% and the top 100 ~77.41% of COMP. (CoinLore) Tokenomics also allocated large shares to insiders (about ~24% shareholders/investors and ~22% founders & team, alongside ~42% to protocol users). (Tokenomist)
One broader limitation: many DAOs have on-chain votes but off-chain influence and plutocracy (whales, delegates, insiders) mean “autonomy” often reflects capital concentration rather than collective self-rule.
Curve DAO (CRV / veCRV)
Curve governance is vote-escrowed: users lock CRV to receive veCRV, and voting power scales with both amount locked and lock duration. veCRV holders vote on on-chain DAO proposals and, crucially, on gauge weights that direct where CRV emissions go (a major economic lever). (docs.tally.xyz)
On decentralization: allocation is “community-heavy” over the long run (57% emissions + 5% early users + 5% reserve ≈ 67% community), but insiders are still large (~26.43% core team + 3.57% investors). (resources.curve.finance) Recent moves to add emergency multisig powers also show reliance on trusted actors in practice. (Outposts)
One limitation of DAO frameworks: safety requires escape hatches (multisigs/guardians), which undermines “true” autonomy exactly during crises.
Lido DAO (LDO / stETH ecosystem)
Lido governance is token-based: LDO holders propose and vote through a regular flow that uses established tools (research/forum → off-chain signaling → on-chain actions), with newer designs adding Dual Governance safeguards tied to stETH holders. (Lido)
On decentralization: voting is open, but ownership is concentrated. Etherscan’s holder chart shows the top 100 LDO holders control ~84.05% of supply, and several top holders are large entities/exchanges or labeled Lido-related vesting addresses—suggesting meaningful centralization risk in practice. (Ethereum (ETH) Blockchain Explorer) Lido also formalizes “ops multisigs” policies, underscoring reliance on committees for execution/security. (Lido Governance)
One limitation of DAO frameworks: even with on-chain voting, execution and crisis response often depend on multisigs/committees, so autonomy degrades into trusted governance when stakes are high.
ENS DAO (Ethereum Name Service)
ENS governance is token-based and delegate-heavy. Proposals are developed on the ENS forum, refined into governance proposals, and then—when ready—submitted as Executable Proposals for an on-chain vote using common governance UIs (e.g., Agora/Tally). If passed, execution happens through the DAO’s on-chain governance system (typically via timelock-style execution). (docs.ens.domains)
On decentralization: the initial distribution emphasized community/airdrop and a DAO treasury, but current holdings appear highly concentrated; one recent rich-list snapshot estimates the top 10 addresses hold ~84.72% of ENS. (ens.mirror.xyz)
One limitation of DAO frameworks: even with on-chain voting, delegation + whale concentration can centralize agenda-setting and outcomes, weakening “true autonomy.”
10 other well-known DAOs by treasury size
- Optimism Collective — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- Arbitrum DAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- Mantle DAO (ex-BitDAO) — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- GnosisDAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- ENS DAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- The Graph DAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- MakerDAO (Maker Protocol) — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- Lido DAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- dYdX DAO — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
- Polygon (protocol governance / community treasury) — DeepDAO treasury leaderboard (AUM ranking) (DeepDAO)
The DeepDAO leaderboard page is the canonical place to click through (it updates continuously). (DeepDAO)