DAO Governance Is an Oligarchy: 1% Holds 90% of Voting Power
Empirical data across hundreds of proposals shows extreme centralization, with top voters dominating outcomes and participation below 20%. Hybrid pipelines and delegation markets are the real locus of control.
The data is clear: DAO governance lives on-chain but functions like an oligarchy. A study of over 370 proposals found that three to five voters can swing most outcomes, and 1% of tokenholders control 90% of voting rights[^2544]. A 2026 economics paper backs this up—the top decile of voters accounts for 76.2% of total realized voting power[^2545]. Participation is pathetic. Average turnout across major DAOs sits at 17%, and Uniswap routinely sees 2–3% against a 4% quorum[^2543]. In many DAOs, fewer than 10% of eligible token holders decide most outcomes[^2559].
The cost of voting makes this worse. Passing a proposal runs $500–$10,000 in gas and campaigning, which naturally tilts power toward large holders and professional delegates[^2550]. Delegation structures help, though. DAOs with formal delegation see 30–50% higher voting efficiency—faster decisions, more throughput[^2550][^2556]. This has spawned a new actor class: professional delegates who wield structural influence over upgrade votes and treasury allocation in protocols like Aave, Maker, and Arbitrum.
The ENS DAO crisis of June–July 2026 is a live stress test. Founder Nick Johnson used roughly 50% of active voting supply to block the renewal of the ENS Security Council on June 30[^2546]. The draft proposal that followed hardened the timelock cancel threshold from 4/8 to 5/8 and introduced legal appointment agreements and a removal mechanism[^2547]. This fits a broader pattern: hybrid governance pipelines—forum temperature check, gasless Snapshot vote, then on-chain binding vote with timelock—are now standard[^2553][^2548]. Tools like SafeSnap bind off-chain votes to on-chain execution through the Reality.eth oracle[^2554]. Practical control often sits in these off-chain social and delegation structures, not in raw token contracts.
Alternative voting mechanisms offer a way out. DAOs that switched from one-token-one-vote to quadratic or delegated models saw voter turnout jump from 2.8% to 11.4% and proposal implementation success scores rise 34%[^2555]. Yet adoption is limited—only 15 major DAOs use quadratic voting. The technical trend is toward hybrid off-chain/on-chain governance tied together with oracles and legal wrappers. DAO governance is becoming a specialized layer of protocol infrastructure that combines smart contracts, delegation markets, and compliance-focused councils rather than pure token-voting democracies.
It is a concentrated system where a tiny minority controls outcomes, and the real action happens off-chain through delegation and social coordination. Protocols should harden supermajority rules, adopt anti-whale voting primitives, and watch for governance attacks via self-delegation. The ENS drama is a warning for any protocol using multisig security councils over timelocked upgrade contracts.
Provenance ledger
8 span-verified · 9 web-cited8 claims below are locked to a verbatim span re-verified against the source. The remaining 9 are web citations: the URL was checked, but the excerpt is the researcher's summary and was not re-derived from the page. Citation markers in the text jump here.
[1] Across major DeFi DAOs, voter participation in governance proposals averaged 17% in 2025, with top DAOs exceeding 22% turnout on critical votes, while Uniswap DAO routinely recorded actual turnout of 2–3% against a 4% quorum requirement. web-cited
“DAO voter participation averages 17% in 2025, with top DAOs exceeding 22% for critical votes.” and “Uniswap actual turnout: 2–3% vs 4% required quorum.”[4][9]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[2] In a large-scale empirical study of more than 370 governance proposals and millions of on-chain events up to August 2024, as few as three to five voters were sufficient to sway most DAO proposals, with 1% of tokenholders controlling 90% of voting rights. span-verified
“Analysing more than 370 governance proposals and millions of on-chain events from inception to August 2024, we found substantial centralisation of voting power: as few as three to five voters were sufficient to sway most proposals… In reality, 1% of tokenholders control 90% of voting rights.”[10]
7e55051d4a421d386936dd29f435ffdac9f1aa73d38923b4fa156db9e93e67f3 [3] A 2026 economics paper on DAO governance finds that at the voting stage, the top decile of voters accounts for 76.2% of total realized voting power, highlighting extreme concentration despite nominally decentralized structures. web-cited
“At the voting stage, the top decile of voters accounts for 76.2% of the total realized voting power … we find low participation rates and high concentration in proposal and voting.”[6]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[4] ENS DAO’s governance crisis in June 2026 demonstrated that a single actor could exercise roughly 50% of active voting supply to block an on-chain proposal: ENS founder Nick Johnson used about half of the protocol’s active voting power to stop the renewal of the ENS Security Council on June 30, 2026. web-cited
“Nick Johnson voted against the on-chain renewal proposal using roughly 50% of the active voting supply…” and “The on-chain vote ended on June 30, 2026, after Johnson did not take part in the earlier off-chain Snapshot vote.”[13]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[5] Following the failed ENS Security Council renewal vote, a draft proposal on the ENS governance forum recommended replacing the council with eight members and raising the threshold to cancel timelocked proposals from a 4/8 to a 5/8 supermajority, coupled with a public mandate, removal mechanism, and legal appointment agreements for each member. web-cited
“The draft proposal would create an eight-member council with a 5/8 supermajority threshold for canceling timelocked proposals, up from 4/8, and would add a public mandate, a removal mechanism, and a legal appointment agreement for each member, according to the ENS governance forum.”[13]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[6] An ENS DAO temp-check proposal published on June 22, 2026, outlines a governance flow where Meta-governance reforms must pass through three stages: (1) forum discussion for 1–2 weeks, (2) off-chain Snapshot signal vote, and (3) a separate on-chain constitutional vote with higher quorum and approval thresholds for binding action; the temp check itself is explicitly non-binding. web-cited
“Her draft is a temp check, the first stage of ENS DAO's governance flow, which routes proposals through forum discussion before a Snapshot signal vote and, for binding action, a constitutional vote on-chain… The temp check is not a binding vote… constitutional changes of this scope require a separate on-chain vote with higher quorum and approval thresholds.”[11]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[7] Alex Van de Sande’s July 2026 ENS DAO Meta-governance draft proposes delegating 5 million ENS tokens from the DAO’s dormant community treasury to individual participants, with constraints that delegates cannot sell the tokens, an undelegation trigger after six months of inactivity, a potential additional 5 million tokens next year, and a full sunset after two years. web-cited
“Alex Van de Sande… proposed… that the ENS DAO delegate 5 million ENS tokens from its dormant community treasury to individual participants… participants would not own or be able to sell the delegated tokens… floated adding another 5 million tokens next year, an undelegation trigger after six months of inactivity, and a full sunset of the arrangement after two years.”[2]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[8] A 2026 survey of DAO governance reports that the average cost to pass a proposal—including gas and campaigning—ranges from $500 to $10,000, while DAOs that use formal delegation structures achieve 30–50% higher voting efficiency (measured by reduced decision time and increased throughput). span-verified
“The average cost to pass a proposal (including gas, campaigning) is $500-$10,000 in many DAOs… Delegation structures shorten decision time and increase throughput. DAOs with delegations had 30-50% higher voting efficiency.”[4]
f836496a2e7a998f4a015a59114c61ba1e40d46f025ef164954020839cb49c7b [9] Live governance trackers for DeFi DAOs in 2025–2026 show concrete voting distributions; for example, one Uniswap proposal to deploy $500k in UNI incentives passed with 11,543,905.16 votes (56.45%) in favor and 8,906,122.12 votes (43.55%) against, while another governance decision passed with 533,936.14 votes (100.00%) for and 0 against. span-verified
“Deploy $500k in UNI Incentives:11,543,905.16 (56.45%)… Do not deploy incentives:8,906,122.12 (43.55%)… ‘For:533,936.14 (100.00%)’.”[1]
f10e93775bf2cdb02cf11da9e3cf75b1636d67a79a63249d995e3d89e25b7c90 [10] Concrete treasury and parameter decisions in smaller DAOs continue to be made via simple yes/no on-chain votes: in BTF’s Governance DAO, a proposal to increase the transaction tax burn rate from 10% to 12% for 30 days (ID #0013) passed with 75% Yes and 25% No by January 15, 2026, while a proposal to reduce flexible staking pool APY was rejected with 92% Against. web-cited
“Increase Burn Rate to 12%… increase the transaction tax burn from 10% to 12% for a period of 30 days. Yes75% No25% Passed ID: #0013 Ended Jan 15, 2026… Reduce Staking APY… Result: Rejected92% Against.”[5]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[11] Hybrid off-chain/on-chain governance pipelines have become standardized in DeFi DAOs: typical flows consist of (1) an off-chain forum temperature check, (2) a gasless off-chain Snapshot signal vote, and (3) an on-chain proposal with a 3–10 day voting period, followed by a 24–72 hour timelock queue for execution; many protocols enforce minimum quorum and proposal thresholds. span-verified
“DeFi governance proposals typically follow a standardised process with several stages: 1. Temperature check (off-chain)… 2. Snapshot vote (off-chain)… 3. On-chain proposal… 4. Voting period: Typically 3–10 days… 5. Timelock queue: If passed, the proposal enters a timelock delay (24–72 hours)… Most protocols require a minimum quorum.”[12]
6a26463b6cf11d67ccf7b3b3f2dcfb7900f199115ad9214f1eec729d3eca5725 [12] Tools like SafeSnap bind off-chain Snapshot votes to on-chain treasury execution by using the Reality.eth crowdsourced oracle to verify the outcome of the off-chain vote before authorizing multisig transactions. web-cited
“Snapshot proposals are not binding… SafeSnap uses the Reality.eth crowdsourced oracle to verify the outcome of the off-chain vote.”[15]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[13] By 2025–2026, alternative voting mechanisms such as quadratic voting, conviction voting, and delegated proof-of-stake had been adopted in production by multiple DAOs; one survey reports that DAOs switching from one-token-one-vote to delegated or quadratic models increased voter turnout from 2.8% to 11.4% on average and boosted proposal implementation success scores by 34%. span-verified
“Quadratic voting… has been adopted by 15 major DAOs, including Gitcoin and Optimism… DAOs that switched from one-token-one-vote to delegated or quadratic models saw voter turnout increase from 2.8% to 11.4% on average, with proposal quality scores (measured by implementation success) rising 34%.”[3]
a93974ad7b43aa82d2804371a55dc7c3bada27e83f5fe2d6e19c91060d8e0c58 [14] A 2026 overview of DAO activity notes that DAOs with formal delegation structures experience 30–50% higher voting efficiency and that major DAOs show proposal passage rates above 70%, while smaller DAOs exhibit lower passage rates due to limited participants and higher relative coordination costs. span-verified
“Delegation structures shorten decision time and increase throughput. DAOs with delegations had 30-50% higher voting efficiency… For major DAOs, the passage rate of proposals is often > 70%; smaller DAOs show lower passage rates due to fewer participants.”[4]
831d0a6418a52a4f103058e1e0d28a0d9d45ae2eeeb67284c0281d5e72fdd4f0 [15] On-chain participation requirements in DeFi governance remain materially different from off-chain signalling: voters must hold governance tokens in a self-custody wallet, connect to the protocol’s governance portal, pay a gas fee for each on-chain vote, and meet protocol-specific proposal thresholds (minimum token balance) for submitting upgrades; voting options are typically For/Against/Abstain. span-verified
“Participating in on-chain governance requires holding governance tokens in a self-custody wallet… visit the protocol's governance portal… On-chain voting requires a small gas fee… Requires meeting a proposal threshold (minimum token balance to submit)… Token holders vote For/Against/Abstain.”[12]
d9ee7b6385561776229ee643115c8d6eb1e76cbc748625a0048f62d9abaf285c [16] Recent ENS governance disputes show meta-governance risk: delegates publicly labeled a June 2026 ENS Foundation authority-handover temp-check as a ‘governance attack’ after lead developer Nick Johnson self-delegated a large ENS voting weight to support the measure, effectively controlling the outcome window. web-cited
“ENS DAO delegates publicly rejected the ENS Foundation's authority-handover proposal Tuesday, branding the plan a governance attack after lead developer Nick Johnson self-delegated a meaningful share of voting weight in the same window.”[8]
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[17] Average DAO turnout among decision-makers in 2026 hovers around 20%, with some analyses indicating that less than 10% of eligible token holders decide most outcomes and that, in many DAOs, 1% of tokenholders effectively control 90% of voting rights. span-verified
“Average turnout among decision-makers hovers around 20%, and in many cases only one in ten actually uses their right to vote… In reality, 1% of tokenholders control 90% of voting rights.”[10]
97e427451b0b812c511df4a86cc65b1203ffb4fa5403dbfd2c5e6997ad4ab448 Sources
- https://sqmagazine.co.uk/decentralized-autonomous-organizations-statistics/
- https://forklog.com/en/good-or-nothing-how-daos-are-faring-in-2026/
- https://www.aeaweb.org/conference/2026/program/paper/D55S6dSe
- https://www.cryptopolitan.com/ens-dao-governance-reform-proposal-drama/
- https://thedefiant.io/news/defi/ens-dao-temp-check-empowering-ens-foundation-treasury-handover
- https://thedefiant.io/news/defi/ens-co-founder-proposes-delegating-5m-ens-tokens-to-reform-dao-governance
- https://www.tokendataview.com/governance
- https://www.btftech.io/Governance.aspx
- https://cryptogrowsnews.com/learn/dao-governance-voting-proposals-delegation/
- https://cointelegraph.com/news/gnosis-and-snapshot-create-tool-to-bind-defi-governance-votes-on-chain
- https://pen-caforr.org/2026/04/15/dao-governance-2026-hybrid-models-legal-wrappers-and-the-end-of-token-voting/
- https://thedefiant.io/newsletter/defi-daily/ens-dao-governance-attack-ef-restructuring-2026-06-23