On-chain governance is a model for making protocol-level decisions – such as parameter changes, upgrades, treasury allocations, and new feature implementations – through a formal, transparent voting process recorded directly on the blockchain. Token holders (or delegated representatives) submit proposals, discuss them publicly, and then cast votes using their governance tokens. While some proposals are automatically executed by smart contracts, many protocols use hybrid models where approved results are manually implemented via multisig wallets. On-chain governance contrasts with “off-chain governance” (such as Bitcoin’s social consensus model, where changes are informally discussed and merged by developer teams), offering greater formal transparency and automation, though as the examples below illustrate, on-chain execution doesn’t automatically prevent contested or controversial outcomes. Major implementations include Compound’s Governor system, Uniswap’s Governance framework, Optimism’s two-chamber model (Token House and Citizens’ House), and the Sky (formerly MakerDAO) protocol. On-chain governance is a defining feature of decentralized autonomous organizations (DAOs) but faces persistent challenges around voter apathy, plutocracy (token concentration), and governance attacks.
A History of On-Chain Governance
| Date | Event |
|---|---|
| 2013-2014 | The concept of a “decentralized autonomous organization” is described in Ethereum’s early whitepaper materials by Vitalik Buterin, as one of several potential applications for a general-purpose smart contract platform |
| 2016 | “The DAO,” a specific and highly-funded venture-capital-style investment DAO built on Ethereum, launches as one of the first large-scale on-chain governance experiments; it is hacked in June for roughly $50-60 million worth of ETH, ultimately leading to Ethereum’s contentious hard fork (creating the ETH/ETC split) that July |
| 2018 | Tezos launches with on-chain protocol governance built into its base layer, allowing token holders/bakers to vote on protocol upgrades |
| 2020 | Compound launches the COMP token and its Governor system in June, becoming one of the first major DeFi protocols with formal on-chain governance |
| 2020 | MakerDAO’s governance process, including on-chain executive votes, oversees a protocol with a substantial and growing amount of DAI in circulation and associated collateral parameters |
| 2020 | Yearn Finance and several other protocols adopt governance frameworks conceptually similar to Compound’s, contributing to a broader wave of DeFi governance token launches |
| 2021 | Uniswap governance requires a high UNI threshold to submit proposals; persistently low voter participation becomes a recognized structural issue |
| 2021 | ConstitutionDAO raises roughly $47 million in about a week via crowdfunded on-chain contributions, aiming (unsuccessfully) to bid on an original copy of the U.S. Constitution at a Sotheby’s auction |
| 2022 | Optimism launches its Token House and Citizens’ House bicameral governance structure, an early prominent example of a two-chamber DAO governance model |
| 2022 | The Beanstalk protocol suffers a flash-loan-enabled governance attack in April, in which an attacker borrows enough governance tokens to pass a malicious proposal in a single transaction and drains roughly $182 million |
| 2023 | The Arbitrum DAO forms following the March ARB airdrop, governing a treasury in the billions of dollars; its early AIP-1 controversy (an attempt to allocate 750 million ARB to the Arbitrum Foundation without full DAO approval) becomes an early, widely-cited case study in DAO governance friction |
| 2024 (July) | A widely-discussed Compound governance controversy unfolds: a delegate group (“Golden Boys,” led by a pseudonymous whale known as “Humpy”) passes Proposal 289 by a narrow margin (roughly 51-52% in favor), reallocating about $24 million worth of COMP from the DAO treasury toward a yield vault the group controlled. The proposal was not defeated by opposing votes – it passed – and was only reversed after Humpy voluntarily agreed to cancel it in exchange for a negotiated settlement involving a new staking product for COMP holders |
| 2024-2026 | On-chain governance design continues evolving, with delegate systems, veTokenomics-style time-locked voting, and hybrid multisig/on-chain combinations becoming increasingly common across protocols |
How On-Chain Governance Works
| Governance Model | Description | Example |
|---|---|---|
| Token-weighted voting | 1 token = 1 vote; large holders have proportionally large influence | Uniswap, Compound |
| ve-Tokenomics | Time-locked tokens receive boosted voting power | Curve (veCRV), Balancer |
| Delegation | Token holders assign their voting power to a trusted delegate | Optimism’s Citizens’ House and various delegate-based systems |
| Multi-sig | A small, trusted group executes decisions on behalf of the community | Common in earlier-stage or smaller protocols |
| Bicameral | Two distinct voting chambers with separate roles | Optimism (Token House + Citizens’ House) |
On-Chain Governance Explained Simply
Democracy on a blockchain, with caveats: Instead of a CEO or board unilaterally deciding protocol changes, token holders vote. Approved proposals can execute automatically via smart contracts in some systems – though as the Compound Proposal 289 episode shows, “the vote passed” doesn’t always mean the outcome reflects broad community consensus, since voting power itself can be concentrated.
Proposal, Discussion, Vote, and Execute: The lifecycle is generally transparent and public. Anyone can read the proposal, see how addresses voted, and verify that execution matches the approved code.
Delegation: Most people don’t vote directly on every proposal, since doing so is time-consuming and often requires real technical expertise. Delegation lets you assign your voting power to a trusted party who votes on your behalf, conceptually similar to proxy voting in traditional corporate governance.
The plutocracy problem: In “one token = one vote” systems, holders with a large share of tokens control a correspondingly large share of voting power. In practice, VCs, early investors, or accumulating whales with large token allocations can meaningfully influence or even dominate protocol decisions – a dynamic often described as a form of digital plutocracy.
The apathy problem: Many DeFi governance votes see quite low overall token participation. When participation is low, proposals can pass with only a small fraction of total token holders actively engaged, which can make governance capture by a motivated, well-resourced minority easier.
Real-World Examples of On-Chain Governance
| Scenario | Implementation | Outcome |
|---|---|---|
| Uniswap fee switch debate | Multiple proposals to activate protocol-level trading fees, requiring a high UNI threshold to formally propose | The debate has spanned years with various proposals and votes, shaped by both regulatory caution and quorum/participation challenges; the specifics have evolved over time and are worth checking current Uniswap governance forums for the latest status |
| Compound Proposal 289 (2024) | A delegate group led by “Humpy” accumulated enough COMP to pass a proposal reallocating treasury funds to a vault they controlled | The proposal passed by a narrow margin (roughly 51-52%) in July 2024, despite community objections; it was not defeated by a governance vote, but was later voluntarily rescinded by Humpy in exchange for a negotiated staking-product settlement – the episode became a widely-cited case study in low-quorum governance vulnerability |
| Sky (formerly MakerDAO) Endgame plan | A multi-year restructuring of MakerDAO’s governance and branding, including a shift toward SubDAOs and a rebrand of DAI/MKR-adjacent products | Approved through on-chain governance processes; represents one of the larger voluntary governance restructurings undertaken by a major DeFi protocol |
| Arbitrum DAO AIP-1 controversy (2023) | A proposal to allocate 750 million ARB to the Arbitrum Foundation was interpreted by parts of the community as circumventing full DAO approval | Significant community pushback followed; the episode became an early, frequently-referenced example of the tension between foundation discretion and DAO governance authority in a newly-launched DAO |
| Optimism Citizens’ House | Non-transferable “Citizen” NFTs allocated to community members, who vote on public goods funding allocations | Retroactive public goods funding (RetroPGF) rounds have distributed substantial sums through this non-token-weighted chamber, intended as a partial counterbalance to plutocratic token-based voting |
Advantages of On-Chain Governance
| Advantage | Detail |
|---|---|
| Transparency | Votes, proposals, and (where applicable) executions are publicly verifiable on-chain |
| Automation | Approved proposals can self-execute in some systems, without needing to rely solely on a development team’s discretion |
| Permissionless participation | Any qualifying token holder can generally vote without needing approval from a central authority |
| Resistance to unilateral override | No single party can simply block a legitimately passed proposal from executing, in systems designed this way |
| Record-keeping | Governance actions are generally permanently recorded, supporting an auditable history |
| Composability | Governance systems can integrate with DAOs, multisigs, and other on-chain mechanisms |
Disadvantages and Risks
| Risk | Explanation |
|---|---|
| Voter apathy | Participation is often quite low in practice, making governance more susceptible to capture by smaller, organized, motivated groups |
| Plutocracy | Token concentration (among VCs, whales, or coordinated groups) can create a de facto oligarchic governance structure even within a nominally democratic framework |
| Governance attacks | Malicious or self-interested actors can acquire governance tokens (sometimes via flash loans, sometimes via straightforward market accumulation) to pass proposals that benefit themselves – and, as Compound’s Proposal 289 demonstrates, such proposals can genuinely pass rather than simply being attempted and failing |
| Slow iteration speed | Formal proposal-vote-timelock cycles can take days to weeks, which may be too slow for urgent emergency responses |
| Technical barrier | Complex proposals often require technical expertise that many token holders lack, which can lead to low-information voting even among engaged participants |
| Regulatory uncertainty | Regulatory scrutiny of governance tokens (including questions about potential securities classification) creates ongoing uncertainty for active governance participants in some jurisdictions |
Governance Health Metrics
- Monitor voter participation rates – low participation, however “low” is defined for a given protocol’s scale, generally signals elevated capture risk
- Track delegate concentration: a small number of delegates controlling a majority of delegated voting power is a meaningful plutocracy/capture risk signal, as Compound’s Proposal 289 illustrated in practice
- Review timelock duration: longer timelocks generally provide more opportunity to react to a rushed or malicious proposal before it executes, though excessively long timelocks trade off against responsiveness
- Check quorum thresholds: quorums set too low make attacks easier; quorums set too high can make it difficult to pass even legitimate, broadly-supported proposals
- Assess treasury management: large, unallocated treasuries are natural targets for governance capture attempts
Frequently Asked Questions
Q: What is a governance attack? A: A governance attack involves acquiring sufficient governance tokens (sometimes via flash loans, sometimes via straightforward market accumulation over time) to pass a proposal that benefits the attacker at the protocol’s or community’s expense – for example, redirecting treasury funds to an attacker-controlled address or vault. Notable examples include the Beanstalk protocol’s flash-loan governance attack (roughly $182 million drained in April 2022) and the July 2024 Compound Proposal 289 episode, in which a delegate group actually succeeded in passing a proposal reallocating treasury funds before it was voluntarily reversed through negotiation rather than through a governance vote defeating it.
Q: What is the difference between on-chain and off-chain governance? A: On-chain governance uses smart contracts to record votes and, in many systems, execute approved proposals automatically or semi-automatically. Off-chain governance (as in Bitcoin) relies more heavily on social consensus – developers propose changes (BIPs), miners and node operators signal support, and users choose whether to upgrade their software. Off-chain approaches tend to be more conservative and slower-moving but are sometimes viewed as less prone to the kind of concentrated-token-driven capture that on-chain systems can experience.
Q: What is vote delegation? A: Token holders can assign (“delegate”) their voting power to another address without transferring ownership of their underlying tokens. The delegate then votes on their behalf in governance processes. Delegation lets more passive holders participate indirectly and allows engaged, informed delegates to accumulate meaningful voting influence from many individual delegators.
Q: Can governance votes be reversed? A: Once a proposal has executed on-chain, it generally cannot be reversed without a subsequent governance vote (and some changes may be effectively permanent depending on the protocol’s design). This is a key reason timelocks (a delay between a vote passing and its execution) matter – they give the community a window to react to a proposal they believe is malicious or mistaken before it actually takes effect. Notably, in the Compound Proposal 289 case, the proposal was reversed not through a governance mechanism defeating it, but through direct negotiation with the party who had pushed it through.
Q: What is a “rage quit” in DAO governance? A: In Moloch-style DAOs, “rage quit” allows dissenting members to exit the DAO with their proportional share of the treasury before an approved proposal they disagree with is executed. This is designed to protect minority members from being forced to remain bound by decisions they fundamentally object to.
Related Terms
- DAO (Decentralized Autonomous Organization) – the organizational structure that on-chain governance enables.
- Governance Token – a token conferring voting rights within an on-chain governance system.
- Governor Contract – the smart contract implementing a proposal-vote-execution lifecycle.
- Timelock – a smart-contract-enforced delay between governance approval and execution.
- Delegation – the assignment of voting power to a representative without transferring the underlying tokens.
- Plutocracy – a governance dynamic where wealth or token concentration confers disproportionate decision-making power.
- Quorum – the minimum participation threshold required for a governance proposal to be considered valid.
Sources
- Compound Finance Governance Documentation – https://compound.finance/governance
- Uniswap Governance
- Optimism Governance Documentation
- Sky (formerly MakerDAO) Governance
- Tally – On-Chain Governance Analytics
- Snapshot – Off-Chain + On-Chain Governance Platform – https://snapshot.org
- The Block, “$24 million Compound Finance proposal passed by whale over DAO objections”
- The Defiant, “Compound Agrees To Distribute 30% Of Reserves To COMP Stakers To End Alleged Governance Attack” – https://thedefiant.io/news/defi/compound-agrees-to-distribute-30-of-reserves-to-comp-stakers-to-end-alleged-governance-attack
