A vote market is a marketplace where anyone can pay the holders of a governance token to vote a particular way. The payments are universally – and only half-jokingly – called bribes. The mechanism is not an attack or an exploit: it is an openly-run, smart-contract-mediated auction for voting power, and for several protocols it has become the single largest recurring flow of value in their governance. Vote markets are the clearest working example of a broader phenomenon – once a vote carries a predictable cash reward, that vote becomes an asset that can be priced, rented, and traded independently of the tokens underneath it.
Vote markets are a downstream consequence of vote-escrow tokenomics. In the ve-model pioneered by Curve, locked tokens (veCRV) do more than signal preference – they direct gauge weights, the on-chain vote that decides how newly-minted token emissions are split across liquidity pools (Convex: voting and gauge weights). Because a gauge vote literally routes real money, a project that wants deep liquidity has two options: buy and lock the governance token itself, or pay existing lockers to point their votes at its pool. The second is almost always cheaper, and the market that formed to price it became known as the Curve wars (Votium: why Votium).
The same logic reappears anywhere emissions are vote-directed: Balancer and its Aura wrapper, Aerodrome on Base, Frax, and any protocol that copies the ve(3,3) gauge design inherits a vote market whether or not it intends to.
Every vote market has three roles:
The delegation step is what turns a scattered electorate into a liquid, auction-clearing bloc: the venue aggregates thousands of small holders and sells their combined weight to the highest bidder each round. Voters do no work and simply collect a yield on tokens they were already locking.
Vote markets are efficient precisely because they are cheap for the buyer. Through much of the Curve wars, a protocol could direct more than a dollar of emissions to its pool for less than a dollar of incentives – a positive "bribe efficiency" that made renting votes strictly cheaper than earning them by buying and locking the token outright (Votium). For the voter, the incentive frequently exceeded the value of the emissions their own tokens would have earned passively, so delegating to a vote market became the default yield strategy for ve-lockers.
The result is a clean separation of two things that token-voting is supposed to bundle together: economic exposure (holding the token and its cash-flows) and control (deciding where the protocol's money goes). In a healthy vote market these can be leased apart by the week.
The same separation that makes vote markets efficient is what makes them contentious:
Defenders counter that a transparent, on-chain bribe is strictly better than the alternative it replaces: an opaque one. Formalising the vote market at least puts the price of influence on a public ledger where everyone can see it.
The largest voting platform built the option in and then left it unused. Snapshot's Boost feature, launched in March 2024, let anyone attach token rewards to a proposal, and a space could choose to reward only voters who picked a given outcome. Snapshot called that option a bribe, left it off by default and recommended it only for decisions such as how to split liquidity-mining rewards (Snapshot Labs). It removed Boost from its interface in August 2026 as unused (sx-monorepo #2294).
The intuitive counter – hide the ballots – does not work either. The shielded voting DAOs actually deploy decrypts and publishes every vote the moment the proposal closes, which leaves the proof of delivery a briber needs perfectly intact; only receipt-freeness removes it, and no DAO runs receipt-free voting at scale (private and shielded voting).
A caper does hold votes – since 29 August 2026 a proposal is decided by a ranked ballot before anything executes – so the question this page asks is live rather than moot. What a caper does not have is a weight that can be leased apart from the position underneath it, and that is a property of the weight formula rather than a rule against bribery.
Voting weight combines two holdings: the voter's balance of the caper's own token, and their balance of that caper's vote token, which is soulbound and earned only by voting. A ballot counts only if the voter holds both, and each resists a vote market by a different mechanism.
depositor role admits only the caper's own state component and the corresponding updater is deny_all, so a peer transfer aborts on the receiving side and the rule can never be relaxed. It is earned rather than held – one per ballot cast, and on nothing else – and the only address it can be paid to is the contract that burns it on exit. There is no veCRV analogue to delegate to a venue, because there is no transferable weight token to delegate.min(balance at cast, balance now), so a stake borrowed for a vote and returned before the tally is counted at whatever is left (fold_ballot_batch, contracts/logic/src/lib.rs). Renting weight for a single proposal is the ordinary shape of a governance-lending pool like Paladin's; here the loan's own repayment prices it out.The second phase makes the same point about money. A ballot that clears its supermajority does not execute; it opens the market window, and the action lands only if the token's time-weighted price across that window holds at or above the trailing average locked at the trigger. So an actor who does buy an outcome – and buying the token is the one route to weight that remains – has to still be holding when the window closes. A vote market's whole premise is that control is cheaper than ownership; a caper's answer is to make the cheapest available control a purchase you have to keep.
The exposure this trades into is a different one and worth naming: governance here is only as hard to move as the market is deep, so a thin order book is the attack surface a vote market would otherwise have been – the same critique this page makes of cheap bribe efficiency, relocated from the bribe to the curve. See the bonding curve for how a caper's price is formed and optimistic governance for the family the second phase belongs to.