Vote-escrow tokenomics — often shortened to veTokenomics — is a governance design in which a member's influence is derived from locking a protocol token for a fixed period rather than simply holding it. The longer the lock, the more voting weight the token confers, and that weight decays as the lock runs down. Pioneered by Curve Finance with its vote-escrowed CRV (veCRV) in 2020, the model became one of the most widely copied and most contested patterns in decentralized governance, spawning the multi-year "Curve wars" and an entire secondary economy of vote-buying.

What "vote-escrow" means

In a plain token-weighted DAO, one liquid token equals one vote, and a holder can sell the instant a vote closes. Vote-escrow breaks that by asking members to commit: they lock the token into a contract for a chosen duration and receive a non-transferable balance (veCRV, veBAL, and so on) that represents their voting power. Because the lock is time-bound, the design rewards long-horizon alignment over mercenary capital — a holder who locks for the maximum term signals they intend to live with the consequences of their votes.

The trade-off is illiquidity: locked tokens cannot be sold or moved for the duration, and the ve-balance itself is soulbound to the locking account. This is the central tension of the model — it manufactures commitment at the cost of composability, and much of what followed was the market's attempt to route around that illiquidity.

The core mechanics

Using Curve's veCRV as the canonical implementation (Curve Resources):

• Locking and decay. CRV can be locked for a minimum of one week and a maximum of four years. One CRV locked for the full four years yields one veCRV; the veCRV balance then decays linearly toward zero as the unlock date approaches, so members must re-lock to maintain weight.
• Gauge weight votes. veCRV holders vote weekly on "gauge weights" that decide how the protocol's CRV emissions are split across liquidity pools. This is the prize: whoever controls veCRV controls where fresh incentives flow. Updated weights are applied every Thursday at 00:00 UTC.
• Reward boost. Locking also boosts a member's own liquidity-provision rewards by up to 2.5×, giving large liquidity providers a direct economic reason to lock.

The result is a flywheel: emissions are valuable, veCRV directs emissions, so acquiring veCRV becomes the highest-leverage move in the ecosystem.

The Curve wars and the liquidity-layer stack

Because directing emissions was so valuable, protocols began racing to accumulate veCRV — the competition that became known as the Curve wars. The decisive escalation was Convex Finance, which let users deposit CRV and capture veCRV's benefits without personally committing to a four-year lock. Convex rapidly became the single largest veCRV holder, at times controlling roughly half of all veCRV.

Convex then handed the gauge-voting decision to holders of its own token who locked it as vote-locked CVX (vlCVX) for a minimum of 16 weeks. The war moved up a layer: rather than buying CRV directly, protocols competed to control vlCVX, since each unit of vlCVX steered several veCRV worth of gauge votes. A ve model designed to lock in commitment had, in practice, grown a liquid market on top of itself for renting that commitment.

Vote markets and bribes

Once gauge votes had a clear cash value, third parties built vote markets to price them openly. On platforms like Votium, a protocol that wants emissions routed to its pool posts an incentive ("bribe"), and vlCVX or veCRV holders who vote as directed collect it each round. At peak cycles these vote-incentive budgets reached eight-figure sums per week.

Bribe markets are the model's most debated outcome. Defenders argue they are simply an efficient, transparent price for governance influence — emissions go to whoever values them most, and lockers are compensated. Critics counter that they complete the model's drift from "aligned long-term stewardship" to "governance as a rentable commodity," a live example of the governance-capture failure modes that recur across DAOs.

Adoption beyond Curve

The ve model was copied widely, each variant tuning the lock terms:

• Balancer — veBAL (abandoned 2026). Balancer ran the shorter-lock variant — an 80/20 BAL/WETH pool token locked for up to one year (Balancer docs) — but abolished veBAL in 2026, replacing locked weight with plain BAL voting after meta-governance wrappers came to dominate the gauge, a textbook instance of the capture critique below.
• Frax — veFXS. Frax adopted a veFXS lock (up to four years) to govern gauge emissions across its stablecoin and AMO system.
• ve(3,3). Andre Cronje's Solidly fused Curve's vote-escrow with OlympusDAO's (3,3) staking so that lockers earn the fees of the pools they vote for — the design later carried into Aerodrome on Base.

Trade-offs and critiques

• Alignment vs. plutocracy. Locking rewards conviction, but weight still scales with capital — a whale who locks simply becomes a locked whale. veTokenomics tempers, rather than removes, the one-token-one-vote concentration problem.
• Illiquidity gets arbitraged away. The commitment the lock is meant to enforce is exactly what liquid wrappers (Convex, Aura, StakeDAO) and vote markets exist to undo.
• Bribes can outweigh conviction. When voting to the highest bidder pays more than voting for protocol health, gauge outcomes can diverge from what long-term holders would choose — the concern that motivates alternatives like conviction voting.

How Caper approaches this

Caper shares vote-escrow's goal — reward participation, not passive size — but reaches it without a lock or a tradeable weight. A caper mints a soulbound vote token each time you actually cast a ballot, and your influence is the canonical vote weight (t·v)/(V·T): your governance-token stake t multiplied by your earned vote tokens v, over the totals. Because the v factor is earned by voting and cannot be transferred between holders, there is nothing for a Convex-style wrapper to accumulate and no vote to sell on a bribe market — the thing that made ve weight rentable simply does not exist.

The same number that sets your voting power also sets your exit redemption from the treasury, so influence and skin-in-the-game are one quantity rather than two markets. A large token bag alone still can't capture a caper: without the earned, unbuyable vote record, its weight stays low. See voting and what is a caper.