How to Prevent Vote Buying in Token Governance
Vote buying quietly hollows out governance by turning votes into a market. Here is how bribery works in token voting and the designs that actually blunt it.
Vote buying is when someone pays voters to cast their ballot a certain way, or rents their voting power outright, turning governance into a market for influence. It is the quiet killer of token governance because it usually looks legitimate: a bribe can be dressed as a bounty, a reward, or a lending market. The defense is not to pretend it will not happen. The defense is to make bought votes either impossible to verify for the buyer or worthless once cast. If a briber cannot confirm you voted their way, they will not pay, and the whole market collapses.
Why vote buying is worse than it looks
The damage is not just one rigged vote. It is that governance stops representing the community and starts representing whoever is willing to pay. Decisions drift toward whoever can afford to buy them, small holders get priced out of mattering, and the outcomes lose all legitimacy. Once a community suspects votes are for sale, it stops participating honestly, and you are left with governance that is theater over a market underneath.
It also blurs into normal activity in nasty ways. In token-weighted systems, buying tokens to gain influence is legitimate ownership. Renting those tokens for a single vote is manipulation. The line between them is thin, and attackers live in that gap.
The mechanisms that enable it
Verifiable votes. The root enabler. If a briber can prove how you voted, they can pay on delivery. Public on-chain ballots make bribery trivially enforceable because the receipt is right there on the ledger.
Vote-lending markets. Protocols that let holders lend voting power turn influence into a rentable commodity. Even when framed as yield, the effect is that anyone can assemble a temporary majority for the price of the rent.
Bribery-as-a-service. Tooling that pays holders to delegate or vote a certain way, marketed as incentives. It is bribery with a dashboard, and it is common enough to have a business model.
How to design against it
There is no perfect cure, but there are strong defenses, and they stack.
Break verifiability of the individual ballot. This is the counterintuitive one. Secret ballots exist in public elections precisely to defeat vote buying: if you cannot prove how you voted, nobody will pay you for it. Techniques that keep the aggregate count verifiable while hiding individual votes attack bribery at the root. The buyer loses the receipt, so the market for votes dries up.
Snapshot the electorate before the vote. If voting power is fixed at a block before the proposal is even known, you cannot rent tokens to swing a specific vote you already see coming. Timing defeats a lot of rented-majority attacks.
Lock or vest voting power. When influence requires tokens to be staked or time-locked, short-term renting becomes expensive or impossible. You have to actually hold the position through the decision, which reunites influence with real exposure.
Cap concentration. Limiting the maximum weight any single address can wield reduces the payoff of assembling a bought majority in the first place. Smaller ceilings mean bribery has to corrupt more people, which is harder and more visible.
Watch the graph. Sudden delegation spikes, coordinated last-minute swings, wallets funded from a common source right before a vote. These are the fingerprints of a bought outcome. Log them and audit for them after the fact.
The honest tradeoff
Notice the tension. Full public verifiability of every ballot is exactly what enables vote buying, but opacity is what enables rigging. You cannot have naive versions of both. The resolution is a system where the count is provably correct but individual choices are private, which defeats bribery and fraud at the same time. That is hard cryptography, not a checkbox, and most tools skip it.
On MintVote I treat vote buying as a design constraint from the start: snapshotting, the option of ballot privacy with a verifiable tally, and monitoring for the coordination patterns that give bought outcomes away. You will not stop every bribe. But you can make the receipt worthless and the rented majority expensive, and that is enough to keep governance representing the community instead of the highest bidder.