Skip to content

Quadratic Voting + Reputation Weighting

The formula

A voter i's effective vote weight on any proposal is:

wi=staked_POPi×ri

Where staked_POP_i is i's currently staked $POP, and r_i ∈ [0, 1] is their Reputation Score.

Why quadratic

Pure token-weighted voting is plutocratic: a 100x holder has 100x influence. Quadratic voting (Lalley & Weyl, 2018) dampens this:

Token holdingLinear weightQuadratic weight
1 $POP11
100 $POP10010
10,000 $POP10,000100
1,000,000 $POP1,000,0001,000

A whale must hold 1M $POP to get 1,000x the voting weight of a small holder — versus 1M× in linear voting. This preserves whale influence (they have more than small holders) without granting outright veto power.

Why reputation-weighted

Pure quadratic voting can be Sybil-attacked: split a holding across 100 sub-accounts to multiply effective weight. Reputation gating defeats this: each sub-account's reputation r_i would need to be cultivated independently over time, at marginal cost roughly matching the linear-voting whale weight. The Sybil multiplier collapses.

Worked example

Three voters on a proposal:

VoterStaked $POPLinear weightQuadratic weightReputation r_iEffective w_i
Whale1,000,0001,000,0001,0000.6600
Mid holder10,00010,0001000.990
Small holder100100100.858.5

The whale has ~7x the weight of the mid holder, not 100x. The small holder has measurable weight (8.5 vs whale's 600 = ~1.4%) — much better than the 0.01% they'd have in linear voting.

Staking

To cast a vote, $POP must be staked. Staking is a lock that:

  • Lasts at least the proposal's voting period (7 days).
  • Cannot be unstaked during an active vote on any proposal the user voted on.
  • Earns a share of protocol fee distribution (see Value Accrual).

Unstaking has a 7-day unbonding period (cooldown). This prevents flash-loan governance attacks (acquire massive position, vote, exit immediately).

Reputation in governance

Reputation Score is the same r_i used in Round settlement (see Cryptographic Protocol). It reflects historical honesty and protocol engagement. Importantly:

  • New accounts have r_i = 0.5. They can vote, but with half weight.
  • Long-term honest participants approach r_i = 0.99. They get near-full weight.
  • Accounts flagged for malicious behavior have r_i → 0. They effectively cannot vote.

This means governance influence is earned through sustained protocol engagement, not just bought.

Founders Club governance bootstrap

For the first 6 months post-TGE, Founders Club members (L7) have 2x voting weight. This is a structural defense against governance capture by exchanges or large early buyers who could acquire $POP at TGE but not Reputation Score. After 6 months, the bonus expires.

Limitations

  • Voter turnout: typical DAO turnout is ~5–15% of holders. We design assuming this baseline.
  • Whale collusion: explicit collusion among whales is not prevented; quadratic only dampens individual whale weight.
  • Reputation gaming: an adversary could try to inflate Reputation through Round participation specifically to vote later. This is expensive but possible; we accept it as a design trade-off.

References

  • Lalley, S.P., Weyl, E.G. (2018). Quadratic Voting: How Mechanism Design Can Radicalize Democracy. AEA P&P 108.
  • Buterin, V., Hitzig, Z., Weyl, E.G. (2019). A Flexible Design for Funding Public Goods.

Built on TON.