Quadratic Voting in DAOs: Theory & Real Implementations

DeFi Intel Research · Updated 2026-07-02 · intermediate

Quadratic voting (QV) has emerged as a compelling alternative to the simple one-token-one-vote systems that dominate most DAOs. In a world where governance often devolves into plutocracy—where wealth dictates outcomes—QV offers a mathematical bridge between the ideal of 'one person, one voice' and the practical reality of token-weighted voting. The core insight is elegant: instead of paying linearly for votes, participants pay quadratically. This means that while a single voter can cast multiple votes on an issue, the cost per additional vote rises steeply, effectively empowering broad consensus over concentrated wealth.

In this guide, you will learn exactly how quadratic voting works, with a focus on its real-world implementations in DAOs such as Gitcoin Grants and how Aave's community explored (and ultimately hesitated to adopt) the mechanism. We'll dissect the anti-Sybil properties that make QV viable—and the tools like Gitcoin Passport that enforce them. We'll also confront the significant limitations: collusion risks, privacy trade-offs, and the need for credible identity verification. By the end, you'll understand not just the theory, but the gritty implementation challenges and why QV remains a powerful but fragile tool for decentralized governance.

The Core Mechanism of Quadratic Voting

Quadratic voting (QV) replaces the traditional linear cost model with a quadratic cost function. In standard token-based voting, buying one vote costs one token; buying 100 votes costs 100 tokens. In QV, the cost to purchase a given number of votes v is proportional to v². So casting one vote costs 1 credit, two votes cost 4 credits, three votes cost 9 credits, and so on. This simple change has profound implications.

To illustrate, consider a DAO where 100 voters each have 1 token. Under linear voting, a whale with 100 tokens can cast 100 votes and outmuscle any single small voter. Under QV, the whale can cast at most √(100) ≈ 10 votes (spending all 100 credits), while the small voter can cast 1 vote. The ratio shrinks from 100:1 to 10:1—still unequal, but far less extreme. Moreover, QV encourages voters to allocate tokens across many issues rather than concentrating them on one, because the marginal cost of each additional vote on the same issue rises steeply.

The key trade-off: QV protects against majority tyranny by a wealthy few, but it relies on the assumption that voters are unique individuals. Without identity verification, an attacker can create thousands of Sybil accounts, each buying one cheap vote, effectively bypassing the quadratic cost. This is why every real QV implementation includes an anti-Sybil layer—typically some form of proof of personhood or reputation.

Quadratic Voting in Gitcoin Grants: The Largest Live Implementation

Gitcoin Grants has been running periodic rounds of quadratic funding—a close cousin of QV—since 2019. Here, donors allocate voice credits to projects they like; the matching pool is distributed to projects in proportion to the square of the sum of square roots of donations per project. But the voting mechanism itself is an approximation of QV: each donor receives a fixed budget of credits (e.g., 100 per round) and can spend them on one or more projects. The cost to give v votes to a project is v² (rounded). This forces donors to spread their support rather than dump all credits on a single project.

What makes Gitcoin's implementation noteworthy is its anti-Sybil layer: Gitcoin Passport. Passport collects verifiable credentials (e.g., a verified Twitter account, a BrightID connection, a Proof of Humanity registration) to build a unique human identity. A Sybil attacker would need to fabricate multiple real-world identities to create many accounts, each with a single vote—a costly and non-scalable attack. Gitcoin's QV has funded thousands of public goods projects and is widely considered the most successful application of the mechanism in crypto.

"Quadratic voting is not a silver bullet. Its success depends entirely on the quality of the identity system that stands behind it." — Vitalik Buterin, Ethereum founder

However, Gitcoin's model is not pure QV for every decision. It's a quadratic funding mechanism applied to grant allocation. For governance decisions (e.g., protocol parameter changes), Gitcoin itself still uses token-weighted voting (on Snapshot). This distinction is important: QV is used for allocating a shared budget, not for every DAO proposal.

Aave's Exploration of Quadratic Voting: Why It Didn't Stick

The Aave governance community has actively discussed introducing quadratic voting for certain proposals, particularly those affecting risk parameters or treasury allocations. In 2021, a forum post proposed a quadratic voting system on Aave’s Snapshot space, with the goal of reducing the influence of large stkAAVE holders. The proposal received significant attention but was never implemented.

Why did Aave hesitate? Two primary reasons: collusion and complexity. First, without a robust identity layer, whales could simply split their AAVE across multiple wallets and vote with each wallet at the first low-cost vote—defeating the quadratic cost curve. Aave had no ready-made Sybil resistance solution. Second, many community members argued that a pure token-weighted system is simpler and more predictable; QV introduces new game-theoretic variables that might discourage participation. Aave ultimately stuck with linear voting, though the discussion remains alive for future parameter-specific implementations.

This example illustrates a critical lesson: QV is not a one-size-fits-all. It requires a trusted identity system, which most DeFi protocols lack. Moreover, QV works best when there is a fixed budget to allocate (like in Gitcoin) rather than binary yes/no governance votes, where the non-linear cost can be confusing and lead to lower voter turnout.

Anti-Sybil Properties: The Identity Layer Is Everything

The mathematical beauty of quadratic voting is that it makes Sybil attacks—creating many fake accounts to buy cheap votes—exponentially expensive. To see why: without identity, an attacker with 100 credits could either buy 10 votes under QV (costing 100 credits) or create 100 Sybils each with 1 credit and 1 vote, spending the same 100 credits but getting 100 votes. The quadratic cost is rendered useless.

To exploit QV without identity, you simply split your budget into many single-vote accounts. Therefore, every viable QV system requires an identity layer that ensures one person = one account (or at least one valid credit budget). Current tools include:

• Gitcoin Passport: Aggregates verifiable credentials from BrightID, Proof of Humanity, ENS, and social accounts. Each unique human gets one passport score.
• Proof of Personhood (PoP): Biometric-or verification systems like Idena or Worldcoin's iris scanning. These are strong but raise privacy concerns.
• BrightID: A social-graph-based Sybil resistance that creates a unique identity by linking people through connection verifications.
• Trusted third parties: Some DAOs rely on reputation or role-based access, which is less decentralized but effective.

The anti-Sybil strength directly determines QV's viability. No identity system is perfect, but a layered approach (multiple credentials) improves robustness.

Limitations and Criticisms of Quadratic Voting in DAOs

Despite its theoretical appeal, quadratic voting suffers from several well-documented shortcomings that have prevented its widespread adoption in DAOs.

• Collusion and vote buying: QV does not prevent a whale from bribing many small identity-verified voters to cast a particular vote. Since each small voter's vote is cheap (1 credit), a briber can pay 1 credit plus a bribe per voter and get many votes, effectively simulating a sybil attack through persuasion. This is known as a 'Sybil of the willing'.
• Voter confusion and low turnout: The quadratic cost curve is unintuitive for many users. DAO voters are accustomed to simple yes/no token voting. Requiring users to compute costs or allocate credits across options can deter participation, reducing the democratic legitimacy the system aims to increase.
• Privacy vs. identity trade-off: To enforce anti-Sybil, the system often requires users to tie their wallet to real-world identity or reputation. This clashes with the pseudonymous ethos of crypto and can expose users to surveillance or coercion.
• Computational complexity: For on-chain QV, computing square roots and managing credit allocation can be gas-inefficient. Most implementations (e.g., Gitcoin) run off-chain voting interfaces and submit only the aggregated result on-chain.
• Inapplicability to binary decisions: QV is most natural for allocating a budget across multiple options (quadratic funding). For simple up/down governance votes, the non-linear cost does little to improve outcomes—whales can still dominate by buying a few expensive votes, and small voters hardly move the needle. Many DAOs have found no clear benefit.

"Quadratic voting fails when there is no fixed budget and no credible identity. In those cases, simpler mechanisms like linear voting with delegation outperform it." — Vitalik Buterin, in 'Governance, Part 2: Plutocracy Is Still Bad'

These limitations are not fatal, but they mean QV is best suited for specific contexts: participatory budgeting, public goods funding, and governance of small, aligned communities with strong identity verification.

Real-World Implementation Patterns and Hybrid Approaches

Rather than pure QV, most DAOs have adopted hybrid mechanisms that borrow elements of quadratic weighting while respecting practical constraints. Common patterns include:

• Quadratic delegation: Users delegate their voting power to representatives, who then vote using a quadratic curve on delegation-weighted votes. This simplifies the user experience while retaining some anti-whale properties.
• Liquid democracy + QV: A user can vote with credits on some issues and delegate other issues—all under a quadratic budget. Tools like Vocdoni's governance interface explore this.
• Time-weighted quadratic voting: Voting credits decay over time, further lowering the influence of large holders who do not participate consistently.
• Off-chain scoring with on-chain settlement: As seen in Gitcoin, all quadratic calculations happen off-chain (through a coordinator like Allo Protocol or a dedicated frontend), and the final allocation is executed via a smart contract. This reduces gas costs and allows for complex identity verification.

No implementation is perfect—each must trade off between decentralization, privacy, security, and user experience. The most successful QV deployments prioritize a strong identity layer (Gitcoin Passport) and a clear use case (grant allocation). For general governance, many DAOs remain skeptical.

The Future: Zero-Knowledge Proofs and Reputation-Based QV

New cryptographic tools promise to address some of QV's limitations. Zero-knowledge proofs (ZKPs) could allow a voter to prove they are a unique human without revealing their identity—solving the privacy trade-off. For example, a ZK circuit can verify a voter's credentials (e.g., from a trusted issuer) and output a nullifier that prevents double voting, all without leaking which human is voting. This could enable a Sybil-resistant, private QV system.

Another frontier is reputation-weighted quadratic voting. Instead of granting every human equal voice, the community assigns weights based on contribution, tenure, or stake. The quadratic curve then applies to the weight, so a high-reputation user can still have disproportionate influence but at an increasing cost. This blends the anti-plutocracy benefits of QV with the meritocratic reality that not all participants are equally well-informed.

Protocols like Clr.fund and charmverse continue to iterate on QV implementations. Expect to see more DAOs adopt QV for specific treasury decisions, but widespread use as the primary governance mechanism remains unlikely until identity solutions become both robust and privacy-preserving.

Frequently asked questions

Related reading

• Aave Complete Guide 2026
• Blackrock Buidl Tokenized Treasury Fund 2026 Guide
• Dao Governance Voting Power Optimization 2026 Guide
• Perpetual Funding Rate Arbitrage 2026 Guide
• Token Vs Coin Difference 2026