The Problem with How We Fund Public Goods
Government grants go to institutions with grant writers. Traditional philanthropy mirrors donor preferences, not community need. Corporate sponsorships track marketing value. Crowdfunding rewards distribution reach.
None of these mechanisms reliably answer the question that actually matters: what does the community most want funded?
The problem is structural. Traditional capital allocation optimizes for whoever controls the capital — governments, philanthropists, corporations — rather than for the aggregate preferences of the people who benefit from the public goods being funded. The result is systematic underfunding of infrastructure, education, open-source software, and research: things with broad, diffuse beneficiaries and no obvious revenue model.
Quadratic funding is a mathematical solution to this problem. It was designed from first principles to make capital allocation reflect actual community preferences. And since 2019, it has been operating at scale.
The Mechanism: How It Works
Quadratic funding was formalized in a 2018 paper by Vitalik Buterin (co-founder of Ethereum), Zoe Hitzig (Harvard economist), and E. Glen Weyl (Microsoft Research/RadicalxChange). The paper, "A Flexible Design for Funding Public Goods," proposed a mechanism that could theoretically produce optimal public goods provision under certain conditions.
The formula: the total funding a project receives equals the square of the sum of the square roots of individual contributions.
Written out: if project X receives contributions c₁, c₂, c₃... from individual donors, the total allocation from a matching pool is:
(√c₁ + √c₂ + √c₃ + ...)²
This is easier to understand through examples:
Scenario A: One donor gives $100 to a project.
- Sum of square roots: √100 = 10
- Squared: 10² = 100
- Matching contribution from pool: proportional to 100
Scenario B: 100 donors each give $1 to a project.
- Sum of square roots: 100 × √1 = 100
- Squared: 100² = 10,000
- Matching contribution from pool: proportional to 10,000
Same total donated ($100 in both cases). Dramatically different matching — the project with 100 small donors receives 100x more from the matching pool than the project with one large donor.
The implication: quadratic funding weights the number of contributors far more heavily than the size of contributions. A project that resonates broadly with many people — even if those people have little money — gets substantially more total funding than a project that appeals primarily to wealthy donors.
This is not arbitrary. It corresponds to a mathematical result in mechanism design: under certain assumptions about how public goods benefit populations, this allocation mechanism produces the socially optimal level of provision for each good. The formal result, called "optimal public goods provision," had been theorized in economics for decades. Buterin, Hitzig, and Weyl showed that quadratic funding implements it.
Why This Is Different from Traditional Philanthropy
Traditional philanthropy has a concentration problem. A billionaire who donates $10 million to their preferred cause moves more capital than a million people who each donate $10. The billionaire's preferences determine the allocation. The million people's preferences don't.
This is not a moral criticism of wealthy donors — it is a structural feature. Capital concentration in philanthropy mirrors capital concentration in the economy. The preferences of whoever controls the money dominate.
Quadratic funding inverts this. The matching pool — typically funded by foundations, corporations, or protocols — is distributed not according to donor-size but according to community-revealed preferences measured by participation breadth. A wealthy donor can still give a large amount, but their influence on matching allocation is bounded by the square root of their contribution. The mathematical dampening of donor size is the point.
Compared to government grants, quadratic funding also has structural advantages. Government grants typically go through application processes that reward institutional capacity over actual need or community value. Quadratic funding bypasses this: projects with genuine community support attract contributions directly, and the matching pool responds proportionally. There is no grant committee to persuade, no application cycle to navigate, no political relationship to cultivate.
Compared to venture capital, quadratic funding does not require projects to be profit-generating. Open-source software, public research, community infrastructure, and educational resources — things that create enormous value but cannot easily monetize it — can receive substantial capital if the community values them.
Gitcoin: The Real-World Test
Theory is useful. Empirical results are more useful.
Gitcoin Grants launched in 2019 as the primary real-world implementation of quadratic funding. It operates as a platform where anyone can submit a project, anyone can donate, and a matching pool provided by sponsors amplifies donations according to the quadratic formula.
The cumulative data as of early 2026:
- $70 million+ distributed to public goods projects
- 5,000+ projects funded across 20+ grant rounds
- Hundreds of thousands of unique donors participating
- Web3 infrastructure, open-source tooling, education, public research as primary categories
The distributions consistently demonstrate the mechanism working as intended. Projects with broad community support — widely-used open-source libraries, educational resources, developer tooling — attract large numbers of small donations and receive outsized matching. Niche projects with narrow but intense donor bases receive smaller matching allocations despite potentially larger total donations.
Notable funded projects include Ethereum documentation, cryptographic libraries, privacy tooling, and developer education resources — exactly the category of public goods that markets systematically underprovide because their benefits are non-excludable (you can't easily charge for them) and non-rival (one person using them doesn't reduce availability for others).
Gitcoin has also expanded beyond Web3. Later grant rounds have included climate research, open civic infrastructure, media projects, and scientific research tools — demonstrating that the mechanism is not inherently tied to blockchain applications.
The Recipients Who Become Funders
One of the most observed effects of quadratic funding at scale is a flywheel dynamic: projects that receive funding become funders of the next round.
This is not an accidental feature. Gitcoin has actively encouraged grant recipients to donate to other projects, and the culture of the ecosystem has developed around the norm of reciprocal support. A developer whose open-source library received $15,000 in matching funds is likely to donate to the next round's projects — both out of genuine support and because participation maintains their standing in the ecosystem.
The flywheel has a compounding effect. As more people receive funding and become funders, the donor pool grows. As the donor pool grows, community-preferred projects receive larger matching contributions. As those projects succeed and their creators become donors, the cycle continues.
This mirrors what we see in healthy ecosystems more broadly: value flows back into the system that produced it, strengthening rather than depleting the foundation. The mechanism design encodes regen logic into the financial infrastructure.
The Sybil Problem
Quadratic funding has one significant and well-documented vulnerability: sybil attacks.
If the mechanism rewards the number of unique donors, an obvious manipulation strategy is to create multiple fake identities and make small donations from each one — inflating the apparent breadth of support without genuine community backing. One person controlling 100 wallets could receive the same matching multiplier as 100 genuinely distinct donors.
Gitcoin has addressed this through Passport, a decentralized identity verification system that aggregates trust signals from multiple sources: GitHub activity, Twitter/X presence, Proof of Humanity, BrightID, ENS domain ownership, on-chain transaction history, and others. Each contributor's Passport generates a "humanity score" based on the number and quality of trust signals they can verify. Contributions from low-trust Passport holders receive reduced weight in the quadratic formula.
As of 2024, Gitcoin Passport had issued over 900,000 passports and was processing identity verification for hundreds of thousands of grant round participants. The system is imperfect — determined sophisticated actors can still acquire multiple verified identities — but it substantially raises the cost of sybil attacks.
The broader sybil problem is unsolved at the theoretical level. Quadratic funding requires distinguishing genuine humans from coordinated fake identities, which is a version of the hard problem in identity verification. Biometric approaches, trusted hardware attestations, and social graph analysis are all being explored. None is complete.
Collusion is a related vulnerability: donors who coordinate to support each other's projects and split the matching gain. This is harder to detect than sybil attacks because the donors are real people. Gitcoin employs data science analysis to identify suspicious correlation patterns in donation behavior, but collusion at small scales remains difficult to prove.
These limitations are real. They do not invalidate the mechanism — they define its current operational envelope and the research agenda for improving it.
Optimism RetroPGF: The Next Evolution
Gitcoin's quadratic funding model asks the community to predict which projects will be valuable, then funds them. This prediction problem is hard. Many funded projects do not deliver. The mechanism is forward-looking and therefore uncertain.
Optimism's Retroactive Public Goods Funding (RetroPGF) model inverts the direction of time. Rather than funding predictions of future value, it rewards demonstrated past value.
The premise: it is easier to evaluate what has already created value than to predict what will. If a developer tool has been used by 50,000 developers for three years, we know it created value. We don't need to predict it — we can observe it. RetroPGF pays for proven impact after the fact.
The Optimism Collective — the governance structure of the Optimism protocol — allocates tokens to fund public goods retroactively. Rounds are organized by category, with a committee of "badgeholders" evaluating projects and voting on allocations. The evaluation criteria focus on what was actually built and used, not on promises.
Results through early 2026:
- RetroPGF Round 1 (December 2021): $1 million to infrastructure and tooling projects
- RetroPGF Round 2 (March 2023): $10 million across 195 projects
- RetroPGF Round 3 (January 2024): $30 million to 501 projects
- RetroPGF Round 4 (2024): $10 million (pivoting to smaller, more targeted round)
- Cumulative: $200 million+ across all rounds
The model has attracted significant attention beyond the Ethereum ecosystem. The core insight — evaluate demonstrated impact rather than predicted impact — generalizes beyond blockchain. Government "milestone-based" grant structures partially implement this logic. Prize models (X-Prize, Wellcome Leap) similarly reward demonstrated achievement. RetroPGF is a more systematic implementation of retrospective funding at scale.
The limitation of RetroPGF is the chicken-and-egg problem: projects must be built before they receive funding. This works for developers who can sustain themselves while building, or who have other revenue sources. It does not work for projects that need upfront capital to exist at all. A combination model — some quadratic forward funding, some retroactive reward — may be optimal.
What This Means for Capital Allocation at Scale
Quadratic funding and retroactive public goods funding are not just interesting experiments. They represent a fundamental rethinking of how capital allocation can be structured to reflect collective preferences rather than concentrated power.
The traditional modes — government grants, philanthropy, venture capital — each carry the preferences of whoever controls the capital. Quadratic funding shifts the allocation signal to community breadth. RetroPGF shifts the allocation signal to demonstrated impact. Both move authority away from prediction by experts toward evidence from communities.
This matters beyond Web3. The mechanism design insights from quadratic funding are being studied by economists, policy designers, and philanthropists. The question "how do we allocate capital to reflect what communities actually value rather than what capital holders prefer?" is relevant to city budgeting, research funding, infrastructure investment, and public media.
The experiments running on Ethereum are generating empirical data on that question at scale. The early results are instructive: communities fund what they actually use, allocation efficiency improves with identity verification, and the flywheel of recipients becoming funders is real and self-reinforcing.
The Regen Finance Thread
Quadratic funding is the clearest implementation of a core regen principle: positive-sum mechanisms produce more total value than zero-sum ones.
In a traditional grant allocation, one project winning means another losing. The total pool is fixed; the competition is for shares of it. In quadratic funding, every donor's contribution increases the matching allocation across all projects, and the mechanism design rewards broad coordination over narrow competition. The optimal strategy for any project is to create genuine value for as many people as possible — which is precisely the behavior we want to incentivize.
This is what regenerative financial mechanisms look like at the design level: incentive structures where the dominant strategy is to create value for others, because doing so is the best way to receive value in return.
The math elegantly encodes what the degen→regen shift looks like in practice. Concentrate capital: extractive, zero-sum. Distribute participation: generative, positive-sum. Reward demonstrated impact: aligned with actual value creation.
The trillion-dollar question — examined in the previous issue — is whether mainstream capital allocation can shift in this direction. Quadratic funding and RetroPGF are two of the most rigorous answers anyone has yet tried to build.