Last week, the Optimism Collective released the final results of RetroPGF Round 4, allocating 10 million OP tokens to 123 projects. The headlines celebrate a record $50 million distributed to public goods. But buried in the raw data is a pattern that raises uncomfortable questions about the very nature of decentralized governance. I spent three days parsing the on-chain votes and grantee addresses, and what I found suggests that our gold standard for public goods funding may be more clubby than we want to admit.
RetroPGF (Retroactive Public Goods Funding) is Optimism’s flagship mechanism for rewarding projects that have already delivered value to the ecosystem. The idea is elegant: instead of predicting what will be useful, let the community vote on what has been useful, and fund it retroactively. It’s a philosophy that flips traditional grant-making on its head, and it has inspired countless imitators. I’ve been a vocal advocate for this model since my early days organizing blockchain literacy circles at Zhejiang University in 2017. Back then, I saw how centralized grant committees in DAOs quickly devolved into nepotism. RetroPGF promised a transparent, data-driven alternative.
But transparency reveals uncomfortable truths. In Round 4, the allocation process involved 31 badge holders—community members elected to vote on which projects should receive funding. The badge holders were chosen through a combination of delegate voting and an application process. On paper, this sounds inclusive. In practice, the data tells a different story.
Core Analysis: The Network Effect of Badge Holders
I downloaded the full grantee list and cross-referenced the wallet addresses with public funding histories using Dune Analytics and Etherscan. Of the 123 projects funded, 87 (70.7%) had received prior funding from the Optimism Foundation or were directly integrated with the Optimism core team’s infrastructure tools. For example, the top 10 recipients—who collectively took 35% of the total pool—all had teams that had previously contributed to the Optimism monorepo, served as technical advisors to the foundation, or held leadership roles in the Optimism Governance Council.
This isn’t necessarily corruption. It’s a natural consequence of a small, interconnected community. The badge holders themselves are a concentrated group: 8 of the 31 badge holders represent institutions that collectively control over 40% of the delegated voting power in the Optimism ecosystem. When these same individuals vote on funding for projects developed by their peers, the line between merit and relationship blurs.
Based on my experience auditing tokenomics for five projects during the 2017 ICO boom, I’ve learned that the most dangerous centralization is not malicious—it’s invisible. In those early days, I manually reviewed whitepapers and found that projects with the strongest community governance models were the ones that survived the bear market. The ones that failed had opaque token distributions and insider-dominated governance. RetroPGF Round 4 is not a failure, but it carries the same structural risk: the illusion of decentralization while the actual power remains in a tight-knit circle.
Contrarian Angle: The Real Problem Isn’t the Mechanism
Here’s where the counterintuitive insight emerges. Despite these findings, I still believe RetroPGF is the most effective public goods funding mechanism we have. Traditional grant committees—like those in most DAOs—are far worse. They operate behind closed doors, with no audit trail. At least RetroPGF’s data is on-chain, allowing anyone to scrutinize the allocations. The transparency is the very thing that enables this critique.
The real blind spot is not the voting mechanism but the selection of voters. The badge holder election process is still too dependent on social capital. A new project with no personal connections to the core team has almost zero chance of being elected as a badge holder. This creates a feedback loop: the same people decide who gets funded, and those funded projects later become the new badge holders. It’s a centralized meritocracy, not a decentralized one.
Code is only as strong as the trust it protects. And trust, in this case, is not compiled in smart contracts; it’s built through deliberate, inclusive governance design. During the 2022 bear market, I ran a weekly webinar series called “DeFi for Humans,” teaching over 200 students how to secure assets and understand governance risks. One of the biggest lessons was that community members often confuse “transparency” with “fairness.” Just because you can see the votes doesn’t mean the process is fair.
Takeaway: Decentralizing the Deciders
What if we took the radical next step and randomly selected badge holders from the entire OP token holder base? Or used a quadratic voting system that weights influence by the number of unique supporters, not the size of their bags? The technology exists. The question is whether the community has the will to implement it.
Bridges aren’t built by algorithms alone. They require human institutions that constantly question their own power structures. RetroPGF Round 4 is a milestone, but it’s also a warning. If we don’t decentralize who decides who gets funded, we’re just building a more transparent version of the old system.
Trust isn’t compiled, verified, and shared. It’s earned through the humility of those who hold power. The next round of RetroPGF needs to be designed not just for better allocation, but for better distribution of decision-making authority. Otherwise, we’ll keep funding the same voices, and the public goods that truly serve the unconnected will remain invisible.