Why DAO Grant Committees Are Inherently Flawed

Committee members are agents spending principal (treasury) funds. Their incentives diverge from the DAO's long-term health, leading to status-quo signaling and low-risk, high-consensus grants that rarely fund breakthrough work.\n- Misaligned KPIs: Rewarded for activity (grants issued), not outcomes (value created).\n- Social Consensus Bias: Favors proposals from known entities, stifling new builders.

Committees act as informational bottlenecks. They lack the domain expertise to evaluate niche technical proposals (e.g., ZK-circuits, MEV research) and create a high-friction application process that filters out global, anon talent.\n- Expertise Gap: Generalists judging specialists leads to poor capital allocation.\n- Process Friction: Weeks-long cycles kill momentum for fast-moving builders.

Retroactive Public Goods Funding (RetroPGF) inverts the model: fund what has already proven valuable. This aligns incentives with outcomes, as seen in Optimism's rounds. However, it introduces new challenges: subjectivity in valuation and retrospective collusion.\n- Pro: Funds tangible results, not promises.\n- Con: Requires robust, Sybil-resistant reputation systems.

The original MolochDAO exposed the flaw: slow, committee-based grants cause voter apathy and capital stagnation. The solution was rage-quitting and forking—a nuclear option that allows dissenting members to exit with their capital. This proves the need for continuous, fluid alignment.\n- Fork as Feature: Highlights the failure of static governance.\n- Capital Efficiency: Locked capital in committees has near-zero velocity.

The original grant DAO became a victim of its own success. High-profile committees attracted low-quality, high-volume proposals from grantees gaming the system. The result was capital misallocation and a ~90%+ rejection rate, creating a hostile environment for genuine builders.

• Key Problem: Signal-to-noise ratio collapses under public fame.
• Key Problem: Committee members become political targets.

Uniswap's formal grants program was shut down after ~2 years despite a $74M+ treasury. The committee model failed to scale, suffering from proposal fatigue and an inability to measure ROI on funded work. This led to the pivot towards more focused, ecosystem-specific initiatives.

• Key Problem: No accountability for grant outcomes.
• Key Problem: Committee becomes a bottleneck for all decisions.

Optimism's bicameral governance exposes the core flaw: separating fund allocation (Citizen House) from protocol direction (Token House) creates incentive divergence. Citizens, not directly staking OP tokens, lack skin-in-the-game, leading to grants that may not align with long-term tokenholder value.

• Key Problem: Delegators vs. Beneficiaries incentive mismatch.
• Key Problem: Retroactive funding (RetroPGF) is still a political process.

Replace subjective committees with verifiable, on-chain KPIs. Fund projects based on smart contract usage, fee generation, or user growth—metrics that are publicly auditable and impossible to game politically. This aligns funding with delivered value, not persuasive proposals.

• Key Benefit: Eliminates committee politics and bias.
• Key Benefit: Creates direct accountability for builders.

Delegate capital allocation to sub-DAOs and ecosystem experts (e.g., Aave Grants DAO, Compound Grants). These specialized groups have higher context and faster decision loops. The core DAO treasury provides funding tranches based on the sub-DAO's proven track record, not individual proposals.

• Key Benefit: Decisions made by domain experts, not generalists.
• Key Benefit: Scales governance through subsidiarity.

Shift from speculative grants to retroactive rewards for proven work, as pioneered by Optimism's RetroPGF. This funds outputs, not promises. While not perfect (it still uses a voting committee), it inverts the incentive, forcing builders to build first and prove value to the community.

• Key Benefit: Funds what worked, not what might work.
• Key Benefit: Reduces upfront capital risk for the DAO.