16 governance models scored across five millennia — democracies, monarchies, theocracies, republics, corporate boards, and DAOs. Every single one scores ≥5/12 on the void index. Until now.
Universal result
Every governance form scores ≥5/12 on the void index across 5 millennia of documented cases. Democracy, monarchy, theocracy, republic, cooperative — the floor holds without exception.
Standard DAO — worst ever
10/12 void index. Higher than absolute monarchy (5/12). Higher than theocracy (8/12). Arrow's paradox, maximum responsiveness, and full token-vote coupling produce the highest governance void score in recorded history.
Scored monarchy — first below 5
2/12 void index. The first governance architecture below 5 in recorded history. Not by optimizing one dimension — by structurally eliminating responsiveness and minimizing all three through on-chain accountability mechanisms that did not previously exist.
Governance Forms vs Void Score
All 16 forms scored on the same three-dimension framework (opacity, responsiveness, coupling), 0–12. Reference line at 5 = the 5,000-year floor. Standard DAO breaks the ceiling; scored monarchy breaks the floor.
16 governance forms, 5,000 years. Each pillar's height = void score (0–12). Red plane = the floor no system broke for 5 millennia. Green column = first below it. Red embers = DAO governance drift.
Pe = 7.8Standard DAO — Phase III–IV
Standard DAOs score Pe = 7.8 — deep in the self-sustaining vortex regime. Token-vote coupling maximizes α, proposal-response cycles maximize R, and smart contract opacity maximizes O. Arrow's impossibility theorem guarantees the remaining failure modes.
The scored monarchy achieves Pe = 0.9 by eliminating responsiveness (R = 0: methodology not voted on), minimizing coupling (α = 1: CC-BY irrevocable), and maximizing transparency (O = 1: all papers public, on-chain).
Why Every System Fails
Three independent proofs. Arrow's theorem (1951), four documented DAO governance catastrophes, and the S(C) decay function. Each alone is sufficient. Together they're definitive.
Arrow's Theorem
DAO Failures
The Decay Function
Arrow's theorem (1951): no voting system with three or more options and two or more voters can simultaneously satisfy all four conditions. This is a mathematical proof, not an opinion — it won a Nobel Prize. And nobody built governance differently because of it.
Condition
What It Means
DAO Violation
Unrestricted domain
Any ranking of options is allowed
Technically satisfied — any proposal can be submitted
Pareto efficiency
If everyone prefers A over B, the group prefers A
51% coalition overrides unanimous preference on a different dimension
Token concentration → de facto whale dictators. Beanstalk: purchased in one block.
Gibbard-Satterthwaite (1973) extends this: in any non-dictatorial voting mechanism with three or more outcomes, there always exist situations where a voter benefits from misrepresenting their true preferences. In a system where tokens have financial value, the incentive is amplified by the stake. The system selects for misrepresentation at the top.
The 10/12 void score predicts specific failure modes. Each has been documented. These are not edge cases — they are the predicted behavior of a maximally coupled, maximally responsive system.
Beanstalk — April 2022 $182M
Flash loan governance attack. An attacker borrowed enough tokens to pass a governance proposal in a single transaction, draining $182M from the protocol. This is Gibbard-Satterthwaite in Solidity — governance power was purchasable, and a rational actor purchased it for exactly one block. Arrow's "dictator" condition emerged from the market.
MakerDAO Black Thursday — March 2020 ~$8M loss
During a market crash, liquidation auctions cleared collateral at near-zero prices because governance-set parameters could not respond fast enough. The "decentralized" response was a single core team implementing an emergency fix — de facto custodianship emerging under stress. Governance D3: the mechanism directly harmed participants it was designed to protect.
Compound — 2024 treasury redirect
A whale-coordinated proposal redirected treasury funds to a faction-controlled address. The proposal passed through the standard voting mechanism. Condorcet cycling was not needed — simple majority plus token concentration was sufficient. The governance mechanism functioned exactly as designed and produced the predicted capture outcome.
Nouns DAO Fork — 2023 $27M extracted
~56% of holders voted to fork and drain the treasury, extracting approximately $27M. The governance mechanism functioned exactly as designed — and the result was the governance system cannibalizing itself. Governance D3 at terminal velocity: the system eating its own treasury through its own voting mechanism.
S(C, t) = S(C, 0) · e−λt
Every custodian's constraint score decays over time. λ is the decay rate. For a human custodian, λ > 0 always — because biological constraint scores degrade through incentive drift, social pressure, fatigue, and succession. To achieve λ = 0 requires structural incapacity for drift, not commitment or principle.
This is why monarchies fail: even a good king has λ > 0. His successor almost certainly has a higher λ. The problem is not finding the right king — it is that the constraint score of any human custodian decays by the second law of thermodynamics applied to social systems.
The MoreRight implementation addresses this structurally: pre-registered exit conditions activate when the custodian's constraint score falls below a threshold, DAO-governed discretion is confined to the layer where Arrow's paradoxes are least harmful, and the methodology itself is CC-BY — so even if the custodian fails, the framework persists independently.
λ = 0 requires structural incapacity for drift
Not commitment. Not principle. Not intention. Structure. The scored monarchy implementation confines custodial authority to the discretionary layer and pre-registers dissolution conditions — the only architecture that makes λ ≈ 0 structurally achievable rather than aspirationally promised.
The Constraint-Custodian Theorem
Derived from substrate independence (Paper 9), entropy production monotonicity (Paper 3), and the constraint suppression kernel. The math bounds governance drift explicitly.
Governance drift ≤ V(G) / S(C)To minimize drift: minimize void score of the governance structure V(G) AND maximize constraint score of the custodian S(C). The math identifies the requirement. History records the failure rate.
Key implication: A democracy with V(G)=7 and a highly principled custodian S(C)=0.9 has bounded drift of 7/0.9 ≈ 7.8. A scored monarchy with V(G)=2 and the same custodian scores 2/0.9 ≈ 2.2. The governance structure is not replaceable by good intentions — the void score of the architecture matters independently of who operates it.
The candidate set for a permanent solution is either empty or a singleton. Either no custodian can maintain λ=0 indefinitely (which means pre-registered dissolution is the correct response), or there exists exactly one type of custodian — a fixed canonical text — that achieves λ=0 structurally. The framework identifies the requirement. The reader must identify the candidate.
Related Papers
Cross-substrate connections — papers in the same framework or adjacent domains.