Understanding Decentralization
in DAO Law
What decentralization means, why it matters, how it is measured mathematically, and how New Hampshire and other jurisdictions define it in law.
Why Decentralization Matters
Decentralized autonomous organizations take their name from a structural property. But decentralization is not valuable in itself. It matters because of what happens when you have it — and what goes wrong when you don’t. The consequences fall into three categories: what decentralization does to the network, what it unlocks in law, and why it benefits people beyond the network’s own participants.
Technical Consequences
Network Security
Without decentralization, a party controlling more than half the network’s decision-making power can rewrite transaction history, double-spend assets, or halt the network. Distributing power makes such attacks increasingly expensive and logistically improbable. See Srinivasan & Lee (2017).
Systemic Resilience
Without decentralization, the network has a single point of failure. If that actor exits, is compromised, or is sanctioned, the system goes down. Distributed architecture ensures the network survives the loss of any individual participant.
Legal Consequences
Securities Law Classification
Without sufficient decentralization, a digital asset is likely an “investment contract” under SEC v. W.J. Howey Co., 328 U.S. 293 (1946), because purchasers rely on the “efforts of others.” As SEC Director Hinman observed in 2018, a “sufficiently decentralized” network may cause an asset to exit securities classification. The proposed CLARITY Act would formalize this: decentralization is the mechanism by which tokens graduate from security to commodity status. See Oranburg, Replacing Howey with CLARITY (2025).
Legal Entity Recognition
Without decentralization structured around a human governance floor, a DAO cannot obtain legal personality, limited liability, or standing to sue. It defaults to an unincorporated association or general partnership, exposing every token holder to personal liability. The Ooki DAO and Sarcuni cases illustrate the consequences. State DAO statutes condition legal recognition on both decentralization and minimum human governance. See Oranburg, Human in the Decentralized Loop (forthcoming).
Antitrust Defense
Without genuine decentralization, a DAO with market power looks like a cartel or monopoly to antitrust enforcers. With it, dispersed governance provides structural rebuttal evidence — within the meaning of the 2023 DOJ/FTC Merger Guidelines § 3 — that anticompetitive coordination is structurally implausible. See Oranburg, Market Power and Governance Power, Competition Policy International (2025).
Legal Certainty
The New Hampshire legislature stated its purpose directly: “the state’s economy and its citizens will benefit from, and the development of business and investment in New Hampshire will be facilitated by, making available a unique legal entity form” for DAOs. Statutory recognition of decentralized organizations “will introduce greater certainty for treatment… under state law.” RSA 301-B:2.
Economic and Social Consequences
Governance Legitimacy
Concentrated governance breeds suspicion about whether the system serves its community or merely its controlling faction. Broadly distributed decision-making power is more likely to be perceived as legitimate, sustaining participation and investment. When participants lose confidence in governance, they exit. See Oranburg, Governance as Club Good (forthcoming).
Worker Protection
In gig economy platforms, concentrated governance enables algorithmic wage-setting, exclusionary gating, and monopsony power over workers. Decentralized governance — high Nakamoto coefficient, low Gini — structurally prevents these harms by distributing control. A truly decentralized gig DAO provides workers a factual basis for collective organization without triggering per se antitrust liability. See Oranburg, Market Power and Governance Power (2025).
A note on circularity. When lawmakers ask whether a DAO is “meaningfully” decentralized, the answer cannot be “because the statute calls it that.” The value of decentralization rests on these consequences — security, legal classification, entity recognition, competition defense, governance legitimacy, and worker protection. A legal definition of decentralization should track these consequences, not merely create a formal category.
New Hampshire’s Three-Part Statutory Test
New Hampshire provides one of the most detailed statutory definitions of what it means for a blockchain network to be “decentralized.” The definition appears in § 301-B:5 IX and establishes that a “decentralized network” must satisfy three independent prongs — each looking at a different dimension of power and covering a different lookback period.
Under § 301-B:15(j)–(k), every registered New Hampshire DAO must be a “decentralized network” with a “decentralized governance system” at all times — making these ongoing compliance requirements, not one-time qualifications.
No Unilateral Control
During the previous 12 months, no person (other than the DAO itself) may have had unilateral authority to: (1) control or materially alter the blockchain system’s functionality; or (2) restrict any person from using digital assets, deploying software, participating in governance, or operating a node or validator.
🕑 12-month lookbackToken Concentration Limit — the 20% Threshold
During the previous 12 months, no person may have (1) directly or indirectly beneficially owned 20% or more of freely transferable related digital assets; or (2) had unilateral authority to direct the voting of 20% or more of outstanding voting power in the governance system.
🕑 12-month lookbackSource Code Independence
During the previous 3 months, the digital asset issuer, any affiliated person, or any related person must not have implemented or contributed intellectual property that materially alters the blockchain system — unless (1) it addressed vulnerabilities or cybersecurity risks, or (2) it was adopted through decentralized governance consensus.
🕑 3-month lookbackRSA 301-B:5 IX(b)(1) — The 20% Bright Line
“No person, excluding the subject DAO itself, directly or indirectly beneficially owned, in the aggregate, 20 percent or more of the total amount of units of a related digital asset that — (A) Can be created, issued, or distributed in such blockchain system; and (B) Were freely transferrable or otherwise used or available to be used for the purposes of such blockchain network.”
Prong B is the statute’s most directly quantifiable test. The 20% threshold means that no single beneficial owner may hold one-fifth or more of the network’s freely transferable tokens. This is a bright-line rule: it does not require a case-by-case analysis of whether someone actually exercised control — holding that percentage share is itself the disqualifying fact.
By comparison, Prong A is qualitative: it asks whether someone had the authority to exercise unilateral control, regardless of their token percentage. And Prong C is behavioral: it asks whether certain actors actually contributed code that materially changed how the system works.
What 20% implies for decentralization measures. If no single holder may own 20% or more, the minimum Nakamoto coefficient consistent with the statute is 3 — you need at least three of the largest holders working together to exceed 50% of tokens. In practice, because the statute requires no person to hold ≥ 20%, a compliant network with five equal holders at 19.9% each would have NC = 3, while a compliant network with 100 tiny holders could have NC well above 50. The threshold sets a floor on concentration, not a ceiling on decentralization.
Mathematical Measures of Decentralization
A statutory bright-line like 20% is only one way to characterize concentration. Two widely used mathematical measures — the Nakamoto coefficient and the Gini coefficient — offer complementary views of the same distribution. Each captures something the other misses, and both belong in any rigorous analysis of whether a network is genuinely decentralized.
The Nakamoto Coefficient
Introduced by Balaji Srinivasan and Leland Lee in 2017, the Nakamoto coefficient directly addresses the attack threshold question: how few independent entities would need to collude to seize majority control?
Let s1 ≥ s2 ≥ … ≥ sn be the holdings of n entities, sorted in descending order. The Nakamoto coefficient is:
In words: N is the smallest number of top holders whose combined holdings strictly exceed half the total supply.
Step-by-step example calculation
Suppose five entities hold the following token percentages: 25%, 20%, 20%, 20%, 15%. The total is 100%.
Sort descending: 25, 20, 20, 20, 15.
- After entity 1: cumulative = 25% < 50%.
- After entity 2: cumulative = 45% < 50%.
- After entity 3: cumulative = 65% > 50%. ✓
Therefore N = 3. Three entities, acting together, hold 65% — enough for majority control. Any smaller coalition falls short.
Notice that the largest single holder (25%) fails the NH 20% test, so this example network would not qualify as a “decentralized network” under RSA 301-B:5 IX(b)(1). The calculator below lets you explore how changing the distribution affects both the Nakamoto coefficient and the NH compliance status simultaneously.
Why 50%? A note on the threshold choice
The 50% threshold used in the Nakamoto coefficient definition corresponds to a simple majority — the point at which a coalition can outvote the entire rest of the network. This is the threshold relevant for Nakamoto-style proof-of-work consensus (the “51% attack”).
In Byzantine-fault-tolerant (BFT) consensus systems — which many modern DAO governance mechanisms use — the relevant threshold is one-third: a coalition controlling more than 1/3 of votes can halt the network (liveness attack), and a coalition controlling more than 2/3 can compromise safety. The Nakamoto coefficient can be calculated at any threshold; the choice of 50% is conventional.
The NH statute’s 20% single-holder limit implies that the minimum NC at the 50% threshold is 3 — at least three separate holders must collaborate to control the majority. At the 1/3 threshold, the minimum NC is 2 under the statute (two holders of 19.99% each barely exceed 1/3).
The Nakamoto coefficient is an attack-oriented measure: it tells you how organized and coordinated an adversary would need to be. A higher NC is better. An NC of 1 means the network is effectively centralized. An NC in the dozens or hundreds provides meaningful security against collusion.
The Gini Coefficient
Originally developed by Italian statistician Corrado Gini (1912) to measure income inequality across a population, the Gini coefficient has been adapted to measure how unequally token holdings are distributed across a DAO’s participants.
While the Nakamoto coefficient focuses on the top of the distribution (“how few large holders can seize control?”), the Gini coefficient characterizes the entire distribution’s shape — how skewed are holdings overall?
The Gini coefficient is defined in terms of the Lorenz curve. To construct it:
- Sort all holders by their holdings from smallest to largest.
- Plot the fraction of holders (x-axis) against the cumulative fraction of total holdings they collectively own (y-axis).
- The resulting curve always starts at (0, 0) and ends at (1, 1). Under perfect equality, every x% of holders would own exactly x% of tokens — so the curve would be a straight diagonal line.
- In practice, the bottom 50% of holders own less than 50% of tokens, so the curve bows below the diagonal.
The Gini coefficient is defined as twice the area between the equality diagonal and the actual Lorenz curve.
Let y1 ≤ y2 ≤ … ≤ yn be the holdings of n entities, sorted in ascending order. Let T = Σ yi be the total supply. Then:
The coefficient G lies between 0 (perfect equality: all holders own the same share) and 1 (perfect inequality: one entity owns everything).
Step-by-step example calculation
Consider three holders with percentages 10%, 30%, 60% (n = 3, T = 100).
Sorted ascending: y1 = 10, y2 = 30, y3 = 60.
Apply the formula term by term:
- i = 1: (2×1 − 3 − 1) × 10 = −2 × 10 = −20
- i = 2: (2×2 − 3 − 1) × 30 = 0 × 30 = 0
- i = 3: (2×3 − 3 − 1) × 60 = 2 × 60 = 120
Sum = −20 + 0 + 120 = 100.
G = 100 / (3 × 100) = 100 / 300 ≈ 0.333.
A Gini of 0.333 indicates moderate-to-high inequality in this three-holder example. By comparison, for equal holders (33.3%, 33.3%, 33.3%), the sum of (2i−n−1)×yi equals zero and G = 0.
Verifying the formula at the extremes
Perfect equality (G = 0): If yi = c for all i (everyone holds the same amount), then Σ(2i−n−1)×c = c×Σ(2i−n−1). Because Σi=1n(2i−n−1) = 2×(n(n+1)/2) − n×n − n = n(n+1) − n² − n = 0, the numerator is 0 and G = 0. ✓
Perfect inequality (G → 1 as n grows): If one entity holds everything (yn = T, all others zero), only the last term survives: (2n−n−1)×T = (n−1)×T. Then G = (n−1)×T / (n×T) = (n−1)/n → 1 as n→∞. ✓
Connecting the Measures
The Nakamoto coefficient and Gini coefficient are complementary lenses, not substitutes:
| Measure | What it asks | Sensitive to | Range |
|---|---|---|---|
| Nakamoto coefficient (N) | How many top holders could seize majority control? | Top of the distribution | 1 (most concentrated) → n (perfectly equal) |
| Gini coefficient (G) | How unequal is the overall distribution? | Entire distribution | 0 (perfectly equal) → 1 (monopoly) |
| NH 20% test | Does any single beneficial owner hold ≥ 20%? | The single largest holder | Pass / Fail |
A network can have a respectable Nakamoto coefficient (say, 5) while still having a very high Gini (0.85) if there are five similarly large holders and thousands of tiny ones. The use of the interactive calculator below will help illustrate how different distributions score differently across all three measures simultaneously.
To ground these abstractions in reality, the chart below plots the only verified empirical measurements available for major DAOs.
Interactive Calculator
Enter any token distribution to compute the Nakamoto coefficient, the Gini coefficient, and the result of New Hampshire’s Prong B test (the 20% bright-line). Values are normalized automatically — you can enter raw token counts or percentages.
Decentralization Metric Calculator
Enter token holdings as comma-separated numbers (any units — they are normalized to percentages).
The Aggregation Problem: When Ownership Doesn’t Equal Control
All three measures above — the Nakamoto coefficient, the Gini coefficient, and the NH 20% test — work with observable token balances at specific wallet addresses. But there is a fundamental challenge: the same underlying entity can hold tokens across many distinct wallet addresses, or can exercise control over tokens held nominally by others.
Identifying six wallet addresses, none holding more than 19%, does not establish that those six wallets are controlled by six independent actors. They might all belong to the same person, company, or affiliated group — satisfying the letter of the statute while violating its spirit.
Wallet Fragmentation
A single entity fragments its holdings across dozens of self-controlled wallets to keep each below 20%. Observable on-chain data shows many small holders; the true picture is a single dominant actor. This is addressable in principle by blockchain analytics that cluster addresses using behavioral heuristics, but is hard to prove conclusively.
Layered Corporate Ownership
Entity A owns entities B, C, and D. Each of B, C, D holds 18% of the DAO’s tokens. On-chain, these are three separate addresses. Legally, A beneficially owns 54% — far exceeding the 20% threshold. The NH statute addresses this with the phrase “directly or indirectly beneficially owned,” but tracing beneficial ownership through corporate structures can be difficult and requires off-chain investigation.
Coordinated Voting Without Common Ownership
Separate entities that are legally independent can agree (formally or informally) to vote their tokens in concert. This is coordination, not common ownership — and it is precisely the pattern that blockchain governance has historically struggled to police. The NH statute’s governance test (no single actor may direct 20% of voting power) partially addresses this, but proving coordination without a written agreement is difficult.
What this means for the 20% rule. The statute’s use of “directly or indirectly beneficially owned, in the aggregate” is a deliberate attempt to capture aggregated holdings. This is sophisticated drafting. But legal requirements and computational verifiability are different things. Any quantitative decentralization analysis should be understood as measuring observable holdings — a necessary but not sufficient condition for genuine decentralization. Robust compliance programs will combine on-chain data analysis with off-chain governance monitoring and beneficial ownership investigation.
How Different Jurisdictions Define Decentralization
This page is the first in a series examining how different legal frameworks approach the concept of decentralization. New Hampshire is distinguished by offering the most numerically specific definition. Other states that have enacted DAO legislation rely primarily on qualitative or functional descriptions.
Note on sources. The statutory descriptions below are based on enacted text as available. Readers should verify citations against current official state code compilations. Future pages in this series will provide deeper analysis of each jurisdiction.
| Jurisdiction | Law | Key Decentralization Standard | Numerical Threshold? | Approach |
|---|---|---|---|---|
| NH New Hampshire | RSA ch. 301-B (2024) | § 301-B:5 IX: Three-part test — no unilateral control, no ≥ 20% concentration, no unilateral code changes. Required “at all times” per § 301-B:15(j). | 20% (assets & votes) | Quantitative bright-line + qualitative control tests |
| WY Wyoming | W.S. § 17-31-101 et seq. (2021) | DAO LLCs may be “algorithmically managed” or “member managed.” Decentralization is reflected in the blockchain governance structure but is not defined with a percentage threshold. The law focuses on enabling smart-contract governance rather than testing for decentralization. | None specified | Functional / structural (smart-contract governance) |
| UT Utah | Utah Code § 48-5-101 et seq. (2023) | Creates the DAO LLC entity type, organized under the LLC Act. Definitions in § 48-5-102 reference decentralization conceptually but do not specify a numerical concentration threshold. Closer to Wyoming’s functional approach than to New Hampshire’s bright-line rule. | None specified | Functional / structural |
| TN Tennessee | Tenn. Code Ann. § 48-250-101 et seq. (2023) | Creates a distinct DAO entity type. Uses “decentralized” as a descriptor of the organizational form but does not define a concentration threshold. Governance through smart contracts is recognized without specific decentralization metrics. | None specified | Descriptive / organizational form |
| VT Vermont | 11 V.S.A. ch. 25 — BBLLC (2018) | Vermont’s Blockchain-Based LLC (BBLLC) was a precursor concept that allows an LLC to incorporate blockchain-based governance. It is not a DAO-specific statute and does not impose decentralization requirements. | N/A | Enabling / permissive (LLC overlay) |
| US Federal (proposed) | GENIUS Act; CLARITY Act (proposed) | Federal proposals reference “sufficiently decentralized” as a concept relevant to whether a digital asset is a security or a commodity, but define the term through multi-factor qualitative tests rather than percentage thresholds. This series will analyze federal proposals in a dedicated installment. | Multi-factor test | Securities law exemption standard |
The contrast is striking: New Hampshire is currently the only state DAO statute that offers a specific numerical threshold (20%) for token concentration. Other states rely on functional descriptions of blockchain governance structures. Each approach carries distinct trade-offs: numerical thresholds provide certainty and measurability but can be gamed through fragmentation; qualitative tests are more flexible but harder to comply with and harder to enforce.
This series will continue with deeper analyses of each jurisdiction, showing how the same mathematical tools applied here to New Hampshire can be used to interrogate any quantitative standard that emerges in law.
📊 Antitrust Perspective: The Herfindahl-Hirschman Index
The Herfindahl-Hirschman Index (HHI) is the standard metric used by the U.S. Department of Justice and the Federal Trade Commission to assess market concentration in antitrust analysis. It has been proposed as a tool for evaluating token concentration in blockchain networks, drawing an analogy between token-holders and market participants.
Let pi = (si / T) × 100 be entity i’s percentage market share (0–100). Then:
Range: near 0 (many equal participants) to 10,000 (monopoly). The DOJ guidelines classify markets as: unconcentrated (< 1,500), moderately concentrated (1,500–2,500), and highly concentrated (> 2,500).
Worked Example
Consider a five-holder distribution at 19%, 19%, 19%, 19%, 24%:
- p1 = 24, p2–p5 = 19 each
- HHI = 24² + 4×19² = 576 + 4×361 = 576 + 1,444 = 2,020
This falls in the “moderately concentrated” range. Note that this distribution also fails the NH statute (24% > 20%), so HHI analysis and the NH bright-line test are picking up the same issue through different lenses.
HHI and the 20% Threshold
If one entity holds exactly 20% of tokens, its HHI contribution alone is 20² = 400. A network with five entities at exactly 20% each would have HHI = 5×400 = 2,000 — right in the moderately-concentrated range. A network that just barely passes the NH test (each holder at 19.99%) with five equal holders would have HHI ≈ 2,000 as well. This suggests that the NH threshold and antitrust concentration guidelines are calibrated at similar levels of concern.
Limitations of HHI in the DAO Context
HHI was designed for product market competition analysis, where market shares are assumed to be stable and where the relevant actors are firms competing for sales. Token holdings in a DAO are different in important ways:
- Token holdings can be highly volatile and change rapidly.
- The relationship between token holding and decision-making power depends on the specific governance mechanism.
- Unlike markets where concentration is presumptively bad, some degree of token concentration may be necessary for governance efficiency.
These limitations suggest HHI is most useful as a supplementary measure rather than the primary analytical tool for DAO decentralization. This is why it appears here in an expandable section rather than as the centerpiece of the analysis. For deeper discussion of the antitrust analogy and its implications for DAO law, see the work cited in the resources section below.
Resources and Further Reading
This page is part of the NH DAO Law Explorer project, which provides open, interactive tools for understanding DAO legislation across jurisdictions. Future installments will apply the same analytical framework to Wyoming, Utah, Tennessee, Vermont, and federal proposals.
Primary Sources
- N.H. Rev. Stat. Ann. ch. 301-B — Full statutory text with interactive annotations (this repository)
- New Hampshire Secretary of State — Official state filings and registry
- HB 645 — Bill Status and Legislative History
Scholarship
- Oranburg, S.C. (2025). Market Power and Governance Power: New Tools for Antitrust Enforcement in the Decentralized Gig Economy. Competition Policy International. — Proposes the dual-metric framework (HHI + Nakamoto/Gini) and the four-quadrant organizational typology.
- Oranburg, S.C. (2024). Antitrust Law for Blockchain Technology. 49 J. Corp. L. 379. — Precursor analysis of decentralized-distributed network structure.
- Fritsch, R., Müller, M. & Wattenhofer, R. (2022). Analyzing Voting Power in Decentralized Governance: Who Controls DAOs? arXiv:2204.01176. — Empirical measurements: Compound Gini 0.987, NMC 8; Uniswap Gini 0.995, NMC 11.
- Cong, L.W. et al. (2020). Decentralized Governance in Blockchain-Based Organizations. Oxford Review of Economic Policy 36(2). — Introduces the Nakamoto coefficient as a governance metric.
Mathematical Foundations
- Srinivasan, B. and Lee, L. (2017). Quantifying Decentralization. — The original formulation of the Nakamoto coefficient across multiple blockchain subsystems.
- Gini, C. (1912). Variabilità e mutabilità. — The foundational work on the inequality coefficient.
Related Projects
- NH Token Commission — Commission to Study Stable Tokens (RSA 383:26, HB 310)
- UNH InterOperability Laboratory — NH DAO Registry Administrator
- GitHub Repository — Source, issues, and contributions
Maintained by Seth C. Oranburg, Professor of Law, UNH Franklin Pierce School of Law. Statutory text is public law. Educational analysis and display tool © 2026 Seth C. Oranburg. MIT License.