# Arbitrum Delegate Concentration: The Numbers Behind the Power Asymmetry *A Research Collaboration | MconnectDAO x ChainSightsOne | March 2026* By [MconnectDAO.eth Research](https://paragraph.com/@mconnectdaoresearch) · 2026-03-26 arbitrum, governance, defi, dao, web3, delegate, on-chain data, token governance, research, crypto, arb, defi governance, blockchain research, mconnectdao, chainsightsone --- **TL;DR** --------- * Arbitrum governance power is structurally concentrated in a small set of large wallets, not broadly distributed across the community. * 40% of the top 20 delegates are pure self-delegations, with effectively zero external accountability from delegators. * Delegation was meant to solve participation, but with this concentration pattern it mostly launders capital power rather than encoding real community trust. * * * **Key Takeaways** ----------------- * Token-weighted, delegate-based governance can look decentralized while being controlled by a narrow band of capital at the top. * A sharp gap between Tier 1 and Tier 2 delegates shows a structural break, not a smooth community-driven gradient of trust and voting power. * Part 1 measures who holds the power; Part 2 will measure what they actually do with it via longitudinal voting and participation data. * * * **Introduction** ---------------- In March 2026, MconnectDAO published a framework arguing that Arbitrum's governance crisis is not a participation problem at its core — it is a structural design problem. Token-weighted governance concentrates power at the top. The people with the most to gain financially vote. The people who build, research, and contribute often hold fewer tokens and therefore less say. This produces a system that looks decentralized on paper but operates through a narrow band of capital. That argument was based on observation. What ChainSightsOne has now done is measure it. This article presents the first quantitative layer of the accountability asymmetry thesis. We are co-authoring it because the framework and the data need each other. A thesis without numbers is opinion. Numbers without a framework are noise. This is Part 1. It covers delegate concentration — who holds the power and how fragile the accountability structure around that power actually is. Part 2 will cover participation rates — what those delegates actually do with the power they hold — once ChainSightsOne's API rolls out longitudinal voting data in the next 4–6 weeks. * * * **What the Data Shows** ----------------------- Snapshot: March 24, 2026 | Source: Tally API / Arbitrum Governor `= * self-delegation (1 delegator)` **Top 20 Delegates (Tracked)** ------------------------------ **Rank** **ARB Delegated** **Share (Top 20)** **Delegators** **Tier** **Name / Address** 1 15,000,002 11.0% 6 T1 0x11dA8Ae2... 2 14,986,062 11.0% 4 T1 0x465f3922... 3 9,295,469 6.8% 1 \* T1 0x19D373d0... 4 9,003,825 6.6% 24 T1 0x53689948... 5 9,000,000 6.6% 1 \* T1 0x030d6830... 6 8,007,118 5.9% 9 T2 Ian Lapham 7 8,002,092 5.9% 78 T2 jessewldn 8 8,000,001 5.9% 13 T2 0x1d8F369F... 9 7,300,000 5.4% 1 \* T2 0x5F246D7D... 10 7,250,000 5.3% 9 T2 0x88E15721... 11 7,000,000 5.1% 5 T2 0xcb70D1b6... 12 5,311,814 3.9% 2 T2 0x4421aec4... 13 5,000,000 3.7% 1 \* T2 Kain / Synthetix 14 5,000,000 3.7% 7 T2 0x2193c9a8... 15 4,000,000 2.9% 2 T2 0xCFaaE1aE… 16 3,100,000 2.3% 1 \* T2 0x2fD0eB27… 17 2,900,000 2.1% 1 \* T2 0xaEbb1719… 18 2,891,295 2.1% 5 T2 Boris Stanic (The Stable) 19 2,750,000 2.0% 1 \* T2 0x069E40A2… 20 2,646,496 1.9% 1 \* T2 0x9c980d9E… * * * **Finding 1: Self-delegation is more common than the ecosystem acknowledges** ----------------------------------------------------------------------------- 8 out of the top 20 tracked delegates have exactly 1 delegator — ranks 3, 5, 9, 13, 16, 17, 19, and 20. That is 40% of the tracked top 20. In practice, this means a single address has delegated to itself and holds that governance power without any external accountability mechanism. There is no group of delegators who could revoke. There is no signal of community trust behind the position. There is simply capital, pointed at itself. This is not low accountability. It is zero accountability by design. * * * **Finding 2: The top of the ranking reflects capital concentration, not community trust** ----------------------------------------------------------------------------------------- Ranks 1 and 2 each hold approximately 15 million ARB, supported by only 4 and 6 delegators respectively. The arithmetic here is worth sitting with: one of the most powerful voices in Arbitrum governance is accountable to fewer people than fit in a small meeting room. This is not an attack on those delegates personally. It is an observation about structural legitimacy. Community-built trust accumulates through many small decisions, made by many different people, over time. Four to six delegators does not describe that process. It describes concentrated capital choosing where to park its governance power. * * * **Finding 3: The Tier 1 vs Tier 2 gap confirms the concentration gradient** --------------------------------------------------------------------------- **Metric** **Tier 1 (Rank 1–5)** **Tier 2 (Rank 6–20)** Avg ARB per delegate 11,457,072 ARB 5,277,254 ARB Avg delegators 7.2 9.1 Total voting power 57.3M ARB (42%) 79.2M ARB (58%) The cliff factor is 2.17x. Tier 1 delegates hold more than twice the voting power per seat compared to Tier 2, with fewer delegators per seat. That gap is not a gradient. It is a structural break. The fingerprint here is fund and whale capital dominating the top tier, rather than organic community delegation scaling upward. * * * **Why This Matters More Than It Looks** --------------------------------------- The delegation mechanism was designed to solve the participation problem. Instead of every token holder needing to vote on every proposal, they could delegate their voting power to someone they trust to represent their interests. This is a reasonable design. It fails when the delegates themselves are primarily self-delegating large wallets with minimal external accountability. In that condition, the delegation layer does not represent community trust. It launders capital concentration through the appearance of a participation system. The result: a DAO where proposals can pass or fail based on the decisions of wallets that are accountable to almost no one. * * * **What We Are Not Saying** -------------------------- We are not saying these delegates are acting in bad faith. Several of them may have strong governance track records. We are not able to evaluate that from concentration data alone. That is precisely why Part 2 matters. A delegate with 15 million ARB who votes on 90% of proposals with clear reasoning is a very different governance participant from one who votes 20% of the time with no stated rationale. The concentration data cannot tell us which is which. What the concentration data can tell us: the accountability structure around these positions is structurally thin. Whether or not they are using that power responsibly, almost no one has the practical ability to revoke it. That asymmetry — power without revocation pressure — is the core governance risk this research is documenting. * * * **Coming in Part 2** -------------------- ChainSightsOne's API is currently building delegate-level participation tracking for Arbitrum. In 4–6 weeks, we will have longitudinal voting data that answers the questions this article cannot: * What percentage of proposals do the top delegates actually vote on? * Do delegates with more delegators vote more consistently than self-delegating wallets? * Does participation correlate with concentration, or are they independent variables? Part 2 will complete the accountability asymmetry argument with that behavioral layer. If the concentration problem is structural, participation data will tell us whether it is also operational. * * * **Closing** ----------- Governance infrastructure is only as trustworthy as the accountability it encodes. Arbitrum's current delegate structure, as this snapshot shows, encodes very little revocation pressure at the top. Power concentrates. Delegators are few. The community's ability to course-correct through the delegation mechanism is weaker than the system's design implies. This is a solvable problem. But solving it requires first measuring it honestly. That is what this collaboration is trying to do. MconnectDAO — Governance Researcher, India. Framework and analysis. ChainSightsOne — On-chain data and API infrastructure. March 2026 | Independent research. No financial interest in any protocol mentioned. --- *Originally published on [MconnectDAO.eth Research](https://paragraph.com/@mconnectdaoresearch/arbitrum-delegate-concentration-the-numbers-behind-the-power-asymmetry)*