Imagine a democracy where ordinary citizens turn up to vote in record numbers.

But the elected representatives they trusted with their power? Nowhere to be found.

That's exactly what's happening inside Arbitrum DAO right now. And the April 2026 data makes it impossible to ignore.

Human Participation Rate: 9.9 / 10 (A+)

Delegate Engagement Index: 1.9 / 10 (F)

This isn't just a gap. This is a governance crisis hiding in plain sight.

Everyday community members are showing up at record levels. Yet the delegates entrusted with millions of ARB tokens in voting power watched their engagement score collapse from 7.0 in February to 1.9 in March in a single month during an active Security Council Election.

The people with the most power are the least present.

1. 6,000 ETH Treasury Transfer 6,000 ETH sits idle in Arbitrum's L2 Treasury Timelock — earning zero yield, generating zero strategic value. Entropy's proposal: move it into the Treasury Management Portfolio. Estimated annual return: 288.6 ETH at ~4.81% yield Current result: Dominant — overwhelming delegate support

2. Security Council Member Election Running April 12 → May 3. A vote decay function is now live — the longer delegates wait, the less voting power they have. And still, many haven't voted.

3. ArbOS 60 Elara — Constitutional upgrade Passed 4. Arbitrum Audit Program Improvements Passed 5. Code of Conduct — Living Documents Passed

Nearly every April proposal passed as "Dominant." No real contest. No genuine debate.

• Entropy: 8.5M → 25.7M ARB (+201.6%) — biggest absolute gain

• 0xfrisson.eth: 4.2M → 7.2M ARB (+72%)

• MUX: 7.4M → 12.4M ARB (+67%)

• dk.3anon.eth: 6.5M → 10.6M ARB (+62%)

• blockworksres.eth: 6.8M → 10.8M ARB (+58.7%)

• jojothecow.eth: 0.08M → 1.05M ARB (+1,156%) — biggest percentage jump

• 0x-ultra-gov.eth: -0.17M ARB (-18.7%)

• woofi.eth: -0.09M ARB (-16.5%)

• l2beat.eth: -0.20M ARB (-1.3%)

Some addresses dropped to zero — fully exiting governance with no explanation.

Looking at Arbitrum's full governance history:

• Dominant outcomes: 65.5% of all proposals

• Majority outcomes: 19.3%

• Contentious (close vote): only 15.3%

Proposals by category:

• Incentives: 49.6% — by far the most common

• Governance: 13.2%

• Grants: 11.5%

• Protocol Upgrades: 7.9%

• Treasury: 4.4%

• Operations: 4.0%

The pattern is clear: delegates strongly support narrowly scoped, transparent proposals. Broad funding requests get rejected. The DAO rewards specificity and punishes ambiguity.

ChainSights — an independent platform tracking 53+ DAOs — published this snapshot for March 2026:

• Active Delegates (65%+ participation rate): just 87

• Inactive Delegates: 108

• Ghost Delegates (received delegation, never voted): 805

805 ghost delegates.

Token holders trusted these addresses with their voting power. And those addresses simply don't participate.

Arbitrum's overall governance score sits at 6.2 / 10 ranked #17 of 53 DAOs. The infrastructure is there. The community is engaged. But the Power Dynamics Index scores just 4.8 / 10 (D) voting power is dangerously concentrated.

1. Token-weighted voting has an accountability gap More tokens = more power. But absence doesn't reduce power. A delegate can hold 25 million ARB, vote on nothing, and face zero consequences. The system doesn't punish disengagement.

2. Financial incentives alone don't drive participation The common theory: delegates show up when the stakes are high. But during a Security Council Election — which determines who guards the protocol engagement collapsed. If financial incentives were the primary driver, this shouldn't happen.

3. Concentration is the real risk The top 50 delegates control 56% of Arbitrum's voting power. When they disengage, the remaining 44% doesn't matter. What looks like decentralization is structurally fragile.

Arbitrum has responded with two moves:

The Rewarding Active Delegates (RAD) Program — delegates earn rewards based on their voting record and published rationale. Quality over quantity.

Updated DAO Procedures (April 2, 2026) — predictable voting schedules, mandatory discussion periods, and a structured governance calendar to reduce delegate fatigue.

These are real steps. But they raise a deeper question:

Can incentive programs fix broken governance or do we need to rethink token-weighted voting entirely?

Arbitrum has world-class infrastructure. A passionate community. Billions in treasury. A proven track record of shipping protocol upgrades.

But 805 ghost delegates sit at the center of its governance. Holding power they don't use. Accountable to no one.

Decentralization isn't just written in code. It has to show up in participation.

The community already does. Now it's the delegates' turn.