Most people see a 95.8% proposal pass rate and think: healthy governance. I see it and think: who's actually voting?

I spent the last few weeks analyzing Entropy Advisors' Dune Dashboard and Arbitrum's official RAD Reports (January–March 2026). What I found should concern every ARB holder, delegate, and Web3 governance enthusiast.

The Numbers Look Good — Until You Dig Deeper

Out of 84 total proposals in 2026:

• Passed → 69 (82.1%)

• Defeated → 13 (15.5%)

• 🚫 Cancelled → 1 (1.2%)

On the surface, this looks like an engaged, functioning DAO. But here's the uncomfortable truth hiding behind these numbers.

50 Delegates Control ~90% of Governance

• Top 10 delegates hold 40.70% of total voting power

• Delegates ranked 11–50 hold another 49.30%

That's 50 wallets deciding the fate of a billion-dollar protocol. When they vote yes, proposals pass. It's that simple and that fragile.

The Ghost Delegate Problem Is a Governance Emergency

Arbitrum has ~986 registered delegates. Sounds decentralized, right?

Look closer:

• Active delegates (65%+ participation rate) → only 91

• 🔴 Ghost delegates (near-zero activity) → 787

787 wallets hold real ARB voting power. 787 wallets never show up.

This isn't voter apathy. This is structural failure. Real token holders delegated their voting power in good faith and it's sitting dormant, unrepresented.

Is the RAD Program the Fix?

Arbitrum launched the RAD (Rewarding Active Delegates) program — $1.5M/year in participation incentives. The intent is right. The execution is still catching up.

February 2026 results:

• 37 enrolled delegates

• Only 29 actually voted

• 140 million ARB cast

• 5 delegates missed the 75% threshold → disqualified from rewards

March 2026 showed improvement, but many delegates still missed the 5-day public rationale deadline and lost their rewards.

RAD is a step forward. But it only touches 37 delegates. The 787 ghosts remain completely unaddressed.

What Arbitrum Actually Needs

A high pass rate doesn't mean strong governance. It means the same small group keeps agreeing with each other.

For Arbitrum to earn the label "decentralized," the DAO needs to seriously pursue:

• DIP 2.0 implementation — reform delegation mechanics at the protocol level

• Contribution-based voting weight — reward consistent, quality participation

• Ghost delegate cleanup — auto-expiry or re-delegation triggers for inactive wallets

• Broader community onboarding — especially from underrepresented regions like India, Southeast Asia, and Africa

The data is public. The tools exist. The Entropy Dashboard shows everything in real time.

What's missing isn't information it's political will.

A Personal Note

As India's DAO governance researcher and an active Arbitrum delegate, I believe governance quality is the most underrated risk in Web3. Smart contracts can be audited. Tokenomics can be modeled. But a governance system captured by 50 wallets? That's a vulnerability no audit will catch.

If you hold ARB — check who you've delegated to. Are they showing up?

If you're a delegate — your rationale matters more than your vote count.

If you're building in Web3 — governance is not an afterthought. It's the protocol.

📌 Data Source: Entropy Advisors Dune Dashboard | Arbitrum Forum RAD Reports (Jan–March 2026)