The criticism follows naturally from how proof of stake looks on the surface. Stake more ETH, earn more ETH. That sounds like a mechanism designed to compound advantage — the large get larger, those who arrive with less capital fall further behind.
The honest answer is: sort of, but probably not in the way the criticism implies. The yield mechanics are more egalitarian than they appear. The genuine concern lives elsewhere.
How Staking Rewards Actually Work
In a proof of stake network, validators lock up capital — on Ethereum, a minimum of 32 ETH per validator — and earn rewards for participating in block proposals and attestations. More validators means more absolute earnings.
But the yield rate is the same for everyone.
A solo validator running 32 ETH earns roughly the same annual percentage as someone operating 100 validators with 3,200 ETH. As of May 2026, Ethereum staking yields approximately 3–5% annually, adjusting with the total amount of ETH staked (more stake participation lowers the per-validator rate). That rate is applied uniformly. There's no VIP tier, no institutional rate, no volume discount built into the protocol.
What this means in practice: someone staking 32 ETH earns about 1–1.5 ETH per year. Someone staking 3,200 ETH earns about 100–150 ETH per year. Larger holders earn more ETH in absolute terms, but not because the system is tilted on rates. They earn more because they have more capital — exactly as they would in any yield-bearing asset, from government bonds to index funds.
The yield structure isn't the mechanism creating inequality. Existing wealth is.
Where the Criticism Has Real Weight
This is where the honest version of the critique lives.
The 32 ETH minimum is a real barrier. At current prices, that's a substantial capital commitment most retail participants can't easily clear. Solo staking is concentrated among larger holders by design — running a production validator with high uptime requires dedicated hardware and a meaningful capital commitment. This is a practical access problem, not a yield problem.
MEV creates genuine asymmetry. Maximal Extractable Value — the profit validators can extract by ordering or inserting transactions in a block — isn't uniformly distributed. Validators running MEV-Boost software and connected to sophisticated block-building networks extract more per block than those running standard configurations. Scale and tooling matter here. This is a real mechanism through which larger operators capture returns that smaller operators miss. It's on top of base staking rewards and it's not trivially accessible at every scale.
Liquid staking concentrates governance. Protocols like Lido, Rocket Pool, and Coinbase's cbETH exist to solve the minimum-stake problem — they pool funds, issue liquid tokens (stETH, rETH), and let anyone participate with any amount. That's genuinely useful. But Lido alone represents roughly 28–32% of all staked ETH as of early 2026. The node operators who run the underlying validators for Lido control that stake, not the individual stETH holders. This creates a concentration of network influence that has nothing to do with yield rates and everything to do with who controls which validators.
So: the yield mechanic is flat and proportional. The access mechanic and the governance mechanic are where the concentration risk actually lives.
Proof of Work Doesn't Escape This
One assumption embedded in the PoS critique is that proof of work was more egalitarian. It wasn't, particularly.
Bitcoin mining requires capital for ASICs plus ongoing operational costs (electricity). Industrial-scale miners with access to cheap power and bulk hardware pricing have substantial cost-per-unit advantages over individuals. Mining rewards flow proportionally to hash rate contributed, and the industrial concentration in Bitcoin mining is well-documented. Large operations in North America and elsewhere account for a significant share of global hash rate.
The mechanisms differ, but the inequality dynamic is similar. Capital-intensive systems tend to benefit those with more capital — whether that capital is hardware or staked tokens.
What's Changing
Distributed Validator Technology (DVT) is being developed to address the concentration concern directly. Protocols like Obol Network and SSV Network split validator keys across multiple independent nodes, enabling groups of participants to jointly run a validator with shared responsibility. DVT is still early in adoption but is designed to reduce the risk that large liquid staking operators control meaningful validator shares without distributed oversight.
Liquid staking governance reform is ongoing. Rocket Pool's permissioned-to-permissionless node operator transition, Lido's governance debates around its validator set, and EigenLayer's restaking primitives are all shifting how staking influence is allocated. These structural changes matter more for the inequality question than the basic yield arithmetic.
Confirmation Signals
The clearest signals to watch: Lido or any single entity's market share approaching 33% of total staked ETH (the theoretical threshold for interference with finality). DVT adoption rates across liquid staking protocols — if they remain near zero, concentration risk isn't being mitigated. MEV extraction distribution between large and small validators: narrowing suggests democratization; widening suggests the opposite.
What Would Break This Picture
If liquid staking consolidation continues unchecked and DVT adoption stalls, the concentration critique sharpens considerably — not through yield mechanics, but through governance influence over a meaningful fraction of the network. Separately, if Ethereum protocol changes lower the 32 ETH minimum validator threshold (not currently proposed), the access barrier narrows directly and channels fewer retail participants through intermediaries.
Timing
Now: Staking yields are flat-rate across all validators; MEV asymmetry favors sophisticated operators; liquid staking concentration is the live risk that matters most. Next: DVT adoption trajectory over the next 12–18 months is the primary indicator of whether concentration risk improves. Later: Protocol-level changes to minimum validator deposits and MEV architecture are longer-horizon questions, likely years away if they happen at all.
Boundary
This covers the mechanics of PoS reward distribution and the structural access and governance risks. It doesn't address whether ETH or any staking arrangement is appropriate for any particular situation.
The yield structure is proportional but equal-rate — not tilted toward large holders on returns. The concentration risk is real but lives in governance and MEV, not in the base staking formula. These are different claims, and conflating them makes both harder to evaluate.
