Proposer-builder separation - PBS - started as a technical fix for a specific MEV problem: validators were capturing maximal extractable value directly, which created perverse incentives to reorganise blocks for profit. The solution was to split the job. Builders assemble blocks. Proposers just pick the highest-value one and sign it. Clean division of labor.
What nobody fully anticipated was how rapidly the builder market would concentrate.
The Centralisation No One Voted For
As of mid-2026, a small number of block builders account for the overwhelming majority of Ethereum blocks under the mev-boost relay architecture. This isn’t a conspiracy - it’s a structural outcome. Sophisticated builders with access to more order flow, better algorithms, and tighter latency can extract more value per block. Validators rationally choose whoever bids highest. The market selects for scale.
This matters because builders aren’t neutral infrastructure. They decide which transactions get included, in what order, and which MEV opportunities to exploit or pass over. A validator following PBS has effectively outsourced that entire decision. The validator signs a block it never actually saw built.
What ePBS Changes - and What It Doesn’t
Enshrined PBS (ePBS) would bring the proposer-builder relationship into Ethereum’s core protocol rather than leaving it to the off-chain mev-boost stack. The theoretical benefit is that the current relay layer - an intermediary that validators trust to deliver honest bids - disappears. Proposers would receive cryptographic commitments directly, reducing trust assumptions.
But ePBS doesn’t fix the underlying economics that drove consolidation in the first place. It formalises the separation at the protocol level without addressing why three or four builders consistently dominate. If anything, enshrining PBS could entrench those players more permanently by making the architecture official.
The FOCIL Angle
Fork-choice enforced inclusion lists (FOCIL) represent a different line of attack. Rather than controlling who builds blocks, FOCIL forces builders to include a set of transactions pre-selected by a committee of attesters. The goal is censorship resistance - ensuring that builders can’t quietly exclude transactions from specific addresses or protocols.
FOCIL and ePBS are meant to work together, and that combination does change the builder’s power in a meaningful way. Builders retain ordering rights - which is still substantial MEV territory - but lose veto power over inclusion.
Whether that’s enough of a constraint is genuinely unclear. The ordering layer alone is where a large share of extractable value lives, and nothing in the current roadmap eliminates the economic gravity pulling block construction toward a handful of specialised actors. The architecture is becoming more sophisticated. The centralisation question is still open.