Proposer-Builder Separation - PBS - gets mentioned in MEV discussions and then quietly dropped. That’s a mistake. It’s one of the more structurally significant changes to how Ethereum’s block production actually works, and the implications run well past gas optimization.
What the Split Actually Means
Before PBS, validators did two things: they built blocks (selecting and ordering transactions) and they proposed those blocks to the network. Combining those roles created a problem. Validators with sophisticated MEV extraction software could earn dramatically more than those without it, concentrating rewards and putting smaller validators at a structural disadvantage.
PBS separates these roles. Specialized actors called builders compete to construct the most profitable block they can, then submit sealed bids to proposers (validators). The proposer simply picks the highest bid, takes the payment, and signs the block without seeing its contents until after commitment. The builder captures MEV; the proposer captures the bid. Neither needs to trust the other.
In practice, Ethereum has been running a version of this through MEV-Boost - middleware built by Flashbots - since shortly after the Merge. MEV-Boost introduces a third actor, the relay, which acts as an escrow between builder and proposer, verifying the block’s value before the proposer commits. By late 2023, a significant majority of Ethereum blocks were being built through MEV-Boost relays rather than locally by validators.
The Centralization Problem It Creates
Here’s the part that tends to get glossed over: PBS doesn’t eliminate centralization pressure, it relocates it. Block building is now an extremely concentrated market. A handful of builders - some affiliated with large trading firms - produce a disproportionate share of Ethereum blocks. They have the infrastructure, the order flow relationships, and the capital to consistently outbid smaller competitors.
This is a known concern, not a speculation. Ethereum researchers have been tracking builder market concentration openly, and it’s reflected in public relay data.
Why Enshrined PBS Matters
The current MEV-Boost setup is opt-in and relies on trusted relays - entities that can theoretically censor transactions or fail. Enshrined PBS (ePBS), which would bake the proposer-builder split directly into Ethereum’s protocol, removes relay trust assumptions and makes the separation a first-class primitive rather than middleware.
ePBS has been in Ethereum research discussions for years. Its absence from recent hard forks reflects genuine design difficulty, not lack of urgency. The relay-based status quo works well enough that pressure to finalize ePBS remains lower than it probably should be, given how much of Ethereum’s block production now flows through a small number of off-protocol intermediaries.