Proposer-Builder Separation (PBS) introduces a Principal-Agent Problem, in which the proposer (the Principal) requests a builder (the Agent) to produce a block for them. But if the builder does not have the same value function as the proposer, it can lead to blocks being built that do not respect what a proposer would want for them, e.g., censoring transactions.
We are looking for a model of the cost of censorship under different assumptions.
- What is the cost of censorship of a builder who wants to “turn off” block production on Ethereum, under inclusion lists or partial block building?
- What is the cost of censorship of a builder who wants to censor a single transaction, under inclusion lists or partial block building?
- In these models, what are contracts that rational/greedy builders and proposers would enter into, e.g., “I won’t add this tx to my inclusion list unless you pay me $x$” or “I won’t build a partial block containing the transaction unless you pay me $y$”. Is $x$ lower than $y$?
Resources
Deliverables