Transaction Fee Mechanism Design

The five-minute read

One action, walked through

1. Users submit bids according to the mechanism's message space.
2. The miner chooses a valid block, possibly after adding its own transactions.
3. Payments and burns are applied by the rule, not by the miner's discretion, if the mechanism is followed.
4. The paper asks whether following it is in the miner's interest, given the bids.
5. Separate theorems vary how strategic the users are allowed to be.

The argument, unpacked

What has to be true

• The miner is myopic in the sense the theorem states.
• Users match the strategic model of that theorem, which is not always 'fully strategic'.
• The chain enforces the burn. If it does not, the mechanism is a different one.
• Bids are the messages the mechanism defined. Off-chain side payments are a different game.

What happened after the paper

This is the paper to read next to EIP-1559, and before believing a new fee rule is 'incentive compatible' without saying for whom. Flash Boys 2.0 is the companion on ordering, which this paper largely sets aside.

What to check before you use the idea

• Which of allocation, payment and burn changed?
• Is the miner in the claim myopic?
• Does the claim survive side payments?
• Is the citation being used as a price or a valuation? It should not be.

Terms

Burning rule

The part of a fee mechanism that destroys value rather than paying it to the block producer.

Myopic miner

A producer who maximises the payment of the current block, ignoring longer strategies.

The problem the paper names

A fee rule is a mechanism: users choose bids, a miner or builder chooses a block, and payments are split. The paper asks which rules survive a miner who wants to be paid more.

What the design proposes

• A transaction fee mechanism specifies an allocation rule, a payment rule, and a burning rule.
• Myopic miners are the strategic agents in the formal question. Users may be price-takers in the simplest theorems.
• EIP-1559 is analysed as one point in that space, next to a first-price auction.

How the mechanism is specified

• The interesting constraint is that the miner should not want to include fake transactions or to collude to rewrite the posted price.
• Burning is a tool in that argument. It is not, in the paper, a monetary-policy recommendation.
• The results are theorems under the paper's equilibrium notions. They are not measurements of a mempool.

What this page does not treat as proven

• Myopic means the miner is not playing a long game across blocks. A builder with a private order flow is a harder agent than the model.
• The paper does not set a fee you should pay, and it does not value a token.
• Later fee markets, including blobs, need their own argument. They are not corollaries printed here.

Why a venture studio still reads it

Use the paper as the vocabulary: allocation, payment, burn. Then ask which of those a new chain actually specified, and which strategic miner the proof assumed.