Gas, the fee market and EIP-1559

Wonder

1. How does a protocol determine the price of computation without a central administrator setting rates?
2. What is the mechanical consequence of splitting a transaction fee into two distinct streams with different destinations?
3. Why would developers accept a more complex fee structure in exchange for predictable costs and supply deflation?

Explore

1. In the early years of Ethereum, users paid a single gas price to include transactions in blocks. This created a volatile auction where fees spiked unpredictably during high demand, often rendering the network unusable for average participants. The London upgrade, activated on 5 August 2021 at block 12,965,000, introduced EIP-1559 to solve this instability. It replaced the simple bid with a two-component fee structure that separates the protocol’s revenue from the validator’s incentive, fundamentally altering how value flows within the system.
2. The first component is the base fee, which is calculated by the protocol itself rather than set by users or validators. This fee adjusts block by block to target blocks that are half-full on average. If a previous block exceeded this capacity, the base fee rises; if it was underfilled, the fee drops. Crucially, this entire base fee is burned—removed from circulation permanently. It does not go to the validator who produced the block, nor to any treasury. This mechanism ties network activity directly to ETH supply reduction.
3. The second component is the priority fee, or tip, which users specify to incentivize validators to include their transactions in the current block rather than waiting for a future one. Unlike the base fee, this amount goes entirely to the block producer. Users set a maximum willingness to pay (maxFeePerGas) that covers both components. If the market base fee is lower than their max, they only pay the actual base fee plus their tip; any unused portion of their max bid is refunded immediately. This caps user exposure while maintaining competitive ordering.
4. This structure contrasts sharply with Bitcoin’s fee model (lesson 2.2.3), where fees are a simple auction for block space and every fee goes to the miner. Bitcoin has no burn mechanism; its supply schedule is fixed by halving events. Ethereum’s EIP-1559 introduced a dynamic variable: when network activity is high, the base fee rises and burns more ETH, potentially making the asset deflationary. This turned fees from pure transaction costs into a macroeconomic lever affecting total supply.
5. The predictability of the base fee allows developers and users to plan costs with greater accuracy than in the pre-London era. However, it did not eliminate congestion; high demand still drives the base fee up, making transactions expensive. The complexity of managing two fee parameters (maxFeePerGas and maxPriorityFeePerGas) adds friction for casual users compared to a single price tag. The burn also tied fee demand to ETH’s supply, which the next lesson follows.
6. Open a block explorer like Etherscan and inspect any transaction from August 2021 onwards. Look for the ‘Base Fee’ and ‘Priority Fee’ fields in the transaction details. You will see the exact amount burned (equal to the base fee times gas used) and the tip paid to the validator. Compare this with a pre-August 2021 transaction, which only displays a single ‘Gas Price.’ This visual distinction confirms the mechanical split of fees introduced by EIP-1559.

Key facts

Exercise nine, six, three, one

Fact check

Dates, figures and claims checked against the published sources listed below.

• Ethereum Improvement Proposal 1559 (EIP-1559) Specification
• Ethereum Foundation, 'London Mainnet Announcement' (July 2021)
• Mastering Bitcoin, 3rd ed. (for fee model contrast)