The history of Ethereum’s EIP-1559 fee burn

Mechanics of EIP-1559

Vitalik Buterin, Eric Conner, Rick Dudley, Matthew Slipper, Ian Norden, and Abdelhamid Bakhta authored the EIP-1559 proposal to replace the inefficient first-price auction model. This reform activated at block 12,965,000 during the London hard fork on August 5, 2021. Before this, users guessed gas prices, which led to overpayment of 2 to 5 times the market-clearing rate. The protocol now sets a base fee per block based on the fullness of the previous block. If a block fills above the target, which equals half the gas limit, the base fee rises by up to 12.5% in the next block. If the block fills below the target, the base fee falls by up to 12.5%. The mathematical formula dictates the change: $base_fee_next = base_fee_prev + base_fee_prev * (gas_used – target) / target / 8$. For example, if the base fee is 50 gwei and a block uses five thirds of the target gas, the next base fee becomes 54.17 gwei. To manage costs, users specify a maximum fee and a maximum priority fee. Because of this mechanism, the fee remains predictable for the next block. The minimum base fee remains 7 wei. Most wallets like MetaMask, Rabby, Coinbase Wallet, and Phantom handle the maxFeePerGas and maxPriorityFeePerGas split internally. The new Type 2 transaction format uses these specific fields to prevent the errors common under the legacy GASPRICE mechanism.

Scaling and Fee Compression

Three major upgrades since the 2021 London fork reshaped the fee market without altering the core formula. Dencun arrived in March 2024 and introduced EIP-4844, which gave rollups a separate blob fee market. This meant rollup data stopped competing with ordinary transactions for L1 blockspace. Pectra arrived on May 7, 2025, and included EIP-7691, which raised the blob target from 3 to 6 per block and the maximum from 6 to 9. It also included EIP-7702 for wallet batching and EIP-7251, which allows validators to consolidate stake up to 2,048 ETH. Fusaka followed on December 3, 2025, with PeerDAS and EIP-7935, which raised the L1 gas limit from 30 million to 60 million. These shifts moved data to blobs and increased capacity, which pushed typical base fees much lower. In May 2026, Etherscan showed standard gas around 0.15 gwei. I observe that the transition from 8 to 50 gwei during the 2021 congestion to the sub-1 gwei levels of 2026 proves the effectiveness of the scaling roadmap, a fact you would recognize if you watched the 2021 NFT boom. Most EVM-equivalent L2s, including Arbitrum, Optimism, Base, Polygon zkEVM, and zkSync, now implement 1559-style mechanics to manage their own markets. Arbitrum sets its pricing algorithm targets as gas per second over a rolling window rather than per block. L2 execution fees typically run in the sub-cent to low-cent range because of these updates.

The Deflationary Milestone

The Ethereum network reached a $12 billion deflationary milestone in November 2026 because the cumulative burn reached unprecedented levels. Because the protocol burns the base fee instead of paying it to validators, every single transaction reduces the total ETH supply, a mechanism that helped the network reach a $12 billion deflationary milestone in November 2026. Ultrasound Money’s 4.2 million ETH destroyed tracker captures this massive reduction in supply. Since the London upgrade, the network torched over 4.6 million ETH, which equals roughly $13.57 billion in value over 1,438 days. This averages 2.22 ETH going up in flames every minute. While the burn reduces supply, the network still issues approximately 3,000 ETH per day as of September 2026. Staked ETH reached about 43 million as of the same month. Ethereum enters periods of net deflation when the burn outpaces this daily issuance. This condition requires sufficient transaction demand to drive the base fee above minimal levels. In 2021, the network burned over 800,000 ETH, which was worth $3.8 billion. Ethereum’s median issuance rate since the London hard fork sits at 0.801%, which is lower than Bitcoin’s current rate of 0.809%. Does the massive shift of activity to Layer 2 networks threaten the ability of the base fee burn to maintain the ultrasound money narrative?

MEV, Validators, and Consensus

Flashbots maintains dominance through its MEV-Boost relay, which manages how validators receive priority fees and maximal extractable value. EIP-1559 removed the base fee from validator revenue, leaving only the tip as a behavioral lever. This change aligns validator economics with users. Validators, who replaced miners after the September 2022 Merge, earn consensus rewards and tips. In 2026, average priority fees during normal traffic sit around 1 to 2 gwei. During congestion, such as large NFT mints or airdrop claims, priority fees can spike to 50 gwei or more. This fee split allows the priority fee market to act as an auction while the base fee remains deterministic. The reduction in miner incentives to manipulate inclusion order to extract more fees helps protect the network from selfish mining attacks. I find that the coordination between builders and relayers prevents the instability seen in earlier auction models. This structural shift aligns validator economics with users, alongside research from Paradigm on MEV markets. Justin Drake continues to advocate for the endgame consensus roadmap throughout the community.