EIP-1559
Ethereum Improvement Proposal (EIP) 1559 is an attempt to address numerous issues with Ethereum’s UX. The implications of the proposal however stretch much further than a better user experience as it fundamentally affects creation of ETH in the Ethereum Network.
Transaction fees on Ethereum are currently determined via a simple auction mechanism, where users on the network sending transactions will specify a gas limit and a gas price for their transcctions in an attempt to get it chosen by miners who will always prefer the highest bids. For information on gas fees on Etherem check out the other article on this blog “Smart Contracts”.
Current Issues with Fees
The current system provides several large sources of inefficiency:
Delays for Users
Due to the gas limit per block, volatility of transaction volume and preference for miners to prioritise higher gas fee transactions users will see their transactions pending for a long time in the mempool before being added (assuming they aren’t overpaying gas for priority). Users can see their transactions delayed by several blocks before inclusion (each new block is filled with the highest paying transactions since the last block) rendering it difficult to estimate when it will be included. This is inefficient for the ecosystem of Ethereum.
Inefficient Auction
On top of delays, due to the the auctioning system of transactions which is very complex and inefficient users find themselves often either overpaying the fees (to avoid cancellation) or losing ETH from having their transactions cancelled for underpaying.
The EIP 1559 Changes
Overall EIP-1559 removes the first-price auction nature of transactions and replaces it with a fixed-price sale. The protocol will quote a price, and the user chooses to either pay that price or not to transact on the network. The EIP proposes the following changes:
Block Size
Under EIP-1559, blocks can be up to “200% full” meaning that if the gas limit is set to 15m in gas, blocks up to 30m gas will be valid. Blocks will also include the minimum price required for transactions to be included in them which is the base fee.
Base Fee
The base fee amount (mandatory to be paid in ETH) is adjusted accordingly by the protocol based on network congestion. The base fee increases when network congestion is high (measured by per block gas usage, 50% of the maximum so “100%” of the gas limit of the block) and vice versa. This is similar to how gas limits for blocks are adjusted except instead of consensus across miners the base fee is based off a controller. The maximum difference in base fee from block to block is predictable allowing wallets to auto-adjust the gas fees for users in a reliable fashion grealty removing situations of overpaying and underpaying.
Priotity Fee (Tip)
The transaction fee will encompass the base fee and a small (automatically set) priority fee (which is also optional). The priority fee is so that transactions that are not included in the block due to being outside the “100%” capacity can be added to the maximum (“200%”). This compensates miners for taking on uncle block risk. The priority fee adjusts the users position in the queue like the current system with miners prefering to take transactions with higher gas. Miners should still prefer higher tip transactions over those with a lower tip, purely from a selfish mining perspective.
Miners Transction Rewards
Miners only get to keep the priority fee. The base fee is always burned. This burn reduces inflation of ETH and can potentially render Ethereum deflationary. Removing the base fee from the miner revenue stream is important to reduce miner incentive to manipulate the network in order to extract more fees from users.
When blocks are “200% full” the fee market will revert to an auction based on transactions with the highest tips. Thus the current system of auction becomes a worst case scenario in post EIP-1559 Ethereum.
Economic Effects
Cements ETH in transactions
Under the current fee market, there is no guarantee that transaction fees are paid to miners in ETH or paid via a transaction over the Ethereum Network. A user can simple communicate with a miner to include their transaction on a block (a cost of 0 gas price) and pay the miner off-chain or with another asset on chain (a stable coin). EIP-1559 solidifies Ethereum as the only source of currency for transactions on the Ethereum network and prevents the possibility of ETH not being the primary currency on Ethereum.
Deflationary
By burning the base fee, there is no guarantee of a fixed token supply. The more people want to use Ethereum, the higher the base fee (more demand for block space) the more ETH gets burnt. If moreETH is burned on base fees than on generating mining rewards then ETH will be deflationary and vice versa. The deflationary element also contributes to ETH being a productive asset (and a more attractive stored of wealth) as burning the basefee internalises some of the revenue of miners.
Better User Experience
The base fee provides a predictable fee system for all users, reduces unexpected wait times for transaction confirmations and automates the fee bidding system. These improvements could lead to higher adoption of Ethereum.
Impact on Gas Fees
Fees are primarily the result of supply and demand. EIP-1559 reduces overpaying transaction fees, but it doesn’t lower fees. The total capacity in a block remains the same, and the supply of transactions doesn’t change.In fact, it might perform the contrary: due to the better user experience, users are willing to pay more for transactions than currently. Only scaling up Ethereum (Optimism and Layer 2) can gas fees be reduced in the long run.
Impact on MEV
Since the position of a transaction on a block can still be determined via the tip thus MEV will continue to occur post EIP-1559. However due to the burning of transaction fees it can be argued that EIP-1559 may incentivise miners who aren’t extracing MEV to actively seek out user-harming MEV strategies previously viewed as unethical.
Security Effects
Complexity around increased Block Size
The increased maximum block size may cause problems for miners who are unable to process a block fast enough. Miners may need to increase technical mining resources to avoid (an out of memory or similar issue).
Reducing Spam transactions
Currently an attacker attempting an attack on the Ethereum Network can collaborate with a miner and launch attacks with a 0 gas price. With EIP-1559, if an attacker wants to spam the network using significant blockspace, they would be required to pay a gas price on every transaction (unable to bypass the base fee). The attack would also cause the base fee to increase implying that the attack gets costlier over time.
Security Budget
Ethereum has a social contract called Minimum Viable Issuance which states that Ethereum must maintain a security budget that scales with the network. This roughly translates to if ETH goes up in value so should mining revenue. However in the case of Ethereum, due to the inflationary nature of the asset which contributes to the security budget, holders of ETH are the ones taxed. Users of the network believe Ethereum is overpaying for security (once MEV gain are accounted for) thus EIP-1559 will aim to reduce this security budget both in placing deflationary pressure and potentially reducing overall hashpower.
Miners Opinions
The proposal has garnered some of the largest support to date from Ethereum application creators and users alike, given the current difficulty of selecting a correct transaction fee. Miners and mining pools, on the other hand, have been gathering in opposition against the proposal as it progressed toward mainnet.
The revenue that miners receive for their work can be broken down into three sources:
- The reward of 2 ETH that they are paid for mining a block (as well as smaller reward for uncle block processes).
- The fees received from users who bid for a spot on the block irrespective of their final position in the block. (For a user this is the gas fee for your transactions on ETH)
- The value (dependent on the nature of the transactions executed) miners extract by inserting (or not inserting) transactions at specific points in the block.
Whilst 1 and 3 will still still remain post EIP-1559, a signficant proportion of their revenue is lost with the reduction in fees received for transactions. What are the alternatives for miners who do not wish to cooperate with the post EIP-1559 Ethereum environment?
Miners attempt to Fork Ethereum
Forking the Ethereum state is complex, unlike BTC Ethereum has a far richer ecosystem of decentralised applications, smart contracts and tokens that rely on it. The state can be copied however replicating the claim on a real world asset is much more difficult i.e. WBTC are claims of BTC. Likewise AMMs that rely on collateral would be dysfunctional. This duplication would render then worthless on the fork chain. Furthermore, off-chain infrastructure like oracles, liquidation bots would all need to be reworked.
Putting the missing infrastruture aside, if the fork does not aim to preserve the state of Ethereum but only the token distribution the creation of this “fresh state” would result in new ETH in the hands of potentially hostile users (those who prefer the EIP-1559 chain and want to remove competitors to ETH) who could use it to depress the price.
Miners join the new chain but block EIP-1559
In this scenario, miners would join users on the new Ethereum blockchain, but then suppress the EIP-1559 mechanism from burning any ETH by manipulating the basefee to zero. Miners can attempt this by controlling how many transactions they include per block, and hence can controlling the blocksize which determins the gas used per block, thereby controlling the basefee. If miners only mined blocks that were less than half full the basefee would never increase above zero. However, basic game theory dictates that the competition between different miners makes this strategy impossible in practice since self-interested miners want to include as many transactions as possible (to take advantage of MEV) increasing the base fee and forcing other miners to do the same an maximise the block sizes.
The likelihood of avoiding EIP-1559 for miners is very low and miners will inevitably choose to continue mining on Ethereum. It can also be argued that with the deflationary element of EIP-1559, miner revenue is not necessary lower than pre EIP-1559 since miner revenue depends heavily on the price of ETH. For example if the price of ETH to USD increases proportionately to the mining revenue lost from the burn miner revenue is exactly the same in USD terms.
Current Status
The proposal is rolling out via the London Hard Fork and the first test has been produced on the Ropsten test network (a full proof of work environment and the most realistic of the test nets).
