Ethereum Transaction Fee Market Reform: Will EIP 1559 Raise Fairness Questions?

Editor's Note: This article comes fromEthereum enthusiasts (ID: ethfans)Editor's Note: This article comes from

Ethereum enthusiasts (ID: ethfans)

, Author: Pintail, translation & proofreading: Min Min & A Jian, reproduced by Odaily with authorization.

The first four sections of this paper describe the progress of EIP 1559 to date. Links to some important articles are included in the article, which can be used as a reference. If you are already familiar with the background of EIP 1559, you can skip directly to Section 5 and start reading. Sections 5-7 of this paper discuss the fairness of this proposal. If you are unfamiliar with the concepts of "Gas", "Gas Cap" and "Gas Price", you can find related articles on EthHub.

secondary title

1. Background

These ideas have certainly sparked interest among members of the Ethereum community, but without a concrete implementation proposal, these ideas are limited to academic discussion. At the time, reforming the transaction fee market was not a pressing issue because Ethereum transaction fees were not high. Indeed, throughout the early stages of development of Ethereum, the Gas price was far below the estimated 'rational' minimum estimated by miners. Although during the ICO craze in 2017, short-term demand peaks caused network congestion, but in general, users only need to pay extremely low transaction fees to easily complete transactions.

However, in early 2018, this changed. In the past, there was always unused Gas in the blocks, and now the Gas in each block is used up and it is not enough to cope. Due to the high demand for Gas in applications such as CryptoKitties, and the rise of FCoin, a token exchange under the banner of "transaction is mining", the Ethereum network has been congested for a long time. The price of Gas required to pack into blocks within the range has skyrocketed.

For new users, understanding Ethereum's gas price mechanism is always a challenge, and it is even more difficult when there is network congestion and wild gas price fluctuations. Smart users can use new tools such as ethGasstation.info to judge reasonable Gas prices, and more and more wallets like Metamask have built-in Gas price estimation tools. Still, according to blockchain data, some users are overpaying for gas, and recurring congestion reflects that existing solutions do not fully address the problem.

secondary title

• 2. EIP 1559 Proposal

• It is against this background that Vitalik published an article in July 2018 detailing a proposal for transaction fee market reform (later, the content of this proposal was written into an article on resource pricing). This proposal introduces a new transaction fee mechanism, which divides the transaction fee paid by users into two parts: one part is the "base fee" determined by the protocol and handed over to the protocol, and the other part is paid to miners Commissions (set by the user, now often referred to as "tips" or "bribes"). The goal of this mechanism is to keep the block usage rate half full (and a full block is twice the size of the current Ethereum block), and adjust the base fee rate according to demand (so that Actual usage tends to be half full). Since the subsidy required by miners to package transactions hardly changes with demand, users only need to pay a lower tip, so most users do not need to change. In this way, the transaction fee will become much simpler, and users no longer need to understand the complex and abstract Gas mechanism. This is a big step forward.

• So the question is, how should the protocol handle the base fee? On the face of it, there seem to be four possibilities:

• Send the base fee to a fund (e.g. developer fund);

Reward the base fee to the miners who dug out the current block;

The base fee is shared by the miners who mined the block afterwards;

Option 1 can be ruled out in the first place, because centralized funds may bring about a governance nightmare. Option 2 is also not feasible, as it would incentivize miners to fill up blocks with spam transactions, which would drive up the base fee regardless of user demand for block space. Only options 3 and 4 can be selected.

The final choice is option 4, one of the reasons is that this will bring another benefit, that is, ETH will become the fuel for transactions on Ethereum. This helps avoid the threat of economic abstraction, which some fear would cause ethereum to undergo a death spiral, greatly undermining the security and viability of the entire network. Users use other tokens instead of ETH to pay fees). There are other reasons why Option 4 was chosen and are discussed in Sections 5 to 7 below.

This time, with a concrete proposal in hand, the Ethereum community generated strong interest and based on that, EIP 1559 was proposed. Unsurprisingly, ETH holders are very optimistic about the economic impact of burning the base fee. This is equivalent to a share buyback, and all ETH holders can benefit from it based on their holdings. This enthusiasm is reflected in the establishment of the EIP 1559 Community Crowdfunding Fund. The fund has received the largest number of grants in the sixth round of the Gitcoin Grants grant program to date.

secondary title

Nevertheless, EIP 1559, once implemented, will be the most significant reform of the economic mechanism since the birth of Ethereum. In addition, it will create a lot of work for wallet developers, who will need to adjust the user interface accordingly. Therefore, experienced Ethereum developers are very cautious about EIP 1559. Among them, the views of Dan Finlay of Metamask are often cited. He pointed out that since EIP 1559 leaves the "base fee" parameter to miners to decide, in order to avoid a reduction in revenue, miners will try to reduce the base fee as much as possible. If the miners succeed in doing this, EIP 1559 will bring the transaction fee market back to the highest price auction mode, and users still need to bid to get their transactions packaged on the chain.

In opposition to EIP 1559, Dan proposed the "Escalator" proposal. This proposal is also to solve the user experience problem, which can achieve more efficient price discovery, and at the same time does not need to hand over the control of the basic fee parameters to the miners. The solution proposed by this proposal is also elegant, but has not received as much support as EIP 1559 so far. Part of the reason may be that EIP 1559 also addresses other issues such as economic abstraction. Of course, the Escalator proposal is not very popular in the Ethereum community, because members of the Ethereum community tend to hold ETH for a long time, so they prefer the proposal of destroying the base fee. By the way, another solution is to combine EIP 1559 and Escalator proposal to take the best of the two, but it has not yet been refined into EIP.

Other senior ethereum developers, such as Nick Johnson of the Ethereum Name Service (ENS) team, also expressed concerns about the momentum behind EIP 1559, as it fundamentally changes the game-theoretic basis of the ethereum transaction fee market. The impact and possible risks are unknown. Clearly, it would be foolish to overhaul Ethereum's core economics without fully understanding the implications.

secondary title

4. Research advancement

In 2020, due to the popularity of stablecoins, decentralized exchanges, and other DeFi applications, Ethereum's demand for Gas will continue to increase. Therefore, from June to August 2020, Ethereum transaction fees were higher than Bitcoin transaction fees almost every day, except for 8 days. Counter-intuitively, such high transaction fees could spell trouble for the Ethereum blockchain. A group of researchers argue that as the focus of miners' income shifts from block subsidies to transaction fees, Ethereum will become unstable and its security will be greatly threatened because attackers can launch some miners without 51% of the computing power. attacks that can be achieved. At the same time, the violent volatility of the Gas market and the corresponding user experience issues remain unresolved.

At the same time, the client implementation work related to EIP 1559 is also advancing. As of the end of August 2020, two Ethereum clients have implemented EIP 1559, and an early testnet is running to address specification and implementation issues. In addition to the core Ethereum client, the Filecoin and Celo networks are also testing implementations.

Significant research resources have been devoted to EIP 1559, and no viable attack vectors have been identified so far. One conceivable attack is hashrate collusion. However, as Micah Zoltu mentioned in the article, even if an interest group with 51% of the computing power forms and tries to manipulate the base fee, the individual private interests of the miners will keep them out of the cartel. This is different from a 51% block censorship attack on a PoW blockchain, which can double the payoff by ignoring blocks mined by non-members of the syndicate and only putting blocks mined by syndicate members on-chain.

While research on EIP 1559 continues to advance, we may discover serious flaws in its economic mechanism. If no serious flaws are found, we can perhaps be optimistic that EIP 1559 is enforceable.

secondary title

Therefore, we can tentatively conclude that we can safely implement EIP 1559 on Ethereum. But should we do this? This is the question of fairness mentioned in the title of this article. Just because miners take on the job of processing transactions, are they entitled to all transaction fees as rewards?

Ethereum's 2013 white paper states that this is not the case. Unlike Bitcoin miners, Ethereum miners take on the responsibility of safeguarding network security and receive block rewards permanently as subsidies. Currently, the block reward is fixed at 2 ETH per block, in addition to the benefits from uncle blocks (blocks that are not included in the legal chain, but the miners of uncle blocks can get part of the block reward). Therefore, there is no requirement that transaction fees should be sufficient to cover the cost of securing the network. This is an advantage because transaction fees vary widely and there is no guarantee that it will meet the security needs of the network.

In addition, not all transaction processing is done by miners, because the cost of state bloat caused by transactions is borne by the entire network. While all full nodes bear this cost, only miners are rewarded. Users also pay an opportunity cost because they are unable to execute otherwise profitable transactions when gas prices are too high. Therefore, Ethereum's block space does not belong to miners in any sense, but should be regarded as "commons". Miners charging high transaction fees is like collecting rent on "public land", because the cost for miners to package transactions is much lower than the benefits they receive.

secondary title

6. Fair Compensation

It is difficult to get accurate data on the cost of miners packaging transactions, and this data will change with the improvement of client implementation and the change of the amount of gas consumed to perform specific operations on the Ethereum virtual machine. As a result, the related research by Vitalik mentioned above is outdated, and it is difficult to obtain data to arrive at an up-to-date estimate. Nonetheless, improved client performance means that miners may be cheaper than before, yet the transaction fees they receive have increased substantially.

Fortunately, the new transaction fee market proposed by EIP 1559 can solve this problem for us. Rational miners will only pack transactions that give enough tips/bribes. Users who are not in a hurry to get their transactions included can pay lower tips, and arbitrage traders who are in a hurry may pay higher tips. Generally speaking, tips do not fluctuate as wildly as gas prices. Most users can accept the default value of the wallet, and no longer need to worry about the problem of gas price surge.

Anything beyond fair compensation can be considered a "fee surplus" or a "transaction dividend". This is the base fee for EIP 1559. If it does not belong to the miners, then who should the base fee go to?

secondary title

7. How to distribute transaction fee surplus

Among the four schemes mentioned in the second section, only the "destruction base fee" is left to choose. After the base fee is burned, all ETH holders will benefit. Is this what they deserve? I think so, and for two reasons.