Coin Metrics co-founder: The design choices of Ethereum are essentially political

Written by: Nic Carter, Partner at Castle Island Ventures, Co-founder of Coin Metrics

Compiled by: Katie Gu, Planet Daily

At the beginning of August, the highly anticipated Ethereum London upgrade went into effect, including a series of Ethereum Improvement Proposals, or EIPs. The most critical of these is EIP-1559, which has disrupted the core fee logic.

Before the change, Ethereum block space was auctioned based on "first price." Users had to predict the potential settlement price of block space. To ensure their transactions were successfully confirmed, they often paid high prices, leading to inefficiency.

Thus, EIP-1559 was introduced. Every transaction in each block will pay the same price, aiming to simplify the fee logic, especially for blockchain developers, increasing predictability and reducing the variance in gas fee prices.

According to the data we have so far, the EIP-1559 proposal has indeed been effective. Fees no longer fluctuate wildly as they did before. However, prices are not cheap, which has frustrated some users. Ethereum users witnessing this process insist that this adjustment was never intended to reduce block space. Insiders attribute the surge in fees mainly to the frenzy over NFTs. In fact, rough analyses provide some support for this theory.

Median fees have increased with the rise in NFT transactions (Coin Metrics)

The recent surge in NFT popularity coincided with the rollout of EIP-1559, so we cannot untangle the causal relationship until the NFT craze subsides. Early analyses suggest that EIP-1559 has effectively set a de facto recent price floor for gas prices, eliminating downtime when block space prices are low (typically at night or on weekends).

Visualizations of gas prices after the implementation of EIP-1559, created by data scientist Takens Theorem, show that this change seems to have eliminated most downward fluctuations in gas prices but did not eliminate later volatility. Fee surges still occur during transaction congestion.

Gas prices before and after EIP-1559 (from Takens Theorem)

This change strips power from users who would set low "limit orders" for gas. In effect, this change homogenizes the way users bid for gas, pushing out those who are price-sensitive and would prefer to wait for lower gas prices.

Then there is the burning of fees. The base fee is now burned, while miners collect some leftover "tips." The reasons for incorporating this mechanism are as follows: first, it positions Ethereum as the sole native currency, similar to how only the dollar can eliminate U.S. tax obligations. Additionally, it eliminates the attack vector where miners might fill blocks with high-fee transactions to raise gas settlement prices and force users to pay extortionate fees. When the base fee is burned, fee recovery attacks become ineffective.

However, the design choice of fee burning also brings corresponding economic consequences. First, it is detrimental to miners, depriving them of most of their previous earnings. Due to this change, currently, only about a quarter of the fees users pay for accessing the blockchain goes to miners. Previously, all fees went into miners' pockets.

The remaining fees are burned. In the 41 days following the proposal's implementation, over 300,000 ETH were consumed, reducing the annual supply by 2.3%. On days of high usage, the fee consumption is sufficient to completely offset new inflation.

If you look at the charts, it is clear that miners' fee income has significantly decreased, while most fees have been burned.

Changes in fee nature before and after EIP-1559 (Coin Metrics)

Thus, this change redistributes miners' economic profits (in which they compete for slim margins) to existing ETH holders. By burning these fees, Ethereum also reduces its security expenditure. While Ethereum appears to be secure, it is not the first instance of fee distribution occurring on a smaller proof-of-work blockchain.

Ethereum is still okay at this level of security expenditure, but it underestimates the contributions of miners. Their job is to create effective linear block space without forks. Reducing payments to the "mercenaries" who defend your domain may make it easier for attackers to exploit, and further declines in miner income will put Ethereum at risk.

A more serious problem caused by the change in fee logic is that it pits token holders against block space consumers

Insiders in Ethereum are well aware of the impact of this policy on miners; in fact, weakening miners' power has been on the agenda. Miners are not part of Vitalik Buterin's long-term plans. In the future of Ethereum, which is slowly moving towards complete proof-of-stake, miners will not play a role at all. Therefore, when it comes to protocol changes, their interests are often not considered; they are merely short-term contractors who will ultimately be discarded.

This poses some issues. If you treat miners as expendable, they will not feel a sense of responsibility for the chain. Currently, miners still occupy a privileged position in the transaction lifecycle because they control the order of transactions in a block. It is well known that reordering (or selectively including or omitting) transactions provides economic returns to entities that control transaction selection.

Some miners have delayed leveraging the so-called "miner extractable value" (MEV) from this principle, which is hostile to end users. Now that they are being pushed out, they have no reason to seek to maximize the long-term value of the protocol. In fact, especially when their fee income has significantly decreased, they are incentivized to seek any possible gains, even at the expense of users. Less direct protocol revenue may mean more income based on "miner extractable value" (MEV) and more exploitation of end users.

A more serious problem caused by the change in fee logic is that it creates a conflict between token holders and block space consumers. Holders now directly benefit from persistently high fees. The higher the fees, the more units of ETH they accumulate. This may increase the unit price of ETH. However, this comes at the expense of actual users of the blockchain, who, all else being equal, would prefer lower fees.

In a world where token holders are also users (and vice versa), this is not an issue. But as Ethereum matures and tries to attract mainstream enterprise users, these users may seek simple transactions rather than holding. This apparent extractive relationship may make it difficult to buy and sell. To draw an analogy, a utility company raises prices, forcing consumers to buy back its stock.

Ethereum has entered a new phase of life. From an inactive newcomer to a mature protocol now extracting rent from consumers of block space. The best evidence is the chart comparing monthly active users to unit price.

Monthly active Ethereum addresses compared to ETHUSD (Coin Metrics)

In fact, throughout history, these two variables have been synchronized. More usage means greater demand for Ethereum itself. But since July of this year, this relationship has decoupled. Monthly activity has steadily declined, yet Ethereum has rebounded strongly. This increase is largely attributed to the hype surrounding EIP-1559 and the subsequent high burning fee rates. But one wonders how long this rebound will last if the practice of burning fees combined with high fees has alienated large block space consumers.

Even if EIP-1559 is not directly responsible for high fees, it is hard to say that Ethereum insiders who made these decisions and hold large amounts of ETH are particularly dissatisfied with the fees, as they now benefit economically from the more deflationary environment brought about by this change. If you are the top consumer of block space, you might feel angry that your fees are not used to pay for security but are directed to token holders.

This could reach a peak in future conflicts, such as block size. I can easily imagine a faction of traders lobbying for an increase in block size, trying to create more block space and drive down gas settlement prices. However, would large token holders, who benefit from the deflationary effects of high fees, want to reduce the scale of the user-funded subsidy? The fact is that due to the mechanism of burning fees, this is clearly quite good for the price of ETH, and individuals controlling the design of the Ethereum protocol now have an incentive to permanently maintain high fees.

Arbitrary changes to monetary policy are both short-term and dangerous

From precedent, arbitrary changes to monetary policy are both short-term and dangerous. It is short-term because it deprives traders of power, and the investment case surrounding Ethereum is built by traders. If you cannot attract companies and users to hold ETH as a medium of exchange, the source of demand for reservations will gradually disappear. Speculators ultimately guess that the protocol will continue to be used in the future. The value of ETH partly depends on the future expectations of token holders, which in turn depend on future adoption. To realize these expectations, they must continue to persuade users that Ethereum is a stable and non-rent-seeking trading venue.

This also sets a disturbing precedent. Of course, prices rebound when expectations change, and a large amount of ETH has been burned, but this decision is equally arbitrary. Sound monetary policy involves establishing rules that cannot be interfered with.

So far, Ethereum cannot commit to a stable monetary rule. If issuance can be reduced after just a few months of discussion among insiders, then this "throttle" may be opened later. Ethereum's decision-makers should consider establishing a rule that can guide policy for decades rather than months. Given that this policy is redistributive, people wonder what the next steps are for insiders in the protocol. Of course, all these decisions carry an unacknowledged political color, and these decisions are merely described as engineering choices.

In this context, "Ethereum killers" are thriving. Alternative platforms like Algorand, Avalanche, Near, and Solana are marketed based on low fees and fast transactions. Of course, to achieve this goal, they need to make decentralization trade-offs (which may be unacceptable), and these blockchains themselves have issues. But if existing smart contract protocols charge exorbitant fees (it is not uncommon for a single Uniswap transfer to cost $500), and may extract rent in the process, then the rationale for choosing alternatives becomes more compelling.

In an environment where stablecoins are highly portable across chains and investors are eager to fund wallets, infrastructure, and DeFi applications on new blockchains, liquidity is no longer a moat. For large consumers of block space, moving their transactions to a new protocol has never been easier. The fee pressure on Ethereum (where large token holders now anomalously benefit from Ethereum) is likely to drive traders to seek greener pastures elsewhere. Why choose to trade on a blockchain where you will inevitably face high fees, thereby enriching insiders through the burning mechanism, when you can find cheap or even subsidized block space on another blockchain?

EIP-1559 may be a good engineering solution in terms of achieving its core goal of reducing the pricing disparity of block space. However, its fundamentally unacknowledged externality of redistributing fees from miners to token holders is problematic. First, it is an arbitrary decision made by a few insiders that affects the core economics of the protocol.

Second, it undermines the monetary credibility of the network. If issuance can be adjusted downward, it can be increased later. EIP-1559 has also alienated miners, who still occupy an important position in the network and are now less willing to be fair stewards of the protocol. Finally, it compares the interests of large token holders with those of current and future block space consumers. In the long run, if alternatives to Ethereum can better align incentive mechanisms and gain traction, this is likely to harm Ethereum.