The Merge is almost here. Transitioning Ethereum’s consensus mechanism from PoW to PoS is the first step in the blockchain’s 5-stage upgrade plan. In a bid to solve the blockchain trilemma of security, decentralization and scalability, Vitalik and Co. have identified a rollup-centric map as the way forward.

Rollups like Optimism, Arbitrum, zkSync and L2 solutions have steadily gained popularity over the last few months. This has led to many within the Ethereum community speculating that the growth of Layer 2s will be parasitic to the L1 mainchain. They claim that by offering significantly faster and cheaper transactions, L2s will attract so many users that the L1 will lose value. But will that really be the case? Let’s go through the more common arguments and see if we can clear things up.

1. L2s will draw activity away from L1

With L2s being cheaper to use, everyone will move away from L1, reducing the price of ETH

This seems to the be the most common anti-rollup argument, but people forget: this is the goal! Vitalik Buterin has clearly stated that after the Surge (the 2nd stage of the upgrade), Ethereum should ideally serve only as the consensus and data availability layer, with all executions taking place on Layer 2 scaling solutions, specifically rollups. But what about all the gas? Yes, Ethereum will no longer earn gas fees from users if they transact on L2s, but remember that the L2s themselves have to settle on the L1 mainchain, and that costs gas too.

Currently, Ethereum has high gas fees and a low throughput of around 15 TPS. Rollups, on the other hand, have low gas fees, but they’re offset by a high throughput (around 1500–4000 TPS with compression). Granted that the revenue is going to the L2 rather than the L1, but the rollups have to post data to Ethereum and the gas paid for that is in ETH. And as individual users transacting on the mainnet reduce in number, more L1 block space opens up for the rollups, which can then post more batches to Ethereum, thus driving its revenue up.

Another thing to keep in mind is that the L1 won’t become a ghost town with only L2s pushing data to it. Despite what Ethereum wants, people will only move to L2 if it’s for a specific application or if the L1 is too congested and hence too expensive to use. Even then, they might have reason to stay. For example, optimistic rollups have withdrawal periods of 1–2 weeks, so users who need to transact quickly will prefer to pay high gas fees on L1 rather than use an L2. A final point: as users move to L2s, the congestion on the mainnet will reduce, taking the gas price down with it. And that will bring users back to L1.

2. L2s become so big that they branch of as independent L1s

Once L2s gain enough users, they might spin off as independent blockchains and steal large chunks of Ethereum’s user base

This seems to be a prominent concern in the community, but it overlooks just how difficult a task it is to make a blockchain truly decentralized and secure. Ethereum’s Beacon chain has over 350,000 validators who will soon maintain the network’s crucial aspects of security and decentralization. Since rollups exist as smart contracts deployed on the mainchain, they inherit Ethereum’s security and pay for it with gas when they post batches to the L1.

If a rollup was to branch off as an independent L1, it would have to set up a large validator set and figure out a token issuance mechanism to reward them. Currently, these rollups pay Ethereum for the security and the amounts are not large. Arbitrum, for example, pays around $15,658 in a day while Loopring pays just $584. Furthermore, the introduction of proto-danksharding and Danksharding will further reduce the security fees paid by L2s to the L1. So Arbitrum could 1000x its throughput and still pay the same fees that it does today despite earning way more revenue.

These costs are trivial when compared to the cost of creating a sovereign blockchain — one which will lose out on the vast ecosystem of dApps, DeFi protocols, Metaverse platforms, and other services offered by Ethereum. It’s clear that the upside of moving away (if there is any) is insignificant when compared to the benefits of staying. But even if one of them did leave, Ethereum will be virtually unscathed because L2s are far more dependent on the L1 than the other way around.

3. L2s have their own tokens, which will decrease ETH demand

As more users transact on L2s and use their native tokens, the demand for ETH will drop and reduce its value

Layer 2s need native tokens to decentralize their sequencers (in the case of Optimistic Rollups) and to incentivize growth and liquidity. To this end, Optimism launched its OP token this May and zkSync has announced that it will eventually launch a native token as well. The argument that L2 native tokens are parasitic to ETH would be considerably stronger if the native tokens were instantiated as the sole acceptable currency on L2s, but that isn’t the case. ETH still serves as the de-factor money on rollups and the security fees to settle on L1 are all paid in ETH. Also, 70–80% of these security fees are burned, which directly drives more value to ETH.

As mentioned earlier, the introduction of proto-danksharding and danksharding will enable L2s to earn higher revenue without paying extra security fees to Ethereum. This definitely lowers the revenue earned by the L1 mainchain, but that brings us to the question: are MEV (maximal extractable value)and fee generation the only factors determining ETH’s value? I don’t think so.

ETH price is based on supply and demand, not just revenue earned. And while the max supply isn’t capped, new ETH tokens are issued at an ever-decreasing rate. The demand, meanwhile, will only go up. People buy ETH not just to pay for gas, but because it serves as the base currency for the entire ecosystem:

• NFTs are priced in ETH
• The largest liquidity pools on Uniswap are paired with ETH
• Users bridging to Ethereum from other L1s like Cosmos or Avalanche will swap tokens for ETH
• The Metaverse is being built on Ethereum and will most certainly use ETH along with any native currency the platforms have
• ETH is a reliable collateral and will continue to be staked on lending/borrowing protocols

And once L2s are inter-connected through bridges, ETH will be the go-to token of choice when it comes to transacting in a cross-rollup Ethereum ecosystem. L2s also enable a higher velocity of money which leads to the expansion of the Ethereum economy and thus further drives up the value of ETH.

4. dApps will be created exclusively for L2s

Some L2s will become so big that dApps will be created specifically for them rather than the L1

This will undoubtedly happen, but why is it an issue? The Ethereum ecosystem already has applications that operate exclusively on Layer 2s. The prominent decentralized exchange dYdX (which intends to become an independent L1 in Cosmos) earlier switched from the Ethereum L1 to StarkWare, which is a ZK-rollup L2. This allowed them to lower trading fees, decrease the minimum trade size, and add more trading pairs — all of which increased its usage. While Ethereum lost the direct gas fees from dYdX operating on the L1, the cost was offset to a degree by the increased number of rollup settlements on the mainchain while the network’s congestion, and hence fees, was lowered.

An application-specific blockchain will far outperform a general-purpose counterpart for its particular use case. Ethereum can’t afford to be application-specific, which is exactly why they’re pursuing a rollup-centric roadmap.

The creation of dApps specifically for L2s is not a problem, it’s the success metric!

The case against having application-specific L2s is weakened when we consider the Ethereum ecosystem as a whole greater than the sum of its parts. When L2s are bridged directly to each other instead of the L1, user experience will be drastically improved by having streamlined application-specific chains which handle transactions more effectively than a slow and bloated general-purpose chain. And such a multi-chain, L2-centric ecosystem would have ETH serving as the base currency for all transactions, playing the same role as the US dollar in today’s global economic system.

So, will Ethereum L2s be parasitic to L1? No.

Saying that is like saying California is parasitic to the USA because it accounts for ~15% of the national GDP (thanks Bankless for the analogy!). The relationship between a state and its country is similar to the relationship between a Layer 1 blockchain and its Layer 2 solutions: It is symbiotic, not parasitic. There are mutual benefits gained by both parties in the relationship, and a healthy give-and-take ensures that neither prospers at the other’s expense.