Ethereum gas fees used to be the single biggest complaint about using the network. During peak congestion in 2021 and 2022, a simple token swap could cost $50-200 in gas. Layer 2 solutions have fundamentally changed that equation, and the shift has implications that go beyond just cheaper transactions.
Layer 2 networks (Arbitrum, Optimism, Base, zkSync, and others) process transactions off the main Ethereum chain while inheriting its security through periodic settlement. The result is transactions that cost pennies instead of dollars, with confirmation times measured in seconds instead of minutes.
The adoption numbers tell the story clearly. In 2023, L2 transaction volume surpassed Ethereum mainnet for the first time. By 2025, some L2s individually process more transactions than the L1. This migration has dramatically changed the gas fee landscape on mainnet. Average gas prices have dropped to levels not seen since 2020, not because demand for Ethereum decreased but because demand shifted layers.
This creates an interesting economic dynamic for ETH. Lower gas fees mean less ETH burned through EIP-1559 (which burns a portion of each transaction fee). During periods of low mainnet activity, ETH issuance through staking can exceed the burn rate, making ETH slightly inflationary. This challenges the "ultrasound money" narrative that was prominent when gas fees were high.
For traders, L2 adoption has changed where DeFi activity happens. Uniswap on Arbitrum or Base often offers better execution than mainnet for mid-size trades because the lower gas costs make market making more efficient. Lending protocols on L2s can offer slightly better rates because the lower transaction costs reduce the friction of managing positions.
The competition between L2s has created a new dynamic. Each L2 effectively competes for users and liquidity. Some offer incentive programs (token airdrops, fee rebates) to attract activity. This competition benefits users through lower costs but fragments liquidity across networks. The same token might have separate pools on Arbitrum, Optimism, Base, and zkSync, each with different depth.
Bridge risk is the primary concern with multi-L2 usage. Moving assets between L2s requires bridges, which have historically been major attack targets. The Wormhole, Ronin, and Nomad bridge exploits collectively lost billions. Native bridges through the L2's own rollup mechanism are generally safer but slower. Third-party bridges are faster but carry smart contract risk.
Data availability costs, which L2s pay to post transaction data to Ethereum L1, have become a key economic factor. Ethereum's EIP-4844 (proto-danksharding) introduced blob space that dramatically reduced these costs for L2s. This made L2 transactions even cheaper and accelerated adoption, but it also further reduced L1 fee revenue.
The long-term question is whether L2s are additive or extractive relative to Ethereum L1 value. Optimists argue that L2 activity generates demand for ETH as the settlement layer and gas token across L2s. Skeptics note that reduced L1 fee revenue and potential for L2s to develop independent ecosystems could diminish ETH's value capture. This tension is worth monitoring as it affects ETH's investment thesis.
For practical purposes, using L2s is now the default for most on-chain activity unless you specifically need mainnet security or liquidity. Keeping working capital across multiple L2s, understanding bridge risks, and monitoring where liquidity concentrates for the tokens you trade are all part of operating effectively in the current multi-layer landscape.