Why 'Set Your Own Fee' Empowers Bad Actors

Fee market sovereignty is a security vulnerability. Protocols like Ethereum and Solana let users set their own gas fees, which transfers the burden of network security to the user. Most users lack the expertise to price congestion, defaulting to wallet-suggested fees that are often wrong, creating a predictable attack surface.

Predictable fees enable denial-of-service attacks. When fee estimation is transparent, as with public mempools or RPC endpoints from Infura or Alchemy, bots front-run and spam transactions. This predictability allows attackers to cheaply flood the network, as seen in NFT mint wars on Ethereum or Solana's repeated congestion crises.

The solution is protocol-enforced fee abstraction. Systems like EIP-1559's base fee or Sui's storage fund remove fee-setting from users. They use algorithmic fee markets that adjust based on real-time demand, preventing users from accidentally underpaying and eliminating the low-cost attack vector of predictable pricing.

Permissionless fee-setting is broken. It transforms transaction ordering from a public good into a private auction, where the highest bidder dictates the state. This is the fundamental flaw in current Proposer-Builder-Separation (PBS) models on networks like Ethereum.

Validators are rational profit-maximizers. Faced with a choice between a standard fee and a bribe to exclude a transaction, the economically dominant strategy is to accept the bribe. This creates a prisoner's dilemma where cooperative behavior (fair ordering) is unstable.

Real-world censorship is the outcome. Protocols like Tornado Cash have faced de facto blacklisting, not by protocol rules, but because validators and builders like Flashbots respond to OFAC compliance pressure. The 'set your own fee' model provides the mechanism.

The counter-intuitive fix is enforced neutrality. Solutions like MEV-Boost++ and MEV-Smoothing propose protocol-level rules that remove the validator's discretion to choose based on fee origin, mandating a commitment to the canonical chain order.

Optional fees become de facto mandatory. When protocols like Ethereum or Solana make priority fees optional, rational users bid zero to minimize cost. This creates a race to the bottom where only spam or arbitrage bots pay, subsidizing their profits by crowding out regular transactions.

Honest users subsidize MEV bots. In a free-for-all fee market, sophisticated actors like Flashbots searchers optimize for latency and bundle profitability, not base fee payment. Their high-volume, low-fee spam forces regular users to overpay during congestion or have transactions fail.

The system rewards hostility. Protocols with user-specified fees like some Cosmos SDK chains incentivize validators to reorder or censor non-paying transactions. This transforms a neutral mempool into a pay-to-play arena, where the default user experience is transaction failure.

Evidence: On Solana, during the Jupiter airdrop, over 70% of failed transactions were due to insufficient priority fees, while MEV bots successfully spammed the network with sub-cent fees to snipe claims.

Permissionless innovation requires permissionless pricing. A core blockchain tenet is that no central party should dictate user actions. Fixed fee models, like those on Ethereum L2s or Solana, are a form of centralized coordination that gatekeeps access based on a single entity's cost model.

User-set fees optimize for diverse utility. A developer batching 10,000 NFTs values speed differently than a hobbyist sending a test transaction. EIP-1559's fee market already demonstrates that user bids create efficient price discovery; taking the final step to remove the protocol's minimum is consistent.

The 'bad actor' framing misunderstands incentives. Spamming a chain with zero-fee transactions is an attack on block space, not a pricing failure. Networks like Solana handle this via local fee markets and slashing, not by prohibiting low fees. The solution is better spam resistance, not user restrictions.

Evidence: Intent-based architectures like UniswapX and CowSwap thrive by decoupling execution from user-specified constraints. Their success proves that sophisticated users, not protocols, are best positioned to define their own cost-benefit trade-offs for maximal extractable value (MEV) and latency.

Unbounded fee delegation is a security vulnerability. It allows a user to sign a transaction with a gas limit and fee price set by a third party, creating an incentive mismatch where the signer bears no cost for resource consumption.

This empowers MEV bots and spam attackers. A malicious actor can submit millions of low-value transactions, delegating fees to a subsidizer, to bloat the mempool or front-run users without personal cost, directly attacking networks like Arbitrum and Base.

The solution is a bounded primitive. Protocols like EIP-3074 invokers or Solana's versioned transactions must enforce hard caps on sponsored gas, creating a verifiable on-chain commitment that limits liability and makes abuse economically irrational.

Evidence: The 2023 Arbitrum gas spike, where spam transactions congested the network, demonstrated how unbounded models fail. A bounded design, analogous to a UniswapX solver's quote, provides cryptographic proof of cost limits before execution.