Blockchains Are Inherently Unfair, But That’s OK
Blockchains are unique because they charge fixed transaction fees, regardless of the total value of a transaction. Most TradFi systems charge percentage fees, for two reasons: They are single asset (dollars, land, etc) so it’s easy to determine objective value, and they are centralized so a rent-maximizing intermediary charges high value transactions more.
In crypto, not having an opinion on the total value of a transaction is crucial for censorship-resistance. We don’t want miners/validators to have an opinion on the worth of a transaction — doing so can lead to security issues. We just want them to confirm the validity of every transaction, as determined by the protocol.
But now we have a fairness problem, because blockchains don’t scale as well as centralized systems and have no choice but to auction off the limited blockspace that’s available. High-value transfers (as determined by their users) are naturally willing to pay more for security and win the auction.
To summarize: Limited capacity + value agnostic settlement = high-value transfers crowding out low-value ones during periods of congestion.
We’ve seen this phenomenon play out time and again, first in Bitcoin, then Ethereum, now other chains. I’m sympathetic to why people find this inherently unfair. In a perfect world, a teenager buying $10 worth of SOL and a company doing a billion-dollar stablecoin transfer get equal access and pay the same fee.
But we live in a world of limited blockspace, and the most efficient solution is to give this valuable and scarce resource to whoever needs it most, as determined by their willingness to pay. That’s capitalism, and why land in manhattan costs more than it does in Kansas, or why oil shocks can really hurt low-income drivers.
Historically, all attempts to make a market more fair, aka socialism, have ended in misery.
The modular approach to scaling starts with the practical assumption that while everyone might want to do stuff on chain, not everyone needs the same amount of settlement guarantees. So we build a vertical stack where users decide how much security they need and communicate their desire via the fees they are willing to pay.
The monolithic approach is more ideological. It wants to give everyone equal access and keep them on the main chain, regardless of how much security they need. So it tries to optimize messaging, consensus, hardware, parallelization etc etc to accomodate all commers. It even introduces local fee markets as a form of rationing. “Just because wales are doing wale things,” it says, “doesn’t mean individual users should have to go elsewhere — we’ll just charge only the wales more.”
There are two problems with this approach. The first is it’s inherently inefficient. Like any kind of socialism, not auctioning off all of a limited resources to the highest bidder leads to waste.
The second is that it chronically underestimates demand, in part because it keeps inducing more. Improving messaging or upgrading the software to accomodate today’s demand leads to more tomorrow. Put differently, there’s no shortage of stupid memecoins. And when more show up, there is no other layer to send them to.
So a monolithic chain will always be under the duress of more blockspace demand than it can handle. It can attract a lot of brilliant devs who work tirelessly to scale it (as Solana has) but it will never be enough. They’ll be like Boxer in Animal Farm, always committing to work harder.
Meanwhile, the tradeoffs made to achieve more throughput hurt decentralization and liveness. High value transfers defect, dooming Solana to be a low value-chain that is always operating on the edge of something breaking.
The modular approach to scaling is more practical and sustainable long term. It’s unfair, but that’s OK, because we are all still better off. As Churchill said, The inherent vice of capitalism is the unequal sharing of blessings
