Thoughts on Bitcoin as a “settlement layer”

It’s well known that block chains have terrible scaling attributes. You basically have two choices if you want to scale to millions of users; 1) you push everyone on to secondary or tertiary layers and have the base layer have a smaller block size and high transaction fees which allows you to have a decentralized and censorship resistant network at the base layer, or 2) you do *everything* on chain and have massive centralization and trust in miners, with very few people running full nodes.

Many people in the Bitcoin community advocate for the first choice, they say Bitcoin should remain as decentralized as possible at the base layer and “coffee transactions” should happen on the other layers. The Bitcoin Cash community goes to the other extreme and talk about things like 1GB block sizes. I’ve always preferred the first choice, but lately I’ve been starting to lose faith in it.

It’s true that keeping that base layer as decentralized as possible is a noble goal. But we don’t want it to turn into something like SWIFT where the ordinary people don’t use it or interact with it, and it’s just mainly used by financial institutions. The same people that say they’re OK with $100+ transaction fees one day are also the ones talking about the importance of making sure the base layer is as decentralized as possible so full nodes can function. But have they thought about who will run those full nodes if transaction fees are so high that it’s not worth transacting on the base layer?

A full node is only economically relevant if you send and receive transactions through it. You won’t do that if the transaction fees are so high that it’s better to use a secondary layer for 99% of transactions. We’ll end up with full nodes only being used by financial institutions, big companies and whales. We might have very strong network decentralization, and I guess on a technical level it’ll be easier to run a full node, but in practise it’ll be extremely economically centralized because the only relevant full nodes will be giant banks and corporations.

When they say they want Bitcoin to be a settlement layer, you should be alarmed because the settling will be mostly between banks and big money. Not you and other ordinary users. You’ll be pushed to other layers where you are economically insignificant and powerless to have any say in the protocol. I think this talk of having layer 1 be a settlement layer is promoting centralization, but in a very different way. It’s taking economic power away from the ordinary user and giving it to banks (who will probably be the only ones transacting on layer 1 in any meaningful sense).

People get scared when full nodes decline in number. “OMG! We need full nodes to keep the network decentralized sufficiently!”. But they say nothing when transaction fees rise, in fact they seem strangely happy about it, as if it means Bitcoin is succeeding or something. But I think we should be seriously concerned with both. Right now, a lot of people from the third world are priced out of using the block chain. Soon people in first world countries will be priced out. Then small to middle sized companies will be priced out. Do we really want just large financial entities using the block chain? Are we sure we want to give them the power of running the majority of economically relevant full nodes?

We never have these discussions because everyone is so focused on attacking the Bitcoin Cash people for how stupid it is to have huge blocks, and how this will lead to centralization, but we don’t ever talk about how “layer 1 as a settlement layer” will also lead to centralization in its own way. High transaction fees are just as dangerous as higher block sizes. Transaction fees should always remain low enough such that ordinary users can use and access the block chain for medium to large purchases.