The block size debate, and the arguments on each side

The argument for raising the limit rested on the claim that Bitcoin was built to be used, and that a network which prices out ordinary payments has failed at its purpose. If blocks were full, the straightforward response was to make them bigger. Bandwidth and storage were becoming cheaper every year, and the growth in demand was, in this view, a sign of success rather than a problem to be rationed.

The second strand was about fees. A permanently full block space means users compete on price for inclusion, and the resulting fee level is set by the most desperate sender rather than by the cost of processing a transaction. For a system intended as a payment network, that is a regressive outcome: the people with the least ability to pay are the ones excluded. Raising the limit would relieve the pressure and keep fees low.

The third strand was about governance. If the limit could be raised only with the agreement of everyone, and agreement was not forthcoming, then the mechanism for changing the rules was effectively broken. Bitcoin XT, Bitcoin Classic and Bitcoin Unlimited each proposed a larger limit and each gathered support among miners and businesses. The argument that the network should follow the majority of its economic participants was central to the position, and it is the part of the debate that was really about who decides.

The opposing argument began with the cost of validation. A larger block takes longer to propagate across the network and more resources to verify, and the nodes doing that work are not miners. They are ordinary participants running software on ordinary hardware, and their independence is what makes the rules enforceable. If blocks grow faster than consumer bandwidth and storage, the set of people able to validate the chain shrinks toward those with data-centre resources, and the network becomes easier to capture.

The second strand was about centralisation pressure on mining. Larger blocks favour miners with better connectivity, because a miner that receives a block late is mining on a stale tip and losing revenue. Over time, that advantage compounds, and the argument was that a change intended to make Bitcoin more useful could make it less decentralised in the process. The cost of a mistake here is not recoverable by a later patch.

The third strand was that the limit was not the only way to increase capacity. Segregated Witness, then in development, increased the effective capacity of a block without changing the base limit, and second-layer designs could move ordinary payments off the base chain entirely. The position was that the base layer should be optimised for settlement and verifiability, with throughput handled above it. The SegWit page covers what that change actually did.

The disagreement was not resolved by argument. SegWit activated in August 2017 through the version-bits mechanism, and the part of the community that had argued for a larger base block limit implemented one as a hard fork, producing a chain split at block 478,558. The sequence of events is set out on the Bitcoin Cash fork page.

What the episode established is less a technical conclusion than a procedural one. A consensus change that a substantial part of the network opposes cannot be imposed, and a group that wants a different rule can have it only by leaving. The conservatism that followed — the reluctance to attempt contentious changes at all — is a direct consequence of how much this one cost.

It is worth being precise about what remains open. The block size question was settled for Bitcoin in the sense that the base limit was not raised, but the underlying trade-off between throughput and validation cost has not gone away. It recurs in every proposal that would increase what the base layer must carry, and the arguments on both sides are the ones set out above.