The main values of public blockchains are: their ability to resist censorship and their ability to operate in trustless environments. In fact, both are forms of Byzantine fault tolerance, a keystone of distributed systems. Unfortunately, the generic hype in the field of “blockchain” technologies has led to a strange, but popular view of private or permissioned blockchains as state of the art in the field. However, such blockchains exclude the concept of censorship resistance; a very core concept which has led, for example, to Bitcoin’s big success. In this brief article we try to explore the issue of censorship-resistance in public blockchains from several angles and demonstrate why it is so important for designing cryptoeconomic systems.
Firstly, censorship-resistant transactions need to be unalterable, with no ways to rewrite blockchain’s history. This is similar to the idea that “the law does not work backwards”. Effectively, it’s a way to secure the monetary balances (along with other information that enters the blockchain) and prevent any involuntary redistribution of funds.
Secondly, censorship-resistance implies that everyone can transact with the network on the same terms, regardless of their personal identity or qualities that are not described in the network protocol. This is much like considering the account as a completely separate “net entity” than the physical entity of its owner. The protocol disseminates rewards and cuts collateral stakes automatically, following known rules and according to account’s actions in the network. If we want to achieve real censorship-resistance, then users, even by supermajority, should not be able to exclude another user from getting equal terms of network service. Probably, the best results would be achieved by providing zero-knowledge anonymity to any transactions. In any case, it is essential to ensure that the network is governed solely by predetermined economic incentives, while issues like moral concerns, political affiliations and social prejudices etc. remain strictly personal matter and can not be imposed on others through the blockchain network. This inability of any individual or their group (however large) to impose their opinions on others is essential part of censorship-resistance. It is understandable that, given a choice, some people will prefer to sacrifice censorship-resistance to achieve e.g. ethical goals or regulatory compliance; this is why a censorship-resistant blockchain is not giving to users that choice, except perhaps only for the person’s own self (if “opting out” of default perks does not compromise others that he transacts with).
Thirdly, there is the question of how censorship-resistance could be implemented on a network that does not stop to evolve and change. Does the governance model allow to change the rules, to what extent, with what incentives, what are the predictable future tendencies for the network? Looking more broadly, this includes not only possible changes to blockchain’s unalterability or non-discrimination between users, but also other important rule changes. For example, involuntary redistribution of funds could take many insidious forms, not being limited to outright confiscation, but also involving taxation through inflation, or perhaps major changes to usefulness of funds within the network (e.g. by raising the stakes required for mining node creation, or by adding burdening bureaucracy).
It is much more convincing for trustless environments when censorship-resistance is embedded in the network design and in the economic incentives, rather than when it is based solely on governance model that could potentially be altered in the future. Therefore, the limits of governance, and whether the governance model of blockchain can be altered, need to be known beforehand. Finally, it is worth mentioning that the possibility of a fork is often useful as the last-resort protection against censorship.