A recent meeting of key players related to Ethereum (arguably the most used blockchain and current industry standard for smart contract deployment) has re-opened the debate on how blockchains should be governed.
The statement released by the group suggests the formation of a governance body and the development of a model. Of course, if a group of companies meet up and try to govern a project, some element of controversy is inevitable, as can be seen in the Reddit discussion on the subject.
The problem of blockchain governance is how to effectively make decisions on the future directions of projects that are designed to be decentralized. The Bitcoin scaling debate that has been going on for years, has shown how hard it is to find practical solutions for decision making in decentralized systems.
In theory, consensus in blockchain systems is easy: In case of a dispute on the direction a project should take, each node in the system decides for itself which proposal to support, and the majority chain wins. In proof of work, this means voting with computational resources.
In practice, however, blockchains are not as decentralized as they may appear, at least not for the purpose of governance. Even though consensus is distributed and decentralized at the protocol level, there several types of centralization in blockchains:
Mining Centralization
Miners require powerful hardware and low electricity cost to be profitable. This has led to a situation in which most Bitcoin mining is done in China and the bulk of Ethereum mining is concentrated in three major mining pools.
In practice, miners with their hash power decide which chain is supported in the case of a split. In the original Bitcoin model, miners where end-users, so their goals were aligned with what users conceived to be best for the platform. However, with mining as a business model, miners are profit-driven and will support the chain that provides most profit, not the chain that most benefits its users.
Why is this relevant to governance? Clearly, any governance model that might be conceived will only work if miners back the decisions that come out of the model.
Wealth Centralization
In proof of stake blockchains, the problem is similar. Instead of hash power, centralization is based on wealth. Whilst, it may seem easier for a community to decide on the rules and incentivize validators to support their proposed model, validators can still concentrate their voting power, in particular in delegate proof of stake models, where a few wealthy individuals can “campaign” to be voted into power.
Developer Centralization
Blockchain code is usually open source and maintained in public repositories. However, only a few individuals have the authorization to modify the code base and commit changes. These individuals decide which proposals are accepted and integrated into the protocol implementation.
Of course, anyone can clone the repositories and create their own implementation, but there is a branding advantage to owning the original repository.
Branding Issues
Branding is important because the public recognizes brand names. Who gets to own the name Bitcoin matters, because people recognize this name and any cryptocurrency allowed to use this name is likely to have an easier time getting people to use it than other versions under a different name. Technological superiority is not the deciding factor in brand adoption. For this reason, certain key domain names and twitter handles are very powerful tools for their owner.
Choosing a Governance Model
Where does all this leave us in terms of governance? It is clear that finding a correct governance model is very difficult. A fully centralized model is unlikely to work. However, a fully democratic (or anarchistic?) model is also a utopian idea. Many projects are backed by a foundation, but the role in governance may or may not be clear. Other blockchains have moved from a completely collaborative open source model to being controlled almost exclusively by a single company. Founders and development teams usually have most decision power.
A chosen model must balance the views of developers, key industry players, miners and the end-user community.
Once a model has been designed there are at least three ways the model can be implemented:
- Governance processes could be enforced at the protocol level, meaning the protocol forces node operators to upgrade their nodes according to some built-in rules.
- Alternatively, things could continue as off-chain Node operators are “encouraged” to follow governance decisions.
- In between these two models, there are hybrid approaches that implement some of the governance processes in smart contracts. These may involve proposal voting, predictions markets or other signals on which node operators can base their decision.
Conclusion
In almost 10 years of blockchain experience, many unexpected problems and conflicts have arisen. No blockchain governance model has been found to deal with the decisions that constantly have to be taken. With blockchains growing larger, the need for effective governance processes is becoming ever over pressing.
