Blockchain governance serves 3 important purposes within an ecosystem:
- To fix bugs and vulnerabilities
- To upgrade the underlying technology
- To repair damages from attacks
Governance models usually consist of 2 crucial components:
- The rules of the protocol (epitomised in code)
- The incentives of the network (economics)
There are 3 general types of network actors within an ecosystem:
- Miners — keep the network going
- Users — derive benefits from the network
- Developers — update the network
In short, a governance model of a protocol aims to determine who amends the code. According to the Western democratic governance tradition, it is the users who should have the most voice in deciding about the future course of blockchain ecosystems. Thus, most models attempt to give as much power as possible to actors who derive most benefits from the network.
So far, we have identified 2 types of blockchain governance:
- Off-chain Governance — Wider consensus has to be reached in order to implement changes in the protocol, e.g. Ethereum, if consensus is not reached, hard forks occur.
- On-chain Governance — Voting power is determined by the amount of tokens an actor holds.
The governance model used by Bitcoin is focused on reaching consensus among network participants, especially those operating a full node. Developers can coordinate updates via BIP (the Bitcoin Improvements Proposals), which happens off-chain. Miners create the chain, and thus in a way participate in governance on-chain. Developers submit improvements and miners can decide whether to implement them. All of them tend to hold some stake in the Bitcoin ecosystem, and are incentivised to maintain it to the best of their abilities. Some of the risks of such an arrangement include a certain amount of incentives for developers to cheat. Contrary to miners, many developers do not hold that big a stake in the Bitcoin ecosystem and are thus less likely to lose out in case of an attack or just plain negligence. Also, they often leave and create their own protocols, which leads to fragmentation of the ecosystem and creation of numerous Bitcoin alternatives.
In its governance model, Ethereum draws many aspects from Bitcoin. Developers are able to provide their updates via the EIP — the Ethereum Improvement Proposal. It is a basic application process which allows developers to express ideas on improvement of the code. Developers can take this and incorporate the proposals, the community (full nodes) can subsequently install and start using it. However, in order to get their proposals implemented, developers have to make sure there is consensus around them. This includes persuading miners and community members that the update is to their benefit. The media is also a very powerful actor in the ecosystem. Influential journalist webs such as Coindesk can indirectly participate in governance systems by shaping the sentiment of the community. The Ethereum Foundation is another force that influences Ethereum governance. It serves as an important mitigator in this process by coordinating protocol upgrades, by providing their vision and by governing software repositories. However, off-chain governance systems are prone to situations in which some people are better connected and have more power than others. Some critics of the Ethereum governance model claim that Ethereum Foundation has too much power, as it is the organiser of regular calls for developers and can decide who joins. One of the advantages of the Ethereum governance model over Bitcoin is that it’s less rigid — exercising governance is more about socially coordinating the change and convince people to come to your side. Some of the main risks that have been identified is over reliance on Vitalik Buterin, the Ethereum founder, in determining the future direction for the ecosystem. However, Vitalik claims to be working hard on solutions which will allow him to take a less active role.
On-chain governance is a type of governance model in which the governance process is laid out in the protocol and updates to the protocol are implemented via smart contracts. Even if full node operators do nothing, they will upgrade automatically. Compared to on-chain governance models, this has a significant disenfranchising effect on full nodes. Also, it creates the danger of sliding into “coinvoting” — voting process in which 1 coin = 1 vote that results in whales having excessive amounts of power. Many experts argue that on-chain governance enforces rigidity and it will not be dynamic enough to follow the what individual communities want. One of the main criticisms is that it is still very difficult to write a smart contract that is going to coordinate between all these stakeholders, as well as that on-chain governance models do not fully recognise the importance of social consensus.
Some of the most popular projects that attempt to implement an on-chain governance model include the following:
Dash is a blockchain based on Bitcoin focusing on instant and secure processing of payments. Dash Masternodes function as full nodes in Bitcoin. They process transactions in exchange for 45% of all block rewards.
When looking at the Dash governance model, it can be said that it is a type of dDAO, a delegated DAO. This means that actors delegate voting rights on community projects to Masternodes. Masternodes are actors which are able to prove their interest in the system by staking some collateral. Masternodes in the Dash ecosystem get voting rights in exchange for locking down 1,000 dash and hosting the service node. 1 Masternode = 1 vote.
Dash operates in a series of voting cycles, which come once for a certain no. of blocks. That roughly corresponds to once every 30 days. Any community member wishing to submit product update can pay a proposal fee and a proposal is hashed into blockchain. Masternodes then have 30 days to vote on the proposal. If that proposal has 10% more “yes” votes than “no” votes, it is considered to have passed. There are currently ca. Dash 5,000 Masternodes in the ecosystem.
Maker Dao is a crypto-collateralized stablecoin made via a smart contract platform on Ethereum that backs and stabilises the value of Dai (its native stablecoin) using a system of Collateralized Debt Positions (CDPs). Governance in MakerDao is exercised via the MKR token. MKR, or Makercoin. MKR is a token on the Ethereum blockchain (like the rest of the Maker ecosystem) that has governance rights over the Maker smart contracts. MKR holders can elect an Active Proposal — a smart contract allowed to edit the Maker Platform. This proposal can be in 2 forms:
- Single Action Proposal Contracts (can be executed only once after gaining approval, they delete themselves after use and cannot be re-used)
- Delegating Proposal Contracts (continuously utilise root access).
Black Swan events are the biggest threat to this ecosystem. MakerDao’s solution for situations like this is as follows:
In return for regulating the system, MKR holders are rewarded with fees. However, they also function as buyers of last resort. If the collateral in the system is not sufficient to cover the amount of Dai in existence, MKR is created and sold in the open market to raise the additional collateral. This acts as a strong incentive for MKR holders to responsibly regulate the parameters at which CDPs can create Dai, as it is the money of MKR holders to be at risk if the system fails, not that of Dai holders.
On-chain governance may seem more stable, formalised and clean in terms of the decision making processes. However, many experts in the field, including Vlad Zamfir, claim that the risks outweigh the rewards. On-chain systems are generally easier to game, i.e. exploit using game theoretical frameworks. In off-chain governance, a full node decides for itself whether to make sense to download and implement an update. Presence of autonomous full nodes provides an important set of checks and balances within the ecosystem. In on-chain governance, a full node follows the decision made by the governance process. If on-chain governance process fails, i.e. implements an update detrimental to the protocol, all full nodes will have implemented it and put the whole network in danger.