Decentralization refers to a structure of control and coordination where the participants are part of a distributed network, and members or entities on that network do not need to emphasize trust in any particular entity. The system is designed to minimize trust while ensuring the network's functionality to attain its goals. Governance is one approach to achieving decentralization by utilizing blockchain technologies that can power the decision-making process, operation, and development of protocols typically through smart contracts and tokens that the governance participants own.

The remainder of this article will present a comparative analysis of various approaches to solving governance undertaken by Optimism and other comparative Layer 2 networks such as Arbitrum, Starknet, and ZKsync.

Optimism’s core approach to governance is a non-plutocratic system that is resistant to capture by any one entity or small group of entities. This means token holders do not have the only say concerning protocol upgrades, resource allocation, and innovation. The accumulation of OP governance tokens will not help any entity capture the value of the network as Optimism’s governance is based on a bi-cameral architecture where the second component known as the Citizens’ House acts as a checks and balance mechanism to the Token House and vice versa.

The Optimism Collective bills itself as an experiment in governance and is designed to be iterative because of a belief that vision can sometimes be in opposition to value creation. The Token House consists of holders of OP governance tokens who can participate directly in voting using their tokens or can delegate the tokens to entities to vote on their behalf who they believe share their values. The Citizens’ House on the other hand is made up of citizens, individuals who have demonstrated good faith actions within the optimism ecosystem. Citizenship is conferred by a soul-bound NFT token and is not transferrable.

In contrast with the Token House where a single entity can wield enormous votes occasioned by the number of tokens they own, citizens are only entitled to one vote, further deepening the democratic process. The main role of the Token House is to vote on protocol upgrades and project incentives while the Citizens’ House mainly governs the Retroactive Public Goods Funding, in addition to ensuring that the protocol pursues long-term visions and is not captured by any entity by ensuring a human-centric governance structure. The Token House can veto citizenship eligibility as a check on the powers of the Citizen’s House. A central modus operandi of the Optimism Collective is that one house has a set of core responsibilities that can be vetoed by the other and vice versa.

The Arbitrum Layer 2 network is governed by the Arbitrum DAO, which comprises a community of $ARB token holders who can propose and vote on changes to the technologies developed for the network. Token holders can participate directly in governance or through elected representatives called delegates. Arbitrum DAO just passed a proposal to introduce the $ARB token staking and transform it from a pure governance token into a dual utility token. Governance on Arbitrum will then be based on the liquid staking token called staked ARB token ($stARB). The rationale for the proposal is to accrue value to $ARB and increase participation in governance since $ARB locked-in DeFi protocols were not compatible with governance.

Arbitrum hopes to increase governance participation as only about 10% of the circulating supply of $ARB was actively used in governance. The Arbitrum DAO is implemented as smart contracts and is responsible for managing the built-in treasury system. It also has a Security Council mechanism that can bypass the process of governance to perform protocol upgrades in cases of emergency. The Security Council, a key part of the governance structure, is responsible for making decisions in these critical situations. The members of the Security Council are elected by the Arbitrum DAO. The general governance structure of the Arbitrum is plutocratic as $ARB token holders are the main arbiters of the system.

Starknet's governance mechanism is unique in the Layer 2 landscape because it has a dual utility token that is not used directly for voting but abstracts it as voting power. STRK, the base token of Starknet, is not a direct governance token as it is used as the fee payment token on the Starknet network. To enable its dual utility, STRK has to be wrapped as vSTRK to be used in governance. STRK tokens are incompatible with governance; only vSTRK is used directly to vote for proposals or to designate a delegate with the equivalent voting power of the vSTRK tokens.

vSTRK tokens can be unwrapped to STRK, but every wrapping and unwrapping operation incurs gas costs. The other fundamental difference of Starknet's governance mechanism is "progressive governance," an approach that constantly moves the network's components toward full decentralization. As a result of this philosophy, Starknet has various entities and councils that shepherd certain aspects of the protocol. Some of those entities include the Starknet Governance Committee, the Builder's Council, and the Starknet Foundation.

ZKsync's governance approach is built around three fundamental principles: resilience, distribution, and mission alignment. It ensures that no single entity holds unilateral power over the protocol, promoting long-term stability and community control. The system is decentralized, with decision-making spread across three bodies: the Token Assembly, Security Council, and Guardians.

The Token Assembly comprises token holders and Delegates, who can propose and vote on protocol changes. The Security Council oversees technical upgrades and has the authority to address security risks. The Guardians ensure all proposals align with ZKsync's core values and can veto misaligned decisions. This layered governance structure creates checks and balances, promoting transparency, security, and alignment with ZKsync's mission

The table above shows that Optimism, Arbitrum, ZKsync and Starknet all retain emergency powers that can bypass the governance process to implement changes during emergencies. Those emergency powers reside in the Security Council of the protocols. Similarly, all the networks reviewed have treasuries controlled by the governance process. Optimism and ZKsync have non-plutocratic governance structures that de-emphasize economic power.

Merely owning more governance tokens will not lead to an outsized influence. Starknet has a dual utility token and Arbitrum’s governance token may soon have a function outside voting if the staking proposal is implemented. Optimism and ZKsync governance tokens on the other hand remain pure governance tokens for the time being.

This article explored the current governance landscape across four leading Ethereum Layer 2 networks - Optimism, Arbitrum, Starknet, and ZKsync. It started with a brief explanation of decentralization and governance before analyzing the governance mechanisms of the various protocols. The main takeaways are that Optimism operates a bicameral system where the power of token holders is moderated by citizens and vice versa. Arbitrum governance is led by Arbitrum DAO through $ARB token holders in a plutocratic system that emphasizes the amount of tokens held. Arbitrum is trying to increase participation and drive value accrual to governance token holders by introducing staking to its governance ecosystem.

Starknet governance is also plutocratic with a dual utility governance token, however, it has various councils and committees that shape the protocol.The ZKsync governance system is designed around the principles of separation of powers and checks and balances. By design, there is not a single person or entity has the power to make changes to the ZKsync protocol.

Though decentralization is a spectrum, protocols should try to be as decentralized as possible to avoid censorship and capture by malicious actors. All the protocols reviewed can further decentralize the emergency powers of security councils and explore ways of eliminating such powers through innovative experiments that place the community at the center of governance.

Decentralization refers to a structure of control and coordination where the participants are part of a distributed network, and members or entities on that network do not need to emphasize trust in any particular entity. The system is designed to minimize trust while ensuring the network's functionality to attain its goals. Governance is one approach to achieving decentralization by utilizing blockchain technologies that can power the decision-making process, operation, and development of protocols typically through smart contracts and tokens that the governance participants own.

The remainder of this article will present a comparative analysis of various approaches to solving governance undertaken by Optimism and other comparative Layer 2 networks such as Arbitrum, Starknet, and ZKsync.

Optimism’s core approach to governance is a non-plutocratic system that is resistant to capture by any one entity or small group of entities. This means token holders do not have the only say concerning protocol upgrades, resource allocation, and innovation. The accumulation of OP governance tokens will not help any entity capture the value of the network as Optimism’s governance is based on a bi-cameral architecture where the second component known as the Citizens’ House acts as a checks and balance mechanism to the Token House and vice versa.

The Optimism Collective bills itself as an experiment in governance and is designed to be iterative because of a belief that vision can sometimes be in opposition to value creation. The Token House consists of holders of OP governance tokens who can participate directly in voting using their tokens or can delegate the tokens to entities to vote on their behalf who they believe share their values. The Citizens’ House on the other hand is made up of citizens, individuals who have demonstrated good faith actions within the optimism ecosystem. Citizenship is conferred by a soul-bound NFT token and is not transferrable.

In contrast with the Token House where a single entity can wield enormous votes occasioned by the number of tokens they own, citizens are only entitled to one vote, further deepening the democratic process. The main role of the Token House is to vote on protocol upgrades and project incentives while the Citizens’ House mainly governs the Retroactive Public Goods Funding, in addition to ensuring that the protocol pursues long-term visions and is not captured by any entity by ensuring a human-centric governance structure. The Token House can veto citizenship eligibility as a check on the powers of the Citizen’s House. A central modus operandi of the Optimism Collective is that one house has a set of core responsibilities that can be vetoed by the other and vice versa.

The Arbitrum Layer 2 network is governed by the Arbitrum DAO, which comprises a community of $ARB token holders who can propose and vote on changes to the technologies developed for the network. Token holders can participate directly in governance or through elected representatives called delegates. Arbitrum DAO just passed a proposal to introduce the $ARB token staking and transform it from a pure governance token into a dual utility token. Governance on Arbitrum will then be based on the liquid staking token called staked ARB token ($stARB). The rationale for the proposal is to accrue value to $ARB and increase participation in governance since $ARB locked-in DeFi protocols were not compatible with governance.

Arbitrum hopes to increase governance participation as only about 10% of the circulating supply of $ARB was actively used in governance. The Arbitrum DAO is implemented as smart contracts and is responsible for managing the built-in treasury system. It also has a Security Council mechanism that can bypass the process of governance to perform protocol upgrades in cases of emergency. The Security Council, a key part of the governance structure, is responsible for making decisions in these critical situations. The members of the Security Council are elected by the Arbitrum DAO. The general governance structure of the Arbitrum is plutocratic as $ARB token holders are the main arbiters of the system.

Starknet's governance mechanism is unique in the Layer 2 landscape because it has a dual utility token that is not used directly for voting but abstracts it as voting power. STRK, the base token of Starknet, is not a direct governance token as it is used as the fee payment token on the Starknet network. To enable its dual utility, STRK has to be wrapped as vSTRK to be used in governance. STRK tokens are incompatible with governance; only vSTRK is used directly to vote for proposals or to designate a delegate with the equivalent voting power of the vSTRK tokens.

vSTRK tokens can be unwrapped to STRK, but every wrapping and unwrapping operation incurs gas costs. The other fundamental difference of Starknet's governance mechanism is "progressive governance," an approach that constantly moves the network's components toward full decentralization. As a result of this philosophy, Starknet has various entities and councils that shepherd certain aspects of the protocol. Some of those entities include the Starknet Governance Committee, the Builder's Council, and the Starknet Foundation.

ZKsync's governance approach is built around three fundamental principles: resilience, distribution, and mission alignment. It ensures that no single entity holds unilateral power over the protocol, promoting long-term stability and community control. The system is decentralized, with decision-making spread across three bodies: the Token Assembly, Security Council, and Guardians.

The Token Assembly comprises token holders and Delegates, who can propose and vote on protocol changes. The Security Council oversees technical upgrades and has the authority to address security risks. The Guardians ensure all proposals align with ZKsync's core values and can veto misaligned decisions. This layered governance structure creates checks and balances, promoting transparency, security, and alignment with ZKsync's mission

The table above shows that Optimism, Arbitrum, ZKsync and Starknet all retain emergency powers that can bypass the governance process to implement changes during emergencies. Those emergency powers reside in the Security Council of the protocols. Similarly, all the networks reviewed have treasuries controlled by the governance process. Optimism and ZKsync have non-plutocratic governance structures that de-emphasize economic power.

Merely owning more governance tokens will not lead to an outsized influence. Starknet has a dual utility token and Arbitrum’s governance token may soon have a function outside voting if the staking proposal is implemented. Optimism and ZKsync governance tokens on the other hand remain pure governance tokens for the time being.

This article explored the current governance landscape across four leading Ethereum Layer 2 networks - Optimism, Arbitrum, Starknet, and ZKsync. It started with a brief explanation of decentralization and governance before analyzing the governance mechanisms of the various protocols. The main takeaways are that Optimism operates a bicameral system where the power of token holders is moderated by citizens and vice versa. Arbitrum governance is led by Arbitrum DAO through $ARB token holders in a plutocratic system that emphasizes the amount of tokens held. Arbitrum is trying to increase participation and drive value accrual to governance token holders by introducing staking to its governance ecosystem.

Starknet governance is also plutocratic with a dual utility governance token, however, it has various councils and committees that shape the protocol.The ZKsync governance system is designed around the principles of separation of powers and checks and balances. By design, there is not a single person or entity has the power to make changes to the ZKsync protocol.

Though decentralization is a spectrum, protocols should try to be as decentralized as possible to avoid censorship and capture by malicious actors. All the protocols reviewed can further decentralize the emergency powers of security councils and explore ways of eliminating such powers through innovative experiments that place the community at the center of governance.