Measuring the Concentration of Power in the Collective - Governance Design 📐 - Optimism Collective

https://gov.optimism.io/t/measuring-the-concentration-of-power-in-the-collective/8956

This research analysis was conducted for the Optimism Collective as part of the Foundation Grant 2.

TL;DR

This research report presents the Concentration of Power Index (CPI), an adapted version of the traditional Herfindahl-Hirschman Index (HHI), developed to assess the concentration of power within the Optimism Collective. By incorporating both individual delegate voting power and the influence of governance bodies - including Houses, Councils, and Committees (HCC) - the CPI provides a detailed view of governance dynamics. The report highlights key findings on power distribution, the role of delegates, and the CPI’s ability to adjust to changes in governance structures, showing its potential to guide policy changes and support decentralization efforts.

Acknowledgement

We sincerely thank the Optimism Collective for sponsoring and supporting this research. Their support allowed us to explore the governance structures within the Collective. We also want to thank @optimistic_emily & @elizaoak for their support and guidance, which were key to our research’s success. We thank @v3naru_Curia and the curiaLab team for their valuable data contributions wherever needed. This report could not have been completed without the collaboration of everyone involved, and we deeply appreciate their efforts. Lastly, we thank all survey participants for measuring each HCC’s influence.

Team - @ARDev097 @Chain_L @Mary17

Introduction

The governance structure of the Optimism Collective plays a key role in supporting decentralized decision-making and ensuring fair representation of stakeholders. However, with ongoing developments within the governance system, evaluating the concentration of power among delegates and governance bodies becomes important. This research report introduces the Concentration of Power Index (CPI), which enhances the traditional Herfindahl-Hirschman Index (HHI) to provide a clearer view of power dynamics. By assessing both individual voting power and the roles of different governance bodies, the CPI aims to check for potential risks related to power concentration while promoting fair governance practices.

Methods

The research began by gathering data on delegates’ voting power from Dune Analytics, offering insights into the influence of individual delegates within the Optimism Collective’s Token House. Information about the various councils and committees, including their members and responsibilities, was sourced from the Optimism Forum.

To accurately measure power concentration, the Concentration of Power Index (CPI) was created to factor in both the individual voting power of delegates and their participation across multiple governance bodies, such as Houses, Councils, and Committees (HCC). This dual approach enables a more detailed assessment of power dynamics and their effects on governance within the Optimism ecosystem.

Method for Measuring the Influence of Optimism Governance Houses, Councils, and Committees (HCC)

The framework for measuring influence is based on several key factors, including Decision-Making Authority, Scope of Influence, Community Engagement, Operational Independence, Voting Power, and Veto Power. Each factor is weighted to reflect its significance in the overall evaluation.

The governance structure comprises several Houses, Councils, and Committees (HCCs) such as the Token House, Citizens’ House, and Grants Councils. The weighted score of each HCC is calculated using specific formulas:

• Token House:

Likewise, the weighted scores for other HCCs are calculated using similar formulas:

Method for Calculating the CPI:

The CPI was determined by combining the influence of Token House delegates with other governance layers, including various Houses, Councils, and Committees. This approach provides a more comprehensive measure of power concentration within the Optimism Collective.

CPI Formula: The Concentration of Power Index (CPI) is mathematically defined as follows:

This detailed methodology ensures that the CPI accurately captures the complex relationship between individual voting power and the collective influence of governance bodies, offering a reliable framework for analyzing power concentration within the Optimism Collective.

Results

Influence of Governance Bodies Within Optimism Collective

The table accompanying this report outlines the influence of each governance body, calculated using the weighted parameter scoring method. The Citizens’ House has the highest influence at 34.59%, followed closely by the Token House at 32.33%. The Grants Council (Builders & Growth Experiments Sub-committee) contributes 10.15%, while the Security Council accounts for 13.17%. Other bodies, such as the Grants Council (Milestone & Metrics Sub-committee) at 2.82%, the Code of Conduct Council at 4.32%, and the Developer Advisory Board at 3.01%, have comparatively lower influence, reflecting the distribution of decision-making authority across the different governance bodies.

Concentration of Power Index (CPI) Across Different Rounds and Seasons

The Concentration of Power Index (CPI), measured across different rounds and seasons, is shown in the graph below. The initial CPI for the Token House was 329.25. After including data from Citizen House Round 2 and the councils for Season 3, the CPI significantly dropped to 140.13 and remained stable during Season 4. With the addition of data from Season 5, the CPI further decreased to 91.81, reflecting a clear trend toward decentralization. This trend continued with Citizen House Round 3 at 81.53 and Round 4 at 80.90, highlighting the ongoing decentralization within the governance structure.

Herfindahl-Hirschman Index (HHI) Across Different DAOs

The Herfindahl-Hirschman Index (HHI) values for Optimism, Compound, Aave, and Uniswap are displayed in the graph below, showing the concentration of voting power among delegates within each DAO. The HHI for Optimism was calculated using only Token House delegates, without considering other governance layers, for easier comparison. The HHI values for Compound, Aave, and Uniswap were determined based on the distribution of voting power among their respective delegates.

Concentration of Power Index (CPI) Across Different DAOs

The graph below displays the CPI values for Optimism, Compound, Aave, and Uniswap, representing the concentration of power in each DAO. Optimism’s CPI of 80.90 shows a decentralized power structure, while Compound’s 320.72 and Aave’s 627.34 indicate more centralized governance. Uniswap’s CPI of 215.29 suggests a balance between decentralization and concentration but is still less decentralized than Optimism. For other DAOs, only their delegate’s voting power is considered for this comparison.

Nakamoto Coefficient Across Different DAOs

The Nakamoto Coefficient, which measures the minimum number of members needed to control over 51% of the voting power, is represented in the bar chart below for Optimism, Compound, Aave, and Uniswap. Optimism requires 21 members to surpass this threshold, indicating relative decentralization. In contrast, Compound requires 13 members and Aave only 8, reflecting a higher concentration of power. Uniswap’s Nakamoto Coefficient is 17, suggesting a decentralized power structure, though slightly more concentrated than Optimism.

Conclusion

The Concentration of Power Index (CPI) is a valuable tool for evaluating power distribution within the Optimism Collective. It improves on the traditional Herfindahl-Hirschman Index (HHI) by considering the influence of various governance bodies, such as Houses, Councils, and Committees. The CPI highlights areas where power is concentrated among a small group of delegates, raising concerns about over-centralization, while also identifying areas where power is more evenly spread, supporting decentralization. It provides a clearer understanding of the impact of delegates involved in multiple governance roles, offering a more detailed view of decision-making power. Additionally, the CPI can adapt to changes in governance structures, allowing for real-time assessments of power dynamics. By pinpointing periods of concentrated power and successful moves toward equitable distribution, the CPI helps guide policy changes and governance improvements. Ultimately, the CPI is key to maintaining decentralization and promoting a fair, inclusive governance structure within the Optimism Collective. The temporal chart of CPI can provide the direction in which our Collective is moving in terms of centralization as CPI is modified applying the same benchmark as HHI will not be an accurate measure.

Source:- CPI Site 3

Resources

• Detailed Report Link - Report 6
• Frontend Dashboard - Frontend Dashboard 3
• Influence Calculator - Influence Calculator 3
• Presentation Link - Presentation 1
• Observation File - Observation File 1
• Behind the Scenes Documentation - Behind the Scenes 2
• GitHub Repo - Repo Link 3

Likewise, the weighted scores for other HCCs are calculated using similar formulas:

Herfindahl-Hirschman Index (HHI) Across Different DAOs

Concentration of Power Index (CPI) Across Different DAOs

Nakamoto Coefficient Across Different DAOs

A big shoutout to the team for putting in the effort on this research! This index is super important, and the insights are spot-on. Love how you included all the governance bodies—it’s a really thorough job. Great work!

optimistic_emily

1d

Thanks for sharing these results @Chain_L

This is a new and interesting way of approaching the measurement of the concentration of power in DAOs with more complex governance structures.

I personally would love to be able to play around with the inputs to see how the outcomes are affected - like a kind of interactive simulator.

I also suspect that we could get 80% of the accuracy with 20% of the complexity - for example by taking a shortcut to determine the “influence” of each of the houses - instead of using weights I wonder if crowdsourcing people’s intuitions would produce results of a similar accuracy.

Anyway, thanks @Chain_L for the great work

Disclaimer: I work for the Optimism Foundation, but views are my own

Chain_L

Thank you for your appreciation.

optimistic_emily:

That’s a great feedback, we will get the simulator ready.