Exploring the Limits of Decentralized Governance

Introduction

Decentralized governance has been a long-discussed topic – it aims at helping to prevent corruption, empowering a community to have a voice, and ensuring that a crypto network is run fairly and transparently.

But one of the biggest challenges is achieving consensus among the network members: in a decentralized system, there is no central authority to make decisions, so all members must agree on the best course of action. Coordinating and communicating among many decentralized users can be challenging, as there is no central point of control. The potential for malicious actors to try to undermine the governance process, either by trying to sway the consensus or by attacking the network itself, is another threat.

In February 2020, Compound, by introducing the $COMP token, started a trend of DeFi governance tokens. In Robert Leshner’s words, it consists of “a governance system that will replace the Compound protocol’s administrator with community governance — allowing you to suggest, debate, and implement changes to Compound — without relying on, or requiring, our team whatsoever”.

Since then, governance has been one of the spaces that have seen the least amount of innovation. Most governance models are still based on the one-token-one-vote model, which is very vulnerable to market manipulations.

While there may be challenges and limitations to decentralized governance, we believe that there is still significant potential for experimentation and exploration in this area. We have only just begun to uncover the possibilities that the space offers, and it is important to recognize that there is no single, universal solution to every problem. Our aim with this article is not to present a definitive answer or 'silver bullet' solution, but rather to consider the issues and opportunities surrounding governance and consider how we might move forward.

The effectiveness of a governance system will depend on a variety of factors, such as the specific goals and needs of the organization, the size, and complexity of the organization, and the level of participation and engagement of its members.

Current Approaches to Decentralized Governance

1. Decentralization (One Token, One Vote)

The risk of plutocracy in “coin voting” has been widely discussed over the last few years. Small groups of whales have too much influence over large groups of small-holders because the former type of voter has no incentive to actually vote.

As an example, the MakerDAO Endgame Plan, proposed by Rune, passed despite controversy. Even delegation, a political tool created thousands of years ago, doesn’t solve that centralization issue because they are paid in line with the number of tokens they get delegated – it creates a situation where a small number of powerful actors can sway decisions in their favor rather than allowing the broader community to have a say in the decision-making process. Paying delegates has been a source of debate we will not cover for the sake of simplicity.

Delegation can create a dependency on the third parties to whom tokens are delegated – if a large number of token holders delegate their voting rights to a single individual or organization, that third party may effectively become the "gatekeeper" for the entire community, making decisions on behalf of all the delegated token holders.

Consequently, delegation has been relatively underutilized in the ecosystem to date, with less than ten threads created in the Uniswap Delegation Pitch section. Nonetheless, we notice a strong trend towards a delegate class with professional delegates like Flipside, StableNode or Llama. Even some of the biggest DAOs out there like ENS heavily rely on delegation today

Many token holders’ interests are tied to the token price, even though communities are composed of members with different values and goals. This, very often, leads to over-considering the value of the token during voting processes – many proposals aim to reduce the token supply, increase the value of fees paid in the token, or implement a buyback program to increase demand for the token.

But is the fact that tradeable asset is used to vote on proposals necessarily the cause of these pitfalls?

According to data from Messari Governor, the wide majority of governance proposals today pass (or don’t) with dozens or a few hundred individual voters only. For example, 390 people voted to elect the new Sushi Head Chef (vs. 108k+ token holders at the time of voting), and 0.69% of UNI holders voted to create the Uniswap Foundation, four months ago.

The participation rate depends on several factors:

• Fatigue – token holders might face a burden of having to vote frequently on a large number of issues.
• Alienation – token holders may feel that their efforts to influence a protocol are futile or ineffective.
• Privacy – token holders may feel uncomfortable expressing their true opinions publicly.
• Options – even if there have been previous discussions on the topic at hand, token holders may still feel uneasy about the potential choices available to them and may be hesitant to express their opinions.
• UX – token holders might encounter UX problems that deter them from participating in the voting process. Gas fees are still major pain point to on-chain governance, despite interesting projects being developed, such as the Biconomy gasless relayer SDK.

This list is not meant to be comprehensive but rather to demonstrate that hard data on voting can be complex to analyze and influenced by various factors that are often similar to what causes voter apathy in political elections.

There have been various innovations in decentralized governance, but their success relies on a sufficient level of participation in governance – not enough people get involved. Can we use incentives that are native to the Web3 environment to reduce voter apathy?

2. Participation Incentivization

Paying people to vote is not exclusive to crypto – in a study conducted by Fordham University political scientist Costas Panagopoulos, offering cash incentives of $25 increased voter turnout in a local election from 14.9% to 19.2%.

Some crypto-native models have been designed to find the most efficient way to pay people to vote. To increase voter turnout in a project, it is necessary to consider the cost of voting, or the amount that a project would need to pay token holders to motivate them to participate. This model, while straightforward, can provide valuable information about voter characteristics and behaviors when implemented by projects. However, there is a potential for voters to use bots to collect rewards without fully understanding the voting process, which could result in an increase in voter ignorance.

Governance mining is an innovative concept that aims to distribute a portion of the token supply to groups or individuals who contribute to the protocol, based on the value of their contributions. This approach could improve decentralization in governance by offering rewards in the form of governance tokens to those who work to address governance-related issues. While it does not specifically address voter apathy, governance mining could potentially increase participation by providing incentives for individuals to engage in governance activities. But distributing tokens is also a flawed approach – as Jacob himself put it, “people may spam governance in hopes of capturing some portion of tokens.”**
**

The "how do we get people to vote" might be the wrong problem to focus on – participation rates can surely be improved, but as long as speculation will be at the core of any behavior in crypto, voting will never be a top priority.

3. Representational Systems

What if token holders voted on representatives who would then handle day-to-day votes?

A few projects like C.R.E.A.M. Finance, Kwenta, and Synthetix, have been experimenting with term-based governance systems.

Council-based governance is a different approach to the main governance models of token-weighted and delegated governance, which are more like representative democracy and a board of directors. Council-based governance aims to fix the problems of these models by having a smaller group of people with equal voting power make decisions. Token-weighted and delegated governance often have low voter participation and a concentration of voting power. To choose council members for the DAO, Synthetix uses a quadratic voting mode.

As shown by a recent Messari report, there may be some areas for improvement in Synthetix Governance. Still, the Synthetix Councils have generally been successful in adding value to the Synthetix protocol and acting in the best interests of the Synthetix community (808 possible votes could have been cast by the eight Spartan Council members in Q2 and Q3 2022, during each epoch – of these 808 possible votes, 112 votes (13.86%) Spartan Council votes were not cast). The upcoming voting and implementation of Synthetix V3 SIPs (300-312) mean that the Synthetix Community can expect a lot of positive developments in the coming year.

Some would argue that the devil seems to be in the details with representational systems – term length, type of votes handled by the representatives, council members removing process, voting weight mechanisms design, etc. are all parameters that could need adjustments for each protocol to work well.

But there are also a few potentially more serious issues with native-term representation in decentralized governance that may need to be addressed:

• Limited participation – in some cases, a native-term representation may exclude certain individuals or groups from participating in governance processes, potentially leading to a lack of diversity and inclusivity.
• Misalignment of incentives – native-term representation may result in a misalignment of incentives between those who hold native tokens and those who do not, potentially leading to conflicts of interest.
• Complexity – decentralized governance systems that rely on native-term representation may be complex and difficult for some users to understand and navigate, potentially limiting their effectiveness.
• Vulnerability to manipulation – native-term representation may be vulnerable to manipulation by those who hold large amounts of native tokens, potentially leading to unfair or biased decision-making.

4. Gamified Governance

As said before, governance can become a stressful and unpleasant experience while necessary. This is the main reason why governance-minimization frameworks exist – they aim at reducing the power and reliance on governance wherever possible.

However, we’ve recently seen an interesting approach to make governance enjoyable – jokedao allows platforms that want to take action to set up contests and ask users for ideas. Users can then vote on their favorite idea. Participants use tokens to cast their votes based on their unique addresses.

For example, Layer3 used jokedao for its own user-generated roadmap—letting its users nominate and pick the next protocols it would partner with for quests

Gamified governance means that decision-making processes must be collaborative and offer clear rewards for participating.

Gamifying governance is, in a sense, not new – Compound, in August 2021, started drafting ideas to implement a points-based governance system to incentivize users to participate in governance and make the whole experience more enjoyable.

While we think that Web3, by nature, can create a brand new voting experience far from the boring political process by creating a gamified governance process, the clear lack of standardization would make it hard in the short term. Also, decentralized networks often have a smaller user base compared to centralized platforms, which can make it difficult to create a critical mass of users to make gamification effective. And gamification elements may not be easily integrated into existing governance processes, which can make it difficult to implement. We also believe that gamified voting is not sustainable in the long term as the issues mentioned above still exist; despite people’s motivation becoming greater.

While these different experiments have pushed the governance design space forward, none of them has established itself as the golden standard. A fair governance process must rely on a system that creates and maintains inclusivity, transparency, consistency, security, and efficiency.

Other Internet (decentralized applied research organization that received a grant from Uniswap) proposed a few changes to the Uniswap Protocol a few months ago, going in that direction.

At first glance, governance minimization could be a straightforward solution to any governance-related inefficiencies. Whether the “un-governance” model proposed by Vitalik turns out to be true or not, this doesn’t really solve anything at the community and ecosystem level.

Human coordination has always been – and will always be – a source of issues, and the most widely-used protocols generally have to highest levels of informal governance.

Many experiments have been done on non-essential problems – these are necessary but will not create efficient governance systems per se. The Nouns community has been very active in the field, with tools like HeyAnoun that lets nouners give feedback on NounsDAO proposals

We think that governance must be rebuilt from the ground up with governance token distribution, voter incentives, and limited governance complexity as core principles.

So, how did we get there? One (partial) explanation is that decentralized governance was also introduced for regulatory arbitrage. By introducing decentralized governance, teams could kind of skirt the Howey test by saying that it fails on the common enterprise prong, without caring much about the potential issues related to governance.

Mapping The Next Big Things in Governance

1. Get Away From Token-Based Voting With Soulbound Tokens

We should build reputation systems by rewarding voters based on the success of a proposal – souldbound tokens are an interesting tool for this, as ERC-20-based voting has many pitfalls we, at least partially, explored above. Governance power should not be transferable or bought by anyone with no competence to govern. As Vitalik put it, “governance is fundamentally a public good, not a private good”.

SBTs provide a way to recognize and acknowledge the governance-related contributions of individuals in a non-financial manner. They encourage governance to consider alternative methods of retaining valuable members and thanking them for their efforts, rather than just relying on financial incentives.

Some projects like Gitcoin have been experimenting with ideas around Quadratic Voting, but its limitations are due to the possibility for a voter to misrepresent himself as multiple individuals in an attempt to gain an unfair advantage. Since soulbound tokens are tied to a specific address and cannot be transferred, it would not be possible for a single individual to create multiple accounts or identities in an attempt to misrepresent themselves as multiple voters.

Optimism, has proposed using the SBT voting mechanism. It has a bicameral governance model with two divisions: Token House and Citizens House. Members of the Citizens House are involved in governance decisions such as determining how revenue generated by the protocol should be allocated. In its first stage, the House is solely responsible for voting on retroactive public goods funding though.

Even if the risks of poor distribution and the risk of centralization don’t disappear with SBTs, by removing pure financial incentives, they represent an interesting tool to embrace non-financialized decentralized governance.

2. Simplicity Is Key to Onboarding New Governance Participants

Ease of access is a major challenge in the cryptocurrency and governance space. Many people are unfamiliar with the process of purchasing cryptocurrencies and setting up wallets, and may struggle to participate in governance activities such as Snapshot votes and on-chain voting. There are several ways governance processes could be made simpler:

• Reduce barriers to participation – removing high token requirements, or expensive gas fees would make it easier for people to get involved.
• Use user-friendly interfaces – people would more easily participate in governance activities if interfaces were easy to navigate.
• Educating the community – providing resources and information to help people understand how decentralized governance works and how they can participate can go a long way in making it more accessible.
• Regularly reviewing and streamlining processes – per time, governance systems can become cluttered with unnecessary processes and procedures. Regularly reviewing and streamlining these processes can help to keep the system simple and efficient.

3. Multiple Autonomous Centers of Authority Within an Organization

MetaDAOs can be efficient in bringing people together into groups that share similar values and goals, which reduces the chances of conflicts arising due to differences in power or political beliefs. MetaDAOs are in fact an evolution of subDAOs, defined as new organizational structures operating with complete autonomy while remaining aligned with the superDAO.

Rune Christensen tried to demonstrate that voter apathy could be reduced by creating sub/child DAOs that are aligned to MakerDAO but able to seek profit and compete with other MetaDAOs. They would run their own governance processes on top of the Maker Core governance infrastructure. It increases the chances that some of them will be able to achieve strong and widespread success without experiencing corruption or problems caused by freeloading.

The idea of having smaller autonomous groups is not new. Yearn Finance, in a Governance 2.0 proposal, proposed to implement “yTeams”. Yearn contributors are organized into small, self-governed groups that YFI holders grant certain decision-making powers. These groups are responsible for acting in the best interest of Yearn within a specific area of responsibility and have the authority to make specific, defined decisions.

Metropolis is an interesting project that puts working groups on-chain, which creates an organizational "source of truth" for DAOs. Besides adding organizational clarity, Metropolis helps you understand and make sense of the information within your organization. It offers insights on the members, finances, and permissions of each working group, called “Pods”, making it easier to comprehend and navigate the data in an organization.

But there are still grey areas around the emergence and viability of MetaDAOs. In its current form, it’s still hard to scale MetaDAOs. As MetaDAOs are nested within a parentDAO, they can add additional layers of complexity to the governance structure

SubDAOs can also lead to fragmentation as they may have conflicting goals or priorities. This can make it difficult for the parentDAO to make cohesive decisions, leading to reduced scalability. Managing multiple subDAOs can be resource-intensive, requiring coordinating with multiple subDAO leaders and managing multiple sets of smart contracts. This can add overhead and reduce scalability.

We need the right set of tooling and interface to limit organizational complexity and optimize governance efficiency.

4. Exploring the Possibility of a Hybrid Governance Model for Decentralized Autonomous Organizations

We believe that one solution to the plutocracy issue is a hybrid model that combines elements of traditional governance structures with the benefits of decentralized technology.

Combining direct democracy with representative democracy, while ensuring that the decentralized governance process remains inclusive, diverse and fair, by giving representation to communities with different characteristics and interests, based on both on-chain and off-chain insights. With more and more companies building analytics tools, that allow to understand, segment, and activate user data, community members are more likely to elect representatives that truly represent their interests.

As relying on representatives means adding an additional governance layer, we think that easy-to-navigate platforms that improve the searchability of governance-related content should be created. Governance is characterized by silos with poor communication and data, which makes it difficult for the average voter to understand how the system works and its advantages.

Closing Thoughts

Decentralized governance presents a unique and exciting opportunity for experimentation and exploration. While there are certainly challenges and limitations to this approach, we believe that there is significant potential for innovation and progress.

Decentralized governance doesn’t necessarily make sense for early-stage startups – it very often reduces the pace of innovation and can be very distractive for founders whose focus should be on finding early market validation, and potentially product-market fit. To not hold teams back, decentralized governance should always be progressive and focused on specific dimensions instead of the entire governance stack.

As we expect more and more teams to work on more fair token distribution mechanisms at a later stage of their development process, governance powers will remain in the hands of a centralized team longer, and be distributed in smarter ways.

If you are building in any above mentioned areas, please reach out to pierre@lattice.fund.