Liquid Delegation in DAO Governance

Liquid delegation is a collective decision-making mechanism that combines direct participation with representation. The core idea is simple: every voter can either vote directly on a given question or delegate their voting power to others, while maintaining the right to revoke their delegation at any time.

Although invented already in the 19th century, liquid delegation became more popular in the last couple of decades as advancements in digital technology significantly reduced the cost of administering voting systems. The most recent wave of experimentation in liquid delegation has been driven by decentralized autonomous organizations (DAOs) that combine software-based automation and tokens on public blockchains with distributed, mostly online, collaboration. Before turning to the learnings from and ongoing challenges of delegated decision-making in DAO governance, let’s first summarize the main advantages and disadvantages of liquid delegation as commonly understood.

Advantages of Liquid Delegation

A positive view of liquid delegation assumes the desirability of representative governance: those affected by a decision should have a say in making it. Typical examples include representative democracies where political power is allocated through general elections in which every voter has a single vote, and the governance of business organizations where voting power is usually proportional to shares of ownership. Adding liquid delegation to these voting systems would result in either liquid democracy (the most common example of liquid delegation discussed in the literature [1]) or liquid plutocracy (in the sense that the concentration of decision-making power derives directly from the concentration of economic ownership). In both cases, liquid delegation offers the following advantages over more conventional forms of governance:

• Specialized division of voting labor without disempowering individual voters who may prefer to rely on their own expertise, resources, and subjective disposition when deciding on certain matters. Therefore, liquid delegation represents a middle ground between representative elitism/technocracy and direct democracy.
• By formally empowering every member of the voting population to either delegate their voting power or to vote on their own behalf, liquid delegation enables a more granular expression of preferences, thereby supporting more inclusive and representative decision-making.
• If voters can allocate their voting power among multiple delegates based on expertise in specific subject matters, and these delegates can further distribute voting power among other delegates (a practice known as meta-delegation), liquid delegation enables the distribution of decision-making authority based on field-specific ability and merit.
• By allowing voters to overturn the votes of their delegates and continuously re-allocate their voting power, liquid delegation establishes a direct feedback and accountability mechanism between individual voters and their delegates. It also allows individual voters to become more or less active decision-makers based on their constantly changing personal life situations.
• Liquid delegation avoids locking voters into narrow decision-making regimes for extended periods of time, thereby minimizing the likelihood of policies that lack majority support. Under liquid delegation, whenever a highly contentious question arises, voters don’t have to settle with the answer that their most recently appointed representatives choose, but can always impose their personal preferences instead.

Disadvantages of Liquid Delegation

All governance systems involve trade-offs, and liquid delegation is certainly no exception. Here are the main disadvantages of liquid delegation, many of which stem directly from the theoretical advantages listed above:

• It is unrealistic for the majority of voters to participate directly in most votes. [2] While this encourages delegation, the lack of specialized knowledge required in voting on individual questions also extends to identifying the most competent delegates. In practice, liquid delegation may therefore end up looking indistinguishable from traditional forms of representative governance, characterized by large information asymmetries between voters and their proxies, as well as an excessive role of electoral politics in driving decisions and outcomes.
• Similar to more traditional forms of representative governance, rivalry among delegates may gradually transform into a popularity contest as opposed to a battle between the quality and practicality of competing policy ideas. As a result, decision-making power may concentrate in the hands of a small number of delegates whose popularity among individual voters is more a product of their skill in manipulating public opinion than their ability to make sound policy decisions.
• Simply enabling delegation does not guarantee a pool of competent delegates for voters to choose from. Depending on the incentives provided to delegates, the role may not attract a sufficient number of candidates, or it may attract candidates without the required skills, experience, and integrity. Without binding obligations on the most capable members of the voting population, the quality of decision-making under liquid delegation will always depend on who at any given moment is willing to participate in voting.
• While many decisions do indeed require specialized knowledge, it is usually impossible to assess ex ante the merits of expert opinion relative to the majority opinion. Similarly, without predefined and broadly agreed-upon ways of measuring outcomes, individual voters may find it difficult to assess whether their delegates are doing a good job or not, which is particularly problematic whenever the intuitions of the former diverge from the conclusions of the latter.
• In its most basic form, liquid delegation results in the tyranny of the majority of voting power. By allowing individual voters to overturn and revoke votes at any time, liquid delegation encourages delegates to vote according to what their delegators prefer (lest they risk losing their votes), and thus lacks an in-built mechanism to defend the interests of the minority against those of the majority. In other words, as long as there are no limits to the voting rights of individual voters (i.e., they are free to vote according to their subjective preferences on all questions at all times), delegates are likely to act more as agents of their delegators as opposed to stewards of some higher principle or objective that may need to be shielded from the volatility of popular opinion.
• By allowing individual voters to overturn votes and revoke their delegations at any time, liquid delegation may result in inconsistent decision-making. Both popular opinion and the distribution of voting power may change at any moment, making it more challenging to engage in long-term planning and developing synergies between different policies.
• Liquid delegation is difficult to implement at scale (i.e., with a large voter and delegate population) without the use of digital technology. As a result, liquid delegation is unsuitable in situations where many voters don’t have access to digital technology or lack the skills to use it.
• Despite efficiency gains from digitization, liquid delegation may still involve considerable administrative overhead. Facilitating the exchange of key information, managing delegations (especially when voting power can be distributed among multiple delegates based on subject matter), coordination among voters and their delegates, monitoring delegate activity, administering the voting process, and measuring the impact of policy decisions all increase the cost and complexity of liquid delegation. This cost and complexity can be particularly difficult to justify when there is little evidence that liquid delegation is increasing voter participation or the quality of decisions being made.

Decentralized Autonomous Organizations: A Defition

Decentralized autonomous organizations (DAOs) are organizations that combine automation (primarily via public blockchain networks and smart contracts) with crypto-economic incentives (primarily via blockchain-based tokens) and human collaboration (primarily web-based and often but not necessarily pseudonymous). As most DAOs aspire to minimize single points of control and failure, a more compact way to describe DAOs is simply as censorship-minimized digital organizations that rely on public blockchain networks as their core administrative infrastructure.

A commonly used but a rather narrow definition of DAO governance focuses on collective decision-making regarding the DAO’s rules (especially rules inscribed in software protocols) and the distribution of its resources (usually in the form of digital assets that are used to pay for contributions to the DAO). However, governance can also be defined more broadly as the process of applying any design feature or control mechanism that maintains and steers a system. Under this definition, DAO governance includes not only human decision-making in the context of changing the DAO’s rules but also the continuous effects of past design choices, especially those related to software-based automation for administering information and coordinating human action.

DAOs vary in purpose and governance but, as of this writing, three types of DAOs stand out as most common:

1. Public blockchain networks, such as Bitcoin and Ethereum, which are often not referred to as DAOs but nonetheless fall under the definitions provided above. The governance of these DAOs is focused on changing and maintaining blockchain protocols and is analogous to that of traditional open-source software projects. It is usually dominated by off-chain structures and processes, which start out as spontaneous and informal but tend to formalize as the network matures. [3] Some public blockchain networks, such as Polkadot, have also implemented sophisticated on-chain governance mechanisms that include tokenized voting, enabling anyone with the network’s tokens to participate in DAO governance.
2. Middleware protocol DAOs, such as MakerDAO and Uniswap, which combine governing smart contracts deployed on public blockchains with coordinating the activities of a diverse set of DAO contributors and other stakeholders. This can take a variety of forms, some more decentralized than others, but typical components of governing protocol DAOs are complex off-chain organizations and token-weighted on-chain voting on changing smart contract parameters and/or allocating the DAO’s resources. [4]
3. Online community DAOs, such as gaming guilds, fan collectives, NFT clubs, or any other shared-interest community that coordinates using blockchain-based automation and/or tokens. These DAOs typically have simpler and less formal off-chain governance mechanisms but, similar to protocol DAOs, tend to also rely on tokenized on-chain voting as their primary decision-making mechanism.

In all three cases, a distinguishing feature of DAOs compared to traditional organizations is their openness. Becoming a member of a DAO is almost always much easier than becoming a member of other types of organizations. This openness also extends to the governance process. While DAOs tend to carefully regulate the implementation of technical changes such as upgrading software, there are usually minimal barriers to entry when it comes to making proposals. This means that anyone with internet access can put forward ideas to improve a DAO and there is a realistic path for any such proposal to become implemented.

The remaining discussion will focus on DAOs that have introduced or plan to introduce a delegated voting mechanism that extends beyond the task of adding new data to blockchains. Specifically, it will exclude two types of DAOs where delegated voting either doesn’t apply or serves a very narrow purpose: (1) DAOs that have no formal voting system and instead rely on the informal ‘vote of approval’ by its contributors and users whenever they continue engaging with the DAO after changes have been made to its core software or other operative structures; and (2) proof-of-stake blockchain networks where delegation refers exclusively to individual holders delegating their tokens to network operators who then use these tokens solely to participate in the block production process without acquiring any additional governance powers.

Liquid Delegation in DAO Governance

Perhaps the most consequential — and for anyone familiar with traditional voting systems, unsurprising — takeaway from DAO governance thus far is that simply giving a large distributed constituency the power to vote on what are often highly specialized questions does not guarantee participatory governance, and may even negatively affect the efficiency and quality of collective decision-making. However, as most DAOs have publicly committed to the goal of decentralized governance via tokenized (and usually token-weighted) voting, low voter turnout inevitably raises concerns over decisions being controlled by an active minority of large token holders. This then is the context in which liquid delegation has emerged as a way to streamline and professionalize voting in DAOs without disempowering the distributed set of individual token holders commonly seen as the key stakeholder group in most DAOs.

While practical implementations of liquid delegation in DAOs may differ in their details, the considerations listed below can be considered as universally applicable.

Core principles

• Prior to implementing a tokenized voting system, it is essential to consider the bigger picture of the DAO’s governance. What types of decisions will the DAO be confronted with in the future? What will be the frequency of these decisions? Who or what determines the size and composition of the key decision-making bodies within the DAO? Should these bodies be formally organized and what is the DAO’s approach to issues relating to legal liability and regulatory compliance? [5] Does the DAO need a pre-defined conflict resolution mechanism? How important is it for the software components of the DAO to be upgradeable and would upgradability pose any centralization or censorship risk? Answering these and similar questions will help determine the initial rules of engaging with and managing change within the DAO.
• For a DAO voting/delegation mechanism to be trustworthy, its code must be available for public scrutiny and closed source voting software should always be viewed as a red flag. Similarly, the core principles of how proposals are created and put to vote, what the voting procedure looks like, what happens after the vote concludes, how delegation works, etc., should all be clearly defined and available for DAO participants to examine and improve upon, if necessary. DAOs that aspire to have a rules-based governance procedure should additionally consider producing a foundational document similar to a national constitution or corporate charter. [6]
• The most fundamental aspect of voting in DAOs is how voting power is distributed. Historically, token-weighted voting (1 token = 1 vote) has served as the simplest starting point for most DAOs. Adding delegation to such a system reinforces a dynamic similar to corporate governance where a higher share of ownership means more voting power and many governance-related challenges center around the principal-agent problem. However, concerns over centralization and the mismatch between the distribution of power and day-to-day contributions have driven some DAOs to explore combining or even replacing token-weighted voting with alternatives such as allocating voting power based on personal identification (1 person = 1 vote) or some other verifiable credential or criteria. Typical examples of such criteria include membership in a particular stakeholder group, proof of qualifying action, or some commonly agreed upon measure of merit. Other iterations on decision-making in DAOs include quadratic voting, conviction voting, futarchy, and randomized selection of voters for each voting round, but these increase the complexity of the system and have thus far seen only limited adoption.
• The second most fundamental aspect of voting in DAOs is adding checks and balances to the permissionless proposal-making and tokenized voting mechanisms, especially in relation to mitigating the risk of ‘governance attacks’ through malicious proposals or the majority of voting power acting contrary to the core objectives of the DAO. Examples include a filtering mechanism to limit which proposals can be put to vote, requiring locking tokens in escrow to participate in governance, defining the conditions under which votes pass or fail, multi-signature schemes to ratify and execute passed proposals (as opposed to automatic execution which would enhance the automated nature of the DAO but is also considered riskier), time delays between key steps in the governance process, and specialized governance bodies with veto or other emergency powers.
• Finally, liquid delegation can be viewed as part of a broader temptation, especially in more commercially oriented DAOs, to replicate traditional forms of corporate structure and governance. DAOs that combine delegated voting with other practices typical of centralized organizations while simultaneously trying to stay true to the raison d’être of DAOs — i.e., the ability to operate without centralized points of control and failure — should be prepared to run into problems that culminate in an organizational identity crisis. That is not to say that liquid delegation is incompatible with the principle of decentralization. However, DAOs should carefully assess the long-term implications of trends in their current governance and remain intellectually honest about whether these are consistent with the core principles of DAOs as a unique organizational form.

Privacy and transparency

• Voting systems often involve a trade-off between personal privacy and public transparency. In representative democracies, for example, access to voting in general referenda requires personal identification, whereas voting is private (i.e., the identity behind each individual vote is not publicized). However, the votes of delegates in representative governance bodies tend to be public, because otherwise it would be impossible for voters to hold them accountable. In DAOs, on-chain voting is public by definition but, in most cases, individual voters can remain pseudonymous. While some DAOs also allow for pseudonymous delegates, for increased security and accountability, many have either required or recommended that delegates reveal their identities and publish regular commentary on their voting activity. [7]
• The way transparency is implemented in the context of a DAO’s public discussion and voting procedure can strongly influence its governance outcomes. For instance, if individual voters tend to conform to the opinion of the majority or those considered authoritative within the DAO, publishing the live tally of an ongoing vote, along with the list of how various delegates have voted, could easily bias those who cast their votes later in the process. An alternative would be to publish the final tally only after the vote has concluded. However, prominent DAO participants would then likely use other channels to signal their preferences in an attempt to influence other voters. This is just one example of how backroom politics and the art of manipulating public opinion play a role in DAO governance. Nevertheless, this is not an argument against a transparent deliberation and voting process, which, despite its limitations, forces all participants to defend their views in public and democratizes the ability to initiate change. While an opaque and informal governance process can temporarily defend a DAO against external attacks by obscuring points of control over decision-making, in the long run, it is incompatible with the principles of decentralization and credible neutrality.
• In the future, the privacy and security of voting in DAOs will be significantly enhanced through the use of zero knowledge cryptography, allowing DAOs to increase the privacy of individual votes without undermining the trustworthiness of tallying and publishing results. [8] However, DAOs will likely choose to limit secret ballots only to direct general referenda, while requiring much more transparency in the case of delegated voting.

User experience

• A good user experience (UX) of participating in DAO governance starts with the accessibility and quality of governance-related information. DAOs rely on online forums, social media channels, voting dashboards, and other resources to administer and disseminate such information, the most important of which are documents describing the core purpose and governance process of the DAO. In DAOs that have implemented liquid delegation, these resources should make it easy for DAO participants to understand how to engage in voting, delegate voting power, and become a delegate.
• One of the most significant UX-related costs of implementing liquid delegation in DAOs is the development and maintenance of voting/delegation interfaces. DAOs can either build their own custom interfaces or rely on existing platforms that serve as generalized frameworks and aggregators for both off-chain and on-chain components of DAO governance. [9] While an easy-to-use governance interface does not guarantee an engaged community and a high voter turnout, not having one can seriously undermine the decentralization and participatory potential of a DAO.
• Unless a DAO is intentionally designed to have a centralized governance process that does not require coordination among a large number of people, or the frequency of software and other upgrades is very low, the cost and complexity of DAO governance can easily exceed the ability of its contributors. In the future, natural language processing artificial intelligence can be used to structure governance-related information and make it easier to communicate and comprehend. For example, it can provide voters and delegates with recommendations based on their past voting behavior or incentive alignment with other voters, thereby reducing the cost of governance and facilitating higher levels of participation. [10] However, DAOs that rely heavily on artificial intelligence in their governance should also critically assess the associated risks and, at a minimum, implement reviews and sanity-checks that are independent of machine intelligence.

Inclusivity and representation

• A core value proposition of DAOs is their inclusivity and openness, allowing anyone with an internet connection to participate, regardless of their geographic location, socioeconomic status, or level of expertise. But many DAOs have learned that, without predefined rules to structure the activities of their contributors, the governance process can quickly become inefficient and rife with interpersonal conflict. The more resources a DAO commands and the larger its contributor base grows, the more there is at stake and the more difficult it becomes to coordinate and reach consensus. While empowering representative structures can certainly help alleviate these issues, it also undermines decentralization and makes DAOs more akin to the traditional, hierarchical organizations they’re trying to improve upon.
• The open and permissionless ethos of DAOs should also extend to the accessibility of becoming a delegate. But even with minimal formal requirements, the gradual professionalization of the delegate role combined with a natural bias towards re-electing the most well-known delegates can easily lead to the concentration of voting power in the hands of a small number of individuals. One way to ensure a higher diversity of thought and a healthy rotation of delegates over time is through random selection of voters for each round and fixed delegate term limits. However, such rules only work if the DAO is able to continuously attract a large enough pool of candidates with the necessary knowledge and skills to serve as delegates — something that’s very difficult to achieve without offering adequate financial compensation (discussed below).
• To ensure goal- and incentive-alignment between different stakeholders, DAOs should have a set of core values that are communicated to all participants. Ideally, DAO governance should not be too contentious, because everyone is aligned on what and how the DAO is trying to achieve. But in reality, as the context around and people within the DAO change, differences of opinion are guaranteed to emerge. Clearly communicated values and goals help DAO participants better navigate situations of conflict by distinguishing between short-term interests of individual stakeholders and the long-term mission of the DAO as a whole. This is especially important in the case of delegates who, on the one hand, serve as representatives of individual voters, but on the other hand, should do so with the interests of the whole DAO in mind.

Metrics and accountability

• All delegated voting systems require an accountability mechanism whereby those who delegate their voting power can assess the actions of their delegates and decide whether to continue delegating to them or not. Depending on the DAO, this may involve establishing success metrics and a cost-efficient method for individual voters to apply these metrics when assessing the outcomes of past governance decisions. This can be combined with field-specific delegation, i.e. distributing voting power between multiple delegates based on subject matter, which can be particularly useful in DAOs that engage in a large variety of activities that may be difficult to measure under the same rubric. In addition to helping DAOs measure their success internally, metrics are also a great way to communicate the reason and value behind DAOs to the rest of the world.
• When developing metrics for assessing the quality of delegates’ decision-making, DAOs should think carefully about the distinction between short- and long-term goals and strategy. One of the downsides of allowing voters to override the votes of their delegates or redelegate their vote at any time is that it can make the governance process too erratic, especially if the long-term strategy of the DAO is poorly defined or it’s not clear when and how its effectiveness can actually be measured. Often, the true value of a particular strategy becomes evident over longer periods of time and if the success metrics of the DAO fail to take this into account, it becomes more difficult to assess whether a particular strategy is working or not.
• If one of the goals of a DAO is to encourage decentralization by increasing the number of active contributors and governance participants, developing a highly complex strategy with a large number of success metrics can have the opposite effect, especially if the DAO does a poor job at knowledge curation. One way to make it easier for participants to understand how well the DAO is doing is by making the most important information available in a highly condensed and actionable format. Typical examples include short un-opinionated summaries of the most relevant recent developments and community dashboards for tracking core metrics, progress on specific objectives, or delegate activity (which can be encapsulated in a reputation score, for example). Using these and similar techniques can help bring more eyes on how the DAO is functioning and thereby increase both decentralization and accountability.

Competition and compensation

• Public blockchain and smart contract protocols are powerful technologies of collaboration. However, the DAOs that form around these technologies are also fields of competition over ideas and resource allocation, and delegated tokenized voting is one way to formalize this competition. By establishing a delegative voting mechanism that is maximally inclusive, DAOs create a more democratic and transparent mechanism for choosing between competing ideas and decision-makers. Without such a mechanism, DAO governance is more prone to technocratic centralization — a tendency that becomes increasingly problematic as the global user base of the services provided by DAOs grows.
• In order to encourage fair and merit-based competition among delegates, it is important to establish clear guidelines on how delegates are selected. Ideally, the open and permissionless ethos of DAOs should also extend to the accessibility of becoming a delegate. In practice, however, DAOs may prefer to distinguish between different groups of delegates based on their voting power to allocate special rights and responsibilities or decide which delegates should receive compensation from the DAO. If performing delegate duties requires highly specialized knowledge, DAOs may also introduce additional requirements based on professional qualifications or past experience. The more prescriptive these requirements are, however, the smaller the competitive landscape of delegates becomes.
• The role of a delegate does not have to be performed by a single individual and many DAOs already allow for collective entities to serve as delegates. This includes student organizations, investment funds, research/consulting firms, or groups of DAO participants with aligned interests. By fostering a diverse community of both individual and group delegates, DAOs can better distribute governance labor and tap into a broader range of perspectives, which can improve the quality of their decision-making, but only if combined with a clearly defined and efficient governance process.
• To attract the most capable delegates and ensure that they have enough time to perform their duties, DAOs should compensate their delegates. However, allowing delegates to participate in votes that determine the details of this compensation creates a conflict of interest. One of the benefits of liquid delegation in this context is that individual voters can always override the votes of their delegates if they feel that delegates are taking excessive advantage of their position. Alternatively, delegated voting can be disabled in certain situations to rely on a general referendum or an independent third party to make the decision instead. In other words, if we accept that most social systems, including DAOs, are ultimately governed by a well-organized elite, the best way to minimize conflicts of interest is through institutional checks and balances that are based on objective, collectively agreed upon rules as opposed to subjective will or opinion.
• One of the more controversial aspects of tokenized DAO governance is the existence of short-term lending markets for voting power. The problem of low voter turnout also extends to delegation and vote markets are one way for DAOs to mobilize voting power. From a practical point of view, lending out votes is not that different from delegation, except that token holders are indifferent on who exactly ends up using their voting power as they’re willing to give it to the highest bidder. However, vote markets also increase the risk of malicious governance attacks, especially if the number of votes available for borrowing is high and the cost of borrowing votes is low. To mitigate this risk, DAOs can introduce measures that increase the cost of participating in governance with borrowed votes, such as requiring voters to lock their tokens for an extended period of time to gain the right to participate or tying voting power to measures other than the number of tokens held.

Conclusion and Looking Ahead

For most organizations, the ability to learn and manage change is a key determinant of long-term survival and success. This presents a particularly difficult challenge for DAOs who are faced with a unique organizational dilemma: how to collectively make decisions without undermining the principles of decentralization, censorship-resistance, and the autonomous delivery of digital services? For some DAOs, the answer is to trust the early discretion of their creators, ossify as many system parameters as possible, and whenever someone does manage to implement a change, let each user and contributor decide individually whether to continue engaging with the DAO or not. But for many others, tokenized voting with liquid delegation has emerged as the go-to method for deciding between competing paths and proposals.

There are certainly benefits to liquid delegation in DAO governance. It brings formal clarity to the decision-making process, empowers individual DAO participants, encourages more transparent and merit-based competition, and establishes a direct feedback and accountability mechanism between different stakeholders. However, it also triggers electoral politics, introduces an element of instability, and without checks and balances, can increase the risk of centralization. In some cases, liquid delegation may be one among many reasons why a particular organization is best not described as a DAO at all. However, liquid delegation is far from incompatible with the ethos of DAOs and there will be many that manage to stay true to this ethos while relying on delegated voting as their primary collective decision-making mechanism. Liquid delegation in DAO governance is not an oxymoron. While it may signal the beginning of the end for some organizations as DAOs, for others, it is just the beginning.

Footnotes

[1] See for example Blum, C. & Zuber, C. I. (2016). Liquid Democracy: Potentials, Problems, and Perspectives. The Journal of Political Philosophy, 24(2), 162–182 (available here).

[2] Consistently well-informed decisions with high levels of voter turnout are realistic only if the frequency of votes is very low (for example in the context of governing highly stable systems that don’t need to evolve as their environment changes), or the content of votes doesn’t presume any specialized knowledge (for example matters of subjective taste/opinion as opposed to objective expertise/veracity).

[3] I’ve discussed the similarities between traditional FOSS and blockchain governance here.

[4] Disclosure: The author’s work is funded by Placeholder, an investor in BTC, ETH, DOT, and MKR. For a full list of Placeholder’s investments, see here.

[5] For an overview of legal and regulatory challenges of DAOs, see Wright, A. (2021). ’The Rise of Decentralized Autonomous Organizations: Opportunities and Challenges’. Stanford Journal of Blockchain Law & Policy (available here). For a more opinionated view on the most appropriate legal framework for DAOs, see Kerr, D. & Jennings, M. (2022). ‘A Legal Framework for Decentralized Autonomous Organizations’. A16z Crypto (available here); and Jennings, M. & Kerr, D. (2022). ‘A Legal Framework for Decentralized Autonomous Organizations. Part II: Entity Selection Framework’. A16z Crypto (available here). As the institutional field of DAOs matures, their legal structuring is likely to converge towards an increasingly limited set of standard types. I’ve previously discussed this and other isomorphic tendencies in DAO governance here.

[6] See this overview by the Metagovernance Project to learn more about the use of constitutions in DAO governance. See also Zargham, M., Alston, E., Nabben, K., & Ben-Meir, I. (2023). ‘What Constitutes a Constitution?’ Medium (available here).

[7] Transparency is widely considered as a core pillar of modern representative democracies. However, it also makes the highest levels of political decision-making more susceptible to lobbying — a practice that is usually not accessible to socio-economically marginalized or vulnerable groups. See here for research on the potential downsides of legislative transparency.

[8] See the Aragon ZK Research blog for examples of applying zero-knowledge cryptography in DAO voting. Disclosure: The author’s work is funded by Placeholder, an investor in ANT. For a full list of Placeholder’s investments, see here.

[9] Examples include Aragon, Boardroom, Council Kit (by Element Finance), and Tally. Disclosure: The author’s work is funded by Placeholder, an investor in ANT, Element, and Tally. For a full list of Placeholder’s investments, see here.

[10] Early examples of using natural language processing to ease the administrative burden of DAO governance are Governatooorr (by Autonolas) and Missio.

Related Writings

Isomorphism in DAO Governance (May 2021)

The Evolving Landscape of Digital Governance (February 2021)

Ten Theses on Decentralized Network Governance (September 2020)

Institutional Participation in Token-Weighted Network Governance (June 2020)