Voting governance in the DAO Tool

Author: Black Soil, Source: Three Views of Crooked Neck

With the development and popularization of blockchain technology, distributed autonomous organizations (DAOs) have gradually become a new form of social organization, however, DAO pays special attention to automated governance in theoretical research, but there is insufficient automation in practice. The purpose of this paper is to analyze the current situation of DAO governance, mainly analyze the relevant theories and practices of voting governance, discuss the strategic model of voting governance from the perspective of rational people, summarize several basic issues that DAO should pay attention to when choosing voting governance tools, and elaborate on it in combination with the current situation in China.

01 Historical Background

There are many different definitions of DAO in practice and literature research, and it is generally shortened to a distributed autonomous organization. The core characteristics of DAOs include two points: distributed and autonomous. In the context of most of the current theory and practice:

Distributed is generally understood as the participants are equal to each other in terms of rights and obligations, there is no authoritative node, and the decision-making process is based on irreversible blockchain technology;

Autonomy is generally understood as the process of participation among participants in the formation of consensus decisions is independent and free, and its implementation is done automatically through smart contracts.

Theoretically, the so-called DAO governance is the study of how to establish the consensus mechanism of the stake distribution relationship and the deployment of related smart contracts within the DAO. However, in practice, many DAOs still use offline governance and do not rely on smart contracts, the main reason for this is still technical, that is, the lack of full DAO Tool support. The main reasons why DAO Tool is out of touch with theoretical research are:

On the one hand, it is difficult for researchers to completely separate subjective consciousness from objective facts, so that they are easily influenced by personal values; on the other hand, due to the diversity, complexity and susceptibility of human behavior to external factors, it is difficult to explain related problems with simple causality, and it is impossible to make accurate predictions through experiments and observations like natural science.

Taking reputation or payroll management as an example, in a centralized system, the leader can simply set an objective standard and then strictly enforce it, but in a decentralized system, it is difficult to achieve consensus among all employees on the objective standard itself. Due to the cognitive differences of participants at the background level, different people may have different value evaluations for the same objective problem, and the reason for this objective cognitive conflict problem may be because of language, culture, The difference in customs (and the resulting differences in values, beliefs or positions, etc.) may also be due to differences in knowledge backgrounds (and resulting differences in information acquisition and processing methods), and in order to deal with cognitive conflicts between people and reach decision-making consensus and benefit distribution, collective decision-making based on voting is a basic means of processing.

Voting governance in a broad sense can generally be regarded as consisting of six steps: proposal, review, voting, execution, dispute, and arbitration. Among them, the proposal and review can be regarded as the pre-processing stage of voting, the voting processing process is the core of voting governance, and the execution can generally be completed by smart contracts, when some members are dissatisfied with the voting results or think that there is cheating in the voting process, it will be the link of dispute, and if the dispute is accepted, then it will enter the arbitration stage. There is a DAO tool that focuses on disputes and arbitration (such as decentralized court-like platforms). We believe that after the full development of smart contract technology, the above process can be reduced to two stages of proposal, voting, or even only one proposal voting (or voting) stage.

This article mainly discusses the issue of voting governance in the DAO Tool. As a consensus mechanism, the voting strategy tries to replace a bad decision with a good decision, or replace the decision that benefits the minority with a decision that benefits the majority, and its common models generally have the following ten forms: one person, one vote and the principle of majority, one currency, one vote system and the principle of majority, representative voting, and flow democratic voting. Quadratic Voting, RageQuitting, ConvictionVoting, HolographicConsensus, Weighted Voting and Reputation Voting, Knowledge Background Voting.

In traditional social practice, different voting governance models reflect the synthesis of multiple issues, including economic issues, cultural issues, social and institutional issues, etc. HOWEVER, FOR THE DAO ORGANIZATION, ITS BIRTH COINCIDED WITH THE VIGOROUS DEVELOPMENT OF WEB3, SO THERE IS NO SOCIAL SYSTEM OR ECONOMIC SYSTEM PROBLEM LEFT OVER FROM HISTORY, AND BEHIND DIFFERENT DECISION-MAKING MODELS, IT IS MORE REFLECTED IN CULTURAL BELIEFS AND MECHANISM CORRECTION PROBLEMS CAUSED BY DYNAMIC GAMES. Therefore, in the following article, we would like to discuss how to design relevant governance decision-making models from the perspective of rational people.

02 Related Questions and Business Scope

(1) Voting strategy and problem analysis

Among the voting models discussed above, the one-person-one-vote and majority principle are the simplest and easiest strategic models, but their disadvantages are also obvious:

(1) The strategy is unfair to the members of the organization because different members have different resources and expected benefits for the organization;

(2) this strategy cannot prevent relevant interest groups from buying votes to carry out malicious attacks based on the majority principle (or governance attacks, some DAO organizations have been disbanded because of such attacks);

(3) the strategy lacks appropriate incentives for voter participation, which can easily lead to governance apathy when voters pay more than they incentivize voting (which is why some organizations often fail to raise the legal minimum number of votes when holding referendums);

(4) The rights and interests of minority opinion holders are ignored.

In view of the shortcomings (1), a feasible improvement is the implementation of the principle of one coin, one vote and the majority system, although the improved strategy considers the economic fairness of the holders, but also discourages the voting enthusiasm of small holders. Although quadratic voting can reduce the influence of whale holders to a certain extent, there is still a lack of effective strategies to improve the voting enthusiasm of small holders, that is, it is difficult to solve the shortcomings (2), (3), and (4).

In response to disadvantage (2), a possible improvement would be the introduction of a representative system. The representative system shields malicious attacks based on the majority principle through the agent mechanism, improves the efficiency of decision-making, and makes collective decision-making more rational, but it also introduces a new problem: that is, it is impossible to ensure that the agent can protect the interests of all the agents, because the determination of the interests of the represented itself requires governance decisions. Therefore, there is a need to establish an incentive system between the agent and the represented. To deal with this problem, a possible improvement strategy is to implement mobile democratic voting, in which the delegate can choose to vote directly or re-delegate another proxy to vote when the agent is not satisfied with the vote of the agent (including indirect agents who have been delegated multiple times). However, mobile democratic voting still does not completely solve the problem of incentives for agents to protect the rights and interests of the represented, and similarly, it cannot solve problems (3) and (4).

Another strategy for disadvantage (2) is to use belief voting or weighted voting, which attempts to reduce governance attacks or effectively collect community preferences by increasing the holding cost of voters, but this solution obviously exacerbates the problem (3) and (4), and in addition to this, the introduction of reputation voting or knowledge token voting is not effective in solving the problem (4). AT THE SAME TIME, REPUTATION OR KNOWLEDGE ITSELF IS ALSO A VALUE JUDGMENT ISSUE THAT REQUIRES DECISION-MAKING, AND THERE IS A TENDENCY TO DEVIATE VOTING GOVERNANCE FROM THE RULE OF LAW TO THE RULE OF MAN, AND DEVIATE FROM THE DIRECTION OF WEB3 GOVERNANCE CONTRACTUALIZATION.

One strategy for disadvantage (3) is to use holographic voting, which attempts to incentivize participants to act in the interests of the majority by having members stake on proposals, creating a prediction market in parallel with the voting mechanism. However, the logic of bettors’ bets is whether a proposal will be passed, rather than whether a proposal should be passed, further distorts the market and exacerbates the problem (4).

One strategy for disadvantage (4) is to use angry retreat voting, which can theoretically ensure that no member can control the funds of other members and improve the ideological unity of the organization, but this inconsistent exit method is somewhat in conflict with the WEB3 development concept, nature is harmonious, and WEB3 organizations should also be able to accommodate different cultures and beliefs. When members disagree, it is only necessary to negotiate an appropriate compensation mechanism, and there is no need to withdraw directly from the organization.

The other two strategies for disadvantage (4) are the eden pattern and the CDao pattern. Both of these strategies are based on the strategy model of sharing the cake, and have not yet been officially opened to the public, and we will analyze their governance models separately later.

(2) Allocation strategy that satisfies everyone

When n participants share a “cake” C, there are multiple distribution schemes that make all individuals “satisfied”:

No jealousy distribution: A jealousy-free distribution is one if each participant believes that they have received no less than anyone else’s share;

Strong No Envy Distribution: If each participant thinks they are getting a greater share than anyone else, the distribution is a strong no envy distribution;

Fair distribution: A fair distribution is made if each participant believes that they have received a share that is not less than the average;

Strong fair distribution: If each participant thinks that they have received a share greater than the average, the distribution is a strong fair distribution;

Super-fair distribution: A strong fair distribution scheme constructed according to the super-fair distribution theorem, which is a strong fair distribution scheme obtained by N fair distribution schemes after neutral processing.

Taking the distribution of irregular cakes between two people as an example (the result is similar when there are more than two people, the process is complicated), and the details of the distribution plan are given to give several detailed differences:

Suppose the two participants P1 and P2 prefer the cake as follows: P1 prefers the red part and P2 prefers the black part, and the two possible distributions are as follows:

P1 cuts the cake, P2 chooses the cake, and the fair cutting point is assumed to be at A.

P2 cuts the cake, P1 chooses the cake, and the fair cutting point is assumed to be at B.

Obviously, A and B are the critical points of fair distribution or no jealousy distribution, and if you cut any of the middle points of A and B, the distribution scheme is a strong fair distribution and a non-jealous distribution. Super-fair distribution is a neutral treatment of two fair distributions, looking for the middle c, (the specific position of c depends on the preference strength of P1 and P2).

Eden is based on a jealousy-free distribution model, and its dispute resolution is as follows:

Within small groups, “both the prosecution and defense propose a dispute resolution, and then a randomly selected jury decides which is “fairer”.” This is similar to having one child divide the cake into two portions while the other child chooses which one of them wants. When the number of participants is large, Eden reaches a consensus through a small group (e.g. 3 to 5 people), elects a small group leader, and then recursively solves the consensus of the large group and forms a final decision.

CDao is based on a hyper-fair distribution model, and its dispute resolution method is as follows, taking co-creation NFT as an example:

A, B, C, D, and E created an NFT, and after the creation is completed, the five people plan to divide the NFT equally.

The relevant processing process is as follows:

(1) The platform conducts a time-limited voting valuation for five people, assuming that the five people evaluate the NFT at a price of 3 tokens, 5 tokens, 7 tokens, 8 tokens, and 9 tokens.

(2) The platform calculates the contract and gives the following processing results:

E pays 6.68 tokens to get the NFT.

A gets 1.12 tokens, B gets 1.52 tokens;

C gets compensated 1.92 tokens;D gets compensated 2.12 tokens;

(3) The hyper-fairness analysis of the relevant governance results is as follows:

In the subjective world of E: the NFT price is 9 tokens, and according to the principle of equality, it is reasonable to compensate others 4*(9/5)=7.2 tokens, and the actual payment is 6.68 tokens, and the benefits exceed the fair value: 7.2-6.68=0.52 tokens;

In the subjective world of A: the NFT price is 3 tokens, according to the principle of equality, it should be compensated 3/5 = 0.6 tokens, and the actual compensation is 1.12 tokens, and the benefits exceed the fair value: 1.12-0.6 = 0.52 tokens;

In B’s subjective world: the NFT price is 5 tokens, according to the principle of equality, it should be compensated 5/5=1 tokens, and the actual compensation is 1.52 tokens, and the benefit exceeds the fair value: 1.52 - 1=0.52 tokens;

In C’s subjective world: the NFT price is 7 tokens, according to the principle of equality, it should be compensated 7/5 = 1.4 tokens, and the actual compensation is 1.92 tokens, and the benefit exceeds the fair value: 1.92-1.4 = 0.52 tokens;

In D’s subjective world: the NFT price is 8 tokens, according to the principle of equality, you should get compensation 8/5 = 1.6 tokens, and you actually get compensation 2.12 tokens, and the benefits exceed the fair value: 2.12-1.6 = 0.52 tokens;

In summary, in the subjective world of A, B, C, D, and E, each person benefits more than 0.52 tokens of fair value.

The CDao believes that in addition to the equality of all people and the equality of votes, each bill (based on different cultural beliefs or modes of thinking) should also be equal, so when no one opposes a bill, it can be considered that no one’s rights and interests are harmed, because the rights and interests of the proposer are increased, and in this case, according to the Pareto improvement principle, the proposal should be automatically executed by the system. However, if an objection is raised, it indicates that the interests of some members will be harmed, and the rights and interests of the victim should be assessed.

Since all proposals are equal in terms of the rights and interests to be enforced, and in reality only one strategy is allowed to be executed, therefore, according to the above allocation strategy, the proposals that are not executed should be compensated. By having all members vote on the proposal, the amount of support for all proposals is the subjective value of the equity to be enforced: the voting tokens can be used as compensation for other proposals (when the proposal you support is accepted by the collective decision), or you can calculate the compensation you can get (when the proposal you support is not accepted by the collective decision).

Since a proposal must have the broadest possible support for it to be accepted, every member who supports the motion has an incentive to vote for the bill to pass, and similarly, every bill that is not accepted and its supporters will be compensated, which will also motivate community members to raise reasonable objections (which can improve the effectiveness of the proposal by setting the criteria for the proposal), and there is no voting compensation for members who do not participate in the vote.

Therefore, the relevant governance strategies not only increase the willingness of participants to participate, avoid the phenomenon of governance apathy, but also protect the rights and interests of minority opinions.

(3) Problem analysis of DAO organizations

For a DAO organization, different domains should be able to choose different voting governance strategies, however, as the DAO organization expands, the DAO organization will also have different business units like the company, and different business units should also adopt different voting governance strategies, so we should pay more attention to which voting types are adapted to different business units in the DAO organization than the domain type of DAO.

First of all, let’s take a look at the business that needs to be voted and governed by the DAO organization, generally speaking, the DAO organization has the following six types of business that need to be governed:

(1) Modify the rules of the DAO: For example, modify the DAO’s charter, governance protocol, voting mechanism, etc.

(2) Member management: such as deciding to admit new members, removing members, and setting the rights and obligations of members.

(3) Treasury fund management: such as deciding which projects to invest in, the amount to invest, the duration of the investment, token issuance, etc.

(4) Project profit distribution: how to distribute profits after the end of project investment;

(5) Community governance: such as deciding how to deal with community disputes, dealing with violations, formulating community rules, etc.

(6) Other decisions: such as deciding on the development direction, partners, brand image, etc. of the DAO.

Among the governance services in (5) above, there are about the following types:

The first category involves the core goals of the DAO:

This type of governance is the foundation of the DAO and must be passed by all members, such as (1);

The second category does not involve the core goals of the DAO, but involves the distribution of benefits:

This type of governance decision often affects the interests of some people, and usually the beneficiary belongs to a small group within the DAO, such as the project team, such as (3) and (4);

The third category, which does not involve core objectives and does not involve the current distribution of benefits:

This type of governance often belongs to a binding normative system, which is binding on all members after the establishment of the relevant system, and the beneficiaries belong to all members of the DAO, such as (2), (5), and (6)

As can be seen from the above discussion:

(1) For governance involving the core goals of the DAO, the more appropriate strategy is to adopt the angry retreat voting;

(2) For the governance of issues that do not involve core objectives and benefit distribution, the more appropriate strategy is to use weighting and reputation voting, but the difficulty here lies in how to reach a consensus on the reputation strategy model, and in this regard, the EDEN model may be referred to;

(3) For the governance of issues that do not involve the core goals of the DAO, but involve the distribution of benefits, the angry retreat voting is also a feasible strategy, but it is not conducive to the development of the DAO. A feasible strategy is to improve the operation mode of the DAO and allow the DAO to form a sub-DAO in the form of a fork, so that when there is a difference of opinion, the sub-DAO can avoid retreat, and at the same time, the sub-DAO can continue to fork to form the next level of sub-DAO. When the child DAO completes the task and is disbanded, it can return to the parent DAO. On the whole, for governance issues involving benefit distribution, sub-DAO and CDao strategy are relatively feasible strategies.

03 Current status of development in the field of voting tools

(1) Existing tool platform

In the field of DAO voting governance, some tool teams and platforms have occupied an important position in the market.

（1）Snapshot

Snapshot and SnapshotX are two products led by Balancer Labs, which are used for offline and online voting governance respectively, mainly relying on the Ethereum ecosystem, and currently occupy a large market share in the field of DAO voting governance.

Support a variety of voting systems: single selection, approval voting, proxy voting, secondary voting, etc.;

（2） Tally

Tally has a certain market share in the DAO voting governance space.

This is a frontend for governance contracts that provides voting governance services to the DAO team. Specific features include: proposals, voting, proxy delegations, allocation of DAO treasury funds, and smart contract upgrades. Support OpenZeppelin governance contract library, interfaces include: event signature, function signature, quorum setting, voting delay, voting cycle, etc.

（3）Paladin

Paladin seeks to turn voting power into an asset, working to improve decentralized governance from a neutral perspective. It focuses on aligning incentives to enable token holders to vote or elect representatives to participate in voting. Helping high-calibre community members build their reputation and influence.

（4） Sybil

Sybil is an Ethereum-based DAO voting governance platform led by the Sybil Labs team, which mainly provides governance tools for discovering, finding, and delegating representatives.

（5）Commonwealth

Commonwealth is a comprehensive platform for DAO voting governance that provides discussion, voting, and project funding services for the on-chain community. Help users achieve the unity of community activities and governance.

（6）Broadroom

Broadroom is a governance (both on-chain and off-chain) management portal, providing a common governance interface and SDK that provides DAO organizations with more than 350 cross-chain high-performance APIs, including services such as querying proposals, delegates, discussions, voting, and more.

(2) Problems and Analysis

In addition to some of the voting governance tools introduced above, Aragon, Stake DAO and other organizations also provide some DAO voting governance tools, however, from the perspective of relevant voting tool functions, most of the existing mainstream toolkits focus on the application of traditional governance strategies, or in other words, these tools do not involve the exploration of DAO governance strategies in nature, and they cannot be well solved for several important fundamental problems currently faced by DAO organizations in governance, including (security is not discussed here, Legal and other issues):

(1) Centralization of governance tokens: In DAO, how to ensure the best interests of DAO members when a small number of members control a large number of governance tokens;

(2) Governance apathy: In DAOs, many participants may not have the interest or time to actively participate in the decision-making process, which is essentially a problem of lack of incentives;

(3) Protection of minority rights and interests: In DAO, how to ensure the rights and interests of minorities is not as easy to deal with as in real life, because DAO governance is often programmed and automated, and the governance attacks involved are often related to the life and death of DAO organizations.

For the above fundamental issues, we still look forward to the development of more innovative tools (including eden and CDao), which will provide the most basic guarantee for the rapid development of DAOs.

(3) Selection of DAO organizations

For a DAO organization, how to choose a suitable voting tool platform needs to have a comprehensive consideration, generally speaking, the following aspects need to be focused on:

(1) Whether it supports the selection of governance policies and the setting of governance policy parameters;

(2) whether anonymous voting or identification is supported;

(3) whether to support proxy voting;

(4) whether to support weighted voting;

(5) whether the quadratic model is supported;

(6) whether to support the separation of voting rights and ownership (support for reputation voting, expert voting, or knowledge voting);

(7) whether to support voting incentives (to avoid governance apathy);

(8) Whether to support the bill to compensate (protect the rights and interests of minorities)…

Of course, in the above options, the weight of each function item also depends on the setting of the white paper or meta-rules within the organization, and sometimes a specific function even has a veto power, and this choice itself is a matter of cultural beliefs or opinions. When the existing tools cannot meet their own needs, independent development is also a feasible solution, and at the same time, using the API provided by the voting governance platform for secondary development is also an alternative (if the relevant platform can provide the corresponding underlying logic).

In addition, for domestic DAOs, there are also their own unique environmental problems:

First of all, due to the problem of centralization caused by the unsmooth network and related policies, in essence, it is a problem of institutional supervision, which requires the cooperation of all parties in society, and it is difficult to solve it only from a technical point of view.

Secondly, the formalism in the voting process is serious, the form is greater than the content, and many voters do not care about the content of the topic at all, resulting in the phenomenon of canvassing and swiping votes, which also deviates from the original intention of voting design. The reason for this is a matter of the design of voting strategies, which, in theory, give participants an incentive to actively cheat if they can bring benefits. In this regard, incentive-compatible governance strategies (as in the case of the pie-sharing model above) are necessary, and just as the fight against bribery is inevitable against corruption, ensuring transparency and fairness in power is the root cause (which is exactly the inaction governance strategy advocated by the CDAO).

(4) Ideal voting governance scheme

As can be seen from the above analysis, in order for a DAO to achieve the goals of distribution (which can be briefly described as equality or fairness) and autonomy (freedom), it must pay enough attention to two aspects: one is the consensus mechanism and the other is the distribution of rights. The former requires the establishment of a consensus mechanism that is incentively compatible, while the latter requires the implementation of a Pareto improvement strategy. Specifically, the following strategies are mandatory:

(1) It can support the governance token strategy and ensure that the members who invest a lot of resources in the organization obtain the corresponding ownership, which can be used as a reference to the corporate system.

(2) The relationship between governance and ownership should be straightened out, which should be different from the traditional corporate system, on the one hand, governance comes from ownership, and on the other hand, the separation of governance and ownership must be more favorable to the owner, that is, the owner can voluntarily choose to transfer or not transfer the governance right.

(3) In order to maintain the free choice of members and protect the rights and interests of all members, when a unique consensus cannot be formed, the DAO organization should be free to fork and destroy freely:

If there is a change in the treasury funds, when the sub-DAO is burned and returned, the treasury funds should also have a corresponding processing contract, and there should be a plan for incentives compatible with profit and loss projects, so as to avoid the sub-DAO from falsely reporting profits and losses.

(4) Delegated proxies should be the basic rights of members, and the system should support proxy voting:

First of all, when entrusting an agent, the agent should be able to freely distribute its tokens to different agents and can recover the undisposed part at any time, and when entrusting an agent, it should also be clear whether the transfer of agency rights is allowed;

Second, when there is a proposal to dispose of treasury funds, the smart contract can only dispose of the part owned by the agent who supports the proposal and its agent, and if the supported share is not the full share owned by the agent, it must be spent from its proxy share in proportion;

Finally, when a representative delegates authority, he or she should delegate a proportionate share of all principals, and a notice should be given to all principals who have not signed the delegation of authority.

(5) When proposing a bill, it is necessary to ensure that the incentive of the bill is compatible, which will help to avoid the phenomenon of governance apathy.

(6) When the bill is voted on, the Pareto improvement strategy should be implemented to protect the rights and interests of minorities.

04 Development Prospects

Although the concept and technology of DAO have begun to take shape, there are still many challenges to its practical application and development. These challenges can be broadly divided into three categories: social, technical, and ethical.

The main manifestation of social problems is the legal status and regulatory issues of DAOs, and it is predictable that DAOs will inevitably enter the legal and regulatory system like companies in the future;

In terms of technology, on the one hand, we expect that in the future, there will be a fair exchange mechanism for governance tokens between different DAOs that does not rely on third-party stablecoins, or a fair cross-chain transfer mechanism for tokens, and on the other hand, we look forward to the emergence of more advanced security technologies and more stable smart contract development methods.

On the ethical side, we expect two things to do: first, we want the governance strategy to be incentive-compatible for the bill itself, and second, we want the proposal to be voted on in accordance with the Pareto-improved strategy.

From the current relevant research, it seems that the above three aspects are likely to be solved in the next few years, and further, we can expect that the whole society may be able to operate as a large DAO in the future.

In general, as an emerging form of organization, the future development prospects of DAO are promising, however, new problems may also arise at any time, so the research and practice of DAO will still be a long-term process. With the development of technology, the governance process realized by using voting technology may become more and more standardized and modular, a good voting strategy will help to straighten out the governance mechanism, develop a logical and self-consistent DAO Tool, excellent DAO Tool can avoid governance vulnerability attacks, and assist in building a complete and stable smart contract, and the mutual cooperation of voting strategy, DAO tool, and smart contract will surely provide us with a fair, transparent, free and safe DAO society.