In most DAOs, Tokens carry a role similar to that of stocks for joint-stock companies, including ownership, governance decision-making rights, income rights, and so on. For a long time, the main form of voting rights of DAO is the simplest form of one token, one vote (1T1V). However, this arrangement has many drawbacks. Users with a small proportion of currency holdings take a “rational indifference” attitude towards participating in governance, just like a saying: “If you say it, you don’t say it”; this further leads to the centralization of voting rights in the DAO. , and instead serve the interests of the minority; even for a DAO organization with a sufficiently decentralized currency holding structure and a good participation culture, if there is a lack of a benign voting decision-making mechanism, it will lead to shortsightedness and stupidity of the “rabble crowd”.
The crypto industry will reward those who make even the smallest innovations generously. With DAO as the carrier, people began to explore various governance coordination mechanisms. Due to the programmability of blockchain and Token, DAO was born in just a few years, and has already contributed rich solutions to governance issues, and the low trial and error cost of DAO organization also provides an experimental stage for different governance mechanisms.
DAO is a charismatic organizational form that has unparalleled advantages in realizing common ownership of public goods, eliminating corruption risks in the public welfare sector, and achieving non-profit organizational goals . However, these advantages can only be realized with a benign voting decision-making mechanism, a fair contribution measurement and reward mechanism, and strong governance scalability. This article will introduce some typical and interesting voting decision-making mechanisms that we think are typical. Although these voting mechanisms are not silver bullets to solve all problems, their governance ideas have extraordinary positive significance.
1. Vote delegation/Liquid democracy
Every decision needs to occupy our thinking bandwidth, which is one of the reasons why we choose “rational indifference” to public issues in reality, but we have a better choice, that is to entrust decision-making to experts, which is better than taking on these matters ourselves. The cost is lower. In addition, in public governance, entrusting power to a few people can effectively improve efficiency. But authorization is not omnipotent, and experts are not necessarily completely reliable. Therefore, in the principal-agent relationship, supervision is indispensable.
Power needs to be centralized, but it needs to be more fluid. Liquid democracy is similar to representative systems overall, but has a stronger “liquid” attribute than representative systems as a whole. Users with voting rights can choose to delegate their voting rights to people they trust and consider professionals, but they can withdraw their delegation at any time and exercise their voting rights in person, or transfer the delegation to others. In addition, delegation is multi-level, and the votes you delegated may be delegated to others again.
The liquid democracy mechanism was first adopted in Aragon. In this way, the participation rate of token holders can be increased, thereby improving the legitimacy of voting decisions. Although there are still problems such as vote bribery and collusion in liquid democracy. However, through a more responsive and more hierarchical delegation mechanism, liquid democracy increases everyone’s right to choose, and also significantly increases the supervisory role of the delegator over the delegatee.
2. Quadrac Voting
This is a voting mechanism between 1T1V and 1P1V (one person, one vote) proposed by Vitalik Buterlin. The key role of this voting mechanism is to allow a single voting subject to repeatedly vote for the same option to express the strength of his or her will. , at the same time, in order to avoid giant whales monopolizing the right to speak, the marginal cost of repeated voting for the same option shows a decreasing trend, which means that the cost of the n+1th vote is higher than the nth vote. For example, one vote for the same option will consume 1 Token, 2 votes will consume 4 Tokens, 3 votes will consume 9 Tokens, and so on, 10 votes will consume 100 Tokens. Token. As the number of repeated votes for an option increases, the utility of the vote is only the square root of the number of tokens voted.
Quadratic voting can be used for rights holders to vote on the allocation of public resources, a typical application case is Gitcoin. Gitcoin is a Grant DAO that uses a quadratic voting mechanism to decide which projects to fund with funds from the Ethereum Foundation. Before Gitcoin, which projects the Ethereum Foundation sponsored was decided by a centralized committee. Gitcoin provides a channel for community users to express their opinions. Community users can vote in the form of “donations” to the projects they support, and the sum of the square roots of each donation obtained by a project ultimately determines the amount of sponsorship that can be obtained.
In practice, Gitcoin has efficiently funded many high-quality Ethereum projects. The quadratic voting mechanism allows projects that are supported by more people to stand out. Gitcoin’s quadratic voting mechanism eliminates the evaluation process of the centralized committee, making the allocation of funds more efficient, while creating an open culture of community participation .
The weakness of quadratic voting is that it must rely on a strict identity proof mechanism to ensure its fairness. If the fake identity is allowed, its operation result will be no different from 1T1V. In terms of identification, Gitcoin adopts diversified methods such as binding social accounts and connecting DID services. Each time a user completes an item, they can get a certain number of points, and after the points reach 150 points, they can achieve identity authentication.
Gitcoin’s authentication interface
3. Holographic Consensus
Holographic consensus is a solution provided by DAOstack to increase governance scalability. Governance scalability can be understood as the adaptability of a governance mechanism to the expansion of the DAO scale. We know that with the expansion of the DAO scale, the cost of coordination increases dramatically, and it is obviously unrealistic for everyone to vote for every proposal, and the attention of participants inevitably becomes the most scarce resource.
DAOstack believes that for a governance system to be efficient, there must be a mechanism for managing collective attention that ensures that the most important proposals get attention, and that small groups that get voted tend to act in the interests of the majority. The original meaning of the word “holographic” refers to the technical means of recording all the information of three-dimensional objects on a two-dimensional plane, and the goal of holographic consensus is to allow small groups to accurately express the will of the public. This is achieved through the Attention Token – GEN.
GEN is the token for the monetization of attention in the DAOstack ecosystem. With GEN, you can’t vote, but you can bet on any proposal, if the proposal you bet on passes, you will get more GEN, and if the proposal you bet on doesn’t pass, you will lose GEN. This way of betting is equivalent to establishing a prediction market in parallel with the voting mechanism.
The governance process of holographic consensus can be divided into four steps:
- Initiate a proposal: any user who meets the reputation threshold can initiate a proposal;
- Proposal enhancements: GEN holders select proposals that they think have a high probability of passing to bet. Proposals that do not receive enough GEN bets will be ignored and will not enter the next stage;
- Voting decision: Groups with voting rights vote on the proposal. If the proposal is passed, the betting user can get GEN reward, otherwise, GEN will be lost;
- On-chain execution: The approved proposal is officially effective and executed on the chain.
The above is the governance voting framework of holographic consensus provided by DAOstack, which has strong governance scalability and provides usable governance solutions for large-scale DAOs.
Questions about the framework are mainly focused on two points: First, does the framework really screen out the most noteworthy proposals, or just screen out popular proposals with a spreading effect? Second, the judgment of bettors is based on whether a proposal will be passed, not whether a proposal should be passed. Bettors with voting qualifications will inevitably participate in the voting. Will they eventually distort the voting results?
These drawbacks need to be tested and corrected in actual operation. In any case, DAOstack has provided a useful practice of combining prediction markets and DAO governance. By using the framework of DAOstack, DXdao, NecDAO and Prime DAO are already practicing holographic consensus mechanism.
4. Conviction Voting
Belief voting is a dynamic voting mechanism based on voter beliefs proposed by Aragon. When users choose to vote for a proposal, their voting utility increases over time. This increase process is not uniform, but decelerates, and will eventually reach a maximum value and no longer increase. Users can withdraw or transfer their votes to other proposals at any time, but their voting utility on the previous proposal will not disappear immediately, but will gradually decrease. This decreasing process is not uniform, but accelerated. To make the presentation more figurative, Aragon created the concept of “belief” to refer to voting utility.
Conviction voting is a novel decision-making process that is now used by the 1HIVE community for community decision-making in funding distribution. Faith voting has the following characteristics:
- A user can distribute his votes among multiple in-progress proposals at any time with no clear deadline;
- The voting utility is not only related to the number of votes, but also increases the time function, which will gradually increase over time;
- Users can withdraw their votes at any time, and their voting utility is not removed immediately, but gradually diminishes over time;
- Each proposal will have a threshold based on the amount of funding it applies for, and once the “belief” gathered by the proposal reaches the threshold, the proposal will be approved and funds will be disbursed.
Conviction voting fundamentally changes the form of voting, community users will not be required to vote under a time limit, nor will they be required to vote for proposals they do not understand. Users can fully express their preferences, instead of always “making decisions”, they can vote for proposals that they know about and are interested in. Conviction voting does not require users to reach a majority consensus on the same issue, but instead gives full play to the diversity of users’ will and allows multiple paths in parallel.
Of course, belief voting is not applicable to all decision-making scenarios. From the current practice, belief voting is very suitable for budget decision-making. It effectively lowers the threshold of voting, and allows organizations to effectively collect community preferences and respond sensitively to the will of the group.
5. Rage Quitting
We think that a better translation is “if you don’t feel comfortable”, this mechanism is derived from Moloch and is now widely used in many DAO platforms or DAO organizations that use the Moloch framework, including DAOhaus.
In theory, an organization that relies on a majority vote to decide the disposition of funds is at risk. For example, an owner who holds 70% of the voting rights votes for a proposal and embezzles the funds of the other 30% of the voting rights. Although such an extreme situation has not yet occurred, it is not uncommon for major shareholders to take advantage of decision-making power and information to reap the interests of small shareholders in joint-stock companies. For an investment DAO (Venture DAO), it is very necessary to prevent a small group with decision-making power from harming the interests of other owners, and the anger retreat mechanism can effectively achieve this.
For the DAO of the Moloch framework, any member can withdraw from the DAO organization at any time, destroy his own Share or Loot (Share is a voting share, Loot is a non-voting share), and retrieve the corresponding share of funds in the DAO . Anger withdrawal refers specifically to the withdrawal behavior in the governance voting process.
Taking DAOhaus as an example, the governance process is divided into the following steps:
- Submit a proposal: Anyone (not limited to DAO members) can submit a proposal;
- Sponsored Proposals: Proposals must receive sufficient sponsorship to enter the voting stage. The meaning of sponsorship is that those who hold Share vote to express support for this proposal. At this stage, meaningless or unimportant proposals can be filtered;
- Queue: After the proposal has received more than the threshold of sponsorship, it enters the queue and waits for voting. Through the queuing mechanism, to ensure that proposals are collected into the voting pool in an orderly manner;
- Voting: By the voting deadline, a proposal must receive enough affirmative votes to pass;
- Buffer period: After the voting is passed, there is a 7-day grace period (Grace Period) before the voting results are executed. During this period, shareholders who are dissatisfied with the voting results can withdraw in anger;
- Execution: The proposal is marked as complete and executed on-chain.
We found that under the anger withdrawal mechanism, no member can control the funds of other members, and theoretically, the interests of any member cannot be harmed through governance voting. In fact, the anger retreat mechanism can not only protect the interests of members, but also improve the unity of the organization in thought and improve the efficiency of organizational coordination.
6. Weighted Voting & Reputation-based Voting
Governance tokens often have dual functions, on the one hand, they have to undertake the function of governance, and on the other hand, they have to undertake the function of value circulation. Tokens must have sufficient liquidity to capture financial value. However, the financial value of the governance token will inevitably bring about the tendency of financialization . This means that governance tokens may not only appear in the trading market, but also in the lending market, and may even generate derivative assets.
Attackers can borrow from the lending market, or rent delegates from the possible delegate bribery market, and can gain a large number of voting rights in a short period of time to launch governance attacks on DAO.
In order to avoid such problems, weighted voting is a good way. Some DAO organizations link the voting utility to the holding time (coin age). The lock-up duration is linked, such as Polkadot: users can choose not to lock the position, but the voting utility is extremely small. Choosing a longer lock-up period can improve the voting utility.
Governance voting interface for Kusama parachain Bifrost
Weighted voting can increase the financial cost of a governance attack, making the attack less economically credible. More thorough approach is the financial value of pass through certificates and certificate authority control completely decoupled , which we might call “reputation.”
Reputation is a non-transferable, non-negotiable point. You may gain reputation by holding or locking tokens, but it is reputation, not tokens, that has voting rights. Reputation can also be earned by contributing to an organization. It should be noted that reputation holders do not have absolute ownership of reputation. The reputation you have acquired can be destroyed through code rules or governance voting. For example, when a reputation holder acts in a way that harms the interests of the organization, its reputation may be deducted; for another example, in order to avoid the continued influence of the reputation acquired in the early stage and undermine fairness, the reputation can be gradually deducted over time. deduction, or lapse due to expiration.
Although reputation is not completely resistant to malicious bribery (for example, you can still sell your private key to indirectly transfer reputation), it is undoubtedly extremely difficult to form an efficient reputation buying and selling market, or to financialize it.
Starting with DAOstack, other governance frameworks have also supported the scheme of reputation voting. Reputation voting gives DAO the ability to adjust the voting weight based on its community ecological distribution and token distribution, and also avoids governance attacks and fairness issues brought about by token-based voting. By customizing the calculation rules and obtaining methods of reputation value, DAO organizations can practice the “democracy” that they understand and recognize.
7. Knowledge-extractable Voting (KEV)
We need to understand a concept called “Furtachy” first, because its idea of ”betting on belief” has a profound impact on the DAO governance voting mechanism. Futarchy is a governance theory proposed by Robin Hanson in 2000, which is a model that combines governance voting with prediction markets. The core idea is to give voting behavior the quality of betting, thus rewarding those who voted for the correct option and punishing those who voted for the wrong option. The holographic consensus mentioned above can be understood as a variant of Furtachy.
This section will introduce another variant of Furtachy: Knowledge Extractable Voting (the translation is a literal translation of Knowledge-extractable Voting, which has not yet formed an industry consensus), referred to as KEV.
The core of KEV is to give knowledgeable experts more voting power. KEV’s thinking stems from a reflection on populism in realpolitik. In the Brexit referendum, there are actually many “insighted people” who have a deep understanding of international politics and trade. They generally do not support Brexit, but they do not have more voting rights. Like the public, they only have One vote of equal weight. In order to improve this situation, the KEV mechanism introduces a new type of knowledge token, which is somewhat similar in nature to reputation, cannot be traded, and can be issued and confiscated through established rules. But the token is not used directly for voting, but by influencing the voting weight.
In the KEV mechanism, proposals will be divided into different topics, and different topics will correspond to different knowledge tokens. Possessing a certain type of knowledge token can have greater voting rights in the proposal of this type of topic. If Alice has knowledge tokens on many tax law topics, Alice’s votes will have greater weight when voting on tax law proposals.
If Alice’s vote matches the final voting result, Alice will be rewarded with more knowledge tokens corresponding to the topic. On the contrary, if Alice’s vote is inconsistent with the final voting result, her knowledge tokens will be deducted. The KEV mechanism encourages people who have more professional knowledge of a proposal to vote, and also encourages people who do not have enough information and knowledge about a proposal not to vote.
The idea of ”knowledge influences power” of the KEV mechanism is undoubtedly positive, but in the final analysis, to judge whether the expert’s choice is correct, it still depends on the final voting result itself, not whether the voting decision itself has a positive impact . Out of Furtachy’s frame.
The KEV mechanism is currently only being adopted by the dit protocol, and the reasonable setting of KEV-related parameters is also being explored.
summary
What is a good governance voting mechanism? For private companies, efficiency may come first, and corporate leaders with absolute power and strict bureaucracy are the standard. Before that, I think “democracy” should be considered first.
As a research team in a non-Western cultural background, we do not take democracy as an unthinking, natural and correct value pursuit . The reason why DAO organizations need to choose democracy as their governance ethics is in line with practical interests. Let us first define democracy in the context of DAOs: broad and fair participation in decision-making by many stakeholders. A broad and fair decision-making mechanism can help DAO organizations achieve the following two goals:
1. Mobilize more wisdom and resources
A group can be expressed as “group unconsciousness” or “group wisdom”, depending on whether there is a good coordination mechanism. A well-functioning and well-functioning DAO organization can allow many owners to choose to create value instead of “free-riding” , so as to effectively explore group wisdom and group resources and serve the organization’s goals. For traditional centralized organizations, I am afraid that this can be done at a higher cost.
2. Develop towards “goodness” and persist for a long time
Since the birth of private ownership, it has provided a huge impetus to the development of human civilization, but with the development of productivity, many organizations that provide public goods have emerged. Whether it is public utilities such as hospitals and schools, or Internet services such as Uber and Facebook, they all have the nature of public goods. However, when private companies operate public services, they will unilaterally pursue the interests of shareholders and harm other broader stakeholders, such as excessive charges by private hospitals, and big data killing by Internet companies. Humans have explored many mechanisms to solve such problems, such as trade union participation in governance, state ownership, cooperatives, Steward Ownership, etc. However, based on the current situation, we find that DAO is the perfect answer . DAO is almost the most ideal organizational form for operating public goods, and even DAO may create more public goods . One might ask, won’t the DAO’s owners conflict with the interests of other stakeholders, but I think the question should be thought of the other way around: How does a DAO transform a wider range of stakeholders into owners?
The development of DAO is not only an organizational form itself, it can overcome endogeneity, make organizational goals better achieved, and generate good externalities. One of them is the innovation and exploration of governance mechanisms. The power of innovation and exploration comes from newly created DAO organizations, but more importantly, it comes from DaaS service platforms such as Aragon, Moloch, DAOstack, and SubDAO.
The design of the governance mechanism of DAO has gone beyond the technical boundaries of the blockchain, but returned to thinking about the origin of human organizational forms. We prefer to believe that the fundamental purpose of DAO is to reduce the cost of organizational coordination through technology. For example, quadratic voting is not only used in DAOs – the Colorado Democrats are also experimenting with quadratic voting to determine their program, which is not only a small improvement to the voting mechanism, but also a new social governance technology. A reform attempt, which will provide an important reference for decision-making on public issues; the idea of monetization of attention in holographic consensus, the philosophy of knowledge amplifying power in KEV, the diversified core of belief voting, the anger and retreat mechanism and the fairness of reputation voting All ideas have profound inspiration for our reality governance.