Seven Voting Mechanisms in Daos
Written by MIDDLE.X @ Paka Labs
Reviewed by Owen @ Paka Labs; Shawn Lin @ 1PAR Research
In most DAOs, tokens carry a role similar to that of shares in a joint-stock company, including ownership, governance decision-making, usufructs, and so on. For a long time, the main form of voting rights in DAOs was one token one vote (1T1V). However, it has many drawbacks, as users with a small percentage of token holdings adopt a “rational indifference” attitude towards participation in governance, as to refuse to be involved in, further leads the voting power in DAOs to be centralized and serves the interests of the few. Even for DAOs with a very decentralized token holding form and a good participation culture, the lack of a healthy voting mechanism will lead to a “rabble” of short-sightedness and stupidity.
The crypto industry rewards those who make even small innovations. With DAO, we are exploring various coordination mechanisms for governance. DAOs have already contributed a wealth of solutions to governance issues in just a few years of their birth, and the lower trial and error cost of DAO promotes continuous exploration with different governance mechanisms.
DAOs are a fascinating form of organization that has unparalleled advantages in achieving ownership-in-common of public goods, eliminating the risk of corruption in the voluntary sectors, and achieving nonprofit goals somewhere. However, these advantages can only be realized if they are coupled with sound decision-making mechanisms, fair contribution measurement standards, and reward mechanisms, and strong governance scalability. In this paper, we present what we believe are typical and interesting voting mechanisms that have positive implications for governance ideas and initiatives, but they are not the silver bullet.
1. Vote delegation/Liquid democracy
Every decision takes up the bandwidth of our minds, a better alternative is that delegating voting power to the few (experts), not ourselves, and this could make a more effective and efficient decision-making process in public governance. But the few are not always reliable, so oversight is essential in the principal-agent relationship.
Power needs to be centralized, but not always controlled by the same group, needs to be mobile. Liquid Democracy is similar to Representative Democracy as a whole but more mobile. Voters with choose to delegate their voting power to someone they trust and consider as a professional, but they can also withdraw their delegation to exercise the voting power in person or transfer it to someone else at any time. Furthermore, the delegation is multi-layer, the votes you have delegated could be then redelegated to others.
Liquid Democracy was first adopted in Aragon, which increases the participation rate of token holders, thus improving the legitimacy of voting decisions. Although it still suffers from problems such as vote-buying and collusion, Liquid Democracy increases the election right of each individual and enhances the supervisory role of the delegation parties by a more responsive and multi-layered proxy mechanism.
2. Quadratic Voting
This is a voting mechanism proposed by Vitalik Buterlin, which is between 1T1V (one token one vote) and 1P1V (one person one vote). The key function of this voting mechanism is to allow individual voters to make multi-voting for the same option in order to express the strong will, at the same time, the marginal cost of repeated votes for the same option tends to decrease ao as to avoid monopolization of voice by giant whales, which means the cost of the n+1th vote is higher than that of the nth vote. For example, for the same option, 1 vote 1 token, 2 votes 4 tokens, 3 votes 9 tokens, and so on, 10 votes 100 tokens. The utility of voting is the square root of the number of tokens needed.
Quadratic voting can be used to vote on the allocation of public resources, as exemplified by Gitcoin, a Grant DAO that uses quadratic voting to decide which projects to be funded by Ether Foundation. Before that, it was a centralized committee that decided which projects the Ether Foundation would sponsor. Gitcoin provides an avenue for community users to express their opinions. Community users could vote by “donating” to the projects they supported, and the sum of the square root of each donation a project received ultimately determined the amount of sponsorship is received.
Gitcoin’s quadratic voting mechanism eliminates the need for a centralized committee evaluation process, making the distribution of funds more efficient while creating a multiple accommodated culture for community engagement. It has effectively funded many high-quality Ethereum projects.
The weakness of quadratic voting is that it relies on a strict identity proofing mechanism to ensure its fairness, and it would operate no differently than 1T1V if falsified identities were allowed. Gitcoin tying social accounts, connect DID services, and many other approaches for identification. Users can earn a certain number of points for each item they complete and achieve an identity after reaching 150 points.
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 a DAO that is being extended. We know that as a DAO scales, the cost of coordination increases dramatically, and it is impractical to have everyone vote for every proposal, and participants’ attention becomes the scarcest resource inevitably.
DAOstack argues that an efficient governance system must be accompanied by a mechanism to manage collective attention that ensures the most important proposals receive community attention and the small group of participants who join vote tend to act in the interests of the majority. The term “holographic” originally means a technical that records all information about a three-dimensional object on a two-dimensional plane, so as make sense of the Holographic Consensus, it allows the few people’s decisions to accurately express the will of the majority. This is achieved by the attention token, GEN.
GEN is the marker for monetizing attention in the DAOstack ecosystem. With GEN, you cannot vote but can bet on any proposal, and if the proposal you bet on is approved, you will gain more GEN, and vice-versa, if not, you lose GEN. This betting way creates a prediction market parallel to the voting mechanism.
The governance process of Holographic Consensus can be divided into four steps.
(1) Proposal initiation: Any user who meets the reputation threshold could initiate a proposal.
(2) Proposal enhancement: GEN holders select proposals that they believe with a high probability of approving and bet on them; proposals that receive insufficient GEN bets will be ignored and will not proceed to the next stage.
(3) Voting decision: the group with voting rights votes on the proposal, and if the proposal is approved, the user who bets on it will be rewarded with GEN, and loses GEN if it is not.
(4) Execution: The approved proposal take effect and is executed on the chain.
Promises are the governance voting framework of Holographic Consensus by DAOstack, which has strong governance scalability and provides a practical governance scheme for larger DAOs.
They’re focused on two main objections to the framework.
- Does it filter the proposals that deserve the most attention, or does it just filter popular proposals with propagation effects?
- Do bettors, whose judgment is based on whether a proposal will be approved rather than whether a proposal should be approved, and who are bound to participate in the voting with voting eligibility, end up distorting the outcome?
These defects have yet to be tested and fixed in practice. In any case, DAOstack has provided a useful practice of combining prediction markets and DAO governance. DXdao, NecDAO, and Prime DAO have implemented Holographic Consensus by using DAOstack framework.
4. Conviction Voting
Conviction Voting is a dynamic voting mechanism based on voter beliefs and proposed by Aragon. The voting utility of the votes will increase over time but the process is decelerated not uniform until reaching a maximum value and no longer increasing. Votes can be withdrawn or moved to another proposal by the users anytime, but the voting utility on the previous proposal will disappear gradually not immediately and this disappear process is an accelerated process. In order to make the formulation more descriptive, Aragon creates the concept of “conviction” to refer to voting utility.
Conviction Voting is a novel decision-making process that has been used by the 1HIVE community for community decision-making on funding allocation. Conviction Voting has the following features.
- The user could allocate his votes among multiple ongoing proposals at any time, with no clear deadlines
- The voting utility is not only related to the number of votes cast but also adds a time function that grows over time.
- The user can withdraw his votes at any time, and its voting utility is not immediately removed, but gradually diminishes over time
- Each proposal has a “conviction” threshold based on the amount of funding applied, and once the “conviction” gathered by the proposal meets the threshold, the proposal and the funds will be approved and disbursed respectively.
Conviction Voting fundamentally changes the format of voting, as community users will not be asked to vote within a limited time, nor will they be asked to vote for a proposal they do not understand. Instead of always having to “decide”, users can fully express their preferences and vote for proposals they know about and are interested in. Rather than requiring users to reach a majority consensus on the same issue, Conviction Voting instead takes advantage of the diversity of user will and allows multiple paths to run in parallel.
Of course, it is not applicable to all decision scenarios but very suitable for budgetary decision-making. It allows organizations to effectively collect community preferences and respond to group will.
5. Rage Quitting
Rage Quitting comes from Moloch and is now widely used in several DAO platforms or DAO organizations that use the Moloch framework, including DAOhaus.
Theoretically, there is a risk that an organization that relies on a majority vote to decide funds arrangement, such as a small group of people who control 70% voting power with the ability to vote to approve a proposal, this proposal would embezzle funds bElong to the other 30% voting power owner. Although such an extreme situation has not yet arisen, it is common in joint-stock companies that larger shareholders use their decision-making power and information advantage to reap the interests of minority shareholders. For Venture DAOs, it is necessary to prevent small groups with decision-making power from undermining the interests of other owners. A rage quitting mechanism can do this effectively
For a DAO using the Moloch framework, anyone could quit from the DAO at any time, destroy his or her Share or Loot (Share is a voting share, Loot is a non-voting share), and withdraw the funds corresponding to the share/loot. While the term “rage quitting” specifically refers to the quit act in the voting process.
In the case of DAOhaus, the governance process has the following steps.
- Submitting a proposal: Anyone (who is not a DAO organization member) could submit a proposal.
- Sponsoring a proposal: The proposal must receive sufficient sponsorship to proceed to the voting stage. Sponsorship means that the Shareholders vote in favor of the proposal. This stage filters out frivolous or unimportant proposals.
- Queuing: After reaching the sponsorship threshold, the proposal is in the queue and awaits voting.
- Voting: sufficient support votes to pass is needed before the deadline.
- Grace period: after the proposal is approved, there is a seven-day grace period before executing the proposal, shareholders could do rage quitting during this time.
- Execution: The proposal is marked as completed and is executed on the chain.
By rage quitting, no one can control the funds of others, it protects the interests of members and improves the unity of the organization in terms of ideas, and increases the efficiency of organizational coordination.
6. Weighted Voting & Reputation-based Voting
Governance Token often has to perform a dual function, with a governance function and a value circulation function. The token must be sufficiently liquid in order to capture the financial value and this necessarily financialize the token in some sense. This means that governance tokens may be found in trading markets, lending markets, and may even generate derivative assets.
An attacker can borrow from the lending market, or rent proxies from the possible proxy bribery market, and can obtain a large portion of vote rights for a short time and launch a governance attack on a DAO.
Weighted voting is a good way to circumvent such problems, Some DAOs make the voting utility linked to the holding length(token age), the longer the holding time, the more voting power, and some DAOs, make the voting utility linked to the lockup length, such as Polkadot: users with the choice not to lockup but the voting utility is extremely small, or, take a longer lockup length can increase the voting utility.
Weighted voting increases the financial cost of governance attacks.
A more radical solution would be to completely decouple the financial value token and governance token, the governance ability of a token we can call ‘reputation’.
Reputation, a non-transferable, non-circulating credit, could be earned by holding/locking tokens or by contributing to the organization, and the voting power is accompanied by reputation rather than a token. It is important to note that the reputation you had earned can be destroyed through code rules or governance. For example, reputation can be deducted when a reputation holder commits an act that is detrimental to the organization’s interests, or it can be deducted over time or expired in order to avoid the continued influence of an early reputation that may undermine fairness.
While reputation is not completely resistant to malicious bribery (you can still sell your private key to indirectly transfer reputation, for example), but it is very difficult to form an efficient market for buying and selling reputation, or to financialize it.
Starting with DAOstack, Reputation Voting has been accepted gradually by others. Reputation voting gives DAO organizations the ability to adjust voting weights based on community ecological distribution and token distribution and avoids the governance attacks and fairness issues associated with token-based voting. By customizing the rules for calculating and obtaining reputation values, DAO organizations can practice what they understand and recognize as “democracy”.
7. Knowledge-extractable Voting (KEV)
It needs to know the concept “Furtachy”, a governance theory proposed by Robin Hanson in 2000, a model that combines governance voting with prediction markets, which core idea is to give the act of voting a betting nature, rewarding those who vote for the right option and punishing those who votes for the wrong option. The holographic consensus mentioned above can be understood as a variant of Furtachy.
This section introduces another variant of Furtachy: Knowledge-extractable Voting, or KEV for short.
KEV is about giving knowledgeable ones more power to vote. It introduces a new type of knowledge token, which is somewhat similar to reputation token, is non-tradable and can be issued and forfeited through established rules, but it couldn’t be used for voting directly, but works by influencing voting weights.
In the KEV mechanism, proposals are divided into different topics, and different topics are associated with different knowledge tokens, and having a certain type of knowledge token gives you more voting power in the proposals of that type of topic. If Alice has many tax tokens, Alice’s votes will have more weight when she votes on tax proposals. If Alice votes in accordance with the final voting result, Alice will be rewarded with more knowledge tokens corresponding to the topic, otherwise, will be deducted. The KEV mechanism encourages people with more expertise on a proposal to vote and who do not have enough information and knowledge on a proposal not to vote.
The idea of “knowledge influencing power” of KEV is certainly positive, but whether an “expert’s” choice is correct or not depends on the final outcome of the vote, it is not the voting decision itself has a positive impact, which is not outside of Furtachy’s framework.
KEV has currently been used only by Dit protocol, and the reasonable setting of KEV-related parameters has been explored.
What is a good governance voting mechanism? For private companies, efficiency is probably the first priority, and corporate leaders with absolute power and a rigid hierarchy are the standard, but for DAO organizations, although they also need community leaders with appeal and efficient execution, we think “democracy” should be considered first.
The democracy in the DAO context should be accurately described as decision-making is generated by a wide range of stakeholders who are broadly and fairly involved in, which can help DAO organizations (1) involving as many as possible members to work together to create the value for the community rather than “hitchhiking”, so as to effectively explore group wisdom and resources to serve the organization’s goals. It will take much more costs for a traditional centralized organization. And (2) DAOs are almost the ideal form of organization for operating public goods in some sense, to avoid the detriment of interests if the most of the others in return for the benefit of shareholders, even with the ability to create more public goods. One might ask, won’t DAO owners contradict the interests of other stakeholders, but it is better to question that how can DAOs convert a broader range of stakeholders into owners?
The development of DAOs not only overcomes endogenous alienation and promotes a better achievement of organizational goals but also generates an exploration of the governance mechanism that should be found in some DaaS service platforms like Aragon, Moloch, DAOstack, SubDAO.
The DAO is devoted to reducing the cost of organizational coordination through technology. Quadratic voting, for example, is not only used in DAOs, Colorado Democrats also experimented with it. This is a new attempt to reform the social governance methods, which could provide important references for the decision-making of public issues. Moreover, the idea of monetizing attention in the holographic consensus, the knowledge amplifying power in KEV, the voting weight pluralism in belief voting, and the fairness concept of the rage quitting mechanism and reputation voting are all important to us.
 Vitalik Buterin：Moving beyond coin voting governance
 Eric Arsenault：Voting Options in DAOs
 DAOrayaki ：DAO民主投票系列：民主投票机制综述
 火币研究院： DAO哪里了 — — 简析DAO的发展进程及治理机制
 Aragon：Choice Architecture & DAOs
 Vitalik Buterin：Ethereum Community Governance and Gitcoin Quadratic Funding
 Josh Zemel：An Explanation of DAOstack in Fairly Simple Terms
 Conviction Voting TL;DR
 DAOhans documents
 Adam Levi：Reputation vs Tokens
 Marvin Kruse, Sebastian Gajek, and Yannik Goldgräbe：The dit Protocol: Knowledge over Wealth to Sustain Open Source Development
 Nathan Schneider：Startups Need a New Option: Exit to Community
PAKA is a starter’s fund in the Polkadot ecosystem, a DAO Venture co-founded by a group of parachain initiators, aiming to discover and help innovative teams in the Polkadot ecosystem. We hope to bring our experience in entrepreneurship and technology to the Polkadot ecosystem and help the next generation of entrepreneurs through the form of DAO while promoting the vision of Web3.0.