Vitalik: The traditional electoral system is prone to vote splitting, and the quadratic voting method can improve democracy

Author: BlockTempo

Ethereum co-founder Vitalik Buterin participated in the Tempo X Accelerator's Plurality Forum in Taipei on January 19, titled "From Sand to Islands: How Taiwan's Experience of Plural Democracy Advances into the Web3 Era?" During the forum, he explored various voting systems, from traditional voting to quadratic voting, discussing their theoretical and practical applications, how these systems work in different contexts, and their impact on the democratic decision-making process.

The Importance of Voting Mechanisms in Various Fields

Before delving into major voting systems, Vitalik first discussed the diversity of voting systems and their applications in different fields. He mentioned that people often associate voting with national or city elections, but in reality, the voting process spans various scales and contexts. For example, in addition to government elections, there are opinion polls and voting within non-profit organizations. He further emphasized that while opinion polls are theoretically non-binding, their results have a significant impact on discourse and culture.

Next, Vitalik turned to the concept of "micro-democracy" on social media platforms. Using tweets as an example, he explained how when people post content on different platforms (such as X, Farkaster, and Mastodon), likes and shares from other users influence public perception of that content. He believes these interactions are essentially "millions of referendums" happening every day, determining whether a viewpoint deserves broader attention.

Limitations and Drawbacks of Traditional Voting Systems

When discussing the limitations and drawbacks of current traditional voting systems, Vitalik posed a fundamental question: Why is simply voting for A or B not good enough? He illustrated this with a simple example where nine voters support different candidates: A receives four votes, B gets three, and C gets two. In this case, although A appears to win, A is not the most popular option among the public.

Vitalik elaborates on the drawbacks of traditional voting

Vitalik further analyzed the preferences of these voters, explaining that even if A wins the vote, it does not mean he is the first choice of the majority. He pointed out that if a large portion of voters strongly oppose A, and their votes are split between B and C, this could lead to A being mistakenly perceived as the most popular choice.

To clarify, Vitalik referenced "Duverger's Law," explaining why such simple voting systems often result in only two major parties. He cited the United States as a clear example, where the voting system tends to evolve into a competition between two major parties.

The Spoiler Effect Extended from Duverger's Law

From the perspective of Duverger's Law, Vitalik explained the reasons why small parties struggle to succeed in the current political system. He noted that voters often believe the chances of a small party candidate winning are slim because they have never won in the past. Therefore, even if voters are very fond of these small party candidates, they may choose to vote for the major party candidates who are more likely to win.

He pointed out that this mindset leads voters to typically choose between the two major candidates, further entrenching the dominance of the two parties and making it difficult for other candidates to enter the democratic system, known as the "spoiler effect."

Exploring the spoiler effect under Duverger's Law, Vitalik concluded that the result of such electoral systems is that even if neither of the major party candidates is ideal, voters will still vote for the one they perceive as "less bad," making it very difficult to hold stable elections with more than two candidates.

Advantages and Disadvantages of Ranked Choice Voting

Regarding Ranked Choice Voting, Vitalik explained that this system allows each voter to express their overall preference order for candidates, from most to least liked. The counting process involves multiple rounds of elimination, with the candidate receiving the fewest votes eliminated each round until only one candidate remains.

Vitalik used an example to illustrate how this voting method can address certain issues in traditional voting systems. In his example, when candidates A, B, and C compete, Ranked Choice Voting can more accurately reflect voter preferences, ultimately allowing the candidate genuinely supported by the majority to win. However, he also pointed out that the downside of this voting method is its complexity, which may lead to intuitively incorrect results in some cases.

Introducing a case of Ranked Choice Voting

Approval Voting as a Simpler Form

Next, Vitalik explained another voting method: Approval Voting. In Approval Voting, voters can vote for as many candidates as they like, including one, two, three, or even none.

To better understand how this voting method works, Vitalik provided an example: Suppose four people like candidate A, while five others strongly dislike A but have varying degrees of preference for candidates B and C. In this case, the four supporters of A would vote for A, while the five opponents of A would choose to support B and C. This results in B and C each receiving five votes, tying for victory.

Vitalik noted that in real-life scenarios, due to the large number of participants, slight differences in vote counts are likely to occur, ultimately leading to one candidate winning. He emphasized that Approval Voting can yield meaningful results and is much simpler than more complex voting methods like Ranked Choice Voting.

Vitalik explains Approval Voting

The Dilemma Presented by Arrow's Theorem

Afterward, Vitalik discussed Arrow's Theorem and its implications for voting systems. He pointed out that Arrow's Theorem simply states one problem: in any voting scenario with at least three candidates, all voting mechanisms may yield clearly incorrect results in certain cases. This is often due to a violation of the so-called "independence of irrelevant alternatives" principle, meaning that introducing a new candidate C could change the outcome between A and B, which intuitively seems unfair.

Vitalik then explained that Arrow's Theorem suggests that no matter how a voting system is designed, this situation cannot be avoided. However, he noted that Arrow's Theorem has an important assumption: ordinal preference, which means that the voting system can consider whether you prefer A over B but cannot consider how much more you prefer A over B.

In reality, Vitalik explained that as long as a voting system begins to allow for consideration of the differences in voter preferences for candidates, it can circumvent the dilemmas posed by Arrow's Theorem. He mentioned that Approval Voting is an effective method because it recognizes the necessity of considering such degrees of difference. Finally, he mentioned Quadratic Voting, a more complex voting system that allows voters to allocate their preferences based on a fixed number of votes.

Considering the dilemmas encountered by the aforementioned voting mechanisms, Vitalik explained the mathematical logic of Quadratic Voting, where the cost of each vote is proportional to the square of the number of votes cast. This characteristic requires participants to think more carefully about their choices and prevents the overall election results from being manipulated through a large number of low-value votes. This helps to reduce the impact of extreme voting behavior, making the final results more representative and fair.

Vitalik mentioned the practical applications of Quadratic Voting, such as in Gitcoin grants' quadratic funding pools and various DAOs. He believes this voting mechanism can be applied not only in the cryptocurrency space but also in various community and decision-making scenarios.

Finally, Vitalik emphasized the importance of practical experience and encouraged the community to actively participate in and experiment with various voting mechanisms. He believes this will help better understand how voting mechanisms work and improve their design, thereby providing communities with fairer and more representative decision-making methods.

Conclusion Q&A

As the forum drew to a close, Ethereum founder Vitalik particularly emphasized the value of Quadratic Voting, but he also believed that in all voting systems, community participation is extremely crucial, alongside mechanism design. He encouraged experimentation and improvement to achieve fairer and more representative decision-making methods.

Vitalik believes that voting mechanisms can be applied in various ways, which is why people are interested in democracy and politics, and why those focused on cryptocurrency and Web3 find themselves in the same room as political activists, as both groups care about similar issues and face the same challenges.

Regarding democratic voting mechanisms, many attendees at Tempo X eagerly posed questions to Vitalik.

Jimmy's Question

Q: In the voting systems implemented in different communities and crypto ecosystems, I would like to know if there is one that you think is relatively well done, and if so, is there an evaluation framework that can be used to assess these different governance and voting systems?

A: For example, the Optimism Public Incubator Fund is a unique approach that allows people to provide an ideal number and then choose the median. This method differs from other voting mechanisms discussed earlier, but I think they can map to each other to some extent.

Additionally, I believe each decentralized autonomous organization (DAO) has its unique way of voting on proposals, reflecting the broad diversity of voting mechanisms. I also want to remind not to overemphasize the voting mechanism itself; while the voting mechanism is important, what’s more critical is the "Communication Structure" surrounding these mechanisms, which I believe accounts for about 75% of the decision-making process, while the voting mechanism itself only accounts for 25%.

In terms of voting in Optimism, for instance, I support the agent system because it allows people to explain in advance why they vote a certain way. This way, representatives can create lists explaining their voting decisions, and other representatives can choose to follow these lists. This structure exists not just above the voting mechanism; it actually enhances the quality of the mechanism.

In many decentralized autonomous organizations (DAOs), when a mechanism needs to be voted on, members not only vote but also participate in governance-related forums. This is very important to me because they provide a way to understand and engage. Although these governance-related communication structures and mechanisms are difficult to describe with mathematical models, they play a crucial role in the governance process.

Hong Zhijie’s Question

Q: I am curious about the cheating mechanisms in Quadratic Voting (QV), particularly regarding how to avoid or identify such cheating. I understand that under the Quadratic Voting system, if someone wants to acquire 100 votes, they need to spend 10,000 points.

However, I am concerned that if this person finds another way to obtain the same number of votes for only 1,000 points, then this cost is significantly lower than the 10,000 points required to directly purchase 100 votes. Such cheating would not only be unfair to the system but could also harm every participant, especially when others are unaware of it. My question is, how do we identify and avoid such cheating in this system?

A: To address the collusion issue in Quadratic Voting, we can technically make cheating more difficult, similar to Macy's method, but the challenge is that the public nature of individual voting information could be abused, as seen in Gitcoin funding, where people use this information for retrospective airdrops, undermining the entire mechanism.

We also face the issue of protecting individual identity security and need to consider that technical solutions may not be entirely perfect. Therefore, we also need to approach this from the perspective of mechanism design, creating better incentive structures. For example, by giving more voting weight to those with differing opinions on other issues, we can limit the influence of colluders controlling a large number of accounts. So I believe a combination of these two strategies is valuable.

Chen Jingfang’s Question (Partner at Mingfu International Law Firm)

Q: Yes, I have a question about the new voting methods requiring constitutional amendments, which need congressional approval. However, Congress is usually elected through old methods. The existing system is unlikely to choose a voting system that goes against its own interests, so is there a chance to break this cycle?

A: Yes, I believe it does depend on the specific context. For example, in the U.S. electoral context that I am more concerned about, we see how it collapses into a two-party system. In this case, we can discuss whether they would allow the existence of a third party or not.

Even on this issue, I think the incentives may be more open than people imagine. Even within the Republican and Democratic parties, they are not single entities but complex groups with various interests, which absolutely includes those who might want to see some form of a third party exist.

So I believe that incentives are very complex in any system, and I agree that this is one of the main reasons for the entrenchment of the political system. But sometimes, the world may be more complex than it appears, even in good ways. So sometimes, change can happen, you know.