DAOs: How Fair can Decision-Making be and Why is Private Voting Essential?

What are DAOs, what are their challenges and how are they influencing decision-making in Web3? How can a voting system be fairer? What’s the role of privacy in voting? These are some of the questions we’re discussing below, exploring concepts like identity, quadratic voting, and Sybil resistance.

What are DAOs?

Well, let’s start from the beginning. What’s a DAO? A Decentralized Autonomous Organization (DAO) allows multiple people and entities to coordinate on matters which can be enforced by a blockchain, such as the management of digital asset portfolios. In a DAO, there is no central authority in power. The rules of the DAO define how decisions are taken by its members.

As DAOs on public blockchains span beyond national borders and jurisdictions, nowadays they are often used for regulatory arbitrage. However, there are examples of DAOs that chose to fit legal frameworks such as the Swiss association, DAA.

Today, DAOs are almost exclusively used to invest or spend tokens with joint decision-making. They can cover virtually anything, from a jazz foundation to a European-based institution linked to delivering. These are real cases, by the way: Jazz DAO is a Lisbon-based project built on the blockchain protocol NEAR. Their main task is to educate people on jazz and organize performances in the city. Voting on different matters is done on-chain and everything is public.

The European Union is also invested in deploying blockchain for a series of matters via the Sandbox project. Several institutions linked to the EU are using smart contracts to facilitate supply chain management, and decide on, fund, and send relief support to areas in need, for example. Although several governments and leaders — like the Finnish Communications Minister, Timo Harakka — have been asking for it, the EU hasn’t yet created a legal framework regarding DAOs.

The US Government Accountability Office (GAO) has also shown interest in experimenting with DAOs. However, GAO has some concerns regarding blockchain-based projects, arguing they can be prejudicial due to lack of regulation: “(…) blockchain-based financial applications can facilitate illicit activity, may reduce consumer and investor protections compared to traditional finance, and, in some cases, are subject to unclear and complex rules.” — voting, organizational structure, and ID management are some of the areas where GAO feels these organizations can be useful. You can read the full report on the subject here. The paper is a great read, as it sheds light on the benefits and downsides of DAOs.

One Token, One Vote

The most common form of decision-making in a DAO is token voting: The weight of your vote corresponds to the number of tokens you commit to a DAO. The obvious advantage of token voting is that anyone can be allowed to join such a DAO and their voting power will correspond to their “skin in the game”. While such a plutocratic design may be suitable for an investment fund, it is a very bad choice if governance participation and democracy is the goal.

Vitalik Buterin sheds some light on the issue in this article. The subject has been heavily discussed and several new ideas for different voting systems are rising, but traditional coin voting is still the most common one. Buterin mentions proof of personhood, proof of participation, and quadratic voting as possible avenues to be implemented in DAOs.

Combining Decentralized Identity & Quadratic voting

Moving beyond pure token voting isn’t trivial. If egalitarian decision-making is the goal, then a democratic one-person-one-vote would be the most desirable solution. But there are very tricky problems to solve. First, we need to make sure that every person can only vote once: The governance system must be Sybil-resilient. Secondly, there has to be a way to define who is allowed to vote and who isn’t — it is not legitimate to let every human vote on matters that only affect a few. Such a system would likely fall prey to vote-buying or hostile takeover and in turn, give more power to those who can afford to buy votes. Whitelisting voters, on the other hand, is inevitably an act of centralization and undesirable for a DAO.

A middle ground between plutocracy and democracy would be quadratic voting: every human gets a vote which is weighted by the square root of tokens they commit to the vote. Therefore, if you want to double your voting power, you’d need to commit 4x the number of tokens. The advantages of quadratic voting are that it shifts power from capital to personhood but it still requires voters to have “skin in the game” (be affected by decisions). It is important to note, however, that quadratic voting can only work with Sybil-resilience — otherwise, a whale token holder could split their capital into many accounts to gain more voting power.

Enter Personhood Verification

One of the most common Sybil resistance systems is identity verification. However, there are several issues regarding the latter, the first being the fact that this verification usually involves third-party organizations or nation-states, centralizing the process — which isn’t favorable in the case of Web3 governance as one of its core values is decentralization.

One of the established DAOs using quadratic voting is gitcoin which focuses on funding common good open source software and uses gitcoin passport for Sybil-resilience: an aggregator of various profiles which need some effort to maintain: Twitter, Facebook, GitHub, BrightId, and others. Encointer is an even more decentralized protocol for Sybil-resilience based on in-person mutual attestation which doesn’t rely on identity verification at all. Combining quadratic voting with a Sybil-resistance logic might be one way to establish a more democratic voting system in DAOs.

Private Voting

However, we would still be struggling with the issue of privacy — or lack thereof. Whether or not you use token-voting or quadratic voting, your vote is linkable to a pseudonym (your voting wallet address) and therefore your voting behavior may sooner or later be associated with your identity.

Why is privacy a paramount aspect in the act of voting and decision-making?

If voting is not private, people tend to vote how they think others expect them to. That leads to very undesirable behavior of conformity, obedience and submission. In contrast, what I’d like to see in our ecosystem is: creativity, exploration and dissent.

Alain Brenzikofer | Integritee’s CTO in Private Voting on Referenda

There are more problems with public voting: If you can prove to another party how you voted without any doubt (if your voting system isn’t receipt-free), that opens the door to various forms of coercion and vote-buying.

Unfortunately, Sybil-resilience protocols often expose even more privacy risks. But that doesn’t have to be the case.

(…) Solving it [Sybil attacks], need not and should not force us to give up our privacy.

Bryan Ford | Decentralized/Distributed Systems (DEDIS) lab, Swiss Federal Institute of Technology

Various projects are already tackling the privacy issue on public blockchains, and Integritee offers a solution with unique properties. Integritee can bring the voting process of any DAO from a public and transparent L1 blockchain to a confidential L2 sidechain, where any form of voting can be performed privately and receipt-free. Privacy is protected by trusted execution environments which can be generically programmed to apply various voting schemes and Sybil-resilience systems. Thanks to Integritee’s public auditability features, the process is entirely transparent as anyone can verify that the correct code has been executed on a genuine TEE — check more info here.

Accountability vs Privacy

The challenges don’t end here, however. Accountability is important for a governance system as well. If a person or entity has a substantial voting weight in a DAO, be it through capital or delegation, it is in the interest of a DAO that they can be kept accountable for their voting behavior. Otherwise, they can manipulate the DAO without the risk of being exposed.

Fortunately, TEEs offer a solution here as well, because we don’t need to decide between absolute privacy or none at all. With Integritee, a DAO could be designed such that voters with a weight above a specified threshold are exposed publicly (or to other voters only). Other privacy technologies, like cryptographic zero-knowledge proofs — which reveal nothing but the fact that a certain statement is true — don’t offer this flexibility.

There will probably never be one preferred voting system that pleases everyone, but that doesn’t mean we shouldn’t keep working on building a fairer system that harnesses the benefits of different structures and helps balance the scale.