HOW PROTOCOL DAOs SHOULD WORK FROM A CRYPTOLAW(-ISH) PERSPECTIVE

Chain Governance vs Social Governance & Why They Must be Handled Very Differently

TLDR

• DAO tokens only control code—on chain, where code is law

• DAO tokens don’t control people (coders, validators, liquidation bot runners, businesses, etc.), but can be used for sentiment signaling—off-chain, where law is law.

• Developers and others cannot respond to on-chain stakeholders (DAO token holders etc) as if they were off-chain stakeholders (stockholders etc.), because doing so would cause massive legal risk to the developers as well as DAO participants, users and others who depend on the protocol. 

• Developers must keep any major work they are doing on the protocol quiet to avoid becoming legally obligated to token holders and making the entire protocol, DAO and all participants subject to onerous financial regulations.

Protocol DAOs’ Main Purpose 

• The main purpose of protocol DAOs (and the ‘governance tokens’ comprising them) is to give users of autonomous digital infrastructure—aka smart contract systems—a voice in whether and how to change whatever features of those systems might be mutable. This occurs by enabling direct, binding, on-chain control of those systems by the requisite voting majority of the DAO token holders. In the MakerDAO community, these are called ‘executive votes’. 

• In effect, protocol DAOs are just massively multiplayer online games for adjusting the parameters of ownerless, decentralized software systems. This is also why DAO tokens are distributed (through liquidity mining, etc.) primarily to users of those systems—the users need a strong voice in the systems they rely upon. 

• For on-chain governance, DAO tokens and DAO voting are ‘God mode’—they are the first, final and only authority, functioning under a principle of ‘code is law’. Whatever quorum and majority rules are written into the code must be strictly abided by.

Protocol DAOs’ Secondary Purposes 

• A secondary purpose of protocol DAOs is to engage in off-chain, social coordination based on rough social consensus regarding issues more loosely related to the relevant autonomous system. This includes chats, ‘governance’ forums, twitter—the whole gamut of social media platforms on which the community around that system communicates. Within this context, DAO token holders might sometimes hold votes about various things they think should happen on the social layer–such as writing a new major code upgrade, or the community adopting certain social goals. MakerDAO refers to these as ‘signaling votes’ and they include, for example, expressions of alignment on social values, such as “green” environmental tech financing initiatives. 

• These votes are non-binding—they signal sentiment.  No person, group or business is required to ‘follow’ the results of these votes, and the votes also do not need to adhere to strict requirements of quorum or specific majority standards. They are just expressing community sentiment on a topic. 

• Moreover, these votes do not even represent the sentiments of the complete community—only the sentiments of the people who hold governance tokens. Since the full community around a protocol is usually broader—including bot runners, developers, and even validators/miners on the L1, etc.—governance tokens cannot be the sole authority on these social topics. They are just one input into a larger process of “rough social consensus,” albeit it is an important one.

Confusing the Main and Secondary Purposes is Legally Disastrous

• Unfortunately, many people either confuse the primary and secondary purpose, or wish that the secondary purpose worked differently—i.e., they wish that governance token holders held binding authority over certain people on the social layer. There are many reasons why this cannot work and why it is a dangerous and bad idea. I will focus on the legal reasons, though there are non-legal reasons as well. 

• Assets that give their owners binding social voting powers are legal contracts—in most cases, highly regulated “securities”. Corporate stock is regulated in large part because it carries specific legal rights, including the right to appoint the directors of the corporation. Specific legal obligations are owed to stockholders by the elected fiduciaries. For example, directors of a corporation have a fiduciary duty to make their corporate decisions with one goal in mind–maximizing stockholder value. They can be sued if they do not follow this goal. 

• Because DAO tokens are not regulated, treating them similarly to corporate stock or other securities would violate many financial regulations. Worse, unlike directors of a corporation, the purported fiduciaries in this case would have no insurance, no protections and unlimited liability—a very bad outcome for software developers and others involved with a protocol. As a result, everyone involved in DAOs must be very careful about how they handle and respond to sentiment polling. 

Example - Code Upgrades and How to Handle Them 

• Let’s take an example: 

  • A signaling poll passes supporting a certain software upgrade to the protocol, but no one has yet coded that upgrade, it is only described conceptually, and it is not even clear if it will work. It will take significant time, resources and talent to code, test and deploy that upgrade, and even if people try to code it, it is possible it could completely fail and never be adopted.

• In the US, the Howey test says that if some group is relying on the entrepreneurial efforts of another for the value of an asset, then the asset might be a security. Thus, a team which announces it will spend the next six months building an upgrade will be construed under the law as essentially making a promise to token holders to do the work they requested to be done via the sentiment poll, and such team will be turning the DAO token into something more like corporate stock–a security. This is so even if the team is a new one rather than the team that originally built the protocol–the SEC refers to these new teams as “active participants” who can be liable under the securities laws. Not only are such activities risky for the development team, but, as we have seen in the CFTC’s recent case against Ooki DAO, they are bad for the DAO itself, as every participant in the DAO can then be accused of running an off-chain business enterprise.

• What should a development team do in this example, if they want to build the socially supported idea for a code upgrade? In the immortal words of the Notorious B.I.G, “bad boys move in silence and violence.”

  • In the new normal of intense legal risk for DAOs and DeFi, development teams must learn to work secretly until the work is completed and can be submitted to a binding on-chain vote. From an outsider’s perspective, it will be impossible to tell whether they are working on something or not, unless and until it is done. This achieves two important objectives: 

    • reduces legal risk for the team and everyone else in the community ; and

    • avoids speculative pump-and-dumps where the value of a potential, experimental software upgrade becomes built into the token price and where, thus, there will be a price crash if it turns out this upgrade is not or cannot be achieved.

Limiting Formal Governance to On-Chain Issues is Not so Bad, & is Common in Practice

• Although the above dynamics might be frustrating, this is why smart contract systems must be open source. That way, anyone can work on the protocol and if they submit specific actual code to be placed on the blockchain, there can be a binding governance vote about that. But, for social proposals, governance votes just express sentiments and do not guarantee any particular outcome. 

• Many successful protocol communities take this attitude—for example, Ethereum is wildly successful despite having no formal protocol governance, proceeding by rough social consensus, and depending on “core developers” who do not respond to votes of ETH holders, but rather make their own decisions about what to code and how to code it. Bitcoin is the same. DeFi communities are admittedly somewhat inconsistent in this regard, but, for example, the “Yearn manifesto” explicitly states that “Yearn is governed by YFI, but YFI does not govern Yearn’s contributors”. And we have seen that protocol communities which view things differently are getting punished by governments (eg in the CFTC’s lawsuit vs Ooki DAO). 

Conclusion

We have entered a new normal of massive legal risk to anyone involved with freedom technologies—similar to the risks faced by cryptographic technology developers in the 1990s “crypto wars”. This requires everyone involved in DeFi—from casual users, to hardcore degens, to coders, to bot-runners, to validators, to CEXs, to social media leaders—to tighten up their game, be more conscious of what they are doing, how they are doing it, and what its legal consequences might be. In the long run, this is also best for everyone, as it forces us to think about what this technology and the related social formations—DAOs—are really supposed to achieve and how it is best for them to achieve those objectives without falling back into old TradFi models like corporate governance.