In what follows, we will briefly examine the merits of decentralization. We then make the assertion that most community-lead projects have not consistently decentralized their governance, and will spend the remainder of this post on an out-of-the-box legal setup towards full DAOfication.

Web2 Ate Its Own Children

First, the philosophical question: do we need decentralization?

In an early 2018 piece on Why Decentralization Matters, Andreessen Horowitz’ Chris Dixon makes the case that the centralized platforms of the second era of the internet, from the mid-2000s to the present, will ultimately eat their own children: users feel cheated, and “complements” (entrepreneurs, developers, and investors) become wary of building on top of centralized platforms.

Source: https://cdixon.org/2018/02/18/why-decentralization-matters

Decentralized projects avoid this by using open source code and by giving participants voice: on-chain and off-chain channels to engage with the project, agitate and exert influence for change “from within.”

This sense of participation should gradually win the loyalty of entrepreneurs and developers, leading to a far greater mobilization of talent, and ultimately result in centralized players — who can only get better at the rate at which their employees improve their products — getting outclassed.

In short: open-sourced and decentrally-governed projects will lead to more compelling products. Beyond immediate consumer demand, decentralized brains may have a better chance of finding new solutions for the challenges humanity is facing.

Fork off!

Alfred Hirschman’s Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States, written in 1970, describes the twin forces of voice and exit, which act in opposite directions within traditional firms and organizations. These same forces arguably push and pull even harder inside community-driven projects.

Voice is the critical centripetal force binding together dispersed members. In the context of blockchains, voice can be hardcoded by giving voting rights to nodes (or Keepers, block producers, etc.).

As a result, voice can act as a bulwark against the centrifugal forces of exit: a governance breakdown that leads participants who feel alienated to fork a new project off an existing one.

With no IP to steal, this option is always available — and is considered an acquired right by the community. However, forking can be damaging, as it forces loyalty choices on present stakeholders and is likely to scare away future contributors.

Like the Great Schism of the 11th Century, it will leave both the forked-off and newly forked churches diminished by lessening the chance of exponential growth of their respective networks.

The goal of decentralized governance should, therefore, be to prevent schisms by giving people a guaranteed voice without — technologically or otherwise — frustrating their sovereign right to exit.

Decentralize and rule

Many project call themselves “community lead” after they put onchain mechanisms in place that give the community in principle rights to voice.

However, there are 3 issues with this:

 1. Can’t be bothered

As with real-world voting, despite their right to vote, many people just can’t be bothered.

Obama won the Presidentials but lost the mid-terms because his base didn’t turn up.

Community-lead projects typically have very low particpation rates. Even the higher-loyalty projects experience that it’s just difficult to cajole people to vote.

Arguably, as loyalty grows, projects should naturally see the degree of participation in their on-chain governance increase. There is also ongoing experimentation with alternative voting models, e.g. quadratic voting.

However, until participation rates increase, governance is only decentralized in theory.

2. Voter concentration 

A second weakness is that participation is often incestuous.

To paraphrase Churchill, never in the history of blockchain was so much controlled for so many by so few.

A closer look at protocols with market caps in the billions shows that voting is typically a plutocratic affair, with voting blocks concentrated in few hands.

For instance, EOS has come under increased scrutiny on suspicion of collusion between its top block producers, most of whom are based in Asia and China in particular.

Even MakerDAO, the poster child of the Decentralized Finance movement, shows 50% of its voting authority is held by the top 50 holders its MKR governance token, of whom the MakerDAO team and the Maker company are the largest.

The concentration of MKR holdings (updated daily on mkrgov.science)

3. Legal stack 

This on-chain concentration of voting power is typically backed up by off-chain legal mechanisms that — either by sinister motive or as a result of legal naiveté — leaves all levers of power in the hands of the initial founders and a small group of leads and early backers.

Ask yourself of the projects you follow: What do you know about their off-chain legal stack? How many are fully transparent about their governance? Do they disclose their entity cascade, and who sits on their various boards? How much power remains legally in the hands of the founders, even in projects with foundations or trusts that made you as node operator or governance token holder a beneficiary?

From our analysis, we see that the off-chain governance design, i.e. the meta-rules regarding changes or updates to a platform’s operational rules, including how decisions are made in matters on-chain rules do not address, often allows a small group of project leads to act with complete impunity against the majority of their stakeholders.

Web3: Who Do I Call?

Part of this sense of impunity may come from the lack of a central contact.

As US Secretary of State Henry Kissinger once asked: “Who do I call if I want to call Europe?”: Who do you call if that smart contract doesn’t perform as it promises on the tin? There’s no helpline, no [email protected] email.

This “who do I call” test is also part of regulators’ assessment of a project’s native token and whether — or at what point in its lifecycle — it moves within the ambit of securities laws.

If the answer is, say, “Vitalik” in the case of Ethereum, ETH would still be stuck in its pre-network security token phase according to the SEC’s most recent analysis.

As it happens, the ownership of ETH and the protocol’s governance was perceived as “sufficiently decentralized” for ETH to morph into a post-network non-security. Holders no longer primarily rely on the “coordinated efforts of others” (which is broadly understood as the founder(s) or a core group of identifiable project leads), a critical prong of the SEC’s infamous Howey Test.

But who are we kidding? Are we truly no longer relying on the “coordinated efforts of others” if the offchain legal setup behind the onchain governance of many blockchain projects shows a degree of centralization that would make Xi Jinping blush?

To consistently decentralize, above and beyond hardcoded participation rights and transparency about how these rights can be excecised, blockchain projects have to methodically dismantle any vestiges of central power in their legal governance stack.

To DAO or not to DAO, that’s the question

In this last part of our post, we present our research on what we hope could become a largely jurisdiction-agnostic, reference DAO legal stack.

Overview

The brief for our research was to come up with a legal structure that aims to reconcile three objectives:
– Decentralization
– Transparency
– Accountability

The basic structure is not complex. The relationship between the entities, however, has to be carefully calibrated and adding extra dimensions, e.g., tax optimization, would act as complexifiers.

DAOfy your setup: A template for decentralized governance (Source: Otonomos).

1. The Core Component: A Trust or Foundation 

The core component of any community-lead structure is a trust or a foundation.

The trust or foundation’s main purpose of existence is to grow the network, rather than a direct profit motive.

To further the growth of the network, assets such as tokens can be pledged to the trust or a foundation, and grants can be made to anybody who helps the network grow, including its own operational subsidiaries.

The beneficiaries of the trust or foundation are the token holders (nodes, Keepers, Stakers, etc., i.e., anybody who has been give on-chain participation rights). The main economic benefit for beneficiaries comes from token appreciation and not from for-profit activities.

Solving the coordination problem

There are two main reasons why a trust or foundation is best placed to be the guardian and treasury of a decentralized project:

  1. It offers a transmission mechanism for the wishes of a distributed, indeterminate group of token holders and the real world, without forcing those token holders to become shareholders or partners in a company or partnership. By using the trust or foundation as their representative body, the community of token holders can coordinate their actions and appoint trust protectors or foundation guardians, who, in turn, can appoint directors in operational subsidiaries (see below). If they wished to appoint such directors directly, they would only be able to do so by becoming direct shareholders in such operational entity, which is impractical and undesirable.
  2. Trusts and foundations can make grants to operating entities it owns or third parties that help with the growth of the network. This granting mechanism is unique to trusts and foundations and offers much greater freedom compared with how funding would work if it were done out of a company.

Trust or foundation?

There are some significant legal differences between a trust and a foundation that we can skip over in this post, except perhaps that a trust is not a separate legal person under the law, while a foundation is.

The trust or foundation can be set up in a number of jurisdictions. It used to be the case that common law jurisdictions generally use trust structures and civil law countries foundations, but foundations are now also on offer in e.g. Cayman or Jersey.

Tax will be a factor: if it is anticipated that income from the operating entity (see below) will be paid out as a dividend to the trust or the foundation, a tax-optimized jurisdiction may make sense. Mind the optics though: a Seychelles Foundation is a cost-effective and ultra-flexible structure, but to some may reek a bit too offshore.

For instance, one of the earliest crypto projects, Dash, chose to set up their Dash Trust in New Zealand, which has well-defined trust laws.

The US, too, has useful foundations and trusts with long histories and established caselaw. Perpetual charitable trusts, in particular, work well for not-for-profit projects and are legal in all states. The law “smiles upon them”, with any assets pledged to them escaping estate and gift taxes and zero income tax on their earnings.

Blockchain dynasties

Generally, trusts and foundations may be the last sacrosanct space where private parties’ freedom to contract is religiously upheld by the Courts and not meddled with by the State.

The few restrictions laid down by precedent or statute relate mainly to who can be a beneficiary, how to avoid possible conflicts of interest between those beneficiaries and the managers, and a limit on the lifetime of the vehicle:

  1. Structurally, the most crucial part of the setup is that the trust or the foundation is established for the benefit of the network participants, be they master node operators, Keepers, Stakers or token holders generally. Given the liquidity of their token holdings, by their nature, beneficiaries cannot be named by name or appointed individually. Hence, the trust or foundation has to be set up with no particular beneficiaries, which is not possible everywhere. Furthermore, a DAOfied trust or foundation setup should not leave residual powers with its settlor or founder to alter who is a beneficiary. We suspect that most projects that have set up trusts or foundations may have kept these (and many more!) powers with the settlor/founder, which leaves the reins of control firmly in their hands.
  2. Secondly, managers of the trust (i.e., the trustee) or the director(s) of a foundation’s Board or Council should themselves not be beneficiaries. This removes a potential area of conflict of interest since the trustees or directors have a fiduciary duty to the beneficiaries, i.e., the network as a whole, and are therefore obliged to act solely in the network’s best interest. Unfortunately, we know of many projects that have populated the key governance seats to the same insiders, who are typically also the primary holders of native token. This creates a very concentrated power core at the heart of many self-proclaimed community-lead projects.
  3. Finally, the law may intervene with the lifetime of the trust or foundation. The idea is that the entity outlives its creators and initial owners. Some jurisdictions, however, may not allow irrevocable trusts. More generally, few projects — perhaps understandably given the youth of their creators — have given much thought to succession. One project that has, Dash, made sure its trust lives forever. Ironically, when a US trust is set up to survive its settlor, it’s called “dynastic,” so networks can be thought of as decentralized dynasties!

Who is guarding the guardians?

Trusts are managed by a trustee, under the instructions of a group of Trust Protectors. In the case of a foundation, the role of the trustee is taken up by a director or a number of directors on a Board (or “Council”) that acts on the instructions of the Foundation Guardians.

In the context of a DAOfied setup, we propose to refer to them collectively as Stewards/esses: They look after the trust or foundation on behalf of the beneficiaries and owe them a fiduciary duty.

It is this fiduciary duty that ultimately underpins the integrity of the trust or foundation setup: it creates a powerful legal obligation on the Stewards/esses to align the interests of the trust or foundation entity with the interests of the network.

Stewards/esses should only be appointed by the beneficiaries, and the rules about their appointment should be transparent and articulated. Here too, any residual powers with the trust settlor or foundation founder to appoint or remove protectors or guardians would signal a centralizing motive by the project leads.

Remember that the trust or foundation ultimately holds all the assets of the project, so it is important to get the election process of its protectors / guardians right since the beneficiaries do as such have no legal title to the project’s property.

How the Greeks did it: A Kleroterion randomized voting machine.

One smart contract approach we’d like to see applied in the context of the DAO is the randomized election of beneficiaries to take up the role of Stewards/esses.

This approach goes back to how the Ancient Greeks selected candidates for most offices of state from amongst its citizens, using a randomization device called a Kleroterion.

A smart contract version of the Kleroterion has been proposed by Kleros to build a platform for crowdsourced justice. In the context of a DAOfied trust or foundation, such a smart contract would randomly elect token holders to act as Steward/esses for a given period of say 1 or 2 years. Their tokens could be held in trust for them their tenure ends and they become beneficiaries again.

2. The Operating Entity

It is difficult for an ethereal project to be an economic agent in the real world without an entity that can hire staff, pay contractors and bills, raise money, or even lobby governments.

Typically, teams may already have a legacy entity that was used to kickstart their project. Others might have skipped incorporation entirely. The latter is dumb since lack of limited liabilility will expose everybody involved with their personal assets.

Two considerations are important to make sure the operational company fits within a decentralized governance structure:

  1. The operational entity has to be a 100% subsidiary or otherwise controlled by the projects’ trust or foundation. This will make it possible for the trust or foundation’s protectors/guardians, themselves appointed by the network, to in turn appoint the Board of Directors of the operational company. This right by its majority shareholder to hire and fire the Board will guarantee that the operational entity executes the network’s consensus on how it uses the funds it receives by way of grants.
  2. Furthermore, the operational entity has to be able to receive grants, which is how it will cover its operations. Some vehicles such as non-profits are limited in how much they can receive without identifying their donors, so a standard limited company in a jurisdiction with good crypto infrastructure is recommended.

Taxation could be a tertiary consideration, especially if the operating company has revenue. However, it typically acts as a pure cost centre, with the grants it receives cancelling out the expenses it incurs.

If there is profit in sight, an LLC in the US may be a better structure since dividends will pass through to the trust or foundation as its owner.

A C-Corp could be better suited for projects seeking to attract additional funding from a traditional investor base such as VCs. They would then be shareholders in the C-Corp alongside the trust or foundation. This in itself should not detract from the DAOfied nature of the setup, lest they ask for a Board seat!

Further permutations

There are endless permutations in solving the equation for a decentralized, transparent and accountable off-chain governance structure for a DAO.

However, we believe a trust or foundation, of which a project’s token holders are beneficiaries, controlling an operational subsidiary and its Board, will be constants in any permutation.

It will give the community voice while giving token holders the legal freedom to easily exit a project off-chain, and should ultimately lead to higher loyalty.

By Way of Conclusion: And The Winner Is…

Making specific recommendations is difficult without knowing an individual project’s needs, goals, and sensitivities.

A simple setup could be a US trust, for instance, in Nevada, which allows for dynastic trusts that last for up to 365 years and has no state or corporate income tax, with a US LLC e.g., in Wyoming as its operating entity.

This could work for US-based project but arguably even better for non-US teams (see our previous post on LLCs).

However, for clients who perceive the US as high-risk, there are offshore foundations that can be DAOfied by tweaking their constitution. At Otonomos, we set up Foundations mainly in Cayman, which has an excellent Foundation Company structure, but have also worked in other jurisdictions, both to setup clients’ trusts and foundations.

Finally, the “combo” deal of a trust/foundation plus the operational entity is relatively cheap in its setup. The challenge may be to find independent officials with sufficient arms-length from the initial team.

Bonus! An Observation about Code, Law and Gödel

As a bonus, a final observation for those readers who wonder why they cannot stay entirely onchain, far detached from the messiness of on offchain legal setup and the carnal sins of a body corporate.

First, immaculately conceived as they may be, blockchain projects involve coordination between many actors. There is a real risk that in the absence of a formal legal setup, a project would be seen as an unlimited partnership which for many reasons beyond this blog post — personal liability being one of them — is entirely undesirable.

Second, even if projects would go unicorporated — isn’t code law after all? — decentralization only works because it rests on fundamental legal premises outside of the code. Like Gödel’s incompleteness theorem, no onchain governance can claim to be entirely self-sufficient without drawing in concepts of natural law, legal precedent and even acts of legislature. If this is the case, it is probably preferable to base your DAO on articulated agreements rather than relying on the vagaries of the law.

by Han, Founder & CEO Otonomos.com

DISCLAIMER. This blog post is not intended to provide legal, tax or other professional advice and should not be relied upon as such.