Introduction
A DAO is different to a traditional organisation, because power is distributed far more widely and in radically different ways.
Traditional organisations are typically pyramids of power. Impactful decisions on the direction and function of the organisation are shielded from the majority of participants.
A central idea of DAOs is to empower all of its participants to have a meaningful say in how they’re run.
That’s the idea anyway.
In practice, they are often very far from this ideal. In fact, they can demonstrate features of centralisation that even traditional organisations couldn’t muster.
In this article I will provide a framework to identifying common sources of power in DAOs and in future posts I’ll make some design suggestions for how that power can be more effectively distributed.
Power
Power is the ability to get things done and getting things done is what organisations do.
So we want our organisation to be powerful, but we don’t want the structures of power within it to be unduly skewed towards specific actors.
Most DAOs will start with healthy degrees of power concentration, and that’s fine. It is simply impossible for a DAO to pop into existence in its final form with an established engaged and empowered community.
So, a DAO’s core job is to diffuse its power across its network of participants and token holders over time. This is both a social layer leadership problem and a protocol layer technological problem.
On both fronts, we are incredibly early and I’m not convinced we’ve seen any shining examples of what a DAO can truly become yet.
What we have most certainly seen is a number of public disasters and it’s a core goal of this article to try and steer DAO designers towards better DAO practices and to inform DAO participants on key issues to look out for.
A principle heuristic for understanding DAO design is asking the question:
Who’s in Charge Around Here?
If you ask this question in a DAO, what do you think the answer should be?
“No one!” You might be shouting at your computer screen.
This is a common misconception, the answer ideally is everyone. Maybe.
Remember, if no one has power, nothing is getting done.
So we want leaders, we just don’t want them to be the only people with power in the organisation.
In fact there are many stakeholders that will interact with DAOs that should all have some degree of power, the greater issue is often ensuring that they don’t have too much.
So let’s analyse what they are.
Sources of Power
This not necessarily an exhaustive list of where power can concentrate in DAOs. In fact, I want you to identify your own. Tell me. And if they’re good ones I’ll update this post with them. My goal is to use a framework such as this to advise budding DAOs launching with FactoryDAO.
For now we’ve got the ones that I’ve plucked from the top of my mind. In each case we’ll discuss who they are and how they can use and abuse their power.
The leader(s)
These are the “team”. The people who are identified as responsible for directing the DAO along its path towards its desired goals. DAOs are often erroneously called “leaderless” organisations, but this simply isn’t practical in any future state of DAOs that I can see. Every organisation needs stakeholders that are more engaged, more informed, more empowered to act than the crowd.
Leaders are the crew that are responsible for coordination of DAO activity, they lead the DAO in effectively executing its will. They are not a “board of directors” since their action is not determined by any formal construct such as contracts, laws and regulations etc. They’re there because they want to be there, or even, just somehow ended up there.
Leadership dynamics in DAOs should be far more fluid. Hierarchical organisations rot as those at the peak of the pyramid become more secure in their positions at the same time as doing less and less.
In an effective DAO every participant should have a feasible path to being a leader and every leader should have a feasible path to exit from their position.
An effective decentralisation of this power base is simply having more of them. They should be a diverse group, covering many perspectives and disciplines.
The token holders
The token holders should ultimately direct a DAO, but it is erroneous to think they are the de facto locus of control in a DAO. They are only empowered if the right processes exist to confer their token stake into meaningful decision making.
Token weighted voting in its native state is purely plutocratic and the distribution of the tokens amongst participants will determine how much actual power they have in the governance of the network.
Wealth follows power law-like distributions meaning that the extreme majority of participants sit on the “long tail” of wealth. In pure token weighted voting systems these participants are fundamentally disempowered since those participants siting in the “upper tail” of the distribution and can control orders or magnitudes more power than everyone else in the system.
This can be improved by building mechanisms for distributing tokens to the long tail (for work for example), thus empowering them directly both in financial and in governance power. Also, token holders can be empowered by seeking better initial distribution mechanisms and into more unexplored territory through quadratic voting and reputation systems (I’m mega bullish on both of these).
The proposal creators
Even if you’re a token whale, if you can’t set the agenda, your power is fundamentally limited. The proposal creators hold the real governance power in a DAO.
Proposals define what is done and by whom. This is perhaps the greatest source of power asymmetry in a DAO and one of the hardest aspects to solve.
If you go fully un-gated, your proposal space is a mess and in fact can be easily sybilled. Token gating doesn’t necessarily solve the problem either. If it’s low you’re just slightly less noisy, and if it’s high you’re just back to pure plutocracy.
There’s a huge design space to attack this problem. Council structures, temperature checks, reputation, role distribution, emergent governance approaches, git-like amendment systems, curation mechanisms etc.
This has been a fundamental problem in political philosophy since the dawn of democracy and I actually think DAOs will push this conversation forward significantly.
DAOs will drive innovation in democracy
The multi-sig signers
DAOs will use lots of multi-sigs. It’s a core cryptoeconomic primitive.
It’s also a key function of DAOs to remove them. They could be considered the de facto executive power in DAOs and are therefore risky.
They are risky in terms of trust, since regardless of vote outcome (if none executive) the power comes down of whether the multi-sig signers follow ‘the will of the people’ (and they might not). It’s also risky in terms of potential liability if the regulatory environment shifts into hostile directions.
As it stands, multi-sigs are hot zones of trust in DAOs, but this can be offset significantly with more signers. But then, coordination becomes a challenge, anything beyond 5 of 7 (and it needs to be odd to avoid collusion) is desperately difficult to manage.
Management keys
It’s quite common in the early days of a protocol for the deployer to hold the power to change the contracts. If those contracts are upgradable then this power is unlimited.
The degree of power in these keys can be mitigated by having them controlled by a multi-sig and also limiting the design to soft parameter changes (as opposed to full rugability e.g. minting functionality). But again, if these exist it’s the job of a DAO to eliminate them through ossification of the code base and removal of upgradability requirements.
Eventually this power can be offloaded to voting mechanisms, or some kind of game theoretic mechanism design, or broader delegation designs. But in any case, it’s DAO god mode and shouldn’t stick around for long.
The developers
Huge asymmetry exists around the developer set. They are the ones who can create the protocol changes. A DAO can ask for functionality all it wants, but without based devs none of it gets done.
Not only that, the knowledge asymmetry that developers have is a perhaps an insurmountable challenge. The understanding of the nuances and implications of code changes will always reside with devs.
This can be overcome again through progressive ossification of core primitives, extensive and quality documentation work and education programmes. This can smooth the knowledge asymmetry curve substantially, but it’s foolhardy to believe that the crowd can (or wants to) get on the level with the devs.
Some DAOs will get their dev work done early and just stick with it, some more complex DAOs will have a constant organic code base and therefore require a deeper relationship with a developer community. Devs are the priests of DAOs, they will always hold higher implicit power, not a bad thing, but something to be mindful of.
Investors
Behind the scenes, there is often DAO related entities that have their foot in the real world that can materially shape the direction of the DAO through back of house means. It’s possible that there’s equity investment in those entities and those investors hold substantial power.
Their influence is proportional to their esteem and stake and if they’re mercenary enough they can use their leveraged token stake to control the economy.
This can be eliminated completely with highly distributed crowd sales, but this comes with its own risks. Routes to softening this asymmetry can arise through transparency. Clear articulations of the sources of power in entities related to the DAO can at least make token holders aware of the games at foot behind the scenes.
The community
This is a more meaningful idea than the token holders. Vocal members of the community can have very few tokens, but can sway sentiment and attitudes in the DAO significantly. If these community members are persistent enough they can be a bad apple that rots the barrel quickly.
On the more positive side, a collective notion of belonging and community can be the strongest power source a DAO can have. Nurturing a shared culture and mental framework for the DAO can make, or break the game.
Mechanisms for the making the community feel heard are important. Demonstrating to the community in practice that their desires leads to action is empowerment.
Empowerment is the super power of DAOs.
The state
Some DAOs may choose to jurisdictionalise, forgoing some of their autonomy and power to partner with a friendly nation state that has sufficiently codified their rules to play with DAO frameworks. This will allow DAOs to organise IRL: own land, have bank accounts, own basketball teams etc etc.
On a more macro level, nation states determine how adversarial the game really is. We need to worry a lot more about autonomy when nations states take a stance that organising with your friends online is a criminal offence. They can also change their minds. If a state can rug you, do you really have an ‘A’ in your DAO?
For now, it’s mostly friendly and the people getting bundled into the back of the vans are the people running multi billion dollar money printers gone wrong, or just outright frauds. Still, things can change and the state are an ever present substrate to everyone’s DAO lives. A big source of power.
The protocol
Every DAO has one, each with their own foibles and trade-offs. Ethereum is the ultra secure but expensive chain, Solana is the fast new kid that falls over, Polygon the Ethereum lite etc etc
Whichever protocol you choose to base your DAO on, it’s a source of power. It charges a tax for using it. A tax that changes like the weather depending on whether there’s an ape frenzy somewhere. This can have a substantial impact on your community and DAO operations.
Protocols, also have their own culture, values and cults, all of which can permeate your DAO. The smart ones will even bribe DAOs to set up there, but one thing is for sure it’s all base layer power.
Protocols are the nation states of the metaverse. Choose wisely.
The delegates
Some DAOs decide to build parallel forms of the bizarre IRL Punch and Judy shows we call representative democracies, and consequently have much of the same problems. “Protocol politicians” are politicians and therefore have egomaniacal tendencies and are subject to being publicly cancelled, but they do make things a bit more efficient.
This creates an effective power layer between the plebs and the governance class and it means your average token Joe can put their feet up and sit back and enjoy the DAO in whatever it does. But this does stick a new slice in the pyramid and will lower empowerment potential significantly, which can be a problem. Less so if you’re a protocolly DAO, more so if you’re a culture / creative DAO.
The influencers
The lads and ladettes who dominate the feeds of crypto Twitter. The power of these people can be off the charts. A sufficiently powerful influencer says you’re cool and it’s WAGMI time, another says you’re lame and its goblin town.
DAO members will tell you to blow your treasury on bribes to them, some of them may roll through the door and blow your representative governance to pieces with their ego.
It’s not entirely clear how DAOs will cope with this power structure, but it’s cool for now, they largely think DAOs will never work (but this will change). We’ll cross this bridge when we come to it.
The path to decentralisation
DAO designers, leaders, participants and stakeholders of all kinds should be asking the question “who’s in charge around here?” and using it as a means to design a path to decentralisation.
A path to decentralisation looks like a progressive distribution of power to as many DAO stakeholders as possible. This process seeks to eliminate all these sources of power in an on-going fashion.
What you might find however, is that power moves around. This is not a linear path. Limit one source of power and a new one becomes more prominent elsewhere. People seek it and they don’t like letting it go. If you have that mindset DAOs are probably not for you.
My governance spidey sense tells me that an iterative approach will work, and governance optimisation is the way to go. This is basically where you decentralise till it breaks, then wind it back till it doesn’t and repeat ad infinitum. It’s an on-going cyclical process.
We are about to embark on an epoch defining shift in human organisation, where community centric power emerges in locally defined communities at epic scale. But.. in order to get there, we have a lot of work to do, and it’s figuring out how to distribute power in DAOs as far and wide as possible. If this excites you join the FactoryDAO discord and help us solve these problems.
Essential reading for anyone considering joining or starting a DAO. Great article!
Since you asked, I thought of another potential source of power, the malicious actor with a lot of resources to infiltrate & subvert the DAO.
Very solid list of common sources of power in DAOs!