William Mougayar, a CoinDesk columnist, is the author of “The Business Blockchain,” producer of the Token Summit and a venture investor and adviser.
It has become trendy to apply blockchain governance principles to almost anything. But if we want to make blockchain organizations work, I believe we need to make an important distinction between the “governance of blockchains” and “governance by blockchains.”
What works for the technical realm of blockchains does not automatically translate to the running of businesses or social organizations, as attractive as it may be to apply the blockchain novelty factor. Democracy is great for society and government, but it is bad for business.
Over the past few years, we have learned a lot from the “governance of blockchains,” primarily about consensus and decentralization as the two primary characteristics. Consensus in the blockchain context refers to the nodes on the network agreeing to synchronize on the state of transactions for that blockchain. And decentralization is the preferred topology for these nodes: they are geographically distributed and diversely owned. This ensures redundancy of failures as well as a level playing field of participative inclusion, both important outcomes of open blockchains.
Our applied knowledge stalls thereafter.
We know much less about the field of “governance by blockchains,” as this is still at the experimental stage.
The desire to apply consensus methods and decentralization architectures to how we run organizations is an interesting concept. It stems from well-intentioned objectives wanting to mimic the governance of blockchains as a guiding strategy.
In the simplest form, people who have a stake into projects are seen like nodes, and they are given voting power. The more decentralized the better. Voilà. If it works for a blockchain, it should work for organizations, no?
Except that people are not computer nodes and consensus is a dreadful management practice for decision-making.
Many decentralized projects or ideas are slapping the word DAO to their names, in a loose way. We now have a Legal DAO, Marketing DAO, Investment DAO, Jurisdiction DAO, Democracy DAO, and so on (I’m refraining from naming actual names, though we all know who they are). These groups are obsessed with group-voting for decisions, being very transparent about their actions or discussions, and adopting decentralized inclusion from the beginning. As a result, they end up voting on everything and insist on public openness as a modus operandi.
The problem with decentralized decisions
Agreeing by majority voting doesn’t always result in the most optimal decisions, and it often leads you to compromise at the lowest common denominator level amongst the group, resulting in mediocre outcomes. Most decisions carry a degree of ambiguity and uncertainty that experienced decision makers are used to, whereas a group of people with less experience will rather debate these ambiguities for a long time, still without resolving them.
The tough and bold decisions are left un-tackled and never made by consensus because there will always be dissenting voices that will prevent these decisions from taking place. For example, a given idea might be beneficial economically to the majority of users, but not to a minority of them immediately. Do you change the decision so that all users get lower benefits, or do you take the tough decision to optimize for the majority of users first? There are never perfect decisions.
One squeaky voice could lead to endless discussions, halting progress or delaying voting, without regard for urgencies or efficiency in implementation. The voting method can also break down when you end up with a less experienced majority that votes on the wrong decision. When the barriers are low for getting a seat at the table, just being there doesn’t say anything about the experience level of participants. Over-voting can lead to zig-zagging into iterative decisions with an appearance of progress.
Popularity doesn’t always lead to the right decision. Just because it is popular doesn’t mean it’s the best. Do politicians get voted in power because they are really competent or because they become popular enough to garner the required votes? Many elected officials (and even presidents or prime ministers) end up as bad choices after being elected, but the electors are stuck with them until their term is up. Elections by popularity vote are a type of irreversible decision-making that is hard to undo, even as we believe in the “wisdom of the crowd” thesis – that more people deciding equals better deciding.
How about accountability? In the most optimistic scenarios, the outcome of decisions turns out as predicted, but in many cases, things don’t always go as planned. Who is responsible for fixing what happened as a result of a bad decision? We can’t say, “well, we voted on that” and wash our hands. If 20 people voted on something and things don’t work out, who is to blame? Who will pick up the ball and fix things, change course or reverse decisions to ensure success?
Many decentralized blockchain projects do not appreciate the need for central leadership, and they brush that aside. It is not cool to take orders from others, or it’s not decentralized enough. Let the community decide, say decentralization advocates. But you can’t ignore leadership, because it is needed. Decentralization should not be about rejecting hierarchies and people management. The reality is that you can’t vote on everything, and people need to be managed with accountability in mind first.
Then there is the issue of culture. When decentralized groups are instantly assembled and they meet only sporadically, coalescing around a homogenous culture or agreeing on principles is difficult to achieve. In the context of blockchain operations, each decentralized nodes works for its own economic advantage, like a mercenary. But putting mercenaries together in a room doesn’t automatically yield cohesiveness at the required level of sound group dynamics.
Do DAOs work?
We are still in the early generation of DAOs. Almost four years have elapsed since the first DAO malfunctioned, but we haven’t had four years of experience working on iterative models since then. Only recently have we started to tinker again with business-minded DAO and decentralized decision-making across a variety of applied scenarios. It is easy to start a DAO, but much more challenging to devise the right structure to support and operationalize it.
We need to evolve the concept of business-minded DAOs with another level of sophistication among its practitioners. Baking programmable decision-making before enough experience has been gathered or previously acquired in whatever other relevant domain is a sure recipe for disaster.
Governance is a very promising application of blockchains, perhaps after the transfer/ownership of money and the emerging field of decentralized finance. But blockchains and new-age governance are not going to reinvent the process of decision-making by throwing away decades of management soundness and best practices experience.
However, we do need more experiments to see if (and how) we can improve upon what we already have. For example, there are promising ideas in innovative voting mechanisms, such as quadratic voting, capital-constrained liberal radicalism (CLR) voting and other related variations that go beyond the blanket “one person-one equal vote” tradition. Quadratic voting is a distribution method that lets voters assign more than one vote across candidates. CLR voting takes into account the total number of votes and their provenance, not just the weight of that vote. So, for example, 500 votes from a weight of one share counts more than a single vote with a weight of a thousand shares.
Technical consensus in blockchains has been well defined and documented, and is backed by many years of research that culminated in Nakamoto putting it all together into bitcoin and its blockchain. But using blockchain consensus methods just for the sake of using them in business is not necessarily a breakthrough itself.
In business, life and technology, good decisions are good because they produce success. But good decisions are only judged in the rearview mirror, well after outcomes fully unravel.
You can’t easily automate decision-making just by configuring logic into a smart contract and trying to mimic and predict all scenarios that need judgment rather than automation. You must ask - what are we trying to fix or invent? And do we have the right people with the relevant domain experience and expertise at the table, beyond just knowing about blockchains?
You must define and organize your processes early on, and not fall into the trap of wanting to decentralize everything too quickly. Not every process needs to be on-chain. When embarking on a DAO experiment, you need to be clear on what is DAOable and what is not.