Why governance in most crypto project sucks and how to fix it
Notes: This short essay is by no mean an attempt to solve all the problems of governance solutions in distributed networks. It’s an attempt for sharing thoughts and ideas that goes sometimes beyond the crypto scene, to start discussions and maybe upset a few people. Hopefully, there will be more discussion than people being actually upset.
I’m around since a long long time. Long enough time to realize that something is going on and long enough to know that what’s going on in crypto isn’t just about crypto. We could play and draw a line to tell a story; it would go from the roots of Bitcoin and straight to the ongoing debate about governance. The rise of DAO (and DFO), the ongoing creation of new theoretical frameworks where code and linguistic meet economy - those aren’t distant abstractions. This is the very reality we live in every day: Bitcoin started along with the Arab Spring and the Occupy movement, at a time where people all around the world were losing the little trust they had left in the centralized systems we built our societies on. “We don’t want banksters, corrupted governments, our leaders are constantly lying, our democracy is broken!”
It was the motto back then, and it still is. It’s not for no reasons researches about wisdom of the crowds suddenly became famous in the early 2000', something is broken indeed.
Bitcoin was one of the answers to all these issues and questions. It was new. It was weird. It was beautiful. At first, it gathered researchers, IRC nerds, weebs, geeks, and all sort of people living aside of a mainstream reality that completely ignored crypto until the mid-late 2017. Not exactly your usual elite, and this was for the best. The monetary power distribution happened in a way where lucky IRC insomniacs got the biggest share of the cake. Some of the most marginalised ones never really sold. They are the ones who helped to make Bitcoin a holy land and one of the strongest symbols for an anti-system ideology. It shown us a way for a group to govern itself and it opened the gates to leaderless governances where codes could run societies.
But the Bitcoin network is also a neo liberal answer to the old system nobody wants to follow anymore. It has its roots in a political vision that does not answer properly to every problem. This inadequacy leads to situations where projects mimic and misuse some of the Bitcoin characteristics without clearly understanding what this is about, thus they’re completely missing the very problem they often claim to solve.
The Bitcoin’s ghost: From network consensus to governance models
The Bitcoin network work as follow:
- a transaction is sent.
- it is propagated and joins the unconfirmed transactions in a memory pool.
- miners compete to solve a hash functions and pack the unconfirmed transactions from the memory pool to a new block.
- once a hash function is solved the corresponding block is added to the blockchain and the corresponding transactions are confirmed. The network just reached a consensus in a decentralized and trustless fashion.
- the difficulty adjustment algorithm ensures that a new block is added approximately every 10 minutes.
- In case of a fork, the longest chain is considered real.
It works great, and from there everyone seems to think the way hashing power plays in Proof of Work implementations of the Nakamoto consensus could and should be mimicked in governance problems. We have yet to find efficient ways to prove someone’s identity onchain without compromising people’s privacy so why not weighting voting power according to token holding for any given address? This is what we’re doing since years but if we’re looking for decentralisation then no, it simply can’t work. Every single one of your community driven project when being built that way will most likely end up being a cartels playground.
Wikipedia defines governance as, “the way rules, norms and actions are structured, sustained, regulated and held accountable.” The United Nations says, “Governance has been defined as the rules of the political system to solve conflicts between actors and adopt decision.” Polkadot’s team defines governance as follow:
- A system of rules
- To make decisions
- With participants being held accountable
Whereby the Polkadot’s governance is interesting as their loss-weighted and timelock functions are used to benefit long history of voting support for candidates under certain circumstances and increase the voting power of DOT holders, it complexifies a design without addressing the roots of its flaws. Actually, it doesn’t even name it.
RRRRRRRRRRRRRRRRRRRoots
Oddly, governance solutions inherit from the experience we have in distributed network protocols, in the ways they reach consensus which are beautiful but also leave us with a cursed legacy.
They’re beautiful because they’re paving the road for a world where we won’t need to trust leaders who can be corrupted, and they’re cursed because their beauty hide their roots and therefore forbid us from questioning the box we’re thinking in. The idea of wealth distribution and the game theory modern distributed networks consensus are being built upon come from Adam Smith, Friedrich Hayek and the liberal zeitgeist of a market where free individuals exchange and interact for everything to end up fairly balanced. Individuals will exploit their full potential as they’re emancipated from constraints of arbitrary rules. A fair wealth distribution and thus a fair valuation of things will happen this way. The invisible hand of the market value everything at its right price and this is the expression of a natural situation, a spontaneous order that can benefit the individuals and the group as a whole: everyone improve!
Ok good. Everyone improves. But the invisible hand of the market is just people talking to each other through exchange of value mechanisms. Due to different initial situations, some are given the right to talk louder than others from the beginning: some are born rich, some are born poor, we don’t have the same power to discuss. Maybe this is good, maybe this is bad, but either ways it has consequences we shouldn’t simply wave away if we want to build efficient governance solutions that benefit the network. For instance, what happen when financially powerful individuals decide to work together and bend the market to make it fit their agenda?
Bitcoin works almost flawlessly since more than 10 years because miners have incentive to work for the good of the Bitcoin network. Bitcoin running smoothly benefits the network as well as big and small miners. Is it a problem if 56.6% of the hashing power is controlled by only 4 entities?
Well, despite all the rage on social medias about Bitcoin decentralisation, it might not be an issue. Bitcoin assumes reasonable actors will act for their own benefit, and that is keeping the network alive and healthy. Good news, it’s also what benefits the network.
Issues arise when you are copying a consensus designed to work at the protocol level and naively try to make it work at the governance level. It doesn’t work in situations where the interest of the most powerful individuals isn’t what benefit the network and this is highly political. Plus, there is no such thing as a natural Schelling point when voters aren’t looking at the same timeframe, have different political background, position, etc. Here, the centralisation versus decentralisation debate isn’t only about the assumptions we use to draw the thin line between those two concepts, it’s also about two ways to phrase the same problem: how individuals benefit from the network or how the network benefit from individuals.
The BUIDL and ZNN cases 👻
Money is the voting power individuals have within a network. This is what Bitcoin does at the protocol level with the Proof of Work implementation of the Nakamoto consensus where miners have to pay for the hardware and electricity that will provide the hashing power the network need to be secure. This is how Proof of Work networks reach consensus. This is what everyone is doing at the protocol level and this is also what ZNN and BUIDL try to do at the governance level. And this is a bad idea.
BUIDL and ZNN are governance tokens, (ZNN is also a native coin but let’s keep it as simple as possible) which mean people buying one of those tokens are literally buying voting power within the network. When someone make a proposal, token holders can vote, their voting power is weighted according to the amount of tokens they hold. Within the BUIDL network, anyone can vote, for ZNN only people with 15'000 ZNN and more are allowed to vote. 15'000 ZNN is one vote, 30'000 ZNN is two votes etc. There are already powerful individuals with a lot of voting power within the ZNN network and it would be unreasonable to think that they will vote for a proposal that does not fit their vision of what ZNN should be. Holding a trizillion token doesn’t make an individual any smarter not honest, and nobody really want to be forced to trust a small group of people for the future of a network with virtually an infinite of individuals, right?
Even if its non-tiered model should mitigate this issue, BUIDL is still affected by the same voting power concentration: votes that shape the network and decide for its future does not reflect people’s voice but people’s money, voting power will be concentrated within the hands of a few. With enough money, you control the network. One could argue that with time distribution will happen and everything will be for the best, especially with VCs stakes being unlocked after months for both ZNN and BUIDL. It wont. The Bitcoin and Ethereum rich lists should be enough to understand that distribution and accumulation process don’t unfold that way, even when smart money forced to buy late. This is ironically the power concentration problem Bitcoin was meant to solve. Whether or not people deserve their voting power should not be a question: when we talk at the governance level, voting power concentration means central entity. The question should be how to balance individuals interest and the network interest, for the best of a truly decentralized network.
The most rightful wrong 🎵
There are many ongoing researches about democracy and governance models, about what works for groups with different views on what effective individual’s incentives and their effects on different timeframes. From there, we can list three major ways that could undermine the voting power concentration:
- Considering that money shouldn’t vote and therefore you need to rely on a decentralized proof of identity protocol that could make it impossible to game the voting process while not allowing any voting power concentration to happen. This is by far the most elegant solution as it remove the voting power concentration problem by design without compromising the trustlesness of the system.
- Considering that money could vote under certain circumstances and within a balanced system that include counter powers: For each proposal, a quorum could be randomly elected. The voting power is weighted through the tokens being held per address but not everyone can always vote. Once a proposal successfully passes the vote, a quorum could be randomly elected for a veto vote. If the proposal is approved at this second stage, the vote ends here, if the proposal isn’t approved at this second stage, the proposal is out and can only be submitted again after a cooldown penalty. If submitted again, voters in both quorums would have to be different. While cartels can’t impact all the votes on all the proposals this model doesn’t really mitigate the voting power concentration problem, it undermines its effects.
- Considering that voting is burning money. Voting has a cost and a heavy voting power comes with a heavy cost. Big holders vote weights more than small holders votes. But, proportionally, small holders votes cost less than big holders votes. The voting cost is weighted through an exponential curve. This is the most exciting possible solution; instead of offering an incentive for people to vote, it grows the voting penalty as voter’s stake in the network increases. This might seem unfair to individuals concerned about their capital, but at the network level, it’s healthy.
There are many other ways to avoid cartels formation or to mitigate their noxious effect on a given network of users. To do it efficiently, understanding the problem, its history and the ways it impacts individuals expectations and groups dynamic is required but not enough; being able to phrase the problem differently, to get out from the money talk paradigm we’re trapped in, this is the only way to be able to design truly decentralized, unstoppable and leaderless governance systems. What next?
Thanks to Champi#7066 for the talks on the road.
