Broken meritocracy. Rethinking governance with smart contracts
This document explores why most governance systems fail, and it attempts to create a simple frame to spot some of the problems.
It also suggests how decentralized technologies like blockchains and smart contracts could solve some of the issues and significantly improve the results of the decision-making processes.
Redefining
Meritocracy is a political philosophy stating that power should be vested in individuals almost exclusively based on ability and talent.
Within the context of this document, I won't be referring to political systems or governments specifically but to any decision-making process.
Also, because power is not the same as power for making decisions I'll redefine meritocracy to the following:
Meritocracy is the idea that decisions should be taken by the entities better qualified for it.
Meritocracy as a base
All governance systems are a subset of Meritocracy if you understand merit as the feature that defines the right of an entity opinion to be considered (vote).
The ideal merit
For ease of use and simplification, I narrowed down the relevant merit properties to the following
Defines what's of value to the system in order to make decisions.
Can't be a medium of exchange
Needs to be quantifiable
You can find some examples in this document.
1. Merit defines value
For the system to work, the definition of merit needs to be aligned with the system goals.
It should tell what is of value to the system in order to make decisions. In this sense, merit is used as a filter, only the entities with certain capacities can participate in the decision-making process.
If defined broadly or incorrectly the system is likely to have unintended consequences or be gamed.
I'm using capacities as opposed to skills or knowledge because it can also represent things like ideology, IQ, experience, dedication, intention...
2. Merit is not a currency
It is not a medium of exchange.
If you can trade the merit you end up with a system ruled by the entities that have more economic power and any required capacity intrinsic to the merit is gone.
In consequence, the goal of the systems is reduced to maximizing its profit and the entities that rule it don't have the necessary capacities to do so (other than having economic power).
3. Tokenized merit
This is not a desired feature but a limitation. In order to operate in a system that can scale, we need merit in a quantifiable format.
This should be some sort of (not-exchangeable) capacity-for-decision-making token, although it may have other formats, or be the combination of several metrics.
This value will be used to decide if an entity can vote, and what's the weight of its vote.
In the case of Bitcoin, this is how many times a miner has been the first to find the block hash within a defined amount of blocks.
In the case of nation-state elections, this your ID card.
System failure
Most (I really mean most) systems end up compromising its own interest (#1) in order to make the system usable (#3). And so many others have a direct or indirect way to exchange merit (#2) and therefore compromise its integrity.
Tokenetization problem
Tokenizing merit is very hard, and I believe it to be one of the root causes for failure.
Most governances make use of a mix of certificates, test, recommendations, reputation... to validate the capacity of an entity. Then, a special credential is given to the entity that allows for its opinion to be heard.
This is mostly a manual process, making the system slow and prone to mistakes, and on top of that, extra processes are required to periodically re-evaluate the entity capacities.
Smart meritocracy
Governance has been one of the main use-cases for smart contracts.
The transparency of the blockchain and the programmable and authentication capabilities of smart-contracts make them very suitable.
To my knowledge, though, no project has reconsidered how they approach merit.
Whitelisting merit
As previously described in the analog/centralized world, the voting rights are defined in the identity credentials... not for its current merit. That forces the system to constantly have to re-evaluate these credentials.
By white-listing capacities (merit) instead of entities (usual people), we can have a governance system that is identity agnostic.
This would allow for anyone/anything to be a potential participant (kids, AI, other governances, anonymous entities). And because the system is automatized, the number of participants can be huge without being a problem.
It makes the system more adaptable to the needs and circumstances (just redefine the merit equation) while adhering to the system goals, and still being compatible with old governance systems (you can always make that a required capacity is to be someone).
Accessible metrics
How do you abstract all the necessary capacities to make a specialized decision and represent them in a single number while being in a decentralized/trustless environment?
Some capacities are easy to tokenize: time spent working at X, age, IQ, ownership of Y, citizenship. The ones that really matter are too abstract: knowledge, intentions, insight, fidelity, trust... Those are relatively easy in the analog world but a little harder on a trustless environment.
Reputation systems
Like in the analog world we can use the past as an indicator of the future. We can quantify proven capacities, and use them as merit.
Within this context, merit defines a required capacity of an entity to make decisions, while reputation is a proven record of a capacity.
This makes reputation systems very suitable for governance.
Finding consensus
Whatever metrics the system uses to define merit all the participants need to agree on the state of these metrics.
Blockchains
Blockchains are an instinctive answer. For proving identity, projects like UPort and Blockstack make it already possible.
Currently though, for most of the metrics, the data is unlikely to be available in the blockchain. Current blockchains are too expensive, and most of the decentralized platforms tend to minimize what they write in it.
Most importantly, most of the platforms are still highly centralized, and so is its data.
Trusted third parties
An alternative approach is to use a trusted third party to provide the required metric. Just by digitally signing the required data, we can have a bridge to the centralized world.
Subjective reputation
In general, when we talk about reputation, we refer to subjective reputation. How does an entity feel about something? This means that not all the entities may feel the same way.
But the fact that it is subjective doesn't mean it can't be useful. Like the above, participants just need to agree what entity will be used to digitally sign the reputation metric.
Tokenization solution
There is no need to explain how good smart-contracts are for creating tokens.
A smart contract can validate the necessary metrics, compute them and generate a none-exchangeable token that represents the merit of an entity.
If a system can migrate or upgrade the token, we can tweak the system goals in real time.
Now we can use this token to represent the voting rights within a system.
Conclusion
Different systems need different capacities to make decisions. Most governances are corrupted in some way due to the lack of tools to evaluate merit properly.
It is up to each system to decide what metrics to use to compute its own 'ideal' merit token, one that represents the system values and the required capacities to make decisions on it.
By whitelisting merit instead of identities we can:
Keep the goals of the system transparent and clear
Eliminate the process of evaluating the entities capacities.
Make the system identity agnostic, making it open to anyone/anything
The system is backward compatible (the list of whitelisted identities is a subset of the result of the whitelisted merit)
Smart contracts are extremely well suited for it. - Allow the system to easily adapt to new circumstances - Keeps the system transparent - Make the process maintenance free
With the new decentralized paradigm, we can now build systems that define merit with an accuracy that we never had before. This should allow for a substantial increase in the quality of governances.
Additional notes
There are some subjects I naively believe are worth mentioning and/or exploring.
Prediction markets
There has been a lot of talk about the role that prediction markets could have on governance.
If the results of the decisions can be easily measured and the pool of participants is high, futarchy seems a promising idea but it's an approach limited to a very specific type of governances.
These systems are especially interesting because they add a very strong incentive layer. I believe similar incentives could be designed for the proposed model.
Proxy voting
Proxy voting or Liquid democracy are commonly suggested improvements that the new paradigm makes easy to implement.
The way I see it, it makes merit exchangeable, therefore breaking property #2.
Governance for governments
The previous model fits well on voluntary governances, where you participate because you agreed on the principles and goals of the system (an open-source project, a company...)
Nation states and certain types of communities are none-voluntary by nature. Its citizens never agreed on the foundations that drive their nation (maybe a small representation of them, their ancestors, but not the current citizens).
A Constitution represents these pillars. An immutable document that roughly describes the principles that drive an ever-changing society.
These principles are at the very least outdated or unclear, making the government gameable.
The presented model could be used to improve this situation.
It may seem hard to apply because there are no clear values, therefore we can't define merit.
Instead, we could crowdsource what merit means. Each citizen could vote on its own principles and values regarding certain subjects. The aggregated results could then be used to automatically compute the definition of merit on each of the subjects.
The voting wouldn't have to be time-framed. Instead, a citizen could at any given time, change its own values, and the merit for governing the system would be updated in real time.
The definition of merit could be considered some sort of constitution. And the values that each citizen has voted could be used as reputation metrics for other systems.
The idea of the citizens voting on what they are experts on (which is about themselves) is borrowed from Ralph Merkle's draft DAOs, Democracy and Governance
Note: I’m from Catalonia, where according to the Spanish constitution, we are forbidden to express our opinion about our sovereignty.
Another example
I used nation-states as an example because I believe is the most intuitive thing to think when talking about governance. The example is naive in few aspects, but mainly because it underestimates the complexity of a state.
Another example could be a DAO that controls the water supply network of a community. The users could vote on the things they care: price, flavor, added chemicals, health impact, environmental impact...
Depending on the results, the merit will morph adding or subtracting requirements.
In this case, maybe the necessary knowledge to take decessions will remain constant, but a metric representing the intentions or incentives could vary.
A certification agency (some sort of Greenpeace) could be used as a trusted party to hand green-tokens to quantify how much an entity cares about the environment, for example.
Axiarchy
I used the term meritocracy because is a word that is familiar to most, and it almost describes the basics of the described idea.
But as stated, it is not the same as the described model.
Meritocracy makes specific reference to 'power'
Merit, from Latin mereō, and -cracy, from Ancient Greek κράτος kratos "strength, power"
To differentiate the two concepts and to have more accurate description, I suggest using the term Axiarchy
Axía "value, merit", from Greek αξία, and -archy, from Ancient Greek -αρχίᾱ arkhíā "rule, government"
Maybe a philologist could make a better suggestion.
Work in progress
This document is part of the Uumm project.
We're currently trying to set the foundations to design and build better governances.
This document is an attempt to expose the ideology behind it, so the community can iterate and point to any flaw, before moving to next stages.
You're very welcome to join us on this exploration. No merit required.
This document was published simultaneously on my personal blog and Medium, and the full change log is available on Github.
This document was possible thanks to the stress tests, constructive feedback, and support of the following people Daniel Armengol, Greg Albritton, Xavier Vives, Assumpta Riba and Lydia Law.