The software engineers who created the blockchains of today built them to be resilient under conditions of adversity, using cryptography and economic incentives to make them hard to attack. What they didn’t anticipate was how blockchains might perform under conditions of cooperation.
On 11 June 2022, a group of validators, with input from the research collective Metagov, expressed their intention to work together to achieve better governance for Proof of Stake (PoS) blockchains. The starting point, as articulated by Teddy Knox (Stakefish), was simple: “Those with the power in on-chain governance should recognise that they have authority and there should be grassroots organisation within that group”. By coming together and agreeing to implement good governance practices, these validators aim to avoid a “race to the bottom” fuelled by financial incentives.
At the event, the founding group presented a declaration setting out their shared values, including:
· That token holders are the fundamental source of power, and they vest that power in validators when they delegate.
· Delegators deserve to be informed of decisions taken on their behalf; and
· That by participating in governance, validators can provide a check on the undemocratic power of foundations and core teams (read it in full here).
The contents of the declaration were only indirectly discussed at the event. Instead, the conversation focused on the validators’ decision to present as what might be the world’s first blockchain political party. The rationale behind forming a political party (discussed further below) is that validators can choose to signal and demonstrate values through affiliation, thereby differentiating themselves from validators or other entities who don’t share those values or practices. Token holders may then be able to make better decisions about which validators they entrust to participate in blockchain governance processes, which by nature spans infrastructure (hardware and software choices) and on-chain voting.
The main themes and tensions that arose during the Validator Commons event were:
· How the issues confronted by validators have arisen through the design of blockchains;
· The emergence of institutional formations that will produce outcomes for blockchain governance, such as political parties, industry associations and policy-geared research endeavours and why these may be necessary in open systems;
· The difference between coordination and collusion; and
· The need for legibility over the systems that constitute blockchain governance and the challenges in achieving this.
The ideas below arose from the minds in the room that day (I was merely facilitator although I had been observing and participating in the process that led to the declaration as part of my ongoing ethnography of blockchain governance). Aside from people who work for validator companies and Metagov researchers, a number of founders, foundation board members and core developers were also present — mostly from the Cosmos ecosystem although it was never intended to be a Cosmos event. As it turns out, some who are involved in creating and maintaining blockchain software also run validators, which meant that many in the room had experience across governance, infrastructure, cryptoeconomic design and research.
What is a Validator?
In Proof-of-stake (PoS) blockchains, consensus is achieved by individuals or entities running a blockchain’s specific software and depositing tokens into a smart contract to create validating nodes. These nodes receive block rewards and transaction fees for proposing and approving blocks and can be penalised for bad behaviour.
Creating a validator means maintaining computing infrastructure and/or depositing a large sum of capital in the form of tokens. Token holders who are not able or willing to meet the criteria for establishing a validator are able to delegate tokens to an entity who will run validators on their and others’ behalf. Typically, the validator service provider will take a portion of the reward and/or transaction fees in return.
It is important to note that PoS blockchains vary in their design. Some blockchains have a maximum number of validators at any point in time. Cosmos Hub is an example of this; if a validator address falls out of the top 125 addresses with the highest balance it will exit the validator set. Ethereum, on the other hand, is moving to a PoS system that has an unlimited number of validators, but rewards decline as the number increases. Governance also varies across chains. In Cosmos Hub, a delegator can still vote independently in on-chain decision-making processes after staking with a validator and if they fail to vote their vote goes to the validator. Ethereum post-merge will remain governed through the off-chain current EIP process, limiting the governance power of validators to infrastructure choices such as which client they use or implementing an upgrade or fork that changes how things are done or who has power in the system.
What’s the problem?
The Validator Commons arose out of concern among some validators that governance processes are failing. This is evidenced in the concentration of capital and power in particular entities as well as the behaviour or apathy of some validators.
Concentration of power occurs when a small number of validators have a large stake in a particular blockchain such that they might be able to collude to determine the course of that chain. Having significant stake and not voting can be an issue too. For instance, achieving quorum for on-chain governance may be impossible if centralised exchanges that offer staking services to their customers hold a large proportion of staked tokens. Typically, these entities do not participate in voting and those who delegate to them do not have the means to vote in their own right as they do not have custody of the token.
A common and immediate issue involves validators not paying attention to crucial governance events. The validators present spoke of the day-to-day challenges of staying across governance discussions occurring on multiple chains. While those who have been involved in the Validator Commons group expressed willingness to do so, others are seeking advice on how to vote in private Discord channels, suggesting the possibility of collusion in the absence of transparent processes.
One example of where validator responsibilities failed through neglect was Proposal 20 on Juno. In this case, all 125 validators failed to check code that was deployed into a live environment, sending $36m into a dead address. Daniel Hwang from Stakefish pointed out that validators who are unclear of their role in the system are more likely to allow such errors through the gates. No one was held accountable in this instance, even though it was “a copy paste error that even a non-technical person could check”.
A scenario that has not occurred to date but could happen in the future is one where validators with financial interest in a particular blockchain conspire to commit a governance attack on a competitor chain that they also run validators for. As Chris Remus (Chainflow) has written, good faith relationships are typical at the start of a system, but these may degrade over time. Establishing best practice and knowing who the responsible actors are while there is still goodwill, innovation and experimentation across blockchain ecosystems could help to help avoid conflicts of interest turning into outright attacks.
Should this be a political party or something else?
The Validator Commons was initially convened to settle on something akin to “governance standards” or possibly an oversight entity. As the discussion progressed, the idea of a political party emerged, recognising that on-chain governance is both competitive and cooperative. Those who hold tokens should be able to make informed choices when deciding to delegate. A party sends information about values and intentions, whereas standards are intended to be viewed as ubiquitous, anonymous and objective, typically requiring compromises to pass, and creating path dependencies. A political party enables those who are opposed to the party’s values or actions to emerge and define their own position. According Metagov’s own research (led by Joshua Tan), the blockchain ecosystem possesses ideological diversity but not necessarily the institutional structures for those ideologies to inhabit.
In this case, the movement towards the formation of a political party needs to be seen in context of an ecosystem in which core teams and foundations hold considerable power and where the connection to token holders only works if validators participate. The goal of coordination is therefore to amplify and increase participation in governance processes. It is foreseeable that such an endeavour could evolve to a point where coalitions of validators fund the work currently undertaken by foundations, including upgrades and auditing and this might be considered as a more democratic than current scenario where token holders have no power over foundations.
There is also the distinct possibility that blockchain political parties might be done differently from traditional political parties. Just as bitcoin emerged or gained traction through widespread disaffection with the financial system (following the events of 2008 financial crisis), the various formations of blockchain governance are emerging during times of ‘peak cynicism with politics’, as Teddy Knox described it. These party-like formations could be made ‘scientifically productive’ (in Michael Zargham’s words), providing substance and tools for evaluation and decision-making.
Other possible organisational forms were also discussed during the event. Industry self-regulation typically arises in conditions of information asymmetry, as is the case here. Metagov’s Eric Alston pointed out that trademarks evolved from guild marks and that such devices are a way to easily show who is in or out in a market context. Labelling those who abide by certain standards and industry agreed practices means that affiliation can be taken away from an entity when if it falls short. Another approach, suggested by Daniel Hwang was that future systems design could implement means to punish validators who do not participate in governance.
Opposing views and future possibilities
‘Do you pick a US senator based on who is going to give you an NFT drop? That’s literally what’s happening now’, Eric from RHINO commented. The assumption that token holders will make decisions based on the governance participation of validators may ultimately be a waste of time.
Some founders and core developers argued that the code that is run by nodes is and should be the ultimate authority. From this perspective, core developers, typically employed by foundations, know what is best and validators act as administrators who implement that code. The possibility of separating out the role of validators from decision-making entirely was canvassed, whereby token holders would look to protocol politicians to make decisions on their behalf. A system in which sub-DAOs take responsibility for different tasks or spheres of decision-making is being considered within Juno (Optimism is also experimenting with a parliament-like governance body). These possibilities are being considered in order to reduce the effects of raw economics that are emerging in PoS blockchains.
These ideas point to an emerging research agenda for PoS blockchain governance. In addition to questions around how blockchain political parties might differ from traditional parties through their role in facilitating legibility and values, R&D efforts should also be directed at making governance legible. Such an endeavour would require mapping and understanding the roles that already exist to identify gaps and avoid situations where we end up with ‘founder prophets manufacturing consent for proposals’ (in Michael Zargham’s words). Such information will also help guide who is best placed to take on roles including quality assurance or policy formation.
The Validator Commons initiative signifies a juncture in the institutional evolution of blockchains. As Daniel Hwang commented, “Maybe it’s the fault of the designers that they have put this on our lap”, adding “We are not rejecting it”. Some founders appear to be suggesting that the systems they created need redesigning rather than political parties or other institutional formations to resolve failures. As any redesign will ultimately be subject to validators adopting the code, perhaps the only path forward now is cooperation.
I am a Professor at RMIT University in Melbourne, working across the RMIT Blockchain Innovation Hub and the Digital Ethnography Research Centre. I am also an Associate Investigator (AI 🤖) of the ARC Centre of Excellence for Automated Decision-Making and Society. I acknowledge the support of the Australian Research Council, FT19010372. I use this Medium blog for work-in-progress ideas and reflections.