Blockchain Governance: How decentralized do we need to be?
While decentralized governance is not a panacea solution to every use-case, more of it is needed in a world that is mired in growing inequality, nationalism, mistrust in traditional government/banking institutions, global warming, potential nuclear armageddon…the list goes on and on. Although I’m very excited about all of the different Blockchain governance models that will be rapidly experimented with and evolved over the next couple of years, it’s important to highlight that the topic of governance as it relates to government institutions, companies, foundations, and other organizations is an extremely nuanced topic. Depending on the project, this topic should be analyzed and explored with the understanding that varying projects and organizations will land in different places on the spectrum of decentralization.
After reading countless token whitepapers and seeing so many projects throw in the word “decentralized” or “decentralized governance” as a buzzword, I can’t help but ask over and over, why does x particular use-case need to be THAT decentralized? And if not that decentralized, where in the spectrum should it end up at to add the most value to society? To answer these questions, this post will examine a variety of lessons learned from history, other disciplines, and current blockchain governance approaches in order to illustrate that implementing decentralized governance is not only a function of technical designs, but also a function of cultural adoption. As such we should strive to take the best practices from UX design, corporate governance, and innovative company cultures to maximize the value of decentralized governance. Determining the right mix of technical design and best practices of cultural adoption will then inform us on how decentralized our particular use-case needs to be.
To determine where we should end up in the decentralized spectrum, let’s first take a stab at painting the current landscape.
Before reading further, I want to caveat that this initial spectrum is not based on a formula backed up by data to determine where a particular industry or government institution currently falls within this spectrum. In fact I would say that my first pass at painting the current landscape is quite arbitrary and is intended more so as an illustration tactic for us to think about the topic of decentralized governance in a more graduated fashion.
Let’s take the industry of Banking as an example. In the blockchain community, we generally think of Banking as one of the more extreme centralized industries that have many middle men bankers and brokers that are ripe for disintermediation. The mechanism of a bank being the trusted third party that provides the final debit/credit settlement processes is in concept centralized; however, the industry of banking/finance can be thought of as fragmented in the sense that there are many banks that need to operate in a network of standards in order to facilitate transactions between each other (CHIPS for domestic and SWIFT for cross-border). There are also many payment technologies (VISA, Paypal, Venmo, Square, etc.) that work with this existing banking system in order to provide integration and ease of use services. These players can also be thought of as centralized in that they operate “walled gardens” that you must put trust into in exchange for fees. Suffice it to say, the exercise of determining whether an industry should be centralized or decentralized is not quite black and white, and as we explore blockchain technology and governance solutions, we should strive to understand what the strengths and weaknesses of our current systems are before replacing them completely with a decentralized solution.
Understanding where we came from
The purpose of governance
To borrow from the very comprehensive overview of blockchain governance Fred Ehrsam put together:
“As with organisms, the most successful blockchains will be those that can best adapt to their environments. Assuming these systems need to evolve to survive, initial design is important, but over a long enough timeline, the mechanisms for change are most important.”
To facilitate evolutionary progress, the primary function of governance is to act as a self-sustaining mechanism to allow for maximum improvement and growth for all participants (subjective) in the ecosystem. Put another way, governance should serve as functional science in that it provides a conduit to create/refine/maintain knowledge while also acting as a mechanism to decide on the best actions given the current facts and theories of our time. Put a third way, good governance would lead to a result of taking meaningful actions to combat global warming rather than espousing its members towards devolution (whoo hoo, back to the coal mines!).
Understanding the participants
In any new or existing governance system, understanding the participants in the ecosystem and their incentives/motivations is crucial. To dive deeper into motivation, we can look to the study of gamification as it seeks to understand user psychology and motivations for the purpose of designing game ecosystems. This has parallels to designing a governance system in that they both involve mapping out the participants motivations and designing the mechanisms of coordination. Borrowing from Yukai and Andrzej’s post about the 8 user types of Gamification, we can build off of this initial framework and craft an even more comprehensive list of the types of participants in blockchain governance via best practices from the UX world of defining user-personas.
Ultimately, each participant or group in a system will have their own incentives, and more than likely those incentives will not be aligned completely. If we are going to usher in a new era of rapidly testing many blockchain governance designs as it relates to on-chain (layer 1) and off-chain (layer 2) coordination, we might as well take the lessons learned from the gaming world as it relates to rapidly simulating mechanisms of coordination. One especially interesting parallel between gaming and blockchain systems lies in the online nature of both mediums. I’m sure all of us can attest to how different people behave in person versus when they have a layer in between (trolls on forums). There’s a whole litany of research on that topic if you’re interested in that rabbit hole: cyberpsychology post.
The tug of war history between decentralization and centralization
Enterprise IT’s battle between on-prem vs cloud
While gamification psychology as it pertains to blockchain governance design has many avenues to explore, an even more important avenue to first dive into is understanding previous governance models as history often rhymes. From a business perspective, governance in the worlds of B2B vs B2C has had a tug of war history of decentralization vs centralization dating back to the advent of computers. In the enterprise space, Fortune 500 sized companies all over the world have spent massives amounts of resources in the past decade centralizing their data and governance processes into the Cloud. This is all to reverse the previous damage done when on-prem mainframe solutions were touted as a decentralized governance that would increase efficiency and allow individual business divisions to more quickly customize and react to problems. What started off as a grand vision quickly devolved into business divisions customizing their processes and systems too much; thus resulting in wasteful data-entry duplication efforts. For a more comprehensive history on this topic check out this post. When we take a deeper look at history, it becomes clear that centralized vs decentralized governance is nothing new and will always have this back and forth battle.
The creation of Internet 2.0
Remember that the internet initially started as a P2P protocol that had a decentralized ethos. In the ensuing decades, companies such as Amazon, Google, and Facebook created tremendous value in the era of Internet 2.0 aggregating data and processes to deliver convenient and delightful user-experiences. The flip side of this aggregation being that as these players became increasingly centralized and powerful, they invariably ended up further on the centralization spectrum, and as a result, began to introduce other externalities such as invasions of privacy and monopolistic economics.
The media war
Another example of this back and forth tug of war can be seen in the entertainment industry in how technologies such as torrenting tried to decentralize file sharing only to be cracked down by state insitutitions and re-centralized into the current fragmented streaming subscription economy that we have today. History shows that depending on the particular industry, businesses and governance swing back and forth between this spectrum. Unfortunately, history has often violently swung too far in one direction; thereby prompting a new technology or movement towards the other direction. As we explore decentralized governance models, let’s remember where we came from and not inadvertently squash what can be an extremely value adding technology to society by swinging too far towards the decentralized direction. Heeding the lessons of history and cautiously moving forward, let’s dive into why blockchain governance is so exciting and how it can move the centralization needle back towards the right amount of decentralization (as I believe we are currently unbalanced in this transition phase of Internet 2.0 and 3.0).
Why Blockchain decentralized governance is exciting
Fundamentally, blockchain technology is a decentralized ledger that is resistant to censorship, immutable, transparent, verifiable, and secure — with varying degrees to each of these properties depending on the protocol. These properties set it up well to be a prime candidate to implement decentralized governance models in which bottom up innovation can be proposed, voted on, verified, and implemented. While Blockchain governance has so far related mainly to the following:
- Protocol changes and tech upgrades
- Critical bug and vulnerability fixes
- Using pooled funds for R&D
It is possible for decentralized governance models to also be able to be applied towards solving many of the major issues plaguing humanity due to misinformation (global warming as an example). What I’m most excited and optimistic for is blockchain governance’s potential to foster the exercise of getting closer to truth as it acts as an infallible technical system that can be trusted in order for humanity to have a basic set of postulates that we can all agree upon. If a populace cannot even agree on basic facts, how can we hope to have a self-sustaining system that can continually evolve and better knowledge? When governments have the ability to spread misinformation and magnify these effects on centralized social networks designed to feed the masses their daily dose of dopamine, the prospect of decentralized governance models that can reverse the status quo gives me great optimism. Here’s a list of implementation ideas I’m excited about. For brevity I’ve included links to other articles that have already done a fine job summarizing each of these approaches:
- Futarchy — Participants in a system decide its values and those with the most info stake their ideas by betting on the outcome.
- Decentralized Autonomous Organization (DAO) — Allows for group governance through a combination of smart contracts and issuing tokens.
- Liquid Democracy — System where everyone can vote for themselves or delegate their votes.
- Quadratic Voting — System of buying votes where each additional vote costs twice as much.
The pitfalls of the decentralization hype
While many of these future approaches show promise, there are plenty of current blockchain governance models we can look at for insights. In one of Vitalik’s post on blockchain governance, he goes into a lot of the pros and cons of on-chain voting (layer 1) and off-chain voting (layer 2). Tezos and Dfinity are often cited examples of on-chain voting in which anyone can submit changes to implement protocol upgrades. In these cases, votes are automatic and the protocol has all of the logic to implement these changes after reaching a pre-programmed quorum. On the other spectrum lies, layer 2 in which coin voting is loosely coupled in that any upgrades to the protocol that were approved are applied but the participants in the ecosystem (Miners, Developers) have the choice of whether to upgrade their software to be compatible with this new upgrade. Although it sounds enticing to be able to implement a system in which people from all over the world do not have to interact and can self-improve a system rapidly simply by voting and having code automatically take care of previously very bureaucratic processes, I have to strongly agree with Vlad Zamfir’s post from Ethereum who argues:
“In which I argue that “tightly coupled” on-chain voting is overrated, the status quo of “informal governance” as practiced by Bitcoin, Bitcoin Cash, Ethereum, Zcash and similar systems is much less bad than commonly thought, that people who think that the purpose of blockchains is to completely expunge soft mushy human intuitions and feelings in favor of completely algorithmic governance (emphasis on “completely”) are absolutely crazy, and loosely coupled voting as done by Carbonvotes and similar systems is underrated, as well as describe what framework should be used when thinking about blockchain governance in the first place.”
There’s a reason bureaucratic governance processes have survived for so long and until we reach the future Isaac Asimov envisioned where AI can better make rational decisions for us, we have to accept that with any blockchain governance approach, how people will interact and utilize this technology is paramount.
Voter participation
Ultimately, all of the different blockchain governance models revolve around different approaches to voting. With that in mind, we can look at historical voting trends among various countries, foundations, boards of directors, etc. Even comparing the U.S. presidential voting turnout which we don’t typically think of as crazy high participation (~ 55% of eligible voters) with that of the DAO Carbon vote (4.5%) we can see a stark contrast. Granted this is only one example and it would be another interesting post to create a comprehensive list of voting statistics among all of the blockchain governance platforms and slice/dice that data compared to the different segments of current political voting (gender, income, education level, etc.). Despite the large contrast in voting numbers, one common theme in both of these voting examples lies in that they both exhibit elements of the classic tragedy of the commons economic problem where participants in an ecosystem do not have enough incentive to perform an action in the best interest of the majority. This unfortunately results in unintended adverse effects. Sadly, if you interview each of these participants individually, many of them in hindsight would lament not participating given the outcome they could have prevented. Suffice it to say, voter participation is a problem in the blockchain space just as it is with nation-states.
Another problem in voting lies in how the youth’s attention (the exact segment more likely to adopt emerging technologies) is split in an era where they are flooded with irrelevant information. There’s only so many “give a shits” people have in their life, and what’d I’d like to see in blockchain governance is the implementation of the best cultural management lessons to bake in cultural mechanisms that can enhance this initial wave of decentralized governance ethos. Whether it be formal governance or informal governance (core founders/dev teams influence on the blockchain community) as is the case with many of the blockchain communities today, we are dealing with people, and thus the right leadership and cultural values need to be properly espoused in order to spur the vision of blockchain governance becoming a mechanism of improving knowledge and outcomes. Participants in any system need to have skin in the game and just because decentralization has an effect of disintermediating current processes, it doesn’t mean it can’t be organized and productive.
Lessons from the Swiss Code
In the spirit of organization and productivity, we can look to the best practices of the “Swiss Code” which was originally published by economiesuisse for all sectors of the economy in July 2002 and continually amended since then. At a high-level the main mission statement is as follows:
“Corporate governance encompasses all of the principles aimed at safeguarding sustainable company interests. While maintaining decision-making capability and efficiency at the highest level of a company, these principles are intended to guarantee transparency and a healthy balance of management and control.”
In a completely decentralized governance model without the proper levels of organization and leadership, community participation may morph into spam and troll wars which defeats the original purpose of good governance. Although the “Swiss Code” is a dense read, many of the common sense recommendations on properly incentivizing, organizing, and efficiently facilitating discussion are battle tested principles that have stood the test of time. One of the major insights reading through the Swiss Code is that organizations that structure their board of directors in a manner where there is a high degree of skin in the game will of course have very high levels of voter participation. This is such an obvious statement but it’s the key difference in blockchain governance which opens the door for engaging and powerful governance.
Lessons learned from Dash and other DAO’s
Dash
One specific decentralized governance model that I’ve had the opportunity to work with is Dash’s governance model which utilizes 10% of the block reward towards a community fund which can then be applied on a monthly basis towards projects that the community proposes and votes on. For some context, Dash operates as a proof of work coin similar to Bitcoin but differs largely in that it created a 2nd consensus layer called the Masternode network which enables secondary features such as instasend, privasend, and governance. To become a Masternode, one must stake 1,000 Dash (~300k USD as of today) in order to reap the rewards of voting on proposals and receiving a dividend for securing the secondary features. While many consensus purists have pointed out that the network of ~4,700 Masternodes is not truly decentralized, it’s an interesting trade-off that has created the right amount of skin in the game motivation to drive more meaningful governance discussions.
Full disclosure, I’ve been working on a team that is focused on enhancing the proposal generator experience as the current tools are disjointed and can be bettered by applying the best practices of UX design. During this process I’ve been able to read through many proposals and discussions and noticed that the general categories of proposals to date generally fall under: Outreach/marketing, Events, Creativity (films), and Technology. In the Hacker ethos of moving fast, many of the tools and processes have been created and evolved by the community very rapidly in an effort to filter good proposals and community feedback. The staking effect of the Masternode network seems to have had a positive effect on voter participation as I’m observing an average of 25% participation; however, the crux issue that is arguably the issue for most DAO models, is ensuring that informed and quality feedback is being provided for each proposal.
Part of creating a DAO (Decentralized Autonomous Organization) that has meaningful participation, proposal ideas, and feedback involves establishing a strong cultural set of norms around being professional, articulate, and informed. In that matter, the Dash Core team has done an excellent job establishing that culture as well as providing quite a bit of documentation and training material on not only how the platform works, but also what sort of vision and culture they want to live on well beyond their involvement. Originally, the Dash founders sought to democratize the decision making process of bug/code bounties but quickly realized that applying a Zappos like Holacracy process could spur bottom-up innovation in other aspects of the organization.
What’s interesting with a strong core team is that they invariably garner a lot of informal power, and as a result, anytime they throw their approval into a particular proposal, it becomes a sort of stamp of legitimacy which makes it easier for other voters to follow. On the flip side, there are also problems where those that are voting on a technology proposal as an example, may not be career technologists themselves and thus look to other technology influencers in the community for guidance. To me, the challenge of how to get the most qualified and best informed participants to weigh in on the right proposals is the key to making decentralized governance lasting. This challenge is not new as the founding fathers of America sought to answer this same question to avoid mob rule. I would be interested to see potential tweaks to DAO designs in which voting power is not only a % of how much of the token you own, but also a % of how much reputation on a particular area of expertise you’ve accumulated. How to define that reputation system and build data around how to measure “expertise” will be a fascinating design and data science problem that will have to be tweaked heavily in order to avoid the pitfalls of many other reputation systems that can be gamed easily.
Other DAO implementations
Other DAO implementations include SmartCash which has a similar model as Dash but differs in two key ways:
- 70% of the mining rewards go towards community projects as opposed to 10% in Dash.
- All holders of SmartCash can vote and their voting power is based on the % of coin they hold whereas Dash currently only has Masternodes voting.
With a much larger pool of resources and wider community, the types of proposals are more varied and are skewed towards more international outreach use-cases such as bringing SmartCash to Nigeria. Over time, local and regional communities will likely self-organize their own DAO’s similar to what Kore.life is doing.
While we are in this phase of governance experimentation, it’s important to establish the right cultural norms as people are often conditioned to follow traditions just for the sake of tradition. Whether you’ve been in a Fraternity or were one of the founding fathers of America, traditions span the gamut of well thought out constitutions to random/silly acts of passage. Keeping in mind that irrelevant or outdated traditions are difficult to change once in place, we should take advantage of this period of time to create governance models that have lasting core tenets where change is a key tenet (establish a tradition to be able to amend traditions — oxymoron I know).
In addition, we should also incorporate the wealth of research and best practices found in companies and foundations that have studied how best to motivate and engage teams. For example, project aristotle sought out to research what are the common elements of effective teams at Google. One of the key takeaways being that it is more important to assemble a team that fosters psychological safety rather than to assemble an entire team of the top Type A individuals. An eclectic team that fosters this camaraderie can benefit from the innovative ideas that come from brainstorming sessions where ideas may not be as coalesced; yet other members may be able to take that vision and flesh it out. On the other hand, a team full of Type A individuals may exhibit troll like behavior where they become afraid to share any brainstorming ideas for fear of losing their social status. If we are to organize and empower decentralized communities to create value in their respective communities, we need to also take these social lessons and apply them in decentralized governance models.
A balanced approach
Whether we’re dealing with a potential blockchain startup idea or a decentralized governance model, we need to evaluate the use-case we are trying to solve against key criteria to determine how centralized or decentralized we should design our system. In the case of deciding whether to implement a blockchain solution for a supplier network, a set of criteria involves what the degree of trust is in this particular ecosystem and how fast this network needs to have data updated. For networks with very low trust and not needing real-time updates, a decentralized blockchain solution is a perfect fit. In other cases where coordination amongst many fragmented parties is not needed and real-time processing is a must, a central database hosted on something like AWS is the way to go. With this criteria in mind, here’s a potential future industry landscape:
In the case of governance, there will likely need to be a spectrum of on-chain voting, off-chain voting, and all manners of hybrid combinations depending on the primary use-case. Pure on-chain tightly coupled voting can play a role in situations where fixing small bugs as quickly as possible is the priority and thus adding bureaucratic processes is unnecessary. On the other hand, if there are paradigm shifts to evaluate and vote on such as moving from Proof of Work to Proof of Stake, the effects of such a hard fork are far reaching amongst participants and thus needs a comprehensive consensus mechanism in which informed participants can debate and craft a solution that best fits the needs of the ecosystem.
At the end of the day, blockchain technology is an incredible tool and means to an end rather than the end itself. We shouldn’t nod our heads agreeably when someone states they want to be the decentralized version of x. Instead, we should be asking how far on the spectrum and why. The tug of war between centralization and decentralization is nothing new and a result of mankind’s nature to centralize power. In many respects we are building governance systems that will transcend any one individual. So while exciting technical breakthroughs in blockchain are happening, we need to also in parallel apply the best governance and cultural practices that history and other disciplines have to offer.
Thanks to Jeff Hodges, Lillian Wei, and Mitchell Lee for editing this post.
