Yak Online Governance Primer
* Primer: The Yak Online Governance Primer. Venkatesh Rao et al. Yak Collective,
URL = https://www.yakcollective.org/projects/yak-online-governance-primer.html
Description
"How do you do online governance? This primer is intended as a guided tour through a curated set of readings—based on a year of study by the Yak Collective*—that can help groups and organizations navigate this question. In selecting the readings we cast a wide net, but in our discussions we made an effort to consider them from the specific perspective of online governance challenges. We believe the ideas surveyed here are applicable to groups and organizations with widely varied purposes, levels of autonomy, degrees of decentralization, and technological sophistication."
The approach:
"Starting in mid-2020, a Yak Collective study group has been meeting weekly to explore the question of online governance, one reading at a time. This paper is based on 49 readings we studied in our first year. The full list is included in the Annotated Bibliography section at the end of this paper.
At one extreme of the range of answers we find organizations that seek to evolve their traditions in minor, cosmetic ways; they run the risk of digital transformation being pure theater.
At the other extreme, the adoption of genuinely radical organizational forms such as DAOs (decentralized autonomous organizations) forces the deep redesign of traditional governance mechanisms. Such innovation leads to the risk of failure of initiatives that might have succeeded had they been organized along more conventional lines.
We believe this primer will be of value whether you’re part of a mature traditional organization just beginning to develop significant online operations, perhaps due to the impetus of the Covid pandemic, or a web3 native decentralized organization just starting out, propelled by revolutionary spirit.
The readings surveyed are not meant to be exhaustive or even representative. Our goal is to teach you to fish in the waters of online governance traditions for yourself. To the extent that we succeed, after reading this primer you will have developed a basic literacy around the topic, and an awareness of several major trailheads for further exploration.
While we have not entirely avoided conventional sources of wisdom on management and organizations, such as academic management literature, we have deliberately and consciously cast a much wider net. Our Year 1 readings ranged from academic papers and excerpts from classic books to corporate presentations and blog posts. We read about Paleolithic farming cultures and medieval guilds, and about modern open-source movements and platform ecosystems. We read thought-provoking bits of fiction, sampled manifestos, and even discussed essays about biology and wildlife management.
This primer does not include readings that explicitly discuss blockchain-based governance models such as DAOs. While these are both a current focus of study for our group and a possible future governance direction for the Yak Collective itself, we feel a thoughtful and critical appraisal of received traditions of governance is a necessary prerequisite for getting the most out of the rapidly emerging blockchain-focused literature. While blockchain-aware readings will likely feature prominently in a future paper, our intent here is to distill the best of the pre-blockchain past.
As we worked our way through the readings, we began developing an idiosyncratic internal lexicon for talking about online governance, drawn both from the readings and our own discussions. We believe this lexicon, a subset of which can be found at the end of this primer, has helped us level up the sophistication, immediate practical utility, and interestingness of our discussions. We encourage you to use our lexicon to jumpstart your own. We also welcome your suggested additions to ours.
The remainder of this primer is organized as follows. In the next section, we provide overview commentary on the 49 readings, organized around a conceptual map of four governance regimes. We then briefly discuss the challenges of synthesizing an online governance strategy for a particular organization from this universe of disparate, sometimes contradictory ideas. Finally, we offer some notes on our own experiences, along with suggestions for using this primer to guide your own further explorations."
Typology
of governance regimes surveyed:
"We identified two principal axes that seemed the most helpful in sorting the readings:
- from low to high alignment in terms of the intentions and interests of the participants
- from managed to wild in terms of the structures and processes of interaction
In the following sections we discuss each of the four regimes in turn, roughly in order of strength of governance forces:
Hobbesian: Governance ideas that assume wild defaults and low alignment, and attempt to foster progress despite conflict and chaos (readings 1–11)
Gaia: Governance ideas that assume wild defaults and high alignment, and attempt to foster progress by drawing on natural patterns and harmonies (readings 12–26)
Muddler: Governance ideas that assume managed defaults and low alignment, and attempt to foster progress by adding some process structure (readings 27–33)
Citadel: Governance ideas that assume managed defaults and high alignment and attempt to foster progress through top-down coordination (readings 34–49)
Hobbesian Regimes: Wild and Low Alignment
The Hobbesian regime, the lower right quadrant of the 2x2, is the least governed regime and can be explored through readings 1–11.
The Hobbesian governance regime is personified by the archetype of the anarch, explored in the writings of Ernst Junger. The ideal anarch is an individual whose identity cannot be tied to an organization or ideology, and whose actions and principles are driven by a pragmatism that serves both him and the common good.
Hobbesian groups tend to be relatively low in energy and cohesion. As a result, a functional Hobbesian regime takes time to build and faces many risks along the way. One major risk is of individuals being isolated, scapegoated, and persecuted. Junger himself was persecuted by both the Allies and Nazis through World War II.
Early in its history, a wild and low alignment governance regime has to solve for trust and common knowledge. Trust is generally low among individuals who are suspicious of prevailing ideologies and exhibit an inclination towards doing their own research. Common knowledge [2] is needed for a governance regime to be functional and Hobbesian regimes typically suffer from a strong deficit.
Many online communities limit their goals to adopting tools for gathering and communicating and fail to work on building trust and common knowledge, making them vulnerable to Hobbesian failure modes. As a result, a key risk of wild and low alignment organizations is that emergent adversarial behaviors can destroy them. A symptom of such destruction in progress is the presence of multiple charismatic figures competing for influence, especially in heavily politicized contexts. The Intellectual Dark Web, QAnon, and the Occupy movement are good examples from the culture wars of the past decade.
The town of Grafton in New Hampshire [4], which was taken over by Libertarians over the last two decades, is an example of an organization that began with a goal of being in the Gaia quadrant (wild and high alignment) but turned Hobbesian due to lack of alignment with existing residents of the town. Unmanaged emergent effects—bears running amok in Grafton’s case—can be traced to insufficient levels of institutionalized common knowledge.
A more enlightened way to handle Hobbesian conflict might have been to take the approach of Musical.ly founder Alex Zhu [5] who analogizes joining a new social network to moving to a new city—“come for the utility, stay for the community.”
Hobbesian patterns of governance tend to work during the early years of a subcultural scene or organization, but tend to fail as they scale. David Chapman [6] attributes this to an invasion of people seeking social status, or worse, seeking to exploit the chaos for personal gain. This particular pattern of hostile entryism in young Hobbesian organizations or scenes has co-evolved rapidly, over the last century and a half, with technologically mediated mass culture. Anarchists in early 20th-century China, for example, were infiltrated by communists operating within a more effective top-down structure [7]. Operating relatively under the public radar (a strategy sometimes called security through obscurity) can help protect a Hobbesian group from being overrun by sociopaths or invading ideologies.
One way to mitigate such effects is to simply eschew growth and the hierarchical structures it tends to induce. Contemporary organizations have inherited the twentieth-century bias toward growth for the sake of growth, typically measured through metrics such as number of members, revenue, and impact on the zeitgeist. Eschewing growth, however, comes at a cost. Nonhierarchical cultures tend to be poorer, as Sarah Constantin [8] points out in an essay on the relationship between hierarchy and wealth. Hierarchy is expensive because it requires systems and people to manage it. It also tends to incentivize the creation of wealth to pay for itself.
Emerging forms of nonhierarchical organization could potentially offload many of the traditional functions of hierarchical structures to low-cost automation, allowing relatively Hobbesian organizations to cohere and persist at larger scales. Web3 technologies such as DAOs are a development worth watching in this evolutionary direction.
Hobbesian regimes work sustainably if individuals manage to arrive at a high level of shared common knowledge and a shared understanding of the common good before they are undermined by the many risks. More commonly, however, they transition to one of the other three regimes depending on which internal dynamics predominate:
If a Hobbesian organization produces enough common good that effective ongoing management is induced, it starts to have the characteristics of an organization with Muddler governance characteristics.
Where charismatic influence drives higher alignment, a Hobbesian group can transform into one with Gaia governance characteristics.
Where a more organized, hierarchical group invades and takes over at the top, it can turn into a group with Citadel governance characteristics.
Gaia Regimes: Wild and High Alignment
The Gaia regime, the upper right quadrant of the 2x2, is the second-least governed regime and can be explored through readings 12–26.
Gaia, the Greek goddess who personified the Earth, also personifies the fifteen readings in this section. The goddess Gaia was the inspiration for James Lovelock’s Gaia Hypothesis [12], which holds that living things adapt to and transform their environments. Lovelock speculated that the complex process of wild coevolution implies that ecosystems are always optimizing for the continued existence of life. Groups of people who gather in pursuit of a common goal with wild sensibilities can be said to constitute a Gaia governance regime.
The overarching governance question in a Gaia regime is this: will the strong tendency toward organic coevolution make the organization resistant to any form of designed structure, even when it improves functioning or addresses possibly fatal weaknesses? In contrast to the Hobbesian regime, which shares the basic suspicion of formal structure, the presence of high alignment makes different outcomes possible.
It is important to understand the distaste for designed structures that drives organizations in the Gaia quadrant. Ivan Illich, arguably a Gaian philosopher, exhibited a particularly refined form of this distaste. Illich’s work was marked by a tension between progressive, libertarian, and anarchist impulses.
Dave Pollard offers an accessible overview in his post, Ivan Illich: The Progressive-Libertarian-Anarchist Priest [13]. Illich’s work can be understood as a response to the designed institutions which make up modern societies, especially in fields like education and medicine. These institutions are a result of the drive to organize large groups of specialists in industrial modes, with a narrow focus on metrics such as productivity. But there is a dark side to this process: the progressive erosion of individual dignity.
Jerry Pournelle’s Iron Law of Bureaucracy [30] suggests that industrial-mode institutions are doomed to eventual capture by cults of expertise, which leads to cartel-like organizational behaviors. Complex education and training pipelines emerge, to sustainably produce the narrow specialists needed to perpetuate the captured condition. The result is progressive loss of dignity for all participants.
In an effort to reclaim their dignity, individuals often gravitate to anarchist ideas as a response to the burdens of institutional life. Jo Freeman’s essay The Tyranny of Structurelessness [31] outlines what can happen next. First, informal structures will emerge, with unclear norms and unwritten rules. The authority within the group will be concentrated to a select few elites and status games will ensue as individuals in the group jockey for position. Once this happens, the group loses focus on its original goals, making progress next to impossible.
The challenge of the Gaia governance regime is to overcome both the tyranny of systems of control and the tyranny of structurelessness. If individuals can agree on both the mission and the means of accomplishing that mission, they can then decide if they’re willing to give up a certain amount of freedom in order to reap the benefits of being in the group. This alignment is often achieved through a temporary period of charismatic leadership, though such leadership risks becoming a benevolent dictator for life [36].
The most important task in governing a Gaia organization is to clearly define measures of success in a way that manages the tension between the goals of individuals and the goals of the organization. When the tension is too great, imbalance threatens the stability of the organization. Another advantage of defining the measures of success is that it provides a simple criteria to filter potential additions to the organization. Once everyone has an idea of what success looks like, it is possible for individuals to understand how they fit into the bigger picture.
The case of Morning Star, a large tomato processing company, illustrates what wild and high alignment conditions look like [14]. Each Morning Star employee creates a colleague letter of understanding (CLOU) which outlines their personal responsibilities and means for being held accountable. Their CLOU is renegotiated every year. Each person in the company gets an opportunity to define how they contribute to the company’s success.
Transparency is a key enabler in Gaia governance regimes. GitLab [15] models itself after open-source software projects. With employees distributed across the globe, the company has to impose some structure to induce order in the chaos. The GitLab employee handbook stresses the importance of documenting decisions, communicating in public spaces, and breaking work down into the smallest pieces possible. Combined, these elements provide individuals with a holistic view of the company’s operations and their own role within it.
Gaian organizations are sensitive to the limits to structure. Things will not always work as intended. Even the most aligned teams will have dissent. In their famous “culture deck,” Netflix [16] recognizes this and provides employees with a process for resolution: they must be willing to voice their dissent and be able to articulate why they disagree. An informed captain then reviews the issue from all sides and makes a decision. This captain, who often will have to tease out these frustrations, documents the decision for review by the entire organization. Captains are trusted to make informed decisions and don’t need consensus to move forward. The team is expected to rally around the final decision so that the outcome is as successful as possible.
Gaia governance regimes tend to discriminate against candidate members who don’t have a shared understanding of organizational values. In their Handbook for New Employees, Valve [17], for instance, asserts that their Gaia model is scalable, as long as they remain particular about the people they add to the company. Making the wrong hire can be a particularly expensive mistake in Gaia regimes: either you let a promising potential hire get away, or you miss key warning signs and a new team member wreaks havoc on the organization.
Many general principles are revealed by these specific cases. Gaia organizations delegate tasks and authority and demand strong commitment from individuals in return. With such delegation come responsibilities for meeting criteria for success. Over time, individual responsibilities must be switched around to avoid the hoarding of knowledge. Information must be made freely available. All decisions must be documented to provide context to the rest of the team.
When these principles are successfully adopted and practiced, organic coevolution is possible, and the system can operate with high energy and tempo. Where they fail, Gaia organizations can drain energy and migrate into the Hobbesian quadrant, through unmanaged dissent and unraveling alignment. Or they might lose variety and distributed autonomy, adopt stronger, more overt organizational forms, and move into the Citadel quadrant.
Muddler Regimes: Managed and Low Alignment
The Muddler regime, the lower left quadrant of the 2x2, is the second-most-governed regime and can be explored through readings 27–33.
The Muddler governance regime represents a condition of shared and commonly acknowledged ignorance rather than common knowledge, with expectations set accordingly. The critical insight regarding the Muddler regime is that self-deprecating humility and a sense of humor in approaching decentralized orgs is a superpower.
When a group sees itself as ordinary people muddling through in ignorance, doing their mediocre best rather than as Chosen Ones constructing a utopia, things are seen in realistic proportions. The seven readings of this section help foster and anchor the attitudes necessary to govern in the Muddler regime.
The quadrant label comes from Charles Lindblom’s 1959 article, The Science of Muddling Through [27]. In it, Lindblom identifies the method of successive approximations (which bears a strong resemblance to modern agile management methods) as a characteristic of successful organizations. The muddling-through organization is the opposite of the efficient, machine-like organization. Frederick Laloux [28] characterizes this condition in terms of loosened optimality criteria, and heightened appreciation for the benefits of “fatter” unoptimized systems. An example can be found in the interview with Tobi Lüttke, founder of Shopify [29]. The key is understanding the organization as a complex system, and being non-deterministically in harmony with, and attuned to, what’s going on. But this posture must also be oriented toward a purpose, and resist simply surrendering to the system’s natural evolutionary tendencies.
You can tell you’re in the Muddler quadrant if you seem to be making progress, but in a confused, near-random-walk way, with many misunderstandings between people due to misalignment at the level of information rather than values. Over time you notice net positive movement emerging. Forbearance and patience achieve a lot. There is a sense of inefficient and sloppy relentlessness. Knowledge retention and transmission will be lossy, naturally producing the muddling-through process and killing any attempt to do a rational planning process.
A key risk in the Muddler quadrant is bureaucratic capture. Jerry Pournelle’s Iron Law of Bureaucracy [30] and Jo Freeman’s Tyranny of Structurelessness both argue that “management” effectively emerges, for better or worse, even if things look unmanaged. Anarchy in the sense of chaos is unstable.
The tempo of the muddling-through regime—staccato stop-go janky progress—is captured by two short readings. The Hurling Frootmig principle—derived from a passage in The Hitchhiker’s Guide to the Galaxy [32]—states that most work gets done by random people wandering in at lunchtime and seeing something worth doing. The Wind in the Willows principle [33] asserts that people can vanish abruptly and reappear at any time, and that the system should be able to make use of their unpredictable availability anyway. These two principles suggest a key tension between being, on the one hand, open to the serendipity of creative contributions from unexpected new participants and, on the other hand, forgiving of unreliable and unequal participation by existing participants caused by the uncertainties and resources limits of individual lives. This idea harmonizes with what’s become known as Postel’s Law, from The Tao of the IETF [35], which advocates, “be conservative in what you send and liberal in what you accept.” Applied to governance, muddling through requires being conservative in what you police and liberal in the patterns of participation you accept.
Muddling through is a low-energy condition because there is a lot of acknowledged uncertainty and decisions/actions happen despite this uncertainty. Time is spent in experimentation and rework, as well as in sorting out tactical confusions and dealing with shifting patterns of participation.
At the Yak Collective, the Muddler regime tends to be the default quadrant. Other quadrants are inhabited by exception. For example, within a well-defined project we might drive up to Citadel or Gaian levels of energy or alignment; around a contentious issue we might briefly inhabit the Hobbesian quadrant. But much of the time we are muddling through.
What makes the Muddler regime a stable place is that things actually tend to live up to expectations based on good-humored cynicism. The challenge, though, is that these expectations are typically low. The Muddler regime can yield somewhat desultory, stop-go progress. Without spikes of more energized action, staying permanently in a Muddler regime can equal a slow death.
Citadel Regimes: Managed and High Alignment
The Citadel regime, the upper left quadrant of the 2x2, is the most strongly governed regime and can be explored through readings 34–49.
The sixteen readings in this section revolve around organizations that are usually discussed in terms of platforms and ecosystems.
Platforms and ecosystems are ubiquitous today, but are still confusing even for their participants. An ecosystem can be defined as “a dynamic group of largely independent economic players that create products or services that together constitute a coherent solution” [20]. The significant questions in the Citadel quadrant revolve around why and how exactly these independent players can collaborate, what keeps them together, and how their products and services compete.
One of the problems most platform builders face is that intuitions developed while running a traditional organization frequently fails them. For example, hoarding power and value is often adaptive in traditional organizations but can drive platforms to failure. Intelligent ways to distribute power and value can be hard to discover.
Metaphors derived from traditional organizations can be misleading as well. For example, the name Citadel selected for this governance regime, while generally reflective of the underlying aspirations, implies strong walls to protect an ecosystem or a marketplace from external threats. While the metaphor highlights some of the logic of this regime, it misses a key point. Ecosystems and marketplaces need dynamic, rather than static, protection. It is their openness that makes them stronger. What needs protection might be an economic engine defined via network effects, rather than a geographic perimeter or traditional market boundary. It may not be possible to hide such an engine behind walls. And unintended consequences from network effects can break even the strongest of walls.
Threats can also emerge internally in a citadel ecosystem. For example, there is a tension between attempts to ensure that hierarchical control is tractable and the need to respond to novel circumstances in novel ways, leading to the proliferation of new rules, new teams, and unconventional ways of working.
These forces operate at different tempos. Hierarchical forces operate on linear time scales, according to designs and plans. Emergent ones operate on a more organic tempo, characteristic of the the organizations discussed in the Gaia quadrant. These emergent forces are typically weak in the beginning, but can strengthen rapidly as the organization learns. Uneven tempos are a feature, not a bug, but the hierarchical regulation forces that they trigger can fail by being too slow or too fast. The most likely outcome of such misregulation is a ghost town, where serendipity has been squashed.
When this tension is palpable, it is a clear sign that you’re in the Citadel quadrant. It is precisely this tension that produces ecological surprises. An ecological surprise is a turn of events that can’t be predicted based on traditional logic. This happens when people are either wrong about the future, mismanage the system, or discover something surprising about it.
Surprises are inevitable. You can either fight them or learn from them. Some organizations successfully turn this process of constant discovery into an element of their economic engines. In such cases the creative tension drives what can be called “a serendipity engine.” An example is the Pinduoduo marketplace [34]. Individuals with well-developed platform thinking mindsets are able to see most surprises in a serendipitous light. But for people with strongly traditional hierarchical mindsets, every surprise can seem like a threat.
Ecosystem governance is about nurturing and protecting network effects by responding appropriately to surprises. But there are few universally applicable governance principles. Due to winner-take-all effects, every successful example tends to be one-of-a-kind. Lessons from specific examples tend to be hard to generalize.
As a result, attempts to accurately describe what’s going on inside a sufficiently developed citadel ecosystem usually fail. You can, of course, study the principles sincerely articulated by insiders, but they are difficult to port to other settings.
Principles derived from successful examples like the IETF (an organization with both Citadel and Muddler characteristics), make complete sense only in the context of the original circumstances. Furthermore, these circumstances are not static, but change with the development of an ecosystem. For example, the authors of The Tao of the IETF [35] acknowledge that the principle “the IETF recognizes leadership positions and grants power of decision to the leaders, but decisions are subject to appeal” only makes sense in the context of specific historical details, but the details themselves are meaningless to outsiders, making the principle hard to port to other contexts.
Due to the uniqueness of successful platforms, the only general principles available are broad-strokes ones. One such principle can be found in a speech delivered in 2014 by Frank Chimero [36].
In it, Chimero contrasts the cases of managing wolves in farming regions of the Western United States and managing bear populations within Yellowstone National Park. In the former case, which played out early in the history of the region, the interests of the organized ranching industry led to a failed attempt to manage the wolf population through large-scale slaughter, aimed at eradication. In the latter case, a similar problem involving the bear population was successfully addressed through investment in long-term sustainable population management processes. Chimero uses the two cases as motifs of two very different approaches to problems in complex systems emerging from the collision between structured interests and wild ecosystems: “shooting the source of the problem” versus “investing in a process to keep things open and adaptable.“
This is perhaps the essence of managing Citadel-like systems that attempt to create islands of order and civilization within essentially wild ecosystems."
(https://www.yakcollective.org/projects/yak-online-governance-primer.html)