A New Social Decentralization Scale

A New Social Decentralization Scale

Category : Society 5.0

When you think about it, we still live in a medieval society.

There’s an elite layer of society where decision-making and wealth are centralized, and a much, much larger layer who’s responsibility includes obedience, conformity and producing value to will be managed by the centralized layer.

Of course, the terminology has changed, and we have gained a few more liberties over the centuries. But we still essentially work for the benefit of our lords (corporate lords rather than nobles), and while we vote for our representatives, they will have a few years of ruling with little accountability. Our role is still to execute, like good little serfs.

We understand that this is how the world works, and there’s little we can do about it. After all, we all have families to feed, we all have to work long hours, we all have to follow the rules that are there for our own good. Right?


Now, more than ever before in history, we have the tools to build the society we need. More than that, we HAVE to do so, as we are on the cusp of a new social era, Society 5.0, the age of AI. It’s not even a matter of choice anymore. Business as usual just won’t work.

What we need is to redesign our society to provide us with options, with means to create more awareness for us, more agency, more prosperity.

This is why Hivernité was founded. We are a think tank dedicated to exploring the building blocks of those options, so people all around the world can use them creatively and figure out how to create the society we need. Alone, our power is limited, but together there’s little we can’t achieve.

Changing the world involve shifting the distribution of power

Inour society, like in any medieval society, the power of decision is centralized, creating an imbalance of power and meaning that a small group of people is calling the shots.

If we want to be more than mere executing drones, we need more power over our world. I’m not talking about shifting the balance, because it would just move the problem. No, I’m talking about redistributing the power differently, from a centralized model toward distributed models. Models with an “s,” because there is no single model that’s appropriate for every situation.

Decentralization is one of Hivernité’s axis of reflexion. To be able to talk about it and create tools and options, we needed to have a standard, clear definition of it. There are definitions out there, for sure, but they are often confusing, and the same term can apply to vastly different models.

The Social Decentralization Scale

Wecreated a Social Decentralization Scale to gives ourselves a common point of reference. We want to share it with you today, both because we will refer to it in the future, but also because a standard frame of reference helps to foster productive discussions across often vastly different groups of people.

In a nutshell, we could define centralized models as models where strategy and decisional power are owned by a central authority and where the autonomy of the components is limited to execution.

By opposition, decentralized models are models where goals are decided by a central authority and where components are autonomous and can self-manage as they wish, as long as they contribute toward the goal.

What we see as we move from more centralized models toward more decentralized ones, is that people (us!) more from being mere components in the system toward being its active agents. We cease to be at the service of the system and the system start serving us.

We broke down the scale in four broads categories of models:


Acentralized model is centered around a strong leader (autocracy) or a small group of people (oligarchy) who hold all decisional powers over both strategy and execution. All other components of the organization exist to execute on those decisions and usually have little to no input.

True centralized models are mostly seen in small businesses, or organizations build around a strong leader. Startups, private clubs, and street gangs use these models.

As long as it remains small, a centralized model can be quite efficient, especially under a single decision-maker. Information can be relayed quickly, and decisions can be made at the drop of a hat. However, since the centralized control is essentially absolute, there is enormous pressure on the leader, with its limited grasp on the whole context, to make the right decision at all times. Bad calls are bound to happen, and with an entire organization trained to execute rather than dispute decisions, there isn’t a safety net to avoid the outcome.

While modern communication and information-processing technology allow a true centralized organization to operate at a much larger size than ever before, at some point, they will have to turn toward a delegated modelto continue to grow.


Often erroneously called a decentralized model, a delegated model(called devolued when discussing states or nations) delegate specific areas of decisional power over execution across the system following hierarchical model, but both keep some powers within an oligarchic group as well as a “last say” over delegated decisions, ultimately retaining the the real decisional power in centralized. Since decisional power is delegated and not transferred, decision-maker higher up in the hierarchy can invalidate decisions made by lower-ranked decision-makers, or even take back delegated decisional power altogether.

Delegated models are commonly used in governments and large companies. Unwieldy size makes them difficult to control through centralization models and delegating execution can ease the process.

This creates bureaucracy. Delegates are allowed to make limited decisions, but even those need to have their scope clearly defined, so they work for the system rather than against it. To help guide them, regulations are enacted, dictating how things should be decided. In turn, these regulations lead to replacing human-made decisions with processes, considered more efficient because they are pre-made decisions. The less trust there is within an organization, the more regulations and processes will be present, and the more cumbersome the bureaucracy will become.

The main problem with bureaucracies is that they are systems in themselves and quickly become self-serving. Processes remove human initiative and agency, and human agents become at the service of those processes rather than the other way around.

Now, as a result of bureaucracy, large centralized organizations are slow to react and adapt. There’s just too much information to process to quickly form plans or adjust them on the fly. They can be very efficient, but only as long as they exist in an unchanging environment: To operate, they require predictability, which is the ability to accurately predict how things are going to be in the future. When the context is very stable, it is possible to extrapolate on the present to predict the future. But if the context is volatile, uncertain, complex or ambiguous (see VUCA) which, let’s be frank, describe most contexts nowadays, predictability starts looking increasingly like sheer divination.

Delegated models can appear decentralized, but ultimately the true decisional power is still held within an individual or oligarchic group. The rest of the decision-makers end up mired in bureaucracy, where regulations and processes have more power and influence than they do.

Coupled with management through predictability, this reliance on processes and regulations directly leads to a reduction in human agency and autonomy. People get turned into cogs in a machine that need to run in a perfect way to stay predictable, even if both that perfection and the accuracy of the prediction are nothing but delusions.

Again, let’s keep in mind this is the context in which most people are working and being governed.


Adecentralized model involves a weak center providing guidance rather than control, and strong autonomous, self-managed nodes where the majority of the decisional power resides. The center (government, upper management, etc..) determines the general direction of the organization and states it as a clear vision supported by high-level goals. Nodes align with the vision, making all the necessary decisions themselves to works toward the goals while also keeping themselves relevant.

Decentralized models trade control for quickness, adaptability and resilience.

Decisions are made closer to the problems, allowing for more relevant solutions (solutions are punctual, not organization-wide regulations) that can also be implemented faster since any approval occurs at the local level.

Decisions, design, and planning happens very close to the people who will consume the value created by the organization, allowing for more discussions and ultimately an offer that’s more relevant the needs and realities of the client.

Without a clear division of labour between departments, self-managing nodes need to develop a broader range of expertise such as budget and workers management so they can function autonomously. They will have to constantly remain relevant to maintain their existence. Nodes unable to develop those skills or stay relevant will fail as the system cannot support them artificially. More autonomy requires more awareness and broader skills: you can’t just coast.

Without the need for a hierarchical structure, self-managing nodes will have to build direct relationships between themselves and find ways to cooperate effectively. This will transform the organizational structure from a pyramid and toward a net. Redundancy across the entire system forms, losing some efficiency but creating resilience. Some nodes can, and probably will, fail, but the system is resilient enough to continue to function.

This constant replacement of failed or irrelevant nodes isn’t a weakness. It’s a decentralized model way of continually evolving and refreshing itself naturally without necessarily needing conscious intervention.

Dave Gray’s Podularity

In his 2014 book The Connected CompanyDave Gray describe Podularity, which is a practical application of a decentralized model for companies. At its core, podularity is a decentralized model that adds a backbone to the nodes system (called “pods” here).

This backbone defines the rules of collaboration between pods (expectations, duties, available services, how information is shared, etc..). There can be more than one backbone within an organization, and pods can link to multiple backbones as needed. Backbones are in constant evolution driven by the interactions, needs, and contributions of the participating pods. Management in this model works through better through vision and goals rather than control, as in other decentralized models, but it still can actively participate in the dynamic backbone(s) by making sure its needs are taken into account in each backbone.

The beauty of podularity is its sheer adaptability and versatility as a model. The constant evolution of the backbones and constant discussion involving all parties keep the system current, fresh and adapted to the organization’s realities of the moment.

We also found podularity to be an excellent tool for the progressive decentralizing of an existing organization. Any organization or company can move toward such model at its own rhythm, starting with the groups more likely to succeed (Shameless plug: if you are interested in decentralizing your organization using a podular model and need some expert help, drop us a line at Moabi.xyz).


Adistributed model is built from autonomous, self-managed nodes who hold all decisional power over themselves, and who will cooperate and pool resources for endeavour where it makes sense either for an economy of scale or if additional skill sets are needed. Such a group usually have a theme or a set of principles to unite the group. If a decentralized model still uses a top-down approach, with vision and guidance at the top, a distributed model is a bottom-up approach with the only shared element adhered to through consent.

The main difference between a decentralized model and a distributed modelis the removal of the last remains of a centralized decisional oligarchy, moving all decisional powers directly in the hands of the constituting nodes. In a distributed model,

In addition to the precedence of local decisions over organizational decisionsthat we can also see in decentralized models, the lack of strong binds between the nodes in a distributed model means that any of those nodes as the option to opt-out or leave the group if they so wish. That level of liberty significantly eases the issue of having to find a compromise that agrees with everyone or to force some to accept a decision that will hurt them.

Distributed models also have a secret weapon: their loose nature and habit of cooperating in endeavour larger than a single node allow them to scale up very rapidly when needed by enlisting additional nodes in a project, and to scale back down as quickly when the need has passed to the satisfaction of all involved. A powerful example of this ability comes from the adventures of famous privateer Henry Morgan (yes, the rum brand is named for him) as he launched an expedition against Panama City in 1670–1671: from a crew of around 150, he assembled a flotilla counting over 2000 buccaneers, sacked the city, split the treasure, and then release his forces all in a matter of weeks.

Distributed models come in two flavours:


league keeps the bonds between nodes on a voluntary basis only. No organizational superstructure encompasses the entire league. Any collaboration between some of the nodes concern only those nodes which participate.

An excellent example of a league would be an association of professionals, bound only by a shared code of deontology. Apart from this code, they are no link between nodes, unless some actively choose to collaborate.

One of Hivernité’s upcoming projects, The Pyrate Model is build along those lines. In it, each node, called Crews (like a pirate ship crew), share an organizational model but remain entirely autonomous, and can come together to form Ventures, such as projects or companies owned by the participating crews. More information about The Pyrate Model will be available in 2019 in the upcoming book The Pyrate Handbook: community-driven entrepreneurship for the age of AI.


confederation is similar to a league, but with an organizational superstructure present. This superstructure hold shared tools, services or endeavours that concern the entire organization. In a confederation decisional power remains within the nodes, and all decision regarding the superstructure or the entirety of the organization is voted on by all the nodes.

The European Union is an example of a confederacy with a strong superstructure that regulates trade and movement between that participating states, while the Native American Wabanaki Confederacy has a much lighter superstructure designed around cultural activities, education, political advocacy, and activism. Superstructures can vary a lot between different confederacies simply because it is a tool designed to serve the participating entities in a specific fashion.

One of Hivernité’s upcoming projects, The Pyrate Haven is the creation of a confederacy for Crews using The Pyrate Way, offering community tools, shared services, and endeavours. The inspiration is drawn from the historical Republic of Pirate in 18th century Bahamas (which was, technically, a confederation rather than a republic).