The Rise of Autonomous Organizations: The End of the Middle Manager?
For our latest "Bucket List" tour, we traveled all the way to the other side of the world—specifically, Japan. We toured the country for about a month to visit a range of local pioneering organizations. One of our first stops of our Japanese tour was the city of Kyoto, where we arrived via one of Japan's famous bullet trains. Here, we visited the headquarters of Omron to learn more about their late visionary founder, Kazuma Tateishi, and his ideas about the end of the traditional hierarchical workplace.
Upon arrival at Omron's headquarters, and after the usual kind Japanese-style introductions to our hosts, we toured Omron's corporate museum, where we were introduced to the company’s most iconic product innovations and by some of the founder's most inspiring saying such as:
These quotes perfectly set the stage for the conversations we were about to have next with a range of Omron employees and executives.
One particular part of these conversations was a long discussion about Tateishi's visionary perspective on Autonomous Organizations with Shinichi Nakama, the director of Omron's Human Renaissance Institute (HRI).
Omron's Human Renaissance Institute
HRI is a research institute founded in 1990 by the company as a sort of think tank. It concerns itself with "exploring the future possibilities of society and technology with the aim of creating new social needs by looking beyond the traditional industrialized society."
Their aim is to explore new ways of organizing the lives of individuals, new ways of working for organizations, and new ways of organizing society as a whole.
They all do this based on the so-called SINIC Theory, as put forward by Tateishi.
The SINIC Cycle by Tateishi
Yuji Yamada introduced the theory based on the SINIC cycle in a previous article. He explained that the theory can be used by organizations as a tool to take a look at the future, and to think about how to adapt oneself for the future that is to come.
"SINIC theory was first presented at the International Future Research Conference in 1970 by Tateish himself, who was 70 at the time. SINIC stands for Seed-Innovation to Need-Impetus Cyclic Evolution.
According to the theory, science, technology, and society have a cyclical relationship. Scientific breakthroughs yield new technologies that help society move forward, and social needs spur technological development and expectations for new scientific advancement.
These factors affect each other in a cyclical manner, thus propelling further social evolution."
Updated SINIC Cycle by HRI
One of the first things Shinichi Nakama explained to us is that HRI had slightly updated the original SINIC cycle to make it fit the current times.
They argue that when Tateishi first developed the SINIC cycle, people tended to adapt themselves toward society. However, over the last few decades, people increasingly want to participate in how our society is structured and organized.
As a result, they observed that our attitude—especially regarding science—had changed dramatically.
This is extremely important. Due to the development of potentially highly disruptive technologies that have the power to fundamentally change and alter our society—such as Artificial Intelligence and Robotics—they argue that all human beings must be able to have a say in how science will develop itself.
These observations prompted HRI to tweak the cycle of how people, technology, science, and society relate to each other (see image above).
In particular, they are convinced that society needs to dictate science and its dreams—i.e., what do want science to develop for us.
But, perhaps even more importantly, it needs to give back to science its ethics—i.e., what can be done, and what can't.
The arrival of the Autonomous Society
Although HRI had tweaked the SINIC cycle slightly, this had not changed the overall predictions of the original SINIC theory. It still describes the evolution of humanity in 10 distinct stages. (Read Yuji's piece for more background.)
According to this theory, we are currently living in the so-called “Optimizing Society,'” but it also predicts that the “Autonomous Society” is arriving soon—particularly around 2025.
In fact, Shinichi Nakama argues that we are currently in a transition time and that over the coming years, the world will transition from a society dominated by “control” towards a society with no control.
What Tateishi and Nakama are predicting is an Autonomous Society with autonomous people that form Autonomous Organizations. (Interestingly, Nakama and Omron tend to speak about “Teal Organizations” as a synonym).
They predict that Autonomous Organizations can run without the need for “control.” But what do they mean by that?
We believe they point towards hierarchical forms of control—top-down control, i.e., superiors-controlling-subordinates controls.
So, they clearly predict a global transition from hierarchical ways of organizing to more autonomous ways of organizing. But what are these more autonomous ways of organizing?
Moreover, what is an Autonomous Organization in the first place? And how do they differ from the more traditional Hierarchical Organizations we are all used to?
To answer these questions, we can turn to Professor Alan Fiske and his work on what he calls “Relational Models.”
(Note: What you will read next is our perspective on the concept of Autonomous Organizations. It is does not necessarily reflect the vision, or ideas, of Omron, HRI, Tateishi, and Nakama)
Fiske's Relational Models
In 1985, Fiske, an anthropologist, published his “Relational Model Theory” (RMT), where he argues that there are four fundamental ways of relating, known as “relational models.” Since its publication, hundreds of researchers have used RMT in their own work.
RMT argues that we, people of all cultures, coordinate nearly all aspects of our social activities—including the coordination and cooperation that needs to happen in an organization to get work done—by these four fundamental relational models.
We use these models often implicitly and subconsciously. But we are intrinsically motivated to use any of these four models to generate, understand, and evaluate our interactions with others—including the interactions with colleagues in our workplaces.
The four relational models are:
- Authority Ranking
- Market Pricing
- Equality Matching
- Communal Sharing
Let us briefly explain the four elementary relational models.
Authority Ranking relationships describe asymmetrical relations in which people order themselves and others in a linear ranking built up by subordinates and superiors.
The primary feature of these relationships is whether a person ranks above or below each other person in the hierarchy. Those higher in rank, the superiors, hold formal authority, prestige, and privileges over the others.
Those lower in rank, the subordinates, are expected to respect and follow the orders of their superiors. Superiors, in turn, are expected to guide, lead and stand up for their subordinates responsibly.
As we all know, this is the default way employees tend to relate to each other. The majority of our organizations, especially commercial businesses, tend to organize themselves as linear hierarchies. This is also how Omron is organized today.
Market Pricing relationships describe relationships in which people relate to each other concerning some ratio, rate, or proportion—think money or any other form of token.
These are typically relationships between buyers and sellers in a market, often defined by some kind of monetary transaction based on prices.
This is how employees of Omron probably related to each other in the times of Tateishi, when the organization was still highly decentralized featuring autonomous units (with their own P&L responsibility) that needed to collaborate with each other through market-like dynamics.
Equality Matching relationships describe people keeping track of what they give and receive while trying to keep the contributions balanced—i.e., tit-for-tat relationships.
These relationships are typically between friends or peers and are primarily concerned with ensuring the relationship remains in a balanced state. These relationships are also often characterized by different types of one-for-one correspondence, such as turn-taking or reciprocity.
Communal Sharing relationships describe people seeing each other as the same and perceiving each other as equivalents in a particular community.
These are typically intimate relationships between members of the same family, community, or other strongly bounded groups of people. Communal Sharing relationships are often characterized by some kind of shared markings (i.e., identical clothing and logos), rituals, and physical intimacy.
Another essential aspect of the relational models is the four fundamental motives underlying most moral judgments, emotions, and behavior in each model:
These four moral motives provide “control” in each separate model with so-called “relational regulation,” which can be regarded as morally motivated actions of peers to fit in well with a certain dominant relational mode.
Let's explain the four “moral motives” briefly:
1. Authority Ranking - Hierarchy
The relational motive generating and controlling authority ranking relationships is the need for hierarchy. People may also talk about a motivation of power.
2. Market Pricing - Proportionality
The relational motive generating and controlling market pricing relationships is the need for proportionality. People may also talk about a motivation of achievement.
3. Equality Matching — Equality
The relational motive generating and controlling equality matching relationships is the need for equality. People may also talk about “being equal,” “equally matched,” or the need for fairness.
4. Communal Sharing — Unity
The relational motive generating communal sharing relationships is the need for unity. People may also talk about a motivation of a sense of belonging or the need to belong.
Predominant relational modes
Note: It is important to realize that there is no particular order in these models. In other words, one model is not better than another. They are also not “pure,” meaning we use these models in complex and dynamic combinations in daily life.
Also, it is not just the relational models between two members that count in organizations, because the relationship between two members also has an impact—along with moral implications—on the respective member's relationship and with others in the organization.
This means that one individual—like an employee—is just one small node in a larger network of relationships, especially in large organizations. The combination of multiple relational models is what Fiske calls the "metarelational model."
However, in many social environments—such as organizations—there is often one metarelational model that is predominantly guiding most of the behavior of the members of that particular organization—the predominant metarelational model that is shared by the majority of the members of the organization. This model then informs the members how to jointly construct meaningful coordinated action.
For the sake of simplicity, we will not talk about the rather technical term “metarelational models” in the remainder of this article. We will instead discuss predominant relational modes:
- We will refer to “organizations with a metarelational model that consists primarily of Authority Ranking relationships” as “organizations with a predominant Ranking mode.”
- “Organizations with a metarelational model of predominant Communal Sharing relationships” are “organizations with a predominant Sharing mode.”
- “Organizations with a metarelational model of predominant Equality Matching relationships” are “organizations with a predominant Matching mode.”
- “Organizations with a metarelational model of predominant Market Pricing relationships” are “organizations with a predominant Pricing mode.”
A relational-model view
Viewing the Autonomous Society from this “relational-model perspective,” we could argue that Autonomous Organizations have to move away from a predominant Ranking mode of organizing that relies strongly on hierarchy for its internal control.
However, this doesn't mean there is no internal control in Autonomous Organizations—it will simply take a different shape. It will morph from hierarchical top-down-enforced controls to more peer-enforced controls guided by a shared moral sense of proportionality, equality, or unity.
Why? Because getting work done in any organization is morally motivated, meaning the work in organizations gets done because its members exercise moral rights and obligations via “relational regulation” dynamics.
These relational regulation dynamics will inform and motivate members to act according to their predominant relational mode—whether that is Ranking, Pricing, Matching, or Sharing mode—because they feel it is just “the right thing to do” in their particular working environment.
Why? Because one of the core ideas of RMT is that people experience these four types of relationships as intrinsically desirable, fulfilling, meaningful, and even necessary.
This means that all people—also employees—seek to start and engage in relationships that realize these four models and evaluate all their social actions—and their work with others—in reference to them. These are merely the ways to relate to each other.
As such, these four “ideal” relational modes guide employees in what is morally good, what is right and wrong, and even what is obligatory in their workplace.
So, instead of relying on Ranking mode, Autonomous Organizations should implement one of the three alternative modes—i.e., Pricing mode, Matching mode, or Sharing mode—with their moral guidance to organize and coordinate the work of their autonomous members successfully within their organization.
Cultural guidelines and rules
Still, it is obviously not that simple. Just a description of one of these relational modes, with their respective moral guidance, will not provide employees with the necessary context to successfully organize all aspects of their work in an Autonomous Organization.
Because to be able to organize all the work—based on a certain predominant relational mode—employees must still have other cultural guidelines and rules that specify which mode to use, how to use it, with whom to use it, when to use it, and so forth.
Consider the example of the traditional workplace. We all know that organizing work via a hierarchy—based on a predominant Ranking mode—still needs clarity about each employee's relative position in that hierarchy, the rules about when and how those hierarchical ranks are defined, what each rank can decide, and much more.
Similarly, when organizing work in a predominant Sharing mode, employees need to know who is part of the community and who is not, what resources can be shared freely, and so forth. Similar cultural guidelines and rules are needed when organizing work in predominant Matching or Pricing mode.
As such, relational modes are always guided by a set of culturally defined guidelines and rules, as defined by a particular cultural context in which the organization operates.
The rise of Autonomous Organizations
So, by now, you might ask yourself,"What do these Autonomous Organizations look like in reality?" And that is certainly a fair question.
Luckily, over the last five years, during our many Bucket List travels around the globe, we have visited plenty of organizations that we could qualify as Autonomous Organizations.
So yes, we definitely know they are real. And we would argue that they are on the rise because we increasingly encounter radical decentralized organizations that rely on their autonomous members' actions to be successful—without the need for any formal middle management layer.
But before we dive into the practical examples, it is important to stress our main argument again: We argue that in these Autonomous Organizations, hierarchy is no longer the predominant moral motive of its members to get work done.
Instead, as described above, members in Autonomous Organizations are instead guided by predominant moral motives of proportionality, equality, or unity.
We argue that these “alternative” moral motives decide what is acceptable behavior of members in an Autonomous Organization, and what behavior is seen as “not done” or “immoral.”
That doesn't mean, however, that there is no hierarchy in Autonomous Organizations. In fact, these organizations often still have a small top-management team in place that has formal authority over all the other autonomous members in their organization. It is just that the hierarchical moral motive is not predominant in Autonomous Organizations, whereas it is in traditional hierarchical organizations.
Ok, that is enough theory for now. Let's make this more tangible by giving real-life examples of Autonomous Organizations for each of the three non-hierarchical modes.
1. Pricing Mode
In Autonomous Organizations with a predominant Pricing mode, we see employees primarily relating to each other based on internal market mechanisms and dynamics that are all aimed to create a feeling of free-market spirit.
Here, we see employees acting as entrepreneurs—as buyers and sellers—to get their work done via an internal marketplace where employees bid on goals, transact work with each other, and engage in formal binding contracts—all based on a particular (internal) currency.
In this mode, we do not see formal middle managers assigned by the formal hierarchy. Instead, we see leaders that are “chosen” by the “invisible hand” of the (internal) market based on the fact that their bids on the tenders for open leadership positions were considered to be the best by the other members of the organization (so, not necessarily the highest or most competitive bids).
Moreover, in this mode, employees decide many things by consent-based decisions based on demand and supply, and rewards are distributed proportionately to the performance in the market via generous profit-sharing mechanisms.
Example of Pricing mode at scale: Haier
The best example of a large Autonomous Organization with a predominant Pricing mode is the Chinese white-goods manufacturer, Haier. Take a look at the animation below to understand why:
As we wrote recently, Haier started to introduce technologies based on blockchain (i.e., “smart contracts”) to improve the level of trust within the organization by creating increased transparency and security, and to decrease the transaction costs needed to run such a large internal market.
Example of Pricing mode in Japan
We do not only find market-based Autonomous Organizations in China. In fact, Japan itself has a strong history of companies being designed and guided by principles based on this idea of radical decentralization combined with market dynamics.
As we wrote above, back in the days of Tateishi, Omron itself was highly decentralized (they have since been centralized over the last several decades), with plenty of autonomous units featuring their own P&L responsibility. Similar ideas could be found at companies like Kyocera, Japan Airlines, and plenty of others highly influenced by Kazuo Inamori's Amoeba Management.
Although most of these firms have lost their decentralized nature over the last few decades (like Omron did), there are still pioneers to be found in Japan that run with highly autonomous members who organize themselves based on internal market principles.
Disco Corporation might be the best example of such a company. During our tour around Japan, we visited Corporate Directions, a consulting firm based in Tokyo that runs with a predominant Pricing mode. Interestingly, they reminded us strongly of an organization on the other side of the world, the Swedish consultants of Centigo.
2. Matching Mode
In Autonomous Organizations with a predominant Matching mode, we see employees that relate to each other primarily based on feelings of fairness, which creates an environment where all employees feel that they have an equal say and equal opportunities.
Here, we see employees acting as equal peers that get work done via mechanisms based on commitment systems, where peers make voluntary commitments that are dynamic but must eventually be “balanced.”
In this mode, we do not see formal middle managers, nor do they tend to speak about “leaders.” Instead, we see peers that chose their own “representatives.” We often encounter two representatives per team: one that takes care of internal team dynamics and one that represents the team to the outside stakeholders.
Moreover, in this mode, employees tend to make decisions via no-objection-based consent guided by a sense of tolerance or democratic voting when it concerns topics that impact the whole organization (or a large majority). Rewards are often equally divided among all members of the organization or based on loyalty.
Example of Matching mode at scale: NER Group
We can find many examples of Autonomous Organizations running on Matching mode all around the globe, but mainly around Western Europe. Also, organizations running on Sociocratic and Holacratic principles tend to fall into this category.
The Spanish NER Group is the best example of a large Autonomous Organization with a predominant Matching mode. You will see why in the animation below:
Example of Matching mode in Japan
Although the majority of the organizations from our Bucket List run with some predominant form of Matching mode, we did not find many Japanese companies that rely on this mode to organize their Autonomous Organization.
However, we did find a few of them, with the best example being Net Protections. Interestingly enough, we observed that in Japan, these organizations tend not to talk about the concept of “representatives” but rather seem to rely on a concept called “catalysts.”
3. Sharing Mode
In Autonomous Organizations with predominant Sharing mode, we see employees that relate to each other primarily based on community dynamics that are aimed to create a strong sense of unity and a feeling of belonging.
Here, we see employees acting as members of a strong community who want to care for and support each other. What a concept, right? Members contribute to the community without expecting anything in return, getting work done because everyone just pitches in and performs tasks until the job is done.
In this mode, we do not see formal middle managers, leaders, or representatives. In fact, in this mode, teams do not have any assigned formal leaders at all. Instead, all team members are collectively responsible for their team's performance.
Moreover, in this mode, employees tend to make decisions via a proposal-based consensus process guided by a sense of solidarity and empathic compassion that aims to “keep the group together.”
Rewards are mostly intrinsic and tend to revolve around such things as a sense of affection, a sense of responsibility for other people, and a sense of belonging to the community.
Example of Sharing mode at scale: Buurtzorg
We did not come across many Autonomous Organizations that rely on a predominant Sharing mode. But when we do see them, they tend to be active in care-oriented industries like health care, social care, and eduction.
The best example of a large Autonomous Organization with a predominant Sharing mode is the Dutch healthcare organization Buurtzorg. You will see why in the animation below:
Example of Sharing mode in Japan
Although companies run by Sharing mode seem to be rare in general, we still found at least one example in Japan: teamLab, an artistic collective of about 200 members.
The end of the middle manager?
Whew, okay, it's time to wrap this giant piece up.
What we are trying to argue here —by using the academic concepts of RMT and relational regulation—is that Autonomous Organizations can organize successfully without the need for formal middle management layers.
They can do so because their autonomous members are morally motivated to organize the work by starting, deepening, and repairing relationships with others that matter to them—be it pricing, matching, or sharing-based—and, when needed, ending relationships that become useless.
Does this mean we’re nearing the end of the middle manager? Does this mean that Autonomous Organizations will eventually take over traditional hierarchical organizations?
We do not know. You do not know. Tateishi did not know. Nakama does not know. Laloux does not know. Hamel & Zanini do not know. Foss & Klein do not know. Nobody knows.
Only time will tell. And we’ll be watching.