Every person alive carries what might be called a social budget — a finite capacity for maintaining genuine relationships. No matter how extroverted, ambitious, or technologically connected a person becomes, the number of individuals they can meaningfully track and truly know appears to be constrained by the architecture of the human brain. The theory that pins this limit at around 150 people is known as the Dunbar Number, and it has become one of the most cited and debated ideas in social science.
Who Is Robin Dunbar and Where Did the Number Come From?
Robin Dunbar is a British anthropologist and evolutionary psychologist, currently based at the University of Oxford. His research in the early 1990s began not with human beings but with other primates. He was investigating why primates, relative to other mammals, have such large brains — and specifically such large neocortices, the region of the brain responsible for complex cognition, social reasoning, and self-awareness.
The Neocortex Ratio
Dunbar found a striking correlation: the ratio of the neocortex volume to the rest of the brain was strongly predictive of the typical social group size for any given primate species. Gibbons, with a relatively smaller neocortex ratio, live in small family groups of around 6 to 8. Chimpanzees, with a larger ratio, form troops of roughly 30 to 80. Gorillas fall somewhere in between.
When Dunbar extrapolated the regression line to the human neocortex ratio, the predicted social group size came out at approximately 147.8 — rounded to 150 in most popular accounts.
The logic behind the correlation is intuitive once stated. Social living requires keeping track of an enormous amount of information: who is allied with whom, who has wronged whom, what obligations exist between individuals, who can be trusted and in what circumstances, and what each person's current emotional and social status might be. This social bookkeeping requires cognitive resources. The neocortex, Dunbar proposed, evolved in large part to meet this demand. Bigger groups require more powerful social processing hardware.
Validation Against Historical and Anthropological Data
Dunbar did not rely on the primate regression alone. He tested the 150 prediction against a wide range of independent data sources:
- Hunter-gatherer band sizes: Across dozens of documented hunter-gatherer societies worldwide, residential communities average around 150 members before they fragment or reorganize.
- Neolithic villages: Archaeological evidence from early agricultural settlements shows village sizes clustering around 150 before they split into daughter settlements.
- Military unit sizes: Most armies throughout recorded history have used a tactical unit of approximately 100 to 200 soldiers — the Roman maniple, the Hutterite colony limit, the British Army company. Military commanders appear to have independently discovered that direct, trust-based coordination breaks down above this threshold.
- Christmas card lists: In a study of British households, Dunbar found that the average person sent cards to around 153 people — a social network proxy that reflects the same underlying cognitive limit.
"The figure of 150 seems to represent the maximum number of individuals with whom we can have a genuinely social relationship, the kind of relationship that goes with knowing who they are and how they relate to us." — Robin Dunbar, Grooming, Gossip, and the Evolution of Language (1996)
The Nested Layers of Human Social Circles
One of the most practically useful extensions of Dunbar's work is the identification of nested social layers inside and beyond the 150 figure. Dunbar and colleagues, including colleagues Sam Roberts and Will Dunbar, found that human social networks are not flat — they are organized in a series of concentric circles, each roughly three times larger than the one inside it.
| Layer Name | Approximate Size | Relationship Quality |
|---|---|---|
| Support clique | 5 | Closest confidants; whose loss would be devastating |
| Sympathy group | 15 | Those you turn to for emotional and material support |
| Affinity group | 50 | Regular social companions; strong familiarity |
| Dunbar group | 150 | Full social network; stable trust and cooperation |
| Acquaintance layer | 500 | People you know by name and face |
| Recognition layer | 1,500 | People whose face you can recognize |
Each transition represents a qualitative change in relationship type. The innermost group of 5 consists of people with whom you communicate multiple times per week and would turn to in a serious crisis. The 15-person sympathy group comprises those whose death would profoundly affect you. The 50-person group covers those you would invite to a party. The 150-person Dunbar group is the limit for a community in which everyone can know everyone else well enough to cooperate without needing formal rules or enforcement mechanisms.
Why Three Times?
The consistent ratio of approximately three between each layer is not fully explained, but researchers have proposed that it reflects the different frequencies of social contact required to maintain different depths of relationship. Maintaining a close confidant relationship requires very frequent interaction. Maintaining a stable acquaintance requires far less. The threefold scaling may reflect a natural discretization of contact frequency bands.
What Dunbar's Number Means for Organizations
The 150-person limit has significant implications for how organizations function. Research by Dunbar and others, as well as independent organizational data, suggests that companies and groups below 150 tend to operate with informal coordination — social trust, shared norms, and direct personal knowledge of colleagues substitute for formal hierarchy. Above 150, these informal mechanisms break down and formal structures (org charts, written policies, performance management systems) become necessary.
Case Studies in Organizational Design
W.L. Gore and Associates, the manufacturer of Gore-Tex fabric, famously applies this principle. Gore limits factory units to roughly 150 workers. Rather than building large centralized manufacturing plants, the company constructs smaller facilities and adds new buildings rather than expanding existing ones. The company's founder, Bill Gore, cited the observation that below 150 people, everyone knows everyone else and social pressure is a sufficient coordination mechanism — formal management layers are largely redundant.
Hutterite colonies provide a centuries-old natural experiment. This Anabaptist religious group has practiced communal living since the 16th century. Community elders long ago established a rule that when a colony reaches approximately 150 members, it must split and found a new colony. The elders did not derive this from any modern science — they discovered through experience that social cohesion and economic cooperation deteriorated above that threshold.
Military units across cultures and centuries converge on similar numbers. The US Army company consists of 80 to 225 soldiers. The Roman legion's tactical subunit, the maniple, had 120 to 160 men. The consistency is remarkable given that these organizations arose independently and were concerned with battlefield effectiveness rather than social psychology.
Does the Dunbar Number Apply to Online Social Networks?
The rise of social media presented an apparent challenge to Dunbar's thesis. If the cognitive limit is 150, how do people manage networks of 5,000 Facebook friends or 500,000 Twitter followers?
Research suggests the answer is: they do not actually manage them. Studies of Twitter data by Dunbar's group found that despite wildly varying follower counts, active, reciprocal engagement — the kind that constitutes a genuine social relationship — consistently clustered in the 100 to 200 range, regardless of total network size. Large follower counts represent an audience, not a social circle. A celebrity with a million followers is broadcasting, not maintaining a million relationships.
A 2016 study published in Royal Society Open Science analyzed 3.3 million Twitter users and found that the emotionally close network — the people with whom users had consistent, mutual, substantive exchanges — had a median size of approximately 100 to 200, fully consistent with the Dunbar prediction.
This finding has significant implications for how we think about social media platforms. The technology has not expanded human social capacity. It has primarily made it easier to broadcast to, and passively observe, larger audiences of acquaintances while maintaining the same basic social circle size.
Criticisms and Challenges to the Dunbar Number
The Dunbar Number is widely cited, but it is not scientifically unchallenged.
The 2021 Statistical Challenge
A 2021 study by Patrik Lindenfors, Andreas Wartel, and Johan Lind, published in Biology Letters, re-examined the statistical basis of Dunbar's primate regression. They found that when they included a broader range of primate species and used more rigorous statistical modeling, the confidence interval around the human social group size estimate was so wide — ranging from roughly 2 to over 500 — that 150 was not statistically distinguishable from much larger or smaller numbers. The study argued that Dunbar's 150 figure was imprecisely estimated and perhaps overconfidently stated.
Cultural and Contextual Variation
Critics also point out that human group sizes vary enormously across cultures, economic systems, and technological contexts. Industrial cities, online communities, and modern corporate structures all involve navigating far more than 150 relationships, even if most are shallow. Whether the Dunbar Number describes a hard cognitive ceiling or simply a comfortable optimum is debated.
Individual Differences
People vary in their social capacity. Some individuals — particularly those high in extraversion or with professional roles requiring broad network maintenance (salespeople, politicians, community organizers) — appear to actively maintain significantly more relationships than 150. Whether this reflects genuine cognitive differences or simply different calibrations of relationship depth is unclear.
Why the Dunbar Number Keeps Mattering
Despite methodological debates, the Dunbar Number remains a powerful conceptual tool precisely because it captures something intuitively recognizable: there is a point beyond which a social group changes character. The tight-knit small team where everyone knows each other's quirks, trusts each other's judgment, and coordinates through informal conversation is qualitatively different from the large organization where strangers pass in corridors and policy documents govern behavior.
Understanding this transition point helps explain why:
- Startups feel different from corporations even when the product hasn't changed
- Small towns have different social dynamics than cities
- Squad-level teams in military and sports contexts outperform similarly skilled but less cohesive larger groups
- Online communities fragment and formalize as they scale
The human brain has a social budget. Spending it wisely — knowing which relationships to invest in, which layer of the circle to place people in — is one of the most consequential decisions people make, even when they make it unconsciously.
Practical Implications of the Dunbar Number
Understanding the layered structure of social circles has several practical applications:
For leaders: The most critical relationships are in the innermost two circles (the 5 and 15 layers). Leaders who lose connection with their core circle — by becoming isolated by seniority or surrounded only by yes-sayers — often lose touch with organizational reality.
For organizations: When a team or unit grows beyond 150, investing in formal structures before that transition (documented processes, onboarding systems, explicit norms) prevents the social disruption that comes from informal coordination mechanisms suddenly failing.
For individuals: Being deliberate about which of the 50 slots in your affinity group are occupied has an enormous effect on wellbeing. Research on social support consistently shows that the quality and size of this middle-circle group predicts health, longevity, and resilience more strongly than almost any other social variable.
For community builders: Online platforms and communities that want to maintain genuine engagement rather than passive broadcast audiences should consider design choices that facilitate the formation of smaller subgroups — the 15 and 50 circles — rather than optimizing only for total membership numbers.
Conclusion
The Dunbar Number is one of those rare scientific ideas that is both specific enough to be testable and broad enough to illuminate phenomena across biology, anthropology, history, and organizational design. The exact figure of 150 may be debated, and the statistical confidence intervals around it may be wide. But the underlying insight — that human social cognition has real limits rooted in primate brain evolution, and that these limits shape the dynamics of every group from families to armies to online communities — is supported by a remarkable convergence of evidence.
Understanding those limits is not a reason for pessimism. It is an invitation to think carefully about which relationships we invest in, how we design institutions that must function beyond the range of informal trust, and what we are really asking for when we seek more connections in a world that makes accumulating them easier than ever before.
Frequently Asked Questions
What is the Dunbar Number?
The Dunbar Number refers to a theoretical cognitive limit of approximately 150 people with whom any individual can maintain stable, meaningful social relationships. Proposed by British anthropologist Robin Dunbar in the early 1990s, it is derived from the relationship between primate neocortex size and typical social group size. Beyond roughly 150 connections, people lack the cognitive bandwidth to track the mutual obligations, histories, and trust levels that define genuine social bonds.
How did Robin Dunbar calculate the number 150?
Dunbar analyzed the relationship between neocortex ratio (the size of the neocortex relative to total brain volume) and observed social group sizes across 38 primate species. He found a strong correlation, then extrapolated the regression line to the human neocortex ratio to predict a natural human group size of roughly 100 to 230, with 150 as the midpoint estimate. He then validated this against anthropological data from hunter-gatherer societies, historical military units, and village records.
What are the nested layers within the Dunbar Number?
Dunbar and colleagues identified a series of nested social circles with characteristic sizes: an innermost support clique of about 5 people (close confidants), a sympathy group of about 15 (those whose death would profoundly affect you), an affinity group of about 50 (regular social companions), the Dunbar group of 150 (stable social network), and outer layers of approximately 500 (acquaintances) and 1,500 (people you can recognize). Each layer is roughly three times larger than the one inside it.
Does the Dunbar Number apply to online social networks?
Research suggests the Dunbar Number constrains meaningful online relationships even when platforms allow thousands of connections. Studies of Twitter and Facebook data show that active, reciprocal engagement clusters around the same 150 to 200 person range, regardless of total follower counts. Having 5,000 Facebook friends does not appear to expand the number of people any individual genuinely keeps track of — it simply adds passive contacts beyond the cognitive ceiling.
Is the Dunbar Number scientifically settled?
The Dunbar Number is an influential hypothesis, not a fixed biological law. Some researchers dispute the neocortex-to-group-size methodology and note that human group sizes vary enormously across contexts. A 2021 study by Patrik Lindenfors and colleagues challenged the statistical basis of Dunbar's primate regression. Most researchers agree that human social cognition has real limits; they debate whether 150 is the correct figure and whether a single number can capture the range of human social environments.