Essay: Hierachy and Heterarchy

ESB Dossier Ecosystemen voor ondernemen

governance

Essay:
Hierarchy and Heterarchy
Karen
Stephenson
Founder of NetForm
International

W

e are on the cusp of something very
big in the 21st century. Where singular hierarchical governance is illsuited to serve, it can be argued that
the digital landscape will rival nothing less than the
Serengeti plains offered up to our hominid ancestors.
We are compelled to evolve beyond hierarchical governance as we daily witness its failures. What is called for
is a rather disruptive form of innovative governance
– an organizational structure called heterarchy. It is a
network by design of hierarchies. Implementing this
form of governance will require a dramatic rethinking
of leadership and the origins of organizational life.

states. This territorial imperative is largely passive and
rooted in place. Ironically, little thought is given to
space. It’s not surprising, as most of us don’t see the
shape of space, only the objects that occupy it. Walk
into any call center, computer center and workstation.
What do you see? The tools have changed, to be sure,
but the actual layout of an office or a floor plan is not
all that different from Bentham’s late 18th-century
panopticon – which refers not only to the built environment but to its governing philosophy as well. We
delude ourselves into thinking that we work in a postmodern, sexy-cool virtual world, but despite enabling
technologies we have advanced only incrementally in
our ability to substantially change our institutions.

Human imperatives

Three imperatives for human connections form a mutually related set of elements for 21st-century governance. The first two imperatives of place and social
proximity have been the mainstays of human culture
for millennia. People convene in town halls and assemble in city squares because they cannot abandon that
primordial need for intimacy and belonging. A third
digital imperative emerges when we substitute space
for place, generating a hybrid human connection. The
three imperatives are detailed below.
The territorial imperative

Society and economy give primacy to physical structures and artifacts – monuments, buildings and city
54

The social imperative

A more reactive social imperative that has evolved for
millennia is physical propinquity – or proximity. People need people, and they need them to be close. Propinquity is a social imperative that catalyzes our need
to belong, and as a result we assemble in tribes, teams
and organizations. How tightly we are coupled in
these organizational structures depends on our social
structure, which, like the shape of space, we typically
cannot see. There is one form of governing structure
that is visible – that of hierarchy – but we know from
research that hierarchies are poor proxies for revealing
social DNA. Leaders see headcounts; they cannot see
the social landscape but for the Dilbertian cubicles –

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

Jaargang 99 (4698S) 20 november 2014

Governance ESB

commonly referred to as hierarchical planning by architects and designers – in their line of sight. If they
can’t see how people are relationally connected, then
they can’t accurately value, in terms of social capital,
either the people or their relationships. This social
DNA has been the missing link in fully valuing, and
therefore governing, the human asset.
The digital imperative

More recently, a new digital imperative has emerged
– crowd sourcing where humans virtually amass and
connect in space regardless of where they hang their
hat. We connect in a vast Milky Way of virtual connections that is visually represented as a network and mathematically measured. But be warned, virtual chatter
is largely transactional noise. Transactions are like traffic where the rules of engagement are mutually agreed
and well understood. Real value is determining where
the ‘trust’ hides out in all that noise. Trust is a species
of connective tissue that absorbs greater shocks of uncertainty and ambiguity, and that is why trust is such
a sturdy form of connection. This has implications for
governance.
However, trust as a theoretical construct of connection was a bridge too far for most economists.
Connection was envisioned as three different types
of transactions and distinguished by Williamson as:
coordination costs – once the contract has been signed
and in keeping with authority-based exchanges, information costs (looking for the right people or information, which could be considered a form of networking)
and bargaining costs (drawing up a perfect contract, or
the road to trust). Sadly, no one has ever been able to
measure transaction costs because authority and trust
are considered modulators of an exchange, not a species of connection. If we shift the paradigm to construct
a ‘connection trifecta’ consisting of trust, authority and
transactions, then there are existing ways of measuring
trust and authority.
But we haven’t seized this paradigm and have
languished in a state of malaise with our old forms of
governance. This malaise was disrupted by a collision
of the digital with the physical in the Arab Spring.
Jump-started in the Twitter sphere, people gathered in
Tahrir Square, as they tweeted up to meet up. There is
a dynamism in the digital domain that contrasts sharply with a government’s presumption of perpetuity.
Governments have toppled, after all. To truly live long
and prosper, different organizational blueprints are reJaargang 99 (4698S) 20 november 2014

quired. Blueprints of the industrial complex – hierarchical governance – are still needed, but there is also a
new class of blueprint that addresses the construction
of heterarchical governance.
Elementary structures

Most of us lament the existence of big government,
these great barrier reefs of red tape that famously
tout the motto ‘bigger is better’. Imagine if there was
an iterative, underlying pattern governing the way
governments morph into enormous proportions. For
that matter, how does any organization scale? What if
there was a ‘structure’ at the root cause of all this ‘scale’.
That’s exactly what the French anthropologist Levi-

leaders see headcounts;
they cannot see the social landscape

Strauss reasoned like when investigating tribes of family structures in South America. He called these nucleic
structures the “atom of kinship†(Levi-Strauss, 1955;
1969). In the end, his idea proved more provocative
than practical. He wasn’t wrong; he was just looking in
the wrong place. There is a broader classification scheme; its genus is organization, and kinship – with its
biological and fictive lineages – is one of its many hybrid species. What other species populate this genus?
Three elementary species come to mind: hierarchy, markets and networks. Economists believed
hierarchy to be an island of planned coordination in a
sea of market relationships, a pristine paradise inhabited by vertically integrated tribes of employees – adapted from Richardson (1972). At one end of a continuum, markets were considered to be the grand genesis
of commerce. At the other end, there was hierarchy,
the logical endpoint of civilization. A network was a
theoretical interloper and often dismissed by economists as a mixed breed, a doomed hybrid. Those were

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

55

ESB Dossier Ecosystemen voor ondernemen

early days. Decades of theoretical research coupled
with the practical realities of social media have reasserted the importance of the network and its relevance
to cultural genesis and social DNA. Networks are after all, primordial; only later did clans, lineages and
hierarchical nation states develop sufficiently to trade

nuanced and capable of asynchronous
and asymmetric exchanges, networks
deftly elude the visible hand of hierarchy

their surpluses in local, regional and global markets.
Ronald Coase’s classic paper (1937) on the nature
of the firm – e.g. hierarchy – suggests that firms and
markets, while being different organizational structures, nevertheless share common transactional practices.
The distinction, later amplified by Williamson (1976;
1985; 1993), was based on the amount of knowledge
contained in a transaction – asset specificity. Disinterested, non-repetitive one-off exchanges occur as market transactions – simple contracts. The structure of
this exchange is a dyad – two people linked together in
an exchange. Dyads ground most of our assumptions

Elementary structures and exchange rates
Organizational
form
ABCs of exchange

Table 1

Structure

Network

Three or more nodes
arrayed in a triad (a triad is
the smallest structure)

Hierarchy

Routinization through a
governing authority

Chains of nodes

Market

Disinterested, nonrepetitive transactions

Collection of dyads

Heterarchy

56

Mutually interested,
repetitive exchanges

Mutually interested,
collec­tive governance
characterized by
Three or more hierarchies
asymmetric & asynchronous networked together as a
exchanges
triad

about market economies, personal relationships, contract law and marriage.
Exchanges that entail greater uncertainty – and
therefore a proportional amount of asset specificity
– are best sheltered within the firm as a way to mitigate greater risk. Theories of ‘the firm’ were conceived
and characterizations of its hierarchical infrastructure
were developed. Management theorists put their imprimatur on the debate with theories based on hybrid
organizational forms and derivative managerial approaches, such as the U-form and M-form organizations
and Theories X and Y (McGregor, 1957), and Z (Ouchi, 1981) respectively. A literal chain of command is
the overarching structure in hierarchy.
Only much later did other researchers argue
(Stephenson and Zelen, 1989; Stephenson, 1990;
Krackhardt, 1990; Powell, 1990) for the existence of
another organizational structure – a network. Networks aren’t dyads – as in markets – and they aren’t
exclusively chains – as in hierarchies. An elementary
network is a triad, built from adding one more person
to a dyad. But a profound mathematical principle is revealed when this operation occurs. Adding one more
person to a triad doubles the number of linkages and
introduces the first indirect link in the structure. Perhaps this is what Levi-Strauss realized when he asserted that he was less interested in a ten percent increase
in a population of 300 million, than in a two-person
household becoming a three-person household. Triads
absorb greater uncertainty, exceeding hierarchical limitations with aplomb. Nuanced and capable of asynchronous and asymmetric exchanges, networks deftly
elude the visible hand of hierarchy (Chandler, 1977)
and the invisible hand of the market (Smith, 1776).
The inclusion of a fourth organizational species – heterarchies – is recently noted (McCulloch,
1945; Stark, 1999; Stephenson, 2004; 2008; 2009).
Heterarchies are a precise combination of networks
with hierarchies. In the same way that a triad introduces the first indirect link in a three-person network, a
heterarchy introduces its first indirect link in a threehierarchy network. The definition of a heterarchy is as
follows: A heterarchy is comprised of three or more
different organizations – hierarchies –, each with its
own raison d’être and no single entity privileged over
the other. Networked together, these hierarchies share in the collective governance of the whole in order
to achieve a greater good that no single entity could
achieve on its own.

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

Jaargang 99 (4698S) 20 november 2014

Governance ESB

We know about heterarchies because of their
spectacular failures, but heterarchies aren’t dysfunctional by nature. They become dysfunctional when a leader of a single hierarchical entity naively privileges his
interests over the whole. Said differently, leaders mistakenly assume that their special interests are the only
things that matter. This is largely due to how they learned their tradecraft by practicing as CEOs or directors
of singular hierarchies or, alternatively, they may have
been educated in business schools which are grounded
in 19th-century norms of leadership. Either way, these
leaders are generally unprepared to manage and lead in
heterarchies. A re-sequencing of social DNA is summarized in Table 1.
Segmentary system constraints

Heterarchies require much more than a coalition of
the willing; they demand a well-designed and coordinated network to ensure the alignment of tasks across
multiple and competing organizations. When tasks are

not aligned, perverse outcomes surely follow (Box 1).
Heterarchical failures reach back into our primordial beginnings. Humanity began as a hunter-gatherer
band. It was a nuclear family structure, simple enough,
that grew into a tribe or clan. Primordial clans separated into segmentary lineages, then chiefdoms, right up
to the modern-day state – or so the story is told. Regardless of whether this evolutionary tale is a true accounting, it is evident that our governance systems betray
some aspect of this heritage (Sahlins, 1961; 1963). In
the early stages of human organization, leaders were
typically charismatic. A change occurred between the
segmentary lineage and chiefdom, where leadership
shifted from that of a charismatic leader – who built
a following by creating loyalties through generosity,
fearful acquiescence through magic, demonstrated
wisdom, oratorical skill, etcetera – to that of an instituted office authorized by God, coup or chad.
When policies change or new needs arise, teams
or departments are created not from the ground up,

Examples from the United States
Three segmentary examples from the
United States can be described that
could and should have operated as heterarchies, but failed:
US military health care
In 2014, the US military health care system experienced an alarmingly high
number of ‘never events’ – fatalities
which are potentially preventable. The
health care system is organized as a heterarchy comprised of four major players:
Army, Navy, Air Force and the Department of Defense (DoD). When certain
‘never events’ were revealed in a New
York Times’ exposé, each member hierarchy blamed the others when in fact, the
refusal to share patient data across the
heterarchy is what led to the fatalities.
As a former Army policy officer said to
the NYT: “Why should the Army safety
system want to play with DoD? Because
then I have less control over my data,

Jaargang 99 (4698S) 20 november 2014

less control over my kingdom, and potentially DoD is going to tell me what to do.â€
2010 Gulf Oil disaster
The 2010 Gulf Oil disaster was a manmade collision of special interests, which
resulted in a natural disaster of epic proportions. Multiple organizations such as
Halliburton, BP and varied insurers were
locked in contractual relationships when
the disaster struck. In a diaspora of abdicating responsibility, organizations passed the blame. Rarely, if ever, did leaders
see their kingdom as part of a whole network of interacting organizations. When
leaders and boards attend to only their
special concerns, perverse outcomes
emerge. “No one wants this over more
than I do. I would like my life back,†said
CEO Tony Hayward on May 31, 2010, regarding the oil spill disaster that claimed eleven lives and spewed 100 million gallons
of toxic oil into the Gulf of Mexico. “We

box 1

care about the small people,†BP Chairman Carl-Henric Svanberg emoted to reporters in Washington on June 16, 2010.
Department of Homeland Security
A few years after 9/11/2001, an attempt
was made to better protect the nation.
The United States Administration created the Department of Homeland Security (DHS) to oversee national security by
combining three separate government
functions: intelligence (NSA, NRO, CIA
as a few examples), policing (the FBI for
instance) and disaster response (FEMA).
Built from 60+ pre-existing departments,
the new super-ordinate layer of authority
only deepened the tug-of-war between
the agencies for limited resources. The
DHS was widely considered a failure. It
was mismanaged from the outset, because simply putting a higher level of authority over competing organizations is
no guarantee that they will collaborate.

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

57

ESB Dossier Ecosystemen voor ondernemen

but as sub-units of existing segments mimicking cellular division. As the layers of smaller chiefdoms proliferate, they compete against one another, calling a truce
only when a larger chiefdom threatens their existence.
So, within a government department for example, one
team jockeys for position with another, one directorate
attacks another to protect its budget, and the department as a whole fights other departments to defend its
turf. In these systems there is no internal structure or
infrastructure to join the system as a whole; it is simply a collection of hierarchies or vertically integrated
silos. As such, they are never more (and often much
less) than the sum of their parts. Segmentary systems
calculate power by comparing and contrasting their

Most large conglomerates are
heterarchies in name only, as they behave
as coalitions of the willing, until collaboration
clashes with the chain of command

stock or status with that of other segments. If required,
they will cannibalize other parts of the organization to
preserve their part (Douglas, 1986). With additional
stress, they will eat their young. This ruthless survival
tableau describes segmentary systems, not heterarchies. Member hierarchies of a single heterarchy will
suppress the killer instinct in lieu of collaboration with
others because they understand that, if the higher objective is achieved, they will all be successful, and not
at the cost of a peer.
When looking at the three examples described
in the box, one can appreciate how segmentary politics easily and perversely thwarts overall objectives
and goals. It should come as no surprise when people
complain of being stuck in organizational silos because
they are, and these silos are of their own making. What
we don’t realize is that we are hardwired to create these
tribal silos because of segmentary system constraints
held in place by cultural, procedural and measurement
barriers. In segmentary systems there is no pre-planned
58

network in place to allow for the seamless exchange of
vital information laterally among the different hierarchical organizations and no measurement system to
ensure its sustainability. The result is that information
disappears, deadlines are missed, fingers point and
wrongs are papered over. If leaders could step back and
see the whole instead of only their portion, then no
one would have to die, pay amends, or bear the whole
brunt of the blame.
Heterarchies are designed to solve crosscutting
problems by leveraging crosscutting collaboration –
for insight into how hospitals grappled and ultimately
embraced heterarchy through the re-engineering of
readmissions (Sobczak, 2014).
Most large conglomerates are heterarchies in
name only. More likely, they behave as coalitions of the
willing encouraging collaboration up to a point, until
collaboration clashes with the chain of command (e.g.
the hierarchy). Three ways to ensure sustainable collaboration are: account for it in policy, plan for it in
network design, and incentivize it with consistent rewards that are integrated with individual performance
measures.
Concluding Remarks

The directive of this article is to focus on theoretical
concepts. Therefore, in closing, I now comment on
methods that impact measurement, standards and policy.
As for measurement, methods of analyzing heterarchies are dangerously derived from social network
analysis (SNA; sometimes abbreviated as ONA, for
organizational network analysis). Unfortunately social network analysis as a general practice is young,
and most practitioners are reduced to banal attempts
at reading or diagnosing networks: closing gaps and
dispersing cliques. There are no standards for diagnosing networks. and therefore no best practices to offer
guidance on how to translate network metrics to incentives – for example, correlating centrality measures
with performance reviews (Stephenson, 2011). The
goal is to derive a more accurate valuation of the total
human asset to ensure that people effectively work together for the common good.
As for standards, there are as yet no methods
or standards for ‘writing’ networks, i.e. creating and
designing networks as human way stations for connecting hierarchies into sustainable heterarchies. The
21st-century is a small, hot and crowded place, and fer-

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

Jaargang 99 (4698S) 20 november 2014

Governance ESB

tile ground for heterarchies to spontaneously form and
disperse. We have been buffeted by their wicked unruliness in dispensing toxic information through education, health, finance and civil society (e.g. terrorism).
We must develop and deploy best practices for the lifecycle management of these heterarchies, for otherwise
we deserve the fate their failures demonstrate.
As for policy, one could argue that the call for
standards in the practice and measurement of networks
and heterarchies is a perverse outcome of antitrust policy laws enacted in the 19th and 20th centuries. The
Sherman Act of 1890 and the Clayton Act as well as
the Federal Trade Commission Act of 1914 were developed to regulate the conduct of business corporations
– e.g. those hierarchical islands of planned coordination in the sea of market relations. Their alleged purpose
was protecting competition, based on the belief that a
free, unregulated market would inevitably lead to the
establishment of coercive monopolies. The writer Ayn
Rand and other laissez faire economists argued that
if any coercive monopolies existed, they were due to
government intervention, not the lack of it. Antitrust
is now iconic; it insulates hierarchies from collaborating and undermines the credibility and sustainability
of heterarchies. It’s unsettling when government can
break up an AT&T in the early 1980s with one hand,
but cobble together a heterarchy called The Affordable
Care Act in 2010 with the other hand. Clearly we need
to rethink governance in the light of larger needs that
only organizational conglomerates like heterarchies
can meet. At issue is the greater moral good or, said
differently, who is being served.
In the introduction, I suggested that we are onto
something very big regarding 21st-century governance.
The mixing of outdated 19th-century hierarchical policies – for example antitrust laws – with 21st-century
social media has created a leadership vacuum that only
heterarchical governance can fill. Effective hierarchical
leadership is no longer a guarantee of success – our
ever-shrinking world has seen to that. Heterarchical
governance has quite literally evolved to meet our collective needs. Heterarchy unburdens us through better
governance and may prove to be the antidote for outdated hierarchical policies that we could and should
shrug off.

Literatuur
Chandler, A. (1977) The Visible Hand. Cambridge: Harvard University Press.
Coase, R. (1937) The Nature of the Firm. Economica, 6(16), 386.
Douglas, M. (1986) How Institutions Think. Syracuse: Syracuse University Press.
Krackhardt, D. (1990) Assessing the Political Landscape: Structure, Cognition and
Power in Organizations. Administrative Science Quarterly, 35(2). 342–369.
Levi-Strauss, C. (1955) The mathematics of man. International Social Science Bulletin,
6(4), 581–590.
Levi-Strauss, C. (1969) Elementary Structures of Kinship. Boston: Beacon Press.
McCulloch, W.A. (1945) A Heterarchy of Values Determined by the Topology of Nervous Nets. Bulletin of Mathematical Biophysics, 7, 89–93.
McGregor, D. (1957, 2006) The Human Side of Enterprise. New York: McGraw Hill.
Olson, M. (1974) The Logic of Collective Action: Public Goods and the Theory of Groups.
Cambridge: Harvard University Press.
Ouchi, W. (1981) Theory Z. New York: Addison-Wesley Publishing.
Powell, W. (1990) Neither Market Nor Hierarchy: Network Forms of Organization.
Organizational Behavior, 12, 295–336.
Rand, A. (1957) Atlas Shrugged. New York: Plume Publishing (Penguin).
Rand, A. (1962) Antitrust: The Rule of Unreason. The Objectivist Newsletter, 1.
Richardson, G. (1972) The organization of industry. Economic Journal, 82(9),
883–896.
Sahlins, M. (1961) The Segmentary Lineage: An Organization for Predatory Expansion. American Anthropologist, 63, 322–345.
Sahlins, M. (1963) Poor man, rich man, big man, chief: political types in Melanesia
and Polynesia. Comparative Studies in Society and History, 5, 285–303.
Smith, A. (1776) The Wealth of Nations. Oxford: Oxford University Press.
Sobczak, S. (2014) Wisconsin Hospitals Tackle Readmissions with Inside/Outside.
Healthcare Executive. Artikel te vinden op www.issuu.com
Stark, D. (1999) Heterarchy: Distributing Authority and Organizing Diversity. In:
Clippinger, J. (ed.) The Biology of Business: Decoding the Natural Laws of Enterprise.
New York: Jossey-Bass, 153–180.
Stephenson, K. and M. Zelen (1989) Rethinking centrality. Social Networks, 11(1), 1–37.
Stephenson, K. (1990) The emergence of virtual groups. Ethnology, 29(4), 279–296.
Stephenson, K. (2004) Towards a Theory of Government. London: DEMOS. Artikel op
www.demos.co.uk.
Stephenson, K. (2008) Rethinking Governance: Conceptualizing Networks and
Their Implications for New Mechanisms of Governance Based on Reciprocity. In:
Williamson, T. (ed.) The Handbook of Knowledge-Based Policing: Current Conceptions
and Future Directions. London: John Wiley & Sons, 323–341.
Stephenson, K. (2009) Neither Hierarchy Nor Network: An Argument for Heterarchy. People and Strategy, 32(1), 4–13.
Stephenson, K. (2011) From Tiananmen to Tahrir: Knowing one’s place in the 21st
century. Organizational Dynamics, 40(4), 281–291.
Williamson, O. (1975) Markets and Hierarchies. New York: The Free Press.
Williamson, O. (1985) Economic Institutions of Capitalism. New York: The Free Press.
Williamson, O. (1993) The Mechanisms of Governance, New York: Oxford University
Press, 219-250.

Jaargang 99 (4698S) 20 november 2014

De auteur heeft verklaard dit artikel alleen te publiceren in ESB en niet elders
te publiceren in wat voor medium dan ook. Het is wel toegestaan om het artikel voor eigen gebruik
en voor publicatie op een intranet van de werkgever van de auteur aan te wenden.

59