I 156 I Karen S.

Cook, Russell Hardin, and Margaret Levi I

\'ele1-lbane1, Carlm G. 1983. Bond) of M11t1111/ Tru t: rite C11lt11ml li1·,tt·111s of Rott11i11g (:red,t
A5�1,cit111om mmmg Url•an Atr.\imns ,111d Cl1irn11os.• ew Brun,wid., J: Ru1gcr, Univcr,ih
Pre,-.. Part Ill
\\'ath, Dun an J. 2003. Six D,·g,r,•e;;: I ht Sc ieme ofII Co1111ate,I .Age. New York: ,orton.
Y.im.ig1,hi, 1o,hio, and Karen 'i. Cook. 1993. "Generali,cd hchange and SodJI Dilemma�."
\t:,, 1,11 Ard1ology Q11artt·rly '>6 4 ': .235-48.

Institutions and Networks

Yunu,, ,\1uhamma<l. 1998. ",\lbiating Po,crtr through fo.hnolog).� \m•11c,· (October I 6):
409-10.
Yunu-., �luhammad. 1999. 81111ker w tltt Poor: .\ficn>•l endmg and tltt' R11ttle ,\gc1i11st \\'oriel
Po1·rrty. 'cw York: Public Affair-..

Introduction to PJrt Ill

10 Economic Embeddcdnc-,s

11 The Iron (Jgc Rcvi.,itcd

12 Catncts

13 Structural llolcs
 Introduction to Part III

The concept of "institution" is one of the most enduring, and poorly defined,
concepts in sociology. While this theoretical ambiguity has led some theorists to
avoid the term, like "structure," "institution" seems to fill a rhetorical role that
sociologists cannot do without (Sewell, 1992). In what follows, our institutional
analysis focuses on the enduring social patterns that, ultimately, are embedded in
social networks. DiMaggio and Powell's (1983) work on institutional isomorphism
shows why certain organizational features dominate, even when they are not
economically rational. Mark Granovetter similarly approaches individual economic
action by asking how actions are conditioned by network contexts, providing a key
insight into the problem of social embeddedness. The link between enduring social
patterns and networks is made explicit in a classic piece by Harrison White that,
while never published, had been in photocopy circulation for over 40 years. The turn
toward networks is completed with a selection from Ronald Burt's classic Structural
Holes; here we make a move back from large-scale enduring social institutions to a
network mechanism for profit in economic settings. Given the breadth of literature
on "institutions", we start next by briefly reviewing the history of institutions
through Parsons, then introduce each author's wider work.
Most theories of social institutions are rooted in Durkheim's treatment of social
facts, and in most work since Parsons institutions have been used as a foil for
atomistic approaches to social life (DiMaggio and Powell, 1983). As with much of
contemporary social theory, the shadow of Parsons falls heavily on current theories
of institutions. Parsons based his subjective conception of institutions on individual
action, arguing that institutions provide goal-oriented actors with means that are
consistent with the overall value system of the society. That is, institutions reduce to
"a set of regulatory norms that give rise to social structure or organization"
(Coleman, 1990: 334).' Importantly, institutions for Parsons are sets of norms that

Contemporary Sociological Theory, Third Edition. Edited by Craig Calhoun, Joseph Gerteis,
James Moody, Steven Pfaff, and Indermohan Virk. Editorial material and organization
©2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
 160 Introduction to Part
go together; they are "a complex of institutionalized role integrates" (Parsons, 1951: p.
39) and are thus broader than norms.
As we would expect, Parsons argues that institutions are functional, contributing
to the social order by ensuring that individual behavior conforms to societal interests.
In addition, "the 'function' of institutions is not only to keep the means and ends of
the intermediate chain in conformity with ultimate values, but also to endow the
individual with the energy to live up to institutional norms" (Parsons, 1990 [1934]:
325). Institutions, however, are not legitimized solely because people have internalized
the ultimate values implicit in the norm. Instead, institutions persist because "once
really established, a system of institutional norms creates an interlocking of interests"
that help keep it in place, even if individual devotion to the underlying values start to
wane (Parsons, 1990 [1934]: 326), and this "interlocking" of interests is part of the
natural affinity linking institutional questions to social networks.
According to Coleman (1990), Parsons' dual emphasis on the functional aspects
of institutions and the subjective focus of institutional analysis was a fundamental
flaw. By focusing on how individuals experience institutions, it is impossible to
identify how they form, because the creation of norms rests on interdependent
action. That is, showing that an institution is in an actor's long-term interest is not
sufficient to explain the formation of the institution, since it might be in his or her
short-term interest to violate the norm. Thus, "What [Parsons] failed to recognize
was that the path from action to system lies in the relation between different actors'
actions - it is there that the complexities arise and generate different kinds of systems
of action" (p. 338).
Identifying the role of institutions in social life has had something of a renascence
in the late 1980s and early 1990s, especially within economic sociology (see Powell and
DiMaggio, 1991; Winship and Rosen, 1988). Much of this work has been a reaction to
treatments of social life based on rational actor models, and are often "united by little
but a common skepticism toward atomistic accounts of social processes and a common
conviction that institutional arrangements and social processes matter" (Powell and
DiMaggio, 1991: 3). This attack on individualism is made apparent in work by
Williamson (1975; 1985) that identifies weaknesses with standard microeconomic
assumptions. In Williamson's model, actors have limited cognitive capacity, poor
information, often cannot monitor agreements and act opportunistically. Under these
conditions, institutions develop to solve resulting organizational problems. For
example, institutions reduce uncertainty and lower transaction costs, providing an
efficient solution to bounded individual action. How institutions come to be, the
extent to which they are economically efficient, and how they evolve, are all hotly
debated topics within institutional theories of economic sociology.
Work on organizational behavior often starts with a process of rationalization
based on Weber's famous prediction that organizations, through the dual engines of
competition and rationalization, should become ever more bureaucratically rational.
Empirical researchers, however, have had difficulty describing organizational
behavior in such terms. Instead, inefficiency and non-productive ritual behavior is
common (DiMaggio and Powell, 1983). Moreover, while organizations face many
Introduction to Part III 161
different challenges, the range of organizational responses is remarkably small,
leading one to ask why organizations are so similar.
DiMaggio and Powell's answer rests on the process of isomorphism, which they
define as "a constraining process that forces one unit in a population to resemble
other units that face the same set of environmental conditions" (p. 149). They
identify two types of isomorphism, competitive and institutional, and argue that
rationalistic competition is not sufficient to explain most organizational similarity.
Instead, competitive isomorphism must be supplemented by three non-competitive
features of the organizational milieu: "1) Coercive isomorphism that stems from
political influence and the problem of legitimacy; 2) mimetic isomorphism resulting
from standard responses to uncertainty; and 3) normative isomorphism, associated
with professionalization" (p. 150). This approach to understanding organizational
behavior rests on a multi-level model of organizations; pressures at the organizational
level (a search for legitimacy, standard conventions to control uncertainty) are
combined with the socialization process of firm managers (professionalization).
DiMaggio and Powell challenge theories that rest on elite control or simple population
dynamics, arguing "that a theory of institutional isomorphism may help explain the
observation that organizations are becoming more homogeneous, and that elites
often get their way, while at the same time enabling us to understand the irrationality,
the frustration of power, and the lack of innovation that are so commonplace in
organizational life" (p. 157).
Granovetter's "Economic Action and Social Structure: The Problem of
Embeddedness" (1985) is among the most-cited papers ever published in sociology
and turns our discussion of institutions from the institutions themselves to the
problem of action embedded in economic institutions. Granovetter asks how
behavior and institutions are affected by social relations and directly challenges key
questions in the "institutional economics" approach represented by DiMaggio and
Powell. Granovetter, who is well-known for work on the power of seemingly
unimportant "weak" social ties, draws a middle position for economic action
between purely egoistic, autonomous, and economically rational actors
("undersocialized action") and deeply embedded, socially constrained and largely
scripted normative behavior ("oversocialized action"). Instead, Granovetter posits
an "embeddedness argument" that centers on "concrete personal relations and
structures," arguing that social relations generate trust and discourage malfeasance.
Since people's patterns of connections to others vary in strength and extent, this
conception provides a clear mechanism to account for variability in observed trust.
If institutional analysis has, at its heart, the rejection of atomistic action theory as
the starting point for sociology, theory and research on social networks represents this
notion taken to full fruition. Harrison White is among the most well-known
sociologists working in the area of social networks. His early work on the structural
implications of kinship patterns (1963) formalized earlier ideas from structural
anthropology and laid the foundation for his later work on abstracting relevant
structural patterns from observed networks. Kinship systems provide a clear example
of how primary relations (marriage and descent) can extend to other known roles

162 Introduction to Part III
(example: your mother's mother is your "grandmother," and your mother's mother's
daughter's daughter is your cousin). The various concatenations of the two primary
relations can be used to describe the full kinship system. White recognized that you
can extend this argument to other kinds of relations and discover roles in systems by
empirically tracing the most common patterns of extended network ties. Known as
"blockmodeling," this approach founded a long tradition of research building on
earlier role theories. The ultimate, abstract extension of these ideas takes White beyond
individual actors to "identities" and from the specific analyses of particular network
structures to the general strategies actors use to gain control (White, 1992; 2004).
White's second major contribution to social theory is a direct attack on classical
economic market models. In a series of papers culminating in his book Markets
from Networks (2004), White demonstrates that the basic competition model for
commodities that is the foundation for most work on markets is really just a special
case of the many possible ways markets can be organized. Instead of focusing on the
supply and demand for commodities, White focuses on observable relations among
product producers and how they negotiate a tradeoff between the quality of goods
produced and the prices for those goods in comparison to similar other firms. In
White's model, firms choose a position along a quality-price array to offer goods.
For example, WalMart seeks to offer low-quality goods at the lowest prices, while
Target offers slightly higher quality goods at higher prices. As you move up the
quality-price curve, you would find retailers such as Macy s deliberately avoiding
low prices in an effort to signal high quality. This insight has deep implications for
market failures, prices and control.
The piece reprinted below existed as a mimeographed copy in circulation among
White's students (and students of students of students) and has never been published
before. While we are typically used to thinking of categories such as "race" or "sex"
as essential fixed characteristics of people, White argues for a conception of categories
that rests on the correspondence of network ties (Nets) with categories (Cats).
A "catnet" is thus the correspondence between these two features of a population.
Substantively, the idea reflects notions that we regularly observe: category
membership (such as being "male" or "female") is only relevant to the degree that it
shapes our relations with others.
We conclude this section with a chapter from Ronald Burt s book Structural Holes.
Combining the insights of Georg Simmel and Granovetter's work on the strength of
weak ties, Burt examines in detail the social-structural conditions of competition
that lead to an ability to profit from networks. Seen as explicating the mechanism
behind the weak-tie formulation, Burt focuses on the returns to linking others who
are not otherwise connected. Filling a "structural hole - creating a bridge between
two people who are not connected - puts one in a unique position to profit from
their activity. The profit is gained through both information (the efficiency of
networks is built on having non-redundant contacts) and arbitrage - the ability to
extract resources by mediating the interests of the two parties. A modern, formal
representation of Simmel's classic tertius gaudens, Burt's work on structural holes
has become a classic touchstone for modern network analysis.
Introduction to Part III 163
NOTE
1 "This system of regulatory norms, of rules governing actions in pursuit of immediate ends in
terms of their conformity with the ultimate common value-system of community, is what I call
its institutions approached from the subjective point of view." (Parsons, 1990 [1934]: p. 324)
SELECTED BIBLIOGRAPHY
Burt, Ronald S. 2005. Brokerage and Closure: An Introduction to Social Capital.Oxford: Oxford
University Press. (A nice introduction to the wider field of social capital research.)
Burt, Ronald S. 2010. Neighbor Networks: Competitive Advantage Local and Personal. Oxford:
Oxford University Press. (A re-examination of the basic idea of structural hole theory,
attempting to incorporate new evidence, building on the personal work actors have to do
to make use of advantageous network positions.)
Coleman, James. 1990. "Commentary: Social Institutions and Social Theory." American
Sociological Review 55: 333-9. (A discussion of Parsons' paper that includes Coleman's
own understanding of what an institution is and how theory should approach the study of
institutions.)
DiMaggio, Paul J. and Powell, Walter W. 1983. "The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields." American Sociological
Review 48: 147-60. (One of the foundational pieces in the application of institutional ideas
in modern organizational theory.)
Form, William. 1990. "Institutional Analysis: An Organizational Approach." Pp. 257-71 in
Change in Societal Institutions, Maureen T. Hallinan, David M. Klein, and Jennifer Glass.
Plenum Publishing Corporation. (A nice review of the concept of institution, providing a
new approach based on inter-organizational connections.)
Maclean, Alair and Olds Andy. 2001. "Interview with Harrison White" http://www.ssc.wisc.
edu/theory@madison/papers/ivwWhite.pdf (a nice interview with Harrison White that
describes his background and much of his work.)
Parsons, Talcott. 1951. The Social System. Glencoe, IL: Free Press.
Parsons, Talcott. 1990 [1934]. "Prolegomena to a Theory of Social Institutions." American
Sociological Review 55: 319-33. (A recently discovered piece that outlines Parsons early
thinking on institutions.)
Powell, Walter W. and DiMaggio, Paul J. 1991. The New Institutionalism in Organizational
Analysis. Chicago, IL: University of Chicago Press. (The first chapter of this book provides
a nice treatment of institutions as they relate to modern organizational theory.)
Sewell, William H. Jr. 1992. "A Theory of Structure: Duality, Agency, and Transformation."
American Journal of Sociology 98: 1-29. (A good discussion of the meaning of "structure"
in social theory. Insights here can readily be linked to the problems of various meanings of
"institution.")
White, Harrison. 1963. An Anatomy of Kinship: Mathematical Models for Structures of
Cumulated Roles. Englewood-CIiffs, NJ. (An early piece that sets the mathematical stage for
much later work in social networks.)
White, Harrison. 1992. Identity and Control: A Structural Theory of Social Action. Princeton,
NJ; Princeton University Press. (This book is known to be difficult going, but provides a
theoretical model for linking networks to the key questions of institutions and action. The
second edition (2010) is really a distinct book, and a major re-thinking of the ideas.)
White, Harrison. 2004. Markets from Networks: Soaoeconom.c Models of Production.P™«ton
NT: Princeton University Press. (Whites ultimate statement of hts model for markets. A key
text to build a sociological extension of market models.)
Chapter 10
Williamson, Oliver E. 1975. Markets and Hierarchies. New York: Free Press.
Williamson, Oliver E. 1985. The Economic Institutions of Capitalism. New York. Free Press.
(These two books are classics in the application of institutional ideas to economic

Winship, Christopher and Rosen, Sherwin. 1988."Introduction: Sociological and Economic

Economic Embeddedness [1985]
Approaches to the Analysis of Social Structure." American Journal of Sociology 94.sl^sl6.
(Contains a nice discussion of how sociological ideas of institutions contribute to
Mark Granovetter
understanding economic problems.)

Introduction: the Problem of Embeddedness

How behavior and institutions are affected by social relations is one of the classic
questions of social theory. Since such relations are always present, the situation that
would arise in their absence can be imagined only through a thought experiment
like Thomas Hobbes's "state of nature" or John Rawls's "original position." Much of
the utilitarian tradition, including classical and neoclassical economics, assumes
rational, self-interested behavior affected minimally by social relations, thus
invoking"ari idealized state not far from that of these thought experiments. At the
other extreme lies what I call the argument of "embeddedness": the argument that
the behavior and institutions to be analyzed are so constrained by ongoing social
relations that to construe them as independent is a grievous misunderstanding.
[...]
My own view diverges from both schools of thought. I assert that the level of
embeddedness of economic behavior is lower in nonmarket societies than is
claimed by substantivists and development theorists, and it has changed less
with "modernization" than they believe; but I argue also that this level has always
been and continues to be more substantial than is allowed for by formalists and
economists. [...)

Mark Granovetter, "Economic Embeddedness," pp. 481-2, 482-8, 488-9, 490-2, 492-3, 508-10
from Mark Granovetter "Economic Action and Social Structure: The Problem of Embeddedness,"
American Journal of Sociology, 91: 3 (November, 1985). Copyright © 1985 by American Journal of
Sociology. Reprinted by permission of The University of Chicago Press.

Contemporary Sociological Theory, Third Edition. Edited by Craig Calhoun, Joseph Gerteis,
James Moody, Steven Pfaff, and Indermohan Virk. Editorial material and organization
©2012 John Wiley 8c Sons, Ltd. Published 2012 by John Wiley 8c Sons, Ltd.
 166 Mark Granovetter
Over- and Undersocialized Conceptions of Human Action
in Sociology and Economics
I begin by recalling Dennis Wrong's 1961 complaint about an oversocialized
conception of man in modern sociology" - a conception of people as overwhelmingly
sensitive to the opinions of others and hence obedient to the dictates of
consensually developed systems of norms and values, internalized through
socialization, so that obedience is not perceived as a burden. [...]
Classical and neoclassical economics operates, in contrast, with an atomized,
undersocialized conception of human action, continuing in the utilitarian tradition.
The theoretical arguments disallow by hypothesis any impact of social structure
and social relations on production, distribution, or consumption. In competitive
markets, no producer or consumer noticeably influences aggregate supply or
demand or, therefore, prices or other terms of trade. As Albert Hirschman has
noted, such idealized markets, involving as they do "large numbers of price-taking
anonymous buyers and sellers supplied with perfect information ... function
without any prolonged human or social contact between the parties. Under perfect
competition there is no room for bargaining, negotiation, remonstration or mutual
adjustment and the various operators that contract together need not enter into
recurrent or continuing relationships as a result of which they would get to know
each other well" (1982: 1473).
It has long been recognized that the idealized markets of perfect competition
havejurviveiL''nrpllprt";fl attack in part because self-regulating economic structures
are politically attractive to many. Another reason for this survival, less clearly
understood, is that the elimination of social relationsjrom^economic analysis
. Jk removes the problenToforder from the intellectual agenda, at least in the economic
J sphere. In Hobbes's argument, disorder arises because conflict-free social and
W t •' economic transactions depend on trust and the absence of malfeasance. But these
V i are unlikely when individuals are conceived to have neither social relationships
nor institutional context - as in the "state of nature." Hobbes contains the difficulty
by superimposing a structure of autocratic authority. The solution of classical
liberalism, and correspondingly of classical economics, is antithetical: repressive
political structures are rendered unnecessary by competitive markets that make
force or fraud unavailing. Competition determines the terms of trade in a way that
individual traders cannot manipulate. If traders encounter complex or difficult
relationships, characterized by mistrust or malfeasance, they can simply move on to
the legion of other traders willing to do business on market terms; social relations
and their details thus become frictional matters.
In classical and neoclassical economics, therefore, the fact that actors may have
social relations with one another has been treated, if at all, as a frictional drag that
impedes competitive markets. In a much-quoted line, Adam Smith complained that
"people of the same trade seldom meet together, even for merriment and diversion,
but the conversation ends in a conspiracy against the public, or in some contrivance
Economic Embeddedness 167
to raise prices." His laissez-faire politics allowed few solutions to this problem, but
he did suggest repeal of regulations requiring all those in the same trade to sign a
public register; the public existence of such information "connects individuals who
might never otherwise be known to one another and gives every man of the trade a
direction where to find every other man of it." Noteworthy here is not the rather
lame policy prescription but the recognition that social atomization is prerequisite
to perfect competition (Smith, 1979 [1776]: 232-3).
More recent comments by economists on "social influences" construe these as
processes in which actors acquire customs, habits, or norms that are followed
mechanically and automatically, irrespective of their bearing on rational choice.
This view, close to Wrong's "oversocialized conception," is reflected in James
Duesenberry's quip that "economics is all about how people make choices;
sociology is all about how they don't have any choices to make" (1960: 233) and in
E. H. Phelps Brown's description of the "sociologists'approach to pay determination"
as deriving from the assumption that people act in "certain ways because to do so
is customary, or an obligation, or the 'natural thing to do,' or right and proper, or
just and fair" (1977, p. 17).
But despite the apparent contrast between under- and oversocialized views, we
should note an irony of great theoretical importance: both have in common a
conception of action and decision carried out by atomized actors. In the
undersocialized account, atomization results from narrow utilitarian pursuit of self-
interest; in the oversocialized one, from the fact that behavioral patterns have been
internalized and ongoing social relations thus have only peripheral effects on
behavior. That the internalized rules of behavior are social in origin does not
differentiate this argument decisively from a utilitarian one, in which the source
of utility functions is left open, leaving room for behavior guided entirely by
consensually determined norms and values - as in the oversocialized view. Under-
and oversocialized resolutionsof the problem of order thus merge in their atomization
of actors from immediate social context. This ironic merger is already visible in
Hobbes's Leviathan, in which the unfortunate denizens of the state of nature,
overwhelmed by the disorder consequent to their atomization, cheerfully surrender
all their rights to an authoritarian power and subsequently behave in a docile and
honorable manner; by the artifice of a social contract, they lurch directly from an
undersocialized to an oversocialized state.
When modern economists do attempt to take account of social influences, they
typically represent them in the oversocialized manner represented in the quotations
above. In so doing, they reverse the judgment that social influences are frictional but
sustain the conception of how such influences operate. In the theory of segmented
labor markets, for example, Michael Piore has argued that members of each labor
market segment are characterized by different styles of decision making and that the
making of decisions by rational choice, custom, or command in upper-primary,
lower-primary, and secondary labor markets respectively corresponds to the origins
of workers in middle-, working-, and lower-class subcultures (Piore, 1975). Similarly,
Samuel Bowles and Herbert Gintis, in their account of the consequences of American
 Economic Embeddedness 169
168 Mark Granovetter
conception - that of actors behaving exclusively in accord with their prescribed
education, argue that different social classes display different cognitive processes
roles - to implement an atomized, undersocialized view.
because of differences in the education provided to each. Those destined for
A fruitful analysis of human action requires us to avoid the atomization implicit
lower-level jobs are trained to be dependable followers of rules, while those who will
in the theoretical extremes of under- and oversocialized conceptions. Actors do
be channeled into elite positions attend "elite four-year colleges" that "emphasize
not behave or decide as atoms outside a social context, nor do they adhere slavishly
social relationships conformable with the higher levels in the production
to a script written for them by the particular intersection of social categories that
hierarchy. [...] As they 'master' one type of behavioral regulation they are either
they happen to occupy. Their attempts at purposive action are instead embedded
allowed to progress to the next or are channeled into the corresponding level in the
in concrete, ongoing systems of social relations. In the remainder of this article
hierarchy of production" (Bowles and Gintis, 1975: 132).
I illustrate how this view of embeddedness alters our theoretical and empirical
But these oversocialized conceptions of how society influences individual
approach to the study of economic behavior. I first narrow the focus to the
behavior are rather mechanical: once we know the individual's social class or labor
question of trust and malfeasance in economic life and then use the "markets and
market sector, everything else in behavior is automatic, since they are so well
hierarchies" problem to illustrate the use of embeddedness ideas in analyzing
socialized. Social influence here is an external force that, like the deists' God, sets
things in motion and has no further effects - a force that insinuates itself into the this question.
minds and bodies of individuals (as in the movie Invasion of the Body Snatchers),
altering their way of making decisions. Once we know in just what way an individual
has been affected, ongoing social relations and structures are irrelevant. Social
Embeddedness, Trust, and Malfeasance in Economic Life
influences are all contained inside an individual s head, so, in actual decision
Since about 1970, there has been a flurry of interest among economists in the
situations, he or she can be atomized as any Homo economicus, though perhaps with
previously neglected issues of trust and malfeasance. Oliver Williamson has noted
different rules for decisions. More sophisticated (and thus less oversocialized)
that real economic actors engage not merely in the pursuit of self-interest but also
analyses of cultural influences (e.g., Fine and Kleinman 1979; Cole 1979, chap. 1)
in "opportunism" - "self-interest seeking with guile; agents who are skilled at
make it clear that culture is not a once-for-all influence but an ongoing process,
dissembling realize transactional advantages. Economic man ... is thus a more
continuously constructed and reconstructed during interaction. It not only shapes
subtle and devious creature than the usual self-interest seeking assumption reveals"
its members but also is shaped by them, in part for their own strategic reasons.
(1975, p. 255).
Even when economists do take social relationships seriously, as do such diverse
But this points out a peculiar assumption of modern economic theory, that
figures as Harvey Leibenstein (1976) and Gary Becker (1976), they invariably
one's economic interest is pursued only by comparatively gentlemanly means. The
abstract away from the history of relations and their position with respect to
Hobbesian question - how it can be that those who pursue their own interest do not
other relations - what might be called the historical and structural embeddedness
do so mainly by force and fraud - is finessed by this conception. Yet, as Hobbes
of relations. The interpersonal ties described in their arguments are extremely
saw so clearly, there is nothing in the intrinsic meaning of "self-interest" that excludes
stylized, average, "typical" - devoid of specific content, history, or structural
force or fraud.
location. Actors' behavior results from their named role positions and role sets;
[...]
thus we have arguments on how workers and supervisors, husbands and wives, or
What has eroded this confidence in recent years has been increased attention
criminals and law enforcers will interact with one another, but these relations are
to the micro-level details of imperfectly competitive markets, characterized by small
not assumed to have individualized content beyond that given by the named
numbers of participants with sunk costs and "specific human capital" investments.
roles. This procedure is exactly what structural sociologists have criticized in
In such situations, the alleged discipline of competitive markets cannot be called
Parsonian sociology - the relegation of the specifics of individual relations to a
on to mitigate deceit, so the classical problem of how it can be that daily economic
minor role in the overall conceptual scheme, epiphenomenal in comparison with
life is not riddled with mistrust and malfeasance has resurfaced.
enduring structures of normative role prescriptions deriving from ultimate
In the economic literature, I see two fundamental answers to this problem and
value orientations. In economic models, this treatment of social relations has
argue that one is linked to an undersocialized, and the other to an oversocialized,
the paradoxical effect of preserving atomized decision making even w en
conception of human action. The undersocialized account is found mainly in the
decisions are seen to involve more than one individual. Because the analyzed set of
new institutional economics - a loosely defined confederation of economists with
individuals - usually dyads, occasionally larger groups - is abstracted out of social
an interest in explaining social institutions from a neoclassical viewpoint. (See,
context, it is atomized in its behavior from that of other groups and from the
e.g., Furubotn and Pejovich 1972; Alchian and Demsetz 1973; Lazear 1979; Rosen
history of its own relations. Atomization has not been eliminated, merely
1982; Williamson 1975,1979, 1981; Williamson and Ouchi 1981.) The general story
transferred to the dyadic or higher level of analysis. Note the use of an oversocialized

170 [MarkGranovetter[
told by members of this school is that social institutions and arrangements previously
thought to be the adventitious result of legal, historical, social, or political forces
are better viewed as the efficient solution to certain economic problems. The tone
is similar to that of structural-functional sociology of the 1940s to the 1960s,
and much of the argumentation fails the elementary tests of a sound functional
explanation laid down by Robert Merton in 1947. Consider, for example, Schotters
view that to understand any observed economic institution requires only that we
"infer the evolutionary problem that must have existed for the institution as we
see it to have developed. Every evolutionary economic problem requires a social
institution to solve it" (1981, p. 2).
Malfeasance is here seen to be averted because clever institutional arrangements
make it too costly to engage in, and these arrangements- many previously interpreted
as serving no economic function - are now seen as having evolved to discourage
malfeasance. Note, however, that they do not produce trust but instead are a
functional substitute for it. The main such arrangements are elaborate explicit and
implicit contracts (Okun 1981), including deferred compensation plans and
mandatory retirement - seen to reduce the incentives for "shirking on the job or
absconding with proprietary secrets (Lazear 1979; Pakes and Nitzan 1982)- and
authority structures that deflect opportunism by making potentially divisive
decisions by fiat (Williamson 1975). These conceptions are undersociahzed in that
they do not allow for the extent to which concrete personal relations and the
obligations inherent in them discourage malfeasance, quite apart from institution
arrangements. Substituting these arrangements for trust results actua y in a
Hobbesian situation, in which any rational individual would be motivated to develop
clever ways to evade them; it is then hard to imagine that everyday economic life
would not be poisoned by ever more ingenious attempts at deceit.
Other economists have recognized that some degree of trust must be assumed to
operate, since institutional arrangements alone could not entirely stem force or
fraud. But it remains to explain the source of this trust, and appeal is sometimes
made to the existence of a "generalized morality." Kenneth Arrow, for example,
suggests that societies, "in their evolution have developed implicit agreements to
certain kinds of regard for others, agreements which are essential to the survival
of the society or at least contribute greatly to the efficiency of its working (1974,
p. 26; see also Akerlof [1983] on the origins of"honesty").
Now one can hardly doubt the existence of some such generalized morality;
without it, you would be afraid to give the gas station attendant a 20-dollar bill when
you had bought only five dollars' worth of gas. But this conception has the
oversocialized characteristic of calling on a generalized and automatic response,
even though moral action in economic life is hardly automatic or universal (as is
well known at gas stations that demand exact change after dark).
The embeddedness argument stresses instead the role of concrete personal
relations and structures (or "networks") of such relations in generating trust
and discouraging malfeasance. The widespread preference for transacting with
Economic Embeddedness 171
individuals of known reputation implies that few are actually content to rely on
either generalized morality or institutional arrangements to guard against trouble.
Economists have pointed out that one incentive not to cheat is the cost of damage to
one's reputation; but this is an undersociahzed conception of reputation as a
generalized commodity, a ratio of cheating to opportunities for doing so. In practice,
we settle for such generalized information when nothing better is available, but
ordinarily we seek better information. Better than the statement that someone is
known to be reliable is information from a trusted informant that he has dealt with
that individual and found him so. Even better is information from one s own past
dealings with that person. This is better information for four reasons: (1) it is cheap;
(2) one trusts one's own information best - it is richer, more detailed, and known to
be accurate; (3) individuals with whom one has a continuing relation have an
economic motivation to be trustworthy, so as not to discourage future transactions,
and (4) departing from pure economic motives, continuing economic relations
often become overlaid with social content that carries strong expectations of trust
and abstention from opportunism.
It would never occur to us to doubt this last point in more intimate relations,
which make behavior more predictable and thus close off some of the fears that
create difficulties among strangers. Consider, for example, why individuals in a
burning theater panic and stampede to the door, leading to desperate results.
Analysts of collective behavior long considered this to be prototypically irrational
behavior, but Roger Brown (1965, chap. 14) points out that the situation is
essentially an n-person Prisoner's Dilemma: each stampeder is actually being quite
rational given the absence of a guarantee that anyone else will walk out calmly,
even though all would be better off if everyone did so. Note, however, that in the
case of the burning houses featured on the 11:00 P.M. news, we never hear that
everyone stampeded out and that family members trampled one another. In the
family, there is no Prisoner's Dilemma because each is confident that the others can
be counted on. T
In business relations the degree of confidence must be more variable, but
Prisoner's Dilemmas are nevertheless often obviated by the strength of personal
relations, and this strength is a property not of the transactors but of their concrete
relations. Standard economic analysis neglects the identity and past relations of
individual transactors, but rational individuals know better, relying on their
knowledge of these relations. They are less interested in general reputations than in
whether a particular other may be expected to deal honestly with them - mainly a
function of whether they or their own contacts have had satisfactory past dealings
with the other. One sees this pattern even in situations that appear, at first glance, to
approximate the classic higgling of a competitive market, as in the Moroccan bazaar
analyzed by Geertz (1979).
Up to this point, I have argued that social relations, rather than institutional
arrangements or generalized morality, are mainly responsible for the production of
trust in economic life. But I then risk rejecting one kind of optimistic functionalism
for another, in which networks of relations, rather than morality or arrangements,

172 Mark Granovetter
are the structure that fulfills the function of sustaining order. There are two ways
to reduce this risk. One is to recognize that as a solution to the problem of order,
the embeddedness position is less sweeping than either alternative argument,
since networks of social relations penetrate irregularly and in differing degrees in
different sectors of economic life, thus allowing for what we already know: distrust,
opportunism, and disorder are by no means absent.
The second is to insist that while social relations may indeed often be a necessary
condition for trust and trustworthy behavior, they are not sufficient to guarantee
these and may even provide occasion and means for malfeasance and conflict on a
scale larger than in their absence. There are three reasons for this.
1. The trust engendered by personal relations presents, by its very existence,
enhanced opportunity for malfeasance. In personal relations it is common
knowledge that "you always hurt the one you love"; that person's trust in you results
in a position far more vulnerable than that of a stranger. (In the Prisoner s Dilemma,
knowledge that one's coconspirator is certain to deny the crime is all the more
rational motive to confess, and personal relations that abrogate this dilemma may
be less symmetrical than is believed by the party to be deceived.) This elementary
fact of social life is the bread and butter of "confidence" rackets that simulate certain
relationships, sometimes for long periods, for concealed purposes. In the business
world, certain crimes, such as embezzling, are simply impossible for those who have
not built up relationships of trust that permit the opportunity to manipulate
accounts. The more complete the trust, the greater the potential gain from
malfeasance. That such instances are statistically infrequent is a tribute to the force
of personal relations and reputation; that they do occur with regularity, however
infrequently, shows the limits of this force.
2. Force and fraud are most efficiently pursued by teams, and the structure of
these teams requires a level of internal trust -"honor among thieves — that usually
follows preexisting lines of relationship. Elaborate schemes for kickbacks and
bid rigging, for example, can hardly be executed by individuals working alone, and
when such activity is exposed it is often remarkable that it could have been kept secret
given the large numbers involved. Law-enforcement efforts consist of finding an
entry point to the network of malfeasance - an individual whose confession
implicates others who will, in snowball-sample fashion, finger still others until the
entire picture is fitted together.
[...]
3. The extent of disorder resulting from force and fraud depends very much on
how the network of social relations is structured. Hobbes exaggerated the extent of
disorder likely in his atomized state of nature where, in the absence of sustained
social relations, one could expect only desultory dyadic conflicts. More extended
and large-scale disorder results from coalitions of combatants, impossible without
prior relations. We do not generally speak of "war" unless actors have arranged
themselves into two sides, as the end result of various coalitions. This occurs only if
there are insufficient crosscutting ties, held by actors with enough links to both main
potential combatants to have a strong interest in forestalling conflict. The same is
Economic Embeddedness 173
true in the business world, where conflicts are relatively tame unless each side can
escalate by calling on substantial numbers of allies in other firms, as sometimes
happens in attempts to implement or forestall takeovers.
Disorder and malfeasance do of course occur also when social relations are absent.
This possibility is already entailed in my earlier claim that the presence of such
relations inhibits malfeasance. But the level of malfeasance available in a truly
atomized social situation is fairly low; instances can only be episodic, unconnected,
small scale. The Hobbesian problem is truly a problem, but in transcending it by the
smoothing effect of social structure, we also introduce the possibility of disruptions
on a larger scale than those available in the "state of nature."
The embeddedness approach to the problem of trust and order in economic life,
then, threads its way between the oversocialized approach of generalized morality
and the undersocialized one of impersonal, institutional arrangements by following
and analyzing concrete patterns of social relations. Unlike either alternative, or
the Hobbesian position, it makes no sweeping (and thus unlikely) predictions of
universal order or disorder but rather assumes that the details of social structure
wrtll ^^tArminp whirVi ic fnunn
REFERENCES
Akerlof, George. 1983. "Loyalty Filters." American Economic Review 73 (1): 54-63.
Alchian, Armen, and Harold Demsetz. 1973. "The Property Rights Paradigm." Journal of
Economic History 33 (March): 16-27.
Arrow, Kenneth. 1974. The Limits of Organization. New York: Norton.
Becker, Gary. 1976. The Economic Approach to Human Behavior. Chicago, IL: University of
Chicago Press.
Bowles, Samuel, and Herbert Gintis. 1975. Schooling in Capitalist America. New York: Basic.
Brown, Roger. 1965. Social Psychology. New York: Free Press.
Cole, Robert. 1979. Work, Mobility and Participation: A Comparative Study of American and
Japanese Industry. Berkeley and Los Angeles, CA: University of California Press.
Duesenberry, James. 1960. Comment on "An Economic Analysis of Fertility." In Demographic
and Economic Change in Developed Countries, edited by the Universities-National Bureau
Committee for Economic Research. Princeton, NJ: Princeton University Press.
Fine, Gary, and Sherryl Kleinman. 1979. "Rethinking Subculture: An Interactionist Analysis."
American Journal of Sociology 85 (July): 1-20.
Furubotn, E., and S. Pejovich. 1972. "Property Rights and Economic Theory: A Survey of
Recent Literature." Journal of Economic Literature 10 (3): 1137-62.
Geertz, Clifford. 1979. "Suq: The Bazaar Economy in Sefrou." Pp. 123-225 in Meaning and
Order in Moroccan Society, edited by C. Geertz, H. Geertz, and L. Rosen. New York:
Cambridge University Press.
Hirschman, Albert. 1982. "Rival Interpretations of Market Society: Civilizing, Destructive
or Feeble?" Journal of Economic Literature 20 (4): 1463-84.
Lazear, Edward. 1979. "Why Is There Mandatory Retirement?" Journal of Political Economy
87 (6): 1261-84.
Leibenstein, Harvey. 1976. Beyond Economic Man. Cambridge, MA: Harvard University
Press.
 174 Mark Granovetter

Merton, Robert. 1947. "Manifest and Latent Functions." pp. 19-84 in Social Theory and Social
Structure. New York: Free Press.
Okun, Arthur. 1981. Prices and Quantities. Washington, D.C.: Brookings. Chapter 11
Pakes, Ariel, and Schmuel Nitzan. 1982. "Optimum Contracts for Research Personnel," NBER
Working Paper no. 871. Cambridge, MA: National Bureau of Economic Research.
Phelps Brown, Ernest Henry. 1977. The Inequality of Pay. Berkeley, CA: University of
California Press.
Piore, Michael. 1975. "Notes for a Theory of Labor Market Stratification." Pp. 125-50 in The Iron Cage Revisited: Institutional
Labor Market Segmentation, edited by R. Edwards, M. Reich, and D. Gordon. Lexington,
MA: Heath. Isomorphism and Collective Rationality
Rosen, Sherwin. 1982. "Authority, Control and the Distribution of Earnings." Bell Journal of
Economics 13 (2): 311-23. in Organizational Fields [1983]
Schotter, Andrew. 1981. The Economic Theory of Social Institutions. New York: Cambridge
University Press. Paul J. DiMaggio and Walter W. Powell
Smith, Adam. 1979(1776). The Wealth of Nations, edited by Andrew Skinner. Baltimore:
Penguin.
Williamson, Oliver. 1975. Markets and Hierarchies. New York: Free Press.
Williamson, Oliver. 1979. "Transaction-Cost Economics: The Governance of Contractual
Relations." Journal of Law and Economics 22 (2): 233-61.
Williamson, Oliver. 1981. "The Economics of Organization: The Transaction Cost Approach."
American Journal of Sociology 87 (November): 548-77.
In The Protestant Ethic and the Spirit of Capitalism, Max Weber warned that the
Williamson, Oliver, and William Ouchi. 1981. "The Markets and Hierarchies and Visible
rationalist spirit ushered in by asceticism had achieved a momentum of its own and
Hand Perspectives. Pp. 347-70 in Perspectives on Organizational Design and Behavior,
that, under capitalism, the rationalist order had become an iron cage in which
edited by Andrew Van de Ven and William Joyce. New York: Wiley.
Wrong, Dennis. 1961. The Oversocialized Conception of Man in Modern Sociology." humanity was, save for the possibility of prophetic revival, imprisoned "perhaps
American Sociological Review 26 (2): 183-93. until the last ton of fossilized coal is burnt" (Weber, 1952: 181-2). In his essay on
bureaucracy, Weber returned to this theme, contending that bureaucracy, the
rational spirits organizational manifestation, was so efficient and powerful a means
of controlling men and women that, once established, the momentum of
bureaucratization was irreversible (Weber, 1968).
The imagery of the iron cage has haunted students of society as the tempo of
bureaucratization has quickened. But while bureaucracy has spread continuously in
the eighty years since Weber wrote, we suggest that the engine of organizational
rationalization has shifted. For Weber, bureaucratization resulted from three related
causes: competition among capitalist firms in the marketplace; competition among
states, increasing rulers' need to control their staff and citizenry; and bourgeois
demands for equal protection under the law. Of these three, the most important was
the competitive marketplace. "Today," Weber (1968: 974) wrote:

Paul J. DiMaggio and Walter W. Powell, "The Iron Cage Revisited," pp. 147-60 from Paul J. DiMaggio
and Walter W. Powell,"The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality
in Organizational Fields," American Sociological Review, 48: 2 (1983). Copyright © 1983 by American
Sociological Review. Reprinted by permission of the authors and the American Sociological Association.

Contemporary Sociological Theory, Third Edition. Edited by Craig Calhoun, Joseph Gerteis,
James Moody, Steven Pfaff, and Indermohan Virk. Editorial material and organization
© 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
 176 | Paul J. DiMaggio and Walter W. Powell
it is primarily the capitalist market economy which demands that the official business
of administration be discharged precisely, unambiguously, continuously, and with as
much speed as possible. Normally, the very large, modern capitalist enterprises are
themselves unequalled models of strict bureaucratic organization.
We argue that the causes of bureaucratization and rationalization have changed.
The bureaucratization of the corporation and the state have been achieved.
Organizations are still becoming more homogeneous, and bureaucracy remains the
common organizational form. Today, however, structural change in organizations
seems less and less driven by competition or by the need for efficiency. Instead, we
will contend, bureaucratization and other forms of organizational change occur as
the result of processes that make organizations more similar without necessarily
making them more efficient. Bureaucratization and other forms of homogenization
emerge, we argue, out of the structuration (Giddens, 1979) of organizational fields.
This process, in turn, is effected largely by the state and the professions, which have
become the great rationalizers of the second half of the twentieth century. For
reasons that we will explain, highly structured organizational fields provide a context
in which individual efforts to deal rationally with uncertainty and constraint often
lead, in the aggregate, to homogeneity in structure, culture, and output.
Organizational Theory and Organizational Diversity
Much of modern organizational theory posits a diverse and differentiated world of
organizations and seeks to explain variation among organizations in structure and
behavior (e.g., Woodward, 1965; Child and Kieser, 1981). Hannan and Freeman
begin a major theoretical paper (1977) with the question, "Why are there so many
kinds of organizations?" Even our investigatory technologies (for example, those
based on least-squares techniques) are geared towards explaining variation rather
than its absence.
We ask, instead, why there is such startling homogeneity of organizational forms
and practices; and we seek to explain homogeneity, not variation. In the initial stages
of their life cycle, organizational fields display considerable diversity in approach
and form. Once a field becomes well established, however, there is an inexorable
push towards homogenization.
Coser, Kadushin, and Powell (1982) describe the evolution of American college
textbook publishing from a period of initial diversity to the current hegemony of
only two models, the large bureaucratic generalist and the small specialist. Roth man
(1980) describes the winnowing^jf several competing models of legal education into
two dominant approaches. Starr (1980) provides evidence of mimicry in the
development of the hospital field; Tyack (1974) and Katz (1975) show a similar
process in public schools; Barnouw (1966-8) describes the development of dominant
forms in the radio industry; and DiMaggio (1982) depicts the emergence of dominant
organizational models for the provision of high culture in the late nineteenth century.
The Iron Cage Revisited 177
What we see in each of these cases is the emergence and structuration of an
organizational field as a result of the activities of a diverse set of organizations; and,
second, the homogenization of these organizations, and of new entrants as well,
once the field is established.
By organizational field, we mean those organizations that, in the aggregate,
constitute a recognized area of institutional life: key suppliers, resource and product
consumers, regulatory agencies, and other organizations that produce similar services
or products. The virtue of this unit of analysis is that it directs our attention not simply
to competing firms, as does the population approach of Hannan and Freeman (1977),
or to networks of organizations that actually interact, as does the interorganizational
network approach of Laumann et al. (1978), but to the totality of relevant actors.
In doing this, the field idea comprehends the importance of both connectedness (see
Laumann et al., 1978) and structural equivalence (White et al., 1976).'
The structure of an organizational field cannot be determined a priori but must
be defined on the basis of empirical investigation. Fields only exist to the extent
that they are institutionally defined. The process of institutional definition, or
"structuration," consists of four parts: an increase in the extent of interaction among
organizations in the field; the emergence of sharply defined interorganizational
structures of domination and patterns of coalition; an increase in the information
load with which organizations in a field must contend; and the development ot a
mutual awareness among participants in a set of organizations that they are involved
in a common enterprise (DiMaggio, 1982).
Once disparate organizations in the same line of business are structured into an
actual field (as we shall argue, by competition, the state, or the professions), powerful
forces emerge that lead them to become more similar to one another. Organizations
may change their goals or develop new practices, and new organizations enter the
field. But, in the long run, organizational actors making rational decisions construct
around themselves an environment that constrains their ability to change further in
later years. Early adopters of organizational innovations are commonly driven by a
desire to improve performance. But new practices can become, in Selznick's words
(1957: 17), "infused with value beyond the technical requirements of the task at
hand." As an innovation spreads, a threshold is reached beyond which adoption
provides legitimacy rather than improves performance (Meyer and Rowan, 1977).
Strategies that are rational for individual organizations may not be rational if adopted
by large numbers. Yet the very fact that they are normatively sanctioned increases the
likelihood of their adoption. Thus organizations may try to change constantly; but,
after a certain point in the structuration of an organizational field, the aggregate effect
of individual change is to lessen the extent of diversity within the field.2 Organizations
in a structured field, to paraphrase Schelling (1978:14), respond to an environment
that consists of other organizations responding to their environment, which consists
of organizations responding to an environment of organizations' responses.
~ Zucker ancfTolbert s (1981) work on the adoption of civil-service reform in the
United States illustrates this process. Early adoption of civil-service reforms was
related to internal governmental needs, and strongly predicted by such city
178 Paul J. DiMaggio and Walter W. Powell
characteristics as the size of immigrant population, political reform movements,
socioeconomic composition, and city size. Later adoption, however, is not predicted
by city characteristics, but is related to institutional definitions ot the legitimate
structural form for municipal administration. Marshall Meyer's (1981) study of the
bureaucratization of urban fiscal agencies has yielded similar findings: strong
relationships between city characteristics and organizational attributes at the turn of
the century, null relationships in recent years. Carroll and Delacroix's (1982) findings
on the birth and death rates of newspapers support the view that selection acts with
great force only in the early years of an industry's existence. Freeman (1982:14)
suggests that older, larger organizations reach a point where they can dominate their
environments rather than adjust to them.
The concept that best captures the process of homogenization is isomorphism. In
Hawley's (1968) description, isomorphism is a constraining process that forces one
unit in a population to resemble other units that face the same set of environmental
conditions. At the population level, such an approach suggests that organizational
characteristics are modified in the direction of increasing compatibility with
environmental characteristics; the number of organizations in a population is a
function of environmental carrying capacity; and the diversity of organizational
forms is isomorphic to environmental diversity. Hannan and Freeman (1977) have
significantly extended Hawley's ideas. They argue that isomorphism can result because
nonoptimal forms are selected out of a population of organizations or because
organizational decision makers learn appropriate responses and adjust their behavior
accordingly. Hannan and Freeman's focus is almost solely on the first process: selection.
Following Meyer (1979) and Fennell (1980), we maintain that there are two types
of isomorphism: competitive and institutional. Hannan and Freeman's classic paper
(1977), and much of their recent work, deals with competitive isomorphism, assuming
a system rationality that emphasizes market competition, niche change, and fitness
measures. Such a view, we suggest, is most relevant for those fields in which free and
open competition exists. It explains parts of the process of bureaucratization that
Weber observed, and may apply to early adoption of innovation, but it does not
present a fully adequate picture of the modern world of organizations. For this
purpose it must be supplemented by an institutional view of isomorphism of the sort
introduced by Kanter (1972:152-4) in her discussion of the forces pressing communes
toward accommodation with the outside world. As Aldrich (1979: 265) has argued,
"the major factors that organizations must take into account are other organizations."
Organizations compete not just for resources and customers, but for political power
and institutional legitimacy, for social as well as economic fitness. The concept of
institutional isomorphism is a useful tool for understanding the politics and ceremony
that pervade much modern organizational life.
Three mechanisms of institutional isomorphic change
We identify three mechanisms through which institutional isomorphic change
occurs, each with its own antecedents: (1) coercive isomorphism that stems from
The Iron Cage Revisited 179
political influence and the problem of legitimacy; (2) mimetic isomorphism resulting
from standard responses to uncertainty; and (3) normative isomorphism, associated
with professionalization. This typology is an analytic one: the types are not always
empirically distinct. For example, external actors may induce an organization to
conform to its peers by requiring it to perform a particular task and specifying the
profession responsible for its performance. Or mimetic change may reflect
environmentally constructed uncertainties.3 Yet, while the three types intermingle in
empirical setting, they tend to derive from different conditions and may lead to
different outcomes.
Coercive isomorphism
Coercive isomorphism results from both formal and informal pressures exerted on
organizations by other organizations upon which they are dependent and by cultural
expectations in the society within which organizations function. Such pressures may
be felt as force, as persuasion, or as invitations to join in collusion. In some
circumstances, organizational change is a direct response to government mandate:
manufacturers adopt new pollution control technologies to conform to environmental
regulations; nonprofits maintain accounts, and hire accountants, in order to meet
tax law requirements; and organizations employ affirmative-action officers to fend
off allegations of discrimination. Schools mainstream special students and hire
special education teachers, cultivate PTAs and administrators who get along with
them, and promulgate curricula that conform with state standards (Meyer et al.,
1981). The fact that these changes may be largely ceremonial does not mean that
they are inconsequential. As Ritti and Goldner (1979) have argued, staff become
involved in advocacy for their functions that can alter power relations within
organizations over the long run.
The existence of a common legal environment affects many aspects of an
organization's behavior and structure. Weber pointed out the profound impact of a
complex, rationalized system of contract law that requires the necessary
organizational controls to honor legal commitments. Other legal and technical
requirements of the state - the vicissitudes of the budget cycle, the ubiquity of
certain fiscal years, annual reports, and financial reporting requirements that ensure
eligibility for the receipt of federal contracts or funds - also shape organizations in
similar ways. Pfeffer and Salancik (1978:188-224) have discussed how organizations
faced with unmanageable interdependence seek to use the greater power of the
larger social system and its government to eliminate difficulties or provide for needs.
They observe that politically constructed environments have two characteristic
features: political decisionmakers often do not experience directly the consequences
of their actions; and political decisions are applied across the board to entire classes
of organizations, thus making such decisions less adaptive and less flexible.
Meyer and Rowan (1977) have argued persuasively that as rationalized states and
other large rational organizations expand their dominance over more arenas of
social life,organizationalstructures increasingly come to reflect rules institutionalized
 The Iron Cage Revisited 181
180 Paul J. DiMaggio and Walter W. Powell

problem with ambiguous causes or unclear solutions, problemistic search may yield
and legitimated by and within the state (also see Meyer and Hannan, 1979). As a
a viable solution with little expense (Cyert and March, 1963).
result, organizations are increasingly homogeneous within given domains and
Modeling, as we use the term, is a response to uncertainty. The modeled
increasingly organized around rituals of conformity to wider institutions. At the
organization may be unaware of the modeling or may have no desire to be copied, it
same time, organizations are decreasingly structurally determined by the constraints
merely serves as a convenient source of practices that the borrowing organization
posed by technical activities, and decreasingly held together by output controls.
may use. Models may be diffused unintentionally, indirectly through employee
Under such circumstances, organizations employ ritualized controls of credentials
transfer or turnover, or explicitly by organizations such as consulting firms or
and group solidarity.
industry trade associations. Even innovation can be accounted for by organizational
Direct imposition of standard operating procedures and legitimated rules and
modeling. As Alchian (1950) has observed:
structures also occurs outside the governmental arena. Michael Sedlak (1981) has
documented the ways that United Charities in the 1930s altered and homogenized
the structures, methods, and philosophies of the social service agencies that While there certainly are those who consciously innovate, there are those who, in their
imperfect attempts to imitate others, unconsciously innovate by unwittingly acquiring
depended upon them for support. As conglomerate corporations increase in size
some unexpected or unsought unique attributes which under the prevailing
and scope, standard performance criteria are not necessarily imposed on
circumstances prove partly responsible for the success. Others, in turn, will attempt to
subsidiaries, but it is common for subsidiaries to be subject to standardized
copy the uniqueness, and the innovation-imitation process continues.
reporting mechanisms (Coser et al., 1982). Subsidiaries must adopt accounting
practices, performance evaluations, and budgetary plans that are compatible with
One of the most dramatic instances of modeling was the effort of Japans
the policies of the parent corporation. A variety of service infrastructures, often
modernizers in the late nineteenth century to model new governmental initiatives
provided by monopolistic firms - for example, telecommunications and
on apparently successful western prototypes. Thus, the imperial government sent its
transportation - exert common pressures over the organizations that use them.
officers to study the courts, Army, and police in France, the Navy and postal system
Thus, the expansion of the central state, the centralization of capital, and the
in Great Britain, and banking and art education in the United States (see Westney,
coordination of philanthropy all support the homogenization of organizational
forthcoming). American corporations are now returning the compliment by
models through direct authority relationships.
implementing (their perceptions of) Japanese models to cope with thorny
We have so far referred only to the direct and explicit imposition of organizational
productivity and personnel problems in their own firms. I he rapid proliferation of
models on dependent organizations. Coercive isomorphism, however, may be more
quality circles and quality-of-work-life issues in American firms is, at least in part,
subtle and less explicit than these examples suggest. Milofsky (1981) has described
an attempt to model Japanese and European successes. These developments also
the ways in which neighborhood organizations in urban communities, many of
have a ritual aspect; companies adopt these"innovations" to enhance their legitimacy,
which are committed to participatory democracy, are driven to developing
to demonstrate they are at least trying to improve working conditions. More
organizational hierarchies in order to gain support from more hierarchically
generally, the wider the population of personnel employed by, or customers served
organized donor organizations. Similarly, Swidler (1979) describes the tensions
by, an organization, the stronger the pressure felt by the organization to provide the
created in the free schools she studied by the need to have a "principal" to negotiate
programs and services offered by other organizations. Thus, either a skilled labor
with the district superintendent and to represent the school to outside agencies. In
force or a broad customer base may encourage mimetic isomorphism.
general, the need to lodge responsibility and managerial authority at least
Much homogeneity in organizational structures stems from the fact that despite
ceremonially in a formally defined role in order to interact with hierarchical
considerable search for diversity there is relatively little variation to be selected from.
organizations is a constant obstacle to the maintenance of egalitarian or collectivist
New organizations are modeled upon old ones throughout the economy, and
organizational forms (Kanter, 1972; Rothschild-Whitt, 1979).
managers actively seek models upon which to build (Kimberly, 1980). 1 hus, in the
arts one can find textbooks on how to organize a community arts council or how to
Mimetic processes start a symphony women's guild. Large organizations choose from a relatively small
set of major consulting firms, which, like Johnny Appleseeds, spread a few
Not all institutional isomorphism, however, derives from coercive authority.
organizational models throughout the land. Such models are powerful because
Uncertainty is also a powerful force that encourages imitation. When organizational
structural changes are observable, whereas changes in policy and strategy are less
technologies are poorly understood (March and Olsen, 1976), when goals are
easily noticed. With the advice of a major consulting firm, a large metropolitan
ambiguous, or when the environment creates symbolic uncertainty, organizations
public television station switched from a functional design to a multidivisional
may model themselves on other organizations. The advantages of mimetic behavior
structure. The stations' executives were skeptical that the new structure was more
in the economy of human action are considerable; when an organization faces a

182 Paul J. DiMaggio and Walter W. Powell
efficient; in fact, some services were now duplicated across divisions. But they were
convinced that the new design would carry a powerful message to the for-profit
firms with whom the station regularly dealt. These firms, whether in the role of
corporate underwriters or as potential partners in joint ventures, would view the
reorganization as a sign that "the sleepy nonprofit station was becoming more
business-minded" (Powell and Freidkin, 1986). The history of management reform
in American government agencies, which are noted for their goal ambiguity, is
almost a textbook case of isomorphic modeling, from the PPPB of the McNamara
era to the zero-based budgeting of the Carter administration.
Organizations tend to model themselves after similar organizations in their field
that they perceive to be more legitimate or successful. The ubiquity ol certain kinds
of structural arrangements can more likely be credited to the universality of mimetic
processes than to any concrete evidence that the adopted models enhance efficiency.
Marshall Meyer (1981) contends that it is easy to predict the organization of a newly
emerging nation's administration without knowing anything about the nation itself,
since "peripheral nations are far more isomorphic - in administrative form and
economic pattern - than any theory of the world system of economic division ot
labor would lead one to expect."
Normative pressures
A third source of isomorphic organizational change is normative and stems primarily
from professionalization. Following Larson (1977) and Collins (1979), we interpret
professionalization as the collective struggle of members ot an occupation to
define the conditions and methods of their work, to control "the production ot
producers" (Larson, 1977:49-52), and to establish a cognitive base and legitimation
for their occupational autonomy. As Larson points out, the professional project is
rarely achieved with complete success. Professionals must compromise with
nonprofessional clients, bosses, or regulators. The major recent growth in the
professions has been among organizational professionals, particularly managers and
specialized staff of large organizations. The increased professionalization of workers
whose futures are inextricably bound up with the fortunes ot the organizations that
employ them has rendered obsolescent (if not obsolete) the dichotomy between
organizational commitment and professional allegiance that characterized
traditional professionals in earlier organizations (Hall, 1968). Professions are subject
to the same coercive and mimetic pressures as are organizations. Moreover, while
various kinds of professionals within an organization may differ from one another,
they exhibit much similarity to their professional counterparts in other organizations.
In addition, in many cases, professional power is as much assigned by the state as it
is created by the activities of the professions.
Two aspects of professionalization are important sources of isomorphism. One is
the resting of formal education and of legitimation in a cognitive base produced by
university specialists; the second is the growth and elaboration of professional
networks that span organizations and across which new models diffuse rapidly.
The Iron Cage Revisited 183
Universities and professional training institutions are important centers for the
development of organizational norms among professional managers and their staff.
Professional and trace associations are another vehicle for the definition and
promulgation of normative rules about organizational and professional behavior.
Such mechanisms create a pool of almost interchangeable individuals who occupy
similar positions across a range of organizations and possess a similarity of
orientation and disposition that may override variations in tradition and control
that might otherwise shape organizational behavior (Perrow, 1974).
One important mechanism for encouraging normative isomorphism is the
filtering of personnel. Within many organizational fields filtering occurs through
the hiring of individuals from firms within the same industry; through the
recruitment of fast-track staff from a narrow range of training institutions; through
common promotion practices, such as always hiring top executives from financial or
legal departments; and from skill-level requirements for particular jobs. Many
professional career tracks are so closely guarded, both at the entry level and
throughout the career progression, that individuals who make it to the top are
virtually indistinguishable. March and March (1977) found that individuals who
attained the position of school superintendent in Wisconsin were so alike in
background and orientation as to make further career advancement random and
unpredictable. Hirsch and Whisler (1982) find a similar absence of variation among
Fortune 500 board members. In addition, individuals in an organizational field
undergo anticipatory socialization to common expectations about their personal
behavior, appropriate style of dress, organizational vocabularies (Cicourel, 1970;
Williamson, 1975) and standard methods of speaking, joking, or addressing others
(Ouchi, 1980). Particularly in industries with a service or financial orientation
(Collins, 1979, argues that the importance of credentials is strongest in these areas),
the filtering of personnel approaches what Kanter (1977) refers to as the homosexual
reproduction of management." To the extent managers and key staff are drawn from
the same universities and filtered on a common set of attributes, they will tend to
view problems in a similar fashion, see the same policies, procedures and structures
as normatively sanctioned and legitimated, and approach decisions in much the
same way.
Entrants to professional career tracks who somehow escape the filtering process -
for example, Jewish naval officers, woman stockbrokers, or Black insurance
executives - are likely to be subjected to pervasive on-the-job socialization. To the
extent that organizations in a field differ and primary socialization occurs on the job,
socialization could reinforce, not erode, differences among organizations. But
when organizations in a field are similar and occupational socialization is carried
out in trade association workshops, in-service educational programs, consultant
arrangements, employer-professional school networks, and in the pages of trade
magazines, socialization acts as an isomorphic force.
The professionalization of management tends to proceed in tandem with the
structuration of organizational fields. The exchange of information among
professionals helps contribute to a commonly recognized hierarchy of status, of
 The Iron Cage Revisited 185
184 Paul J. DiMaggio and Walter W. Powell

entry and exit. Lee (1971: 51) maintains this is why hospital administrators are less
center and periphery, that becomes a matrix for information flows and personnel
concerned with the efficient use of resources and more concerned with status
movement across organizations. This status ordering occurs through both formal
competition and parity in prestige. Fennell (1980) notes that hospitals are a poor
and informal means. The designation of a few large firms in an industry as key
market system because patients lack the needed knowledge of potential exchange
bargaining agents in union - management negotiations may make these central
partners and prices. She argues that physicians and hospital administrators are the
firms pivotal in other respects as well. Government recognition of key firms or
actual consumers. Competition among hospitals is based on attracting physicians,
organizations through the grant or contract process may give these organizations
who, in turn, bring their patients to the hospital." Fennell (p. 505) concludes that:
legitimacy and visibility and lead competing firms to copy aspects of their structure
or operating procedures in hope of obtaining similar rewards. Professional and
Hospitals operate according to a norm of social legitimation that frequently conflicts
trade associations provide other arenas in which center organizations are recognized
with market considerations of efficiency and system rationality. Apparently, hospitals
and their personnel given positions of substantive or ceremonial influence. Managers
can increase their range of services not because there is an actual need lor a particular
in highly visible organizations may in turn have their stature reinforced by
service or facility within the patient population, but because they will be defined as fit
representation on the boards of other organizations, participation in industry-wide
only if they can offer everything other hospitals in the area offer.
or inter-industry councils, and consultation by agencies of government (Useem,
1979). In the nonprofit sector, where legal barriers to collusion do not exist, These results suggest a more general pattern. Organizational fields that include a
structuration may proceed even more rapidly. Thus executive producers or artistic large professionally trained labor force will be driven primarily by status competition.
directors of leading theatres head trade or professional association committees, sit Organizational prestige and resources are key elements in attracting professionals.
on government and foundation grant-award panels, or consult as government This process encourages homogenization as organizations seek to ensure that they
or foundation-financed management advisors to smaller theatres, or sit on smaller can provide the same benefits and services as their competitors.
organizations' boards, even as their stature is reinforced and enlarged by the grants
their theatres receive from government, corporate, and foundation funding sources
(DiMaggio, 1982).
Predictors of Isomorphic Change
Such central organizations serve as both active and passive models; their policies
and structures will be copied throughout their fields. Their centrality is reinforced as
It follows from our discussion of the mechanism by which isomorphic change
upwardly mobile managers and staff seek to secure positions in these central
occurs that we should be able to predict empirically which organizational fields will
organizations in order to further their own careers. Aspiring managers may undergo
be most homogeneous in structure, process, and behavior. While an empirical test ot
anticipatory socialization into the norms and mores of the organizations they hope
such predictions is beyond the scope of this paper, the ultimate value of our
to join. Career paths may also involve movement from entry positions in the center
perspective will lie in its predictive utility. The hypotheses discussed below are not
organizations to middle-management positions in peripheral organizations.
meant to exhaust the universe of predictors, but merely to suggest several hypotheses
Personnel flows within an organizational field are further encouraged by structural
that may be pursued using data on the characteristics of organizations in a field,
homogenization, for example the existence of common career titles and paths (such
either cross-sectionally or, preferably, over time. The hypotheses are implicitly
as assistant, associate, and full professor) with meanings that are commonly
governed by ceteris paribus assumptions, particularly with regard to size, technology,
understood.
and centralization of external resources.
It is important to note that each of the institutional isomorphic processes can be
expected to proceed in the absence of evidence that they increase internal
organizational efficiency. To the extent that organizational effectiveness is enhanced, A. Organizational-level predictors
the reason will often be that organizations are rewarded for being similar to other
There is variability in the extent to and rate at which organizations in a field change
organizations in their fields. This similarity can make it easier for organizations to
to become more like their peers. Some organizations respond to external pressures
transact with other organizations, to attract career-minded staff, to be acknowledged
quickly; others change only after a long period of resistance. The first two hypotheses
as legitimate and reputable, and to fit into administrative categories that define
derive from our discussion of coercive isomorphism and constraint.
eligibility for public and private grants and contracts. None of this, however, insures
Hypothesis A-l: The greater the dependence of an organization on another
that conformist organizations do what they do more efficiently than do their more
organization, the more similar it will become to that organization in structure, climate,
deviant peers.
and behavioral focus. Following Thompson (1967) and Pfeffer and Salancik (1978),
Pressures for competitive efficiency are also mitigated in many fields because the
this proposition recognizes the greater ability of organizations to resist the demands
number of organizations is limited and there are strong fiscal and legal barriers to
 186 Paul J. DiMaggio and Walter W. Powell
of organizations on whom they are not dependent. A position of dependence leads
to isomorphic change. Coercive pressures are built into exchange relationships. As
Williamson (1979) has shown, exchanges are characterized by transaction-specific
investments in both knowledge and equipment. Once an organization chooses a
specific supplier or distributor for particular parts or services, the supplier or
distributor develops expertise in the performance of the task as well as idiosyncratic
knowledge about the exchange relationship. The organization comes to rely on the
supplier or distributor and such transaction-specific investments give the supplier
or distributor considerable advantages in any subsequent competition with other
suppliers or distributors.
Hypothesis A-2: The greater the centralization of organization A's resource supply,
the greater the extent to which organization A will change isomorphically to resemble
the organizations on which it depends for resources. As Thompson (1967) notes,
organizations that depend on the same sources for funding, personnel, and
legitimacy will be more subject to the whims of resource suppliers than will
organizations that can play one source of support off against another. In cases where
alternative sources are either not readily available or require effort to locate, the
stronger party to the transaction can coerce the weaker party to adopt its practices
in order to accommodate the stronger party's needs (see Powell, 1983).
The third and fourth hypotheses derive from our discussion of mimetic
isomorphism, modeling, and uncertainty.
Hypothesis A-3: The more uncertain the relationship between means and ends the
greater the extent to which an organization will model itself after organizations it
perceives to be successful. The mimetic thought process involved in the search for
models is characteristic of change in organizations in which key technologies are
only poorly understood (March and Cohen, 1974). Here our prediction diverges
somewhat from Meyer and Rowan (1977) who argue, as we do, that organizations
which lack well-defined technologies will import institutionalized rules and
practices. Meyer and Rowan posit a loose coupling between legitimated external
practices and internal organizational behavior. From an ecologist's point of view,
loosely coupled organizations are more likely to vary internally. In contrast, we
expect substantive internal changes in tandem with more ceremonial practices, thus
greater homogeneity and less variation and change. Internal consistency of this sort
is an important means of interorganizational coordination. It also increases
organizational stability.
Hypothesis A-4: The more ambiguous the goals of an organization, the greater the
extent to which the organization will model itself after organizations that it perceives to
be successful. There are two reasons for this. First, organizations with ambiguous or
disputed goals are likely to be highly dependent upon appearances for legitimacy.
Such organizations may find it to their advantage to meet the expectations of
important constituencies about how they should be designed and run. In contrast to
our view, ecologists would argue that organizations that copy other organizations
usually have no competitive advantage. We contend that, in most situations, reliance
on established, legitimated procedures enhances organizational legitimacy and
The Iron Cage Revisited 187
survival characteristics. A second reason for modeling behavior is found in situations
where conflict over organizational goals is repressed in the interest of harmony; thus
participants find it easier to mimic other organizations than to make decisions on
the basis of systematic analyses of goals since such analyses would prove painful or
disruptive.
The fifth and sixth hypotheses are based on our discussion of normative processes
found in professional organizations.
Hypothesis A-5: The greater the reliance on academic credentials in choosing
managerial and staff personnel, the greater the extent to which an organization will
become like other organizations in its field. Applicants with academic credentials have
already undergone a socialization process in university programs, and are thus more
likely than others to have internalized reigning norms and dominant organizational
models.
Hypothesis A-6: The greater the participation of organizational managers in trade
and professional associations, the more likely the organization will be, or will become,
like other organizations in its field. This hypothesis is parallel to the institutional view
that the more elaborate the relational networks among organizations and their
members, the greater the collective organization of the environment (Meyer and
Rowan, 1977).
B. Field-level predictors
The following six hypotheses describe the expected effects of several characteristics
of organizational fields on the extent of isomorphism in a particular field. Since the
effect of institutional isomorphism is homogenization, the best indicator of
isomorphic change is a decrease in variation and diversity, which could be measured
by lower standard deviations of the values of selected indicators in a set of
organizations. The key indicators would vary with the nature of the field and the
interests of the investigator. In all cases, however, field-level measures are expected
to affect organizations in a field regardless of each organizations scores on related
organizational-level measures.
Hypothesis B-l: The greater the extent to which an organizational field is dependent
upon a single (or several similar) source(s) of support for vital resources, the higher the
level of isomorphism. The centralization of resources within a field both directly
causes homogenization by placing organizations under similar pressures from
resource suppliers, and interacts with uncertainty and goal ambiguity to increase
their impact. This hypothesis is congruent with the ecologists' argument that the
number of organizational forms is determined by the distribution of resources in
the environment and the terms on which resources are available.
Hypothesis B-2: The greater the extent to which the organizations in afield transact
with agencies of the state, the greater the extent of isomorphism in the field as a whole.
This follows not just from the previous hypothesis, but from two elements of state/
private-sector transactions: their rule-boundedness and formal rationality, and the
emphasis of government actors on institutional rules. Moreover, the federal
 188 Paul J. DiMaggio and Walter W. Powell
government routinely designates industry standards for an entire field which require
adoption by all competing firms. John Meyer (1979) argues convincingly that the
aspects of an organization which are affected by state transactions differ to the extent
that state participation is unitary or fragmented among several public agencies.
The third and fourth hypotheses follow from our discussion of isomorphic
change resulting from uncertainty and modeling.
Hypothesis B-3: The fewer the number of visible alternative organizational models
in a field, the faster the rate of isomorphism in that field. The predictions of this
hypothesis are less specific than those of others and require further refinement; but
our argument is that for any relevant dimension of organizational strategies or
structures in an organizational field there will be a threshold level, or a tipping point,
beyond which adoption of the dominant form will proceed with increasing speed
(Granovetter, 1978; Boorman and Levitt, 1979).
Hypothesis B-4: The greater the extent to which technologies are uncertain or
goals are ambiguous within a field, the greater the rate of isomorphic change.
Somewhat counterintuitively, abrupt increases in uncertainty and ambiguity
should, after brief periods of ideologically motivated experimentation, lead to
rapid isomorphic change. As in the case of A-4, ambiguity and uncertainty may be
a function of environmental definition, and, in any case, interact both with
centralization of resources (A-1, A-2, B-l, B-2) and with professionalization and
structuration (A-5, A-6, B-5, B-6). Moreover, in fields characterized by a high
degree of uncertainty, new entrants, which could serve as sources of innovation
and variation, will seek to overcome the liability of newness by imitating established
practices within the field.
The two final hypotheses in this section follow from our discussion of professional
filtering, socialization, and structuration.
Hypothesis B-5: The greater the extent of professionalization in a field, the greater
the amount of institutional isomorphic change. Professionalization may be measured
by the universality of credential requirements, the robustness of graduate training
programs, or the vitality of professional and trade associations.
Hypothesis B-6: The greater the extent of structuration of a field, the greater the
degree of isomorphics. Fields that have stable and broadly acknowledged centers,
peripheries, and status orders will be more homogeneous both because the diffusion
structure for new models and norms is more routine and because the level of
interaction among organizations in the field is higher. While structuration may not
lend itselt to easy measurement, it might be tapped crudely with the use of such
familiar measures as concentration ratios, reputational interview studies, or data on
network characteristics.
1 his rather schematic exposition of a dozen hypotheses relating the extent of
isomorphism to selected attributes of organizations and of organizational fields
does not constitute a complete agenda for empirical assessment of our perspective.
We have not discussed the expected nonlinearities and ceiling effects in the
relationships that we have posited. Nor have we addressed the issue of the indicators
that one must use to measure homogeneity. Organizations in a field may be highly
The Iron Cage Revisited 189
diverse on some dimensions, yet extremely homogeneous on others. While we
suspect, in general, that the rate at which the standard deviations of structural or
behavioral indicators approach zero will vary with the nature of an organizational
field's technology and environment, we will not develop these ideas here. The point
of this section is to suggest that the theoretical discussion is susceptible to empirical
test, and to lay out a few testable propositions that may guide future analyses.
NOTES
1 By connectedness we mean the existence of transactions tying organizations to one
another: A set of organizations that are strongly connected to one another and only
weakly connected to other organizations constitutes a clique. By structural equivalence we
refer to similarity of position in a network structure.
2 By organizational change, we refer to change in formal structure, organizational culture,
and goals, program, or mission.
3 This point was suggested by John Meyer.
REFERENCES
Alchian, Armen (1950) "Uncertainty, evolution, and economic theory." Journal of Political
Economy 58: 211-21.
Aldrich, Howard (1979) Organizations and Environments. Englewood Cliffs, NJ:
Prentice-Hall.
Barnouw, Erik (1966-8) A History of Broadcasting in the United States, 3 volumes. New York:
Oxford University Press.
Boorman, Scott A. and Paul R. Levitt (1979)"The cascade principle for general disequilibrium
dynamics." Cambridge/New Haven: Harvard-Yale Preprints in Mathematical Sociology.
Number 15.
Carroll, Glenn R. and Jacques Delacroix (1982) "Organizational mortality in the newspaper
industries of Argentina and Ireland: an ecological approach. Administrative Science
Quarterly 27: 169-98.
Child, John and Alfred Kieser (1981) "Development of organizations over time." Pp. 28-64 in
Paul C. Nystrom and William H. Starbuck (eds.), Handbook of Organizational Design.
New York: Oxford University Press.
Cicourel, Aaron (1970) "The acquisition of social structure: toward a developmental
sociology of language." pp. 136-68 in Jack D. Douglas (ed.), Understanding Everyday Life.
Chicago: Aldine.
Collins, Randall (1979) The Credential Society. New York: Academic Press.
Coser, Lewis, Charles Kadushin and Walter W. Powell (1982) Books: The Culture and
Commerce of Book Publishing. New York: Basic Books.
Cyert, Richard M. and James G. March (1963) A Behavioral Theory of the Firm. Englewood
Cliffs, NJ: Prentice-Hall.
DiMaggio, Paul (1982) "The structure of organizational fields: an analytical approach and
policy implications." Paper prepared for SUNY-Albany Conference on Organizational
Theory and Public Policy. April 1 and 2.
Fennell, Mary L. (1980) "The effects of environmental characteristics on the structure of
hospital clusters." Administrative Science Quarterly 25: 484-510.
 190 1 Paul J. DiMaggio and Walter W. Powell
Freeman, John H. (1982) "Organizational life cycles and natural selection processes."
pp. 1-32 in Barry Staw and Larry Cummings (eds.), Research in Organizational Behavior.
Vol. 4. Greenwich, CT: JAI Press.
Giddens, Anthony (1979) Central Problems in Social Theory: Action, Structure, and
Contradiction in Social Analysis. Berkeley, CA: University of California Press.
Granovetter, Mark (1978) "Threshold models of collective behavior." American Journal of
Sociology 83: 1420-43.
Hall, Richard (1968) "Professionalization and bureaucratization." American Sociological
Review 33: 92-104.
Hannan, Michael T. and John H. Freeman (1977) "The population ecology of organizations."
American Journal of Sociology 82: 929-64.
Hawley, Amos (1968) "Human ecology." pp. 328-37 in David L. Sills (ed.), International
Encyclopedia of the Social Sciences. New York: Macmillan.
Hirsch, Paul and Thomas Whisler (1982) "The view from the boardroom." Paper presented
at Academy of Management Meetings, New York, NY.
Kanter, Rosabeth Moss (1972) Commitment and Community. Cambridge, MA: Harvard
University Press.
Kanter, Rosabeth Moss (1977) Men and Women of the Corporation. New York: Basic Books.
Katz, Michael B. (1975) Class, Bureaucracy, and Schools: The Illusion of Educational Change in
America. New York: Praeger.
Kimberley, John (1980) "Initiation, Innovation, and Institutionalization in the Creation
Process." Pp. 18-43 in John Kimberley and Robert B. Miles (eds.), The Organizational Life
Cycle. San Francisco, CA: Jossey-Bass.
Larson, Magali Sarfatti (1977) The Rise of Professionalism: A Sociological Analysis. Berkeley, CA:
University of California Press.
Laumann, Edward O., Joseph Galaskiewicz and Peter Marsden (1978) "Community structure
as interorganizational linkage." Annual Review of Sociology 4: 455-84.
Lee, M. L. (1971) "A conspicuous production theory of hospital behavior." Southern Economic
Journal 38: 48-58.
March, James C. and James G. March (1977) "Almost random careers: the Wisconsin school
superintendency, 1940-72." Administrative Science Quarterly 22: 378-409.
March, James G. and Michael Cohen (1974) Leadership and Ambiguity: The American College
President. New York: McGraw-Hill.
March, James G. and Johan P. Olsen (1976) Ambiguity and Choice in Organizations. Bergen,
Norway: Universitetsforlaget.
Meyer, John W. (1979) "The impact of the centralization of educational funding and
control on state and local organizational governance." Stanford, CA: Institute for Research
on Educational Finance and Governance, Stanford University, Program Report
No. 79-B20.
Meyer, John W. and Michael Hannan (1979) National Development and the World System:
Educational, Economic, and Political Change. Chicago, IL: University of Chicago Press.
Meyer, John W. and Brian Rowan (1977) "Institutionalized organizations: formal structure as
myth and ceremony." American Journal of Sociology 83: 340-63.
Meyer, John W„ W. Richard Scott and Terence C. Deal (1981) "Institutional and technical
sources of organizational structure explaining the structure of educational organizations."
In Herman Stein (ed.), Organizations and the Human Services: Cross-Disciplinary
Reflections. Philadelphia, PA: Temple University Press.
The Iron Cage Revisited 191
Meyer, Marshall (1981) "Persistence and change in bureaucratic structures." Paper presented
at the annual meeting of the American Sociological Association, Toronto, Canada.
Milofsky, Carl (1981) "Structure and process in community self-help organizations."
New Haven: Yale Program on Non-Profit Organizations, Working Paper No. 17.
Ouchi, William G. (1980) "Markets, bureaucracies, and clans." Administrative Science
Quarterly 25: 129-41.
Perrow, Charles (1974) "Is business really changing?" Organizational Dynamics Summer:
31-44.
Pfeffer, Jeffrey and Gerald Salancik (1978) The External Control of Organizations: A Resource
Dependence Perspective. New York: Harper & Row.
Powell, Walter W. (1983) "New solutions to perennial problems of bookselling: whither the
local bookstore?" Daedalus: Winter.
Powell, Walter W. and Rebecca Freidkin (1986) "Politics and Programs: Organizational
Factors in Public Television Decision-making." pp. 245-69 in Nonprofit Enterprise in the
Arts, ed. P. J. DiMaggio. New York: Oxford University Press.
Ritti, R. R. and Fred H. Goldner (1979) "Professional pluralism in an industrial organization."
Management Science 16: 233-46.
Rothman, Mitchell (1980) "The evolution of forms of legal education." Unpublished
manuscript. Department of Sociology, Yale University, New Haven, CT.
Rothschild-Whitt, Joyce (1979) "The collectivist organization: an alternative to rational
bureaucratic models." American Sociological Review 44: 509-27.
Schelling, Thomas (1978) Micromotives and Macrobehavior. New York: W. W. Norton.
Sedlak, Michael W. (1981) "Youth policy and young women, 1950-1972: the impact
of private-sector programs for pregnant and wayward girls on public policy." Paper
presented at National Institute for Education Youth Policy Research Conference,
Washington, D.C.
Selznick, Philip (1957) Leadership in Administration. New York: Harper & Row.
Starr, Paul (1980) "Medical care and the boundaries of capitalist organization." Unpublished
manuscript. Program on Non-Profit Organizations, Yale University, New Haven, CT.
Swidler, Ann (1979) Organization Without Authority: Dilemmas of Social Control of Free
Schools. Cambridge, MA: Harvard University Press.
Thompson, James (1967) Organizations in Action. New York: McGraw-Hill.
Tyack, David (1974) The One Best System: A History of American Urban Education.Cambridge,
MA: Harvard University Press.
Useem, Michael (1979) "The social organization of the American business elite and
participation of corporation directors in the governance of American institutions."
American Sociological Review 44: 553-72.
Weber, Max (1952) The Protestant Ethic and the Spirit of Capitalism. New York: Scribner.
Weber, Max (1968) Economy and Society: An Outline of Interpretive Sociology. Three volumes.
New York: Bedminster.
Westney, D. Eleanor (Forthcoming) Organizational Development and Social Change in Meiji,
Japan.
White, Harrison C., Scott A. Boorman and Ronald L. Breiger (1976) "Social structure from
multiple networks. I. Blockmodels of roles and positions." American Journal of Sociology
81: 730-80.
Williamson, Oliver E. (1975) Markets and Hierarchies, Analysis and Antitrust Implications:
A Study of the Economics of Internal Organization. New York: Free Press.
192 Paul J. DiMaggio and Walter W. Powell

Williamson, Oliver E. (1979) "Transaction-cost economics: the governance of contractual

relations." Journal of Law and Economics 22:233-61.
Woodward, John (1965) Industrial Organization, Theory and Practice. London: Oxford
Chapter 12
University Press.
Zucker, Lynne G. and Pamela S. Tolbert (1981) "Institutional sources of change in the formal
structure of organizations: the diffusion of civil service reform, 1880-1935." Paper
presented at American Sociological Association annual meeting, Toronto, Canada.
Catnets [1966]
Harrison White

Net

A "net" can be drawn as a set of points with straight lines connecting various pairs
oT pointsTin our usage the points represent persons or social parties and each line
segment indicates a given kind of social relation between a pair of persons. If the
line segment has an arrowhead, it represents an asymmetric relation. In any case,
the relation must be familiar and clearly defined in the eyes of the people in the set
so that it makes sense to say (1) each pair of people either is or is not in that relation
and (2) the relation between one pair has about the same type of content and
significance as the relation between another pair. This is the obvious part of our
definition of a "net."
We sharpen this obvious definition of a net by adding further stipulations to make
it more useful as a concept in sociology. The basic idea is that a net is more than an
observer's recording of a set of relations between pairs of people, but yet is less than
a map known and attended to by any or all of the people in the net. We stipulate:

(i) the persons in the net accept the idea that they have meaningful indirect relations
with anyone paired with a person with whom they are paired (i.e., most of all
persons, when considered as the ego viewing the net, accept "composition" as a
meaningful "operator" on the net, to use mathematical language);
(ii) the relation diagramed implies sufficient familiarity with the other in a pair to
have some idea of whom else he is related to;

Harrison White, "Catnets," from Notes on the Constituents of Social Structure (unpublished
manuscript, 1966).

Contemporary Sociological Theory, Third Edition. Edited by Craig Calhoun, Joseph Gerteis,
James Moody, Steven Pfaff, and Indermohan Virk. Editorial material and organization
©2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley 8c Sons, Ltd.
 Catnets 195
194 Harrison White

suggestion confirmed with some regularity. Suppose one type of relation has positive
(iii) yet the many possible indirect connections among people, at various removes
connotations, as in friendship; then if one man is joined by it to each of two other
(through various numbers and patterns of intermediary persons) are not
men, they will tend to come into a friendly relation with each other too. Suppose a
recognized as falling into distinct new types of relations with their
second type of relation has negative connotations of enmity, etc. Then the reasoning
own definitions and contents, i.e., the indirect connections are not
is that a man's enmity with each of two other men will be a force tending to generate
"institutionalized." (If they were, one would have really a superposition of
a positive bond between them. More abstractly, in any triangle either none or two of
nets, one for each kind of relation possible between a pair of people and it is
the pair relations will tend to be negative, the other three or one of the relations
better to give this more complex type of structure a different name, as we
(lines) being of positive quality.
shall below.)
This idea of balanced relations, directly parallel to the psychological ideas in
consonance theory treated last term, can be extended to refer to larger clusters of
In other words, persons in a net recognize indirect connections implied by the set of
points and longer circuits of relations in the graph. One prediction of such extended
pair relations, but these connections are recognized only in part, and over a limited
theories is the formation within the population of two opposed factions such that
number of removes, and they are reacted to in concrete terms rather than as
all positive relations are within the factions and all negative relations between
well-defined new types of relations.
them. The idea can also be extended to take into account the direct nature of the
A second representation of nets helps in thinking about them and particularly in
relation when the two parties perceive it differently. Furthermore, third types of
seeing their connection with category systems. Set up a table with one row for each
relations, such as familiarity, can be introduced and propositions set forth about
person in the net and also one column. For visual clarity it helps to arrange the
their impact on allowable configurations of the positive and negative qualities.
persons in the rows in the same order as the columns. Then the existence ot an arrow
No satisfactory theory of these complex interrelations has yet been proposed, and
from one person to another is indicated by a check or a one in the cell formed b\
in particular no way of consistently introducing measures of the strengths of
the intersection of the row of the former with the column of the other. In this
the relations has been devised. But this discussion can serve as a warning that
representation it is natural to think of making the concept of net more flexible by
our neglect hereafter of the existence and interaction of nets of different qualities
attaching a measure of strength to each arrow, just as the arrow is the more natural
is dangerous.
idea here than the undirected line- which can only be shown by two entries of unity
in the table. The network can be regarded as formally similar to a cross-tabulation,
defined below, with each person being regarded as a category and the relations being
Limited systems
the items to be cross-tabulated.
A question used commonly in sociometric work, for example the Coleman et al.
study of drug diffusion, calls for the names of your three best friends or closest
Net Systems colleagues, etc. The implication is, and it seems to be confirmed in actual nets
derived from direct observation, that there is an upper limit on the number of links
A net continues indefinitely: its structure is essentially local, a matter of pair relations. a person has and should have in a net, and also a lower limit. When the population
Yet there must be a common culture to define a type of relations sharply and clearly, is large such constraints limit enormously the possible nets; in place of n(n-l)
if there is to be a net defined by the presence or absence of that relation between possible arrows in a population of n persons, there can be at most 3n. In abstract
pairs of persons. The implication of a common culture and the implication that a terms one can say that the set has a maximum "local degree."
net is defined over some set of persons both lead one to define the population over The population for the net system and the requirements of balance among
which the net is perceived. Population is essentially a categorical concept; so the relations are also of course limits on net systems.
concept of net is not in fact independent of the concept of the cat. Once a concrete
population is defined for the net the idea of other types of relation existing among
the population requiring representation by other nets cannot be avoided.
Cats
Balanced net systems: If several types of relation are defined on a population, the
obvious question is whether there are any limitations on the co-existence of different
Concretely, a "cat" is a bunch of people alike in some respect, from someone's point
relations on the same pair. The next question in order of complexity is whether in
oFviewT
any set of three people there are necessary connections between the relations ot two
Abstractly, a category is the respect in which persons may be grouped as similar -
pairs and the relation to be found between the third pair on the triangle.
a "box" into which people do or do not fall. These "persons" can of course be any
One very simple set of propositions about the interconnections of relations on a
population of things or social parties.
triangle has gained wide currency, although it is not so much a law as a plausible

"A single category" is therefore a meaningless idea: at a minimum there is at least
the residual category or box into which fall "all others," those not alike in the given
respect. To say category is to mean a system of categories.
The different aspects of the cat are emphasized by two different ways of
formulating the cat in network terms, where we now use "network" in its formal
sense.
Formally, a network is a collection of points in which some pairs are connected by
lines.
1. Let one point be set apart from the others as the center and identified as the
attribute or "respect" which the people, the other points, can have. Then those points
connected to the center by lines form the cat. (The center need not be in a particular
position since the confluence of lines on it is sufficient identification.)
Additional possible connotations of category are suggested by variations on this
representation: Let the lines be directed by arrowheads. Arrows leading out from the
center connote an attribute "received" by the points, say sex or citizenship. If all
arrows run toward the center an "achieved" attribute, one "projected" by the persons,
is implied, say political party or good health.
2. Let the subset of points which have the attribute be fully connected by lines:
each line then represents the sharing of the attribute.
An implicit assumption has slipped in with representation (2) above that "sharing
an attribute" is between pairs of people. Why not triplets? Then three persons in a
category could be represented by where the dotted lines from each person to the
"relation" between the other pair in the triad indicates that all three persons are
bound up in each relation (see Figure 12.1). And so on through the quartet, the
quintet, etc. as the fundamental unit of similarity. The point is that drawing a
complete network of lines between pairs really indicates only one special, limited
possible meaning of jointly belonging to a category. And the advantage of having
tried to represent a cat in this network fashion is that it brings out complexities in
the possible meanings of category.
Representation (1) can be seen as the limiting case of representation (2) in which
not pairs, triads, etc. but the total group of persons in the cat is the smallest unit in
Catnets 197
which it is meaningful to speak of the sharing of the attribute which defines the
category. Call this the total integration. Representation (1) can also be seen as the
other extreme case of representation (2), wherein each person has the attribute as an
individual with no connotation of sharing anything with others in the cat. Call this
the isolate representation.
As in the definition of net so also here we are forced to introduce psychological
and cultural elements in our very definitions of the elementary terms of social
structure. The total representation of a cat is appropriate where there is an intense
consciousness of joint categorical membership and/or clear recognition in the
culture of common possession of the attribute. The isolate representation is most
appropriate where the category is defined by some observer in a way not visible to
the persons involved even though the category may in fact sharply describe their
behavior. The distinction is an old and familiar one, of course, e.g., is "the proletariat"
a self-conscious class of a construct of the observer?
Category Systems
Ad hoc
"Have attribute" versus "don't have it" is the simplest ad hoc "system" of categories.
Any arbitrary set of categories such that any person in the population belongs to
exactly one of them is an ad hoc system.
Generic
Mail addresses form a typical generic system of categories: a person is located within
successively finer subdivisions of an initial large category. The first level of largest
categories can be regarded as an ad hoc system, as can each level of subcategories
within a next-larger category.
Cross-tabulation
A person is cross-tabulated by being placed simultaneously within two or more ad
hoc category systems, each of which covers the entire population.
If he is placed first in his proper category within one system, and then in his proper
category in the second system, and so on, the result looks like a generic category
system. But here any one of the ad hoc systems can be applied first as the level of
largest categories, and any other be applied next to refine the largest categories, and
so on. Whereas in the generic system proper only one set of categories can be used as
the first or highest level, and the breakdown into finer categories may be made in
different ways within each of the biggest categories. Thus, location of an adult by the
industry, company, division and title of his work affiliation is a generic system, whereas
his specification by sex, age, religion, location, and income is a cross-tabulation.
 Catnets 199
198 Harrison White

Table 12.1 Tabular representation of Figure 12.2.

Male

/
8 6 2 5 <— Youth

7 3 1 4

\ Wealth

Figure 12.2 Venn diagram showing membership in three classes.

the four classes. You can see from this brief discussion that assumptions bootlegged
Two different representations of the "cells" or boxes in the cross-tabulation bring
in from our knowledge of two-dimensional spaces can bias our assumptions about
out different possible connotations. (The representations used for cats can also be
such an elementary matter as the overlap of logical classes.
used here, but are too clumsy to give much insight.)
The first question one asks of a cross-tabulation is how much dependence it
reveals between the category one belongs to in one ad hoc system and the category
1 The common procedure in social surveys is a "lay out" of the boxes as the
one belongs to in the other. Suppose there is no dependence at all: the two ad hoc
intersections of rows for one ad hoc system with columns for the other. (Where
systems are not only logically but also factually completely unrelated. Suppose 100
three ad hoc systems are cross-tabulated one needs a cube, or on the printed page
men of a tabulated population of 1,000 men belong to a given category in one ad
a series of such two-dimensional lay-outs.) The order of the rows, and the order
hoc system, and 50 men belong to a certain category in the other ad hoc system.
of columns, and which system is laid out by rows and which by columns is
How many men would one expect to find who belonged to both categories? Clearly,
immaterial.
1/10 of the 50 men: but, equally clearly, 1/20 of the 100 men. Fortunately, there is no
2 A Venn diagram is the standard way to represent the overlapping memberships
disagreement - one would expect to find five men belonging to both categories.
in different logical classes, which is just the special case of cross-tabulation with
two classes (in vs. out) in each ad hoc system. The typical diagram for membership
in three classes (say male sex, youth, and wealth) is shown in Figure 12.2. Contextual
The eight numbered areas can equated to the eight cells in the cube for
Placement in a cat in this system is meaningful only within the context of the whole
representation 1.
structure formed by the categories. A hierarchical system of social classes is an
example. One is not upper class because of some intrinsic attribute but in contrast
The distinctive thing in this representation is that one of the two categories of each
to being lower class. The actual criteria of upper class membership can change, and
ad hoc system (e.g., male in male vs. female above) is singled out by being shown as
even become inverted in a given society over time, so that membership in the
the inside of a circle. Thus the eight areas above are shown in distinctively different
category is a matter of the context. In this simple case of two social classes one could
shapes, all centered around area 1 as the core of the classification system. Thus the
just say membership is a relative matter, but the word context better conveys the
Venn diagram is a relativistic or subjective way to present a cross-tabulation.
complexity of assignment in more complex systems of categories which form
Note that representation 2 for cross-tabulations, using overlapping circles, breaks
structures. One example of a more complex system would be the schools and cliques
down if more than three classes are to be used. With four classes one needs 16
in which artists are viewed as falling.
intersections (2*2*2*2 = 16), 16 different combinations of the inside and outside
Both the generic and cross-category systems can be regarded as contextual in a
areas of the four circles, and these cannot be found using circles on a plane. The
primitive way.
contrast between representations 1 and 2 is made clearer by using rectangles in
Neighbordom is an aspect of each of these types of category systems: which
place of circles in the Venn diagram. Thus, the three-circle diagram in Figure 12.2
categories are most akin? The answer is built into the definition of generic systems.
can be shown as a table (see Table 12.1) and a similar diagram for four overlapping
It is usually clear in contextual systems.
rectangles (plus a corner as above for the residual category) can handle the core of
 200 Harrison White Catnets 201

Cities may be categorized in an ad hoc system by the size of their populations. In Personchosen
this and similar cases a measure of neighbordom is implied by the recognition in that A B C D
culture of a structure of similarity among the attributes defining the categories. There Number of
a b c d e f g h i j k 1 m
is no necessary implication of a closeness perceived by the persons in the categories. choices made
a_

b
Initial cat system A
c
_d
Only cats which are recognized as such by their occupants are considered here.
The fundamental difficulty of a cat system is recruitment, the agreed definition of Person ® c; p.
who 9
new members. There is no hidden hand available to drop a bucket of blue paint over
chooses ~ h
a new person and thus define unambiguously his possession of the attribute blue C i
and his membership in the blue cat. Instead new ties, representing the relation _ j
between sponsor and protege, must continually be spun between people in the cats
in a system, whether ad hoc or more complicated. Over time the ties between
D
~
_
k
1 %
m
particular cats tend to become a tradition and one ends up with a network defined Number of Total number
over the set of categories as nodes. choices received °f choices
Wilensky has a concept of career systems as the backbone of our social structure.
Figure 12.3 Mapping a neighborhood system.
He describes career systems in terms of stages of advancement and thus presupposes
as foundation a contextual system of categories. On the other hand, the idea of the closeness of points on a net. Neighborhoods in one city may be best represented
regularized paths of advancement suggests the spinning out of network ties among on a sphere with highest and lowest social status neighborhoods at the poles, and the
people in successive stages.
broadest spectrum of distances between neighborhoods at the same class level being
In a large society a catnet system tends to become perceived in a broader view as at the equator area representing the middle-class neighborhoods. In another city
itself a category, one of an ad hoc system that grows links and evolves into a catnet factors such as ethnicity, class and political views, which help set and summarize the
system on a bigger scale. Lumpability is a simple term to use to describe the likelihood distances among neighborhoods, may interact in such a complicated pattern that a
of a catnet system being lumped together as a single category in a broader system. As doughnut (toroid) shape might be required for an accurate plot of neighborhoods.
an example, we all repeatedly have the experience of leaving one life stage and The second representation of a net, as a cross-tabulation of each relation by the
moving on to another; as we do our previous highly refined and detailed perception person making and the person receiving it, provides a basis for mapping a
of the internal social structure for the stage we have left is quickly drilled into a neighborhood system (see Figure 12.3). Let the rows for people thought to belong to
lumped representation as a category or stage. each neighborhood be grouped together. Then if the columns are arranged in the
same order a heavy clustering of entries in various-sized blocks along the diagonal
Neighborhood systems should be apparent: Here small letters label individuals, capitals label the
neighborhoods found, and the hatched areas indicate where an unusually high
Hie idea of neighborhood has several connotations, all consistent with and sharpening density of relations is found (in this case so few persons are concerned that it is
the genera] idea of a catnet system evolved from an initial net. One neighborhood hardly necessary to talk of a density of entries). In actually carrying out such a study
supplies others linked to it through some overlap in members and numerous relations a number of difficulties arise:
between its members, and the others in turn have their neighboring neighborhoods.
In the most common usage neighborhoods are residential areas spread out on the (i) Finding the best ordering of the individual rows to bring out the neighborhood
ground. Interconnections are implied by juxtaposition on a two-dimensional map. clustering could take an enormous number of trials (note: 10!=10*9*8*7*6*
Yet we know from our own lives that the density of social relations between 5*4*3*2*1 is about 3 million, the number of ways just 10 people can be
neighborhoods often has little to do with physical closeness. ordered).
The question is, what kind of space can best serve to map the relative closeness of (ii) Is it absolute or relative density of choice that should be the criterion? Suppose
a given system of neighborhoods? It is a very general question that we could also the individuals in a possible neighborhood make only one of the two "choices"
have asked about networks, except that the yes-or-no character of the relation apiece, but practically all of these choices are given to others in the small
between a pair of individuals hardly provides the basis for an elaborate discussion of fraction of the total population. Is it a neighborhood?
 202 Harrison White
The boundaries between neighborhoods will not be sharp of course: this is a catnet
system with ties connecting different cats. The interconnections will appear as
entries scattered over the table above away from the main diagonal. In a catnet
system these entries too will tend to cluster in a few rectangular areas; the assumption
is that persons in a cat tend to evolve relations with the same sets of others. By
rearranging the groups of rows and columns corresponding to the neighborhoods
found it is possible to make apparent the pattern of closeness among the
neighborhoods, so to speak the neighborhood structure of the set of neighborhoods.
If in the figure above the areas of high density are clustered adjacent to the blocks
shown along the main diagonal, one can see the neighborhoods are strung out as if
along a railroad track in a Western town. If in addition the extreme lower left and
upper right corners are also areas of high density the neighborhoods should be
arranged in a ring to represent their social closeness.
Frame
People develop culture in part to meet their needs to visualize, operate in and modify
the social structure to which they belong. Some nets persist for a very long time. The
pair relation on which the net is based remains stable and clearly defined. New
persons are added to and leave the net, but according to clear-cut rules. In such a net
it is natural for the simplest kinds of indirect relations to be "institutionalized," that
is, recognized in that culture as a distinctive new kind of relation. Indeed, the rules
of admission to the net regulate an indirect relation.
If the basic pair relation is any friendship, the man who is a friend of my friend
may come to be defined as being in a definite relation to me, say acquaintance.
Problems arise because of the enormous variety of types of indirect relations. For
example, I also know of men who are friends with not just one but several friends of
mine, who in turn may or may not be friends with one another. The natural step is
to categorize many varieties together as a single new type of indirect relation: thus
all the above types of friends at one remove can be lumped as acquaintances.
The end state of such an evolution of the cultural definition of a net is what we
mean by a frame. The characteristics of a frame are:
(i) Indirect ties reckoned through at least one and possibly indefinitely many
removes are grouped by a regular scheme into a smaller number of types.
(ii) Each type of indirect relation is recognized as a distinct and well-defined
relation in its own right, with its own emotional quality and set of rights and
duties. It is "institutionalized."
(iii) The types of indirect relations form a cumulative structure: just as the "sum"
or composition of two direct relations may define a simplest type of indirect
relation, so more complex types of indirect relations are regarded as the
composition of simpler types.
(iv) The structure of indirect relations is a theory held in that culture. That is, it
exists independent of any concrete net of relations existing in the society. In
Catnets 203
particular, the structure is keyed to a central "person" which is really an abstract
role; we call it ego. The structure is the same no matter who is assigned the role
ego in a particular realization of it; a given person as ego will recognize every
type of indirect relation even though no actual person in the existing population
of the net is in that role to him.
(v) Since it is a theory, a frame tends to be used as an ideal type. Even in the abstract
the frame may not be as fully developed and rationalized in a given culture as
is implied by the preceding description. Certainly, many of the people who use
it may have only a confused understanding of it. But even if it is fully developed,
and fully understood by all users, the complexities and irregularities in the
concrete net of relations may make it difficult to apply in a fully coherent way.
These "types" of indirect relations are categories, of course, and the frame can be
seen as an extreme case of a contextual system of categories. But the frame is a
floating system of categories which can be anchored on any person and applied from
his point of view; in that sense it is subjective. Moreover, one type of indirect relation
can be seen as the sum or composition of other types; in this sense the frame has the
quality of a network as much as a category system.
Balanced net systems are also akin to frames. The latter can carry the process of
cumulating indirect relations much further since they deal with a potential network
of relations from the point of view of only one node in it at a time. For the same
reason the frame does not deal with possible inconsistencies in the different types of
relations which may coexist among a set of people; the frame is a statement about the
structure of abstract types of relations, not about the combinations of different
relations with ego or each other that are allowable for actual persons in the population.
A frame can be built using more than one basic type of relation. Indeed it is
plausible to assert that a frame is more likely to develop when at least two distinct
basic relations are present. The need for an ordering and simplification of indirect
relations is more pressing then since the possible types that can result from
combinations of two basic relations are much more numerous and intricate. And a
culture rich enough to define two basic relations in an aspect of social life is more
likely to be able to generate a frame.
Another way to search out patterns of neighbordom in an ad hoc system is
through considering transitions of persons between the categories. Consider a cross-
tabulation of the population in an ad hoc system by which category they are in at
one time and which category they are in at a later time. There may be many more
transitions between some than other pairs of categories; the relative frequency of
transitions, with allowance made for the absolute sizes of categories, is a natural
measure of neighbordom.
Catnets
The catnet evolves from a simpler system, either net or cat.

Chapter 13
Structural Holes [1992]
Ronald S. Burt
The Social Structure of Competition
Opportunity and Capital
A player brings at least three kinds of capital to the competitive arena. Other
distinctions can be made, but three are sufficient here. First, the player has financial
capital: cash in hand, reserves in the bank, investments coming due, lines of credit.
Second, the player has human capital. Your natural qualities - charm, health,
intelligence, and looks - combined with the skills you have acquired in formal
education and job experience give you abilities to excel at certain tasks.
Third, the player has social capital: relationships with other players. You have
friends, colleagues, and more general contacts through whom you receive opportunities
to use your financial and human capital. I refer to opportunities in a broad sense, but
I certainly mean to include the obvious examples of job promotions, participation in
significant projects, influential access to important decisions, and so on. The social
capital of people aggregates into the social capital of organizations. In a firm providing
Ronald S. Burt. "Structural Holes," pp. 8-10, 15-23, 30-4, 38-40, 44-5, 46 from Ronald S. Burt,
Structural Holes: The Social Structure of Competition.Cambridge, MA: Harvard University Press, 1992.
Copyright © 1992 by the President and Fellows of Harvard College. Reprinted with permission
of Harvard University Press.
Contemporary Sociological Theory, Third Edition. Edited by Craig Calhoun, Joseph Gerteis,
James Moody, Steven Pfaff, and Indermohan Virk. Editorial material and organization
©2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley 8c Sons, Ltd.
Structural Holes 205
services - for example, advertising, brokerage, or consulting - there are people valued
for their ability to deliver a quality product. Then there are "rainmakers," valued for
their ability to deliver clients. Those who deliver the product do the work, and the
rainmakers make it possible for all to profit from the work. The former represent
"the financial and human capital of the firm. The latter represent its social capital.
More generally, property and human assets define the firm's production capabilities.
Relations within and bevond the firm are social capital.
Distinguishing social capital
Financial and human capital are distinct in two ways from social capital. First, they are
the property of individuals. They are owned in whole or in part by a single individual
defined in law as capable of ownership, typically a person or corporation. Second,
they concern the investment term in the market production equation. Whether held
by a person or the fictive person of a firm, financial and human capital gets invested
to create production capabilities. Investments in supplies, facilities, and people serve
to build and operate a factory. Investments of money, time, and energy produce a
skilled manager. Financial capital is needed for raw materials and production facilities.
Human capital is needed to craft the raw materials into a competitive product.
Social capital is different on both counts. First, it is a thing owned jointly by the
parties to a relationship. No one player has exclusive ownership rights to social
capital. Tf you or your partner in a relationship withdraws, the connection, with
whatever social capital it contained, dissolves. If a firm treats a cluster of customers
poorly and they leave, the social capital represented by the firm-cluster relationship
is lost. Second, social capital concerns rate of return in the market production
equation. Through relations with colleagues, friends, and clients come the
opportunities to transform financial and human capital into profit.
Social capital is the final arbiter of competitive success. The capital invested to
bring your organization to the point of producing a superb product is as rewarding
as the opportunities to sell the product at a profit. The investment to make you a
skilled manager is as valuable as the opportunities - the leadership positions - you
get to apply your managerial skills. The investment to make you a skilled scientist
with state-of-the-art research facilities is as valuable as the opportunities - the
projects - you get to apply those skills and facilities.
More accurately, social capital is as important as competition is imperfect and
investment capital is abundant. Under perfect competition, social capital is a
constant in the production equation. There is a single rate of return because capital
moves freely from low-yield to high-yield investments until rates of return are
homogeneous across alternative investments. When competition is imperfect,
capital is less mobile and plays a more complex role in the production equation.
There are financial, social, and legal impediments to moving cash between
investments. There are impediments to reallocating human capital, both in terms of
changing the people to whom you have a commitment and in terms of replacing
them with new people. Rate of return depends on the relations in which capital is
 206 RonaldS. Burt
invested. Social capital is a critical variable. This is all the more true when financial
and human capital are abundant - which in essence reduces the investment term in
the production equation to an unproblematic constant.
These conditions are generic to the competitive arena, which makes social capital
a factor as routinely critical as financial and human capital. Competition is never
perfect. The rules of trade are ambiguous in the aggregate and everywhere negotiable
in the particular. The allocation of opportunities is rarely made with respect to a
single dimension of abilities needed for a task. Within an acceptable range of needed
abilities, there are many people with financial and human capital comparable to your
own. Whatever you bring to a production task, there are other people who could do
the same job - perhaps not as well in every detail, but probably as well within the
tolerances of the people for whom the job is done. Criteria other than financial and
human capital are used to narrow the pool down to the individual who gets the
opportunity. Those other criteria are social capital. New life is given to the proverb
that says success is determined less by what you know than by whom you know. As a
senior colleague once remarked (and Cole, 1992: chaps. 7-8, makes into an intriguing
research program), "Publishing high-quality work is important for getting university
resources, but friends are essential." Of those who are equally qualified, only a select
few get the most rewarding opportunities. Of the products that are of comparably
high quality, only some come to dominate their markets. The question is how.
Benefit-rich networks
A player with a network rich in information benefits has contacts: (a) established in
the places where useful bits of information are likely to air, and (b) providing a
reliable flow of information to and from those places.
Selecting contacts
The second criterion is as ambiguous as it is critical. It is a matter of trust, of
confidence in the information passed and the care with which contacts look out for
your interests. Trust is critical precisely because competition is imperfect. The
question is not whether to trust, but whom to trust. In a perfectly competitive
arena, you can trust the system to provide a fair return on your investments. In the
imperfectly competitive arena, you have only your personal contacts. The matter
comes down to a question of interpersonal debt. If I do for her, will she for me?
There is no general answer. The answer lies in the match between specific people.
If a contact feels that he is somehow better than you - a sexist male dealing with a
woman, a racist white dealing with a black, an old-money matron dealing with an
upwardly mobile ethnic - your investment in the relationship will be taken as
proper obeisance to a superior. No debt is incurred. We use whatever cues can be
found for a continuing evaluation of the trust in a relation, but we never know a
debt is recognized until the trusted person helps us when we need it. With this kind
Structural Holes 207
Network A Network B Network C
Figure 13.1 Network expansion.
of uncertainty, players are cautious about extending themselves for people whose
reputation for honoring interpersonal debt is unknown. (...)
Siting contacts
(...) Everything else constant, a large, diverse network is the best guarantee of having a
contact present where useful information is aired. This is not to say that benefits must
increase linearly with size and diversity, a point to which I will return (Figure 13.5), but
only that, other things held constant, the information benefits of a large, diverse
network are more than the information benefits of a small, homogeneous network.
Size is the more familiar criterion. Bigger is better. Acting on this understanding,
people can expand their networks by adding more and more contacts. They make more
cold calls, affiliate with more clubs, attend more social functions. Numerous books and
self-help groups can assist them in "networking" their way to success by putting them
in contact with a large number of potentially useful, or helpful, or like-minded people.
The process is illustrated by the networks in Figure 13.1. The four-contact network at
the left expands to sixteen contacts at the right. Relations are developed with a friend of
each contact in network A, doubling the contacts to eight in network B. Snowballing
through friends of friends, there are sixteen contacts in network C, and so on.
Size is a mixed blessing. More contacts can mean more exposure to valuable
information, more likely early exposure, and more referrals. But increasing network
size without considering diversity can cripple a network in significant ways. What
matters is the number of nonredundant contacts. Contacts are redundant to the
extent that they lead to the same people, and so provide the same information benefits.
Consider two four-contact networks, one sparse, the other dense. There are no
relations between the contacts in the sparse network, and strong relations between
every contact in the dense network. Both networks cost whatever time and energy
is required to maintain four relationships. The sparse network provides four
nonredundant contacts, one for each relationship. No single one of the contacts gets
the player to the same people reached by the other contacts. In the dense network,
each relationship puts the player in contact with the same people reached through
the other relationships. The dense network contains only one nonredundant contact.
Any three are redundant with the fourth.
 208 Ronald S. Burt
I he sparse network provides more information benefits. It reaches information
in four separate areas of social activity. The dense network is a virtually worthless
monitoring device. Because the relations between people in that network are strong,
each person knows what the other people know and all will discover the same
opportunities at the same time.
The issue is opportunity costs. At minimum, the dense network is inefficient in the
sense that it returns less diverse information for the same cost as that of the sparse
network. A solution is to put more time and energy into adding nonredundant contacts
to the dense network. But time and energy are limited, which means that inefficiency
translates into opportunity costs. If I take four relationships as an illustrative limit on
the number of strong relations that a player can maintain, the player in the dense
network is cut off from three fourths of the information provided by the sparse network.
Structural Holes
I use the term structural hole for the separation between nonredundant contacts.
Nonredundant contacts are connected by a structural hole. A structural hole is a
relationship of nonredundancy between two contacts. The hole is a buffer, like an
insulator in an electric circuit. As a result of the hole between them, the two contacts
provide network benefits that are in some degree additive rather than overlapping.
Empirical indicators
Nonredundant contacts are disconnected in some way — either directly, in the sense
that they have no direct contact with one another, or indirectly, in the sense that one
has contacts that exclude the others. The respective empirical conditions that
indicate a structural hole are cohesion and structural equivalence. Both conditions
define holes by indicating where they are absent.
Under the cohesion criterion, two contacts are redundant to the extent that they
are connected by a strong relationship. A strong relationship indicates the absence of
a structural hole. Examples are father and son, brother and sister, husband and wife,
close friends, people who have been partners for a long time, people who frequently
get together for social occasions, and so on. You have easy access to both people if
either is a contact. Redundancy by cohesion is illustrated at the top of Figure 13.2.
The three contacts are connected to one another, and so provide the same network
benefits. The presumption here - routine in network analysis since Festinger,
Schachter, and Back's (1950) analysis of information flowing through personal
relations and Homans's (1950) theory of social groups - is that the likelihood that
information will move from one person to another is proportional to the strength of
their relationship. Empirically, strength has two independent dimensions: frequent
contact and emotional closeness (see Marsden and Hurlbert, 1988; Burt, 1990).
Structural equivalence is a useful second indicator for detecting structural holes.
Two people are structurally equivalent to the extent that they have the same contacts.
Structural Holes 209
Redundancy Y0U
by Cohesion
Redundancy
by Structural YOU
Equivalence
o
Figure 13.2 Structural indicators of redundancy.
Regardless of the relation between structurally equivalent people, they lead to the
same sources of information and so are redundant. Cohesion concerns direct
connection; structural equivalence concerns indirect connection by mutual contact.
Redundancy by structural equivalence is illustrated at the bottom of Figure 13.2.
The three contacts have no direct ties with one another. They are nonredundant by
cohesion. But each leads you to the same cluster of more distant players. The
information that comes to them, and the people to whom they send information,
are redundant. Both networks in Figure 13.2 provide one nonredundant contact at
a cost of maintaining three.
The indicators are neither absolute nor independent. Relations deemed strong are
only strong relative to others. They are our strongest relations. Structural equivalence
rarely reaches the extreme of complete equivalence. People are more or less
structurally equivalent. In addition, the criteria are correlated. People who spend a
lot of time with the same other people often get to know one another. The mutual
contacts responsible for structural equivalence set a stage for the direct connection
of cohesion. The empirical conditions between two players will be a messy
combination of cohesion and structural equivalence, present to varying degrees, at
varying levels of correlation.
Cohesion is the more certain indicator. If two people are connected with the
same people in a player's network (making them redundant by structural
^equivalence), they can still be connected with different people beyond the network
(making them nonredundant). But if they meet frequently and feel close to one
another, then they are likely to communicate and probably have contacts in
common. More generally, and especially for field work informed by attention to
network benefits, the general guide is the definition of a structural hole. There is a
structural hole between two people who provide nonredundant network benefits.
If the cohesion and structural equivalence conditions are considered together,
redundancy is most likely between structurally equivalent people connected by a
strong relationship. Redundancy is unlikely, indicating a structural hole, between
 210 RonaldS. Burt
^ N,
YOU YOU
% little loss in information benefits from the cluster and a gain in the reduced effort
Network A' Network B' Network C'
Figure 13.3 Strategic network expansion.
total strangers in distant groups. I will return to this issue again, to discuss the
depth of a hole, after control benefits have been introduced.
The efficient-effective network
Balancing network size and diversity is a question of optimizing structural holes. The
number of structural holes can be expected to increase with network size, but the holes
are the key to information benefits. The optimized network has two design principles.
Efficiency
The first design principle of an optimized network concerns efficiency: Maximize
the number of nonredundant contacts in the network to maximize the yield in
structural holes per contact. Given two networks of equal size, the one with more
nonredundant contacts provides more benefits. There is little gain from a new
contact redundant with existing contacts. Time and energy would be better spent
cultivating a new contact to unreached people. Maximizing the nonredundancy of
contacts maximizes the structural holes obtained per contact.
Efficiency is illustrated by the networks in Figure 13.3. These reach the same
people reached by the networks in Figure 13.1, but in a different way. What expands
in Figure 13.1 is not the benefits, but the cost of maintaining the network. Network
A provides four nonredundant contacts. Network B provides the same number. The
information benefits provided by the initial four contacts are redundant with
benefits provided by their close friends. All that has changed is the doubled number
of relationships maintained in the network. The situation deteriorates even further
with the sixteen contacts in network C. There are still only four nonredundant
contacts in the network, but their benefits are now obtained at a cost of maintaining
sixteen relationships.
With a little network surgery, the sixteen contacts can be maintained at a fourth
of the cost. As illustrated in Figure 13.3, select one contact in each cluster to be a
primary link to the cluster. Concentrate on maintaining the primary contact, and
allow direct relationships with others in the cluster to weaken into indirect relations
through the primary contact. These players reached indirectly are secondary
Structural Holes 211
contacts. Among the redundant contacts in a cluster, the primary contact should be
the one most easily maintained and most likely to honor an interpersonal debt to
you in particular. The secondary contacts are less easily maintained or less likely to
work for you (even if they might work well for someone else). The critical decision
obviously lies in selecting the right person to be a primary contact. The importance
of trust has already been discussed. With a trustworthy primary contact, there is
needed to maintain the cluster in the network.
Repeating this operation for each cluster in the network recovers effort that would
otherwise be spent maintaining redundant contacts. By reinvesting that saved time
and effort in developing primary contacts to new clusters, the network expands to
include an exponentially larger number of contacts while expanding contact
diversity. The sixteen contacts in network C of Figure 13.1, for example, are
maintained at a cost of four primary contacts in network C' of Figure 13.3. Some
portion of the time spent maintaining the redundant other twelve contacts can
be reallocated to expanding the network to include new clusters.
Effectiveness
The second design principle of an optimized network requires a further shift in
perspective: Distinguish primary from secondary contacts in order to focus resources
on preserving the primary contacts. Here contacts are not people on the other end
of your relations; they are ports of access to clusters of people beyond. Guided by the
first principle, these ports should be nonredundant so as to reach separate, and
therefore more diverse, social worlds of network benefits. Instead of maintaining
relations with all contacts, the task of maintaining the total network is delegated to
primary contacts. The player at the center of the network is then free to tocus on
properly supporting relations with primary contacts and expanding the network to
include new clusters. The first principle concerns the average number of people
reached with a primary contact; the second concerns the total number of people
reached with all primary contacts. The first principle concerns the yield per primary
contact. The second concerns the total yield of the network. More concretely, the
first principle moves from the networks in Figure 13.1 to the corresponding networks
in Figure 13.3. The second principle moves from left to right in Figure 13.3. The
target is network C' in Figure 13.3: a network of few primary contacts, each a port
of access to a cluster of many secondary contacts.
Figure 13.4 illustrates some complexities in unpacking a network to maximize
structural holes. The "before" network contains five primary contacts and reaches a
total of fifteen people. However, there are only two clusters of nonredundant contacts
in the network. Contacts 2 and 3 are redundant in the sense of being connected with
each other and reaching the same people (cohesion and structural equivalence
criteria). The same is true of contacts 4 and 5. Contact 1 is not connected directly to
contact 2, but he reaches the same secondary contacts; thus contacts 1 and 2 provide
redundant network benefits (structural equivalence criterion). Illustrating the other
 212 RonaldS. Burt Structural Holes 213

together, which gives you the opportunity to coordinate their activities. These
benefits are compounded by the fact that having a network that yields such benefits
makes you even more attractive as a network contact to other people, thus easing
your task of expanding the network to best serve your interests.

Control and the Tertius Gaudens

I have described how structural holes can determine who knows about opportunities,
when they know, and who gets to participate in them. Players with a network optimized
for structural holes, in addition to being exposed to more rewarding opportunities,
are also more likely to secure favorable terms in the opportunities they choose to
pursue. The structural holes that generate information benefits also generate control
benefits, giving certain players an advantage in negotiating their relationships.
To describe how this is so, I break the negotiation into structural, motivational, and
outcome components (corresponding to the textbook distinction between market
structure, market conduct, and market performance; for example, Caves, 1982). The
social structure of the competitive arena defines opportunities, a player decides to
pursue an opportunity, and is sometimes successful. I will begin with the outcome.

Tertius gaudens
Sometimes you will emerge successful from negotiation as the tertius gaudens. Taken
from the work of Georg Simmel, the tertius role is useful here because it defines
extreme, contacts 3 and 5 are connected directly, but they are nonredundant because
successful negotiation in terms of the social structure of the situation in which
they reach separate clusters of secondary contacts (structural equivalence criterion).
negotiation is successful. The role is the heart of Simmel's (1922) later analysis of the
In the after network, contact 2 is used to reach the first cluster in the "before
freedom an individual derives from conflicting group affiliations (see Coser, 1975,
network and contact 4 is used to reach the second cluster. The time and energy saved
for elaboration). The tertius gaudens is "the third who benefits" (Simmel, 1923:154,
by withdrawing from relations with the other three primary contacts is reallocated
232). The phrase survives in an Italian proverb, Far i due litiganti, il terzo gode
to primary contacts in new clusters. The "before" and "after" networks are both
(Between two fighters, the third benefits), and, to the north, in a more jovial Dutch
maintained at a cost of five primary relationships, but the "after" network is
wording, de lachende derde (the laughing third). Tertius, terzo, or derde, the phrase
dramatically richer in structural holes, and so network benefits.
describes an individual who profits from the disunion of others.
Network benefits are enhanced in several ways. There is a higher volume of
There are two tertius strategies: being the third between two or more players after
benefits, because more contacts are included in the network. Beyond volume,
the same relationship, and being the third between players in two or more relations
diversity enhances the quality of benefits. Nonredundant contacts ensure exposure
with conflicting demands. The first, and simpler, strategy is the familiar one that
to diverse sources of information. Each cluster of contacts is an independent source
occurs in economic bargaining between buyer and seller. When two or more players
of information. One cluster, no matter how numerous its members, is only
want to buy something, the seller can play their bids against one another to get a
one source of information, because people connected to one another tend to know
higher price. The strategy extends directly: a woman with multiple suitors or a
about the same things at about the same time. The information screen provided by
professor with simultaneous offers of positions in rival institutions.
multiple clusters of contacts is broader, providing better assurance that you, the The control benefits of having a choice between players after the same relationship
player, will be informed of opportunities and impending disasters. Further, because extends directly to choice between the simultaneous demands of players in separate
nonredundant contacts are only linked through the central player, you are assured relationships. The strategy can be seen between hierarchical statuses in the enterprising
ot being the first to see new opportunities created by needs in one group that could subordinate under the authority of two or more superiors: for example, the student
be served by skills in another group. You become the person who first brings people
 214 RonaldS. Burt
who strikes her own balance between the simultaneous demands of imperious faculty
advisors. The bargaining is not limited to situations of explicit competition. In some
situations, emerging as the tertius depends on creating competition. In proposing
the concept of a role-set, for example, Merton (1957:393-394) identifies this as
a strategy to resolve conflicting role demands. Make simultaneous, contradictory
demands explicit to the people posing them, and ask them to resolve their - now
explicit - conflict. Even where it doesn't exist, competition can be produced by
defining issues such that contact demands become contradictory and must be
resolved before you can meet their requests. Failure is possible. You might provide too
little incentive for the contacts to resolve their differences. Contacts drawn from
different social strata need not perceive one another's demands as carrying equal
weight. Or you might provide too much incentive. Now aware of one another, the
contacts could discover sufficient reason to cooperate in forcing you to meet their
mutually agreed-upon demands (Simmel, 1902:176,180-181, calls attention to such
failures). But if the strategy is successful, the pressure on you is alleviated and
is replaced with an element of control over the negotiation. Merton (1957:430) states
the situation succinctly: the player at the center of the network,"originally at the focus
of the conflict, virtually becomes a more or less influential bystander whose function
it is to highlight the conflicting demands by members of his role-set and to make it a
problem for them, rather than for him, to resolve their contradictory demands."
The strategy holds equally well with large groups. Under the rubric "divide and
rule," Simmel (1902:185-186) describes institutional mechanisms through which
the Incan and Venetian governments obtained advantage by creating conflict
between subjects. The same point is illustrated more richly in Barkey's (1991)
comparative description of state control in early seventeenth-century France and
Turkey. After establishing the similar conditions in the two states at the time, Barkey
asks why peasant-noble alliances developed in France against the central state while
no analogous or substitutable alliances developed in Turkey. The two empires were
comparable with respect to many factors that scholars have cited to account for
peasant revolt. They differed in one significant factor correlated with revolt - not in
the structure of centralized state control, but in control strategy. In France, the king
sent trusted representatives as agents to collect taxes and to carry out military
decisions in provincial populations. These outside agents, intendants, affected
fundamental local decisions and their intrusion was resented by the established local
nobility. Local nobility formed alliances with the peasantry against the central state.
In Turkey, the sultan capitalized on conflict among leaders in the provinces. When a
bandit became a serious threat to the recognized governor, a deal was struck with the
bandit to make him the legitimate governor. Barkey (1991:710) writes: "At its
extreme, the state could render a dangerous rebel legitimate overnight by striking a
bargain that ensured new sources of revenue for the rebel and momentary relief
from internal warfare and, perhaps, an army or two for the state." The two empires
differed in their use of structural holes. The French king, assuming he had absolute
authority, ignored them. The Turkish sultan, promoting competition between
alternative leaders, strategically exploited them. Conflict within the Turkish empire
Structural Holes 215
remained in the province, rather than being directed against the central state. As is
characteristic of the control obtained via structural holes, the resulting Turkish
control was more negotiated than was the absolute control exercised in France.
It was also more effective.
The essential tension
There is a presumption of tension here. Control emerges from tertius brokering
tension between other players. No tension, no tertius.
[...]
The tension essential to the tertius is merely uncertainty. Separate the uncertainty
of control from its consequences. The consequences can be life or death, in the
extreme, or merely a question of embarrassment. Everyone knows you made an
effort to get that job, but it went to someone else. The tertius strategies can be applied
to control with severe consequences or to control of little consequence. What is
essential is that control is uncertain, that no one can act as if he or she has absolute
authority. Where there is any uncertainty about whose preferences should dominate
a relationship, there is an opportunity for the tertius to broker the negotiation for
control by playing demands against one another. There is no long-term contract
that keeps a relationship strong, no legal binding that can secure the trust necessary
to a productive relationship. Your network is a pulsing swirl of mixed, conflicting
demands. Each contact wants your exclusive attention, your immediate response
when a concern arises. All, to warrant their continued confidence in you, want to see
you measure up to the values against which they judge themselves. Within this
preference webwork, where no demands have absolute authority, the tertius
negotiates for favorable terms.
The connection with information benefits
Structural holes are the setting for tertius strategies. Information is the substance.
Accurate, ambiguous, or distorted information is moved between contacts by the
tertius. One bidder is informed of a competitive offer in the first tertius strategy.
A player in one relationship is informed of demands from other relationships in the
second tertius strategy.
The two kinds of benefits augment and depend on one another. Application of
the tertius strategies elicits additional information from contacts interested in
resolving the negotiation in favor of their own preferences.The information benefits
of access, timing, and referrals enhance the application of strategy. Successful
application of the tertius strategies involves bringing together players who are willing
to negotiate, have sufficiently comparable resources to view one another s preferences
as valid, but won't negotiate with one another directly to the exclusion of the tertius.
Having access to information means being able to identify where there will be an
advantage in bringing contacts together and is the key to understanding the resources
and preferences being played against one another. Having that information early is
 Structural Holes 217
216 RonaldS. Burt

©
A. Market
the difference between being the one who brings together contacts versus being just
another person who hears about the negotiation. Referrals further enhance strategy.
It is one thing to distribute information between two contacts during negotiation. It YOU
is another thing to have people close to each contact endorsing the legitimacy of the
information you distribute.

B. Clique

Secondary Holes
YOU

[...] I have linked opportunity to structural holes, but not with respect to the
whole domain of relevant holes. Thus far, a network optimized for entrepreneurial
opportunity has a vine-and-cluster structure. As illustrated in Figures 13.3 and
13.4, a player has direct relations with primary contacts, each a port of access to a C. Hierarchy

cluster of redundant secondary contacts. Structural holes between the primary

contacts, primary structural holes, provide information and control benefits. But YOU

the benefits they provide are affected by structural holes just beyond the border of
the network. Structural holes among the secondary contacts within the cluster
Figure 13.5 Contact clusters with and without secondary structural holes.
around each primary contact play a role in the tertius strategies. These are
secondary structural holes.

the connection between redundant contacts, your only recourse is to live with your
Control benefits and secondary structural holes contact's demands, dominate the cluster, or cut the cluster from your network.
Network density is not the issue here. Situation C is a relatively low density cluster
The ultimate threat in negotiating a relationship is withdrawal: either severing your
(43% of the 28 relations within the cluster are marked with a line as strong), but
link to a former contact's cluster or transferring the primary relationship to a new
contacts within the cluster are coordinated through their joint ties to two leaders in
person in the cluster. This threat depends on two things. First, there must be
the center. It doesn't make sense to negotiate the price of a purchase in a department
alternatives, secondary contacts who are redundant with your primary contact and
store by playing one sales clerk against another. They both answer to a higher
capable of replacing the primary contact in your network. Examples include an
authority. You have to make a purchase sufficiently large that it allows you to deal
alternative spouse in the case of negotiating a conjugal relationship, an alternative
with someone higher in the organization. Then, as in Situation C, you can develop
job in the case of negotiating with a truculent supervisor, or an alternative supplier
the structural hole between the two leaders at the center of the circle and play one
in the case of a firm renewing a contract with a past supplier. Second, there must
be structural holes among thesecondary contacts. If there are no contacts substitutable leader against the other.
for your current primary contact, he or she is free to impose demands - up to the
limit of structural holes between primary contacts. If your current primary contact
is in collusion with whatever substitutes exist, which eliminates structural holes
you might exploit, he or she is free to impose demands - again, up to the limit of Structural Autonomy
structural holes between primary contacts.
The argument can now be summarized with a concept defining the extent to which
Consider Figure 13.5. You are negotiating with a primary contact in a cluster of
a player's network is rich in structural holes, and thus rich in entrepreneurial
redundant contacts indicated by dots enclosed by a gray circle. Situation A illustrates
opportunity, and thus rich in information and control benefits. The concept is
the familiar negotiation between buyer and seller. You use the offer from one buyer
structural autonomy and the general argument is sketched in Figure 13.6. [...]
to raise the other's offer.
The argument began with a generic production equation. Profit equals an
Situation B illustrates the exact opposite condition. Here the redundant contacts
investment multiplied by a rate of return. The benefits of a relationship can be
are all connected by strong relations. This is the situation of negotiating with a
expressed in an analogous form: time and energy invested to reach a contact
member of a social clique or cult. In the absence of holes over which you can broker
 218 RonaldS. Burt Structural Holes 219

entrepreneurial opportunities depends on having numerous structural holes

Competitive Advantage of Structural Holes
A Question of around your contacts and none attached to yourself.
Production
Kind of Substance of Social Structural Condition Responsible These considerations come together in the concept of structural autonomy.
Advantage Advantage for the Advantage Players with relationships free of structural holes at their own end and rich in
Profits
structural holes at the other end are structurally autonomous. These are the players
Investment • Rate of Return Information Access, Timing Contact Redundancy & Structural Holes best positioned for the information and control benefits that a network can provide.
Benefits and Referrals network trust, size & diversity,
cohesion & structural equivalence [...] Structural autonomy summarizes the action potential of the tertiuss network.
efficient-effective networks The budget equation for optimizing structural autonomy has an upper limit set by
A Question of structural holes & weak ties
Opportunity the time and energy of the tertius, and a trade-off between the structural holes a new
contact provides versus the time and energy required to maintain a productive
Control Tertius Gaudens, Structural Autonomy
Benefits Entrepreneurial holes & entrepreneurial opportunity relationship with the contact.
Motivation primary holes & constraint
secondary holes & constraint
hole signature & structural autonomy

Conclusion: Players with contact networks optimized for structural holes - players with
REFERENCES
networks providing high structural autonomy - enjoy higher rates of return on their
investments because they know about, have a hand in, and
Barkey, Karen. 1991. Rebellious alliances: the state and peasant unrest in early seventeenth
^exercise control over, more rewarding opportunities.
century France and the Ottoman empire. American Sociological Review 56: 699-715.
Burt, Ronald S. 1990. Kinds of relations in American discussion networks, pp. 411-451 in
Figure 13.6 Argument.
Structures of Power and Constraint, ed. C. Calhoun, M. W. Meyer, and W. R. Scott.
New York: Cambridge University Press.
Caves, Richard E. 1982. American Industry: Structure, Conduct, Performance. Englewood
multiplied by a rate of return. A player's entrepreneurial opportunities are enhanced
Cliffs, N.J.: Prentice-Hall.
by a relationship to the extent that: (a) the player has invested substantial time and Cole, Stephen. 1992. Making Science. Cambridge, MA: Harvard University Press.
energy to secure a connection with the contact, and (b) there are many structural Coser, Rose Laub. 1975. The complexity of roles as a seedbed of individual autonomy.
holes around the contact ensuring a high rate of return on the investment. More Pp. 237-263 in The Idea of Social Structure, ed. L. A. Coser. New York: Harcourt, Brace,
specifically, rate of return concerns how and whom you reach with the relationship. Jovanovich.
Time and energy invested to reach a player with more resources generates more Festinger, Leon, Stanley Schachter, and Kurt W. Back. 1950. Social Pressures in Informal
social capital. For the sake of argument, as explained in the discussion of social Groups. Stanford, CA: Stanford University Press.
Homans, George C. 1950. The Human Group. New York: Harcourt, Brace and World.
capital, I assume that a player with a network optimized for structural holes can
Marsden, Peter V. and Jeanne S. Hurlbert. 1988. Social resources and mobility outcomes:
identify suitably endowed contacts. My concern is the how of a relationship, defined
application and extension. Social Forces 67: 1038-1059.
by the structure of a network and its connection with the social structure of the
Merton, Robert K. (1957) 1968. Continuities in the theory of reference group behavior.
competitive arena. Thus the rate of return keyed to structural holes is a product of
pp. 335-440 in Social Theory and Social Structure. New York: Free Press.
the extent to which there are: (a) many primary structural holes between the contact Simmel, Georg. 1902. The number of members as determining the sociological form of the
and others in the player's network, and (b) many secondary structural holes between group, II. Trans. A. Small. American Journal of Sociology 8: 158-96.
the contact and others outside the network who could replace the contact. Simmel, Georg. (1922) 1955. Conflict and Web of Group Affiliations. Trans. K. H. Wolff and
There is also the issue of structural holes around the player. As the holes around R. Bendix. New York: Free Press.
contacts provide information and control benefits to the player, holes around the Simmel, Georg. (1923) 1950. The Sociology of Georg Simmel. Trans. K. H. Wolff. New York:
player can be developed by contacts for their benefit. [...] Your contacts have the Free Press.
option of replacing you with one of your colleagues who provides the same
network benefits that you do. To manage this uncertainty, you might develop
relationships with your colleagues so that it would be difficult to play them off
against you (an oligopoly strategy), or you might specialize in some way so that
they no longer provide network benefits redundant with your own (a differentiation
strategy). [...] The point here is that your negotiating position is weaker than
expected from the distribution of structural holes around contacts. Developing