Power and Its Pathologies

The Politics, Power, and Pathologies
of International Organizations
Michael N. Barnett and Martha Finnemore
Do international organizations really do what their creators intend them to do? In the
past century the number of international organizations (IOs) has increased exponen-
tially, and we have a variety of vigorous theories to explain why they have been
created. Most of these theories explain IO creation as a response to problems of
incomplete information, transaction costs, and other barriers to Pareto efficiency and
welfare improvement for their members. Research flowing from these theories, how-
ever, has paid little attention to how IOs actually behave after they are created. Closer
scrutiny would reveal that many IOs stray from the efficiency goals these theories
impute and that many IOs exercise power autonomously in ways unintended and
unanticipated by states at their creation. Understanding how this is so requires a
reconsideration of IOs and what they do.
In this article we develop a constructivist approach rooted in sociological institu-
tionalism to explain both the power of IOs and their propensity for dysfunctional,
even pathological, behavior. Drawing on long-standing Weberian arguments about
bureaucracy and sociological institutionalist approaches to organizational behavior,
we argue that the rational-legal authority that IOs embody gives them power indepen-
dent of the states that created them and channels that power in particular directions.
Bureaucracies, by definition, make rules, but in so doing they also create social
knowledge. They define shared international tasks (like ‘‘development’’), create and
define new categories of actors (like ‘‘refugee’’), create new interests for actors (like
‘‘promoting human rights’’), and transfer models of political organization around the
world (like markets and democracy.) However, the same normative valuation on
impersonal, generalized rules that defines bureaucracies and makes them powerful in
We are grateful to John Boli, Raymond Duvall, Ernst Haas, Peter Haas, Robert Keohane, Keith Krause,
Jeffrey Legro, John Malley, Craig Murphy, M. J. Peterson, Mark Pollack, Andrew Moravcsik, Thomas
Risse, Duncan Snidal, Steve Weber, Thomas Weiss, and two anonymous referees for their comments. We
are especially grateful for the careful attention of the editors of International Organization. Earlier ver-
sions of this article were presented at the 1997 APSA meeting, the 1997 ISA meeting, and at various fora.
We also acknowledge financial assistance from the Smith Richardson Foundation and the United States
Institute of Peace.
International Organization 53, 4, Autumn 1999, pp. 699–732

r 1999 by The IO Foundation and the Massachusetts Institute of Technology
@xyserv2/disk3/CLS_jrnl/GRP_inor/JOB_inor53-4/DIV_087k03 cind
700 International Organization
modern life can also make them unresponsive to their environments, obsessed with
their own rules at the expense of primary missions, and ultimately lead to inefficient,
self-defeating behavior. We are not the first to suggest that IOs are more than the
reflection of state preferences and that they can be autonomous and powerful actors
in global politics.1 Nor are we the first to note that IOs, like all organizations, can be
dysfunctional and inefficient.2 However, our emphasis on the way that characteristics
of bureaucracy as a generic cultural form shape IO behavior provides a different and
very broad basis for thinking about how IOs influence world politics.3
Developing an alternative approach to thinking about IOs is only worthwhile if it
produces significant insights and new opportunities for research on major debates in
the field. Our approach allows us to weigh in with new perspectives on at least three
such debates. First, it offers a different view of the power of IOs and whether or how
they matter in world politics. This issue has been at the core of the neoliberal-
institutionalists’ debate with neorealists for years.4 We show in this article how neo-
liberal-institutionalists actually disadvantage themselves in their argument with real-
ists by looking at only one facet of IO power. Global organizations do more than just
facilitate cooperation by helping states to overcome market failures, collective action
dilemmas, and problems associated with interdependent social choice. They also
create actors, specify responsibilities and authority among them, and define the work
these actors should do, giving it meaning and normative value. Even when they lack
material resources, IOs exercise power as they constitute and construct the social
world.5
Second and related, our perspective provides a theoretical basis for treating IOs as
autonomous actors in world politics and thus presents a challenge to the statist ontol-
ogy prevailing in international relations theories. Despite all their attention to inter-
national institutions, one result of the theoretical orientation of neoliberal institution-
alists and regimes theorists is that they treat IOs the way pluralists treat the state. IOs
are mechanisms through which others (usually states) act; they are not purposive
actors. The regimes literature is particularly clear on this point. Regimes are ‘‘prin-
ciples, norms, rules, and decision-making procedures;’’ they are not actors.6 Weber’s
insights about the normative power of the rational-legal authority that bureaucracies
embody and its implications for the ways bureaucracies produce and control social
knowledge provide a basis for challenging this view and treating IOs as agents, not
just as structure.
1. For Gramscian approaches, see Cox 1980, 1992, and 1996; and Murphy 1995. For Society of States
approaches, see Hurrell and Woods 1995. For the epistemic communities literature, see Haas 1992. For IO
decision-making literature, see Cox et al. 1974; Cox and Jacobson 1977; Cox 1996; and Ness and Brechin
1988. For a rational choice perspective, see Snidal 1996.
2. Haas 1990.
3. Because the neorealist and neoliberal arguments we engage have focused on intergovernmental
organizations rather than nongovernmental ones, and because Weberian arguments from which we draw
deal primarily with public bureaucracy, we too focus on intergovernmental organizations in this article and
use the term international organizations in that way.
4. Baldwin 1993.
5. See Finnemore 1993 and 1996b; and McNeely 1995.
6. Krasner 1983b.
Pathologies of International Organizations 701
Third, our perspective offers a different vantage point from which to assess the
desirability of IOs. While realists and some policymakers have taken up this issue,
surprisingly few other students of IOs have been critical of their performance or
desirability.7 Part of this optimism stems from central tenets of classical liberalism,
which has long viewed IOs as a peaceful way to manage rapid technological change
and globalization, far preferable to the obvious alternative—war.8 Also contributing
to this uncritical stance is the normative judgment about IOs that is built into the
theoretical assumptions of most neoliberal and regimes scholars and the economic
organization theories on which they draw. IOs exist, in this view, only because they
are Pareto improving and solve problems for states. Consequently, if an IO exists, it
must be because it is more useful than other alternatives since, by theoretical axiom,
states will pull the plug on any IO that does not perform. We find this assumption
unsatisfying. IOs often produce undesirable and even self-defeating outcomes repeat-
edly, without punishment much less dismantlement, and we, as theorists, want to
understand why. International relations scholars are familiar with principal-agent
problems and the ways in which bureaucratic politics can compromise organizational
effectiveness, but these approaches have rarely been applied to IOs. Further, these
approaches by no means exhaust sources of dysfunction. We examine one such source
that flows from the same rational-legal characteristics that make IOs authoritative
and powerful. Drawing from research in sociology and anthropology, we show how
the very features that make bureaucracies powerful can also be their weakness.
The claims we make in this article flow from an analysis of the ‘‘social stuff’’ of
which bureaucracy is made. We are asking a standard constructivist question about
what makes the world hang together or, as Alexander Wendt puts it, ‘‘how are things
in the world put together so that they have the properties they do.’’9 In this sense, our
explanation of IO behavior is constitutive and differs from most other international
relations approaches. This approach does not make our explanation ‘‘mere descrip-
tion,’’ since understanding the constitution of things does essential work in explain-
ing how those things behave and what causes outcomes. Just as understanding how
the double-helix DNA molecule is constituted materially makes possible causal argu-
ments about genetics, disease, and other biological processes, so understanding how
bureaucracies are constituted socially allows us to hypothesize about the behavior of
IOs and the effects this social form might have in world politics. This type of consti-
tutive explanation does not allow us to offer law-like statements such as ‘‘if X hap-
pens, then Y must follow.’’ Rather, by providing a more complete understanding of
what bureaucracy is, we can provide explanations of how certain kinds of bureau-
cratic behavior are possible, or even probable, and why.10
We begin by examining the assumptions underlying different branches of organi-
zation theory and exploring their implications for the study of IOs. We argue that
assumptions drawn from economics that undergird neoliberal and neorealist treat-
7. See Mearsheimer 1994; and Helms 1996.

8. See Commission on Global Governance 1995; Jacobson 1979, 1; and Doyle 1997.
9. See Ruggie 1998; and Wendt 1998.
10. Wendt 1998.
ments of IOs do not always reflect the empirical situation of most IOs commonly
studied by political scientists. Further, they provide research hypotheses about only
some aspects of IOs (like why they are created) and not others (like what they do).
We then introduce sociological arguments that help remedy these problems.
In the second section we develop a constructivist approach from these sociological
arguments to examine the power wielded by IOs and the sources of their influence.
Liberal and realist theories only make predictions about, and consequently only look
for, a very limited range of welfare-improving effects caused by IOs. Sociological theories,
however, expect and explain a much broader range of impacts organizations can have and
specifically highlight their role in constructing actors, interests, and social purpose.
We provide illustrations from the UN system to show how IOs do, in fact, have such
powerful effects in contemporary world politics. In the third section we explore the dysfunc-
tional behavior of IOs, which we define as behavior that undermines the stated goals of
the organization. International relations theorists are familiar with several types of
theories that might explain such behavior. Some locate the source of dysfunction in
material factors, others focus on cultural factors. Some theories locate the source of
dysfunction outside the organization, others locate it inside. We construct a typology,
mapping these theories according to the source of dysfunction they emphasize, and
show that the same internally generated cultural forces that give IOs their power and
autonomy can also be a source of dysfunctional behavior. We use the term pathologies to
describe such instances when IO dysfunction can be traced to bureaucratic culture. We
conclude by discussing how our perspective helps to widen the research agenda for IOs.
Theoretical Approaches to Organizations
Within social science there are two broad strands of theorizing about organizations.
One is economistic and rooted in assumptions of instrumental rationality and effi-
ciency concerns; the other is sociological and focused on issues of legitimacy and
power.11 The different assumptions embedded within each type of theory focus atten-
tion on different kinds of questions about organizations and provide insights on dif-
ferent kinds of problems.
The economistic approach comes, not surprisingly, out of economics departments
and business schools for whom the fundamental theoretical problem, laid out first by
Ronald Coase and more recently by Oliver Williamson, is why we have business
firms. Within standard microeconomic logic, it should be much more efficient to
conduct all transactions through markets rather than ‘‘hierarchies’’ or organizations.
Consequently, the fact that economic life is dominated by huge organizations (busi-
ness firms) is an anomaly. The body of theory developed to explain the existence and
power of firms focuses on organizations as efficient solutions to contracting prob-
lems, incomplete information, and other market imperfections.12
11. See Powell and DiMaggio 1991, chap. 1; and Grandori 1993.
12. See Williamson 1975 and 1985; and Coase 1937.
This body of organization theory informs neoliberal and neorealist debates over
international institutions. Following Kenneth Waltz, neoliberals and neorealists un-
derstand world politics to be analogous to a market filled with utility-maximizing
competitors.13 Thus, like the economists, they see organizations as welfare-improv-
ing solutions to problems of incomplete information and high transaction costs.14
Neoliberals and realists disagree about the degree to which constraints of anarchy, an
interest in relative versus absolute gains, and fears of cheating will scuttle interna-
tional institutional arrangements or hobble their effectiveness, but both agree, implic-
itly or explicitly, that IOs help states further their interests where they are allowed to
work.15 State power may be exercised in political battles inside IOs over where, on
the Pareto frontier, political bargains fall, but the notion that IOs are instruments
created to serve state interests is not much questioned by neorealist or neoliberal
scholars.16 After all, why else would states set up these organizations and continue to
support them if they did not serve state interests?
Approaches from sociology provide one set of answers to this question. They
provide reasons why, in fact, organizations that are not efficient or effective servants
of member interests might exist. In so doing, they lead us to look for kinds of power
and sources of autonomy in organizations that economists overlook. Different ap-
proaches within sociology treat organizations in different ways, but as a group they
stand in sharp contrast to the economists’ approaches in at least two important respects:
they offer a different conception of the relationship between organizations and their
environments, and they provide a basis for understanding organizational autonomy.
IOs and their environment. The environment assumed by economic approaches

to organizations is socially very thin and devoid of social rules, cultural content, or
even other actors beyond those constructing the organization. Competition, ex-
change, and consequent pressures for efficiency are the dominant environmental char-
acteristics driving the formation and behavior of organizations. Sociologists, by con-
trast, study organizations in a wider world of nonmarket situations, and, consequently,
they begin with no such assumptions. Organizations are treated as ‘‘social facts’’ to
be investigated; whether they do what they claim or do it efficiently is an empirical
question, not a theoretical assumption of these approaches. Organizations respond
not only to other actors pursuing material interests in the environment but also to
normative and cultural forces that shape how organizations see the world and concep-
tualize their own missions. Environments can ‘‘select’’ or favor organizations for
reasons other than efficient or responsive behavior. For example, organizations may
be created and supported for reasons of legitimacy and normative fit rather than
efficient output; they may be created not for what they do but for what they are—for
what they represent symbolically and the values they embody.17
13. Waltz 1979.
14. See Vaubel 1991, 27; and Dillon, Ilgen, and Willett 1991.
15. Baldwin 1993.
16. Krasner 1991.
17. See DiMaggio and Powell 1983; Scott 1992; Meyer and Scott 1992, 1–5; Powell and DiMaggio
1991; Weber 1994; and Finnemore 1996a.
Empirically, organizational environments can take many forms. Some organiza-

tions exist in competitive environments that create strong pressures for efficient or
responsive behavior, but many do not. Some organizations operate with clear criteria
for ‘‘success’’ (like firms that have balance sheets), whereas others (like political
science departments) operate with much vaguer missions, with few clear criteria for
success or failure and no serious threat of elimination. Our point is simply that when
we choose a theoretical framework, we should choose one whose assumptions ap-
proximate the empirical conditions of the IO we are analyzing, and that we should be
aware of the biases created by those assumptions. Economistic approaches make
certain assumptions about the environment in which IOs are embedded that drive
researchers who use them to look for certain kinds of effects and not others. Specify-
ing different or more varied environments for IOs would lead us to look for different
and more varied effects in world politics.18
IO autonomy. Following economistic logic, regime theory and the broad range of
scholars working within it generally treat IOs as creations of states designed to fur-
ther state interests.19 Analysis of subsequent IO behavior focuses on processes of
aggregating member state preferences through strategic interaction within the struc-
ture of the IO. IOs, then, are simply epiphenomena of state interaction; they are, to
quote Waltz’s definition of reductionism, ‘‘understood by knowing the attributes and
the interactions of [their] parts.’’20
These theories thus treat IOs as empty shells or impersonal policy machinery to be
manipulated by other actors. Political bargains shape the machinery at its creation,
states may politick hard within the machinery in pursuit of their policy goals, and the
machinery’s norms and rules may constrain what states can do, but the machinery
itself is passive. IOs are not purposive political actors in their own right and have no
ontological independence. To the extent that IOs do, in fact, take on a life of their
own, they breach the ‘‘limits of realism’’ as well as of neoliberalism by violating the
ontological structures of these theories.21
The regimes concept spawned a huge literature on interstate cooperation that is
remarkably consistent in its treatment of IOs as structure rather than agents. Much of
the neoliberal institutionalist literature has been devoted to exploring the ways in
which regimes (and IOs) can act as intervening variables, mediating between states’
pursuit of self-interest and political outcomes by changing the structure of opportuni-
ties and constraints facing states through their control over information, in particu-
lar.22 Although this line of scholarship accords IOs some causal status (since they
demonstrably change outcomes), it does not grant them autonomy and purpose inde-
18. Researchers applying these economistic approaches have become increasingly aware of the mis-
match between the assumptions of their models and the empirics of IOs. See Snidal 1996.
19. Note that empirically this is not the case; most IOs now are created by other IOs. See Shanks,
Jacobson, and Kaplan 1996.
20. Waltz 1979, 18.
21. Krasner 1983a, 355–68; but see Finnemore 1996b; and Rittberger 1993.
22. See Keohane 1984; and Baldwin 1993.
pendent of the states that comprise them. Another branch of liberalism has recently
divorced itself from the statist ontology and focuses instead on the preferences of
social groups as the causal engine of world politics, but, again, this view simply
argues for attention to a different group of agents involved in the construction of IOs
and competing for access to IO mechanisms. It does not offer a fundamentally differ-
ent conception of IOs.23
The relevant question to ask about this conceptualization is whether it is a reason-
able approximation of the empirical condition of most IOs. Our reading of detailed
empirical case studies of IO activity suggests not. Yes, IOs are constrained by states,
but the notion that they are passive mechanisms with no independent agendas of their
own is not borne out by any detailed empirical study of an IO that we have found.
Field studies of the European Union provide evidence of independent roles for
‘‘eurocrats.’’24 Studies of the World Bank consistently identify an independent cul-
ture and agendas for action.25 Studies of recent UN peacekeeping and reconstruction
efforts similarly document a UN agenda that frequently leads to conflict with mem-
ber states.26 Accounts of the UN High Commission on Refugees (UNHCR) routinely
note how its autonomy and authority has grown over the years. Not only are IOs
independent actors with their own agendas, but they may embody multiple agendas
and contain multiple sources of agency—a problem we take up later.
Principal-agent analysis, which has been increasingly employed by students of
international relations to examine organizational dynamics, could potentially provide
a sophisticated approach to understanding IO autonomy.27 Building on theories of
rational choice and of representation, these analysts understand IOs as ‘‘agents’’ of
states (‘‘principals’’). The analysis is concerned with whether agents are responsible
delegates of their principals, whether agents smuggle in and pursue their own prefer-
ences, and how principals can construct various mechanisms to keep their agents
honest.28 This framework provides a means of treating IOs as actors in their own
right with independent interests and capabilities. Autonomous action by IOs is to be
expected in this perspective. It would also explain a number of the nonresponsive
and pathological behaviors that concern us because we know that monitoring and
shirking problems are pervasive in these principal-agent relationships and that these
relationships can often get stuck at suboptimal equilibria.
The problem with applying principal-agent analysis to the study of IOs is that it
requires a priori theoretical specification of what IOs want. Principal-agent dynamics
are fueled by the disjuncture between what agents want and what principals want. To
produce any insights, those two sets of interests cannot be identical. In economics
this type of analysis is usually applied to preexisting agents and principals (clients
23. Moravcsik 1997.

24. See Pollack 1997; Ross 1995; and Zabusky 1995; but see Moravcsik 1999.
25. See Ascher 1983; Ayres 1983; Ferguson 1990; Escobar 1995; Wade 1996; Nelson 1995; and
Finnemore 1996a.
26. Joint Evaluation of Emergency Assistance to Rwanda 1996.
27. See Pollack 1997; Lake 1996; Vaubel 1991; and Dillon, Ilgen, and Willett 1991.
28. See Pratt and Zeckhauser 1985; and Kiewit and McCubbins 1991.
hiring lawyers, patients visiting doctors) whose ongoing independent existence makes
specification of independent interests relatively straightforward. The lawyer or the
doctor would probably be in business even if you and I did not take our problems to
them. IOs, on the other hand, are often created by the principals (states) and given
mission statements written by the principals. How, then, can we impute independent
preferences a priori?
Scholars of American politics have made some progress in producing substantive
theoretical propositions about what U.S. bureaucratic agencies want. Beginning with
the pioneering work of William Niskanen, scholars theorized that bureaucracies had
interests defined by the absolute or relative size of their budget and the expansion or
protection of their turf. At first these interests were imputed, and later they became
more closely investigated, substantiated, and in some cases modified or rejected alto-
gether.29
Realism and liberalism, however, provide no basis for asserting independent util-
ity functions for IOs. Ontologically, these are theories about states. They provide no
basis for imputing interests to IOs beyond the goals states (that is, principals) give
them. Simply adopting the rather battered Niskanen hypothesis seems less than prom-
ising given the glaring anomalies—for example, the opposition of many NATO and
OSCE (Organization for Security and Cooperation in Europe) bureaucrats to those
organizations’ recent expansion and institutionalization. There are good reasons to
assume that organizations care about their resource base and turf, but there is no
reason to presume that such matters exhaust or even dominate their interests. Indeed,
ethnographic studies of IOs describe a world in which organizational goals are strongly
shaped by norms of the profession that dominate the bureaucracy and in which inter-
ests themselves are varied, often in flux, debated, and worked out through interac-
tions between the staff of the bureaucracy and the world in which they are embed-
ded.30
Various strands of sociological theory can help us investigate the goals and behav-
ior of IOs by offering a very different analytical orientation than the one used by
economists. Beginning with Weber, sociologists have explored the notion that bureau-
cracy is a peculiarly modern cultural form that embodies certain values and can have
its own distinct agenda and behavioral dispositions. Rather than treating organiza-
tions as mere arenas or mechanisms through which other actors pursue interests,
many sociological approaches explore the social content of the organization—its
culture, its legitimacy concerns, dominant norms that govern behavior and shape
interests, and the relationship of these to a larger normative and cultural environ-
ment. Rather than assuming behavior that corresponds to efficiency criteria alone,
these approaches recognize that organizations also are bound up with power and
social control in ways that can eclipse efficiency concerns.
29. See Niskanen 1971; Miller and Moe 1983; Weingast and Moran 1983; Moe 1984; and Sigelman
1986.
30. See Ascher 1983; Zabusky 1995; Barnett 1997b; and Wade 1996.
The Power of IOs
IOs can become autonomous sites of authority, independent from the state ‘‘princi-
pals’’ who may have created them, because of power flowing from at least two sources:
(1) the legitimacy of the rational-legal authority they embody, and (2) control over
technical expertise and information. The first of these is almost entirely neglected by
the political science literature, and the second, we argue, has been conceived of very
narrowly, leading scholars to overlook some of the most basic and consequential
forms of IO influence. Taken together, these two features provide a theoretical basis
for treating IOs as autonomous actors in contemporary world politics by identifying
sources of support for them, independent of states, in the larger social environment.
Since rational-legal authority and control over expertise are part of what defines and
constitutes any bureaucracy (a bureaucracy would not be a bureaucracy without them),
the autonomy that flows from them is best understood as a constitutive effect, an
effect of the way bureaucracy is constituted, which, in turn, makes possible (and in
that sense causes) other processes and effects in global politics.
Sources of IO Autonomy and Authority

To understand how IOs can become autonomous sites of authority we turn to Weber
and his classic study of bureaucratization. Weber was deeply ambivalent about the
increasingly bureaucratic world in which he lived and was well-attuned to the vices
as well as the virtues of this new social form of authority.31 Bureaucracies are rightly
considered a grand achievement, he thought. They provide a framework for social
interaction that can respond to the increasingly technical demands of modern life in a
stable, predictable, and nonviolent way; they exemplify rationality and are techni-
cally superior to previous forms of rule because they bring precision, knowledge, and
continuity to increasingly complex social tasks.32 But such technical and rational
achievements, according to Weber, come at a steep price. Bureaucracies are political
creatures that can be autonomous from their creators and can come to dominate the
societies they were created to serve, because of both the normative appeal of rational-
legal authority in modern life and the bureaucracy’s control over technical expertise
and information. We consider each in turn.
Bureaucracies embody a form of authority, rational-legal authority, that modernity
views as particularly legitimate and good. In contrast to earlier forms of authority
that were invested in a leader, legitimate modern authority is invested in legalities,
procedures, and rules and thus rendered impersonal. This authority is ‘‘rational’’ in
that it deploys socially recognized relevant knowledge to create rules that determine
how goals will be pursued. The very fact that they embody rationality is what makes
bureaucracies powerful and makes people willing to submit to this kind of authority.
31. See Weber 1978, 196–97; Weber 1947; Mouzelis 1967; and Beetham 1985 and 1996.
32. See Schaar 1984, 120; Weber 1978, 973; and Beetham 1985, 69.
According to Weber,
in legal authority, submission does not rest upon the belief and devotion to
charismatically gifted persons. . . or upon piety toward a personal lord and
master who is defined by an ordered tradition. . . . Rather submission under legal
authority is based upon an impersonal bond to the generally defined and
functional ‘‘duty of office.’’ The official duty—like the corresponding right to
exercise authority: the ‘‘jurisdictional competency’’—is fixed by rationally
established norms, by enactments, decrees, and regulations in such a manner that
the legitimacy of the authority becomes the legality of the general rule, which is
purposely thought out, enacted, and announced with formal correctness.33
When bureaucrats do something contrary to your interests or that you do not like,
they defend themselves by saying ‘‘Sorry, those are the rules’’ or ‘‘just doing my
job.’’ ‘‘The rules’’ and ‘‘the job’’ are the source of great power in modern society. It is
because bureaucrats in IOs are performing ‘‘duties of office’’ and implementing ‘‘ra-
tionally established norms’’ that they are powerful.
A second basis of autonomy and authority, intimately connected to the first, is
bureaucratic control over information and expertise. A bureaucracy’s autonomy de-
rives from specialized technical knowledge, training, and experience that is not im-
mediately available to other actors. While such knowledge might help the bureau-
cracy carry out the directives of politicians more efficiently, Weber stressed that it
also gives bureaucracies power over politicians (and other actors). It invites and at
times requires bureaucracies to shape policy, not just implement it.34
The irony in both of these features of authority is that they make bureaucracies
powerful precisely by creating the appearance of depoliticization. The power of IOs,
and bureaucracies generally, is that they present themselves as impersonal, techno-
cratic, and neutral—as not exercising power but instead as serving others; the presen-
tation and acceptance of these claims is critical to their legitimacy and authority.35
Weber, however, saw through these claims. According to him, the depoliticized char-
acter of bureaucracy that legitimates it could be a myth: ‘‘Behind the functional
purposes [of bureaucracy], of course, ‘ideas of culture-values’ usually stand.’’36 Bu-
reaucracies always serve some social purpose or set of cultural values. That purpose
may be normatively ‘‘good,’’ as Weber believed the Prussian nationalism around him
was, but there was no a priori reason to assume this.
In addition to embodying cultural values from the larger environment that might
be desirable or not, bureaucracies also carry with them behavioral dispositions and
values flowing from the rationality that legitimates them as a cultural form. Some of
these, like the celebration of knowledge and expertise, Weber admired. Others con-
cerned him greatly, and his descriptions of bureaucracy as an ‘‘iron cage’’ and bureau-
crats as ‘‘specialists without spirit’’ are hardly an endorsement of the bureaucratic
33. Gerth and Mills 1978, 299 (italics in original).

34. See Gerth and Mills 1978, 233; Beetham 1985, 74–75; and Schaar 1984, 120.
35. We thank John Boli for this insight. Also see Fisher 1997; Ferguson 1990; Shore and Wright 1997;
and Burley and Mattli 1993.
36. Gerth and Mills 1978, 199.
form.37 Bureaucracy can undermine personal freedom in important ways. The very
impersonal, rule-bound character that empowers bureaucracy also dehumanizes it.
Bureaucracies often exercise their power in repressive ways, in the name of general
rules because rules are their raison d’être. This tendency is exacerbated by the way
bureaucracies select and reward narrowed professionals seeking secure careers inter-
nally—people who are ‘‘lacking in heroism, human spontaneity, and inventive-
ness.’’38 Following Weber, we investigate rather than assume the ‘‘goodness’’ of
bureaucracy.
Weber’s insights provide a powerful critique of the ways in which international
relations scholars have treated IOs. The legitimacy of rational-legal authority sug-
gests that IOs may have an authority independent of the policies and interests of
states that create them, a possibility obscured by the technical and apolitical treat-
ment of IOs by both realists and neoliberals. Nor have realists and neoliberals consid-
ered how control over information hands IOs a basis of autonomy. Susan Strange, at
the forefront among realists in claiming that information is power, has emphatically
stated that IOs are simply the agents of states. Neoliberals have tended to treat infor-
mation in a highly technocratic and depoliticized way, failing to see how information
is power.39 As IOs create transparencies and level information asymmetries among
states (a common policy prescription of neoliberals) they create new information
asymmetries between IOs and states. Given the neoliberal assumption that IOs have
no goals independent of states, such asymmetries are unimportant; but if IOs have
autonomous values and behavioral predispositions, then such asymmetries may be
highly consequential.
Examples of the ways in which IOs have become autonomous because of their
embodiment of technical rationality and control over information are not hard to find.
The UN’s peacekeepers derive part of their authority from the claim that they are
independent, objective, neutral actors who simply implement Security Council reso-
lutions. UN officials routinely use this language to describe their role and are explicit
that they understand this to be the basis of their influence. As a consequence, UN
officials spend considerable time and energy attempting to maintain the image that
they are not the instrument of any great power and must be seen as representatives of
‘‘the international community’’ as embodied in the rules and resolutions of the UN.40
The World Bank is widely recognized to have exercised power over development
policies far greater than its budget, as a percentage of North/South aid flows, would
suggest because of the expertise it houses. While competing sites of expertise in
development have proliferated in recent years, for decades after its founding the
World Bank was a magnet for the ‘‘best and brightest’’ among ‘‘development ex-
perts.’’ Its staff had and continues to have impressive credentials from the most pres-
37. See Weber [1930] 1978, 181–83; and Clegg 1994a, 152–55.
38. Gerth and Mills 1978, 216, 50, 299. For the extreme manifestation of this bureaucratic character-
istic, see Arendt 1977.
39. See Strange 1997; and Keohane 1984.
40. See David Rieff, ‘‘The Institution that Saw No Evil,’’ The New Republic, 12 February 1996, 19–24;
and Barnett 1997b.
tigious universities and the elaborate models, reports, and research groups it has
sponsored over the years were widely influential among the ‘‘development experts’’
in the field. This expertise, coupled with its claim to ‘‘neutrality’’ and its ‘‘apolitical’’
technocratic decision-making style, have given the World Bank an authoritative voice
with which it has successfully dictated the content, direction, and scope of global
development over the past fifty years.41 Similarly, official standing and long experi-
ence with relief efforts have endowed the UNHCR with ‘‘expert’’ status and conse-
quent authority in refugee matters. This expertise, coupled with its role in implement-
ing international refugee conventions and law (‘‘the rules’’ regarding refugees), has
allowed the UNHCR to make life and death decisions about refugees without consult-
ing the refugees, themselves, and to compromise the authority of states in various
ways in setting up refugee camps.42 Note that, as these examples show, technical
knowledge and expertise need not be ‘‘scientific’’ in nature to create autonomy and
power for IOs.
The Power of IOs

If IOs have autonomy and authority in the world, what do they do with it? A growing
body of research in sociology and anthropology has examined ways in which IOs
exercise power by virtue of their culturally constructed status as sites of authority; we
distill from this research three broad types of IO power. We examine how IOs (1)
classify the world, creating categories of actors and action; (2) fix meanings in the
social world; and (3) articulate and diffuse new norms, principles, and actors around
the globe. All of these sources of power flow from the ability of IOs to structure
knowledge.43
Classification. An elementary feature of bureaucracies is that they classify and

organize information and knowledge. This classification process is bound up with
power. ‘‘Bureaucracies,’’ writes Don Handelman, ‘‘are ways of making, ordering,
and knowing social worlds.’’ They do this by ‘‘moving persons among social catego-
ries or by inventing and applying such categories.’’44 The ability to classify objects,
to shift their very definition and identity, is one of bureaucracy’s greatest sources of
power. This power is frequently treated by the objects of that power as accomplished
through caprice and without regard to their circumstances but is legitimated and
justified by bureaucrats with reference to the rules and regulations of the bureau-
cracy. Consequences of this bureaucratic exercise of power may be identity defining,
or even life threatening.
Consider the evolving definition of ‘‘refugee.’’ The category ‘‘refugee’’ is not at all
straightforward and must be distinguished from other categories of individuals who
41. See Wade 1996; Ayres 1983; Ascher 1983; Finnemore 1996b; and Nelson 1995.
42. See Malkki 1996; Hartigan 1992; and Harrell-Bond 1989.
43. See Foucault 1977, 27; and Clegg 1994b, 156–59. International relations theory typically disre-
gards the negative side of the knowledge and power equation. For an example, see Haas 1992.
44. Handelman 1995, 280. See also Starr 1992; and Wright 1994, 22.
are ‘‘temporarily’’ and ‘‘involuntarily’’ living outside their country of origin—

displaced persons, exiles, economic migrants, guest workers, diaspora communities,
and those seeking political asylum. The debate over the meaning of ‘‘refugee’’ has
been waged in and around the UNHCR. The UNHCR’s legal and operational defini-
tion of the category strongly influences decisions about who is a refugee and shapes
UNHCR staff decisions in the field—decisions that have a tremendous effect on the
life circumstance of thousands of people.45 These categories are not only political
and legal but also discursive, shaping a view among UNHCR officials that refugees
must, by definition, be powerless, and that as powerless actors they do not have to be
consulted in decisions such as asylum and repatriation that will directly and dramati-
cally affect them.46 Guy Gran similarly describes how the World Bank sets up criteria
to define someone as a peasant in order to distinguish them from a farmer, day
laborer, and other categories. The classification matters because only certain classes
of people are recognized by the World Bank’s development machinery as having
knowledge that is relevant in solving development problems.47 Categorization and
classification are a ubiquitous feature of bureaucratization that has potentially impor-
tant implications for those being classified. To classify is to engage in an act of
power.
The fixing of meanings. IOs exercise power by virtue of their ability to fix mean-
ings, which is related to classification.48 Naming or labeling the social context estab-
lishes the parameters, the very boundaries, of acceptable action. Because actors are
oriented toward objects and objectives on the basis of the meaning that they have for
them, being able to invest situations with a particular meaning constitutes an impor-
tant source of power.49 IOs do not act alone in this regard, but their organizational
resources contribute mightily to this end.
There is strong evidence of this power from development studies. Arturo Escobar
explores how the institutionalization of the concept of ‘‘development’’ after World
War II spawned a huge international apparatus and how this apparatus has now spread
its tentacles in domestic and international politics through the discourse of develop-
ment. The discourse of development, created and arbitrated in large part by IOs,
determines not only what constitutes the activity (what development is) but also who
(or what) is considered powerful and privileged, that is, who gets to do the develop-
ing (usually the state or IOs) and who is the object of development (local groups).50
Similarly, the end of the Cold War encouraged a reexamination of the definition of
security.51 IOs have been at the forefront of this debate, arguing that security pertains
not only to states but also to individuals and that the threats to security may be
45. See Weiss and Pasic 1997; Goodwin-Gill 1996; and Anonymous 1997.
46. See Harrell-Bond 1989; Walkup 1997; and Malkki 1996.
47. Gran 1986.
48. See Williams 1996; Clegg 1994b; Bourdieu 1994; Carr [1939] 1964; and Keeley 1990.
49. Blumer 1969.
50. See Gupta 1998; Escobar 1995; Cooper and Packard 1998; Gran 1986; Ferguson 1990; and Wade 1996.
51. See Matthews 1989; and Krause and Williams 1996.
economic, environmental, and political as well as military.52 In forwarding these

alternative definitions of security, officials from various IOs are empowering a differ-
ent set of actors and legitimating an alternative set of practices. Specifically, when
security meant safety from invading national armies, it privileged state officials and
invested power in military establishments. These alternative definitions of security
shift attention away from states and toward the individuals who are frequently threat-
ened by their own government, away from military practices and toward other fea-
tures of social life that might represent a more immediate and daily danger to the
lives of individuals.
One consequence of these redefined meanings of development and security is that
they legitimate, and even require, increased levels of IO intervention in the domestic
affairs of states—particularly Third World states. This is fairly obvious in the realm
of development. The World Bank, the International Monetary Fund (IMF), and other
development institutions have established a web of interventions that affect nearly
every phase of the economy and polity in many Third World states. As ‘‘rural devel-
opment,’’ ‘‘basic human needs,’’ and ‘‘structural adjustment’’ became incorporated
into the meaning of development, IOs were permitted, even required, to become
intimately involved in the domestic workings of developing polities by posting in-
house ‘‘advisors’’ to run monetary policy, reorganizing the political economy of en-
tire rural regions, regulating family and reproductive practices, and mediating be-
tween governments and their citizens in a variety of ways.53
The consequences of redefining security may be similar. Democratization, human
rights, and the environment have all now become tied to international peace and
security, and IOs justify their interventions in member states on these grounds, par-
ticularly in developing states. For example, during the anti-apartheid struggle in South
Africa, human rights abuses came to be classified as security threats by the UN
Security Council and provided grounds for UN involvement there. Now, that linkage
between human rights and security has become a staple of the post–Cold War envi-
ronment. Widespread human rights abuses anywhere are now cause for UN interven-
tion, and, conversely, the UN cannot carry out peacekeeping missions without pro-
moting human rights.54 Similarly, environmental disasters in Eastern Europe and the
newly independent states of the former Soviet Union and water rights allocations in
the Middle East have also come to be discussed under the rubric of ‘‘environmental
security’’ and are thus grounds for IO intervention. The United Nations Development
Program argues that there is an important link between human security and sustain-
able development and implicitly argues for greater intervention in the management
of environment as a means to promote human security.55
Diffusion of norms. Having established rules and norms, IOs are eager to spread
the benefits of their expertise and often act as conveyor belts for the transmission of
52. See UN Development Program 1994; and Boutros-Ghali 1995.

53. See Escobar 1995; Ferguson 1990; and Feldstein 1998.
54. World Conference on Human Rights 1993.
55. UN Development Program 1994.
norms and models of ‘‘good’’ political behavior.56 There is nothing accidental or

unintended about this role. Officials in IOs often insist that part of their mission is to
spread, inculcate, and enforce global values and norms. They are the ‘‘missionaries’’
of our time. Armed with a notion of progress, an idea of how to create the better life,
and some understanding of the conversion process, many IO elites have as their
stated purpose a desire to shape state practices by establishing, articulating, and
transmitting norms that define what constitutes acceptable and legitimate state behav-
ior. To be sure, their success depends on more than their persuasive capacities, for
their rhetoric must be supported by power, sometimes (but not always) state power.
But to overlook how state power and organizational missionaries work in tandem and
the ways in which IO officials channel and shape states’ exercise of power is to
disregard a fundamental feature of value diffusion.57
Consider decolonization as an example. The UN Charter announced an intent to
universalize sovereignty as a constitutive principle of the society of states at a time
when over half the globe was under some kind of colonial rule; it also established an
institutional apparatus to achieve that end (most prominently the Trusteeship Council
and the Special Committee on Colonialism). These actions had several conse-
quences. One was to eliminate certain categories of acceptable action for powerful
states. Those states that attempted to retain their colonial privileges were increas-
ingly viewed as illegitimate by other states. Another consequence was to empower
international bureaucrats (at the Trusteeship Council) to set norms and standards for
‘‘stateness.’’ Finally, the UN helped to ensure that throughout decolonization the
sovereignty of these new states was coupled with territorial inviolability. Colonial
boundaries often divided ethnic and tribal groups, and the UN was quite concerned
that in the process of ‘‘self-determination,’’ these governments containing ‘‘mul-
tiple’’ or ‘‘partial’’ selves might attempt to create a whole personality through territo-
rial adjustment—a fear shared by many of these newly decolonized states. The UN
encouraged the acceptance of the norm of sovereignty-as-territorial-integrity through
resolutions, monitoring devices, commissions, and one famous peacekeeping epi-
sode in Congo in the 1960s.58
Note that, as with other IO powers, norm diffusion, too, has an expansionary dy-
namic. Developing states continue to be popular targets for norm diffusion by IOs,
even after they are independent. The UN and the European Union are now actively
involved in police training in non-Western states because they believe Western polic-
ing practices will be more conducive to democratization processes and the establish-
ment of civil society. But having a professional police establishment assumes that
there is a professional judiciary and penal system where criminals can be tried and
jailed; and a professional judiciary, in turn, presupposes that there are lawyers that
can come before the court. Trained lawyers presuppose a code of law. The result is a
package of reforms sponsored by IOs aimed at transforming non-Western societies
56. See Katzenstein 1996; Finnemore 1996b; and Legro 1997.

57. See Alger 1963, 425; and Claude 1966, 373.
58. See McNeely 1995; and Jackson 1993.
into Western societies.59 Again, while Western states are involved in these activities
and therefore their values and interests are part of the reasons for this process, inter-
national bureaucrats involved in these activities may not see themselves as doing the
bidding for these states but rather as expressing the interests and values of the bureau-
cracy.
Other examples of this kind of norm diffusion are not hard to find. The IMF and
the World Bank are explicit about their role as transmitters of norms and principles
from advanced market economies to less-developed economies.60 The IMF’s Articles
of Agreement specifically assign it this task of incorporating less-developed econo-
mies into the world economy, which turns out to mean teaching them how to ‘‘be’’
market economies. The World Bank, similarly, has a major role in arbitrating the
meaning of development and norms of behavior appropriate to the task of developing
oneself, as was discussed earlier. The end of the Cold War has opened up a whole
new set of states to this kind of norm diffusion task for IOs. According to former
Secretary of Defense William Perry, one of the functions of NATO expansion is to
inculcate ‘‘modern’’ values and norms into the Eastern European countries and their
militaries.61 The European Bank for Reconstruction and Development has, as part of
its mandate, the job of spreading democracy and private enterprise. The OSCE is
striving to create a community based on shared values, among these respect for
democracy and human rights. This linkage is also strong at the UN as evident in The
Agenda for Democratization and The Agenda for Peace.62 Once democratization and
human rights are tied to international peace and security, the distinctions between
international and domestic governance become effectively erased and IOs have li-
cense to intervene almost anywhere in an authoritative and legitimate manner.63
Realists and neoliberals may well look at these effects and argue that the classifi-
catory schemes, meanings, and norms associated with IOs are mostly favored by
strong states. Consequently, they would argue, the power we attribute to IOs is sim-
ply epiphenomenal of state power. This argument is certainly one theoretical possibil-
ity, but it is not the only one and must be tested against others. Our concern is that
because these theories provide no ontological independence for IOs, they have no
way to test for autonomy nor have they any theoretical cause or inclination to test for
it since, by theoretical axiom, autonomy cannot exist. The one empirical domain in
which the statist view has been explicitly challenged is the European Union, and
empirical studies there have hardly produced obvious victory for the ‘‘intergovern-
mentalist’’ approach.64 Recent empirical studies in the areas of human rights, weap-
ons taboos, and environmental practices also cast doubt on the statist approach by
providing evidence about the ways in which nongovernmental and intergovernmen-
59. Call and Barnett forthcoming.

60. Wade 1996.
61. See Perry 1996; and Ruggie 1996.
62. Boutros-Ghali 1995 and 1996a,b.
63. Keen and Hendrie, however, suggest that nongovernmental organizations and IOs can be the long-
term beneficiaries of intervention. See Keen 1994; and Hendrie 1997.
64. See Burley and Mattli 1993; Pollack 1997; and Sandholtz 1993.
tal organizations successfully promote policies that are not (or not initially) sup-
ported by strong states.65 Certainly there are occasions when strong states do drive IO
behavior, but there are also times when other forces are at work that eclipse or signifi-
cantly dampen the effects of states on IOs. Which causal mechanisms produce which
effects under which conditions is a set of relationships that can be understood only by
intensive empirical study of how these organizations actually do their business—
research that would trace the origins and evolution of IO policies, the processes by
which they are implemented, discrepancies between implementation and policy, and
overall effects of these policies.
The Pathologies of IOs
Bureaucracies are created, propagated, and valued in modern society because of their
supposed rationality and effectiveness in carrying out social tasks. These same con-
siderations presumably also apply to IOs. Ironically, though, the folk wisdom about
bureaucracies is that they are inefficient and unresponsive. Bureaucracies are infa-
mous for creating and implementing policies that defy rational logic, for acting in
ways that are at odds with their stated mission, and for refusing requests of and
turning their backs on those to whom they are officially responsible.66 Scholars of
U.S. bureaucracy have recognized this problem and have devoted considerable en-
ergy to understanding a wide range of undesirable and inefficient bureaucratic behav-
iors caused by bureaucratic capture and slack and to exploring the conditions under
which ‘‘suboptimal equilibria’’ may arise in organizational structures. Similarly, schol-
ars researching foreign policy decision making and, more recently, those interested
in learning in foreign policy have investigated organizational dynamics that produce
self-defeating and inefficient behavior in those contexts.67
IOs, too, are prone to dysfunctional behaviors, but international relations scholars
have rarely investigated this, in part, we suspect, because the theoretical apparatus
they use provides few grounds for expecting undesirable IO behavior.68 The state-
centric utility-maximizing frameworks most international relations scholars have bor-
rowed from economics simply assume that IOs are reasonably responsive to state
interests (or, at least, more responsive than alternatives), otherwise states would with-
draw from them. This assumption, however, is a necessary theoretical axiom of these
frameworks; it is rarely treated as a hypothesis subject to empirical investigation.69
With little theoretical reason to expect suboptimal or self-defeating behavior in IOs,
these scholars do not look for it and have had little to say about it. Policymakers,
however, have been quicker to perceive and address these problems and are putting
65. See Keck and Sikkink 1998; Wapner 1996; Price 1997; and Thomas forthcoming.
66. March and Olsen 1989, chap. 5.
67. See Nye 1987; Haas 1990; Haas and Haas 1995; and Sagan 1993.
68. Two exceptions are Gallaroti 1991; and Snidal 1996.
69. Snidal 1996.
FIGURE 1. Theories of international organization dysfunction
them on the political agenda. It is time for scholars, too, to begin to explore these
issues more fully.
In this section we present several bodies of theorizing that might explain dysfunc-
tional IO behavior, which we define as behavior that undermines the IO’s stated
objectives. Thus our vantage point for judging dysfunction (and later pathology) is
the publicly proclaimed mission of the organization. There may be occasions when
overall organizational dysfunction is, in fact, functional for certain members or oth-
ers involved in the IO’s work, but given our analysis of the way claims of efficiency
and effectiveness act to legitimate rational-legal authority in our culture, whether
organizations actually do what they claim and accomplish their missions is a particu-
larly important issue to examine. Several bodies of theory provide some basis for
understanding dysfunctional behavior by IOs, each of which emphasizes a different
locus of causality for such behavior. Analyzing these causes, we construct a typology
of these explanations that locates them in relation to one another. Then, drawing on
the work of James March and Johan Olsen, Paul DiMaggio and Walter Powell, and
other sociological institutionalists, we elaborate how the same sources of bureau-
cratic power, sketched earlier, can cause dysfunctional behavior. We term this particu-
lar type of dysfunction pathology.70 We identify five features of bureaucracy that
might produce pathology, and using examples from the UN system we illustrate the
way these might work in IOs.
Extant theories about dysfunction can be categorized in two dimensions: (1) whether
they locate the cause of IO dysfunction inside or outside the organization, and (2)
whether they trace the causes to material or cultural forces. Mapping theories on
these dimensions creates the typology shown in Figure 1.
Within each cell we have identified a representative body of theory familiar to
most international relations scholars. Explanations of IO dysfunction that emphasize
the pursuit of material interests within an organization typically examine how com-
petition among subunits over material resources leads the organization to make deci-
70. Karl Deutsch used the concept of pathology in a way similar to our usage. We thank Hayward Alker
for this point. Deutsch 1963, 170.
sions and engage in behaviors that are inefficient or undesirable as judged against
some ideal policy that would better allow the IO to achieve its stated goals. Bureau-
cratic politics is the best-known theory here, and though current scholars of interna-
tional politics have not widely adopted this perspective to explain IO behavior, it is
relatively well developed in the older IO literature.71 Graham Allison’s central argu-
ment is that the ‘‘name of the game is politics: bargaining along regularized circuits
among players positioned hierarchically within the government. Government behav-
ior can thus be understood as . . . results of these bargaining games.’’72 In this view,
decisions are not made after a rational decision process but rather through a competi-
tive bargaining process over turf, budgets, and staff that may benefit parts of the
organization at the expense of overall goals.
Another body of literature traces IO dysfunctional behavior to the material forces
located outside the organization. Realist and neoliberal theories might posit that state
preferences and constraints are responsible for understanding IO dysfunctional behav-
ior. In this view IOs are not to blame for bad outcomes, states are. IOs do not have the
luxury of choosing the optimal policy but rather are frequently forced to chose be-
tween the bad and the awful because more desirable policies are denied to them by
states who do not agree among themselves and/or do not wish to see the IO fulfill its
mandate in some particular instance. As Robert Keohane observed, IOs often engage
in policies not because they are strong and have autonomy but because they are weak
and have none.73 The important point of these theories is that they trace IO dysfunc-
tional behavior back to the environmental conditions established by, or the explicit
preferences of, states.
Cultural theories also have internal and external variants. We should note that
many advocates of cultural theories would reject the claim that an organization can
be understood apart from its environment or that culture is separable from the mate-
rial world. Instead they would stress how the organization is permeated by that envi-
ronment, defined in both material and cultural terms, in which it is embedded. Many
are also quite sensitive to the ways in which resource constraints and the material
power of important actors will shape organizational culture. That said, these argu-
ments clearly differ from the previous two types in their emphasis on ideational and
cultural factors and clearly differ among themselves in the motors of behavior empha-
sized. For analytical clarity we divide cultural theories according to whether they see
the primary causes of the IO’s dysfunctional behavior as deriving from the culture of
the organization (internal) or of the environment (external).
The world polity model exemplifies theories that look to external culture to under-
stand an IO’s dysfunctional behavior. There are two reasons to expect dysfunctional
behavior here. First, because IO practices reflect a search for symbolic legitimacy
rather than efficiency, IO behavior might be only remotely connected to the efficient
implementation of its goals and more closely coupled to legitimacy criteria that come
71. See Allison 1971; Haas 1990; Cox et al. 1974; and Cox and Jacobson 1977.
72. See Allison 1971, 144; and Bendor and Hammond 1992.
73. Personal communication to the authors.
from the cultural environment.74 For instance, many arms-export control regimes
now have a multilateral character not because of any evidence that this architecture is
the most efficient way to monitor and prevent arms exports but rather because multi-
lateralism has attained a degree of legitimacy that is not empirically connected to any
efficiency criteria.75 Second, the world polity is full of contradictions; for instance, a
liberal world polity has several defining principles, including market economics and
human equality, that might conflict at any one moment. Thus, environments are often
ambiguous about missions and contain varied, often conflicting, functional, norma-
tive, and legitimacy imperatives.76 Because they are embedded in that cultural envi-
ronment, IOs can mirror and reproduce those contradictions, which, in turn, can lead
to contradictory and ultimately dysfunctional behavior.
Finally, organizations frequently develop distinctive internal cultures that can pro-
mote dysfunctional behavior, behavior that we call ‘‘pathological.’’ The basic logic
of this argument flows directly from our previous observations about the nature of
bureaucracy as a social form. Bureaucracies are established as rationalized means to
accomplish collective goals and to spread particular values. To do this, bureaucracies
create social knowledge and develop expertise as they act upon the world (and thus
exercise power). But the way bureaucracies are constituted to accomplish these ends
can, ironically, create a cultural disposition toward undesirable and ultimately self-
defeating behavior.77 Two features of the modern bureaucratic form are particularly
important in this regard. The first is the simple fact that bureaucracies are organized
around rules, routines, and standard operating procedures designed to trigger a stan-
dard and predictable response to environmental stimuli. These rules can be formal or
informal, but in either case they tell actors which action is appropriate in response to
a specific stimuli, request, or demand. This kind of routinization is, after all, precisely
what bureaucracies are supposed to exhibit—it is what makes them effective and
competent in performing complex social tasks. However, the presence of such rules
also compromises the extent to which means-ends rationality drives organizational
behavior. Rules and routines may come to obscure overall missions and larger social
goals. They may create ‘‘ritualized behavior’’ in bureaucrats and construct a very
parochial normative environment within the organization whose connection to the
larger social environment is tenuous at best.78
Second, bureaucracies specialize and compartmentalize. They create a division of
labor on the logic that because individuals have only so much time, knowledge, and
expertise, specialization will allow the organization to emulate a rational decision-
making process.79 Again, this is one of the virtues of bureaucracy in that it provides a
way of overcoming the limitations of individual rationality and knowledge by embed-
ding those individuals in a structure that takes advantage of their competencies with-
74. See Meyer and Rowan 1977; Meyer and Zucker 1989; Weber 1994; and Finnemore 1996a.
75. Lipson 1999.
76. McNeely 1995.
77. See Vaughan 1996; and Lipartito 1995.
78. See March and Olsen 1989, 21–27; and Meyer and Rowan 1977.
79. See March and Olsen 1989, 26–27; and March 1997.
out having to rely on their weaknesses. However, it, too, has some negative conse-
quences. Just as rules can eclipse goals, concentrated expertise and specialization can
(and perhaps must) limit bureaucrats’ field of vision and create subcultures within
bureaucracy that are distinct from those of the larger environment. Professional train-
ing plays a particularly strong role here since this is one widespread way we dissemi-
nate specialized knowledge and credential ‘‘experts.’’ Such training often gives ex-
perts, indeed is designed to give them, a distinctive worldview and normative
commitments, which, when concentrated in a subunit of an organization, can have
pronounced effects on behavior.80
Once in place, an organization’s culture, understood as the rules, rituals, and be-
liefs that are embedded in the organization (and its subunits), has important conse-
quences for the way individuals who inhabit that organization make sense of the
world. It provides interpretive frames that individuals use to generate meaning.81
This is more than just bounded rationality; in this view, actors’ rationality itself, the
very means and ends that they value, are shaped by the organizational culture.82
Divisions and subunits within the organization may develop their own cognitive
frameworks that are consistent with but still distinct from the larger organization,
further complicating this process.
All organizations have their own culture (or cultures) that shape their behavior.
The effects of bureaucratic culture, however, need not be dysfunctional. Indeed, spe-
cific organizational cultures may be valued and actively promoted as a source of
‘‘good’’ behavior, as students of business culture know very well. Organizational
culture is tied to ‘‘good’’ and ‘‘bad’’ behavior, alike, and the effects of organizational
culture on behavior are an empirical question to be researched.
To further such research, we draw from studies in sociology and anthropology to
explore five mechanisms by which bureaucratic culture can breed pathologies in IOs:
the irrationality of rationalization, universalism, normalization of deviance, organiza-
tional insulation, and cultural contestation. The first three of these mechanisms all
flow from defining features of bureaucracy itself. Consequently, we expect them to
be present in any bureaucracy to a limited degree. Their severity may be increased,
however, by specific empirical conditions of the organization. Vague mission, weak
feedback from the environment, and strong professionalism all have the potential to
exacerbate these mechanisms and to create two others, organizational insulation and
cultural contestation, through processes we describe later. Our claim, therefore, is
that the very nature of bureaucracy—the ‘‘social stuff’’ of which it is made—creates
behavioral predispositions that make bureaucracy prone to these kinds of behav-
iors.83 But the connection between these mechanisms and pathological behavior is
probabilistic, not deterministic, and is consistent with our constitutive analysis.
Whether, in fact, mission-defeating behavior occurs depends on empirical condi-
80. See DiMaggio and Powell 1983; and Schien 1996.

81. See Starr 1992, 160; Douglas 1986; and Berger and Luckman 1966, chap. 1.
82. See Campbell 1998, 378; Alvesson 1996; Burrell and Morgan 1979; Dobbin 1994; and Immergut
1998, 14–19.
83. Wendt 1998.
tions. We identify three such conditions that are particularly important (mission,
feedback, and professionals) and discuss how they intensify these inherent predispo-
sitions and activate or create additional ones.
Irrationality of rationalization. Weber recognized that the ‘‘rationalization’’ pro-

cesses at which bureaucracies excelled could be taken to extremes and ultimately
become irrational if the rules and procedures that enabled bureaucracies to do their
jobs became ends in themselves. Rather than designing the most appropriate and
efficient rules and procedures to accomplish their missions, bureaucracies often tailor
their missions to fit the existing, well-known, and comfortable rulebook.84 Thus,
means (rules and procedures) may become so embedded and powerful that they
determine ends and the way the organization defines its goals. One observer of the
World Bank noted how, at an operational level, the bank did not decide on develop-
ment goals and collect data necessary to pursue them. Rather, it continued to use
existing data-collection procedures and formulated goals and development plans from
those data alone.85 UN-mandated elections may be another instance where means
become ends in themselves. The ‘‘end’’ pursued in the many troubled states where
the UN has been involved in reconstruction is presumably some kind of peaceful,
stable, just government. Toward that end, the UN has developed a repertoire of instru-
ments and responses that are largely intended to promote something akin to a demo-
cratic government. Among those various repertoires, elections have become privi-
leged as a measure of ‘‘success’’ and a signal of an operation’s successful conclusion.
Consequently, UN (and other IO) officials have conducted elections even when evi-
dence suggests that such elections are either premature or perhaps even counterpro-
ductive (frequently acknowledged as much by state and UN officials).86 In places like
Bosnia elections have ratified precisely the outcome the UN and outside powers had
intervened to prevent—ethnic cleansing—and in places like Africa elections are criti-
cized as exacerbating the very ethnic tensions they were ostensibly designed to quell.
UN peacekeeping might also provide examples. As the UN began to involve itself
in various ‘‘second-generation operations’’ that entailed the management and recon-
ciliation of domestic conflicts it turned to the only instrument that was readily avail-
able in sufficient numbers—peacekeeping units. Peacekeepers, however, are military
troops, trained to handle interstate conflict and to be interposed between two contend-
ing national armies, operating with their consent. Some UN staff, state officials, and
peacekeeping scholars worried that peacekeepers might be inappropriate for the de-
mands of handling domestic security. They feared that peacekeepers would transfer
the skills and attitudes that had been honed for one environment to another without
fully considering the adjustments required. According to some observers, peacekeep-
ers did just that: they carried their interstate conflict equipment and mindset into new
situations and so created a more aggressive and offensively minded posture than
84. Beetham 1985, 76.

85. See Ferguson 1990; and Nelson 1995.
86. Paris 1997.
would otherwise have been the case. The result was operations that undermined the
objectives of the mandate.87
Bureaucratic universalism. A second source of pathology in IOs derives from the

fact that bureaucracies ‘‘orchestrate numerous local contexts at once.’’88 Bureaucrats
necessarily flatten diversity because they are supposed to generate universal rules
and categories that are, by design, inattentive to contextual and particularistic con-
cerns. Part of the justification for this, of course, is the bureaucratic view that techni-
cal knowledge is transferable across circumstances. Sometimes this is a good assump-
tion, but not always; when particular circumstances are not appropriate to the
generalized knowledge being applied, the results can be disastrous.89
Many critics of the IMF’s handling of the Asian financial crises have argued that
the IMF inappropriately applied a standardized formula of budget cuts plus high
interest rates to combat rapid currency depreciation without appreciating the unique
and local causes of this depreciation. These governments were not profligate spend-
ers, and austerity policies did little to reassure investors, yet the IMF prescribed
roughly the same remedy that it had in Latin America. The result, by the IMF’s later
admission, was to make matters worse.90
Similarly, many of those who worked in peacekeeping operations in Cambodia
were transferred to peacekeeping operations in Bosnia or Somalia on the assumption
that the knowledge gained in one location would be applicable to others. Although
some technical skills can be transferred across contexts, not all knowledge and orga-
nizational lessons derived from one context are appropriate elsewhere. The UN has a
longstanding commitment to neutrality, which operationally translates into the view
that the UN should avoid the use of force and the appearance of partiality. This
knowledge was employed with some success by UN envoy Yasushi Akashi in Cam-
bodia. After his stint in Cambodia, he became the UN Special Representative in
Yugoslavia. As many critics of Akashi have argued, however, his commitment to
these rules, combined with his failure to recognize that Bosnia was substantially
different from Cambodia, led him to fail to use force to defend the safe havens when
it was appropriate and likely to be effective.91
Normalization of deviance. We derive a third type of pathology from Diane

Vaughan’s study of the space shuttle Challenger disaster in which she chronicles the
way exceptions to rules (deviance) over time become routinized and normal parts of
procedures.92 Bureaucracies establish rules to provide a predictable response to envi-
ronmental stimuli in ways that safeguard against decisions that might lead to acci-
dents and faulty decisions. At times, however, bureaucracies make small, calculated
87. See Featherston 1995; Chopra, Eknes, and Nordbo 1995; and Hirsch and Oakley 1995, chap. 6.
88. Heyman 1995, 262.
89. Haas 1990, chap. 3.
90. See Feldstein 1998; Radelet and Sachs 1999; and Kapur 1998.
91. Rieff 1996.
92. Vaughan 1996.
deviations from established rules because of new environmental or institutional de-

velopments, explicitly calculating that bending the rules in this instance does not
create excessive risk of policy failure. Over time, these exceptions can become the
rule—they become normal, not exceptions at all: they can become institutionalized
to the point where deviance is ‘‘normalized.’’ The result of this process is that what at
time t1 might be weighed seriously and debated as a potentially unacceptable risk or
dangerous procedure comes to be treated as normal at time tn. Indeed, because of
staff turnover, those making decisions at a later point in time might be unaware that
the now-routine behavior was ever viewed as risky or dangerous.
We are unaware of any studies that have examined this normalization of deviance
in IO decision making, though one example of deviance normalization comes to
mind. Before 1980 the UNHCR viewed repatriation as only one of three durable
solutions to refugee crises (the others being third-country asylum and host-country
integration). In its view, repatriation had to be both safe and voluntary because forced
repatriation violates the international legal principle of nonrefoulement, which is the
cornerstone of international refugee law and codified in the UNHCR’s convention.
Prior to 1980, UNHCR’s discussions of repatriation emphasized that the principles
of safety and voluntariness must be safeguarded at all costs. According to many
commentators, however, the UNHCR has steadily lowered the barriers to repatria-
tion over the years. Evidence for this can be found in international protection manu-
als, the UNHCR Executive Committee resolutions, and discourse that now weighs repatria-
tion and the principle of nonrefoulement against other goals such a peace building. This
was a steady and incremental development as initial deviations from organizational
norms accumulated over time and led to a normalization of deviance. The result was
a lowering of the barriers to repatriation and an increase in the frequency of involun-
tary repatriation.93
Insulation. Organizations vary greatly in the degree to which they receive and
process feedback from their environment about performance. Those insulated from
such feedback often develop internal cultures and worldviews that do not promote
the goals and expectations of those outside the organization who created it and whom
it serves. These distinctive worldviews can create the conditions for pathological
behavior when parochial classification and categorization schemes come to define
reality—how bureaucrats understand the world—such that they routinely ignore in-
formation that is essential to the accomplishment of their goals.94
Two causes of insulation seem particularly applicable to IOs. The first is profes-
sionalism. Professional training does more than impart technical knowledge. It ac-
tively seeks to shape the normative orientation and worldviews of those who are
trained. Doctors are trained to value life above all else, soldiers are trained to sacri-
93. See Chimni 1993, 447; Amnesty International 1997a,b; Human Rights Watch 1997; Zieck 1997,
433, 434, 438–39; and Barbara Crossette, ‘‘The Shield for Exiles Is Lowered,’’ The New York Times, 22
December 1996, 4-1.
94. See Berger and Luckman 1967, chap. 1; Douglas 1986; Bruner 1990; March and Olsen 1989; and
Starr 1992.
fice life for certain strategic objectives, and economists are trained to value effi-
ciency. Bureaucracies, by their nature, concentrate professionals inside organiza-
tions, and concentrations of people with the same expertise or professional training
can create an organizational worldview distinct from the larger environment. Sec-
ond, organizations for whom ‘‘successful performance’’ is difficult to measure—that
is, they are valued for what they represent rather than for what they do and do not
‘‘compete’’ with other organizations on the basis of output—are protected from selec-
tion and performance pressures that economistic models simply assume will operate.
The absence of a competitive environment that selects out inefficient practices coupled
with already existing tendencies toward institutionalization of rules and procedures
insulates the organization from feedback and increases the likelihood of pathologies.
IOs vary greatly in the degree to which the professionals they recruit have distinc-
tive worldviews and the degree to which they face competitive pressures, but it is
clearly the case that these factors insulate some IOs to some degree and in so doing
create a tendency toward pathology. The World Bank, for example, has been domi-
nated for much of its history by economists, which, at least in part, has contributed to
many critiques of the bank’s policies. In one such critique James Ferguson opens his
study of the World Bank’s activity in Lesotho by comparing the bank’s introductory
description of Lesotho in its report on that country to facts on the ground; he shows
how the bank ‘‘creates’’ a world that has little resemblance to what historians, geog-
raphers, or demographers see on the ground in Lesotho but is uniquely suited to the
bank’s organizational abilities and presents precisely the problems the bank knows
how to solve. This is not simply ‘‘staggeringly bad scholarship,’’ Ferguson argues,
but a way of making the world intelligible and meaningful from a particular perspec-
tive—the World Bank’s.95 The problem, however, is that this different worldview
translates into a record of development failures, which Ferguson explores in detail.
Insulation contributes to and is caused by another well-known feature of organiza-
tions—the absence of effective feedback loops that allow the organization to evalu-
ate its efforts and use new information to correct established routines. This is surely a
‘‘rational’’ procedure in any social task but is one that many organizations, including
IOs, fail to perform.96 Many scholars and journalists, and even the current head of the
World Bank, have noticed that the bank has accumulated a rather distinctive record
of ‘‘failures’’ but continues to operate with the same criteria and has shown a marked
lack of interest in evaluating the effectiveness of its own projects.97 The same is true
of other IOs. Jarat Chopra observes that the lessons-learned conferences that were
established after Somalia were structurally arranged so that no information could
come out that would blemish the UN’s record. Such attempts at face saving, Chopra
cautions, make it more likely that these maladies will go uncorrected.98 Sometimes
95. Ferguson 1990, 25–73.

96. March and Olsen 1989, chap. 5; Haas 1990.
97. See Wade 1996, 14–17; Nelson 1995, chaps. 6, 7; and Richard Stevenson, ‘‘The Chief Banker for
the Nations at the Bottom of the Heap,’’ New York Times, 14 September 1997, sec. 3, 1, 12–14.
98. Chopra 1996.
new evaluative criteria are hoisted in order to demonstrate that the failures were not
really failures but successes.
Cultural contestation. Organizational coherence is an accomplishment rather than

a given. Organizational control within a putative hierarchy is always incomplete,
creating pockets of autonomy and political battles within the bureaucracy.99 This is
partly a product of the fact that bureaucracies are organized around the principle of
division-of-labor, and different divisions tend to be staffed by individuals who are
‘‘experts’’ in their assigned tasks. These different divisions may battle over budgets
or material resources and so follow the bureaucratic politics model, but they may
also clash because of distinct internal cultures that grow up inside different parts of
the organization. Different segments of the organization may develop different ways
of making sense of the world, experience different local environments, and receive
different stimuli from outside; they may also be populated by different mixes of
professions or shaped by different historical experiences. All of these would contrib-
ute to the development of different local cultures within the organization and differ-
ent ways of perceiving the environment and the organization’s overall mission. Orga-
nizations may try to minimize complications from these divisions by arranging these
demands hierarchically, but to the extent that hierarchy resolves conflict by squelch-
ing input from some subunits in favor of others, the organization loses the benefits of
a division of labor that it was supposed to provide. More commonly, though, at-
tempts to reconcile competing worldviews hierarchically are simply incomplete. Most
organizations develop overlapping and contradictory sets of preferences among sub-
groups.100 Consequently, different constituencies representing different normative
views will suggest different tasks and goals for the organization, resulting in a clash
of competing perspectives that generates pathological tendencies.
The existence of cultural contestation might be particularly true of high-profile
and expansive IOs like the UN that have vague missions, broad and politicized con-
stituencies, and lots of divisions that are developed over time and in response to new
environmental demands. Arguably a number of the more spectacular debacles in
recent UN peacekeeping operations might be interpreted as the product of these
contradictions.
Consider the conflict between the UN’s humanitarian missions and the value it
places on impartiality and neutrality. Within the organization there are many who
view impartiality as a core constitutive principle of UN action. On the one hand, the
UN’s moral standing, its authority, and its ability to persuade all rest on this principle.
On the other hand, the principles of humanitarianism require the UN to give aid to
those in need—values that are particularly strong in a number of UN relief and
humanitarian agencies. These two norms of neutrality and humanitarian assistance,
and the parts of the bureaucracy most devoted to them, come into direct conflict in
those situations where providing humanitarian relief might jeopardize the UN’s
99. See Clegg 1994a, 30; Vaughan 1996, 64; and Martin 1992.
100. Haas 1990, 188.
vaunted principle of neutrality. Bosnia is the classic case in point. On the one hand,
the ‘‘all necessary means’’ provision of Security Council resolutions gave the UN
authority to deliver humanitarian aid and protect civilians in the safe havens. On the
other hand, the UN abstained from ‘‘taking sides’’ because of the fear that such
actions would compromise its neutrality and future effectiveness. The result of these
conflicts was a string of contradictory policies that failed to provide adequately for
the UN’s expanding humanitarian charges.101 According to Shashi Tharoor, a UN
official intimately involved in these decisions, ‘‘It is extremely difficult to make war
and peace with the same people on the same territory at the same time.’’102
UNHCR provides another possible example of cultural contestation. Historically,
the UNHCR’s Protection Division has articulated a legalistic approach toward refu-
gee matters and thus tends to view the UNHCR and itself as the refugee’s lawyer and
as the protector of refugee rights under international law. Those that inhabit the
UNHCR’s regional bureaus, however, have been characterized as taking a less ‘‘nar-
row’’ view of the organization’s mission, stressing that the UNHCR must take into
account the causes of refugee flows and state pressures. These cultural conflicts have
been particularly evident, according to many observers, when the UNHCR contem-
plates a repatriation exercise in areas of political instability and conflict: protection
officers demand that the refugees’ rights, including the right of nonrefoulement, be
safeguarded, whereas the regional bureaus are more willing to undertake a risky
repatriation exercise if it might serve broader organizational goals, such as satisfying
the interests of member states, and regional goals, such as facilitating a peace agree-
ment.103
Although bureaucratic culture is not the only source of IO dysfunction, it is a

potentially powerful one that creates broad patterns of behavior that should interest
international relations scholars. None of the sources of pathologies sketched here is
likely to appear in isolation in any empirical domain. These processes interact and
feed on each other in ways that will require further theorizing and research. More-
over, while we have highlighted the organization’s internal characteristics, we must
always bear in mind that the external environment presses upon and shapes the inter-
nal characteristics of the organization in a host of ways. Cultural contestation within
an organization frequently originates from and remains linked to normative contra-
dictions in the larger environment. Demands from states can be extremely important
determinants of IO behavior and may override internal cultural dynamics, but they
can also set them in place if conflicting state demands result in the creation of organi-
zational structures or missions that are prone to pathology. As we begin to explore
dysfunctional and pathological behavior, we must bear in mind the complex relation-
ship between different causal pathways, remaining closely attentive to both the inter-
nal organizational dynamics and the IO’s environment.
101. See Barnett 1997a; David Rieff, ‘‘We Hate You,’’ New Yorker, 4 September 1995, 41–48; David
Rieff, ‘‘The Institution That Saw No Evil,’’ The New Republic, 12 February 1996, 19–24; and Rieff 1996.
102. Quoted in Weiss 1996, 85; also see Rieff 1996, 166, 170, 193.
103. See Kennedy 1986; and Lawyers Committee for Human Rights 1991.
Conclusion
For all the attention international relations scholars have paid to international institu-
tions over the past several decades, we know very little about the internal workings
of IOs or about the effects they have in the world. Our ignorance, we suspect, is in
large part a product of the theoretical lens we have applied. From an economistic
perspective, the theoretically interesting question to ask about IOs is why they are
created in the first place. Economists want to know why we have firms; political
scientists want to know why we have IOs. In both cases, the question flows naturally
from first theoretical principles. If you think that the world looks like a microeco-
nomic market—anarchy, firms (or states) competing to maximize their utilities—
what is anomalous and therefore theoretically interesting is cooperation. Conse-
quently, our research tends to focus on the bargains states strike to make or reshape
IOs. Scholars pay very little attention to what goes on subsequently in their day-to-
day operations or even the larger effects that they might have on the world.
Viewing IOs through a constructivist or sociological lens, as we suggest here,
reveals features of IO behavior that should concern international relations scholars
because they bear on debates central to our field—debates about whether and how
international institutions matter and debates about the adequacy of a statist ontology
in an era of globalization and political change. Three implications of this alternative
approach are particularly important. First, this approach provides a basis for treating
IOs as purposive actors. Mainstream approaches in political science that are in-
formed by economic theories have tended to locate agency in the states that comprise
IO membership and treat IOs as mere arenas in which states pursue their policies. By
exploring the normative support for bureaucratic authority in the broader interna-
tional culture and the way IOs use that authority to construct the social world, we
provide reasons why IOs may have autonomy from state members and why it may
make sense analytically to treat them as ontologically independent. Second, by pro-
viding a basis for that autonomy we also open up the possibility that IOs are powerful
actors who can have independent effects on the world. We have suggested various
ways to think about how IOs are powerful actors in global politics, all of which
encourage greater consideration of how IOs affect not only discrete outcomes but
also the constitutive basis of global politics.
Third, this approach also draws attention to normative evaluations of IOs and
questions what appears to us to be rather uncritical optimism about IO behavior.
Contemporary international relations scholars have been quick to recognize the posi-
tive contributions that IOs can make, and we, too, are similarly impressed. But for all
their desirable qualities, bureaucracies can also be inefficient, ineffective, repressive,
and unaccountable. International relations scholars, however, have shown little inter-
est in investigating these less savory and more distressing effects. The liberal Wilso-
nian tradition tends to see IOs as promoters of peace, engines of progress, and agents
for emancipation. Neoliberals have focused on the impressive way in which IOs help
states to overcome collective action problems and achieve durable cooperation. Real-
ists have focused on their role as stabilizing forces in world politics. Constructivists,
too, have tended to focus on the more humane and other-regarding features of IOs,
but there is nothing about social construction that necessitates ‘‘good’’ outcomes. We
do not mean to imply that IOs are ‘‘bad’’; we mean only to point out theoretical
reasons why undesirable behavior may occur and suggest that normative evaluation
of IO behavior should be an empirical and ethical matter, not an analytic assumption.
References
Alger, Chadwick. 1963. United Nations Participation as a Learning Process. Public Opinion Quarterly 27
(3):411–26.
Allison, Graham. 1971. Essence of Decision. Boston: Little, Brown.
Alvesson, Mats. 1996. Cultural Perspectives on Organizations. New York: Cambridge University Press.
Amnesty International. 1997a. In Search of Safety: The Forcibly Displaced and Human Rights in Africa.
AI Index, 20 June, AFR 01/05/97. Available from ⬍www.amnesty.org/ailib/aipub/1997/10100597.htm⬎.
———. 1997b. Rwanda: Human Rights Overlooked in Mass Repatriation. Available from ⬍www.
amnesty.org/ailib/aipub/1997/AFR/147002797.htm⬎.
Anonymous. 1997. The UNHCR Note on International Protection You Won’t See. International Journal
of Refugee Law 9 (2):267–73.
Arendt, Hannah. 1977. Eichmann in Jerusalem: A Report on the Banality of Evil. New York: Penguin.
Ascher, William. 1983. New Development Approaches and the Adaptability of International Agencies:
The Case of the World Bank. International Organization 37 (3):415–39.
Ayres, Robert L. 1983. Banking on the Poor: The World Bank and World Poverty. Cambridge, Mass.: MIT
Press.
Baldwin, David, ed. 1993. Neorealism and Neoliberalism. New York: Columbia University Press.
Barnett, Michael. 1997a. The Politics of Indifference at the United Nations and Genocide in Rwanda and
Bosnia. In This Time We Knew: Western Responses to Genocide in Bosnia, edited by Thomas Cushman
and Stjepan Mestrovic, 128–62. New York: New York University Press.
———. 1997b. The UN Security Council, Indifference, and Genocide in Rwanda. Cultural Anthropology
12 (4):551–78.
Beetham, David. 1985. Max Weber and the Theory of Modern Politics. New York: Polity.
———. 1996. Bureaucracy. 2d ed. Minneapolis: University of Minnesota Press.
Bendor, Jonathan, and Thomas Hammond. 1992. Rethinking Allison’s Models. American Political Sci-
ence Review 82 (2):301–22.
Berger, Peter, and Thomas Luckmann. 1966. The Social Construction of Reality. New York: Doubleday.
Blumer, Herbert. 1969. Symbolic Interactionism: Perspective and Method. Englewood Cliffs, N.J.: Prentice-
Hall.
Bourdieu, Pierre. 1994. On Symbolic Power. In Language and Symbolic Power, edited by Pierre Bour-
dieu, 163–70. Chicago: University of Chicago Press.
Boutros-Ghali, Boutros. 1995. Agenda for Peace. 2d ed. New York: UN Press.
———. 1996a. Global Leadership After the Cold War. Foreign Affairs 75:86–98.
———. 1996b. Agenda for Democratization. New York: UN Press.
Bruner, Jerome. 1990. Acts of Meaning. Cambridge, Mass.: Harvard University Press.
Burley, Anne-Marie, and Walter Mattli. 1993. Europe Before the Court: A Political Theory of Integration.
International Organization 47 (1):41–76.
Burrell, Gibson, and Gareth Morgan. 1979. Sociological Paradigms and Organizational Analysis. Lon-
don: Heinemman.
Call, Chuck, and Michael Barnett. Forthcoming. Looking for a Few Good Cops: Peacekeeping, Peace-
building, and U.N. Civilian Police. International Peacekeeping.
Campbell, John. 1998. Institutional Analysis and the Role of Ideas in Political Economy. Theory and
Society 27:377–409.
Carr, Edward H. [1939] 1964. The Twenty Year’s Crisis. New York: Harper Torchbooks.
Chimni, B. 1993. The Meaning of Words and the Role of UNHCR in Voluntary Repatriation. International
Journal of Refugee Law 5 (3):442–60.
Chopra, Jarat. 1996. Fighting for Truth at the UN. Crosslines Global Report, 26 November, 7–9.
Chopra, Jarat, Age Eknes, and Toralv Nordbo. 1995. Fighting for Hope in Somalia. Oslo: NUPI.
Claude, Inis L., Jr. 1966. Collective Legitimization as a Political Function of the United Nations. Interna-
tional Organization 20 (3):337–67.
Clegg, Stewart. 1994a. Power and Institutions in the Theory of Organizations. In Toward a New Theory of
Organizations, edited by John Hassard and Martin Parker, 24–49. New York: Routledge.
———. 1994b. Weber and Foucault: Social Theory for the Study of Organizations. Organization 1 (1):
149–78.
Coase, Ronald. 1937. The Nature of the Firm. Economica 4 (November):386–405.
Commission on Global Governance. 1995. Our Global Neighborhood. New York: Oxford University
Press.
Cooper, Frederick, and Randy Packard, eds. 1998. International Development and the Social Sciences.
Berkeley: University of California Press.
Cox, Robert. 1980. The Crisis of World Order and the Problem of International Organization in the 1980s.
International Journal 35 (2):370–95.
———. 1992. Multilateralism and World Order. Review of International Studies 18 (2):161–80.
———. 1996. The Executive Head: An Essay on Leadership in International Organization. In Approaches
to World Order, edited by Robert Cox, 317–48. New York: Cambridge University Press.
Cox, Robert, and Harold Jacobson. 1977. Decision Making. International Social Science Journal 29
(1):115–33.
Cox, Robert, Harold Jacobson, Gerard Curzon, Victoria Curzon, Joseph Nye, Lawrence Scheinman, James
Sewell, and Susan Strange. 1974. The Anatomy of Influence: Decision Making in International Organi-
zation. New Haven, Conn.: Yale University Press.
Deutsch, Karl. 1963. The Nerves of Government: Models of Political Communication and Control. Glencoe,
Ill.: Free Press.
Dillon, Patricia, Thomas Ilgen, and Thomas Willett. 1991. Approaches to the Study of International Orga-
nizations: Major Paradigms in Economics and Political Science. In The Political Economy of Interna-
tional Organizations: A Public Choice Approach, edited by Ronald Vaubel and Thomas Willett, 79–99.
Boulder, Colo.: Westview Press.
DiMaggio, Paul J., and Walter W. Powell. 1983. The Iron Cage Revisited: Institutional Isomorphism and
Collective Rationality in Organizational Fields. American Sociological Review 48:147–60.
Dobbin, Frank. 1994. Cultural Models of Organization: The Social Construction of Rational Organizing
Principles. In The Sociology of Culture, edited by Diana Crane, 117–42. Boston: Basil Blackwell.
Douglas, Mary. 1986. How Institutions Think. Syracuse, N.Y.: Syracuse University Press.
Doyle, Michael. 1997. Ways of War and Peace. New York: Norton.
Escobar, Arturo. 1995. Encountering Development: The Making and Unmaking of the Third World. Prince-
ton, N.J.: Princeton University Press.
Featherston, A. B. 1995. Habitus in Cooperating for Peace: A Critique of Peacekeeping. In The New
Agenda for Global Security: Cooperating for Peace and Beyond, edited by Stephanie Lawson, 101–18.
St. Leonards, Australia: Unwin and Hyman.
Feld, Werner J., and Robert S. Jordan, with Leon Hurwitz. 1988. International Organizations: A Compara-
tive Approach. 2d ed. New York: Praeger.
Feldstein, Martin. 1998. Refocusing the IMF. Foreign Affairs 77 (2):20–33.
Ferguson, James. 1990. The Anti-Politics Machine: ‘‘Development,’’ Depoliticization, and Bureaucratic
Domination in Lesotho. New York: Cambridge University Press.
Finnemore, Martha. 1993. International Organizations as Teachers of Norms: The United Nations Educa-
tional, Scientific, and Cultural Organization and Science Policy. International Organization 47: 565–97.
———. 1996a. Norms, Culture, and World Politics: Insights from Sociology’s Institutionalism. Interna-
———. 1996b. National Interests in International Society. Ithaca, N.Y.: Cornell University Press.
Fisher, William. 1997. Doing Good? The Politics and Antipolitics of NGO Practices. Annual Review of
Anthropology 26:439–64.
Foucault, Michel. 1977. Discipline and Punish. New York: Vintage Press.
Gallaroti, Guilio. 1991. The Limits of International Organization. International Organization 45 (2):183–
220.
Gerth, H. H., and C. Wright Mills. 1978. From Max Weber: Essays in Sociology. New York: Oxford
University Press.
Goodwin-Gill, Guy. 1996. Refugee in International Law. New York: Oxford Clarendon.
Gran, Guy. 1986. Beyond African Famines: Whose Knowledge Matters? Alternatives 11:275–96.
Grandori, Anna. 1993. Notes on the Use of Power and Efficiency Constructs in the Economics and Soci-
ology of Organizations. In Interdisciplinary Perspectives on Organizational Studies, edited by S. Lin-
denberg and H. Schreuder, 61–78. New York: Pergamon
Gupta, Akhil. 1998. Postcolonial Developments: Agriculture in the Making of Modern India. Durham,
N.C.: Duke University Press.
Haas, Ernst. 1990. When Knowledge Is Power. Berkeley: University of California Press.
Haas, Ernst, and Peter Haas. 1995. Learning to Learn: Improving International Governance. Global Gov-
ernance 1 (3):255–85.
Haas, Peter, ed. 1992. Epistemic Communities. International Organization 46 (1). Special issue.
Handelman, Don. 1995. Comment. Current Anthropology 36 (2):280–81.
Harrell-Bond, Barbara. 1989. Repatriation: Under What Conditions Is It the Most Desirable Solution for
Refugees? African Studies Review 32 (1):41–69.
Hartigan, Kevin. 1992. Matching Humanitarian Norms with Cold, Hard Interests: The Making of Refugee
Policies in Mexico and Honduras, 1980–89. International Organization 46:709–30.
Helms, Jesse. 1996. Saving the UN. Foreign Affairs 75 (5):2–7.
Hendrie, Barbara. 1997. Knowledge and Power: A Critique of an International Relief Operation. Disasters
21 (1):57–76.
Heyman, Josiah McC. 1995. Putting Power in the Anthropology of Bureaucracy. Current Anthropology 36
(2):261–77.
Hirsch, John, and Robert Oakley. 1995. Somalia and Operation Restore Hope: Reflections on Peace-
making and Peacekeeping. Washington, D.C.: USIP Press.
Human Rights Watch. 1997. Uncertain Refuge: International Failures to Protect Refugees. Vol. 1, no. 9
(April).
Hurrell, Andrew, and Ngaire Woods. 1995. Globalisation and Inequality. Millennium 24 (3):447–70.
Immergut, Ellen. 1998. The Theoretical Core of the New Institutionalism. Politics and Society 26 (1):
5–34.
Jackson, Robert. 1993. The Weight of Ideas in Decolonization: Normative Change in International Rela-
tions. In Ideas and Foreign Policy, edited by Judith Goldstein and Robert O. Keohane, 111–38. Ithaca,
N.Y.: Cornell University Press.
Jacobson, Harold. 1979. Networks of Interdependence. New York: Alfred A. Knopf.
Joint Evaluation of Emergency Assistance to Rwanda. 1996. The International Response to Conflict and
Genocide: Lessons from the Rwanda Experience. 5 vols. Copenhagen: Steering Committee of the Joint
Evaluation of Emergency Assistance to Rwanda.
Kapur, Devesh. 1998. The IMF: A Cure or a Curse? Foreign Policy 111:114–29.
Katzenstein, Peter J., ed. 1996. The Culture of National Security: Identity and Norms in World Politics.
New York: Columbia University Press.
Keck, Margaret, and Kathryn Sikkink. 1998. Activists Beyond Borders. Ithaca, N.Y.: Cornell University
Press.
Keeley, James. 1990. Toward a Foucauldian Analysis of International Regimes. International Organiza-
tion 44 (1):83–105.
Keen, David. 1994. The Benefits of Famine: A Political Economy of Famine and Relief in Southwestern
Sudan, 1983–89. Princeton, N.J.: Princeton University Press
Kennedy, David. 1986. International Refugee Protection. Human Rights Quarterly 8:1–9.
Keohane, Robert O. 1984. After Hegemony. Princeton, N.J.: Princeton University Press.
Kiewiet, D. Roderick, and Matthew McCubbins. 1991. The Logic of Delegation. Chicago: University of
Chicago Press.
Krasner, Stephen D. 1991. Global Communications and National Power: Life on the Pareto Frontier.
World Politics 43 (3):336–66.
———. 1983a. Regimes and the Limits of Realism: Regimes as Autonomous Variables. In International
Regimes, edited by Stephen Krasner, 355–68. Ithaca, N.Y.: Cornell University Press.
Krasner, Stephen D., ed. 1983b. International Regimes. Ithaca, N.Y.: Cornell University Press.
Krause, Keith, and Michael Williams. 1996. Broadening the Agenda of Security Studies: Politics and
Methods. Mershon International Studies Review 40 (2):229–54.
Lake, David. 1996. Anarchy, Hierarchy, and the Variety of International Relations. International Organi-
zation 50 (1):1–34.
Lawyers Committee for Human Rights. 1991. General Principles Relating to the Promotion of Refugee
Repatriation. New York: Lawyers Committee for Human Rights.
Legro, Jeffrey. 1997. Which Norms Matter? Revisiting the ‘‘Failure’’ of Internationalism. International
Organization 51 (1):31–64.
Lipartito, Kenneth. 1995. Culture and the Practice of Business History. Business and Economic History
24 (2):1–41.
Lipson, Michael. 1999. International Cooperation on Export Controls: Nonproliferation, Globalization,
and Multilateralism. Ph.D. diss., University of Wisconsin, Madison.
Malkki, Liisa. 1996. Speechless Emissaries: Refugees, Humanitarianism, and Dehistoricization. Cultural
Anthropology 11 (3):377–404.
March, James. 1988. Decisions and Organizations. Boston: Basil Blackwell.
———. 1997. Understanding How Decisions Happen in Organizations. In Organizational Decision Making,
edited by Z. Shapira, 9–33. New York: Cambridge University Press.
March, James, and Johan P. Olsen. 1989. Rediscovering Institutions: The Organizational Basis of Politics.
New York: Free Press.
Martin, Joan. 1992. Cultures in Organizations: Three Perspectives. New York: Oxford University Press.
Matthews, Jessica Tuchman. 1989. Redefining Security. Foreign Affairs 68 (2):162–77.
McNeely, Connie. 1995. Constructing the Nation-State: International Organization and Prescriptive Ac-
tion. Westport, Conn.: Greenwood Press.
Mearsheimer, John. 1994. The False Promise of International Institutions. International Security 19 (3):
5–49.
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., and W. Richard Scott. 1992. Organizational Environments: Ritual and Rationality. New-
bury Park, Calif.: Sage.
Meyer, Marshall, and Lynne Zucker. 1989. Permanently Failing Organizations. Newbury Park: Sage
Press.
Miller, Gary, and Terry M. Moe. 1983. Bureaucrats, Legislators, and the Size of Government. American
Political Science Review 77 (June):297–322.
Moe, Terry M. 1984. The New Economics of Organization. American Journal of Political Science 28:
739–77.
Moravcsik, Andrew. 1997. Taking Preferences Seriously: Liberal Theory and International Politics. Inter-
national Organization 51 (4):513–54.
———. 1999. A New Statecraft? Supranational Entrepreneurs and International Cooperation. Interna-
Mouzelis, Nicos. 1967. Organization and Bureaucracy. Chicago: Aldine.
Murphy, Craig. 1994. International Organizations and Industrial Change. New York: Oxford University
Press.
Nelson, Paul. 1995. The World Bank and Non-Governmental Organizations. New York: St. Martin’s Press.
Ness, Gayl, and Steven Brechin. 1988. Bridging the Gap: International Organizations as Organizations.
International Organization 42 (2):245–73.
Niskanen, William A. 1971. Bureaucracy and Representative Government. Chicago: Aldine.
Paris, Roland. 1997. Peacebuilding and the Limits of Liberal Internationalism. International Security 22
(2):54–89.
Perry, William. 1996. Defense in an Age of Hope. Foreign Affairs 75 (6):64–79.
Pollack, Mark. 1997. Delegation, Agency, and Agenda-Setting in the European Community. International
Organization 51 (1):99–134.
Powell, Walter W., and Paul J. DiMaggio, eds. 1991. The New Institutionalism in Organizational Analysis.
Chicago: University of Chicago Press.
Pratt, John, and Richard J. Zeckhauser. 1985. Principals and Agents: The Structure of Business. Boston:
Harvard Business School Press.
Price, Richard. 1997. The Chemical Weapons Taboo. Ithaca, N.Y.: Cornell University Press.
Radelet, Steven, and Jeffrey Sach. 1999. What Have We Learned, So Far, From the Asian Financial Crisis?
Harvard Institute for International Development, 4 January. Available from ⬍www.hiid.harvard.edu/
pub/other/aea122.pdf⬎.
Rieff, David. 1996. Slaughterhouse. New York: Simon and Schuster.
Rittberger, Volker, ed. 1993. Regime Theory and International Relations. Oxford: Clarendon Press.
Ross, George. 1995. Jacques Delors and European Integration. New York: Oxford University Press.
Ruggie, John. 1996. Winning the Peace. New York: Columbia University Press.
———. 1998. What Makes the World Hang Together. International Organization 52 (3):855–86.
Sandholtz, Wayne. 1993. Choosing Union: Monetary Politics and Maastricht. International Organization
47:1–40.
Sagan, Scott. 1993. The Limits of Safety: Organizations, Accidents, and Nuclear Weapons. Princeton, N.J.:
Princeton University Press.
Schaar, John. 1984. Legitimacy in the Modern State. In Legitimacy and the State, edited by William
Connolly, 104–33. Oxford: Basil Blackwell.
Schien, Edgar. 1996. Culture: The Missing Concept in Organization Studies. Administrative Studies Quar-
terly 41:229–40.
Scott, W. Richard. 1992. Organizations: Rational, Natural, and Open Systems. 3d ed. Englewood Cliffs,
N.J.: Prentice-Hall.
Shanks, Cheryl, Harold K. Jacobson, and Jeffrey H. Kaplan. 1996. Inertia and Change in the Constellation
of Intergovernmental Organizations, 1981–1992. International Organization 50 (4):593–627.
Shapira, Zur, ed. 1997. Organizational Decision. New York: Cambridge University Press.
Shore, Cris, and Susan Wright. 1997. Policy: A New Field of Anthropology. In Anthropology of Policy:
Critical Perspectives on Governance and Power, edited by Cris Shore and Susan Wright, 3–41. New
York: Routledge Press.
Sigelman, Lee. 1986. The Bureaucratic Budget Maximizer: An Assumption Examined. Public Budgeting
and Finance (spring):50–59.
Snidal, Duncan. 1996. Political Economy and International Institutions. International Review of Law and
Economics 16:121–37.
Starr, Paul. 1992. Social Categories and Claims in the Liberal State. In How Classification Works: Nelson
Goodman Among the Social Sciences, edited by Mary Douglas and David Hull, 154–79. Edinburgh:
Edinburgh University Press.
Strange, Susan. 1997. The Retreat of the State. New York: Cambridge University Press.
Thomas, Daniel C. Forthcoming. The Helsinki Effect: International Norms, Human Rights, and Demise of
Communism. Princeton, N.J.: Princeton University Press.
UN Development Program. 1994. Human Development Report 1994. New York: Oxford University Press.
UN Peacekeeping Missions. 1994. The Lessons from Cambodia. Asia Pacific Issues, Analysis from the
East-West Center, No. 11, March.
Vaubel, Roland. 1991. A Public Choice View of International Organization. In The Political Economy of
International Organizations, edited by Roland Vaubel and Thomas Willett, 27–45. Boulder, Colo.:
Westview Press.
Vaughan, Diane. 1996. The Challenger Launch Decision. Chicago: University of Chicago Press.
Wade, Robert. 1996. Japan, the World Bank, and the Art of Paradigm Maintenance: The East Asian
Miracle in Political Perspective. New Left Review 217:3–36.
Walkup, Mark. 1997. Policy Dysfunction in Humanitarian Organizations: The Role of Coping Strategies,
Institutions, and Organizational Culture. Journal of Refugee Studies 10 (1):37–60.
Waltz, Kenneth. 1979. Theory of International Politics. Reading, Mass.: Addison-Wesley.
Wapner, Paul. 1996. Environmental Activism and World Civic Politics. Albany: State University of New
York Press.
Weber, Max. 1947. Theory of Social and Economic Organization. New York: Oxford University Press.
———. [1930] 1968. The Protestant Ethic and the Spirit of Capitalism. New York: Routledge.
———. 1978. Bureaucracy. In From Max Weber: Essays in Sociology, edited by H. H. Gerth and C. Wright
Mills, 196–44. New York: Oxford
Weber, Steven. Origins of the European Bank for Reconstruction and Development. International Organi-
zation 48(1):1–38.
Weingast, Barry R., and Mark Moran. 1983. Bureaucratic Discretion or Congressional Control: Regula-
tory Policymaking by the Federal Trade Commission. Journal of Political Economy 91 (October):765–
800.
Weiss, Thomas. 1996. Collective Spinelessness: U.N. Actions in the Former Yugoslavia. In The World and
Yugoslavia’s Wars, edited by Richard Ullman, 59–96. New York: Council on Foreign Relations Press.
Weiss, Tom, and Amir Pasic. 1997. Reinventing UNHCR: Enterprising Humanitarians in the Former
Yugoslavia, 1991–95. Global Governance 3 (1):41–58.
Wendt, Alexander. 1995. Constructing International Politics. International Security 20 (1):71–81.
———. 1998. Constitution and Causation in International Relations. Review of International Studies 24
(4):101–17. Special issue.
Williams, Michael. 1996. Hobbes and International Relations: A Reconsideration. International Organiza-
tion 50 (2):213–37.
Williamson, Oliver. 1975. Markets and Hierarchies, Analysis and Antitrust Implications: A Study in the
Economics of Internal Organization. New York: Free Press.
———. 1985. The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. New
York: Free Press.
World Conference on Human Rights. 1993. Vienna Declaration on Human Rights, adopted 14–25 June.
UN Document A/CONF.157/24 (Part I), 20; A/CONF.157/23 DPI/1394/Rev.1 DPI/1676 (95.I.21), 448;
DPI/1707 ST/HR/2/Rev.4, 383.
Wright, Susan. 1994. ‘‘Culture’’ in Anthropology and Organizational Studies. In Anthropology of Organi-
zations, edited by Susan Wright, 1–31. New York: Routledge.
Zabusky, Stacia. 1995. Launching Europe. Princeton, N.J.: Princeton University Press.
Zieck, Marjoleine. 1997. UNHCR and Voluntary Repatriation of Refugees: A Legal Analysis. The Hague:
Martinus Nijhoff.

Power and Its Pathologies

Uploaded by

Copyright:

Available Formats

Power and Its Pathologies

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Power and Its Pathologies

Uploaded by

Copyright:

Available Formats

The Politics, Power, and Pathologies

International Organization 53, 4, Autumn 1999, pp. 699–732

7. See Mearsheimer 1994; and Helms 1996.

Theoretical Approaches to Organizations

IOs and their environment. The environment assumed by economic approaches

Empirically, organizational environments can take many forms. Some organiza-

23. Moravcsik 1997.

The Power of IOs

Sources of IO Autonomy and Authority

33. Gerth and Mills 1978, 299 (italics in original).

The Power of IOs

Classification. An elementary feature of bureaucracies is that they classify and

are ‘‘temporarily’’ and ‘‘involuntarily’’ living outside their country of origin—

economic, environmental, and political as well as military.52 In forwarding these

52. See UN Development Program 1994; and Boutros-Ghali 1995.

norms and models of ‘‘good’’ political behavior.56 There is nothing accidental or

56. See Katzenstein 1996; Finnemore 1996b; and Legro 1997.

59. Call and Barnett forthcoming.

The Pathologies of IOs

FIGURE 1. Theories of international organization dysfunction

80. See DiMaggio and Powell 1983; and Schien 1996.

Irrationality of rationalization. Weber recognized that the ‘‘rationalization’’ pro-

84. Beetham 1985, 76.

Bureaucratic universalism. A second source of pathology in IOs derives from the

Normalization of deviance. We derive a third type of pathology from Diane

deviations from established rules because of new environmental or institutional de-

95. Ferguson 1990, 25–73.

Cultural contestation. Organizational coherence is an accomplishment rather than

Although bureaucratic culture is not the only source of IO dysfunction, it is a

You might also like