Review: Approaches to the State: Alternative Conceptions and Historical Dynamics

Reviewed Work(s): On the Autonomy of the Democratic State by Eric Nordlinger:

Negara: The Theatre State in Nineteenth Century Bali by Clifford Geertz: Building a
New American State: The Expansion of National Administrative Capacities by Stephen
Skowronek: The Formation of National States in Western Europe by Charles Tilly:
Crises of Political Development in Europe and the United States by Raymond Grew:
Revolution from Above: Military Bureaucrats and Development in Japan, Turkey,
Egypt, and Peru by Ellen Kay Trimberger
Review by: Stephen D. Krasner
Source: Comparative Politics , Jan., 1984, Vol. 16, No. 2 (Jan., 1984), pp. 223-246
Published by: Comparative Politics, Ph.D. Programs in Political Science, City University
of New York

Stable URL: https://www.jstor.org/stable/421608

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Comparative Politics, Ph.D. Programs in Political Science, City University of New York is
collaborating with JSTOR to digitize, preserve and extend access to Comparative Politics

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Review Article

Approaches to the State

Alternative Conceptions and Historical Dynamics

Stephen D. Krasner

Eric Nordlinger, On the Autonomy of the Democratic State, Cambridge, Harvard

University Press, 1981.

Clifford Geertz, Negara: The Theatre State in Nineteenth Century Bali, Princeton,
Princeton University Press, 1981.

Stephen Skowronek, Building a New American State: The Expansion of National

Administrative Capacities, New York, Cambridge University Press, 1982.

Charles Tilly, ed., The Formation of National States in Western Europe, Princeton,
Princeton University Press, 1975.

Raymond Grew, ed., Crises of Political Development in Europe and the United
States, Princeton, Princeton University Press, 1978.

Ellen Kay Trimberger, Revolution from Above: Military Bureaucrats and Develop-
ment in Japan, Turkey, Egypt, and Peru, New Brunswick, Transaction Books,
1978.

From the late 1950s until the mid-1970s the term state virtually disappeared from
the professional academic lexicon. Political scientists wrote about government,
political development, interest groups, voting, legislative behavior, leadership, and
bureaucratic politics, almost everything but "the state." However, in the last
decade "the state" has reappeared in the literature. Marxist scholars have made a
self-conscious, theoretically grounded effort to develop a theory of the capitalist
state. In Between Power and Plenty, an edited volume about the foreign economic
policies of advanced industrial countries, Peter Katzenstein developed a typology of
weak and strong states. Theda Skocpol's States and Social Revolution examined
the political conditions, both international and domestic, associated with major
social revolutions. Alfred Stepan's The State and Society in Peru investigated both
the organic statist intellectual tradition and corporatist political structures that gave
"'the state" a major initiative role in the Peruvian and other Latin American political
systems. The state is central to all of the studies reviewed here.'
However, to note that terms have changed, that certain scholars have self-

223

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

consciously adopted a new vocabulary (or readopted an old one), does not neces-
sarily imply that there has been a change in substance. The purpose of this essay is
to examine the ways in which several recent books explicitly concerned with "the
state" differ from and challenge prevailing intellectual approaches that emerged out
of the behavioral revolution of the 1950s.
Recent literature on the state has been concerned with two central issues: the
extent of state autonomy and the degree of congruity between the state and its
environment. The issue of autonomy has generally been cast in a temporarily static
framework in which the state is viewed as an exogenous variable. The central issue
is: can the state formulate and implement its preferences? The issue of congruity
has been placed in a temporally dynamic framework in which the state is viewed as
an intervening variable. The central issue is: how do institutional structures change
in response to alterations in domestic and international environments and then in
subsequent time periods influence these environments?
Cutting across both of these analytic concerns is the prior question of how the
state should be defined. In a recent essay Roger Benjamin and Raymond Duval
argue that the following conceptualizations have appeared in the literature.

1. The state as government," by which is meant the collective set of personnel who
occupy positions of decisional authority in the polity."
2. The state as "public bureaucracy or administrative apparatus as a coherent totality"
and as an institutionalized legal order.
3. The state as ruling class.
4. The state as normative order.2

The state as ruling class is, in one variant or another, the Marxist definition, an
will not be further considered here. The dominant conceptualization in the non
Marxist literature is the state as a bureaucratic apparatus and institutionalized le
order in its totality. The final phrase is critical, for it distinguishes statist ori
entations from the bureaucratic politics approaches which have parcelled the st
into little pieces, pieces that can be individually analyzed (where you stand depen
on where you sit) and that float in a permissive environment (policies are a prod
of bargaining and compromise among bureaus).3 Statist arguments have emphasiz
the overall structure of the bureaucratic apparatus, in particular the degree of
centralization of power at the national level and the extent of state power vis-a-
the society. Among the books reviewed in this essay, the two exceptions to the
generalization that the state is seen as a bureaucratic apparatus and institutionali
legal order taken as a totality are Eric Nordlinger's On the Autonomy of the
Democratic State, which adopts the state as government conceptualization, and
Clifford Geertz's Negara, which views the state as a normative order.
There are five characteristics of the recent statist literature that distinguish it
from orientations associated with the behavioral revolution. First, statist ap-
proaches see politics more as a problem of rule and control than as one of alloca-
tion; they are more concerned with issues associated with preserving order against
internal and external threats than with the distribution of utiles to political actors.

224

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

Politics is not just about "who gets what, when, how;" it is a struggle of us against
them.4
Second, statist approaches emphasize that the state can be treated as an actor in
its own right as either an exogenous or an intervening variable. Whether in its
institutional form or in terms of specific policies, the state cannot be understood as
a reflection of societal characteristics or preferences.
Third, statist orientations place greater emphasis on institutional constraints, both
formal and informal, on individual behavior. This is especially true for authors who
view the state as the bureaucratic apparatus and legal order taken as a totality or as
the normative political order.5 Actors in the political system, whether individuals or
groups, are bound within these structures, which limit, even determine, their
conceptions of their own interest and their political resources. Political outcomes
cannot be adequately understood as simply the resolution of a vector of forces
emanating from a variety of different groups.
Fourth, statist analyses have been more anxious to take what Gabriel Almond has
called the "historical cure."6 It is necessary to understand both how institutions
reproduce themselves through time and what historical conditions gave rise to them
in the first place. Current institutional structures may be a product of some peculiar
historical conjuncture rather than contemporaneous factors. Moreover, once an
historical choice is made, it both precludes and facilitates alternative future choices.
Political change follows a branching model. Once a particular fork is chosen, it is
very difficult to get back on a rejected path. Thus, the kinds of causal arguments
appropriate for periods of crises when institutions are first created may not be
appropriate for other periods.
Fifth, statist arguments are more inclined to see disjunctures and stress within
any given political system. Systems are not composed of interrelated and compati-
ble components. Structures do not exist because they perform certain functions,
and functions do not necessarily give rise to corresponding structures. Rather,
political life is fraught with tensions and conflicts, especially for the state. For
instance, international pressures frequently lead the state to attempt to increase the
level of resource extraction from its own society. But these efforts can engender
negative reactions from social groups who see their economic utility, and even their
sense of justice, undermined by new state policies. Political life is characterized,
not simply by a struggle over the allocation of resources, but also periodically by
strife and uncertainty about the rules of the game within which this allocative
process is carried out.
These five characteristics do not constitute a coherent theory of the state. Only
structural Marxists could credibly make such a claim, and even they are plagued by
deep and probably insoluble difficulties related to the degree of autonomy that can
be accorded to the state before fundamental tenets concerning the determining
character of economic structures are compromised. The studies under review here
do not set out to present a general theory of the state. Eric Nordlinger's basic
objective is to demonstrate that even in democratic polities public officials can
autonomously determine public policy. Clifford Geertz is concerned with the sym-

225

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

bolic attributes of the state as a unifying element for the entire social community.
Stephen Skowronek investigates the ways in which the functional political chal-
lenges posed by nineteenth-century industrialization were met, or not met, in the
context of the fragmented and localized political system that existed in the United
States. Tilly, Trimberger, and some of the authors in Grew emphasize the impact of
external threats on state-building.
Despite their diversity these studies do pose a challenge to the analytic traditions
that have dominated political science in the United States. They see a different
political universe, ask different questions, investigate different empirical phenom-
ena, and offer different kinds of answers.
The second section of this essay reviews Robert Dahl's theory of leadership in his
seminal study, Who Governs, to provide a clearer contrast between pluralist and
statist orientations.' The third section deals with problems of public policy in which
the central issue has been the degree of autonomy of the state. The fourth section
deals with problems of state-building in which the central issue has been the degree
of conformity or congruence of the state with its environment.8 The concluding
section suggests that a model taken from recent evolutionary theory, punctuated
equilibrium, can serve as an appropriate metaphor for understanding changes in the
relationship between states and their environments.

A Contrast with Pluralist Theories of Leadership

One way to illuminate the distinctive characteristics of statist approaches is to

contrast them with the pluralist approach. This is not only because pluralism is the
prevailing model for the study of politics in the United States, but also because
some major authors identified with the pluralist school have been explicitly con-
cerned with the role of public actors. By examining how their studies, in particular,
differ from the books reviewed here, one can clarify the distinct features of a statist
orientation.
The more obvious general differences can be noted first. Pluralism emphasizes
problems of allocation rather than ones of rule and control. Nordlinger points out
that as "portrayed by pluralism, civil society is made up of a plethora of diverse,
fluctuating, competing groups of individuals with shared interests. Many effective

political
their own,resources are available
autonomously to self-interests.
defined them ... ."9 These groups struggle
Cross-cutting to maximize
cleavages and broad
consensus on the rules of the game guarantee moderate political behavior.
Within the literature on American politics the central debate has not been about
the relative power of societal and state actors, but rather about which societal
actors most influence public policy. Conventional pluralists see a very wide array of
interest groups, virtually all of which have some political resources. Neo-pluralists
such as McConnell, Lowi, and especially Lindblom see a more constricted uni-
verse. Only a limited range of groups, among which business is particularly promi-
nent, possess significant political resources.1' But this debate takes place within the

226

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

confines of an intellectual universe that understands politics to be a problem, both

analytically and normatively, of allocation."
Pure interest group versions of pluralism virtually ignore public actors and in-
stitutions. The government is seen as a cash register that totals up and then
averages the preferences and political power of societal actors. Government may
also be seen as an arena within which societal actors struggle to insure the success
of their own particular preferences. The major function of public officials is to make
sure that the game is played fairly. If public institutions are viewed as figurative
cash registers or as literal referees, there is no room for anything that could be
designated as a state as actor with autonomous preferences capable of manipulating
and even restructuring its own society.
However, most pluralist models are much more explicitly concerned with the role
of political leaders and public officials. Robert Dahl's Who Governs offers an
example. The simple answer to the question "Who governs?" is that everyone does,
including political leaders. If there is a hero in Dahl's book it is Richard Lee, a
working-class boy without a college education who made good not only as an
official of Yale University but also as a dynamic and active mayor of New Haven.
Dahl argues that Lee's preferences were critical especially for urban renewal, a
program that came to dominate the city's public life. Lee was motivated initially by
his own concern with conditions in New Haven, not by pressure from any particu-
lar societal group. In fact, there were no groups particularly interested in urban
renewal one way or the other. The concluding sentence of Who Governs says that
there are "complex processes of symbiosis and change that constitute the relations
of leaders and citizens in a pluralistic democracy."'2
What distinguishes Dahl's treatment, a seminal study of pluralist politics, from
the more explicitly and self-consciously state-oriented works of the kind under
consideration in this essay? The most critical difference is that Dahl views the state
as a collection of individuals occupying particular roles, not as an administrative
apparatus or legal order. Benjamin and Duvall have referred to this formulation as
"the state as the government." They note that this "concept of the state is compati-
ble with extreme liberal tenets of the behavioral revolution, in treating the state as
merely a collection of individuals who occupy role positions (those of governing
authority) and who act as a group to govern."'3 Institutional imperatives and
constraints, including general political beliefs, do not play a significant role in
Dahl's formulation. The pluralist universe is atomistic. Different societal groups and
political leaders have different objectives and political resources. Public policy is
the outcome of the resolution of vectors resulting from interests and resources.
Formal, authoritative institutions are of little importance. There is no detailed
discussion in Who Governs of New Haven's governmental structure or of state and
national institutions. The defining characteristics of New Haven's political system
are, for Dahl, "universal suffrage, a moderately high participation in elections, a
highly competitive two-party system, opportunity to criticize the conduct and
policies of officials, freedom to seek support for one's views among officials and
citizens, and surprisingly frequent alternations in office from one party to the

227

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

other . 1. ."14 Dahl points out that at the very moment when Lee was being
reelected by an unprecedented majority the same voters were rejecting proposed
revisions of the city charter that would have invested the office of the mayor with
more power.
Dahl also disputes Tocqueville's proposition that American republican be
embodied in the legal structure are an important factor in American poli
behavior.'5 While there is widespread endorsement of general democratic values
the citizenry and the political stratum, these values are too vague, Dahl argue
have much impact on actual political affairs. There is only limited agreement on
specific application of general principles. Even general consensus is only m
tained by a complex and difficult process involving childhood socialization, pa
ticularly by the schools, and a "recurring process of interchange among polit
professionals, the political stratum, and the great bulk of the population.
process generates enough agreement on rules and norms so as to permit the sys
to operate, but agreement tends to be incomplete, and typically it decays" (itali
original).16
Statist orientations take institutions and political beliefs more seriously. The
political universe is not atomistic. Atoms are bound within stable molecules and
compounds. The preferences of public officials are constrained by the administra-
tive apparatus, legal order, and enduring beliefs. There are only a limited number of
ways in which political actors can combine their resources. The nature of political
resources is itself defined by institutional structures. The ability of a political leader
to carry out a policy is critically determined by the authoritative institutional
resources and arrangements existing within a given political system. Industrial
policy can be orchestrated in Japan through the ministry of international trade and
industry. There is no American institutional structure that would allow a political
leader, regardless of the resources he commanded, to implement a similar set of
policies. Moreover, at least some statist arguments have emphasized the im-
portance of ideology, not simply as an instrument of governance, but as an end in
itself. Ideology may not only coordinate expectations and delineate legitimate
modes of interaction between state institutions and societal actors, but it may also
serve as a basic source of identity, and its preservation may be a consummatory
function of the state. One of the critical purposes of the state is to represent
symbolically the existence and unity of the political community.
A second distinction between sophisticated pluralist and statist views is that
pluralism does not sharply differentiate public actors from their own society. Dahl's
unit of analysis most closely related to the state is the political stratum, composed
of subleaders and leaders. Members of the political stratum possess more political
resources than ordinary citizens. Some of these resources are drawn from the
public arena, including legality, but most derive from the society, including control
over jobs and information. Individuals carry many of these resources with them
both in and out of public office. Many important members of the political stratum
never hold public office at all. Statist perspectives contrast individuals in and out of
public office. Political leadership is closely related to official position. The adminis-
trative apparatus and legal order constrain preferences and provide means of influ-

228

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

ence. Political leaders are state actors pursuing either particular state goals or
collective societal objectives and utilizing resources primarily derived from their
official positions.
A third distinction between pluralist and statist approaches is that pluralist
theories of leadership see public officials as relatively more constrained by societal
pressures. Dahl argues that the support of the business community as well as other
societal elements was a necessary condition for the urban renewal program. Lee
was able to alter the societal forces he confronted by activating slack or latent
resources. However, Lee is not seen as being able to change the underlying
distribution of political resources. The ability of leaders to alter the preferences of
citizens depends upon the extent of agreement among leaders, and this in turn
depends in large part on the degree to which they are drawn from the same "social
strata.""7 Only a coherent leadership cadre can change the desires of citizens. In
contrast, statist orientations see political leaders as less constrained by societal
forces. They can alter preferences using the state's own resources. They may even
be able to change the distribution of political resources possessed by societal
groups.
In sum, the pluralist tradition in America has not simply ignored the role of
political leaders. However, it has seen these leaders as being substantially con
strained by societal forces, commanding resources that are derived from a wide
variety of public and private sources, and functioning in a fluid, institutio
environment which has a limited impact on the power and interests of actors.
These characteristics are shared by most of the major theoretical perspectives
that blossomed in the 1960s, including studies of mass behavior, political develo
ment, and bureaucratic politics, as well as pluralism. The general rejection of "t
state" as a meaningful analytic concept by all of these modes of analysis was no
coincidental. They were part of a larger intellectual change-the behavioral revol
tion. Behaviorism was a reaction against formal legalism, the approach that had
dominated the discipline of political science in the United States from its inceptio
during the last part of the nineteenth century through the 1940s. Formal legalism
virtually identified political life with the state, understood as an institution th
promulgated binding laws and stood in a superior hierarchical position to other
parts of the polity. Formal rules were seen as independent variables. The statemen
of rules was "treated as tantamount to the explanation for behavior."'
Behaviorism rejected the identity between rules and behavior. Empirical studi
did not demonstrate a close relationship between formal rules and political activit
Rules did not necessarily lead to regularities; and regularities existed without rule
In a world that included polities as different as the United States, the Soviet Unio
and Upper Volta, behaviorism suggested that it was impossible to understand muc
about actual political life simply by studying formal legal institutions.19
As a corollary to rejecting formal legalism behaviorist approaches focused on th
society. They reversed the causal relationship that had been posited by form
legalism. Societal forces were viewed as the independent variable. Political ou
comes were determined primarily by the preferences and power capabilities of
societal actors. Furthermore, the introduction of statistical methods and the com-

229

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

puter made it much more attractive to collect data on many variables from a large
number of cases. Such an empirical strategy was peculiarly compatible with a
pluralistic view of the political universe as heterogeneous and atomistic. (Obviously
statistical methodology and theoretical perspectives are not linked in any rigid way.
But computers and statistics did facilitate the acceptance of pluralist arguments by
making it easier to publish and conduct research.)20 Hence pluralism is part of a
more general intellectual orientation that has dominated American political science
for the last twenty years. The new concern with the state must be seen, in the first
instance, as a reaction against prevailing fashion.
Two central concerns have informed this new literature. The first involves the
autonomy of the state, its ability to formulate and implement public policy. The
second involves the extent of congruity between the state and its environment, a
central issue for the study of political development. These issues are addressed in
the next two sections of this essay.

Public Policy: The State as Exogenous Variable

Both Nordlinger and Geertz are concerned with public policy broadly defined, that
is, with authoritative actions taken by public institutions. Nordlinger investigates
the relationship between state and private actors in modern industrialized democra-
cies, arguing that if the autonomy of the state can be demonstrated in this political
setting it should hold in others as well. Nordlinger notes that this is the case most
likely to disprove assertions that public officials can formulate and authoritatively
implement their preferences. Geertz guides his readers into a much more exotic
political environment, nineteenth-century Bali, to show that symbolic activities,
which have been largely ignored by western political theory for several hundred
years, can be the consummatory end of public life and the central attribute of the
state.

Nordlinger's basic objective is to demonstrate that "the preferences of the stat

are at least as important as those of civil society in accounting for what the
democratic state does and does not do; the democratic state is not only frequent
autonomous insofar as it regularly acts upon its preferences, but also markedly
autonomous in doing so even when its preferences diverge from the demands of th
most powerful groups in civil society."21 The preferences of the state, whic
Nordlinger defines as "all those individuals who occupy offices that authorize them
and them alone, to make and apply decisions that are binding upon any and a
segments of society,"22 may be generated within the state itself or from the wid
society. The critical point with regard to state autonomy is that these preferences
not simply be designed to curry favor or avoid punishment from particular societ
groups. The final measure of autonomy is the ability of state officials to translate
their weighted preferences into authoritative actions.
The heart of On the Autonomy of the Democratic State is three chapters de-
lineating state autonomy under different relationships between the preferences o

230

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

state and societal actors. Type III state autonomy refers to situations in which there
is non-divergence between the preferences of the state and the society. Nordlinger
argues that even under these conditions state-oriented accounts can explain au-
thoritative actions at least as well as society-oriented ones. The state may initiate
policy and provide access for particular societal groups. The state can reinforce a
weak level of convergence by manipulating information, inflating the success of
ongoing programs, setting agendas, appealing to widely shared symbols, playing
upon deference to official expertise, and deflecting potential opposition. In all
Nordlinger lists fifteen specific tactics that state officials can use to reinforce
convergence.23
Type II state autonomy refers to situations in which state action changes diver-
gent societal preference to convergent ones. The state can use four general strate-
gies to effect such changes: altering the views of societal opponents; limiting the
deployment of resources by societal opponents; gaining the support of indifferent
actors; and increasing the resources of societal actors holding convergent views.24
Type I state autonomy refers to situations in which state actors translate their
preferences into authoritative action despite divergent societal preferences. They
can accomplish this by using the resources of the state to neutralize societal
opponents by measures such as deploying public capital, threatening to withhold
specific government programs, and masking the state's decision making procedures.
The state may even be able to act peremptorily by relying upon its inherent powers.
Public officials have the authority "to take any and all actions other than those
which violate the constitutional format and other legitimized procedural princi-
ples.'"25
Nordlinger's book is at once the most and least ambitious of the studies under
review. It is the most ambitious with regard to specifying the wide array of
resources that public officials can use to secure their preferences and in arguing for
their autonomy even in those settings where it has been most strongly denied. It is
the least ambitious with regard to delineating the institutional arrangements that
constrain political actors. Nordlinger's view of the political universe is not qualita-
tively different from earlier pluralist positions. Nordlinger does place much greater
emphasis on the role of the state. This is an important departure from prevailing
traditions. But like the pluralists, he sees a world of atomistic political actors, albeit
one in which relatively more of the atoms are public officials. Nordlinger concep-
tualizes the "state as government;" the state is defined as a collection of individuals
in official positions. It includes mayors as well as presidents, prefects as well as
cabinet ministers. These individuals may derive their preferences from a variety of
sources including other public officials, bureaucratic interests, distinctive experi-
ence and information, some conception of the public interest, and a desire for
greater autonomy.26 When more than one state unit is involved in a policy, "state
preferences are based exclusively upon the weighted intrastate resources of those
public officials who have an interest in the issue at hand ....27 Nordlinger en-
dorses the pluralist image of politics as a resolution of vectors.28
Nordlinger explicitly rejects state structures as an explanation for variation in the
degree of state autonomy. He argues that the ability of state officials to carry out

231

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

their preferences is not a function of the institutional structure within which they
must function but rather of the amount of societal resistance that they encounter.29
Nordlinger does note that it may be easier for a strong state than for a weak one to
change preferences, that appeals to common values may enhance support for the
state, and that the inherent power of public officials is limited by constitutional
rules; but he does not give much credence to such institutional constraints. The
political world is fluid; the preferences and capabilities of actors shift across issue
areas and over time. On the Autonomy of the Democratic State is focussed more on
the state than even nuanced pluralist arguments; it is concerned more with state-
society relations than with bureaucratic politics perspectives; but it does not differ
from these orientations in its basic depiction of the atomistic character of political
life.
Clifford Geertz' Negara: The Theater State in Nineteenth Century Bali is con-
cerned with a political world and a conception of politics very different from
Nordlinger's, and for that matter from conventional western political thought. It is a
brilliant, evocative, and poetic book, a book that at once draws the reader into a
totally alien world and draws out of that world enduring insights about the nature of
political life. Geertz shows that the essence of the state in nineteenth-century Bali
was not allocation but ceremony. "Court ceremonialism was the driving force of
court politics; and mass ritual was not a device to shore up the state, but rather the
state, even in its final gasp, was a device for the enactment of mass ritual. Power
served pomp, not pomp power."30 The Balinese state, the negara, was "a constel-
lation of enshrined ideas," whose central precept was that "worldly status has a
cosmic base, that hierarchy is the governing principle of the universe, and that the
arrangements of human life are but approximations, more close or less, to those of
the divine."'31 The main mechanism of political struggle in Bali was the court
ceremony, especially cremation, designed to demonstrate the relative status of a
particular noble house. These ceremonies were "not merely the drapery of political
order, but its substance."'32
The negara was only one of several authoritative, decision making institutions in
Bali. Issues involving civil society, including public works, local security, and civil
disputes, were the preserve of the village. Wet rice cultivation, or more precisely
the allocation of water rights, was regulated by irrigation societies. Religious com-
mitments were defined by the temple congregation. Membership in these three
groups intersected and overlapped.33
The defining authoritative relationship between the negara, composed of the
noble houses, and the rest of Balinese society was the obligation of individual
commoners to specific lords for service in ceremonial functions and war. These
commoners could be drawn from a number of different villages. They could belong
to different irrigation societies and temple congregations. The overriding objective
of each noble house was to control more men and resources in order to stage more
elaborate ceremonies to demonstrate a higher position in the social order, which
was seen as a reflection or mirror of a more encompassing hierarchical cosmos that
linked man and the gods.

232

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

In the concluding chapter of Negara entitled "Bali and Political Theory," Geertz
muses upon the broader implications of his presentation. He argues that since the
sixteenth century western political thought has identified the state with governance
or statecraft, with rule and control. However, Bali "exposes the symbolic dimen-
sions of state power." It "restores our sense of the ordering force of display,
regard, and drama."34 Western analysis has been utterly incapable of providing an
adequate account of this force. The ceremonial functions of the state have been
regarded, as in Hobbes, as a device to reinforce the power of the monarch, or in
Marx or Pareto as "great frauds" designed to facilitate the extraction of resources
by the elite. "Political symbology is political ideology, and political ideology is class
hyprocrisy.''35 Symbolic activities have not been regarded as ends in themselves.
Geertz conceives of the state as a normative order. His observations about the
central importance of symbolic activities in Bali are not merely of antiquarian
interest, and he does not mean them to be so. Even in the modern era they are a
common defining characteristic of the state. In its simplest form, what is univers
and distinctive to what Harry Eckstein has recently called the princely realm, is
that it symbolizes and represents in rules and laws the existence of a separat
society. Rules "stand for the fact that a common, thus, moral life exists, and they
celebrate the common life and make it compelling.''3"6 This symbolic function is th
core activity of barely differentiated political institutions in primitive societies.
While many other activities have entered the princely realm, the visible celebratio
of the polity's identity has not disappeared.
Symbols embodied in the state and representing basic political and ethical senti
ments that permeate the polity can be seen as a fundamental institutional constrain
that channels the behavior of individuals even to the point of endangering or
sacrificing their lives. (Hobbes was too worried about desertion, at least once the
existence of a moral community has been established.) Dahl's democratic creed
cannot be dismissed because of disagreement over specific applications, because
this creed defines a set of non-decisions that cannot be revealed by simple behav
ioral observation.37 David Truman's latent groups are not groups at all, but rathe
generally accepted societal beliefs that are enshrined in the decorative activities an
specific laws of the state and that delineate the acceptable range of political
behavior.
The central importance of symbols illuminated by Geertz provides greater insight
into the much maligned concept of the common good or the national interest. In
allocative terms, promoting the common good can be conceived of as a policy that
makes some or all actors better off without making any worse off. But seeing the
state as the institution where the common identity and moral beliefs of the polity
are embodied in ceremony and practice goes beyond questions of resource distribu-
tion among groups or of rectification of Pareto's suboptimal conditions. Political
activity focussed on the state sustains the ethical and moral needs of citizens, not
just their material ones. The destruction of the state by, for instance, alien con-
quest, is a loss for all citizens because it means the destruction or severe weakening
of the individual's social and moral community.

233

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

Political Development: The State as an Intervening Variable

The arguments presented by Nordlinger and Geertz do not address questions of

political development: they do not explain how various forms of the state that exist
in the contemporary world took their present shape. Especially for Nordlinger, the
focus of concern is the causal arrow from the state to the society. In contrast, for
Skowronek and the other authors discussed in this section the focus of attention is
on the causal arrow from the broader environment, both international and domestic,
to the nature of state institutions. The state (understood as an administrative
apparatus and legal order) is treated initially as the dependent variable, although in
subsequent time periods changed institutional capabilities alter the state's ability to
influence economic and social behavior.
All of the authors discussed here reject functional explanations of state develop-
ment. Institutional structures do not respond in any rapid and fluid way to altera-
tions in the domestic or international environment. Change is difficult. Incongru-
ence between the needs and expressed demands of the state and various societal
groups is the norm, not the exception. Institutional change is episodic and dramatic
rather than continuous and incremental. Crises are of central importance. Skow-
ronek defines a crisis as "a sporadic, disruptive event that suddenly challenges a
state's capacity to maintain control and alters the boundaries defining the legitimate
use of coercion. Crisis situations tend to become the watersheds in a state's
institutional development. Actions taken to meet the challenge often lead to
establishment of new institutional forms, powers, and precedents."38 Durin
riods of crisis politics becomes a struggle over the basic rules of the game r
than allocation within a given set of rules. However, once crises are past ins
tional arrangements tend to rigidify. Institutions reproduce themselves and re
more to their own needs than to those of their domestic society or the intern
environment. Thus, different kinds of causal variables are appropriate for exp
ing the creation, as opposed to the maintenance, of state institutions.
Crisis may be generated internally or externally. Skowronek is concerned
the former case; the other authors discussed here with the latter. Internally
ated crises are precipitated by dynamic changes in the society; externally gen
crises by threats from the international system that lead to state efforts to in
extraction, efforts that can precipitate societal resistance. Hence, two patter
incongruence between the state and its environment can be identified. Patte
public stasis and private dynamism, occurs when the exogenous sources of c
are internal. Pattern II, state demands and societal resistance, occurs when the
exogenous sources of change emanate from the international system.

Pattern I: Public Stasis and Private Dynamism The state as administrative appara-
tus and legal order will not smoothly adjust to changes in its domestic environment.
Once institutions are in place they will perpetuate themselves. Power holders strive
to select their own successors. Elaborate educational structures, such as France's
grands ecoles, may be created to socialize members of the higher civil service.39

234

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

One of Skowronek's major points of emphasis is that, short of a total overthrow of

the old regime, reforms must be carried out in the context of existing institutional
structures. Some individuals occupying positions in these institutions are bound to
resist change because it can undermine their budgetary support, policy scope, and
personal status. The natural path for institutions is to act in the future as they have
acted in the past.
Institutional statis is also encouraged by sunk costs. Once a given set of institu-
tional structures is in place, it embodies capital stock that cannot be recovered.
This stock takes primarily the form of information trust and shared expectations.
Long established institutional structures facilitate the exchange of information and
tacitly coordinate behavior. There is more information, and therefore less uncer-
tainty, about existing programs than proposed ones. If new institutions are created,
these infrastructures must be recreated. Thus, even if there is widespread societal
dissatisfaction with a particular set of institutions, it may be irrational to change
them. The variable costs of maintaining the existing institutions may be less than
the total costs of creating and maintaining new ones.40 Moreover, when pressures
are emanating from a rapidly changing domestic environment, it may be difficult to
know whether to commit substantial resources to create new institutions because it
may not be possible to distinguish transitory from enduring change.
The United States offers an example of a pattern of national development in
which the society has changed more dramatically than the institutional structures of
the state. Skowronek suggests that there are four characteristics that can be used to
describe any state: "the concentration of authority at the national center ..., the
penetration of institutional controls from the governmental center throughout the
territory . . ., the centralization of authority within the national government ...9
the specialization of institutional tasks and roles within the government."41 The
state that evolved in early nineteenth-century America rated low on all of these
dimensions. Most governing tasks were carried out by state and local governments.
The only national institutions of serious import were the parties and the courts. The
courts became "the American surrogate for a more fully developed administrative
apparatus."42 The parties helped to link the national and local levels through
patronage. However, these were weak foundations upon which to develop national
administrative capacity: the courts and parties were centered at the local level, and
their operations were fluid and malleable. Even the Civil War did not lead to any
permanent increase in the institutional capability of central governmental institu-
tions.43
There are straightforward functional explanations for the creation and reinforce-
ment of a weak and fragmented state in eighteenth and early nineteenth century
America. The American Revolution was a reaction against British efforts to in-
crease penetration and control of American society. The United States was not
confronted with any persistent external threat. Aristocratic traditions were weak, so
it was not necessary to have a strong state to facilitate socioeconomic transforma-
tion. The frontier offered a safety valve for societal pressures that might otherwise
have been directed at the central government. The physical abundance of the

235

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

United States made it easier to resolve conflict by increasing the size of the pie
rather than by asking authoritative institutions to alter the relative distribution of
the segments.
These social and economic conditions existed in symbiotic relationship with a
pervasive set of values that legitimated a weak state. Louis Hartz has argued that
the United States is a fragment society whose political beliefs were determined by
its initial settlers. In leaving Europe, these settlers not only escaped from their
contemporary conservative opponents but also from their future socialist ones.44
The American creed emplanted "liberal, democratic, individualistic, and egalitarian
values."45 It is not necessary to hold that this ideology was merely a handmaiden of
economic interest, or conversely that beliefs were entirely independent of societal
conditions. Through the middle of the nineteenth century, Lockean liberal princi-
ples were particularly well suited to the social structure and economic conditions
that existed in the United States. The liberal fragment could take root and flourish
precisely because it legitimated prevailing social relationships.
Thus, at its premier constitutional moment, in the period of its birth, the Ameri-
can state was congruent with its environment. State structures were consistent with
the functional needs of society, the preferences of state officials, and basic political
beliefs.
However, over time the society and economy changed. Industrialization involved
greater economic centralization and concentration in the United States as in other
areas of the world. Commercial networks became more complex. Conflicts between
labor and management became more frequent. Externalities multiplied. Market
imperfections became more common. Information was no longer readily available.
The United States became a significant actor in the international economy. The
fundamentally rural, agrarian society of early nineteenth-century America was
supplemented by an industrialized, urban society populated by diverse ethnic
groups that had unevenly assimilated the American creed. Such an environment
constantly placed strains on the weak and fragmented political system.
The response to industrialization is the major concern of Building a New Ameri-
can State: The Expansion of National Administrative Capacities, 1877-1920. Skow-
ronek examines the way in which institutional structures changed in three major
issue areas: business regulation, the military, and the civil service. A functional
orientation suggests a major adjustment in institutional structures to bring them into
congruence with the changing society; functionalist logic would predict the creation
of a strong state to meet new societal needs and demands. But this did not happen.
Skowronek refers to the period 1877-1900 as one of "state building as patch-
work." New institutions simply patched up the existing polity dominated by locally
oriented political parties and the courts. The Interstate Commerce Commission, the
major institutional innovation in government regulation of business, was gutted by a
series of Supreme Court decisions. A new civil service system was created, but it
was used only to fill in the interstices around the patronage appointments from
which the political parties drew much of their strength. The army was made more
professional, but at the same time the national guard, the representative institution
of political localism, was maintained and even strengthened.

236

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

The development of stronger national capabilities was impeded not simply by the
opposition of specific groups but more fundamentally by the existing institutional
structure. "Institutions and procedures once created to serve socioeconomic devel-
opment now appeared as self-perpetuating perversions of that purpose."''46 Local
interests, especially in the South, opposed a strong, professional, national army.
The political parties resisted a civil service system that would deny them patronage
appointments. The courts "vigorously asserted and jealously guarded the preroga-
tives of the judiciary in regulating economic affairs."47 The outcome of the struggle
between these interests of the old order and the proponents of the new, including
upper middle class professionals, regular army officers, government officials, and
businessmen, was determined not just by the relative power of each faction but by
the institutional structure within which the struggle took place. This was especially
true for the fate of the Interstate Commerce Commission. Skowronek writes that
"the key to understanding the early regulatory effort is not to be found in the
interests themselves but in the structure of the institutions they sought to influence.
In an archetypical case of the pluralist paradigm, each of the interests contesting the
railroad issue found representation in American national government, and each was
able to make its mark on the blank slate of national regulatory policy. Yet, in this
new regulatory situation, the sum of the interests could not possibly serve any of
them, let alone some 'public interest' standing apart from each; it only promised an
incoherent, unworkable policy from which no one stood to benefit. A state that
promoted pluralism promoted a formula for failure in regulation."48 Public policy
cannot be viewed simply as the resolution of a set of vectors. The interests and
political resources of actors are a function of existing institutions, and this may
make it impossible for any given actor or coalition of actors, whether state or
private, to implement their preferences.
Skowronek argues that a more powerful set of state institutions was created
between 1900 and 1920. However, even after the New Deal, more recent bursts of
social legislation, and two world wars, there is still no effective, consistent, and
coherent control of the national administrative apparatus. The old order of localistic
parties and courts was destroyed, but the "reconstituted" American state, to use
Skowronek's terms, did not successfully concentrate power. "Beyond the state of
courts and parties lay a hapless administrative giant, a state that could spawn
bureaucratic goods and services but that defied authoritative control and direc-
tion.'49 In the area of national administrative capability, institutions that were
created or strengthened after 1900 included the Civil Service Commission, the
Bureau of the Budget, and the General Accounting Office. However, the specific
positions of these agencies "within the federal establishment were all somewhat
obscure," and they all functioned in environments characterized by "parallel sets of
controls pitted against each other ... ." In the area of the military "[n]ationalism
came to mean a proliferation of semi-independent and competing power centers at
the national level rather than the establishment of a national center of power."50
Thus, the institutional structures of the past placed constraints on the possibilities
for the future. The preferences and capabilities of political actors cannot be treated
as exogenous variables: they can only be understood within the context of a given

237

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

set of institutional arrangements. Although Skowronek does not place much empha-
sis on enduring political beliefs, these too have played a continuing and central role
in the development of the American state. Huntington has pointed to the persistent
tensions between beliefs and institutional development that have erupted during
periods of creedal passion.51 The attack on the national bureaucracy by the Reagan
administration only serves to underline the enduring, deep resistance to the con-
centration of political power at the center. The sinister connotations of the term
"military-industrial complex" as well as the failure to agree on a consistent policy
toward military service are constant reminders that even in the area of national
defense the United States has failed to develop a coherent and legitimate set of
institutional arrangements. Judicial activism remains a powerful tradition for
American courts. And lest anyone think that localism is dead, consider the follow-
ing statement from the preliminary report of a commission established by the State
of Alaska to study deteriorating relations with the rest of the United States: "If
Alaska wants to protect its resources, its revenues and its state prerogatives, the
state government must vigorously defend [itself] against federal encroachments. It
should not be afraid of suing, of mounting a national information campaign, of
building political coalitions, or taking what otherwise might seem to be drastic
steps, with the exception of secession."52
In sum, Skowronek's study complements the work of other scholars, such as
Huntington and Hartz, who see varying degrees of tension between state institu-
tions and their environment as a constant theme of American political history.
Skowronek's underlying causal model is one in which the outcome of institution-
building during any particular period of crisis is a function of both contemporaneous
environmental factors, such as industrialization, and existing institutional structures
that are a product of past conditions, such as the nineteenth-century state of parties
and courts. Crises reduce incongruence between state structures and the domestic
environment. But during subsequent periods institutional structures reproduce
themselves while society changes, leading to increased tensions which eventually
precipitate another crisis.

Pattern H: State Demands and Societal Resistance Incongruity between the state
and civil society can also arise because the state increases its demands upon the
society. Such policies can lead to great social unrest, not simply because the level
of extraction increases, but also because legitimacy is placed at risk by the imposi-
tion of new state practices. Political leaders do not bring such difficulties upon their
heads for nothing.
Throughout the history of the western state system the most persistent sources of
pressure on the state have been external. It is the threat of invasion, or the desire to
act efficaciously in the international system, that has prompted rulers to increase
their level of extraction from the society. This is a major theme of a number of
studies published during the last decade. In the penultimate volume of the Social
Science Research Council's (SSRC) series Studies in Political Development, The
Development of National States in Western Europe, the editor Charles Tilly argues
that the need to maintain and increase military establishments was the major

238

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

impetus for the growth of the state's administrative apparatus. Greater military
capacity required higher levels of taxation. Higher levels of taxation required a
more extensive bureaucracy. Extractive capacities initially used for military pur-
poses could be applied in other areas as well. Taxes that began as extraordinary
levies to fight specific wars often became normal and continuing sources of reve-
nue.53
One of the repetitive patterns noted in Crises of Political Development in Europe
and the United States, the last volume of the SSRC series, is that external threats
lead to crises of penetration. The articles on Germany and Russia by John R. Gillis
and Walter M. Pintner, respectively, place particular emphasis on this point. Gillis
argues that nineteenth-century German reforms were designed to increase the
power of the state, not to create a more liberal society. The humiliation of Prussia
at Jena in 1806 convinced leading members of the elite that rigid absolutism would
perpetuate military inferiority. Reforms were initiated by military leaders and bu-
reaucrats to mobilize the population in the service of the state. Citizenship was
defined as a set of duties, not rights. Legislative changes and the extension of the
franchise were introduced from above, not only to stave off the possibility of
domestic unrest, but also to increase the state's level of extraction. Similarly
Pintner shows that the perennial Russian problem was that the needs of the state,
prompted by external threats, exceeded its extractive capacity. Russian rulers
strove to maintain a large army on an anemic agrarian economic base. Nineteenth-
century social reforms were prompted by external failures. Defeat in the Crimean
War convinced the ruling elite (composed of the tsar and a small group of high
officials) that basic change was necessary. Economic reform, including the emanci-
pation of the serfs, was initiated from above to increase the level of resources that
the state could secure from its society.
Ellen Kay Trimberger's Revolution from Above: Military Bureaucrats and Devel-
opment in Japan, Turkey, Egypt, and Peru examines four cases of major social
change initiated from above by military bureaucrats. Trimberger defines a revolu-
tion as "an extralegal takeover of the central state apparatus which destroys the
economic and political power of the dominant social group of the old regime."54
There are two basic preconditions for revolution from above: the military and
civilian administration must be bureaucratized in a Weberian sense, and they must
be independent of the dominant class. Given these preconditions a revolution from
above can be pyecipitated "only in a crisis situation-when the existing social,
political, and economic order is threatened by external forces and by upheaval from

below ... ."ss

military's Foreign
position. pressures
National lead to spawns
degradation internal disorderideologies
nationalist which undermines the
that unite the
military internally and provide a rationale for dramatic action. In Japan, Turkey,
Peru, and Egypt the military initiated major social and economic changes after they
had taken power. The purpose of these changes was to enhance the power of the
state by promoting economic development. However, Trimberger is skeptical of the
long-term impact of these initiatives because of the tendency of subsequent leaders
to ally with the dominant economic forces and because of the constraining influence
of the world economic system.

239

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

In sum, a number of recent studies point to external pressures as a major cause of

the development of new state capacity. Such external pressures are almost cer-
tainly more compelling than internal ones. The failure to act effectively, as in
Geertz's negara and in eighteenth-century Poland, means the destruction of the
polity. But in cases of both external and internal environmental pressures, a similar
pattern emerges, characterized by rapid change during periods of crisis followed by
consolidation and stasis. To borrow a metaphor from recent work in evolutionary
biology, such a pattern can be labeled punctuated equilibrium.

Punctuated Equilibrium

All of these studies of political development point to differential rates of change in

social and political structures over time. A basic analytic distinction must be made
between periods of institutional creation and periods of institutional stasis. The
kinds of causal factors that explain why a set of state structures is created in the
first place may be quite distinct from those that explain its persistence over time.
New structures originate during periods of crisis. They may be imposed through
conquest or be implanted by a particular fragment of the existing social structure.
But once institutions are in place they can assume a life of their own, extracting
societal resources, socializing individuals, and even alterting the basic nature of
civil society itself. The causal dynamics associated with a crisis of the old order and
the creation of a new one are different from those involved in the perpetuation of
established state institutions.
Furthermore, once a critical choice has been made it cannot be taken back. There
may be a wide range of possible resolutions of a particular state-building crisis. But
once a path is taken it canalizes future developments. Sidney Verba has referred to
this conceptualization as the branching tree model of sequential development. A
critical choice forecloses other options in part because the "choice to set up a
program in relation to a particular problem area may lead almost inevitably to the
maintenance and even expansion of the program because of the vested interests it
creates."56 It is not possible in human affairs to start de novo with every change in
wants, needs, and power capabilities. Past choices preclude certain strategies or
make them very costly. Institutions generated by functional demands of the past
can perpetuate themselves into a future whose functional imperatives are radically
different.
One of the clearest examples of a branching tree argument is Lipset and Rokkan's
model of the development of European party systems.57 They argue that there were
three crucial historical junctures: first, the Reformation involving the struggle for
control of the church; second, the Democratic Revolution after 1789 involving the
struggle for control of the rapidly expanding educational system; and third the
Industrial Revolution involving conflict between the urban center and agrarian
periphery. At each of these critical junctures the nation-building elite that controlled
the state machinery had alliance options. Once a particular alliance was chosen,
however, it set the agenda for future party development. Thus, the settlement of the

240

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

sixteenth-century struggle between the church and the state "gave a very different
structure to the cleavages of the nineteenth" century in Protestant and Catholic
Europe.8 In a more recent analysis Lipset has argued that "the nature of working
class politics has been profoundly influenced by the variations in the historic
conditions under which the proletariat first entered the political arena." Formative
experiences initiated "'certain trends or institutional patterns that took on a self-
perpetuating character and hence affected ideology, structure, and political out-
comes in later years.""59
In a particularly provocative set of analyses Charles Sabel and his collaborators
have argued that the organization of industrial society has also followed a branching
tree pattern. Fordism, the use of mass production techniques involving special
purpose machines and unskilled labor to produce standardized products, was not
foreordained by the nature of technology. Flexible specialization involving small
firms using skilled workers and general purpose machines to produce more spe-
cialized products was a viable alternative. However, as a result of the distribution
of property rights and income in Britain and the United States, Fordism became the
dominant mode of industrial production. Once this choice was made, economic and
political institutions were shaped to guarantee the existence of stable mass markets.
Thus, an historically contingent set of conditions in the nineteenth century set the
path for the future evolution of industrial society.60
Fernand Braudel has argued that one of the critical determinants of the rise of
capitalism in the West was that for a set of fortuitous reasons cities revived faster
than states in the Middle Ages. Revivals, Braudel maintains," always feature two
runners, the state and the city. The state usually won and the city then remained
subject and under a heavy yoke. The miracle of the first great urban centuries in
Europe was that the city won hands down, at least in Italy, Flanders, and Ger-
many." Braudel goes on to suggest that the existence of free cities created a new
state of mind, "broadly that of an early, still faltering, Western capitalism-a
collection of rules, possibilities, calculations, the art both of getting rich and
living.""6 Thus, small initial differences promoted a new kind of institutional
structure with profound consequences over the long term.
Moreover, choices made by leading states at a particular point in time influence
not only their future range of options, but also the options of later developing
states. The functions that are viewed as proper and legitimate for the state are
influenced by general international norms and practices. In the modern system the
institutional characteristics of states in more industrially developed areas have set
an agenda for states in less developed areas. These characteristics come to be
associated with the essential nature of the "modern" state and cannot be ignored
even by states with very different needs. In his study of social security systems in
Britain and Sweden, Hugh Heclo points out that the options pursued by the Swedes
were heavily influenced by the bureaucracy's assessment of policies that had been
adopted in countries with more developed industrial structures. "Because of a
process of transnational learning Sweden found unemployment insurance on its
national agenda at about the same time that Britain did even though Sweden's level
of economic development and industrialization lagged considerably behind Brit-

241

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

ain's."62 Virtually every country today has, at least on paper, some kind of social
security system, even though the economic resources of many Third World coun-
tries are totally incapable of actually implementing such a program. The organiza-
tion of educational systems throughout the world has not only been seen as a
responsibility of the state (as opposed to the church or some other private organi-
zation), but has also mimicked the characteristics of systems in the more developed
countries.63
Even the state itself can be seen in this light. The concept of a single hierarchical
ruling structure governing a defined territorial area developed out of feudal Europe.
New military technologies in the late Middle Ages were characterized by economies
of scale providing an incentive to form larger territorial units. The revival of trade
offered economic benefits to those political actors that could assure the safe move-
ment of goods over longer distances. States, especially nation-states, were able to
secure more intense affective commitment from their inhabitants than were em-
pires.
Over time the national state has pushed aside all other forms of political organi-
zation. After the second world war demands for decolonization could only be met
by granting formal independence, even though many of the areas that achieved this
status lacked the economic, military, and bureaucratic capability to function effec-
tively. Some intermediate form of political organization that divided functions
between colonial territories and their home governments would almost certainly
have been more politically and economically efficacious. But such solutions were
not possible. They lacked legitimacy. They could not have commanded the support
of colonial populations. The triumph of the national state in Europe became a
triumph of the national state around the globe. Choices made in Europe's past
dictated the possibilities for Africa's future. Once a particular path had been
chosen, other paths, perhaps more functionally appropriate for contemporary
problems, were foreclosed.
To borrow a term from another discipline, an imagery that expects short bursts of
rapid institutional change followed by long period of stasis can be termed
punctuated equilibrium. Punctuated equilibrium refers to a set of arguments about
evolution whose main proponents are Stephen Jay Gould and Niles Eldredge.
Gould and Eldredge have attacked the conventional Darwinian synthesis which
pictures evolutionary progress as a slow, continuous process of change in which
entire species slowly adapt to environmental conditions. They have argued instead
that change tends to take place rapidly in geographically isolated groups which may
then displace their ancestral populations. Such displacements are rare. Generally
species do not change substantially over very long periods of time. Evolutionary
change, Gould and Eldredge argue, is concentrated in geographically instantaneous
events .64
Gould has noted that the gradualist-punctualist debate in the largest sense "is but
one small aspect of a broader discussion about the nature of change: Is our world
(to construct a ridiculously oversimplified dichotomy) primarily one of constant
change (with structure as a mere incarnation of the moment), or is structure primary
and constraining, with change as a 'difficult' phenomenon, usually accomplished

242

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

rapidly when a stable structure is stressed beyond its buffering capacity to resist
and absorb. "65 This description of the basic nature of the debate in evolutionary
theory has its close analog in social and political analysis. Punctuated equilibrium is
an apt description of an analytic stance that sees political institutions enduring over
long periods once they are established.
Critics of the Darwinian synthesis have also made branching tree arguments.
Once a particular evolutionary path is taken, the direction of future evolutionary
developments is constrained by the available genetic pool.66 If the same set of
environmental conditions exists at two different points in time or at two geographi-
cally isolated areas of the globe, they will not give rise to the same set of species.
Speciation is a function not only of the contemporary environment, but also of past
environments, of paths that have been followed and of paths that have not been
followed. While the long time frames and slow change of the conventional Darwin-
ian model offer the possibility of optimal functional adjustment (at least over the
eons), such optimality is not anticipated by alternative approaches. In a world
characterized by punctuated equilibrium there is more uncertainty and chance. A
particular structural development reflecting marginal advantages at a particular
point in time may constrain future evolutionary developments.
The metaphor of punctuated equilibrium suggests a very different world from that
of pluralism and other orientations that emerged out of the behavioral persuasion.
Central to these approaches was not simply a societally oriented focus but also
reservations about institutional constraints. If institutions adjusted relatively
quickly to societal changes, and if formal institutions did not explain political
behavior, there was little point in making them an object of scholarly investigation.
Attention could be focussed on the motivations of individuals or groups. Even the
state, which Geertz refers to as "that master noun of modern political discourse,"
could be ignored.6' But if institutions-the administrative apparatus, legal order,
and political beliefs-are seen as basic determinants of both the interests and the
power of political actors, a different agenda is suggested for political research. How
can political institutions, including the state, be adequately described? How do
institutional structures constrain the behavior of individual actors? What factors
best explain the creation of new institutions? What resources enable institutions,
especially the state, to perpetuate themselves? What is the duration of lags between
different kinds of environmental changes and changes in different kinds of institu-
tional arrangements? When do state institutions fail to change, even when the
polity's survival is at stake? When state institutions are suboptimal or even
counter-productive for those individuals living within a given territory, what pos-
sibilities are there for change'?
The books discussed in this essay have taken the first steps in offering answers to
these questions. Others will follow. The more comfortable and familiar world of the
1950s and 1960s is gone. American global hegemony has eroded. "Enlightened"
policies have not ended social ills. Economic problems do not respond to con-
ventional solutions. Third World countries will not follow the path trod by the
United States. Institutional arrangements that seemed to be part of the basic nature
of things have come undone. In such a world the attention of scholars will turn from

243

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January/ 1984

behavior within a given set of institutional constraints to the constraints themselves.

The work of Nordlinger, Geertz, Skowronek, Tilly, Grew, and Trimberger show
that the agenda is already changing. "The state" will once again become a major
concern of scholarly discourse.

NOTES

An earlier version of this paper received critical and astute comments from Peter Evans, John Ferejo
Peter Katzenstein, Robert Keohane, John Meyer, Dietrich Reuschmeyer, Ronald Rogowski, Charles S
Theda Skocpol, and Charles Tilly. I also benefitted from a series of papers presented at a conference
States and Social Structures held at the Seven Springs Conference Center, Mount Kisco, New York
February 1982, especially "Bringing the State Back In" by Theda Skocpol.
1. Peter Katzenstein, Between Power and Plenty (Madison: University of Wisconsin Press, 1977); Th
Skocpol, States and Social Revolutions (New York: Cambridge University Press, 1979); Alfred Stepan,
State and Society in Peru (Princeton: Princeton University Press, 1978). My own Defending the Natio
Interest (Princeton: Princeton University Press, 1978) applied a statist interpretation to policy in the area o
foreign raw materials investments.
2. Roger Benjamin and Raymond Duvall, "The Capitalist State in Context," unpublished manuscript,
University of Minnesota, pp. 4-4 to 4-8. I have combined two of Benjamin and Duvall's categories,
state as public bureaucracy and the state as legal order, a strategy they also follow.
3. The classic presentation of the bureaucratic politics argument is Graham Allison, Essence of Decis
(Boston: Little-Brown, 1971).
4. For the distinction between politics as allocation and politics as us-against-them, see Gianfra
Poggi, The Development of the Modern State (Stanford: Stanford University Press, 1978), chap. 1. Th
quotation is from Harold Lasswell, Politics: Who Gets What, When, How (Cleveland: World Publish
Company, 1958).
5. Those who view the state as government, as a collection of individuals in authority, do not see t
institutional structure as constraining.
6. Gabriel Almond, "Approaches to Developmental Causation," in Gabriel Almond et al., Crisis, Choi
and Change: Historical Studies of Political Development (Boston: Little-Brown, 1973), pp. 22-24.
7. Robert Dahl, Who Governs: Democracy and Power in an American City (New Haven: Yale Univers
Press, 1961). Pluralism is the major strand of the behavioral approach in American political science.
8. The distinction between autonomy and congruity or conformity is developed in Benjamin and Duva
9. Eric Nordlinger, On the Autonomy of the Democratic State (Cambridge: Harvard University Pre
1981), p. 151.
10. Ibid., p. 44.
11. Who Governs appeared over twenty years ago. However, more recent work by Dahl and Lindblom
as societally oriented, emphasizing in particular the corrupting influence of large corporations on dem
cratic processes. See especially Charles Lindblom, Politics and Markets (New York: Basic Books, 1977).
See John Manley, "Neopluralism: A Class Analysis of Pluralism I and Pluralism II," American Polit
Science Review, 77 (June 1983), for a discussion and critique of this evolution from a Marxist perspect
12. Dahl, p. 325. For the purposes of this discussion I have emphasized Dahi's treatment of Richard L
the most important public actor in Who Governs. However, Dahl's study has generally been understo
more as an effort to refute economically oriented, elite community studies by demonstrating that t
number of societal actors and the dispersion of political resources are greater than Marxist-inspi
formulations would indicate. Who Governs is concerned with demonstrating both that the menu of soc
action is long and varied and that political leadership can influence outcomes.
13. Benjamin and Duvall, p. 4-4.
14. Dahl, p. 311.
15. Ibid., pp. 311-312.
16. Dahl pays more attention to political beliefs in Polvarchv: Participation and Opposition (New Hav
Yale University Press, 1971), where he argues that beliefs are one of the critical determinants in the
formation of an inclusive regime characterized by a high degree of contestation. Most of his treatm
focuses on individual attributes and process rather than substance. However, the last part of the cha
about "The Beliefs of Political Activists," which Dahl entitles "Another Paradigm," offers arguments v
similar to the ones presented in this paper. In this section Dahl discusses shifts in the substantive belief
whole populations during periods of instability and breakdown followed by periods of relative stability
Polyarchy, chap. 8, esp. pp. 180-88.

244

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
 Stephen D. Krasner

17. Dahl, Who Governs, p. 165.

18. Harry Eckstein, "On the 'Science' of the State", Daedalus, 108 (Fall 1979), 4 and passim. While
formal legalism was the dominant intellectual orientation before 1950, it was not the ony one. The work of
scholars such as Schattschneider, Merriam, and Lassell provided starting points for the behavioral revolu-
tion.
19. Ibid., pp. 10-11. See also Gabriel Almond, "Corporatism, Pluralism, and Professional Memory,"
World Politics, 35 (January 1983), 245-47, for a discussion of interest group literature as a reaction to
formal legalism.
20. Thomas Kuhn has particularly emphasized the importance of heuristic power as a criterion for
choosing among competing paradigms. See The Structure of Scientific Revolutions (Chicago: University of
Chicago Press, 1962), chap. 12.
21. Nordlinger, p. 1.
22. Ibid., p. 11.
23. Ibid., chap. 3, esp. p. 91-98.
24. Ibid., chap. 4, esp. pp. 109-117.
25. Ibid., chap. 5, esp. pp. 128-139.
26. Ibid., pp. 32-33.
27. Ibid., p. 16.
28. Theda Skocpol makes a similar argument. See her "Bringing the State Back In," paper presented at
the Conference on States and Social Structures, Mount Kisco, New York, February 25-27, 1982.
29. Nordlinger, pp. 192-97.
30. Clifford Geertz, Negara: The Theatre State in Nineteenth Century Bali (Princeton: Princeton Univer-
sity Press, 1981), p. 13.
31. Ibid., pp. 135 and 102.
32. Ibid., p. 32.
33. Ibid., p. 153.
34. Ibid., p. 122.
35. Ibid.
36. Harry Eckstein, "The Idea of Political Development: From Dignity to Efficiency," World Politics, 34
(July 1982), 472.
37. Peter Bachrach and Morton Baratz, "The Two Faces of Power," American Political Science Review,
56 (December 1962).
38. Stephen Skowronek, Building a New American State: The Expansion of National Administrative
Capacities (New York: Cambridge University Press, 1982), p. 10.
39. For an illuminating discussion of the French case see Ezra Suleiman, Elites in French Society
(Princeton: Princeton University Press, 1978).
40. For discussions of sunk costs see Robert O. Keohane, "The Demand for International Regimes," in
Stephen D. Krasner, ed., International Regimes (Ithaca: Cornell University Press, 1983); and Arthur
Stinchcombe, Constructing Social Theories (New York: Harcourt, Brace and World, 1968), pp. 120ff.
41. Skowronek, p. 20.
42. Ibid., p. 28.
43. Ibid., chap. 2.
44. Louis Hartz, The Liberal Tradition in America (New York: Harcourt, Brace, 1955).
45. Samuel P. Huntington, "American Ideals versus American Institutions," Political Science Quarterly,
97 (Spring 1982), 2.
46. Skowronek, p. 40.
47. Ibid., p. 122.
48. Ibid., p. 131.
49. Ibid., p. 290.
50. Ibid., pp. 208 and 247.
51. Huntington.
52. Quoted in the Wall Street Journal, January 18, 1983.
53. Charles Tilly, "Reflections on the History of European State-Building," in Charles Tilly, ed., The
Formation of National States in Western Europe (Princeton: Princeton University Press, 1975), pp. 42, 54,
73-74.

54. Ellen Kay Trimberger, Revolution from Above: Military Bureaucrats and Development in Japan,
Turkey, Egypt, and Peru (New Brunswick: Transaction Books, 1978), p. 2.
55. Ibid., pp. 4-5.
56. Sidney Verba, "Sequences and Development," in Leonard Binder et al., Crises and Sequences in
Political Development (Princeton: Princeton University Press, 1971), p. 308.

245

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms
Comparative Politics January 1984

57. Seymour M. Lipset and Stein Rokkan, "Cleavage Structures, Party Systems, and Voter Alignments:
An Introduction," in Seymour M. Lipset and Stein Rokkan, eds., Party System and Voter Alignments:
Cross National Perspectives (New York: The Free Press, 1968).
58. Ibid., p. 38.
59. Seymour M. Lipset, "Radicalism or Reformism: The Sources of Working Class Politics," American
Political Science Review, 77 (March 1983), 1 and 16.
60. Charles F. Sabel, Work and Politics: The Division of Labor in Industry (New York: Cambridge
University Press, 1982), esp. chap. 4 and 5; Michael Piore and Charles F. Sabel, The Industrial Divide (New
York: Basic Books, forthcoming); and Charles F. Sabel and Jonathan Zeitlin, "Historical Alternatives to
Mass Production," Past and Present (forthcoming).
61. Fernand Braudel, Civilization and Capitalism, 15th-18th Century, Volume I, The Structures of
Eveyday Life: The Limits of the Possible (New York: Harper and Row, 1981), pp. 511-513.
62. Hugh Heclo, Modern Social Politics in Britain and Sweden (New Haven: Yale University Press,
1974), p. 66.
63. The central importance of transnational values has been emphasized by John Meyer and his associ-
ates. See, for instance, John W. Meyer, "The World Polity and the Authority of the Nation-State," in
Albert Bergesen, ed., Studies of the Modern World-System (New York: Academic Press, 1980); John W.
Meyer and Brian Rowan, "Institutionalized Organizations: Formal Structure as Myth and Ceremony,"'
American Journal of Sociology, 83 (1977); Francisco Ramirez and John Boli-Bennett, "Global Patterns of
Educational Institutionalization," in Philip Altbach et al., Comparative Education (New York: Macmillan,
1982).
64. Niles Eldredge and Stephen Jay Gould, "Punctuated Equilibria: An Alternative to Phyletic
Gradualism," in Thomas J. M. Schopf, ed., Models in Paleobiology (San Francisco: Freeman, Cooper,
1972); Stephen Jay Gould, "Darwinism and the Expansion of Evolutionary Theory," Science, 216 (April 23,
1982); Stephen Jay Gould and Niles Eldredge, "Punctuated Equilibria: The Tempo and Mode of Evolution
Reconsidered," Paleobiology, 3 (1977).
65. Gould, "Darwinism and the Expansion of Evolutionary Theory," p. 383.
66. Ibid., p. 303.
67. Geertz, p. 121.

246

This content downloaded from

201.163.9.162 on Sun, 19 Mar 2023 17:27:23 UTC
All use subject to https://about.jstor.org/terms