3.03.02.33 - U5 Gourevitch - The Second Image Reversed
3.03.02.33 - U5 Gourevitch - The Second Image Reversed
3.03.02.33 - U5 Gourevitch - The Second Image Reversed
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=uwisc. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
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 [email protected].
University of Wisconsin Press is collaborating with JSTOR to digitize, preserve and extend access to
International Organization.
http://www.jstor.org
The second image reversed:the
internationalsources of domestic politics
Peter Gourevitch
881
882 InternationalOrganization
itself become an explanatoryvariable. Instead of being a cause of international
politics, domestic structuremay be a consequenceof it. Internationalsystems, too,
become causes instead of consequences.
Like others working in these fields, I am interestedin the questions posed by
both sensibilities. In this essay, however, I wish to concentrateon those posed by
the comparativist.We all know about interaction;we all understandthat interna-
tional politics and domestic structuresaffect each other. Having recentlyread, for a
variety of purposes, much of the currentliteraturewhich explores this interaction,I
thinkthe comparativist'sperspectivehas been neglected, thatis, the reasoningfrom
internationalsystem to domestic structure.I offer comment and criticism of that
literaturein three interrelateddomains:
First, in using domestic structureas a variablein explainingforeign policy, we
must explore the extent to which that structureitself derives from the exigencies of
the internationalsystem. As a contributionto such exploration, I will examine a
variety of argumentsfound in diverse writings which seek to do this.
Second, in using domestic structureas a variablefor explainingforeign policy,
much of the literatureis "apolitical." It stresses structuralfeatures of domestic
regimes which constrainpolicy regardless of the content of the interests seeking
goals throughpublic policy or the political orientationof the persons in control of
the state machine. I will examine this problem through a brief discussion of the
distinction often made between "strong" and "weak" states as an explanationof
foreign economic policy.
Finally, in exploring the links between domestic and internationalpolitics
much of the literatureargues that a break with the past has occurredsuch that the
present characterof the interactionrepresentsa discontinuitywhich requires new
categories of analysis. In particular,much is made of interdependence,permeabil-
ity, transnationalactors, and the decline of sovereignty. While it is certainthat the
present is not identical to the past, this claim for newness is overstated. Many
features which are considered characteristicof the present (interdependence,the
role of trade, transnationalactors, permeability, conflict within the state over
desirable policy) also seem relevant to past systems and regimes; and conversely,
characteristicsof the past (war, instability, sovereignty, military power, interna-
tional anarchy)seem to be still with us.
The bulk of the paperdeals with the first point. The other two are handledall
too briefly and tentatively.I place them all together, at the risk of overburdeningthe
reader, because I wish to show that there is some reward for the international
relations specialists interested in the second and third points who are willing to
undertakethe voyage of looking at the first. Treatingthem togetherwill compel us
to think differentlyabout the linkage between internationalrelations and domestic
politics.
The internationaleconomy
Recent events and the internationalrelations literaturehave made us acutely
aware of the impact of world marketforces upon domestic politics. Citing the oil
embargo after the Arab-Israeliwar of 1973 makes other examples unnecessary.
These effects though are not something new. The Great Depression alone did not
bringHitlerto power:Germanhistory, institutions,parties, political culture, classes
and key individualsdid that, but it is impossibleto imagine thatwithoutthe millions
thrownout of work by the contractionof the United States economy following the
Crash of 1929 these other forces could have broughtthe result about.
The economic cycle referredto as the Great Depression of 1873-96 also had
dramaticeffects on political life aroundthe world. Immenseincreasesin agricultural
and industrialproductioncaused the prices of both sorts of goods to plummet. In
Britain, the flood of foreign grain drove many persons off the land, underminedthe
landed aristocracy,and hasteneddemocratizationof political life (the secret ballot,
universal suffrage, elected local government, disestablishmentof the Church). In
France and Germany, the drop in prices threatenedlanded and industrialinterests.
In both countries, these groups managed to protect themselves by erecting high
tariff barriers. In France, this served to strengthenthe Republic. In Germany, it
stabilized Bismarck's newly-fashioned ramshackle empire. In both countries,
preindustrialgroups were thereby able to prolong their positions with ultimately
disastrous consequences for constitutionalgovernment:fascism in Germany, the
Frenchcollapse in the thirtiesand Vichy afterwards.Italy, Russia and Southeastern
Europecould no longer provide even subsistence to much of their populationsand
sent a tidal wave of migrants aroundthe world. In America, the late nineteenth
centurydepressionspawnedPopulism, a most powerful challenge to the two-party
system and to the hegemony of industrialinterests.It was ultimatelydefeatedin part
because the immigrants hurled ashore by the crisis in Europe sided with their
Republicanemployers against the Populist and Democratic farmers.4
In all these countries, what we now call transnationalactors were certainly
present (at least in some sense of that term)5: British investors, German steel
manufacturers,French engineers, American missionaries.
That internationalmarketforces affect politics and have done so for a long time
seems incontrovertible.Can we find general arguments which posit systematic
relationshipsbetween such forces and certain configurationsof regime type and
coalition pattern?
8F. Carsten, The Origins of Prussia (Oxford: Clarendon Press, 1968); T. Hamerow, Revolution,
Restoration, Reaction (Princeton:Princeton U.P., 1958); J.H. Clapham, Economic Development of
France and Germany, 4th edition (Cambridge:CambridgeU.P., 1935).
The second image reversed 887
reactionarymethods adoptedin Germanyand Japanwould scarcely have been
possible. Without both the capitalist and reactionaryexperiences, the com-
munist method would have been something entirely different, if it had come
into existence at all.... Althoughtherehave been certaincommonproblemsin
the constructionof industrialsocieties, the task remainsa continuallychanging
one. The historicalpreconditionsof each majorpolitical species differ sharply
from the others.9
9BarringtonMoore, Jr., Social Origins of Dictatorshipand Democracy (Boston: Little, Brown, 1966),
pp. 413-44. For a critique of Moore directed at the failure to develop sufficiently an "intersocial
perspective," see Theda Skocpol, "A CriticalReview of BarringtonMoore's Social Originsof Dictator-
ship and Democracy," Politics and Society (Fall 1973): 1-34.
"0AlbertHirschman, "The Political Economy of Import Substituting Industrializationin Latin
America," in A Bias for Hope (New Haven: Yale University Press, 1971), pp. 85-123, and "The Turn
to Authoritarianismin LatinAmericaand the Searchfor Its Economic Determinants," in the forthcoming
volume on Latin Americaedited by David Collier, and "A GeneralizedLinkage Approachto Develop-
ment, with Special Reference to Staples," Economic Development and CulturalChange, 25 (Supple-
ment 1977): 67-98.
"Guillermo O'Donnell, Modernizationand Bureaucratic Authoritarianism(Berkeley: University of
California,Institutefor InternationalStudies, Politics of Modernization-Series, no. 9, 1973) and "Re-
flections on the GeneralTendencies of Change in Bureaucratic-Authoritarian States," Latin American
Research Review, forthcoming.
888 InternationalOrganization
The system whirls like a fast merry-go-round:Italy, Greece, Spain, or Portugalcan
leap on when certain types of manufacturingfavor their mix of labor supply and
organizationalskills but they lack the maturedstrengthto hold on for good, and
consequently face the constant danger of being thrown off. This situation also
constrainspolitics, pushing and pulling toward and away from liberalism.'2
For this groupof authorswith whom I am associatingGerschenkron's notion of
late development, political outcomes within countries are strongly affected by the
characterof the world economy at the time in which they attemptindustrialization.
Because of competitionand changes in technology, each entrantinto the "industri-
alizationrace" faces a new game, with alteredrules. For all these authors,however,
the impact of each internationalsituationcannot be determinedwithout knowledge
of the internalcharacterof each society. Gerschenkronshowed how certaincharac-
teristics of German and Russian society became advantages as the economy
changed. Moore, Hirschman, O'Donnell, and Kurth all stress internal factors as
well: the characterof social developmentat the point at which the countryis drawn
into the internationaleconomy. This, among other points, differentiatesthem from
the next group, the "dependencia" or "center-periphery"theorists.
13AndreGunder Frank, Capitalism and Underdevelopmentin Latin America (New York: Monthly
Review Press, 1967).
14ImmanuelWallerstein, The Modern WorldSystem (New York: The Academic Press, 1974).
890 InternationalOrganization
Wallerstein'streatmentis complex: it has vices and virtues about which I and
many others have written.15 For the purposeat hand, the importanceof his discus-
sion lies in his insistence on a "world-system" perspective. Wallerstein sees his
work as a break with state-centeredaccounts of economic and political develop-
ment. Commercializedagriculture, early manufacturing,and the factory system
cannot be understoodin disaggregated, national terms. From the beginning, na-
tional economies grew in interactionwith each other. The analystmust seek there-
fore to understandthe propertiesof the system as a whole. Differentiationis one of
the centralpropertiesof that system, one which confirms the necessity of the world
viewpoint since its effects can only be detected from such a perspective. For Wal-
lerstein the very essence of capitalismlies in that sort of differentiation:the opera-
tion of marketforces leads to the accentuationof differences, not their reduction.
Rather slight differences at an early point may explain why one area or country
ratherthananothertakes a particularplace in the system-why for example Western
ratherthan EasternEurope became the core. Once the system begins to articulate
itself, it greatly magnifies the consequences of the early differences.
Also political differentiationmust, for Wallerstein, be viewed from world
system to country, not vice versa. The internationalsystem is the basic unit to be
analyzed, ratherthan units of power which come into being from some process
conceived of quite separately from the operation of the system. States are the
concrete precipitatesfrom the system, not the component units of it. Perhapsit is
not accidentalthat this argumenthas been enunciatedby a sociologist, ratherthan
by a political scientist or more particularlyan internationalrelations specialist. It
amply partakesof Durkheim,as well as Marx, and not, as the internationalrelations
specialist would do, of Hobbes, Rousseau, and Kant. Durkheimderives the indi-
vidual from society: he is as individuallydifferentiatedfrom others as the society
allows, according to its division of labor. Similarly, Wallersteinderives the state
from the system. The internationaldivision of labordetermineshow much variance
in political forms is allowed the componentunits. Position in the division of labor
determinesthe type of form:states at the core must be strong;states at the periphery
must be weak.
Wallerstein wavers from this resolute application of the world-system
frameworkat times. When he takes up the explanationof why "coreness" requires
strength, he notes that the core states needed governments capable of defending
themselves militarilyagainstrivals, of imposing themselves on certainmarketsand
sources of materials,and of creatinglarge uniformmarketsinternally.The reason-
ing is then circular:strong states led to a core position, not a core position to strong
states. That there was interactionmakes sense but reduces the explanatoryleverage
providedby the general argument.The explanationof strengthcan no longer be so
cleanly connected to a system-level argument. Some exploration of internal
18RobertKeohane and Joseph Nye, eds., TransnationalRelations and World Politics (Cambridge:
HarvardU.P., 1971), and Power and Interdependence(Boston: Little Brown, 1977); EdwardMorse,
Modernizationand the Transformationof InternationalRelations (New York: Free Press, 1976). For
criticism, see Kenneth Waltz, "The Myth of National Interdependence,"in Charles P. Kindleberger,
ed., The InternationalCorporation (Cambridge:MIT Press, 1970).
"9GrahamAllison, "Conceptual Models and the Cuban Missile Crisis," American Political Science
Review LXIII (September 1970). Not surprisingly, these debates relate to changes in reality: realism
dominated in a period of war and militaryconfrontations;the easing of Cold War tensions and greater
fluidity in internationalrelationsmeantthe system was less plausiblyconstraining,hence the disaggregat-
ing of the state through bureaucraticanalysis; the salience of internationaleconomic issues in the
seventies led to even furtherdisaggregation,andeven furtherdowngradingof militaryand state-centered
views.
The second image reversed 893
tion of foreign policy was the question. Nye and Keohane, Karl Kaiser, Edward
Morse and othersstressedthe growing role of transnational,internationaland multi-
national actors, and global, non-militaryforces such as technology, trade, com-
munications, and culture, in shaping policy. States were depicted as losing control
over importantissue areas, especially economic ones. Insteadof explaining foreign
policy, which is implicitly state-centered,the emphasis is on explaining "interna-
tional regimes" in various issue areas, and not just the internationalsystem, which
essentially stresses militarypower. Countriesdiffer in these issue areasaccordingto
their "sensitivity" and "vulnerability" in various domains. In their most recent
work, Nye and Keohane call this model "complex interdependence"and explore
the conditionsunderwhich it, ratherthanotherparadigms,is the most applicable.20
Complex interdependencealtersdomestic structuresbecause it entails shifts in
power away from certaingovernmentalinstitutionstowardotherones, or even shifts
outside the governmentto privateactors, or to internationalactors, or other foreign
actors. Policy becomes the outcome of an immense swirl of forces, in which pieces
of governmentbecome componentsalong with companies, unions, pressuregroups,
internationalorganizations,technology and so on. Patternsexist in these outcomes;
the system is not totally anarchic, or at least, unlike the liberals, most of these
authorsdo not wish it to be.
In theirlatest book, Nye and Keohanehave become more cautious. "Complex
interdependence"is not the paradigmof the present, but one model among others,
whose applicabilitymust be empiricallydeterminedcase by case. Keohaneand Nye
accept quite readily that traditionalmodels become more relevanton many issues,
especially those involving conditions of considerable tension between countries,
since military capability partakes more of the realist paradigm. The gains in
applicability which come from these more limited claims are welcome but they
reduce the uniqueness of the interdependenceliterature.
The interdependenceargumenthas been takenfarthestby EdwardMorse in his
recent book, Modernizationand Interdependence.2"Morse sees the two as linked:
modern societies are interdependentones. Hence modernity through interdepen-
dence has alteredthe natureof the internationalsystem so much thatthe "anarchy"
model of sovereign units loses its relevance. All modernsocieties in interdependent
situations acquire certain common political characteristicssuch as strong welfare
pressures,bureaucratization,legitimationproblemswhich increasethe relevance of
domestic politics in foreign policy-makingcomparedto the classic period of diplo-
macy. Thus the internationaland the domestic spheres become more important
while the intermediatelevel, national government, diminishes.
forces. This literaturedoes not deny that the internationaleconomic system con-
strains states, nor that the system affects the content of the policies which they
formulate. Rather it challenges the tendency of some liberals, transnationalists,
Marxists, and dependencistasto make the state wither away.
The leading neomercantilistformulationis RobertGilpin's U.S. Power and the
MultinationalCorporation.22He argues that governmentsof states have and assert
some notion of nationalinterest, be it power, stability, welfare, security, which is
not reducibleto the goals of any one groupor coalition. These governmentshave the
capabilityof acting in a coherent way, at least some of the time, in orderto make
their views prevail over those of other members of the polity. When the state
chooses to act, its power is greaterthanthatof any subunit, including such transna-
tional actors as multinationalcorporations.In general, whenever states assert their
views they are able to prevail over internationalorganizations. Interdependence
derives from state policy, not the other way around;that is, it exists because states
allow it to exist. Should states refuse to do so, the constraining quality of that
interdependencewould be broken. All of these propositionsare truerof some states
than of others and vary accordingto the historicalperiod. "Internationalregimes"
express the configurationof power. If a hegemonic power exists, the international
economy will be open; in a multipolarworld, economic nationalism and protec-
tionism are more likely to prevail. States are constrainedby the internationaleco-
nomic system if they are not the hegemonic power. When there is no hegemonic
power then all states are constrained by the system. Nonetheless, for the neo-
mercantilists,the system leaves some latitude of policy response. At least for the
larger states, the determinationof that response, lies ultimately not in the hands of
private actors but in those of the state.
Marxist writing on the internationaleconomy is generally criticized for an
economic reductionistview of the role of politics and institutions.This is often true
of the most frequently cited literature:Magdoff, Baran, and Sweezy.23 These
writerstend to derive political behaviorfrom strictly constructedeconomic exigen-
cies: the drive to counteractfalling profits, or to obtain resources critical to the
operationof the defense technology of capitalist states, or to export domestic con-
tradictions. The state is the instrumentof the capitalist class.
Some recent Marxist literaturehas sought to set out a far less reductionist
argument about the role of the state. The state is seen as having considerable
autonomyfrom any one sector of the capitalist class or from any narrowlyformu-
lated economic goal such as procuringa particularresourceor protectinga particular
market. The state seeks to preservethe capitalistsystem, a tenet which makes these
authorsMarxists,but in orderto do so it may have to do a greatmany things specific
36StanleyHoffmann, Primacy of World Order: American Foreign Policy Since the Cold War (New
York: McGraw Hill, 1978).
37KarlKaiser, GermanForeign Policy in Transition (London:Oxford U.P., 1968).
900 InternationalOrganization
Revolutions all began with some internationaldisturbancethat overtaxedthe politi-
cal system. For the English, it was the need to fight a war with Scotland;in France,
involvementin the AmericanRevolution;in Russia, the defeats of WorldWarI; for
China, the Japaneseinvasion during World War II. The Civil War was America's
nearestequivalentto a revolutionin its impact upon society and institutions.38The
outbreakand outcome of these revolutionsis unintelligiblewithoutan examination
of internationalfactors. At present, Israel, Lebanon, and the Arab states offer
obvious examples of the impact of these forces.
Thus far argumentshave been presented which discuss the effect of interna-
tional politics on domestic politics. While the argumentsdiffer widely in the type of
relationship posited, in tightness and plausibility, there is certainly enough to
suggest that studentsof comparativepolitics treatdomestic structuretoo much as an
independent variable, underplayingthe extent to which it and the international
system are parts of an interactivesystem.
In this section, I returnto the traditionalquestion: which aspect of domestic
structurebest explains how a countrybehaves in the internationalsphere?The only
type of argument which would render that question unnecessary is one which
derived domestic structurecompletely from the internationalsystem. This is a
thoroughlynonreductionistapproach(in Waltz's language)39and argumentsof that
type are not totally convincing. The internationalsystem, be it in an economic or
politico-military form, is underdetermining.The environment may exert strong
pulls but short of actual occupation, some leeway in the response to that environ-
ment remains. A country can face up to the competition or it can fail. Frequently
more than one way to be successful exists. A purely internationalsystem argument
relies on functionalnecessity to explain domestic outcomes; this is unsatisfactory,
because functional requisites may not be fulfilled. Some variance in response to
externalenvironmentis possible. The explanationof choice among the possibilities
thereforerequires some examinationof domestic politics.
Prussia and Poland, for example, both occupied similar positions in relation-
ship to world economic forces and similarlyvulnerablesecuritypositions. The one
developed a powerful militaryabsolutismwhich eventually conqueredits linguistic
neighbors to form Germany. The other gave rise to the liberum veto, a large
eighteenth century literatureon defective constitutions, and was partitioned.The
difference has to do with internalpolitics. Thus the formationof regime type and
coalition patternrequiresreference to internalpolitics.
38Seethe excellent study by Theda Skocpol, "France, Russia, China: A StructuralAnalysis of Social
Revolutions," ComparativeStudies in Society and History 18 (April 1976): 175-2 10. See also her book
on revolutions to be published by Oxford University Press.
39KennethWaltz, "Theory of InternationalRelations." By non-reductionist,Waltz means an explana-
tion of internationalpolitics at the system level, thirdratherthan second image. Here I am extendingthe
word to distinguishbetween endogenous and exogenous explanationsof regime type.
The second image reversed 901
So do most of the categories used in discussing regime type or domestic
structure.The internationalrelations literaturecontains numerousargumentsabout
the importanceof domestic structure.The debate has centeredaroundwhich aspect
mattersmost: the presence and characterof bureaucracy(Kissinger, Allison, Halpe-
rin); the pressure of the masses on policy making or the lack of such pressure
(Kissinger, Wilson, Lenin);the strengthandautonomyof the state (Gilpin, Krasner,
Katzenstein);the drives of the advancedcapitalist economy (Lenin, Magdoff, Ba-
ran, and Sweezy); the perceptual set of leaders (Jervis, Steinbrunner,Brecher);
nationalstyle (Hoffmann);the logic of industrialdevelopment(Kurth);the character
of domestic coalitions (Gourevitch, Katzenstein);the relative weight of transna-
tional actors in a given polity (Nye and Keohane); the level of modernization
(Morse).40
Having gone throughthe exercise of Part one helps bring out a deficiency in
many of these argumentsor presentformulationsof them. Many argumentsfocus on
process and institutionalarrangementdivorced from politics; on structurein the
sense of procedures, separatefrom the groups and interests which work through
politics; on the formal properties of relationships among groups, ratherthan the
content of the relations among them; on the characterof decisions (consistency,
coherence, etc.) ratherthan the content of decisions. Somehow politics disappears.
Clearly, a careful defense of such broadassertionswould requirean examinationof
each of the argumentsabout which I am critical. There is no space to do that here.
Instead, I shall consider one example, chosen because it deals with foreign eco-
nomic policy, an area in which I have worked. The line of argumentto which I refer
is thatwhich uses as a majorexplanatoryvariablestate strength("strong states" vs.
"weak states," or "state-centeredpolicy networks" vs. "society-centeredpolicy
networks").41
The strong state argumentgoes something like this. In societies with strong
states, or state-centeredpolicy networks,policy-formationcorrespondsto the model
of unitarygovernment:the state, emanatingfrom the public or some other sover-
40Besides works already cited, see: Robert Jervis, Perception and Misperception in International
Politics (Princeton,N.J.: PrincetonU.P., 1976); JohnSteinbrunner,The CyberneticTheoryof Decision:
New Dimensions of Political Analysis (Princeton, N.J.: PrincetonU.P., 1974); Michael Brecher, The
Foreign Policy System of Israel: Setting, Images, Processes (New Haven: Yale U.P., 1972); Michael
Brecher, Decisions in Israel's Foreign Policy (New Haven: Yale U.P., 1975). See also R. Harrison
Wagner, "Dissolving the State: Three Recent Perspectives on InternationalRelations," International
Organization28 (Summer, 1974): 435-466.
41StephenKrasner, Raw Materials Investment and American Foreign Policy (Princeton: Princeton
University Press, forthcoming); Peter Katzenstein, "Introduction" and "Conclusion" to "Between
Power and Plenty: Foreign Economic Policies of Advanced IndustrialStates," InternationalOrganiza-
tion 31 (Autumn 1977) and "InternationalRelations and Domestic Structures:Foreign Economic
Policies of Advanced Industrial States," International Organization 30 (Winter 1976); Stanley
Hoffmann, "The State:For WhatSociety," Decline or Renewal (New York:Viking Press, 1974); Bruce
Andrews, "Surplus Security and National Security: State Policy as Domestic Social Action," Interna-
tional Studies Association, Washington, D.C., February22-26, 1978. John Zysman has some astute
commentsaboutthe connectionbetween institutionalform and the contentof policy towardinternational
competition in his study of the French electronics industry:Political Strategiesfor Industrial Order
(Berkeley: University of CaliforniaPress, 1977).
902 InternationalOrganization
eign, formulatespolicy which is an articulationof collective interests. The state
speaks on behalf of goals broaderthan those of any particulargroup. Its unitary
structureallows it to impose that policy over the objections of particularisticinter-
ests.
In societies with weak states (or society-centered policy networks) policy-
formationcorrespondsto a model of pluralisticgovernment:social forces are well-
organizedand robust. Public institutionsare fragmented;power is formally distrib-
uted among a large number of interdependentbut autonomous agencies. These
pieces of the state are capturedby differentprivateinterests, which are then able to
use them to exercise veto power over public policy or even to acquire a complete
control over public policy in a given domain. Policy is the outcome of the conflict
among these complex public-privatelinkages. The United States is obviously the
most commonly citecdexemplar.
The prevalence of one or the other type of state or networkcan be explained
historically: different state-society relationships prevailed in the process of
modernization. However similar countries have become in other respects, these
differencespersist and remainrelevant. The natureof the networkor state structure
explains, accordingto this school, key aspects of foreign policy. In several remark-
ably comprehensiveessays," Peter Katzenstein argues, for example, that United
States foreign economic policy is less consistent and more dominatedby economic
considerationsthan is French policy, which tends to have more coherence and to
reflect political preoccupationsaboutFrance'sposition in the world. More recently,
he has contrastedthe marketorientationof American, British, and Germanpolicy
with the dirigiste orientationof the French and Japanese.
Lacking space for careful examinationof each of the countries, let me evoke
the problems with such argumentationthrough two examples, one concerning a
"weak state" country, the other concerninga "strong" one. Katzenstein,Krasner,
and other authors propose that in the United States, protectionists' interests are
strongin Congress, while free tradeinterestsare strongerin the executive. The shift
from protectionto free tradeafterWorldWar II thus requiredand was facilitatedby
the shift in power from Congress to the presidency. That the presidency grew in
power and that it favored free tradewhile protectionistssquawkedin CongressI do
not dispute. Was the policy change, though, caused by a shifting balanceof institu-
tions or were both symptomsof somethingelse? The argumentimplies that had the
presidency somehow managed to acquire power earlier there would have been a
shift in United States commercialpolicy. I find thatdubious. The United States and
many other countrieswith widely differing political forms, strong and weak states
alike, pursuedprotectionistpolicies from the 1870s through 1945, with few inter-
ruptions. Would Congressional dominance have prevented the shift to free trade
after World War II? I doubt even that. A great many American interests were
shifting to free trade. Perhapsthe votes in Congress would have been closer, but
would the outcome have been so different?
42Katzenstein,InternationalOrganizationarticles.
The second image reversed 903
For France, the strong state argumentrequiresus to think that it matterslittle
who controls the state; the fact of having a state-centerednetwork becomes more
importantthan the question of the political orientationof the government.Thus the
change from the FourthRepublicto de Gaulle did not matter,this argumentimplies,
since the bureaucracyran the show in both, as presumablyit did under the Third
Republic. Is thatplausible?Would the same be true if the Left had won the legisla-
tive elections of 1978? Would there be no policy change because the strong state
network remains in place, because the state-society balance remains the same? Or
would it change because the state machine was now in the hands of persons with
differentpolicy goals? Even continuityin policy would be hardto interpret.Should
the Left coalition turnout to be prudentand cautious, pursuingpolicies not different
from Giscard's, would that be because the state machine constrainedit, or because
the coalition feared flight of capital, an investor's strike, foreign pressures, labor
militancy, voter discontent:in short, politics of a similar kind, regardless of this
strong-weakstate distinction?The strong state-weak state argumentsuggests that
the type of relationpredominates,hence the identityof the governingcoalition does
not matter. This is a very apolitical argument.
The basic problemwith this line of reasoningis that it provides no explanation
for the orientationof state policy in the supposedly state-dominatedcountries. The
advantageof looking at politics and the state is that it helps us get away from the
well-known problems of pluralist or Marxian reductionism:policy is not simply
traceableto the interests of one or anothergroup. First, powerful groups conflict
among themselves. Second, the interactionamong groups is affected by structures.
Third, politicians and bureaucratswho run the state have some leeway. Hence, the
importanceof politics andthe state. But the notion of a strongstate as presentlyused
escapes from this trap at the cost of heading into another:instead of explaining
society (where the groupsget their orientations,why some are strongerthan others)
we have to explain the state. Why does the state go in one direction ratherthan
another?Why does it articulatea particularconception of the nationalinterestover
another?Why does it use its leverage over particulargroups in some ways and not
others?Why doesn't the Frenchstate use its power to bring aboutworkers'control,
equality of income distribution, stricter pollution control? Why does it support
traditionaland small industries?Why does it also promote concentrationof indus-
try? How does it choose between the claims of small and large enterprises?Any
policy pursuedby the statemust be able to elicit the supportof at least enough social
elements to sustainthe state leaders in power. Hence explanationof the orientation
of state policy requires some examination of the politics behind state action. To
speak of the strong state suggests that politics can be taken out of the equationfor
some states and not for others. The action of the strong state depends as much on
politics as does that of the weak one.
Ignoring politics in fact shapes the discussion in a way those interestedin the
type of state would presumablydislike. The argumentreverts to explanations in-
volving a rationalactor, or at least a unitaryactor model of the state, and towards
realist type reasoningabout the state, in which the state becomes a unitaryactor. If
904 InternationalOrganization
there is no conflict, latentor actual, aboutthe propercourse of policy, and if thereis
no disagreementabout what public power should be used for, then the state or the
networkbecome wholes, respondingin a collective way to externalstimulants.The
analysis of foreign policy is then reducible to the examinationof the international
system. If, on the other hand, conflict within each country exists, then the conse-
quences of having a particulartype of state must be linked with the political struggle
for one or anotherpolicy option. If there is little or no conflict over policy, that too
requiresexplanation. Such constraintcould be derived from a range of causes: the
logic of an internationalsituation(back then to rationalactor analysis), or the logic
of market capitalism (which then requires an analysis of how and why it has
influence, something usually only the neo-Marxistliteratureis interestedin show-
ing).43
I do not wish to argue that the characterof the state structureor the "policy
network" has no impacton policy. On the contrary,by setting down the rules of the
game, institutions reward or punish specific groups, interests, visions, persons.
These effects can be seen case by case, as in the Cuban missile crisis, or over a
whole patternof cases, such as tariff policy. But the impactof structureslies not in
some inherent, self-contained quality, but rather in the way a given structureat
specific historical moments helps one set of opinions prevail over another.
Structureaffects the extent to which a governing coalition must make side-
payments to build up its strength, the extent to which it can impose its views. It
affects the possibility of realizing certainpolicies. Examples abound. The types of
taxes the Italianstate can raise are limited by the weakness of the state bureaucracy;
the types of industrialpolicies Britain can pursue are limited by the fragmented
characterof banking and industry; French dirigismeis surely facilitated by the
position of the grandcorps there; Americanenergy policy is not likely to be some
carefully orchestratedscheme which must get throughCongressin toto or not at all.
In each case, the effect on policy derives from the voice given by structuresto some
point of view: relatively fragmented,or open political systems increasethe number
of veto groups;relativelyunitary,or closed systems, do not eliminateveto groupsor
bargainingbut only limit their range. Whoever controls the state in unitarysystems
has an easier time passing laws, though whetherthis makes the formationof policy
"easier" is not clear. Witness Britain.4
What the strong-weakstate distinctiondoes, along with many other structural
categorizations,is to obscure importantways in which politics, our centralsubject
matter, shapes outcomes. It therebyencouragesvarious forms of reductionism.The
need to secure supportfor a policy affects its final content. Majoritieshave to be
built, coalitions constructed, terms of trade among alliance partnersworked out,
legitimating argumentsdeveloped, and so on. These tasks impose constraintson
48Seethe very interestingand growing neo-Marxistliteratureon the state such as: Fred Bloch, "The
Ruling Class does not Rule: Notes on the MarxistTheory of the State," Social Revolution33 (May-June
1977): 6-28; David Gold, Clarence Lo, and Erik Olin Wright, "Recent Developments in Marxist
908 International Organization
Theories of the CapitalistState," parts 1 and 2, MonthlyReview (Octoberand November 1975); Klaus
Offe, "StructuralProblemsof the CapitalistState," Klaus von Beyme, ed., GermanPolitical Studies,
(Beverly Hills: Sage, 1976); Klaus Offe and Volker Ronge, "Theses on the Theory of the State," New
German Critique 6 (Fall 1975); James O'Connor, Fiscal Crisis of the State (New York: St. Martin's
Press, 1973).
The second image reversed 909
Marlborough(preventinghim from carryingthe war to Paris)andnegotiateda peace
far more beneficial to the French than was thought possible a few years earlier.
What of Prussia, supposed paradigmof the distinction between foreign and
domestic policy? The existence and power of Prussiadependedon militaryorgani-
zation and hence on domesticpolitics. The whole thrustof GordonCraig's excellent
book is to show how repeatedcycles of the army's estrangementfrom the rest of
society gravely weakenedPrussiapolitically and militarily.49Repeatedly,the politi-
cal system and the armyhad to be opened up, new elements let in, and meritocracy
rewarded. Such was the meaning of French dominance between 1789 and 1815:
social organizationchanged militarycapacity. War forced the Prussiansto reform.
Stein Hardenbergmade the defeat of Napoleonpossible. The conflict over the Army
IndemnityBill in 1862 and the subsequentwars leading to Germanunification in
1870 show as thoroughan interpenetrationof constitutional,economic, militaryand
securityissues as one can find before or since. And war among "modem" countries
can hardly be said to have disappeared.
What of internationaltrade?The shift in traderoutes from East to West in the
fifteenth and sixteenth centuriesunderminedtowns, alteredthe social balance, and
contributedto the enserfmentof the entire East Europeanpeasantry. The pull of
Dutch marketsinduced sheep farming in southernEngland, which in turn spurred
the decompositionof feudalism (enclosures, commutationof dues, drivingpeasants
off the land, crumblingof guilds, etc.) which became one of the centralcleavages in
the English Civil War. At the height of the Civil War, despite two centuries of
cooperationbefore and afterwardconcerning opposition to the Catholic-Spanish-
Frenchhegemony, Britainand Hollandtook time to fight a series of naval wars for
control of internationaltrading. One could argue that in some of these examples
policy affected relatively small numbersof people in each country,thoughhardlyin
the case of serfdomvs. the collapse of feudalism. Surely by the nineteenthcentury,
this was no longer true. England'sfactories devastatedthe Indiantextile economy.
What does the international-domesticdistinction mean in the fight over the Corn
Laws, Cobden-Chevalier,or the German, French, Italian, American, Canadian,
Australian,Russian tariffs after 1873?
In other respects the present is not so different from the past. Despite inter-
dependence,the state retainsits ability to controltransnationalactors, if it is able to
musterthe political supportfor doing so. The Soviet Union andChinadependon the
world for many things, but they controlfar more strictly thanthe Westerncountries
the termson which they interact.In the case of Russia surely thatis not because it is
less modernthanthe West, but because its political system is different.The Western
states could do the same but do not.
Bureaucraticpolitics, interdependence,interpenetration,transnationalforces
and actors-seem as relevantfor an analysis of the past as they do for the present.
Louis XIV, Frederickthe Great,andWallensteincould be looked at using these con-
cerns. Internationalanarchy, security, and state power all seem relevant to the
present as well as to the past.
49GordonCraig, The Politics of the Prussian Army (Oxford: Oxford U.P., 1964).
910 InternationalOrganization
I am not suggesting that the presenthas no unique features:the United Nations
and the atom bomb are by far the most distinctive. The latter marks a qualitative
change in internationalrelationsby confusingthe relationshipbetween the existence
of power and its use. The global economy is certainlymore extensive and dramatic
than anythingpreviously witnessed. But is it more thanthe extension of the Roman
model to the whole world? Every period has its distinctive features. Why assert a
discontinuity?The most useful thing the interdependenceschool has done is to work
out models based on principles other than anarchy. Having these allows us to
comparedifferentperiods. Reality will, as usual, prove confusing. It will be hardto
settle argumentsin any clean or definitive way, aboutwhetherwe have moved from
one model to another, and when.
When reality is too confusing to settle arguments, posing a question in the
sociology of knowledge becomes interestingand relevant:the strikingthing is that
very diverse authorswho read very little of each other should be asking similar
questions about widely different periods. Wallerstein wants to know about inter-
dependencein the sixteenth century;Morse about the same thing in the twentieth.
Why are they looking for the same thing? Why doesn't someone do a bureaucratic
politics study of Marlboroughor Napoleon?
The answeris certainlydifferentfor each school of authors,but in both cases it
is markedby deeper issues of value and political outlook. This is not meant as a
criticism, implying that these authorsshould have purgedsuch elements from their
work. Complete value neutralityis unrealizable.Values and the politics connected
to them should be discussed openly, and it is always importantto have a sense of
what values inform various works.
For the interdependence-modernization school, the centralconcern appearsto
be with the dangers of anarchy:one might say that they reject the relevance of
anarchybecause they fear it. Since the world is new, there are new requirementsfor
its maintenance. Leadership is necessary. The United States is the only power
capable of providing it. It should do so. Such leadership is not imperium, since
imperiumis a concept which applies to the unneededhegemony of one power over
others. Now interactionis inescapable;the issue is not whetherit exists, but how it
is to be managed.
For the dependenciaschool, the central concern appearsto be the dangers of
interdependencein its capitalist embodiment. They reject it because it is seen as
incompatible with socialism. Interdependence is linked to capitalism, and
capitalism is old; therefore interdependencegoes back a long way. In its market
forms, it preventssocieties from developing as they see fit, ratherthanaccordingto
the needs of the capitalist system. These theorists worry about the Third World,
althougha subset of them does make the link back to the core; participationin the
system corruptsthe master as well. Interdependencewas an element of capitalism
neglected by Marxianand other theorizingwhich treatedstates and nationaltrajec-
tories of developmentas separate.The explication of the possibilities of socialism,
as well as the dangers, therefore requires analysis of the progress of interdepen-
dence, of the constraintswhich it imposes.
The second image reversed 911
I am personally part of that group which is sympatheticto both concerns. If
hegemony has its nefarious consequences, so does economic and political
nationalism.Leadershipis useful but easily pervertingand pervertible.But this has
always been true. Interdependenceis an old reality, as is anarchy. The argument
ought to be about how interdependent-anarchic situationsdiffer, not whether they
are new.
3
0 ~~~~~~~~~~~!4
u
a 4. Cn Q ^20ve
0
*t $ iM i E R $ S O c X S~~~ci
| u t4. X i E
CN
._ b
-q
~~~~C#~~~~~~~~~~0
~ ~ ON
C)
4mo~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~(
CL.,~~~~~~C
a q t s- 0 a t
C/i~~~~ 3 U i eC S g
4-4 -
PC~~~~~~~~~~~~~~~~~~~~~~~)
M