~

New York COLUMBIA UNIVERSITY PRESS

NEOREALISM AND ITS CRITICS

--- Robert -O~Keohane; editor

PtA~~~~\i«Ef'3~ ~ ~i'fR~ ~li: '?oo~

Kenneth N. Waltz

The trick, obviously, is to link theoretical concepts with a few variables in order to contrive explanations from which hypotheses can then be inferred and tested.

NOTES

I. One must be careful. The above statement is law-like ~nly if it can be verified in various ways. Counterfactual conditions, for example, would have to be met in this way: Person b is in the income category of likely Republicans; if b's income were reduced to a certain level, he would probably become a Democrat. More precisely, the law-like statement establishes these expectations: If b is an R with probability x, and if a is a D withprobabilityj; then-if bbecomes a.: he thereby becomes a D with probability y.

2. The proof is simply presented by Nagel (1961: 116n). One should add that the explanations will not be equally simple and useful.

3. For consideration of testing proce-dures and explanation of their importance, see Stinchcombe (1968 ch. 2).

4. See article 5, part III, for further thoughts about testing.

'. :-:_

: ~ ... ;:'~ . : -s,'

-. -<';

THREE

Reductionist and Systemic Theories

KENNETH N. WALTZ

IN ONE WAY or another, theories of international politics, whether reductionist or systemic, deal with events at all levels, from the subnational to the supranational. Theories are reductionist or systemic, not according to what they deal with, but according to how they arrange their materials. Reductionist theories explain international outcomes through elements and combinations of elements located at national or subnational levels. That internal forces produce external outcomes is the claim of such theories. N ~ X is their pattern. The international system, if conceived of at all, is taken to be merely an outcome.

A reductionist theory is a theory about the behavior of parts. Once the theory that explains the behavior of the parts is fashioned, no further effort is required. According to the theories of imperialism examined in chapter 2 [not reprinted here: ed.], for example, international outcomes are simply the sum of the results produced by the separate states, and the behavior of each of them is explained through its internal characteristics. Hobson's theory (1902), taken as a general one, is a theory about the workings of national economies. Given certain conditions, it explains why demand slackens, why production falls, and why resources are underemployed. From a knowledge of how capitalist economies work, Hobson believed he could infer the external behavior of capitalist states. He made the error of predicting outcomes from attributes. To try to do that amounts to overlooking the difference between these two statements:

"He is a troublemaker" "He makes trouble:' The second statement does not follow from the_first one if the attributes of actors do not uniquely determine outcomes. Just as peacemakers may fail to make peace, so troublemakers may fail to make trouble. From attributes one cannot

z

• • • • :.

~

~

..

Kenneth N. Waltz

predict outcomes if outcomes depend on the situations of the actors as well as on their attributes.

Few, it seems, can consistently escape from the belief that international-political outcomes are determined, rather than merely affected, by what states are like. Hobson's error has been made by almost everyone, at least from the nineteenth century onward. In the earlier history of modem great-power politics, all of the states were monarchies, and most of them absolute ones. Was the power-political game played because of international-political imperatives or simply because authoritarian states are power-minded? If the answer to the latter part of the question were "yes;' then profound national changes would transform internationalpol- . ities. Such changes began to take place in Europe and America most strikingly in 1789. For some, democracy became the form of the state that would make the world a peaceful one; for others, later, it was socialism that would tum the trick. Not simply war and peace; moreover, but international politics in general was to be understood through study of the states and the statesmen, the elites and the bureaucracies, the subnational and the transnational actors whose behaviors and interactions form the substance of international affairs.

Political scientists, whether traditional or modem in orientation, reify their systems by reducing them to their interacting parts. For two reasons, the lumping of historically minded traditionalists and Scientifically oriented modernists together may seem odd. First, the difference in the methods they use obscures the similarity of their methodology, that is, of the lOgiC their inquiries follow. Second, their different descriptions of the objects of their inquiries reinforce the impression that the difference of methods is a difference of methodology. Traditionalists emphasize the structural distinction between domestic and international politics, a distinction that modernists usually deny. The distinction turns on the difference between politics conducted in a condition of settled rules and politics conducted in a condition of anarchy. Raymond Aron, for example, finds the distinctive quality of international politics in "the absence of a tribunal or police force, the right to resort to force, the plurality of autonomous centers of decision, the alternation and continual interplay between peace and war" (1967:192). With this view; contrast J. David Singer's examination of the descriptive, explanatory, and predictive potentialities of two different levels of analysis: the national and the. international (1961). In his examination, he' fails even to mention the contextual difference between organized politics within states and formally unorganized politics among them. If the contextual difference is

.f~-::.

_- t,. t:-

A'

REDUCTION 1ST AND SYSTEMIC THEORIES

49

~.

.:

overlooked or denied, then the qualitative . difference of internal and external politics disappears or never was. And that is indeed the conclusion that modernists reach. The difference between the global system and its subsystems is said to lie not in the anarchy of the former and the formal organization of the latter, but in there being, as Singer puts it, only one international system "on and around the planet Earth" (1969:30). If one believes that, then "the level-of-analysis problem in international relations" is solved by turning the problem into a matter of choice, a choice made according to the investigator's interest (1961:90).

Traditionalists keep harping on the anarchic character of international politics as marking the distinction between internal and external realms, andmodernistsdo not.' If we listen to what members of the two camps sa)\ the gulf between them is wide. If we look at what members of both camps do, methods aside, the gulf narrows and almost disappears. All of them drift to the "subsystem dominant pole?' Their attention focuses on the behaving units. They concentrate on finding out who is doing what to produce the outcomes. When Aron and other traditionalists insist that theorists' categories be consonant with actors' motives and perceptions, they are affirming the preeminently behavioral logic that their inquiries follow. Modernists and traditionalists are struck from the same mold. They share the belief that explanations of international-political outcomes can be drawn by examining the actions and interactions of nations and other actors.

The similarity of traditional and modern approaches to the study of international politics is easily shown. Analysts who confine their attention to interacting units, without recognizing that systemic causes are in play, compensate for the omissions by assigning such causes arbitrarily to the level of interacting units and parcelling them out among actors. The effects of relegating systemic causes to the level of interacting units are practical as well as theoretical. Domestic politics are made into matters of direct international concern. This was clearly shown in 1973 and after when detente became something of an issue in American politics. Could detente, some wondered, survive American pressure on Russian political leaders to govern a little more liberally? Hans Morgenthau, not unexpectedly, turned the argument around. American concern with Russia's internal politics, he claimed, is not "meddling in the domestic affairs of another country. Rather it reflects the recognition that a stable peace, founded upon astable balance of power, is predicated upon a common moral frameworkthat expresses the commitment of all the nations concerned to certain basic moral principles, of which the preservation of

50

Kenneth N. Waltz

REDUCTION 1ST AND SYSTEMIC THEORIES

51

that balance of power is one" (1974:39). If the international-political outcomes are determined by what states are like, then we must be concerned with, and if necessary do something to change, the internal dispositions of the internationally important ones.

As a policymaker, Secretary of State Henry Kissinger rejected Morgenthau's argument. As a political scientist, however, Kissinger had earlier agreed with Morgenthau in believing that the preservation of peace and the maintenance of international stability depend on the attitudes and the internal characteristics of states. Kissinger defined an international order as "legitimate" if it is accepted by all of the major powers and as "revolutionary" if one or more of them rejects it. In contrast to a legitimate order, a revolutionary order is one in which one or more of the major states refuses to deal with other states according to the conventional rules of the game. The quality of the order depends on the dispositions of the states that constitute it. A legitimate international order tends toward stability and peace; a revolutionary international order, toward instability and war. Revolutionary states make international systems revolutionary; a revolutionary system is one that contains one or more revolutionary states (KiSSinger 1957:316-20; 1964:1-6, 145-147; 1968:899). The reasoning is circular, and naturally so. Once the system is reduced to its interacting parts, the fate of the system can be deter- . mined only by the characteristics of its major units. I

Among political scientists, Morgenthau and Kissinger are considered to be traditionalists-scholars turned toward history and concerned more with policy than with theory and scientific methods. The practice in question, however, is common among social scientists of different orientations. Kaplan's reasoning (1957) is Morgenthau's, although Kaplan's vocabulary, borrowed from general-systems theory, has obscured this. Marion Levy, a sociologist who at times writes about international politics, provides another example. He asserts that the "problem foci" of international afTairs "are those of the modernization of the relatively nonmodernized societies and of the maintenance of stability within (and consequently among) the relatively modernized societies" (1966:734).

Inside-out explanations always produce the results that these examples illustrate. Kissinger's saying that international instability and war are caused by the existence of revolutionary states amounts to saying that wars occur because some states are warlike. And yet revolutionary regimes may obey international rules--or, more simply, tend toward peaceful coexistencc--because the pressures of their external situations overwhelm their internally g~nerated aims. Revolutionary international

orders are. at times stable and peaceful. Conversely, legitimate international orders are at times unstable and war prone. Levy's efTort to predict international outcomes from national characteristics leads to similarly i unimpressive results. Saying that stable states make for a stable world

amounts to no more than saying that order prevails if most states are . orderly. But even if every state were stable, the world of states might not be. If each state, being stable, strove only for security and had no designs

· on its neighbors, all states would nevertheless remain insecure; for the means of security for one state are, in their very existence, the means by which other states are threatened. One cannot infer the condition of

· international politics from the internal composition. of states, nor can one

· arrive at an understanding of international politics by summing the for-

· eign policies and the external behaviors of states.

Differences across traditional and modem schools are wide enough to obscure their fundamental Similarity. The similarity, once seen, is striking: . Members of both schools reveal themselves as behavioralists under the skin. Members of both schools ofTer explanations in terms of behaving units while leaving aside the efTect that their situations may have. Veblen (1915) and Schumpeter (1919) explain imperialism and war according to internal social development; Hobson and his vast progeny, by internal economic arrangement. Levy thinks national stability determines international stability. Kaplan declares international politics to be subsystem dominant. Aron says that what the poles of the system are like is more important than how many poles there may be. As scholar, though not as public official, Kissinger identified revolutionary states with internal instability and war. Because he agrees with Kissinger as scholar, Morgenthau advises intervention in the domestic afTairs of other states in the name of international-political necessity. Rosecrance (1963) makes the international system all efTect, and not at all cause, and turns his examination of international politics into a "correlating" of internal conditions and international outcomes and a tracing of sequential efTects. Many modem students spend much of the time calculating Pearsonian coefficients of correlation. This often amounts to attaching numbers to the kinds of impressionistic associations between internal conditions and international outcomes that traditionalists so frequently ofTer. International-political studies that conform to the inside-out pattern proceed by correlational logic, whatever the methods used. Scholars who mayor may not think of themselves as systems theorists, and formulations that seem to be more scientific or less so, follow the same line of reasoning. They examine international politics in terms of what states are like and how they

Kenneth N. llilltz

interact, but not in terms of how they stand in relation to each other. They commit C. F. A. Pantin's "analytic fallacy" by confining their studies to factors that bear on their phenomena without considering that "higher-order configurations may have properties to be studied in their

own right" (1968: 175). .

It is not possible to understand world politics simply by looking inside of states. If the aims, policies, and. actions of states become matters of exclusive attention or even of central concern, then we are forced back to the descriptive level; and from simple descriptions no valid generalizations can logically be drawn. We can say what we see, but we cannot know what it may mean. Every time we think that we see something different or new, we v ViII· have todesigna.te anotherunlt-level "variable" as its cause. If the situation of actors affects their behavior and influences their interactions, then attempted explanation at the unit level will lead to the infinite proliferation of variables, because at that level no one variable, or set of variables, is sufficient to produce the observed result. So-called variables proliferate wildly when the adopted approach fails to comprehend what is causally important in the subject matter. Variables are added to account for seemingly uncaused effects. What is omitted at the systems level isrecaptured-e-if it is recaptured at all-byattributing characteristics, motives, duties, or whatever to the separate actors. The result observed is turned into a cause, which is then assigned to the actors. There is, however, no logically sound. and traceable process by which effects that derive from the system can be attributed to the units. Variables then have to be added subjectively, according to the good or bad judgment of the author. This makes for endless arguments that are doomed to being inconclusive.

In order to take Morgenthau, Kissinger, Levy, and the rest seriously, we would have to believe that no important causes intervene between the aims and actions of states and the results their actions produce. In the history of international relations, however, results achieved seldom correspond to the intentions of actors. Why are they repeatedly thwarted? The apparent answer is that causes not found in their individual characters and motives do operate among the actors collectively. Each state arrives at policies and decides on actions according to its own internal processes, but its decisions are shaped by the very presence of other states as well as by interactions with them. When and. how internal forces. find external expression, if they do, cannot be explained in terms of the interacting

REDUcTlONIST AND SYSTEMIC THEORIES

53

parties if the situation in which they act and interact constrains them from some actions, disposes them toward others, and affects the outcomes of their interactions.

If changes in international outcomes are linked directly to changes in actors, how can one account for similarities of outcome that persist or recur even as actors vary? One who believes that he can account for changes in international politics must also ask how continuities can be explained. International politics is sometimes described as the realm of accident and upheaval, of rapid and unpredictable change. Although changes abound, continuities are as impressive, or more so, a proposition that can be illustrated in a number of ways. One who reads the apoc-

. ryphal book of First Maccabees with events in and after World War I in mind will gain a sense of the continuity that characterizes international politics, Whether in the second century before Christ or in the twentieth century after, Arabs and Jews fought among themselves and over the residues of northern empire, while states outside of the arena warily watched or actively intervened. To illustrate the point more generally, one may cite the famous case of Hobbes experiencing the contemporaneity of Thucydides. Less famous, but equally striking, is the realization by Louis J. Halle of the relevance ofThucydides in the era of nuclear weapons and superpowers (1955, Appendix), In the two world wars of this century, to choose a different type of illustration, the same principal countries lined up against each other, despite the domestic political upheavals that took place in the interwar period. The texture of international politics remains highly constant, patterns recur, and events repeat themselves endlessly. The relations that prevail internationally seldom shift rapidly in type or in quality. They are marked instead by dismaying persistence, a persistence that one must expect so long as none of the competing units is able to convert the anarchic international realm into a hierarchic one.

The enduring anarchic character of international politics accounts for the striking sameness in the quality of international life through the millennia, a statement that will meet with wide assent. Why then do we find such a persistent pull toward reduction? The answer is that usually reduction results not from a scholar's intent but from his errors. The study of interacting units is thought to exhaust the subject, to include all that can be included both at the level of the unit and at the level of the system. Some political scientists claim that a systems perspective draws attention to the relationalaspects of international politics. But

54

Kenneth N. Waltz

interacting states have always been the objects of study. Others say that to complete an analysis done in terms of interacting states one need only add consideration of nonstate actors. They may need to be included, but including them will leave us at the unit level or lower. Interactions occur at the level of the units, not at the level of the system. Like the outcome of states' actions, the implications of interactions cannot be known, or intelligently guessed at, without knowledge of the situation within which interactions occur. The sporadic interactions of states may, for example, be more important than the daily conduct of routine business. The fate of states whose economic and touristic relations are sparse may be closely linked. We know that this holds for the United States and the Soviet

_- Union. We could not reach that conclusion by counting transactions and by measuring the interactions that take place. This does not mean that counting and measuring are useless activities. It does mean that conclusions about the condition of international politics cannot be directly inferred from data about the formal or informal relations of states. In fact, we more often proceed in the opposite direction. We say, for example, that the United States and the Soviet Union, or the United States, the Soviet Union, and China, interact closely because we believe that actions separately taken strongly affect the pair, or the trio, whether or not there are relations to observe and transactions to count. We save ourselves from the absurdity of saying that a low level of observed in- _ teractions between or among certain states indicates the unimportance of their relations by falling back on what we already know.

Continuities and repetitions defeat efforts to explain international politics by following the familiar inside-out formula. Think of the various causes of war discovered by students. Governmental forms, economic systems, social institutions, political ideologies: These are but a few examples of where causes have been found. And yet, though causes are specifically assigned, we know that states with every imaginable variation of economic institution, social custom, and political ideology have fought wars. More strikingly still, many different sorts of organizations fight wars, whether those organizations be tribes, petty principalities, empires, nations, or street gangs. If an indicated condition seems to have caused a given war, one must wonder what accounts for the repetition of wars even as their causes vary. Variations in the quality of the units are not linked directly to the outcomes their behaviors produce, nor are variations in patterns of interaction. Many, for example,havcclaimed that \Vorld War I was caused by the interaction of two opposed and closely balanced coalitions. But the? many have claimed that World War II was

REDUCTION 1ST AND SYSTEMIC THEORIES

55

caused by the failure of some states to right an imbalance of power by combining to counter an alliance in being.

II

Nations change in form and in purpose; technological advances are made; weaponry is radically transformed; alliances are forged and disrupted. These are changes within systems, and such changes help to explain variations in international-political outcomes. In chapter 3 [not reprinted ed.] we.Iound that aspiring systems theorists think of such withinsystem changes as marking shifts from one system to another. Once structure is clearly defined, a task for the next chapter, changes at the level of structure can be kept separate from changes at the level of units. One may wonder, however, whether inadvertent reductions that result in calling unit-level changes structural ones can be remedied by a change of vocabulary. Unfortunately they cannot be. The problem of showing how structural causes produce their effects would be left unsolved.

Low-level explanations are repeatedly defeated, for the similarity and repetition of international outcomes persist despite wide variations in the attributes and in the interactions of the agents that supposedly cause them. How can one account for the disjunction of observed causes and effects? When seeming causes vary more than their supposed effects, we know that causes have been incorrectly or incompletely specified. The repeated failure of attempts to explain international outcomes analytically-that is, through examination of interacting units-strongly signals the need for a systems approach. If the same effects follow from different causes, then constraints must be operating on the independent variables in ways that affect outcomes. One cannot incorporate the constraints by treating them as one or more of the independent variables with all of them at the same level, because the constraints may operate on all of the independent variables and because they do so in different ways as systems change. Because one cannot achieve that incorporation, reduction is not possibly adequate, and an analytic approach must give way to a systemic one. One can believe that some causes of international outcomes are located at the level of the interacting units. Since variations in presumed causes do not correspond very closely to variations in observed

Kenneth N. Wa!cz

outcomes, however, one has to believe that some causes are located at a different level as well. Causes at the level of units and of systems interact, and because they do so explanation at the level of units alone is bound to mislead. If one's approach allows for the handling of both unit-level and systems-level causes, then it can cope with both the changes and the continuities that occur in a system. It can do so, moreover, without proliferating variables and multiplying categories.

From chapter 1 we know how theories are constructed. To construct a theory we have to abstract from reality, that is, to leave aside most of what we see and experience. Students of international politics have tried to get-closer to the reality of international practice and to increase the empirical content of their studies. Natural science, in contrast, has advanced over the millennia by moving away from everyday reality and by fulfilling Conant's previously mentioned aspiration to lower "the degree of the empiricism involved in solving problems" Natural scientists look for simplicities: elemental units and elegant theories about them. Students of international politics complicate their studies and claim to locate more and more variables. The subject matters of the social and natural sciences are profoundly different. The difference does not obliterate certain possibilities and necessities. No matter what the subject, we have to bound the domain of our concern, to organize it, to simplify the materials we deal with, to concentrate on central tendencies, and to Single out the strongest propelling forces.

From the first part of this article, we know that the theory we want to construct has to be a systemic one. What will a systems theory of international politics look like? What scope will it have? What will it be able, and unable, to explain?

Theory explains regularities of behavior and leads one !o expect that the outcomes produced by interacting units will fall within specified ranges. The behavior of states and of statesmen, however, is indeterminate. How can a theory of international politics, which has to comprehend behavior that is indeterminate, possibly be constructed? This is the great unanswered, and many say unanswerable, question of international-political studies. The question cannot be answered by those whose approach is reductionist or behavioral, as we have seen. They try to explain international politics in terms of its principal actors. The dominant behavioral approach to constructing intemational-politlcal theory proceeds by frarn-

" .. :-.

REDUCTION 1ST AND SYSTEMIC THEORIES

57

.: -,

ing propositions about the behavior, the strategies, and the interactions of states. But propositions at the unit level do not account for the phenomena observed at the systems level. Since the variety of actors and the variations in their actions are not matched by the variety of outcomes, we know that systemic causes are in play. Knowing that, we know further that a systems theory is both needed and possible. To realize the possibility requires conceiving of an international system's structure and shmving how it works its effects. We have to bring off the Copernican revolution that others have called for by showing how much of states' actions and interactions, and how much of the outcomes their actions

--- _and interactions produce, can be explained by forces that operate at the level of the system, rather than at the level of the units.

What do I mean by explain? I mean explain in these senses: to say why the range of expected outcomes falls within c~rtain limits; to say why patterns of behavior recur; to say why events repeat themselves, including events that none or few of the actors may like. The structure of a system acts as a constraining and disposing force, and because it does so systems theories explain and predict continuity within a system. A systems theory shows why changes at the unit level produce less change of outcomes than one would expect in the absence of systemic constraints. A theory of international politics can tell us some things about expected international-political outcomes, about the resilience systems may show in response to the unpredictable acts of a varied set of states, and about the expected effects of systems on states.

A theory has explanatory and predictive power. A theory also has elegance. Elegance in social-science theories means that explanations and predictions will be general. A theory of international politics will, for example, explain why war recurs, and it will indicate some of the conditions that make war more or less likely; but it will not predict the oubreak of particular wars. Within a system, a theory explains continuities. It tells one what to expect and why to expect it. Within a system, a theory explains recurrences and repetitions, not change. At times one is told that structural approaches have proved disappointing, that from the study of structure not much can be learned. This is supposedly so for two reasons. Structure is said to be largely a staticconcept and nearly an empty one. Though neither point is quite right, both points are suggesth·e. Structures appear to be static because they often endure for

-'.J"'

'c.:

::;. ~'.

:"".

58

Kennerh N. Walrz

long periods. Even when structures do not change, they are dynamic, . not static, in that they alter the behavior of actors and affect the outcome of their interactions. Given a durable structure, it becomes easy to overlook structural effects because they are repeatedly the same. Thus one expects the same broad range of outcomes to result from the actions of states in an anarchic condition. What continues and repeats is surely not less important than what changes. A constancy of structure explains the recurrent patterns and features of international-political life. Is structure nevertheless an empty concept? Pretty much so, and because it is it gains in elegance and power. Structure is certainly no good on detail. Structural concepts, although they lack detailed content, help to explain

some big, important, and enduring patterns. .

Structures, moreover, may suddenly change. A structural change is a revolution, whether or not Violently produced, and it is so because it gives rise to new expectations about the outcomes that will be produced by the acts and interactions of units whose placement in the system varies with changes in structure. Across systems, a theory explains change. A theory of international politics can succeed only if political structures are defined in ways that identify their causal effects and show how those effects vary as structures change. From anarchy one infers broad expectations about the quality of international-political life. Distinguishing between anarchic structures of different type permits somewhat narrower and more precise definitions of expected outcomes.

Consider, for example, the effects on European states of the shift from a multipolar to a bipolar system. So long as European states were the world's great powers, unity among them could onlybe dreamt of. Politics among the European great powers tended toward the model of a zerosum game. Each power viewed another's loss as its own gain .. Faced with the temptation to cooperate for mutual benefit, each state became wary and was inclined to draw back. When on occasion some of the great powers did move toward cooperation, they did so in order to oppose other powers more strongly. The emergence of the Russian and American superpowers created a situation that permitted wider ranging and more effective cooperation among the states of Western Europe. They became consumers of security, to use .an expressioncommonin the days of the League of Nations. For the first time in modem history, the determinants

;;.
:;;
"i'
-r-.:-
- ~~-
~~;".;
? ...
'.~
';.. :
~:.. e, , .

..

f

REDUCTION 1ST AND SYSTEMIC THEORIES

59

of war and peace lay outside the arena of European states, and the means of their preservation were provided by others. These new circumstances made possible the famous "upgrading of the common interest;' a phrase which conveys the thought that all should work together to improve everyone's lot rather than being obsessively concerned with the precise division of benefits. Not all impediments to cooperation were removed, but one important one was-the feat that the greater advantage of one would be translated into military force to be used against the others. Living in the superpowers' shadow, Britain, France, Germany, and Italy quickly saw that war among them would be fruitless and soon began to believe it impossible . Because the security of allof them came to depend

ultimately on the policies of others, rather than on their own, unity could effectively be worked for, although not easily achieved.

Once the possibility of war among states disappears, all of them can more freely run the risk of suffering a relative loss. Enterprises more beneficial to some parties than others can be engaged in, partly in the hope for the latter that other activities will reverse the balance of benefits, and partly in the belief that overall the enterprise itself is valuable. Economic gains may be granted by one state to another in exchange for expected political advantages, including the benefit of strengthening the structure of European cooperation. The removal of worries about security among the states of Western Europe does not mean the termination of conflict; it does produce a change in its content. Hard bargaining within the European Economic Community (by France over agricultural policies, for example) indicates that governments do not lose interest in who will gain more and who will gain less. Conflicts of interest remain, but not the expectation that someone will use force to resolve them. Politics among European states became different in quality after World War II because the international system changed from a multipolar to a bipolar one. The limited progress made in economic and other ways toward the unity of Western Europe cannot be understood without considering the effects that followed from the changed structure of international politics. The example helps to show what a theory of international politics can and cannot tell us. It can describe the range of likely outcomes of the actions and interactions of states within a given system and show how the range of expectations varies as. systems change. Itcan tell us what

60

Kenneth N. Waltz

pressures are exerted and what possibilities are posed by systems of different structure, but it cannot tell us just how, and how effectively, the units of a system will respond to those pressures and possibilities.

Structurally we can describe and understand the pressures states are subject to. We cannot predict how they will react to the pressures without knowledge of their internal dispositions. A systems theory explains changes across systems, not within them, and yet international life within a given system is by no means all repetition. Important discontinuities occur. If they occur within a system that endures, their causes are found at the unit level. Because something happens that is outside a theory'spurviewa deviatlonfromthe expected occurs:

A systems theory of international politics deals with the forces that are in play at the international, and not at the national, level. This question then arises: With both systems-level and unit-level forces in play, how can one construct a theory of international politics without simultaneously constructing a theory of foreign policy? The question is exactly like asking how an economic theory of markets can be written in the absence of a theory of the firm. The answer is "very easily:' Market theory is a structural theory showing how firms are pressed by market forces to do certain things in certain ways. Whether and how well they will do them varies from firm to firm, with variations depending on their different internal organization and management. An internationalpolitical theory does not imply or require a theory of foreign policy any more than a market theory implies or requires a theory of the firm. Systems theories, whether political or economic, are theories that explain how the organization of a realm acts as a constraining and disposing force on the interacting units within it. Such theories tell us about the forces the units are subject to. From them, we can infer some things about the expected behavior and fate of the units: namely. how they will have to compete with and adjust to one another if they are to survive and flourish. To the extent that dynamics of a system limit the freedom of its units, their behavior and the outcomes of their behavior become predictable. How do we expect firms to respond to differently structured markets, and states to differently structured international-political systems? These theoretical questions require us to take firms as firms, and states as states, without paying attention to differences among them. The questions are then answered by reference to the placement of the units

t:"

REDUCTION 1ST AND SYSTEMIC THEORIES

61

in their system and not by reference to their internal qualities. Systems theories explain why different units behave similarly and, despite their variations, produce outcomes that fall within expected ranges. Conversely, theories at the unit level tell us why different units behave differently despite their similar placement in a system. A theory about foreign policy is a theory at the national level. It leads to expectations about the responses that dissimilar polities will make to external pressures. A theory of international politics bears on the foreign policies of. nations while claiming to explain only certain aspects of them. It can tell us what international conditions national policies have to cope with. To think thatatheory of international politics can in itself say how the coping is likely to be done is the opposite of the reductionist error.

The theory, like the story, of international politics is written in terms of the great powers of an era. This is the fashion among political scientists as among historians, but fashion does not reveal the reason lying behind the habit. In international politics, as in any self-help system, the units of greatest capability set the scene of action for others as well as for themselves. In systems theory, structure is a generative notion; and the structure of a system is generated by the interactions of its principal parts. Theories that apply to self-help systems are , .. Titten in terms of the systems' principal parts. It would be as ridiculous to construct a theory of international politics based on Malaysia and Costa Rica as it would be to construct an economic theory of oligopolistic competition based on the minor firms in a sector of an economy. The fates of all the states and of all the firms in a system are affected much more by the acts and the interactions of the major ones than of the minor ones. At the tum of the century, one who was concerned with the prospects for international politics as a system, and for large and small nations within it, did not concentrate attention on the foreign and military policies of Switzerland, Denmark, and Norway, but rather on those of Britain and Germany, of Russia and France. To focus on great powers is not to lose Sight of lesser ones. Concern with the latter's fate requires paying most attention to the former. Concern with international politics as a system requires concentration on the states that make the most difference. A general theory of international politics is necessarily based on the great powers. The theory once written also applies to lesser states that interact insofar as their interactions are insulated from the intervention of the

62

&nneth N. Waltz

, ,

~.

'!_.

great powers of a system, whether by the relative indifference of the latter or by difficulties of communication and transportation.

III

1-' ; .

"

In a systems theory, some part of the explanation of behaviors and outcomes is found in the system's structure. A political structure is akin to a field of forces in physicS: Interactions within a field have" properties different from those they would have if they occurred outside of it, and as the field affects the objects, so the objects affect the field. How can one give clear and useful political meaning to such aconcept as structure? How do structures work their effects? In considering structures as causes, it is useful to draw a distinction between two definitions.

The term "structure" is now a social-science favorite. As such, its meaning has become all inclusive. In meaning everything, it has ceased to mean anything in particular. Its casual and vacuous uses aside, the term has two important meanings. First, it may designate a compensating device that works to produce a uniformity of outcomes despite the variety of inputs. Bodily organs keep variations' within tolerable ranges despite changes of condition. One's liver, for example, keeps the blood-sugar level within a certain range despite the variety of food and drink ingested. Similarly, negative and progressive income taxes narrow disparities of income despite variations in people's skill, energy, and luck. Because such structures bring leveling processes into play, those who experience the leveling effects need be aware neither of the structure nor of how its effects are produced. Structures of this sort are agents or contrivances that work within systems. They are structures of the sort that political scientists usually have in mind. They do share one quality with structures as I shall define them: They work to keep outcomes within narrow ranges. They differ in being designed by nature or man to operate for particular purposes within larger systems. When referring to such devices, I use' terms such as agent, agency, and compensating device. I use the word "structure" only in its second sense.

In the second Sense structure designates' a set of constraining conditions. Such a structure acts as a selector, but it cannot be seen, examined, and observed at work as livers and income taxes can be. Freely

I'

r

\ ~, :i. ~

.... _,_.

~: ..

"

'f; $;'

REDUCTIONIST AND SYSTEMIC THEORIES

formed economic markets and international-political structures are selectors, but they are not agents. Because structures select by rewarding some behaviors and punishing others, outcomes cannot be inferred from intentions and behaviors. This is simple logic that everyone will understand. What is not so simple is to say just what it is politically that disjoins behavior and result. Structures are causes, but they are not causes in the sense meant by saying that A causes X and B causes Y. X and Y are different outcomes produced by different actions or agents. A and B are stronger, faster, earlier, or weightier than X and Y. By observing the values of variables, by calculating their coyaria~ce, and by tracing sequences.isuch c<ltls.e~.<lre6J(edLI3ecauseAandB. are different, they

- produ~edifferent effects. In contrast, structures limit and mold agents and agencies and point them in ways that tend toward a common quality of outcomes even though the efforts and aims of agents and agencies vary. Structures do not work their effects directly. Structures do not act as agents and agencies do. How then can structural forces be understood? How can one think of structural causes as being more than vague social

propensities or ill-defined political tendencies? "wlf""',:,

Agents and agencies act; systems as a whole do not. But the actions of agents and agendes are affected by the system's structure. In itself a structure does not directly lead to one outcome rather than another. Structure affects behavior within the system, but does so indirectly. The effects are produced in two ways: through socialization of the actors and through competition among them. These two pervasive processes occur in international politics as they do in societies of all sorts. Because they are fundamental processes, I shall risk stating the obvious by explaining each of them in elementary terms.

Consider the process of socialization in the simplest case of a pair of persons, or for that matter of firms or of states. A influences B. B, made different by A's influence, influences A. As Mary Parker Follett, an organization theorist, put it: "A's own activity enters into the stimulus which is causing his activity" (1941: 194). This is an example of the familiar structural-functional logic by which consequences become causes (cf. Stinchcombe 1968:80-101). B's attributes and actions are affected by A, and vice versa. Each is not just influencing the other: both are

" being influenced by the situation their interaction creates. Extending the example makes the logic clearer. George and Martha, the principal char-

That which is George or Martha, individually, does not explain what is compounded between them, nor how To break this whole into individual personality traits ... is essentially to separate them from each other, to deny that their behaviors have special meaning in the context of this interaction-that in fact the pattern of the interaction perpetuates these (1967:156).

Kmnetl: N. Waltz

acters in Edward Albee's play, Who's Afraid if Virginia Woo!!? through their behavior and interaction create a condition that -neither can ··control by individual acts and decisions. In a profound study of Albee's play, Paul Watzla\\1ck and his associates show that George's and Martha's activities cannot be understood without considering the system that emerges from their interactions. They put it this way:

The behavior of the pair cannot be apprehended by taking a unilateral view of either member; The behavior of the pair cannot, moreover, be resolved into a set of two-way relations because each element of behavior that contributes to the interaction is itself shaped by their being a pair. They have become parts of a system. To say simply that George and Martha are interacting, with the action of one eliciting a response from the other, obscures the circularity of their interactions. Each acts and reacts to the other. Stimulus and response are part of the story. But also the two of them act together in a game, which-no less because they have "devised" it-motivates and shapes their behavior.· Each is playing a game, and they are playing the game together. They react to each other and to the tensions their interactions produce.

These are descriptions and examples of what we all know and experience. One may firmly intend to end an argument, may announce the intention, may insist on it, and yet may be carried along by the argument. One may firmly predict one's action and yet be led to act in"\vays that surprise onself as well as oth~rs. Years ago, Gustave Le Bon said this about the effect of the group on the individual:

The most striking peculiarity presented by a psychological crowd is the follOWing: Whoever be the individuals that compose it, however like or unlike be their mode of life, their-occupations, their character, or their

. intelIigence,t~e fact that they have been transformed into a crowd puts them in possession of a sort of collective mind which makes them feel think, and act in a manner. quite different from that in which each

REDUCTIO:-.lIST AND SYSTEMIC THEORIES

individual of them would feel, think, and act were he in a state of isolation {I 896:29-30).

We do not cease to be ourselves when situations strongly affect us, but we become ourselves and something else as well. We become different, but we cannot say that any agent or agency caused us to do so.

Pairs and crowds provide microcosmic and transitory examples of the socialization that takes place in organizations and in societies on larger scales and over longer periods. Nobody tells all of the teenagers in a given school or town to dress alike, but most of them do. They do so, indeed, despite the fact that many people--their parents-are ordinarily telling them not to. In spontaneous and informal ways, societies establish norms of behavior. A group's opinion controls its members. Heroes and leaders emerge and are emulated. Praise for behavior that conforms to group norms reinforces them. Socialization brings members of a group into conformity with its norms. Some members of the group will find this repressive and incline toward deviant behavior. Ridicule may bring deviants into line or cause them to leave the group. Either way the group's homogeneity is preserved. In various ways, societies establish norms and encourage conformity. Socialization reduces variety. The differe~ces of society's members are greater than the differences in their observed behavior. The persistent characteristics of group behavior result in one part from the qualities of its members. They result in another part from the characteristics of the society their interactions produce.

The first way in which structures work their effects is through a process of socialization that limits and molds behavior. The second way is through competition. In social sectors that are loosely organized or segmented, socialization takes place within segments and competition takes place among them. Socialization entourages similarities of attributes and of behavior. So does competition. Competition generates an order, the units of which adjust their relations through their autonomous decis ions and acts. Adam Smith published The I#alth if Nations in 1776. He did not claim to explain economic behavior and outcomes only from then onward. He did not develop a theory that applies only to the economic activities of those who read, understand, and follow his book . His economic theory applies wherever indicated conditions prevail, and it applies aside from the state of producers' and consumers' kriowledge. J

Kenneth N. Walcz

66

This is so because the theory Smith fashioned deals with structural constraints. Insofar as selection rules, results can be predicted. whether or not one knows the actors' intentions and whether or not they under-

. stand structural constraints. Consider an example. Suppose I plan to open a shoe store. Where should I put it? I might notice that shoe stores tend to cluster. Following common political-science reasoning, I would infer either that towns pass laws regulating the location of shoe stores or that shoe-store owners are familiar with the location theory of economists, which tells them generally how to locate their stores in order to _ catch the attention of the largest number of shoppers. Neither inference is justifi~~. Following. common economic reasoning, Lwould say_that market conditions reward those who wittingly or not place their stores in the right places and punish those who do not. Behaviors are selected for their consequences. Individual entreprenuers need not know how to increase their chances of turning a profit. They can blunder along, if they wish to, and rely on the market selector to sort out the ones who happen to operate intelligently from those who do not.

Firms are assumed to be maximizing units. In practice, some of them may not even be trying to maximize anything. Others may be trying, but their ineptitude may make this hard to discern. Competitive systems are regulated, so to speak, by the "rationality" of the more successful competitors. What does rationality mean? It means only that some do better than others-whether through intelligence, skill, hard work, or dumb luck. They succeed in providing a wanted good or service more attractively and more cheaply than others do. Either their competitors emulate them or they fall by the wayside. The demand for their product shrinks, their profits fall, and ultimately they go bankrupt. To break this unwanted chain of events, they must change their ways. And· thus the units that survive come to look like one another. Patterns are formed in the location of firms, in their organization, in their modes of production, in the design of their products, and in their marketing method~. The orderliness is in the outcomes and not necessarily in the inputs. Those who survive share certain characteristics. Those who go bankrupt lack them. Competition spurs the actors to accommodate their ways to the socially most acceptable and successful practices. Socialization and competition are two aspects of a process by which the varietyof behaviors and of outcomes is reduced.

.~ ~~~~" ~. ~:..;

REDUCfIONIST AND SYSTEMIC THEORIES

Where selection according to consequences rules, patterns emerge and endure without anyone arranging the parts to form patterns or striving to maintain them. The acts and the relations of parties may be regulated through the accommodations they mutually make. Order may prevail without an orderer; adjustments may be made without an adjuster; tasks may be allocated without an allocator. The mayor of New York City does not phone the truck gardeners of southern New Jersey and tell them to grow more tomatoes next year because too few were recently supplied. Supply and demand are more sensitively and reliably adjusted through the self-interested responses of numerous buyers and sellers than they

are_by mayors' instructions. An example ofa somewhat different sort is provided by considering Montesquieu's response when presented with a scheme for an ideal society. "Who;' he is said to have asked, "will empty the chamber pots?" As an equivalent question, we might ask: Who will collect the trash? The buyers of the trash-collecting service want to buy the service cheaply. The sellers want to sell their service dearly. What happens? Cities take steps to make the trash detail more attractive: cleaner and simpler through moves toward automation, and socially more acceptable through increasing the status of the job, for example, by providing classy uniforms for the workers. Insofar as trash collecting remains unattractive, society pays more in relation to the talents required than it does for other services. The real society becomes hard to distinguish from the ideal.

IV

·O::·~i:r~:

Different structures may cause the same outcomes to occur even as units

and interactions vary. Thus throughout a market the price of any good or service is uniform if many firms compete, if a few oligopolists engage in collusive pricing, or if the government controls prices. Perfect competition, complete collusion, absolute control: These different causes produce identical results. From uniformity of outcomes one cannot infer that the attributes and the interactions of the parts of a system have remained constant. Structure may determine outcomes aside from changes at the level of the units and aside from the disappearance of some of them and the emergence of others. Different "causes" may

I'

, - ,

lSI - ...•.. -.- ....••.. -

-

~

68

Kenneth N. WaJrz

REDUCfIONIST AND SYSTEMIC THEORIES

It is a concept that takes different values, a concept developed as part of a highly simplified model of some part of the world. _

3. In saying that the theory applies, I leave aside the question of the theory's validity;

produce the same effect; the same "causes" may have different consequences. Unless one knows how a realm is organized, one can hardly tell the causes from the effects.

The effect of an organization may predominate over the attributes and the interactions of the elements within it. A system that is independent of initial conditions is said to display equifinality. If it does, "the system is chen irs own best explanation," and the study of its present organization the appropriate methodology" (Watzlawick et 01. 1967: 129; cf. P: 32). If structure influences without determining, then one must ask how and to what extent the structure of a realm accounts for outcomes and how and to what extent the units accoul1tfor olltcomes.Stlllcturchastobe studied in its own right as do units. To claim to be follO\\ing a systems approach or to be constructing a systems theory requires one to show how system and unit levels can be distinctly defined. Failure to mark and presen-e the distinction between structure, on the one hand, and units and processes, on the other, makes it impossible to disentangle causes of different sorts and to distinguish between causes and effects. Blurring the distinctiori between the different levels of a system has, I believe, been the major impediment to the development oftheories about international politics. The next chapter shows how to define political structures in a way that makes the construction of a systems theory possible.

NOTES

I. What Kissinger learned as a statesman is dramatically different from the conclusions he had reached as a scholar. Statements revealing his new views abound, but one example will suffice. When interviewed while Secretary of State by William E Buckley Jr., Kissinger made the following points in three successive par-agraphs: "Communist societies are morally, in their internal structure, not acceptable to us .•. :' Though our and their ideologies continue to be incompatible, we can nevertheless make practical and peace-preserving accommodations in our foreign policy. We should, indeed, "avoid creating the illusion that progress on some foreign policy questions ... means that there has been a change in the domestic structure" (September 13, 1975, p. 5).

The link between internal attributes and external results is not seen as an unbreakable one. Internal conditions and commitments no longer determine the quality of international life.

2. A variable, contrary to political-science usage, is not just anything that varies.

FOUR Political Structures

KENNETH N. WALTZ

"'\ X TE LEARNED in chapters 2, 3, and 4 [article 3 here] that inter-V V national-political outcomes cannot be . explained reductively.We. found in chapter 3 that even avowedly systemic approaches mingle and confuse systems-level with unit-level causes. Reflecting on theories that follow the general-systems model, we concluded at once that international politics does not fit the model closely enough to make the model useful and that only through some sort of systems theory can international politics be understood. To be a success, such a theory has to show how international politics can be conceived of as a domain distinct from the economic, social, and other international domains that one may conceive of. To mark international-political systems ofT from other international systems, and to distinguish systems-level from unit-level forces, requires showing how political structures are generated and how they afTect, and are affected by, the units of the system. How can we conceive of international politics as a distinct system? What is it that intervenes between interacting units and the results that their acts and interactions produce? To answer these questions, this chapter first examines the concept of social structure and then defines structure as a concept appropriate for national and for international politics.

A system is composed of a structure and of interacting units. The structure is the system-wide component that makes it possible to think of the system as a whole. The problem, unsolved by the systems theorists considered in chapter 3 [not reprinted here: ed.], is to contrive a definition of structure free of the attributes and the interactions of units.

POLITICAL STRUcrURES

71

Definitions of structure must leave aside, or abstract from, the characteristics of units, their behavior, and their interactions. Why must those obviously important matters be omitted? They must be omitted so that we can distinguish between variables at tilt' 1",,·1 of the units and variables at the level of the system. The problem is t9 develop theoretically useful concepts to replace the vague and varying systemic notions that are customarily employed-notions such as environment, situation, context, and milieu. Structure is a useful concept if it gives clear and fixed meaning to such vague and varying terms ..

We know what we have to omit from any definition of structure if the definition is to be useful theoretically. Abstracting from the attributes. of ~nits means leaVing aside questions about the kinds of political leaders, SOCIal and economic institutions, and ideological commitments states may have. Abstracting from relations means leaving aside questions about the cultu~l, economic, political, and military interactions of states. To say w~at ~ ~o be left out does not indicate what is to be put in. The negative pomt IS Important nevertheless because the instruction to omit attributes is often violated and the instruction to omit interactions almost always goes unobserved. But if attributes and interactions are omitted, what is left? The question is answered by considering th~ double meaning of the t~~ "r.elation~' As S. F. Nadel points out, ordinary language obscures a distinction that is important in theory. "Relation" is used to mean both the interaction of units and the positions they occupy vis-a-vis each other (1_957:8-11). To define a structure requires ignoring how units relate. WIth ~ne another (how they interact) and concentrating on how they stand m relation to one another (how they are arranged or positioned). Interactions, as I have insisted, take place at the level of the units. How units stand in relation to one another, the way they are arranged or positioned, is not a property of the units. The arrangement of units is a property of the system.

By leaving aside the personality of actors, their behavior, and th~ir interactions, one arrives at a purely positional picture of society. Three propositions follow from this. First, structures may endure while personality, behavior, and interactions vary widely. Structure is sharply distinguished from actions and interactions. Second, a structural defi~ition

. applies to realms of Widely clifTerel1t substance so long as the arrangement - of parts is similar (cf Nadel: I 04-109). Third, because this is so, theories

Kenneth N. Walcz

72

developed for one realm may with some modification be applicable to other realms as well.

A structure is defined by the arrangement of its parts. Only changes of arrangement are structural changes. A system is composed of a structure and of interacting parts. Both the structure and the parts are concepts, related to, but not identical with, real agents and agencies. Structure is not something we see. The anthropologist Meyer Fortes put this well. "When we describe structure;' he said, "we are in the realm of grammar and syntax, not of the spoken word. We discern structure in the 'concrete reality' of social events only by virtue of having first established structure byabstraction __ from 'concrete r~ality'" (~ortes 1949:56). Since structure is an abstraction, it cannot be defined by enumerating material characteristics of the system. It must instead be defined by the arrangement of the system's parts and by the principle of that arrangement.

This is an uncommon way to think of political systems, although

structural notions are familiar enough to anthropologists, to economists, and even to political scientists who deal with political systems in general but with such of their parts as political parties and bureaucracies. In defining structures, anthropologists do not ask about the habits and the values of the chiefs and the Indians; economists do not ask about the organization and the efficiency of particular firms and the exchanges among them; and political scientists do not ask about the personalities and the interests of the individuals occupying various offices. They leave aside the qualities, the motives, and the interactions of the actors, not because those matters are uninteresting or unimportant, but because they want to know how the qualities, the motives, and the interactions of tribal units are affected by tribal structure, how decisions of firms are influenced by their market, and how people's behavior is molded by the. offices they hold.

"

The concept of structure is based on the fact that units differently juxtaposed and combined behave differently and in interacting produce

POLITICAL STRUCTURES

73

different. outcomes. I first want to show how internal political structure can be defined. In a book on international-political theory, domestic political structure has to be examined in order to draw a distinction between expectations about behavior and outcomes in the internal and external realms. Moreover, considering domestic political structure now will make the elusive international-political structure easier to catch later on.

Structure defines the arrangement, or the ordering, of the parts of a system. Structure is not a collection of political institutions but rather the arrangement of them. How is the arrangement defined? The constitution of a stated_escribes some. parts of the arrangement, but political structures as they develop are not identical with formal constitutions. In defining structures, the first question to answer is this: What is the principle by which the parts are arranged?

Domestic politics is hierarchically ordered. The units-institutions and agencies-stand vis-a-vis each other in relations of super- and subordination. The ordering principle of a system gives the first, and basic, bit of information about how the parts of a realm are related to each other. In a polity the hierarchy of offices is by no means completely articulated, nor all all ambiguities about relations of super- and subordination removed. Nevertheless, political actors are formally differentiated according to the degrees of their authority, and their distinct functions are specified. By "specified" I do not mean that the law of the land fully describes the duties that different agencies perform, but only that broad. agreement prevails on the tasks that various parts of a government are to undertake and on the extent of the power they legitimately wield. Thus Congress supplies the military forces; the President commands them. Congress makes the laws; the executive branch enforces them; agencies administer laws; judges interpret them. Such specification of roles and differentiation of functions is found in any state, the more fully so as the state is more highly developed. The specification of functions of formally differentiated parts gives the second bit of structural information. This second part of the definition adds some content to the structure, . but only enough to say more fully how the units stand in relation to one another. The roles and the functions of the British Prime Minister and Parliament, for example, differ from those of the American

74

Kenneth N. Waltz

, ..

i I'

Ii L i

President and Congress. When offices are juxtaposed and functions are combined in different ways, different behaviors and outcomes result, as I shall shortly show.

The placement of units in relation to one another is not fully defined by a system's ordering principle and by the formal differentiation of its parts. The standing of the units also changes with changes in their relative capabilities. In the performance of their functions, agencies may gain capabilities or lose them. The relation of Prime Minister to Parliament and of President to Congress depends on, and varies with, their relative capabilities. The third part of the definition of structure acknowledges that even while specified functions remain unchanged, units come to stand in different relation to each other through changes in relative capability.

A domestic political structure is thus defined, first, according to the principle by which it is ordered; second, by specification of the functions of formally differentiated units; and third, by the distribution of capabilities across those units. Structure is a highly abstract notion, but the definition of structure does not abstract from everything. To do so would be to leave everything aside and to include nothing at all. The three-part definition of structure includes only what is required to show how the units of the system are positioned or arranged. Everything else is omitted. Concern for tradition and culture, analysis of the character and personality of political actors, consideration of the conflictive' and accommodative processes of politics, description of the making and execution of policy-all such matters are left aside. Their omission does not imply their unimportance. They are omitted because we want to figure out the expected effects of structure on process and of process on structure. That can be done only if structure and process are distinctly defined.

Political structures shape political processes, as can best be seen by comparing different governmental systems. In Britain and America legislative and executive offices are differently juxtaposed and combined. In England they are fused; in America they are separated and in many ways placed in opposition to each other. Differences in the distribution of power and authority among formal and informal agencies affect the chief executives' power and help to account for persistent differences in their performance. I have shown elsewhere how structural differences explain

1:: ..

POLITICAL STRUCTURES

75

contrasts in the patterns of British and American political behavior: Repeating a few points in summary form will make preceding definitional statements politically concrete. I shall take just political leadership as an example and concentrate more on Britain than on America so as to be able to go into some small amount of detail (1967; I draw mainly on chapters 3 and 11).

Prime Ministers have been described, at least since the late nineteenth century, as gaining ever more power to the point where one should no longer. refer to parliamentary or even to cabinet government. The Prime Minister alone now carries the day, or so one is told. One must then wonder why these increaSingly strong Prime Ministers react so slowly to

-- .. events, do the same ineffective things over and over agairi; and iii general govern so weakly. The answers are not found in the different personalities of Prime Ministers, for the patterns I refer to embrace all of them and . extend backward to the 1860s, that is, to the time when the discipline of parties began to emerge as a strong feature of British governance. The formal powers of Prime Ministers appear to be ample, and yet their behavior is more closely constrained than that of American Presidents. The constraints are found in the structure of British government, especially in the relation of leader to party. 1\\'0 points are of major importance: the way leaders are recruited and the effect of their having to manage their parties so carefully.

I~ both countries, directly or indirectly, the effective choice of a chief .. executive lies between the leaders of two major parties. How do they . become the two from whom the choice is made? An MP becomes leader

of his party or Prime Minister by long service in Parliament, by proving

his ability in successive steps up the ministerial ladder, and by displaying the qualities that the House of Commons deems important. The members

of the two major parliamentary parties determine who will rise to the highest office. They select the person who will lead their party when it

is out of power and become Prime Minister when it is triumphant. The MP who would be Prime Minister must satisfy his first constituents, the members of his party who sit in the Commons, that he would be competent and, according to the lights of the party, safe and reliable in office. They will look for someone who has shown over the years that

he wiII displease few of his fellow MPs. Given no limits on length of

Kenneth N. llilJcz

service as Prime Minister, MP's will, moreover, be reluctant to support a younger person, whose successful candidacy might block the road to

the highest office for decades. .. .

Like most countries of settled political institutions, the British apprentice their rulers. The system by which Britain apprentices her rulers is more likely than America's quite different system to produce ~ot only older chief executives but also ones who are safer and surer. Since the Second Reform Act, in 1867, Britain has had 20 Prime Ministers. Their average age in office is 62 years. Their average service in Parliament prior to becoming Prime Ministeris 28 years, during which time they served their apprenticeships in various high Cabinet posts. In. England the one way.of attaining the highest office-tsto-climb -the ministerial ladder. 1_ Since the Civil War, America has had 22 Presidents. Their average age in office is 56 years.! Since Congress is not a direct route to executive preferment, it is pointless to compare congressional with parliam~ntary service. It is, however, safe and significant to say that the Presidency draws on a wider field of experience, occasionally-as with Grant and Eisenhower-on a field not political at all.

The British mode of recruitment creates a condition that serves as a gross restraint on executive power. The Prime ~inister, insofar as ~e has. great powers, is likely to be of an age and experience, a wo~ldly .wlsdom if you like, that makes his exercising them with force and vigor Improbable. If it is true that England muddles through, here is part of the

explanation, a bigger part than the oft-cited national character to w~ich ideological commitment and programmatic politics are supposedly allen.

The limitations that come to bear on Prime Ministers in the very process by ~\'hich they are selected are as important as they are subtle, elusive, and generally overlooked. These qualities also characterize the limitations that derive from the Prime Minister's relation to his party and to Parliament, where his strength is often thought to be greatest. The situation in the two countries can be put as follows: The President can lead but has trouble getting his party to follow; the Prime Minister has the followers but on condition that he not be too far in front of, or to the side of, his party, which makes it difficult for him to lead. The requisite art for a Prime Minister is to manage the party in ways that amid the defiance of the many or the rebellion of the fC\v,ifthose few areirnport~~t, rather than to levy penalties after rebellion has occurred.

.: " .;.

~. .

I.',

I ~ :

POLITICAL STRUCTIlRES

77

.»,

"'-

Most often the Prime Minister's worry is less that some members wiII 3efy him than that his real and effective support will dwindle in the years between general elections, as happened to Churchill and Macmillan in their last governments, and even more obviously to Eden and Heath. It is wrong to see the parliamentary party as a brake on the government only when the party is split and the Prime Minister faces an unruly faction, for a party is never monolithic. A well-managed party will appear to be almost passively obedient, but the managerial arts are difficult to master. The effective Prime Minister or party leader moves in ways that avoid dissent, if possible, by anticipating it. Concessions are made; issues

are postponed and at times evaded entirely. If we think of the two parties _ .... as disciplined armies marching obediently at their leaders' commands,

we not only ignore much important history but we also overlook the infinite care and calculation that goes into getting groups, be they armies, football teams, or political parties, to act in concert. The Prime Minister

can ordinarily count on his party to support him, but only within limits

that are set in part by the party members collectively. The Prime Minister

can only ask for what his party will give. He cannot say: "The trade unions must be disciplined:' He cannot say: "The relations of labor and management must be recast?' He cannot say: "Industry must be rationalized?' He cannot make such statements, even if he believes them. He

can give a bold lead only if he is sure that his party will come around without a major faction splitting off. But by the time a Prime Minister

is sure of that, any lead given is no longer a bold one. One can be a bold Prime Minister only at the cost of being a bad party manager. "A Party has to be managed, and he who can manage it best, will probably

be its best leader. The subordinate task of legislation and of executive government may well fall into the inferior hands of less astute practitloners!" Such were the reflections of Anthony Trollope on the career of

Sir Timothy Beeswax, a party manager of near magical skill~ (1880,3: 169;

cf. I :216). The roles of leader of the country and manager of a party easily come into conflict. In the absence of formal checks and balances

of the American sort, the party that would act can do so. Because the party in power acts on the word of its leader,the leader must be cautious about the words he chooses to utter.

-The leadership problem coupled. with the apprenticeship factor goes far to describe the texture of British politics. The Prime Minister must

Kenneth N. Wa!cz

preserve the unity of his party, for it is not possible for him to perpetuate his rule by constructing a series of majorities whose composition varies from issue to issue. Prime Ministers must be, and must take pains to remain, acceptable to their parliamentary parties. By the political system within which he operates, the Prime Minister is impelled to seek the support of his entire party, at the cost of considerably reducing his freedom of action. He is constrained to crawl along cautiously, to let situations develop until the near necessity of decision blunts inclinations to quarrel about just what the decision should be. Leadership characteristics are built into the system. The typical Prime Minister is a weak national leader but an expert party manager--characteristics that he

.. ordinarily must have in order to gain-office and retain it.---·

In contrast, consider Presidents. Because their tenure does not depend on securing majority support in Congress, because they can be defeated on policies and still remain in office, and because obstruction is an ordinary and accepted part of the system, they are encouraged to ask for what at the moment may well not be granted. Presidents are expected to educate and inform, to explain that the legislation Congress refuses to pass is actually what the interest of the country requires; they may, indeed, ask for more than they want, hoping that the half-loaf they often get will conform roughly to their private estimate of need. The gap between promise and performance, between presidential request and congressional acquiescence is thus often illusory. Prime Ministers get all that they ask, and yet majorsocial and economic legislation in Britain is ordinarily a long time maturing:-·Presidents ask for much that they do not get, and yet the pace of reform is not slower, the flexibllity and response of American government are not less, than those of Great Britain.

Appearances are often deceptive. Prime Ministers are thought to be strong leaders because they are in public so ineffectively opposed. The fusion of powers, however, tempts the Prime Minister to place his concern for the unity of the party above his regard for the public interest and in rendering the party responsible in the eyes of the voter makes the government unresponsive to the needs of the nation. "A public man is responsible" as a character in one of Disraeli's novels once said, "and a responsible man is a slave" (1880: 156). To be clearly responsible is to be highly visible. In America, the congressional show detracts in some mea-

POLITICAL STRUCTURES

79

sure from the attention the President receives; in Britain, the public concentrates its gaze with single-minded intensity on the Prime Minister. Fairly or not, he is praised or blamed for the good or ill healthofthe polity. Responsibility is concentrated rather than diffused. The. leader who is responsible then has to husband his power; the onus for th{risky policy that fails to come off falls entirely on him.

Americans, accustomed to rule by strong Presidents, naturally think only in terms of limits that are institutionally imposed and overlook the structural constraints on British government. Indeed in the two countries, the term "leadership" has different political meanings: in the United States, that strong men occupy the Presidency; In Britain, that thewill

of the PrirneMinlsterbecomes the law of the land. To say i:hatth~~vill of the leader becomes law should not be taken to mean that the system is one of strong leadership in the American sense; instead everything depends on the leader's identity and on the forces that shape his decisions. The British system goes far to ensure that the leader is moderate and will behave with propriety. This is not seen by simply observing political processes. One has first to relate political structure to process, to consider the ways in which political offices and. institutions are juxtaposed and combined. Power is concentrated in the hands of the Prime Minister and yet with great, though informal, checks against its impetuous use: the apprentice system by which parliamentarians rise to office; the subtle restraints of party that work upon the Prime Minister; the habit, institutionally encouraged, of moving slowly with events and of postponing changes in policy until their necessity is widely accepted.

The endurance of patterns over the decades is striking, Think of the Prime Ministers Britain has known since the tum of the century. They are Balfour, Campbell-Bannerman, Asquith, Lloyd George, Bonar Law, Baldwin, MacDonald, Chamberlain, Churchill, Attlee, Eden, Macmillan,

. Home, Wilson, Heath, and Callaghan. 1\vo failed to fit the patternLloyd George and Winston Churchill. Both had long sat in the Commons:

Both had worked their ways up the ladder. They had served their apprenticeships, but doing so had not tamed them. In normal times each of them appeared unreliable at best, and perhaps downright dangerous, to fractions of their parties large enough to deny them the highest office. Back benchers in large number th.ought of them as being unlikely to balance the interests and convictions of various groups within the party,

-

r-

~ '.

80

Kennech N. Hfllcz

to calculate nicely whose services and support merited higher or lower ministerial positions, and to show a gentlemanly respect for the opinions of others even when they were thought to be ill-founded. A few comments on Winston Churchill will show what I mean. Member of Parliament since 1900 and the holder of more ministerial posts than any politician in British history, he was richly qualified for the highest office. But he had been a maverick for most of his political life. A Conservative at the outset of his political career, he became a Liberal in 1906 and did not return to the Conservative fold until the middle 19205. In the 19305, he was at odds with his party on great matters of state policy, first on Indian and then on I:llropean afTair.;._Nothing less thanacrisis big enough to tum his party liabilities into national assets could elevate him to the highest office. The events required to raise him to prime ministerial office, by virtue of their exceptional quality, cause the normal practice to stand out more clearly. Accidents do occur, but it takes great crises to produce them. To pull someone from outside the normal lines of succession is not easily done.

Political structure produces a similarity in process and performance so long as a structure endures. Similarity is not uniformity. Structure operates as a cause, but it is not the only cause in plat How can one know whether observed effects are caused by the structure of national politics rather than by a changing cast of political characters, by variations of nonpolitical circumstances, and by a host of other factors? How can one separate structural from other causes? One does it by extending the comparative method that I have just used. Look, for example, at British political behavior where structure differs. Contrast the behavior of the Labour movement with that of the Parliamentary Labour Party. In the Labour movement, where power is checked and balanced, the practice of politics, especially when the party is out of power, is strikingly similar. to the political conduct that prevails in America. In the face of conflict and open dissension, the leaders of the party are stimulated actually to lead, to explore the ground and try to work out compromises, to set a line of policy, to exhort and persuade, to threaten and cajole, to inform and educate, all with the hope that the parts of the party-the National Executive Committee, the trade unions, and the constituency parties, as

. well as the Members of Parllament-e-can be brought to follow the leader.

Within a country one can identify the effects of structure by noticing

POLITICAL STRUCTIlRES

8r

d1ferences of behavior in differently structured parts of the polity. From one country to another, one can identify the effects of structure by noticing similarities of behavior in polities of similar structure. Thus Chihiro Hosoya's description of the behavior of Prime Ministers in postwar Japan's parliamentary system exactly fits British Prime Ministers (1974:366-369). Despite cultural and other differences, similar structures produce similar effects.

III

I defined domestic political structures first by the principle according to which they are organized or ordered, second by the differentiation of units and the specification of their functions, and third by the distribution of capabilities across units. Let us see how the three terms of the definition apply to international politics.

I. Orderins Principles

Structural questions are questions about the arrangement of the parts of a system. The parts of domestic political systems stand in relations of super- and subordination. Some are entitled to command; others are required to obey. Domestic systems are centralized and hierarchic. The parts of international-political systems stand in relations of coordination. Formally, each is the equal of all the others. None is entitled to command; none is required to obey. International systems are decentralized and anarchic. The ordering principles of the two structures are distinctly different, indeed, contrary to each other. Domestic political structures have governmental institutions and offices as their concrete counterparts, International politics, in contrast, has been called "politics in the absence of government" (Fox 1959:35). International organizations do exist, and in evergrowing numbers. Supranational agents able to act effectively, however, either themselves acquire some of the attributes and capabilities of states, as did the medieval papacy in the era of Innocent III, or they soon reveal their inability to act in important ways except with the support,

_or at least the acquiescence, of the principal states concerned with the matters at hand.\Vhatever elements of authority emerge internationally are barely once removed from the capability that provides the foundation

82

Kenneth N. Walcz

for the appearance· of those elements. Authority quickly reduces to a particular expression of capability. In the absence of agents with systemwide authority, formal relations of super- and subordination fail to develop.

The first term of a structural definition states the principle by which the system is ordered. Structure is an organizational concept. The prominent characteristic of international politics, however, seems to be the lack of order and of organization. How can one think of international politics as being any kind of an order at all? The anarchy of politics internationally is often referred to. If structure is an organizational concept, the terms "structure" and "anarchy" seem to be in contradiction. If international politics is "politics inthe absehceofgovernmeiit;' \\·hat· are \\'e-irfthe· presence of? In looking for international structure, one is brought face to face with the invisible, an uncomfortable position to be in.

The problem is this: how to conceive of an order without an orderer

and of organizational effects where formal organization is lacking. Because ,'""., these are difficult questions, I shall answer them through analogy with microeconomic theory. Reasoning by analogy is helpful where one can move from a domain for which theory is well developed to one where

it is not. Reasoning by analogy is permissible where different domains

are structurally similar.

Classical economic theory, developed by Adam Smith and his followers,

is microtheory. Political scientists tend to think that microtheory is theory about small-scale matters, a usage that ill accords with its established meaning. The term "micro" in economic theory indicates the way in which the theory is constructed rather than the scope of the matters it pertains to. Microeconomic theory describes how an order is spontaneously formed from the self-interested acts and interactions of individual units-in this case, persons and firms. The theory then turns upon the

two central concepts of the economic units and of the market. Economic' units and economic markets are concepts, not descriptive realities or concrete entities. This must be emphasized since from the early eighteenth century to the present, from the SOciolOgist Auguste Comte to

the psychologist George Katona, economic theory has been faulted because its assumptions fail to correspond with realities (Martineau 1853, 2:51-53; Katona 1953). Unrealistically, economic theorists conceive of an economy operating in. isolation from its society and polity. Unrealist-

POLITICAL STRUCTURES

ically, economists assume that the economic world is the whole of the world. Unrealistically, economists think of the acting unit, the famous "economic man;' as a Single-minded profit maximizer. They single out one aspect of man and leave aside the wondrous variety of human life. As any moderately sensible economist knows, "economic man" does not exist. Anyone who asks businessmen how they make their decisions will find that the assumption that men are economic maximizers grossly distorts their characters. The assumption that men behave as economic men, which is known to be false as a descriptive statement, turns out to be useful in the construction of theory.

, Markets are the second. major concept invcnt<!d bymicroeconornic

- - --theoristS: 1\\'0 general questions must be asked about markets: How are they fonned? How do they work? The answer to the first question is this: The market of a decentralized economy is individualist in origin, spontaneously generated, and unintended. The market arises out of the activities of separate units-persons and finns-whose aims and efforts are directed not toward creating an order but rather toward fulfilling their own internally defined interests by whatever means they can muster. The individual unit acts for itself. From the coaction of like units emerges a structure that affects and constrains all of them. Once formed, a market becomes a force in itself, and a force that the constitutive units acting singly or in small numbers cannot control. Instead, in lesser or greater degree as market conditions vary, the creators become the creatures of the market that their activity gave rise to. Adam Smith's great achievement was to show how self-interested, greed-driven actions may produce good social outcomes if only political and social conditions permit free competition. If a laissez-faire economy is harmonious" it is so because the intentions of actors do not correspond with the outcomes their actions produce. What intervenes between the actors and the objects of their action in order to thwart their purposes? To account for the unexpectedly favorable outcomes of selfish acts, the concept of a market is brought' into play. Each unit seeks its own good; the result of a number of units simultaneously doing so transcends the motives and the 'aims of the separate units. Each would like to work less hard and price his product higher. Taken together, all have to work harder and price their products lower. Each finn seeks to increase its profit; the result of many firms doing so drives the profit rate downward. Each man seeks his own end,

Kenneth N. H-&ltz

84 and, in doing so, produces a result that was no part of his inten~on. Out of the mean ambition of its members, the greater good of society

is produced. .

The market is a cause interposed between the economic actors and

the results they produce. It conditions their calculations, their behaviors, and their interactions. It is not an agent in the sense of A being the agent that produces outcome X. Rather it is a stru~ral ca~e. ~market constrains the units that comprise it from taking certainactions and disposes them toward taking others. The market. cre~ted by self~directed interacting economic units, selects behaviors accordmg to thel: con:e- quences (cf. article 3, part III). The market rewards sO.me With .hlg~

_ profits and assigns others to bankruptcy. Since a market IS not an msntution or an agent in any concrete or palpable sense, such statements become impressive only if they can be reliably inferred from a theory as part of a set of more elaborate expectations. They can be. Microeconomic theory explains how an economy operates and why certain effects are to be expected. It generates numerous "if-then" statements ~hat ~an more or less easily be checked. Consider, for example, the following sl~ple.but important propositions. If the money demand for a commodity nses, then so will its price. If price rises, then so will profits. If profits rise, then capital will be attracted and production will increase. If production increases, then price will fall to the level that returns profits to the producers of the commodity at the prevailing rate. This sequence of statements could be extended and refined, but to do so would not serve my purpose. I want to point out that although the stated expectations are now commonplace, they could not be arrived at by economists working in a pre-theoretic era. A.II of the state~ents ~~e, of co~rse, m:de at an appropriate level of generalIty. They require an other thmgs bemg equal" stipulation. They apply, as do statements inferred from any theo?,: only to the extent that the conditions contemplat~d by the the~ry obt~m. They are idealizations, and so th~ ~re ~ever full! borne ?u~ In pra:tlce. Many things-social customs, political interventlons+-will in fact interfere with the theoretically predicted outcomes. Though interferences have to be allowed for, it is nevertheless extraordinarily useful to know what

to expect in general. .

International-political systems, like _economic markets, are formed by the coaction of self-rega:ding units. International structures are defined

POLITICAL STRUCTURES

85

in terms of the primary political units of an era, be they city states, empires, or nations. Structures emerge from the coexistence of states. No state intends to participate in the formation of a structure by which it and others will be constrained. International-political systems, like economic markets, are individualist in origin, spontaneously generated,

and unintended. In both systems, structures are formed by the coaction of their units. Whether those units live, prosper, or die depends on their own efforts. Both systems are formed and maintained on a principle of self-help that applies to the units. To say that the two realms are struc-

turally similar is not to proclaim their identity Economically, the self-~¢ help principle applies withingoverIlIl1entally _contrived limits .. Market economies-are hedged about in ways that channel energies constructively.

One may think of pure food-and-drug standards, antitrust laws, securities

and exchange regulations, laws against shooting a competitor, and rules forbidding false claims in advertising, International politics is more nearly a realm in which anything goes. International politics is structurally similar to a market economy insofar as the self-help principle is allowed to operate in the latter.

In a microtheory, whether of international politics or of economics, the motivation of the actors is assumed rather than realistically described. I assume that states seek to ensure their survival. The assumption is a radical simplification made for the sake of constructing a theory. The question to ask of the assumption, as ever, is not whether it is true but whether it is the most sensible and useful one that can be made. Whether it is a useful assumption depends on whether a theory based on the

... assumption can be contrived, a theory from which important consequences not otherwise obvious can be inferred. Whether it is a sensible assumption can be directly discussed.

Beyond the survival motive, the aims of states may be endlessly varied; they may range from the ambition to conquer the world to the desire merely to be left alone. Survival is a prerequisite to achieving any goals that states may have, other than the goal of promoting their own disappearance as political entities. The survival motive is taken as the ground of action in a world where the security of states is not assured, rather than as a realistic description of the impulse that lies behind every act

. of state. The assumption allows for the fact that no state - always acts exclusively to ensure its survival. It allows for the fact that some states

86

Kenneth N. Walrz

POLITICAL STRUCTURES .

game one has to win ;S defined by the structure that determines the kind of player who is likely to prosper.

.Where selection according to behavior occurs; no enforced standard of behavior is required for the system to operate, although either system may work better if some standards are enforced or accepted. Internationally, the environment of states' action, or the structure of their system, is set by the fact that some states prefer survival over other ends obtainable in the short run and act with relative efficiency to achieve that end. States may alter their behavior because of the structure they form through interaction with other states. But in what ways and why? To answer these questions we must complete the definition of international structure.

2. The Character if the Units

The second term in theQ~fj!.1ition of domestic political structure specifies the functions performed by differentiated units. Hierarchy entails relations of super- and subordination among a system's parts, and that implies their differentiation. In defining domestic political structure the second term, like the first and third, is needed because each term points to a possible source of.structural variation. The states that are the units of international-political systems are not formally differentiated by the functions they perform. Anarchy entails relations of coordination among a system's units, and that implies their sameness. The second term is not needed in defining international-political structure, because so long as anarchy endures, states remain like units. International structures vary only through a change of organizing principle or, failing that, through variations in the capabilities of units. Nevertheless I shall discuss these like units here, because it is by their interactions that internationalpolitical structures are generated.

Two questions arise: Why should states be taken as the units of th~ system? Given a wide variety of states, how can one call them "like units"? Questioning the choice of states as the primary units of international-political systems became popular in the 1960s and '70s as it was at the turn of the. century. Once . one understands what is logically involved, the issue is easily resolved. Those who question the state-centric view do so for two main reasons. First, states arc not the only actors of

may persistently seek goals that they value more highly than survival; they may, for example, prefer amalgamation with other states to their own survival in form. It allows for the fact that in pursuit of its security no state will act with perfect knowledge and \ v isdom-if indeed we could know what those terms might mean. Some systems have high requirements for their functioning. Traffic will not flow if most, but not all, people drive on the proper side of the road. If necessary, strong measures have to be taken to ensure that everyone does so. Other systems have medium requirements. Elevators in skyscrapers are planned so that they can handle the passenger load if most people take express elevators

for the longer runs and locals _ only _fortheshorter .ones. But ifsome people choose locals for long runs because the speed of the express makes them dizzy, the system will not break down. To keep it going, most, but not all, people have to act as expected. Some systems, market economies and international politics among them, make still lower demands. Traffic systems are designed on the knowledge that the system's requirements will be enforced. Elevators are planned with extra capacity to allow for human vagaries. Competitive economic and internationalpoliticalsystems work differently. Out of the interactions of their parts they develop structures that reward or punish behavior that conforms more or less nearly to what is required of one who wishes to succeed in the system. Recall my description of the constraints of the British parliamentary system. Why should a would-be Prime Minister not strike out on a bold course of his own? Why not behave in ways markedly different from those of typical British political leaders? Anyone can, of course, and some who aspire to become Prime Ministers do so. They rarely come to the top. Except in deepest crisis, the system selects others to hold the highest office. One may behave as one likes to. Patterns of behavior nevertheless emerge, and they derive froni the structural con-. straints of the system.

Actors may perceive the structure that constrains them and understand how it serves to reward some kinds of behavior and to penalize others. But then again they either may not see it or, seeing it, may for any of many reasons fail to conform their actions to the patterns that are most often rewarded and least often punished. To say that "the structure selects" means simplytllat those who conform to accepted and successful practices more often rise. to the top and are likelier to stay there. The

88

Kenneth N. Waltz

importance on the international scene. Second, states are declining in importance, and other actors are gaining, or so it is said. Neither reason is cogent, as the following discussion shows.

States are not and never have been the only international actors. But then structures are defined not by all of the actors that flourish within them but by the major ones. In defining a system's structure one chooses one or some of the infinitely many. objects comprising the system and defines its structure in terms of them. For international-political systems, as for any system, one must first decide which units to take as being the parts of the system. Here the economic analogy will help again. The

_structure of a market is defined by the number of firms competing.-If many roughly equal firms contend, a condition of perfect competition is approximated. If a few firms dominate the market, competition is said to be oligopolistic even though many smaller firms may also be in the field. But we are told that definitions of this sort cannot be applied to international politics because of interpenetration of states, because of their inability to control the environment of their action, and because rising multinational corporations and other nonstate actors are difficult to regulate and may rival some states in. influence. The importance of nonstate actors and the extent of transnational activities are obvious. The conclusion that the state-centric conception of international politics is made obsolete by them does not follow. That economists and economically minded political scientists have thought that it does is ironic. The irony lies in the fact that all of the reasons given for scrapping the state-centric concept can be restated more strongly and applied to firms. Firms competing with numerous others have no hope of controlling their market, and oligopolistic firms constantly struggle with imperfect success to do so. Firms interpenetrate, merge, and buy each other up at a merry pace. Moreover, firms are constantly threatened and regulated by, shall we say, "nonfirm" actors. Some governments encourage concentration; others' work to prevent it. The market structure of parts of an economy may move from a wider to a narrower competition or may move in the opposite direction, but whatever the extent and the frequency of change, market structures, generated by the interaction of firms, are defined in terms of them.

Just as economists define markets in terms of firms, so I define international-political structur~s in terms of states. If Charles P. Kindleberger

.~ 1,,;

.. , ~"

POLITICAL STRUCTURES

~ ... .'

were right in saying that "the nation-state is just about through as an economic unit" (1969:207), then the structure of international politics would have to be redefined. That would be necessary because economic capabilities cannot be separated from the other capabilities of states. The distinction frequently drawn between matters of high and low politics is misplaced. States use economic means for military and political ends; and military and political means for the achievement of economic interests.

An amended version of Kindleberger's statement may hold: Some states may be nearly washed up as economic entitites, and others not. That

. -poses no" problem' forintemational-politicalth~ory since international politics is mostly about inequalities anyway. So long as the major states are the major actors, the structure of international politics is defined in terms of them. That theoretical statement is of course borne out in practice. States set the scene in which they, along with nons tate actors, stage their dramas or carryon their humdrum affairs. Though they may choose to interfere little in the affairs of nonstate actors for long periods of time, states nevertheless set the terms of the intercourse, whether by passively permitting informal rules to develop or by actively intervening to change rules that no longer suit them. When the crunch comes, states remake the rules by which other actors operate. Indeed, one may be struck by the ability of weak states to impede the operation of strong international corporations and by the attention the latter pay to the wishes of the former.

It is important to consider the nature of transnational movements, the extent of their penetration, and the conditions that make it harder or easier for states to control them. But the adequate study of these matters, like others, requires finding or developing an adequate approach to the study of international politics. 1\\'0 points should be made about latterday transnational studies. First, students of transnational phenomena have developed no distinct theory of their subject matter or of international politics in general. They have drawn on existing theories, whether economic or political. Second, that they have developed no distinct theory is quite proper, for a theory that denies the central role of states will be needed only if nonstate actors develop to the point of rivaling or surpassing the great powers, not just a few of the minor ones. They show no sign of doing that.

Kenneth N. j%ltz

POLITICAL STRUCTURES

9I

The study of transnational movements deals with important factual··, own courses, make their own decisions about how to meet whatever

questions, which theories can help one to cope with. But the help will needs they experience and whatever desires they develop. It is no more

not be gained if it is thought that nons tate actors call the state-centric contradictory to say that sovereign states are always constrained and often

view of the \varld into question. To say that major states maintain their tightly so than it is to say that free individuals often make decisions

central importance is not to say that other actors of some importance under the heavy pressure of events.

do not exist. The "state-centric" phrase suggests something about the Each state, like every other state, isa sovereign political entity. And

system's structure. Transnational movements are among the processes yet the differences across states, from Costa Rica to the Soviet Union

that go on within it. That the state-centric view is so often questioned from Gambia to the United States, are immense. States are alike, and

merely reflects the difficulty political scientists have in keeping the dis- : r they are also different. So are corporations, apples, universities, and

tinction between structures and processes clearly and constantly in mind. , !~: people. Whenever we put two or more objects in the same category, we

- States are the units whose interactions form the-structure of inter- __ -'; {_~ __ are saying_ that they are alike not in all respects but in some. No two

national-political systems. They will long remain so. The death rate t!> objects in this world are identical, yet they can often be usefully compared

among states is remarkably low Few states die; many firms do. Who is F and combined. "You can't add apples and oranges" is an old saying that

likely to be around 100 years from now-the United States, the Soviet " seems to be especially popular among salesmen who do not want you

Union, France, Egypt, Thailand, and Uganda? Or Ford, IBM, Shell, Un- to compare their wares with others. But we all know that the trick of

ilever, and Massey-Ferguson? I would bet on the states, perhaps even on adding diss~milar objects is to express the result in terms of a category

Uganda. But what does it mean to refer to the 1 SO-odd states of today's • that compnses them. Three apples plus four oranges equals seven pieces

world, which certainly form a motley collection, as being "like units"? t~, of fruit. The only interesting question is whether the category that clas-

Many students of international politics are bothered by the description. :; . sHies objects according to their common qualities is useful. One can add

To call states "like units" is to say that each state is like all other states . up a large number of Widely varied objects and say that one has eight

in being an autonomous political unit. It is another way of saying that million things, but seldom need one do that.

states are sovereign. But sovereignty is also a bothersome concept. Many States vary Widely in size, wealth, power, and form. And yet variations

believe, as the anthropologist M. G. Smith has said, that "in a system of in these and in other respects are variations among like units. In what

sovereign states no state is sovereign?" The error lies in identifying the way are they like units? How can they be placed in a single category?

sovereignty of states with their ability to do as they wish. To say that States are alike in the tasks that they face, though not in their abilities

states are sovereign is not to say that they can do as they please, that to perform them. The differences are of capability, not of function. States

they are free of others' influence, that they are able to get what they perform or try to perform tasks, most of which are common to all of

want. Sovereign states may be hardpressed all around, constrained to act them; the ends they aspire to are similar. Each state duplicates the

in ways they would like to avoid, and able to do hardly anything just as . activities of other states at least to a considerable extent. Each state has

they would like to. The sovereignty of states has never entailed their its agencies for making, executing, and interpreting laws and regulations,

. insulation from the effects of other states' actions. To be sovereign and for raising revenues, and for defending itself. Each state supplies out of

to be dependent are not contradictory conditions. Sovereign states have its own resources and by its own means most of the food, clothing,

seldom led free and easy lives. What then is sovereignty? To say that a hOUSing, transportation, and amenities consumed and used by its citizens.

state is sovereign means that it decides for itself how it will cope with All states, except the smallest ones, do much more of their business at

its internal and external problems, including whether or not to seek homethan abroad. Onehas to beimpressedwith the functional similarity

assistance from others and in doing so to limit its freedom by making of states and, now more than ever before, with the similar lines their

commitments to them. Sjatesdevelop their own strategies; chart their development follows. From the rich to the poor states, from the old to

Kenneth N. Waltz

::, I

92 the new ones, nearly all of them take a larger hand in ma~ters of economic regulation, of education, health, and housing, of ~lt~:e and the a.rts, and so on almost endlessly. The increase of the activities of states IS a

_~trong and strikingly uniform international tre~d. n:e f~nctions of stat~s _ are similar, and distinctions among them anse principally from their _ -. .varied capabilities. National politics consists of difTeren:iated u~its ~rJorming specified functions. International politics consists of like units ~duplicating one another's activities.

3. The Distribution if Capabilities

_ The _ parts oC ~_hierarc:hic _ system.Clre _rt:la_t~d toone another in ways th~t_ are determined both by their functional differentiation and by the extent of their capabilities. The units of an anarchic system .ar.e fu.nctional~y undifferentiated. The units of such an order are then dlstmgUlshed pnmarily by their greater or lesser capabilities for ~erformin~.similar tasks. This states formally what students of international politics have long noticed. The great powers of an era have always been marke~ off from the others by practitioners and theorists alike. Students of national government make such distinctions as that between parliamentary and presidential systems; governmental systems differ in. form '. Studen~. of international politics make distinctions between international-political systems only according to the number of their gre~t p~wers. The s:~cture of a system changes with changes in the distribution of capabll~tles across the system's units. And changes in structure change expectations - about how the units of the system will behave and about the outcomes their interactions will produce. Domestically, the differentiated parts of a system may perform similar tasks. We know from observing the American government that executives sometimes legislate_ and legis_Iatures sometimes execute. Internationally, like units sometimes perform different tasks. Why they do so, and how the likelihood of their doing ,so varies with their capabilities, are matters treated at length in the last three chapters [not reprinted here: ed.]. Meanwhile, two problems should be considered.

The first problem is this: Capability tells us something about units.

Defining structure partly in terms of the distribution of capabilities seems to VIolate my instruction to keep unit attributes out of structural definitions. As I remarked earlier, structure is a highly but not entirely

' .... ;~. "

POLITICAL STRUCJ11RES

93

abstract concept. The maximum of abstraction allows a minimum of content, and that minimum is what is needed to enable one to say how the units stand in relation to one another. States are differently placed by their power. And yet one may wonder why only capability is included in the third part of the definition, and not such characteristics as ideology, form of government, peacefulness, bellicOSity, or whatever. The answer is this: Power is estimated by comparing the capabilities of a number of units. Although capabilities are attributes of units, the distribution of capabilities across units is not. The distribution of capabilities is not a unit attribute, but rather a system-wide concep~. Again, the parallel with market theory is exact. Both firms-and states are like·units. Through all of their variations in form, firms share certain qualities: They are self-

regarding units - that, within governmentally imposed limits, decide for themselves how to cope with their environment and just how to work for their ends. Variation of structure is introduced, not through differences in the character and function of units, but only through distinctions made among them according to their capabilities.

The second problem is this: Though relations defined in terms of interactions must be excluded from structural definitions, relations defined in terms of groupings of states do seem to tell us something about how states are placed in the system. Why not specify how states stand in relation to one another by considering the alliances they form? Would doing so not be comparable to defining national political structures partly in terms of how presidents and prime ministers are related to other. political agents? It would not be. Nationally as internationally, structural' definitions deal with the relation of agents and agencies in terms of the organization of realms and not in terms of the accommodations and conflicts that may occur within them or the groupings that may now and then form. Parts of a government may draw together or pull apart, may oppose each other or cooperate in greater or lesser degree. The.se are the relations that form and dissolve within a system rather than structural alterations that mark a change from one system to another. This is made clear by an example that runs nicely parallel to the case of alliances. DistingUishing systems of political parties according to their number is common. A multiparty system changes if, say, eight parties become two, but not if two groupings of the eight form merely for the occasion of fighting an election. By the same logic, an international-

._- .. :

94

Kenneth N. mUtz

political system in which three or more great powers have split into two alliances remains a multipolar system-structurally distinct fr?m a bipolar system, a system in which no third power is able to challenge the top two. In defining market structure, information about the particular quality of firms is not called for, nor is information about their interactions, short of the point at which the formal merger of firms sign~ficant1y reduces their number. In the definition of market structure, firms are not identified and their interactions are not described. To take the qualities of firms and the nature of their interactions as being parts of market structure would be to say that whether a sector of an economy is oligopolistic or not depends on how the firms are organized internally and how they deal with one another, rather than simply on how many major firms coexist. Market structure is defined by counting firms; international-political structure, by counting states. In the counting. distinctions are made only according to capabilities.

In defining international-political structures we take states with whatever traditions, habits, objectives, desires, and forms of government they may have. We do not ask whether states are revolutionary or legitimate, authoritarian or democratic, ideological or pragmatic. We abstract from every attribute of states except their capabilities. Nor in thinking about structure do we ask about the relations of states-their feelings of friendship and hostility, their diplomatic exchanges, the alliances they form, and the extent of the contacts and exchanges among them. We ask what range of expectations arises merely from looking at the type of order that prevails among them and at the distribution of capabilities within that order. We abstract from any particular qualities of states and from all of their concrete connections. What emerges is a positional picture, a general description of the ordered overall arr~ngement of a society written in terms of the placement of units rather than in terms of their qualities.

IV

I have now defined thetwo essential elements of a systems theory of international politics-the structure of the system and its interacting units. In doing so I have broken sharply away from common approaches.

POLITICAL STRucruRES

95

':> we have seen, some scholars who attempt systems approaches to ~nternat.i~nal politics conceive of a system as being the product of its mteractmg parts, but they fail to consider whether anything at the systems level affects those parts .. Other systems theorists, like students of international politics in general, mention at times that the effects of the international environment must be allowed for; but they pass .over the question of how this is to be done and qUickly return their attention to the level of interacting units. Most students, whether or not they claim to follow a systems approach, think of international politics in the way fig. 4.1 suggests. Nl.2.l are states internally generating their external effects.

.. Xl•2•l are. states. acting. externally. and interacting with each other.· No systemic force or factor shows up in the picture.

Figure 4.1.
Nl Xl
t
N2 X2
t
Nl Xl Because systemic effects are evident, international politics should be seen as in fig. 4.2. The circle represents the Structure of an Internationalpolitical system. As the arrows indicate, it affects both the interactions of states and their attributes." Although structure as an organizational concept has proved elusive, its meaning can be explained simply. While states retain their autonomy, each stands in a specifiable relation to the others. They form some sort of an order. We can use the term "organization" to cover this preinstitutional condition if we think of an organization as simply a constraint, in the manner of W. Ross Ashby (1956:131). Because states constrain and limit each other, international politics can be viewed in rudimentary organizational terms. Structure is the concept that makes it possible to say what the expected organizational effects are and how structures and units interact and affect each other.

r.hinking of structure as Ihave defined it solves the problem of separatIng changes at the level of the units from changes at the level of the system. If one is concerned with the different expected effects of dif-

Kenneth N. Waltz

Figure 4.2.

N~-----i'"

ferent systems, one must be able to distinguish changes of systems from changes within them, something that would-be systems theorists have found exceedingly difficult to do. A three-part definition of structure enables one to discriminate between those types of changes:

• Structures are defined, first, according to the principle by which a system is ordered. Systems are transformed if one ordering principle replaces another. To move from an anarchic to a hierarchic realm is to move from one system to another.

• Structures are defined, second, by the specification of functions of differentiated units. Hierarchic systems change if functions are differently defined and allotted. For anarchic systems, the criterion of systems change derived from the second part of the definition drops out since the system is composed of like units,

• Structures are defined, third, by the distribution of capabilities across units. Changes in this distribution are changes of system whether the system be an anarchic or a hierarchic one.

NOTES

I. The exception, which does not disprove the rule, is Ramsay MacDonald, who, absent from the wartime coalition, and with his party not previously in power,had never served in a ministerial position.

2. All calculations as of July 1978.

3. In some respects a century brings little change. Despite the many harsh

'~"

.• ~

POLITICAL STRUcruRES

97

comments made about Callaghan by Crossman, Wilson, and others, Crossman thought of him as "easily the most accomplished politician of the Labour Party"; and apparently because of that distinction, Callaghan gained Wilson's help in succeeding him as Prime Minister (1977, 3:627-628 er passim).

4. Smith should know better. Translated into terms that he has himself so effectively used, to say that states are sovereign is to say that they are segments of a plural SOciety (1966:122; cf. 1956).

5. No essentials are omitted from fig. 4.2, but some complications are. A full picture would include, for example, coalitions possibly forming on the right-hand side.

FIVE Anarchic Orders and Balances of Power

KENNETH N . WALTZ

J. Violence at Home and Abroad

THE STATEAMONG STATES, it is often said, conducts its affairsin the brooding shadow of violence. Because some states may at any time use force, all states must be prepared to do so--or live at the mercy of their militarily more vigorous neighbors. Among states, the state of nature is a state of war; This is meant not in the sense that war constantly occurs but in the sense that, with each state deciding for itself whether or not to use force, war may break out at any time. Whether in the family, the community, or the world at large, contact without at least occasional conflict is inconceivable; and the hope that in the absence of an agent to manage or to manipulate conflicting parties the use of force will always be avoided cannot be realistically entertained. Among men as among states, anarchy, or the absence of government, is associated with the occurrence of violence.

The threat of violence and the recurrent use of force are said to distinguish international from national affairs. But in the history of the world surely most rulers have had to bear in mind that their subjects might use force to resist or overthrow them. If the absence of government is associated with the threat of violence, so also is its presence. A haphazard list of national tragedies illustrates the point all too well. The most destructive wars of the hundred years following the defeat of Napoleon took place not among states but within them. Estimates of deaths in China's Taiping Rebellion, which began in 1851 arid lasted 13years, " range as high as 20 million .. In. the American Civil War some 600,000

\'.' .~ .

l'"

7',.'

: -, ~.;-.

.,'

:0:'

ANARCHIC ORDERS AND BALANCES OF POWER

99

people lost their lives. In more recent history, forced collectivization and Stalin's purges eliminated five million Russians, and Hitler exterminated 6 mil!ion Jews. In some Latin American countries, coups d'etat and rebellions have been normal features of national life. Between 1948 and 1957, for example, 200,000 Colombians were killed in civil strife. In the middle 1970s most inhabitants of Idi Amin's Uganda must have felt their lives becoming nasty, brutish, and short, quite as in Thomas Hobbes's state of nature. If such cases constitute aberrations, they are uncomfort.ably comm~n ~nes. We easily lose Sight of the fact that struggles to achieve and mamtam power, to establish order, and to contrive a kind of

.lllstice withiI1states'Il1ay~~ bloodie!tbanwars_among them.

If anarchy is identified with chaos, destruction, and death, then the distinction between anarchy and government does not tell us much. Which is more precarious: the life of a state among states, or of a

. government in relation to its subjects? The answer varies with time and place. Am~ng some states at some times, the act'ual or expected occurrence of VIOlence is low Within some states at some times, the actual or expected occurrence of violence is high. The use of force, or the const~nt fear of its use, is not sufficient grounds for distinguishing international from domestic affairs. If the possible and the actual use of force mark both national and international orders, then no durable distinction between the two realms can be drawn in terms of the use or the nonuse of force. No human order is proof against violence.

To discover qualitative differences between internal and external affairs one must must look for a criterion other than the occurrence of violence. The distin:tion between international and national realms of politics is not found m the use or the nonuse of force but in their different structur~s. But if th.e dangers of being violently attacked ~re greater, say, in t~kl~g .an evemng stroll through downtown Detroit than they are in plcmcking along the French and German border, what practical difference does the difference of structure make? Nationally as internationally, contact generates conflict and at times issues in violence. The difference between national and international politics lies not in the use of force but in the different modes of organization for doing something about it. A go~'Crnment, ruling by some standard of legitimacy, arrogates to itself the nght to use force--"that is, to apply a variety of sanctions to control the use of force by its subjects. If some use private force, others may

····1

100

Kenneth N. nl1ltz

appeal to the government. Agm'ernment has no monopoly on the use of force, as is all too evident. An effective government, however, has a monopoly on the leEJitimate use of force, and legitimate here means that public agents are organized to prevent and to counter the private use of force. Citizens need not prepare to defend themselves. Public agencies do that. A national system is not one of self-help. The international system is.

2. Interdependence and lnteqration

The political Significance of interdependence varies depending on whether

a realm iSorganized, wlth relations of authorityspecified and established,or remains formally unorganized. Insofar as a realm is formally organized, its units are free to specialize, to pursue their own interests without concern for developing the means of maintaining their identity and preserving their security in the presence of others. They are free to specialize because they have no reason to fear the increased interdependence that goes with specialization. If those who specialize most benefit most, then competition in specialization ensues. Goods are manufactured, grain is produced, law. and order are maintained, commerce is conducted and financial services are provided by people who ever more narrowly spe~ cialize. In simple economic terms, the cobbler depends on the tailor for his pants and the tailor on the cobbler for his shoes, and each would be ill-clad without the services of the other. In simple political terms, Kansas depends on Washington for protection and regulation and Washington depends on Kansas for beef and wheat. In saying that in such situations interdependence is close, one need not maintain that the one part could not learn to live without the other. One need only say that the cost of breaking the interdependent relation would be high. Persons and institutions depend heavily on one another because of the different tasks they perform and the different goods they produce and exchange. The parts of a polity binds themselves together by their differences (cf. Durkheim 1893:212).

Differences between national and international structures are reflected in the ways the units of each system define their ends and develop the means for rcaching them. In anarchic realms, like units coact.· In hierarchic realms, unlike units interact. In an anarchic realm, the units are

ANARCHIC ORDERS AND BALANCES OF POWER

101

functionally similar and tend to remain so. Like units work to maintain a measure of independence and may even strive for autarchy. In a hierarchic realm, the units are differentiated, and they tend to increase the extent of their specialization. Differentiated units become closely interdependent, the more closely so as their specialization proceeds. Because

of the difference of structure, interdependence within and interdependence among nations are two distinct concepts. So as to follow the logicians' admonition to keep a Single meaning for a given term throughout one's discourse, I shall use "integration" to describe the condition within nations and. "interdependence" . to describe the condition among them.

Although states are like units functionally, they differ vastly in their

- capabilities.-Out of such differences something of a division of labor develops. The division of labor across nations, however, is slight in comparison with the highly articulated division of labor within them. Integration draws the parts ofa nation closely together. Interdependence among nations leaves them loosely connected. Although the integration of nations is often talked about, it seldom takes place. Nations could mutually enrich themselves by further diViding not just the labor that goes into the production of goods but also some of the other tasks they perform, such as political management and military defense. Why does their integration not take place? The structure of international politics limits the cooperation of states in two ways.

In a self-help system each of the units spends a portion of its effort, not in forwarding its own good, but in providing the means of protecting itself against others. Specialization in a system of divided labor works to everyone's advantage, though not equally so. Inequality in the expected distribution of the increased product works strongly against extension of the division of labor internationally. When faced with the possibility of cooperating for mutual gain, states that feel insecure must ask how the gain will be divided. They are compelled to ask not "Will both of us gain?" but "Who will gain more?" If an expected gain is to be divided, say, in the ratio of two to one, one state may use its disproportionate gain to implement a policy intended to damage or destroy the other. Even the prospect of large absolute gains for both parties docs not elicit their cooperation so long as each fears how the other will use its increased capabilities. Notice that the impediments to collaboration may not lie in the character and the immediate intention of either party. Instead, the

102

Kenneth N. Waltz

condition of insecurity-at the least, the uncertainty of each about the other's future intentions and actions-works against their cooperation.

, In any self-help system, units worry about their survival, and the worry conditions their behavior. Oligopolistic markets limits the cooperation of firms in much the way that international-political structures limit the cooperation of states. Within rules laid down by governments, whether firms survive and prosper depends on their own eITorts. Firms need not protect themselves physically against assaults from other firms. They are free to concentrate on their economic interests. As economic entities, however, they live in a self-help world. All want to increase pr()5ts. I[ they ~I"lln ~n~ue_ti~~_i.nt~~ ~IT()rt todCl_ s~~~ey m~!exp~ct

to suffer the consequences. As William Fellner says, it is "impossible to maximize joint gains without the collusive handling of all relevant variables?' And this can be accomplished only by "complete disarmament of the firms in relation to each other" But firms cannot sensibly disarm even to increase their profits. This statement qualifies, rather than contradicts, the assumption that firms aim at maximum profits. To maximize profits tomorrow as well as today, firms first have to survive. Pooling all resources implies, again as Fellner puts it, "discounting the future possibilities of all participating firms" (1949:35). But the future cannot be discounted. The relative strength of firms changes over time in ways that cannot be foreseen. Firms are constrained to strike a compromise between maximizing their profits and minimizing the danger of their own demise. Each of two firms may be better oIT it one of them accepts compensation from the other in return for withdrawing from some part of the market. But a firm that accepts smaller markets in exchange for larger profits will be gravely disadvantaged if, for example, a price war should break out as part of a renewed struggle for markets. If possible, one must resist accepting smaller markets in return for'larger profits (pp.132, 217-218). "It is;' Fellner insists, "not advisable to disarm in . relation to one's rivals" (p. 199). Why not? Because "the potentiality of renewed warfare always exists" (p, 177). Fellner's reasoning is much like the reasoning that led Lenin to believe that capitalist countries would never be able to cooperate for their mutual enrichment in one vast imperialist enterprise. Like nations, oligopolistic firms must be more concerned with relative strength than with absolute advantage.

A state worries about a division of possible gains that may favor others

ANARCHIC ORDERS AND BALANCES OF POWER

103

, more than itself. That is the first way in which the structure of international politics limits the cooperation of states. A state also worries lest it become dependent on others through cooperative endeavors and ex, changes of goods and services. That is the second way in which the , structure of international politics limits the cooperation of states. The more a state specializes, the more it relies on others to supply the " materials and goods that it is not producing. The larger a state's imports and exports, the more it depends on others. The world's well-being would be increased. if an ever more elaborate division of labor were developed, but states would thereby place themselves in situations of ever closer. illterdependence. Some. sta tes .may, n()_~r~~ist. tha L, ~or small and ill-endowed states the costs of doing so are excessively high. But states '. that can resist becoming ever more enmeshed with others ordinarily do

so in either or 'both of two ways. States that are heavily dependent, or closely interdependent, worry about securing that which they depend on. The high interdependence of states means that the states in question experience, or are subject to, the common vulnerability that high interdependence entails. Like other organizations, states seek to control what 'they depend on or to lessen the extent of their dependency. This simple thought explains quite a bit of the behavior of states: their imperial thrusts to widen the scope of their control and their autarchic strivings toward greater self-sufficiency.

, Structures encourage certain behaviors and penalize those who do not respond to the encouragement. Nationally, many lament the extreme development of the division of labor, a development that results in the , allocation of ever narrower tasks to individuals. And yet specialization proceeds, and its extent is a measure of the development of societies. In a formally organized realm a premium is put on each unit's being able to specialize in order to increase its value to others in a system of divided labor. The domestic imperative is "specialize"! Internationally, many la-

o ment the resources states spend unproductively for their own defense 'and the opportunities they miss to enhance the welfare of their people through cooperation with other states. And yet the ways of states change little. In an unorganized realm each unit's incentive is to put. itself in a position to be able to take care of itself since no one else can be counted on to do so. The international imperative is "take care of yourself"! Some leaders of nations may understand that the well-being of all of

----~-----_--_-~-----~---~,- -~--

Kenneth N. Waltz

ANARCHIC ORDERS AND BALANCES OF POWER

105

J. Scrucrures and Scraceaies

and I cannot sensibly follow them unless we are pretty sure that many others will as well. Let us go more deeply into the problem by considering two further examples in some detail.

-- Each of many persons may choose to drive a private car rather than take a train. Cars offer Ilexibility in scheduling and in choice of destination; yet at times, in bad weather for example, railway passenger service is a much wanted convenience. Each of many persons may shop in supermarkets rather than at comer grocery stores. The stocks of super-

- markets are larger, and their prices lower; yet at times the corner grocery ~tore, offering, say, credit and delivery service, is a much wanted convenience. Theresult of most people usually driving their own cars and shopping at supermarkets is to reduce passenger service and to decrease : the number of comer grocery stores. These results may not be what

_ • most people want. They may be willing to pay to prevent services from -_ disappearing. And yet individuals can do nothing to affect the outcomes.

Increased patronage would do it, but not increased patronage by me and ____ few others I might persuade to follow my example.

_. We may well notice that our behavior produces unwanted outcomes, _ we are also likely to see that such instances as these are examples - . ,what Alfred E. Kahn describes as "large" changes that are brought about by the accumulation of "small" decisions. In such situations people _ victims of the "tyranny of small decisions:' a phrase suggesting that '.'if one hundred consumers choose option x, and this causes the market

.' to make decision X (where X equals 100 x), it is not necessarily true that _. those same consumers would have voted for that outcome if that large : decision had ever been presented for their explicit consideration" (Kahn

: 1966:523). If the market does not present the large question for decision, ___ then individuals are doomed to making decisions that are sensible within . their narrow contexts even though they know all the while that in making _ such decisions they are bringing about a result that most of them do

- not want. Either that or they organize to overcome some of the effects' of the market by changing its structure--for example, by bringing consumer units roughly up to the size of the units that are making producers' _ decisions. This nicely makes the point: So long as one leaves the structure - _ unaffected it is not possible for changes in the intentions and the actions of particular actors to produce desirable outcomes or to avoid - undesirable ones. Structures may be changed, as just mentioned, by changing the

them would increase through their participation in a fuller division of labor. But to act on the idea would be to act on a domestic imperative, an imperative that does not run internationally. What one might want to do in the absence of structural constraints is different from what one is encouraged to do in their presence. States do not willingly place themselves in situations of increased dependence. In a self-help system, considerations of security subordinate economic gain to political interest.

What each state does for itself is much like what all of the others are - - dOing. They are denied the advantages that a full division of labor, political as well as economic, would provide. Defense spending, moreover, isunproductive for all andunavoidable __ for most. Rather than increased - well-being, their reward is in the maintenance of their autonomy. States - compete, but not by contributing their individual efforts to the joint production of goods for their mutual benefit.

That motives and outcomes may well be disjoined should now be easily seen. Structures cause actions to have consequences they were not intended to have. Surely most of the actors will notice that, and at least some of them will be able to figure out why. They may develop a pretty good sense of just how structures work their effects. Will they not then be able to achieve their original ends by appropriately adjusting their strategies? Unfortunately, they often cannot. To show why this is so I shall give only a few examples; once the point is made, the reader will easily think of others.

If shortage of a commodity is expected, all are collectively better off if they buy less of it in order to moderate price increases and to distribute shortages equitably. But because some will be better off if they lay ill extra supplies quickly; all have a strong incentive to do so. If one expects others to make a run on a bank, one's prudent course is to run faster than they do even while knowing that if few others run, the bank will remain solvent, and if many run, it will fail. In such cases, pursuit of individual interest produces collective results that nobody wants, yet individuals bybehaving dlfferentlywillhurt themselves without altering outcomes. These two much used examples establish the main point. Some courses of action I canl}ot sensibly follow unless you do too, and you

Kenneth N. Walcz

106 distribution of capabilities across units. Structures may also be changed by imposing requirements where previously people had- to decide for themselves. If some merchants sell on Sunday, others may have to do so in order to remain competitive even though most prefer a six-day "week. Most are able to do as they please only if all are required to keep comparable hours. The only remedies for strong structural effects are structural changes.

Structural constraints cannot be wished away, although many fail to understand this. In every age and place, the units of self-help systemsnations, corporations, or whatever-are told that the greater good, along with their own, requires them to act for the sake of the system and for their own narrowly defined advantage. In the 1950s, as fear of the world's destruction in nuclear war grew, some concluded that the alternative to world destruction was world disarmament. In the 1970s, with the rapid growth of population, poverty, and pollution, some concluded, as one political scientist put it, that "states must meet the needs of the political ecosystem in its global dimensions or court annihilation" (Sterling 1974:336). The international interest must be served; and if that means anything at all, it means that national interests are subordinate to it. The problems are found at the global level. Solutions to the problems continue to depend on national policies. What are the conditions that would make nations more or less willing to obey the injunctions that are so often laid on them? How can they resolve the tension between pursuing their own interests and acting for the sake of the system? No one has shown how that can be done, although many wring their hands and plead for rational behavior. The very problem, however, is that rational behavior, given structural constraints, does not lead to the wanted results. With each country constrained to take care of itself, no one can take

care of the system. I .

A strong sense of peril and doom may lead to a clear definition of ' ends that must be achieved. Their achievement is not thereby made possible. The possibility of effective action depends on the ability to provide necessary means. It depends even more so on the existence of conditions that permit nations and other organizations to follow appropriate policies and strategies. World-shaking problems cry for global solutions, but there is no global agency to provide them. Necessities do

ANARCHIC ORDERS AND BALANCES OF POWER

107

not create possibilities. Wishing that final causes were efficient ones does not make them so.

Great tasks can be accomplished only by agents of great capability.

That is why states, and especially the major ones, are called on to do

. what is necessary for the world's survival. But states have to do whatever they think necessary for their own preservation, since no one can be relied on to do it for them. Why the advice to place the international interest above national interests is meaningless can be explained precisely in terms of the distinction between micro- and macro theories. Among economists the distinction is well understood. Among political scientists jUs not.AsI have explained,a microeconomic theory is a theory of the market built up from assumptions about the behavior of individuals. The theory shows how the actions and interactions of the units form and affect the market and how the market in tum affects them. A macrotheory is a theory about the national economy built on supply, income, and demand as systemwide aggregates. The theory shows how these and other aggregates are interconnected and indicates how changes in one or some of them affect others and the performance of the economy. In economics, both micro- and macrotheories deal with large realms. The difference between them is found not in the size of the objects of study. but in the way the objects of study are approached and the theory to explain them is constructed. A macrotheory of international politics

would show how the international system is moved by system-wide aggregates. One can imagine what some of them might be+-arnount of world GNp, amount of world imports and exports, of deaths in war, of everybody'S defense spending, and of migration, for example. The theory would look something like a macroeconomic theory in the style of John Maynard Keynes, although it is hard to see how the international aggregates would make much sense and how changes in one or some of them .would produce changes in others. I am not saying that such a theory cannot be constructed, but only that I cannot see how to do it in any way that might be useful. The decisive point, an'f'vay, is that a rnacroth-

. eory of international politics would lack the practical implications of macroeconomic theory. National governments can manipulate systemwide economic variables. No agencies with comparable capabilities exist internationally. . Who would act on the possibilities of adjustment that a

108

Kenneth N. Waltz

macrotheory of international politics might reveal? Even were such a theory available, we would still be stuck with nations as the only agents capable of acting to solve global problems. We would still have to revert to a micropolitical approach in order to examine the conditions that make benign and effective action by states separately and collectively more or less likely.

Some have hoped that changes in the awareness and purpose, in the organization and ideology, of states would change the quality of international life. Over the centuries states have changed in many ways, but: the quality of international life has remained much the same. States may seek reasonable and worthy ends, but they cannot figure out how to reach them. The problem is notin their stupidity or ill will, although one does not want to claim that those qualities are lacking. The depth of the difficulty is not understood until one realizes that intelligence and goodwill cannot discover and act on adequate programs. Early in this century Winston Churchill observed that the British-German naval race promised disaster and that Britain had no realistic choice other than to run it. States facing global problems are like individual consumers trapped by the "tyranny of small decisions?' States, like consumers, can get out of the trap only by changing the structure of their field of activity. The message bears repeating: The only remedy for a strong structural effect _is a structural change.

4. The Virtues if Anarch),

To achieve their objectives and maintain their security, units in a condition of anarchy-be they people, corporations, states, or whatever-must rely on the means they can generate and the arrangements they can make - for themselves. Self-help is necessarily the principle of action in an anarchic order. A self-help situation is one of high risk--of bankruptcy in - the economic realm and of war in a world of free states. It is also one in which organizational costs are 10\''1. Within an economy or within an international order, risks may be avoided or lessened by moving from a situation of coordinate action to one of super- and subordination, that is, by erecting agencies with effective authority and extending a system ~f rules. Government emerges where the functions of regulation and management themselves become distinct and spe-cialized tasks. The costs

: ANARCHIC ORDERS AND BALANCES OF POWER

109

of maintaining a hierarchic order are frequently ignored by those who de~lore its absence. Organizations have at least two aims: to get somethmg done and to maintain themselves as organizations. Many of their activities are directed toward the second purpose. The leaders of organizations, and political leaders preeminently, are not masters of the mat-

-- ters their organizations deal with. They have become leaders not by being experts on one thing or another but by excelling in the organizational arts-in maintaining control of a group's members, in eliciting predictable and satisfactory efforts from them, in holding a group together. In

making political decisions, the first and most important concern is not _ - .achieve.the aims the members of an organization may have but to secure the continuity and health of the organization itself (cf. Diesing 1962: 198-204; Downs 1967:262-270).

t: Along with the advantages of hierarchic orders go the costs. In hierarchic orders, moreover, the means of control become an object of strug- gle. Substantive issues become entwined with efforts to influence or _ control the controllers. The hierarchic ordering of politics adds one to the already numerous objects of struggle, and the object added is at a

new order of magnitude.

- v. If the risks of war are unbearably high, can they be reduced by .I!dlllJz.m~ to manage the affairs of nations? At a minimum, management requires controlling the military forces that are at the disposal of states. Within nations, organizations have to work to maintain themselves. As organizations, nations, in working to maintain themselves, sometimes have _ .. to use force against dissident elements and areas. As hierarchical systems, governments nationally or globally are disrupted by the defection of major parts. In a SOciety of states with little coherence, attempts at world government would founder on the inability of an emerging central au_ thority to mobilize the resources needed to create and maintain the unity of the system by regulating and managing its parts. The prospect of _ world government would be an invitation to prepare for world civil war. calls to mind Milovan Djilas's reminiscence of World War II. Ac- cording to him, he and many Russian soldiers in their wartime discussions came to believe that human struggles would acquire their ultimate bitterness jf all men were subject to the same social system, "for the system would be untenable as such and various sects would undertake the reckless destruction of the human race for the sake of its greater 'happiness' "

110

Kenneth N. l%ltz

· ANARCHIC ORDERS AND BALANCES OF POWER

III

(1962:50). States cannot entrust managerial powers to a central agency unless that agency is able to protect its client states. The more powerful the clients and the more the power of each of them appears as a threat to the others, the greater the power lodged in the center must be. The. greater the power of the center, the stronger the incentive for states to engage in a struggle to control it.

States, like people, are insecure in proportion to the extent of their .. freedom. If freedom is wanted, insecurity must be accepted. Organizations that establish relations of authority and control may increase security as they decrease freedom. If might does not make right, whether among people or states, then some institution or agencYchas intervened to- 11ft them out of nature's realm. The more influential the agency, the stronger the desire to control it becomes. In contrast, units in an anarchic order act for their own sakes and not for the sake of preserving an organization and furthering their fortunes within it. Force is used for one's own interest. In the absence of organization, people or states are free to leave one another alone. Even when they do not do so, they are better able, in the absence of the politics of the organization, to concentrate on the politics of the problem and to aim for a minimum .. agreement that will permit their separate existence rather than a max-' imum agreement for the sake of maintaining unity. If might decides, then ' bloody struggles over right can more easily be avoided.

. Nationally, the force of a government is exercised in the name of right and justice. Internationally, the force of a state is employed for the sake of its own protection and advantage. Rebels challenge a government's, claim to authority; they question the rightfulness of its rule. Wars among states cannot settle questions of authority and right; they can only determine the allocation of gains and losses among contenders and settle for a time the question of who is the stronger. Nationally, relations of ' . authority are established. Internationally, only relations of strength result. ' Nationally, private force used against a government threatens the political system. Force used by a state--a public body-s-Is, from the international perspective, the private use of force; but there is no government to overthrow and no governmental apparatus to capture. Short of a drive toward world .. hegemony, the private use of force does not threaten the system of international politics, only some of its members. War pits some states against others in a struggle among similarly constituted entities.

· The power of the strong may deter the weak from asserting their claims, ,. not because the weak recognize a kind of rightfulness of rule on the -, part of the strong, but simply because it is not sensible to tangle with Conversely, the weak may enjoy considerable freedom of action if are so far removed in their capabilities from the strong that the latter are not much bothered by their actions or much concerned by

. increases in their capabilities.

;.') National politics is the realm of authority, of administration, and of law. International politics is. the realm of power, of struggle, and of , accommodation. The international realm is preeminently a political one. national realm is variously described as being hierarchic, vertical, centralized, heterogeneous, directed, and contrived; the international realm, as being anarchic, horizontal, decentralized, homogeneous, undirected, and mutually adaptive. The more centralized the order, the nearer to the top the locus of decisions ascends. Internationally, decisions are at the bottom level, there being scarcely any other. In the vertical

· horizontal dichotomy, international structures assume the prone position. Adjustments are made internationally, but they are made without a formal or. authoritative adjuster. Adjustment and accommodation proceed by mutual adaptation (cf. Barnard 1948:148-152; PolanyiI941:428-456). Action and reaction, and reaction to reaction, proceed by a piecemeal , process. The parties feel each other out, so to speak, and define a .. situation simultaneously with its development. Among coordinate units, , adjustment is achieved and accommodations arrived at by the exchange

• of "considerations;' in a condition, as Chester Barnard put it, "in which the duty of command and the desire to obey are essentially absent"

• (pp, 150-151). Where the contest is over considerations, the parties seek to maintain or improve their positions by maneuvering, by bargaining, .or by fighting. The manner and intensity of the competition is determined by the desires and the abilities of parties that are at once separate and interacting.

Whether or not by force, each state plots the course it thinks will best serve its interests. If force is used by one state or its use is expected, .. the recourse of other states is to use force or be prepared to usc it singly or in combination. No appeal can be made to a higher entity clothed with the authority. and equipped with the ability to act on its own initiative. Under such conditions the possibility that force will be used

112

Kenneth N. WaJcz

II3

ANARCHIC ORDERS AND BALANCES OF POWER

by one or another of the partles looms always as a threat in the background. In polities force is said to be the ultima ratio. In international politics force serves, not only as the ultima ratio, but indeed as the first and constant one. To limit force to being the ultima ratio of politics implies, in the words of Ortega y Gasset, "the previous submission of force to methods of reason" (quoted in Johnson 1966:13). The constant possibility .. that force will be used limits manipulations; moderates demands, and serves as an incentive for the settlement of disputes. One who knows that pressing too hard may lead to war has strong reason to consider • whether possible gains are worth the risks entailed. The threat of force . internationally is comparable to therole __ (~.rthe strike_iI'lI_abor and managemend)ariai~i~g>The few strikes that take place are in a sense;' as . Livernash has said, "the cost of the strike option which produces settlements in the large mass of negotiations" (1963:430). Even if workers ... seldom strike, their doing so is always a possibility. The possibility of - .. industrial disputes leading to long and costly strikes encourages labor and management to face difficult issues, to try to understand each other's .. problems, and to work hard to find accommodations. The possibility that. conflicts among nations may lead to long and costly wars has similarly sobering effects.

students are inclined to see a lessening of anarchy in each outbreak of peace. Since world politics, although not formally organized, is not en. tirely without institutions and orderly procedures, students are inclined to see a lessening of anarchy when alliances form, when transactions

· across national borders increase, and when international agencies multiply. Such views confuse structure with process, and I have drawn attention to that error often enough.

, Second, the two simple categories of anarchy and hierarchy do not seem to accommodate the infinite social variety our senses record. Why insist on reducing the types of structure to two instead of allowing for

· a greater variety? Anarchies are-ordered by the juxtaposition of similar units, but those similar units are not identical. Some specialization by function develops among them. Hierarchies are ordered by the social division of labor among units specializing in different tasks, but the resemblance of units does not vanish. Much duplication of effort contin-

· ues. All societies are organized segmentally or hierarchically in greater - or lesser degree. Why not, then, define additional social types according to the mixture of organizing principles they embody? One might conceive of some societies approaching the purely anarchic, ofothers approaching the purely hierarchic, and of still others reflecting specified mixes of the .- two organizational types. In anarchies the exact likeness of units and the ..• determination of relations by capability alone would describe a realm wholly of politics and power with none of the interaction of units gUided by administration and conditioned by authority In hierarchies the complete differentiation of parts and the full specification of their functions would produce a realm wholly of authority and administration with none of the interaction of parts affected by politics and power. Although such pure orders do not exist, to distinguish realms by their organizing principles is nevertheless proper and important.

Increasing the number of categories would bring the classification of societies closer to reality. But that would be to move away from a theory claiming explanatory power to a less theoretical system promising greater descriptive accuracy. One who wishes to explain rather than to describe should resist moving in that direction if resistance is reasonable. Is it? What does one gain by insisting on two types when admitting three or . four would still be to simplify boldly? One gains clarity and economy of

concepts. A new concept should be introduced only to cover matters

5. Anarco/ and Hierarchy I have described anarchies and hierarchies as though every political order were of one type or the other. Many, and I suppose most, political scientists who write of structures allow for a greater, and sometimes for .a bewildering, variety of types. Anarchy is seen as one end of a continuum whose other end is marked by the presence of a legitimate and competent government. International politics is then described as being flecked wi~h particles of government and alloyed with elements of community-supranational organizations whether universal or regional, alliances, multinational corporations, networks of trade, and what not. Internationalpolitical systems are thought of as being more or less anarchic.

Those who view the world as a modified anarchy do so, it seems, for two reasons. First, anarchy is taken to mean not just the absence of government but also the presence of disorder and chaos. Since world polities, although not ,reliably peaceful, falls short of unrelieved chaos,

I14

Kenneth N. Walcz

ANARCHIC ORDERS AND BALANCES OF POWER

lIS

that existing concepts do not reach. If some societies are neither anarchic nor hierarchic, if their structures are defined by some third "rriprinlT principle, then we would have to define a third system.' All societies mixed. Elements in them represent both of the ordering principles. does not mean that some societies are ordered according to a principle. Usually one can easily identify the principle by which a .n(·iptv~ is ordered. The appearance of anarchic sectors within hierarchies not alter and should not obscure the ordering principle of the system, for those sectors are anarchic only within limits. The and behavior of the units populating those sectors within the larger

system difTer, moreover, from what they would be and how they would . behave outside of it. Firms in oligopolistic markets again are perfect examples of this. They struggle against one another, but because need not prepare to defend themselves physically, they can afford specialize and to participate more fully in the division of economic than states can. Nor do the states that populate an anarchic world

It impossible to work with one another, to make agreements their arms, and to cooperate in establishing organizations. H elements within international structures limit and restrain the exercise of sovereignty but only in ways strongly conditioned by the anarchy of the larger system. The anarchy of that order strongly affects the of cooperation, the extent of arms agreements, and the jurisdiction of international organizations.

But what about borderline 'cases, societies that are neither clearly anarchic nor clearly hierarchic? Do the), not represent a third type? To say that there are borderline cases is not to say that at the border a third' type of system appears. All categories have borders, and if we have any .' categories at all, we have borderline cases. Clarity of concepts does not . eliminate difficulties of classification. Was China from the 1920s to the 1940s a hierarchic or an an anarchic realm? Nominally a nation, China looked more like a number of separate states existing alongside one another. Mao Tse-tung in 1930, like Bolshevik leaders earlier, thought that striking a revolutionary spark would "start a prairie fire:' Revolu- . tionary flames would spread across China, if not throughout the world. Because the interdependence of China's. provinces, . like the interdependence of nations, was insufficiently close, the flames failed to spread. So nearly autonomous were China's provinces that the effects of war in one .:

. part of the country were only weakly registered in other parts. Battles . the Hunan hills, far from sparking a national revolution, were hardly in neighboring provinces. The interaction of largely self-sufficient 'nrl~''''nrf>~ was slight and sporadic. Dependent neither on one another

nor on the nation's center politically, they were not subject

. the close interdependence characteristic of organized and integrated

As a practical matter, observers may disagree in their answers to such

. as just when did China break down into anarchy, or whether countries of \Vestern Europe are slowly becoming one state or stubremaining nine. The point of theoretical importance is that our PVT,pri·"ti,~n~ about the fate of those areas differ widely dependIng on answer to the structural question becomes the right one. Strucdefined according to two distinct ordering principles help to explain irnlnn.-j-",nt aspects of social and political behavior. That is shown in various

in the following pages. This section has explained why two, and

two, types of structure are needed to cover societies of all sorts.

II

"How can a theory of international politics be constructed? Just as any must be. As chapters 1 and 4 [articles 2 and 3 above] explain, first, one must conceive of international politics as a bounded realm or . domain; second, one must discover some law-like regularities within it; '. and third, one must develop a way of explaining the observed regularities. The first of these was accomplished in chapter 5 [article 4]. This chapter . so far has shown how political structures account for some recurrent aspects of the behavior of states and for certain repeated and enduring patterns. Wherever agents and agencies are coupled by force and competition rather than by authority and law, we expect to find such behaviors and outcomes. They are closely identified with the approach to politics suggested by the rubric, Realpolitik. The elements of Realpolitik, . exhaustively listed, are these: The ruler's, and later the state's interest provides the spring of action; the necessities of policy arise from the

unregulated competition of states; calculation based on these necessities can discover the policies that will be~t serve a state's interest; success is

II6

Kenneth N. Waltz

the ultimate test of policy, and success is defined as preserving and " strengthening the state. EYer since Machiavelli, interest and necessity- ' and raison d'ecat, the phrase that comprehends them-have remained the key concepts of Realpolitik. From Machiavelli through Meinecke and Mor~ genthau the elements of the approach and the reasoning remain constant.', Machiavelli stands so clearly as the exponent of Realpolitik that one easily slips into thinking that he developed the closely associated idea of balance'

of power as well. Although he did not, his conviction that politics can be explained in its own terms established the ground on which balance- . of-power theory can be built.

Realpolitik indic.::ates thefriethodsbywhich foreign policy is conducted ' and provides a rationale for them. Structural constraints explain why the ' methods are repeatedly used despite differences in the persons and states who use them. Balance-of-power theory purports to explain the result that such methods produce. Rather, that is what the theory should do. If there is any distinctively political theory of international politics, bal- ," ' ance-of-power theory is it. And yet one cannot find a statement of the theory that is generally accepted. Carefully surveying the copious balanceof-power literature, Ernst Haas discovered eight distinct meanings of the term, and Martin Wight found nine (1953, 1966). Hans Morgenthau, in his profound historical and analytic treatment of the subject, makes use off our different definitions (1948/1973). Balance of power is seen by some as being akin to a law of nature; by others, as simply an outrage. Some view it as a guide to a statesmen; others as a cloak that disguises their imperialist policies. Some believe that a balance of power is the best guarantee of the security of states and the peace of the world; others, that it has ruined states by causing most of the wars they have fought.l

To believe that one can cut through such confusion may seem quixotic, I shall nevertheless try. It will help to hark back to several basic propositions about theory. (1) A theory contains at least one theoretical assumption. Such assumptions are not factual. One therefore cannot legitimately ask if they are true, but only if they are useful. (2) Theories must be evaluated in terms of what they claim to explain. Balance-ofpower theory claims to explain the results of states' actions, under given conditions, and those results may not be foreshadowed in any of the actors' motives or be contained as objectives in their policies. (3) Theory, as a general explanatory system, cannot account for particularities.

ANARCHIC ORDERS AND BALANCES OF POWER

Il7

, Most of the confusions in balance-of-power theory, and criticisms of it, derive from misunderstanding these three points. A balance-of-power theory, properly stated, begins with assumptions about states: They are unitary actors who, at a minimum, seek their own preservation and, at a maximum, drive for universal domination. States, or those who act for , them, try in more or less sensible ways to use the means available in ~rder to achieve the ends in view: Those n?eans fall into two categories: , mternal efforts (moves to increase econorlilc capability, to increase mil-

itary strength, to develop clever strategies) and external efforts (moves to strengthen and enlarge one's own alliance or to weaken and shrink an 'opposing one). The external game ()f alignment and re~lignm:ent're- . quires three or more players, and it is usually said that balance-of-power systems require at least that number. The statement is false, for in a two-power system the politics of balance continue, but the way to compensate for an incipient external disequilibrium is primarily by intensifying one's internal efforts. To the assumptions of the theory we then add the condition for its operation: that two or more states coexist in a self-help system, one with no superior agent to come to the aid of states that may be weakening or to deny to any of them the use of whatever instruments they think will serve their purposes. The theory, then, is built up from the assumed motivations of states and the actions that correspond to them. It describes the constraints that arise from the system that those actions produce, and it indicates the expected outcome:

" namely, the formation of balances of power. Balance-of-power theory is microtheory precisely in the economist's sense. The system, like a market in economics, is made by the actions and interactions of its units, and the theory is based on assumptions about their behavior.

A self-heleJk~~"l)1 is one in which those who do not help themselves, or who do so less effectively than others, will fail to prosper, will lay themselves open to dangers, will suffer. Fear of such unwanted consequences stimulates states to behave in ways that tend toward the creation of balances of power. Notice that the theory requires no assumptions of rationality or, of constancy of will on the part of all of the actors. The theory says simply that if some do relatively well, others will emulate them or fall by the wayside. Obviously, the system won't work if all states lose interest in preserving themselves. It will, however, continue to work if some states do, while others do not, choose to lose their political

Kenneth N. Waltz

IIS identities, say, through amalgamation. Nor need it be assumed that all of the competing states are striving relentlessly to increase their power. The possibility that force may be used by some states to weaken or destroy others does, however; make it difficult for them to break out of

the competitive system.

The meaning and importance of the theory are made clear by ex-

amining prevalent misconceptions that are theoretical, not factual. One of the most common misunderstandings of balance-of-power theory centers on this point. The theory is criticized because its assumptions are erroneous. The following statement can stand for a host of others:

If nations were in fact unchanging units with no permanent ties to each other, and if all were motivated primarily by a drive. to maximize their power, except for a single balancer whose aim was to prevent any nation from achieving preponderant power, a balance of power might in fact result. But we have seen that these assumptions are not correct, and since the assumptions of the theory are wrong, the conclusions are also in error (Organski 1968:292).

The author's incidental error is that he has compounded a sentence some parts of which are loosely stated assumptions of the theory, and other parts not. His basic error lies in misunderstanding what an assumption is. From previous discussion, we know that assumptions are neither true nor false and that they are essential for the construction of theory. We can freely admit that states are in fact not unitary, purposive actors. States pursue many goals, which are often vaguely formulated and inconsistent. They fluctuate with the changing currents of domestic politics, are prey to the vagaries of a shifting cast of political leaders, and are influenced by the outcomes of bureaucratic struggles. But all of this has always been known and it tells us nothing about the merits of balance-

of-power theory.

A further confusion relates to our second proposition about theory.

Balance-of-power theory claims to explain a result (the recurrent formation of balances of power), which may not accord with the intentions of any of the units whose actions combine to produce that result. To contrive and maintain a balance may be the aim of one or more states, but then again it may not be. According to the theory, balances of power tend to form whether so:ne or all states consciously aim to establish and

ANARCHIC ORDERS AND BALANCES OF POWER

1I9

maintain·a balance, or whether some or all states aim for universal domination! Yet many, and perhaps most, statements of balance- of-power theory attribute the maintenance of a balance to the separate states as a motive. David Hurne, in his classic essay "Of the Balance of Power,' offers "the maxim of preserving the balance of power'" as a constant rule of prudent politics (1742:142-144). So it may be, but it has proved to be an unfortunately short step from the belief that a high regard for preserving a balance is at the heart of wise statesmanship to the belief that states must follow the maxim if a balance of power is to be maintained. This is apparent in thefirst of Morgenthau's four definitions of the term: namelY, "a policy aimed at a certain state of affairs?' The reasoning then easily becomes tautological. If a balance of power is to be maintained,

the policies of states must aim to uphold it. If a balance of power is in fact maintained, we can conclude that their aim was accurate. If a balance of power is not produced, we can say that the theory's assumption is erroneous. Finally, and this completes the drift toward the reification of a concept, if the purpose of states is to uphold a balance, the purpose of the balance is "to maintain the stability of the system without destroying the multiplicity of the elements composing it?' Reification has obviously occurred where one reads, for example, of the balance operating "successfully" and of the difficulty that nations have in applying it (1948/1973:167-174,202-207).

Reification is often merely the loose use ofJanguage or the employment of metaphor to make one's prose more pleaSing. In this case, however; the theory has been drastically distorted, and not only by introducing the notion that if a balance is to be formed, somebody must want it and must work for it. The further distortion of the theory arises when rules are derived from the results of states' actions and then illogically prescribed to the actors as duties. A possible effect is turned into a necessary cause in the form of a stipulated rule. Thus, it is said, "the balance of power" can "impose its restraints upon the power aspirations of nations" only if they first "restrain themselves by accepting the system of the balance of power as the common framework of their endeavors?' Only if states recognize "the same rules of the game" and play "for the same

... limited stakes"can the balance of power fulfill "its functions for international stability and national independence" (Morgenthau 19481 1973:219-220).

Kenneth N. Walcz

12.0

The closely related errors that fall under our second proposition about theory are, as we have seen, twin traits of the field of international politics: namely, to assume a necessary correspondence of motive and result and to infer rules for the actors from the observed results of their action. What has gone wrong can be made clear by recalling the economic analogy (pp. 81-87, above). In a purely competitive economy, everyone's striving to make a profit drives the profit rate downward. Let the competition continue long enough under static conditions, and everyone's profit will be zero. To infer from that result that everyone, or anyone, is seeking to minimize profit, and that the competitors must adopt that

. goal· as a . rule in order for the svstem to work. would be absurd .. ~l1d_ yet in international politics one frequently finds that rules inferred from the results of the interactions of states are prescribed to the actors and are said to be a condition of the system's maintenance. Such errors, often made, are also often pointed out, though seemingly to no avail. S. E Nadel has put the matter simply: "an orderliness abstracted from behaviour cannot guide behaviour" (Nadel 1957:148; cf. Durkheim 1893:366,418;

Shubik 1959:11, 32).

Analytic reasoning applied where a systems approach is needed leads

to the laying down of all sorts of conditions as prerequisites to balances of power forming and tending toward equilibrium and as general pre- - conditions of world stability and peace. Some require that the number of great powers exceed two; others that a major power be willing to play the role of balancer. Some require that military technology not change radically or rapidly; others that the majo-r states abide by arbitrarily specified rules. But balances of power form in the absence of the "necessary" conditions, and since 1945 the world has been stable, and the world of major powers remarkably peaceful, even though international conditions have not conformed to theorists' stipulations. Balance-ofpower politics prevail wherever two, and only two, requirements are met: that the order be anarchic and that it be populated by units wishing to

survive.

For those who believe that if a result is to be produced, someone, or

everyone, must want it and must work for it, it follows that explanation turns ultimately on wh;t the separate statesal'-e like. If that is true, then theories at the national level, or lower, will sufficiently explain interna-

. ANARCHIC ORDERS AND BALANCES OF POWER

- .. :J ....

- -

-0'. ~

~-- "

-,

12.1

».

tional politics. If, for example, the equilibrium of a balance is maintained through states abiding by rules, then one needs an explanation of how agreement on the rules is achieved and maintained. One does not need a balance-of-power theory, for balances would result from a certain kind of behavior explained perhaps by a theory about national psychology or bureaucratic politics. A balance-of-power theory could not be constru:ted because it would have nothing to explain. If the good or bad motives of states result in their maintaining balances or disrupting them, then the notion of a balance of power becomes merely a framework organizing one's account of what happened, and that is indeed its cus-

__ tol11aryuse.A.constructionthat starts out to be a theory ends up as a set of categones. Categories - then multiply fapidly to cover events that the embryo ~heory had not contemplated. The quest for explanatory power turns mto a search for descriptive adequacy.

Finally, and related to our third proposition about theory in general, balance-of-power theory is often criticized because it does not explain the particular policies of states. True, the theory does not tell us why state X made a certain move last Tuesday. To expect it to do so would be like expecting the theory of universal gravitation to explain the wayward path of a falling leaf. A theory at one level of generality cannot answe~ ques~io~s about matters at a different level of generality. Failure to notice this IS one error on which the criticism rests. Another is to mistak~ a theory of international politics for a theory of foreign policy. Confusion about the explanatory claims made by a properly stated bal-.: ance-of-power theory is rooted in the uncertai~ty of the distinction drawn between national and international politics or in the denials that the distinction should be made. For those who deny the distinction, for th~se who dev~se explanations that are entirely in terms of interacting unJ:s, explanations of international politics are explanations of foreign policy, and explanations of foreign policy are explanations of international politics. Others mix their explanatory claims and confuse the problem of u~dersta~ding international politics with the problem of understanding f~re~gn polI~y. Mo~genthau, for example, believes that problems of predlctmg foreign policy and of developing theories about it make international-political theories difficult, if not impossible, _ to contrive (1970: 253-258). But the difficulties 'of explaining foreign policy work

~ '_:,_.

12.4

Kenneth N. Waltz

is a difficult and subtle task, made so by the interdependence of fact and theory, by the elusive relation between reality and theory as an instrument for its apprehension. Questions of truth and falsity are somehow involved, but so are questions of usefulness. and uselessness. In the end, one sticks with the theory that reveals most, even if its validity is suspect. I shall say more about the acceptance and rejection of theories elsewhere. Here I say only enough to make the relevance of a few examples of theory testing clear. Others can then easily be thought of. Many are provided in the first part of this chapter and in all parts of the next three, although I have not always labeled them as tests or put them in testable form. __Tests_ a~e ~e~sy to think up, once one has a theory to test, but they are hard to carry through.- Given the difficUlty of testinga-ny theOij,al1d the added difficulty of testing theories in such nonexperimental fields as international politics, we should exploit all of the ways of testing I have mentioned-by trying to falsify, by devising hard confirmatory tests, by comparing features of the real and the theoretical worlds, by comparing behaviors in realms of similar and of different structure. Any good theory raises many expectations. Multiplying hypotheses and varying tests are all the more important because the results of testing theories are necessarily problematic. That a single hypothesis appears to hold true may not be very impressive. A theory becomes plausible if many hypotheses inferred from it are successfully subjected to tests.

Knowing a little bit more about testing, we can now ask whether expectations drawn from our theory can survive subjection to tests. What will some of the expectations be? 1\\'0 that are closely related arise in the above discussion. According to the theory, balances of power recurrently form, and states tend to emulate the successful policies of others. Can these expectations be subjected to tests? In principle, the answer is "yes" Within a given arena and over a number of years, \,:e should find

, the military power of weaker and smaller states or groupings of states' growing more rapidly, or shrinking more slowly, than that of stronger and larger ones. And we should find widespread imitation among competing states. In practice, to check such expectations against historical

observa tions is difficult. '

1\\'0 problems are paramount. First, though. balance-of-power theory offers some predictions, the predictions are indeterminate. Because only a loosely defined and inc,~nstantconditioll' of balance is predicted, it is

ANARCHIC ORDERS AND BALANCES OF POWER

"'~.-" .. "- .. ' .. '.J"'.'"

. -~.

-~-. ;

.. 1.'

"

-~-

1d_

12.5

(:

difficult to say that any given distribution of power falsifies the theory. The theory, moreover, does not lead one to expect that emulation among states will proceed to the point where competitors become identical. What will be imitated, and how quickly and closely? Because the theory does not give precise answers, falsification again is difficult. Second, although states may be disposed to react to international constraints and incentives in accordance with the theory's expectations, the policies and actions of states are also shaped by their internal conditions. The failure of balances to form, and the failure of some states to conform to the successful practices of other states, can too easily be explained away by pointing to effects produced by forces that lie' outside of the theory's

--purview.

In the absence of theoretical refinements that' fix expectations with certainty and in detail, what can we do? As I have just suggested, and as the sixth rule for testing theories set forth in chapter 1 (article 2 above) urges, we should make tests ever more difficult. If we observe outcomes that the theory leads us to expect even though strong forces work against them, the theory will begin to command belief. To confirm the theory one should not look mainly to the eighteenth-century heyday of the balance of power when great powers in convenient numbers interacted and were presumably able to adjust to a shifting distribution of' power by changing partners with a grace made possible by the absence of ideological and other cleavages. Instead, one should seek confirmation through observation of difficult cases. One should, for example, look for instances of states allying, in accordance with the expectations the theory gives rise to, even though they have strong reasons not to cooperate with one another. The alliance of France and Russia, made formal in 1894, is one such instance. One should, for example, look for instances of states making internal efforts to strengthen themselves, however distasteful or difficult such efforts might be. The United States and the Soviet Union following World War II provide such instances: the United States by rearming despite having demonstrated a strong wish not to by dismantling the most powerful military machine the world had ever known; the Soviet Union by maintaining about three million men under arms while striving to acquire a costly new military technology despite the terrible destruction she had suffered in war.

These examples tend to confirm the theory. We find states fo~ing

~- .

Kenneth N. Walcz

balances of power whether or not they wish to. They also show the difficulties of testing. Gennany and Austria-Hungary formed their Dual Alliance in 1879. Since detailed inferences cannot be drawn from the theory, we cannot say just when other states are expected to counter this move. France and Russia waited until 1894. Does this show the theory false by suggesting that states mayor may not be brought into balance? We should neither quickly conclude that it does nor lightly chalk the delayed response off to "friction?' Instead, we should examine diplomacy and policy in the IS-year interval to see whether the theory serves to explain and broadly predict the actions and reactions of states and to

.... see whether the delay is out of accord with the theory. Careful judgment is needed. For this, hi;tori~;;~' accounts serve better thanthe historical summary I might provide.

The theory leads us to expect states to behave in ways that result in balances fonning. To infer that expectation from the theory is not impressive if balancing is a universal pattern of political behavior, as is sometimes claimed. It is not. Whether political actors balance each other or climb on the bandwagon depends on the system's structure. Political parties, when choosing their presidential candidates, dramatically illustrate both points. When nomination time approaches and no one is established as the party's strong favorite, a number of would-be leaders contend. Some of them form coalitions to check the progress of others. The maneuvering and balancing of would-be leaders when the party lacks one is like the external behavior of. states. But this is the pattern only during the leaderless period. As soon as someone looks like the winner, nearly all jump on the bandwagon rather than continuing to build coalitions intended to prevent anyone from winning the prize of power. Bandwagoning, not balancing, becomes the characteristi~ behavior."

Bandwagoning and balancing behavior are in sharp contrast. Internally, lOSing candidates throw in their lots with the winner. Everyone wants someone to win; the members of a party want a leader established even while they disagree on who it should be. In a competition for the position of leader, bandwagoning is sensible behavior where gains are possible even for the losers and where losing does not place their security in jeopardy. Externally, states work harder to increase their own strength, or they combine with others, if they are falling behind. In a competition for the position of leader, balan~i.ng is sensible behavior where the victory of one

ANARCHIC ORDERS AND BAUNCES OF POWER . >27 ------:';j' coalition over another leaves weaker members of the winning coalition .1 at the mercy of the stronger ones. Nobody wants anyone else to win;

none of the great powers wants one of their number to emerge as the

leader. ~

" . If two coalitions form and one of them weakens, perhaps because of

the political disorder of a member, we expect the extent of the other coalition's military preparation to slacken or its unity to lessen. The classic example of the latter effect is the breaking apart of a war-winning coalition in or just after the moment of victory. We do not expect the strong to combine with the strong in order to increase the extent of their power over others, but rather to square off and look for allies who might help them. Iii- anarchy, security is the highest end. Only if ~u~~ival

is assured can states safely seek such other goals as tranquility, profit, and power. Because power is a means and not an end, states prefer to join the weaker of two coalitions. They cannot let power, a possibly useful means, become the end they pursue. The goal the system encourages them to seek is security. Increased power mayor may not serve that end. Given two coalitions, for example, the greater success of one

in draWing members to it may tempt the other to risk preventive war, hoping for· victory through surprise before disparities widen. If states wished to maximize power, they would join the stronger side, and we would see not balances forming but a world hegemony forged. This does not happen because balancing, not bandwagoning, is the behavior induced

by the system, The first concern of states is not to maximize power but.

to maintain their positions in the system.

Secondary states, if they are free to choose, flock to the weaker side; for it is the stronger side that threatens them. On the weaker side, they are both more appreciated and safer, provided, of course, that the coalition they join achieves enough defensive or deterrent strength to dissuade adversaries from attacking. Thus Thucydides records that in the Peloponnesian War the lesser city states of Greece cast the stronger Athens as the tyrant and the weaker Sparta as their liberator (circa 400 B.C., Book v, ch. 17). According to Werner Jaeger, Thucydides thought this "perfectly natural in the circumstances;' but saw "that the parts of tyrant and liberator did not correspond with any permanent moral quality in_these_states but were simply masks which would one day be interchanged to the astonishment of the beholder when the balance of power

128

Kenneth N. Waltz

was altered" (1939,1:397). This shows a nice sense of how the placement of states affects their behavior and even colors their characters. It also supports the proposition that states balance power rather than maximize it. States can seldom afford to make maximizing power their goal. International politics is too serious a business for that,

The theory depicts international politics as a competitive realm. Do states develop the characteristics that competitors are expected to display? The question poses another test for the theory. The fate of each state depends on its· responses to what other states do. The possibility that conflict will be conducted by force leads to competition in the arts and the instruments of force. Competition produces a teI14~ncy_toward the_ sameness of the competitors. Thus Bismarck's startling victories over Austria in 1866 and over France in 1870 quickly led the major continental rowers (and Japan) to imitate the Prussian military staff system, and the failure of Britain and the United States to follow the pattern simply indicated that they were outside the immediate arena of competition. Contending states imitate the military innovations contrived by the country of greatest capability and ingenuity. And so the weapons of major contenders, and even their strategies, begin to look much the same all over the world. Thus at the tum of the century Admiral Alfred von Tirpitz argued successfully for building a battleship fleet on the grounds that Germany could challenge Britain at sea only with a naval doctrine and weapons similar to hers (Art 1973b:16).

. The effects of competition are not confined narrowly to the military realm. Socialization to the system should also occur. Does it? Again, because we can almost always find confirming examples if we look hard, we try to find cases that are unlikely to lend credence to the theory. One should look for instances of states conforming to common international practices even though for internal reasons they would prefer not to. The behavior of the Soviet Union in its early years is one such instance .. The Bolsheviks in the early years of their power preached international revolution and flouted the conventions of diplomacy. They were saying, in effect, "we will not be socialized to this system?' The attitude was well expressed by Trotsky, who, when asked what he would do as foreign minister, replied, "I will issue some revolutionary proclamations to the peoples and then close up the joint" (quoted in Von Laue 1963:235). In a competitive arena, howe.ver, one party may need the assistance of others.

ANARCHIC ORDERS AND BALANCES OF POWER

129

Refusal to play the political game may risk one's own destruction. The pressures of competition were rapidly felt and reflected in the Soviet Union's diplomacy ". Thus Lenin, sending foreign minister Chicherin to the Genoa Conference of 1922, bade him fareweil with this caution:

"Avoid big words" (quoted in Moore 1950:204): Chicherin, who personified the carefully tailored traditional diplomat rather than the simply uniformed revolutionary, was to refrain from inflammatory rhetoric for the sake of working deals. These he successfully completed with that other pariah power and ideological enemy, Germany.

The close juxtaposition of states promotes their sameness through the disadvantages that arise from a failure to conform to successful practices. It is this- "sameness;' aneffectof the system, that is so often attributed to the acceptance of so-called rules of state behavior. Chiliastic rulers occasionally come to power. In power, most of them quickly change their ways. They can refuse to do so, and yet hope to survive, only if they rule countries little affected by the competition of states. The socialization of nonconformist states proceeds at a pace that is set by the extent of their involvement in the system. And that is another testable statement.

The theory leads to many expectations about behaviors and outcomes.

From the theory, one predicts that states will engage in balancing behavior, whether or not balanced power is the end of their acts. From the theory, one predicts a strong tendency toward balance in the system. The expectation is not that a balance, once achieved, will be maintained, but that a balance, once disrupted, will be restored in one way or another. Balances of power recurrently form. Since the theory depicts international . politics as a competitive system, one predicts more specifically that states will display characteristics common to competitors: namely, that they will imitate each other and become socialized to their system. In this chapter,

I have suggested ways of making these propositions more specific and concrete so as to test them. In remaining chapters, as the theory is elaborated and refined, additional testable propositions will appear. .

NOTES

1. Put differently, states face a "prisoners' dilemma?'. If each of two parties follows his own interest, both end up worse ofT than if each acted to achieve joint interests. For thorough examination of ' the logic of such situations, see Snyder and

I30

Kenneth N. WaIcz

Diesing 1977; for bri~f and suggestive international applications, see Jervis, January

1978. h . I .. ·11 id

2. Emile Durkhelrn's depiction of solidary and mec amc~ SOCI~tl:S S~I .P:OVI es

the _ best explication of the two ordering pririciples, an~ hIS logIC m hmlt.mg the types of society to two continues to be compelling despite the efforts of hIS many critics to overthrow it (see esp. 1893). I shall discuss the problem at some length in a future work.

3. Along with the explication of balance-of-power theory in the pages ~~at follow, the reader may wish to consult a historical study of balance-of-power politics in practice. The best brief work is Wight (1973).

4. Looking at states over a wide span of time and space, Dowty concludes that in no case were shifts in alliances produced "by considerations of an overall balance of power" (1969:95).

5. The confusion is widespread and runs both ways. Thus Herbert Simon t~inks -the goal of classical economic theorists is unattainable because he. wrongly. b~heves that they were trying "to predict the ~ehavior of ~at:~nal man WIthout makmg an

empirical investigation of his psychological propertIes (1957: 199). .

6. Stephen Van Evera suggested using "band wagoning" to serve as the opposIte of "bahincing:'

SIX

Continuity and Transformation the World Polity:

Toward a Neoreolist Synthesis

.

In

JOHN GERARD RUGGIE

IN The Rules if Sociological Method. Emile Durkheim sought-to establish the "social milieu;' or- SOciety itself, "as the determining factor of collective evolution" In tum, he took society to reflect not the mere summation of individuals and their characteristics, but "a specific reality which has its own characteristics?' And he attributed this social facticity to "the system formed by [individuals'] association, "by the fact oftheir combination?' Hence, "if the determining condition of social phenomena is, as we have shown, the very fact of association, the phenomena ought to vary with the forms of that association, i.e., according to the ways in which the constituent parts of society are grouped" (Durkheim 18QE: 116, 103, xlvii, 112). In sum, the possibilities for individual action in the short run, and collective evolution in the long run, were to be accounted for by the changing forms of social solidarity. I

Durkheim's methodological premise was controversial from the start, but over the years its influence has' \~aned and come to be felt largely indirectly, as through the analysis of "primitive social structures" by Claude Levi-Strauss (19?7). - Suddenly, it is enjoying a resurgence in the study of a social domain never contemplated by Durkheim: the international system. It is being adopted by - the most unlikely of followers:

American students of comparative and international politics. And it is as controversial as ever. Adherents share Durkheim's views that social totalities are the appropriate uni~ of analysis for the study - of collective phenomena, and forms of association within them the appropriate level

.':'.