Journal of Peace Research http://jpr.sagepub.com/

Stability and Polarity: New Paths for Inquiry Patrick James and Michael Brecher Journal of Peace Research 1988 25: 31 DOI: 10.1177/002234338802500104

The online version of this article can be found at: http://jpr.sagepub.com/content/25/1/31

Published by:

http://www.sagepublications.com

On behalf of: Peace Research Institute Oslo

Journal of Peace Research Replication Data

Additional services and information for Journal of Peace Research can be found at:

Email Alerts: http://jpr.sagepub.com/cgi/alerts

Subscriptions: http://jpr.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://jpr.sagepub.com/content/25/1/31.refs.html

>> Version of Record - Mar 1, 1988

What is This?

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 ISSN 0022-3433 Journal of Peace Research, vol. 25, no. 1, 1988 Stability and Polarity: New Paths for Inquiry* PATRICK JAMES & MICHAEL BRECHER McGill University

One of the most intractable debates in the field of world concerns the linkage of systemic polarity to international stability. Despite many theoretical expositions and limited testing, disagreement persists over which type of structure and distribution of power is most stable. The dialogue on polarity and stability has focused on the relative merits of bipolar and multipolar structures; in other words, which configuration of power centers is more stable, two or more than that number? Advocates of each system have their adherents and, for some time now, have agreed to disagree. Most of the debate on polarity and stability thus far has been cast in terms that do not facilitate its resolution. The objective of this paper is to reformulate the debate, in order to facilitate a more compelling empirical judgment of the competing claims. More specifically, this involves revision of the central concepts. Polarity, it will be asserted, cannot be assessed only in terms of distribution of power. The concept also should incorporate the notion of autonomous decision centers. With respect to instability, war is held to be a less comprehensive measurement than international crisis. It is recommended that renewed testing should focus on the linkage of polarity to stability as so defined.

1. Introduction reformulate the debate, in order to facilitate One of the most intractable debates in the a more compelling empirical judgment of field of world politics concerns the linkage the competing claims. Moreover, revival of of systemic polarity to international stability. this controversy may turn out to be con- Despite many theoretical expositions and structive and innovative in ways that have limited testing, disagreement persists over not been anticipated. which type of structure and distribution of While interesting theories about inter- power is most stable. The dialogue on national stability undoubtedly remain un- polarity and stability has focused on the rela- known, a better understanding of the past tive merits of bipolar and multipolar struc- record can only enhance prospects for their tures ; in other words, which configuration of discovery. Thus the call for a reassessment power centers is more stable, two or more of what is known does not imply a retreat. than that number? Advocates of each system According to one recent evaluation of the have their adherents and, for some time now, literature on international conflict, it may have agreed to disagree. mean exactly the opposite: As will become apparent from an overview theoretical articulation a task that is too of the literature, most of the debate on [is] rarely attended to m the field of international thus far has been cast quantitative polarity and stability relations. Theoretical articulation is the clarification in terms that do not facilitate its resolution. and extension of the theoretical structure of an exist- More specifically, a lack of falsifiable, de- ing paradigm. Clarification ... creates a stronger has impeded structure to expand; extension . demonstrates the ductively-derived propositions of a to and of the connection ability (or inability) paradigm predict, understanding presumed thus explam, previously inexplicable phenomena between systemic polarity and international (Morrow 1985, p. 474). stability. The objective of this study is to This exposition implies that a more creative understanding of polarity and stability * Earher versions of this article were presented at the depends upon a reevaluation of theoretical Annual Meetings of the American Political Science axioms and, where possible, extension of Association, August 1986, and the Peace Science Soci- their explanatory power.2 ety, November 1986. We are grateful to Nils Petter Gleditsch and two anonymous referees for their helpful suggestions, and to the Research Grants Office of 2. The Debate McGill University for word processing services. The first phase of the debate over polarity

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 32

and stability took place in the sixties. Waltz basis. The implications of polarity for stab- (1964) argued that bipolarity is inherently ility, it will be argued, are more complex. more stable than multipolarity for several They cannot be covered adequately by a reasons: first, the two blocs have a shared straightforward comparison of power distri- interest in a global balance of power, and butions with the frequency of war. the bloc patron-leaders have the capability of achieving this goal; secondly, there is less danger of miscalculation, of both capability 3. The Conceptual Basis of Polarity and and intent; thirdly, while crises would recur, Stability their escalation to war could be more effec- Why should the structure of an international tively prevented than in a multipolar struc- system affect its level of stability? Although ture ; and finally, other states on the this question has been posed many times, a periphery would be less able to destabilize convincing answer has not yet been forth- the international system.3 Deutsch & Singer coming. A fundamental problem with the (1964) asserted the contrary and pointed to literature is its tendency to focus only on several processes in support of conventional the international system in recent times. For wisdom on this issue.4 First, the outcome of example, the debate over bipolarity and conflict and war in a system of several major multipolarity two decades ago took place powers is more uncertain, leading to greater against the backdrop of a bipolar system in caution among them; secondly, the greater transition. Thus it focused on bipolarity and likelihood of cross-cutting cleavages among the system that preceded it, the multipolar the participants reduces the rigidity of their inter-World War years, 1919-1939. conflicts; thirdly, alignments are more flex- One of the basic problems in the literature ible ; and finally, arms races escalate more is the uncertain meaning of systemic polarity. slowly because of the reality of changing Nogee (1974, p. 1193) described the prob- alliances. lematic role of polarity in the following The two competing hypotheses were criti- manner: cized by Rosecrance (1966), who contended that an intermediate international system, The debate over whether a bipolar or multipolar which he termed is world is more stable has produced a variety of ’bi-multipolarity’, likely and conclusions... Part to be the most stable. The two responses contradictory preeminent of the difficulty lies in the vagueness and he would in both ambiguity states, argued, engage of the concept of polarity. Surprisingly, few of the cooperative and competitive relations, res- many analysts using polar concepts have explicitly training conflict but preventing hegemony defined them, nor is there a umversally recognized referent for international affairs. by their adversary. Zero-sum games would empirical a ’pole’ m be avoided. Conflicts would be limited in stakes. The possibility of war would be less, Most writers on the concept of polarity have and the consequences of war would be more defined it as the distribution of power/ tolerable.55 resources, or the number of poles/clusters, Although Waltz, Deutsch & Singer, and or the number of autonomous power centers, Rosecrance specified the rationale for their in an international system (Haas 1970, pp. hypotheses about the impact of polarity on 99,100; Bueno de Mesquita 1975,1978; Rap- stability, they did not test their views against kin, Thompson with Christopherson 1979; empirical evidence. From the late 1960s Wayman 1984; Hart 1985, p. 31; Sabrosky onward this gap was partly filled by several 1985). Others have referred to economic studies using Correlates of War (COW) Pro- power (Russett 1965, pp. 2-4) or ideology ject data from 1816 to 1965, by Haas’ (1970) (Rosecrance 1963, pp. 79-101, 169-191, use of evidence about international subsys- 193-215) as the basis of polarity. tems, and by Levy’s (1985) examination of Designation of a threshold for polar status wars over five centuries. These has troubled each of these approaches efforts toward testing, however, proceeded (Nogee 1974, pp. 1205-1211). Instead of from an inadequately specified conceptual looking forward in time to what the system

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 33 might become, those who participated in the such hybrid structure, with two preeminent debate over polarity and stability invariably centers of military power and multiple cen- argued over what existed in the near past ters of political decision. It therefore has and present, resulting in a one-dimensional elements of both bipolarity and multi- assessment of polarity. Casting a wider his- polarity. torical net will facilitate understanding of, As a more systematic exposition of power among other things, the global transition that and decision, consider the matrix which occurred a quarter of a century ago and its appears as Fig. 1. ~ It displays empirical refer- implications for stability. ents for the logically possible combinations More specifically, systemic transforma- of decisional and power distribution. With tions have not always entailed changes in respect to power, the structure is one of the distribution of power. The emergence monopoly (or, more accurately prepon- of new centers of decision, too, can lead to derance), duopoly or oligopoly. These desig- significant changes in an international sys- nations correspond respectively to one, two tem or subsystem without dramatically and more than two, centers of power. At altering the hierarchy or distribution of the same time, centers of decision may be military capability. This is a crucial distinc- distributed differently, across one, two, or tion which permits a more complete concep- more than two nodes. Examples are included tual articulation of polarity and stability. for both the dominant system and various Polarity, as used in this discussion, refers subsystems. to power and decisional stratification, in Illustrations of conjuncture between the other words, the structure of a system. two centers (power and decision) appear on Theoretically, the power dispersion of a sys- the diagonal. Rome at its height provides tem may be unipolar, bipolar, or multipolar, an example of unipolarity; for centuries it the rough counterparts to monopolistic, possessed overwhelming military might and duopolistic or oligopolistic, in the structure had no other independent centers of decision of market systems (Keohane & Nye 1977, p. with which to deal. Similarly, an unchal- 6 20).’ lenged Sweden presided over the subsystem Each of the above-noted distributions of of northern Europe through the better part power has different implications for political of the 18th century. processes at the international level. A uni- With respect to bipolarity, the US and the polar structure is characterized by an USSR led two highly discrete blocs of actors overwhelming concentration of military after World War II and also possessed capability in one entity which thereby shapes preponderant capability. A subsystemic the rules of the system, dominates relations example of bipolarity would be the pairing among lesser actors and asserts its hegemony of India and Pakistan in the South Asian at will from an acknowledged position at subcontinent since 1947. the apex of the power pyramid. Bipolarity Finally, the inter-World War period of indicates a concentration of power in two the twentieth century featured a multipolar relatively equal actors. The two polar centers structure. Along with the US and Japan, the determine the essential rules of the system, principal European states constituted inde- the conditions of stability and the outcomes pendent centers of decision and relatively of its major wars. Multipolarity signifies a equal power bases at the dominant system diffusion of military power among several level. A subsystem counterpart is Central (or relatively equal units within an international East) Africa in the post-colonial era; the system, at least three (tripolar), usually states in these regions are many, and power more. is not highly concentrated among them. Decisional and power stratification are Other cells in the matrix reveal a dis- identical in the three systems just described. juncture between the number of centers of However, it is also possible that the number decision and power. In all cases the number of decisional and power centers will be of decision centers is greater; other com- unequal. Polycentrism is an example of one binations are excluded. It would be incon-

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 34

Fig. 1. Centers of Power and Decision: Distributions at the Dominant and Subsystemic Levels*

* The top of each cell contains the example for the dominant system, while the bottom shows the subsystem example. ** Examples of dominant system monopoly coupled with two or more decisional nodes have not been detected. ceivable that a state with a noteworthy but with each state an independent endowment of power would fail to function decisional unit (Rosecrance 1963). It is also independently in terms of decision; even the tenable to argue that the current (post-1962) choice of isolationism would constitute an global system is polycentric. The US and individual selection of policy. In sum, every USSR stand far above all actors in terms of center of power is, ipso facto, a center of the concentration of military power. How- decision, but the reverse does not obtain. ever, that does not preclude the existence of Consider first an instance of power mono- other important centers of decision, ranging poly (preponderance) with more than one from London and Paris to Beijing and center of decision. In the case, of the Tokyo. At the subsystem level, the Greek Caribbean/Central America from 1959 city-states on the eve of the Peloponnesian onward, the US continued to have pre- War provide an excellent example of poly- ponderant military capability, but the regime centrism : Sparta and Athens were the pre- in Cuba existed outside of the US-led, mil- eminent powers, yet other states functioned itary-security coalition within the region, independently in decision making (Thu- hence two centers of decision. cydides 1930). When a duopoly of power exists along with three or more centers of decision, that is 4. The Axiomatic Basis of Polarity and polycentrism. An excellent illustration is the Stability French Revolution/Napoleonic epoch, with Having specified various configurations of France and the coalition of traditionalist polarity based on centers of decision and European powers - England, Austria, Prus- power, it is appropriate to explore their sia and Russia - exhibiting power bipolarity implications for international stability. The

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 35 fundamental question is this: Is there an Prior to the analysis of systemic costs, one optimal, that is, most stable, combination of further point should be made; it concerns power and decisional concentration? This the axiomatic basis of polarity and stability. leads to a discussion of the theoretical mean- The separation of power and decisional cen- ing of stability. ters will permit a more thorough analysis of One approach is to relate instability to costs, especially when compared to theor- the notion of cost. Instability, somehow etical discussions which have focused exclus- measured through the frequency and inten- ively on the distribution of power. sity of disruptive interaction, may be Fig. 2 displays the hypothetical cost func- regarded as costly to most, if not all, mem- tions based upon a varying number of inde- bers of the international system. It reflects pendent decision nodes.&dquo; There are two a lack of agreement over the allocation of kinds of decisional cost to consider: decision existing political, economic and military time and externalities. resources. In other words, instability is a Expected costs from decision time (i.e., form of preference revelation; conflict will be formulation) increase with the number of more pervasive when there is dissatisfaction independent decision centers. When there with regimes. In this perspective the question are more decisional nodes, all other things of polarity and stability becomes one of being equal, the time required to reach an determining which system is least costly to agreement is greater. More interests must 9 its members.9 be represented, making negotiations more

Fig. 2. Costs of Decision Makmg

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 36 complicated and time-consuming. This might duce fewer externalities. Is one system more also be referred to as the opportunity cost stable (meaning less costly than the other), arising from decision making; as more time given this tradeoff? It is postulated that costs and energy are devoted to bargaining, less of decision time (C) will rise sharply once becomes available for other activities. there are more than two decisional nodes, Costs also arise during formulation outweighing any advantages from reduced because of the creation of externalities. externalities. With only two centers, When very few independent decision centers coalitional dynamics cannot take place. By exist, other states in the system may disap- contrast, three (or more) decision centers prove of some (or all) aspects of an agree- may result in time-consuming shifts in align- ment. These costs, which affect third parties, ment and related bargaining positions. As become less severe with a larger number of for externalities, the inclusion of two inter- independent decision centers, because the ests obviously is less attractive than three or interests represented are more diverse. more, but the extent depends upon the total Thus, as Fig. 2 suggests, there is a tradeoff number of states in the system. At present, involving costs of decision time (C) and for example, there are about 160 members of externalities (D). In any system (or subsys- the . The difference between tem), the number of independent decision two or three centers - or even the his- centers to be consulted ranges from 1 to N, torically prominent five - is negligible com- with N representing all system members. The pared to the proportion of states which, in costs of decision time increase with the num- each instance, is excluded from the apex of ber of centers, while the reverse obtains for decision masking. 12 In sum, a system with two externalities. When curves C and D are sum- decisional nodes is hypothesized to be less mated, the minimal level is reached at an costly. Thus ’K’ in Fig. 2 is regarded as unspecified point K between 1 and N. equal to two, the minimal number beyond Once a decision has been reached, costs of dictatorship. implementation must be considered. These An analysis of decision centers results in costs could include setting up an organization an expectation of greater stability from to monitor compliance with an agreement bipolarity. What, then, are the implications and to enforce rules. The costs might be of variable power distribution for the three shared informally, with the power centers systems? somehow coordinating implementation. In With regard to implementation, there are either instance, concentration of power will two types of costs to consider: fixed and lead to greater efficiency in implementation, variable. Fixed costs are those of establishing all other things being equal. As a problem some venture; in economic terms, that would of collective action, implementation of a refer to the cost of the physical plant, pur- decision is facilitated by a smaller number of chase of equipment and the like. In the power centers, with a power monopoly being present context, the fixed costs of implemen- ideal in the context of this specific problem. tation refer to the obstacles to collective With respect to the structure of inter- action, which always increase as the number national systems, these cost factors can be of essential participants becomes greater. In used to argue that bipolarity will be more terms of these fixed costs, therefore, bipolar- stable than multipolarity or polycentrism.ll ity is equal to polycentrism and preferable Of course, as Fig. 1 illustrates, other systems to multipolarity. are possible, but these, especially the first Variable costs are those which fluctuate two, are the configurations which have been over time. Labor is the most obvious central to the debate. example in the economic realm. The amount Consider first the impact of decision costs. of labor required for production will vary Bipolarity, that is, two independent decision over time. With respect to international centers, promises lower costs of decision relations, periodic disruption (and therefore time while multipolarity (or polycentrism), the need for stabilization) is expected to be that is, many decision centers, should pro- a function of the degree of status incon-

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 37 sistency in the system. In other words, if a ment. Polarity, the independent variable, state functions as an independent center of will be considered first, followed by the decision, but does not have military status, dependent variable, stability. that produces dissonance. Most empirical studies have been con- Such states usually are found in the devel- ducted by COW researchers. They utilized oping world. These system members desire the COW national capabilities index to change, especially in economic relations, yet assess power concentration and thereby lack the power to enforce it on the existing measure polarity. Each major power is international order. Instability therefore assigned a composite power score based on should become greater as the number of six indicators: standing armies, military decision centers increases over and above expenditures, fuel consumption, iron/steel the number of power centers. As a result, production, population and urban pop- polycentrism is expected to have higher vari- ulation. Concentration of capabilities then able costs, because neither bipolarity nor is represented by the standard deviation of multipolarity has such ’dissonant’ system the mean score among the major powers. members. The minimum would be zero, with absolute Taken together, costs of decision and equality among the states, and the maximum those associated with implementation favor would be 1, with one state controlling all of bipolarity. It therefore emerges as less costly the capabilities. Hence polarity is treated as than either polycentrism or multipolarity. a continuous variable. When comparing the latter two systems, The preceding operationalization of po- the combined variable and fixed costs of larity makes it somewhat difficult to apply implementation seem to favor multipolarity. the findings of COW research specifically to Regulating a polycentric system will be more the debate over bipolarity and multipolarity. costly because of status inconsistency result- Since discrete poles are not identified, direct ing from decolonization. It is very likely that comparison of systems with two or more the greater variable costs of polycentrism poles is not facilitated. The COW-oriented will outweigh the somewhat higher fixed studies explored the linkage of stability to costs of multipolarity. Although polycen- polarity, with each being treated as a con- trism has fewer power centers, and therefore tinuous (as opposed to categorical) variable. should be more able to achieve the initial Furthermore, the distribution of decision implementation of an agreement, over the centers is not incorporated in the measure- long-term its costs of maintenance will be ment of polarity. more substantial. Changes from one type of polarity (system Having compared the three systems in structure) to another can be measured, quali- theoretical terms, bipolarity has emerged as tatively, on the basis of transformations in most stable, followed by multipolarity and power distribution or the concentration/dif- polycentrism. The eventual task will be to fusion of decisions in international politics - test the underlying logic with a hitherto or both. Thus, in the period since World unused body of twentieth century data, in War I, the polarity phases are designated order to resolve the ‘great debate’ over struc- as multipolarity (1929-39), bipolarity (1945- ture and conflict. However, it is useful first 62), and polycentrism (1963- ). to examine the empirical findings thus far The entire inter-World War era was a and then specify a measurement strategy to multipolar systenl, with several relatively facilitate testing of the postulated rank order. equal great powers recognized as sharing the apex of the power pyramid. Unlike unipo- 5. Indicators and Measurement Strategy larity, bipolarity or polycentrism, the struc- The debate on polarity and stability has taken ture of the inter-World War system was place primarily at the level of theorizing characterized by flexibility in alliance about logically-expected relationships. Thus pattern, multiple centers of power and it is appropriate to devote some attention to decision, and uncertainty about behavior by the relatively neglected subject of measure- the other great powers in the system. That

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 38 system can be differentiated from the subsystems from 1649 to 1965: ’Multipolarity bipolarity which preceded World War I entails more violence, more countries at war, (1890-1914), with the Dual Alliance con- and more casualties; biopolarity brings fronting the Triple Entente, and which fol- fewer but longer wars.’ Cannizzo (1978, p. lowed World War II (1945-62), a period 957) observed that in the 20th century wars dominated by the US- and USSR-led blocs. involving major powers tended to follow In both cases, basic structural change result- ’periods of parity and rapid change toward ed from a long war of very high intensity. parity...’. And Levy (1985, p. 54), who Thus the onset of the Second World War in discovered a different relationship between September 1939 marks the effective end of polarity and stability in the European great multipolarity and the beginning of a six-year power system from 1495 to 1975, also used transition to a new system structure, war as the sole indicator: ’Bipolar systems bipolarity, with two hostile agglomer- have been historically more stable than ations - and centers - of power and multipolar systems.... The intensity and decision which, inter alia, set the limits of concentration of war have generally been independent behavior by both bloc members lower in bipolar periods, and general wars and unaffiliated state actors. All lasting have never occurred during bipolarity.’ decisions of consequence in world politics These studies are valuable but inadequate emanated from the . in terms of illuminating the polarity/stability The emergence of a distinctive polycentric link. First, they exclude all sources of sys- structure in the early 1960s is associated with temic instability other than war, even less the Cuban Missile Crisis and the Sino-Indian violent interactions among states such as border war, both in October-November serious or minor clashes. Moreover, their 1962. Together they indicated that power focus on the ’central [European] subsystem’ bipolarity and decision bipolarity in world ignores the myriad of local wars in the per- politics are not synonymous and that other ipheries which undermine the stability of centers of decision in the global system could international subsystems comprising the vast no longer be controlled by the US and the majority of states in twentieth-century world USSR. Polycentrism so defined has con- politics. Most important, their underlying tinued through the 1980s. view is conceptually limited and oper- It should be noted that the polarity phases ationally narrow. Although war is the gravest outlined above are based upon non-quanti- type of disturbance to system tranquility, it tative, historical evidence (Brecher & Wil- is not synonymous with instability. kenfeld 1987). Further research is intended A more comprehensive conceptual indi- to reveal whether these time periods vary cator of stability/instability than war prone- with respect to instability levels. If they are ness is disruptive interaction among states, different, that would provide a valuable link- which encompasses the widest possible range age between theory, which has not included of system disturbances. The corresponding references to the quantitative measurement operational indicator is crisis, a much of polarity, and the practice of testing, which broader phenomenon than war. War is, in invariably has followed a quantitative path. 13 reality, a subset of crisis; that is, all wars With respect to the dependent variable, result from crises, but not all crises lead to stability/instability, the absence or presence war. While some crises lead to war (e.g. of war has been the conventional measure- 1939), others do not (e.g. Berlin Wall 1961). ment from the outset (Singer & Small 1968; Moreover, some crises are accompanied by Singer, Bremer & Stuckey 1972; Wallace violence - minor or serious clashes, or full- 1973; Bueno de Mesquita 1975, 1978; Can- scale war - while others are not. Yet all nizzo 1978; Wayman 1984). War has been crises cause disruption within an inter- the litmus test of stability/instability both in national system: that is, they are sources of and outside of the COW research program. instability. It is precisely because crises cause Thus Haas (1970, p. 121) concluded about system disruptions that crisis is an excellent the polarity/stability link in 21 international indicator of instability.

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 39

An international crisis is a situational into World War II Crisis of 1939; (2) involve- change characterized by an increase in the ment by the Powers, ranging from none to intensity of disruptive interactions between direct military activity by all great powers two or more adversaries, with a high prob- (1929-39) or both superpowers (after 1945); ability of military hostilities. Higher-than- (3) geostrategic salience, that is, the location normal conflictual interactions destabilize of an international crisis in terms of its nat- the ongoing relationship of the adversaries ural resources, distance from major power and pose a challenge to the existing structure centers, etc.; (4) heterogeneity, measuring of an international system - global, domi- the number of attribute differences between nant or subsystem(s). A large-scale inquiry adversaries relating to military capability, into military-security crises, across conti- economic development, political regime, nents, cultures, and political and economic and culture; (5) the number and type of systems, uncovered 278 international crises issues in a crisis relating to four issue- from 1929 to 1979, that is, situations which areas - military-security, political- met the conditions specified above. The diplomatic, economic-developmental, and cases extend in time from the first Chaco cultural-status; and (6) the extent of War to the US Hostages in Iran. (They are violence. summarized and extensively analyzed in Overall Severity is calculated for each Brecher & Wilkenfeld 1987.) crisis as an index, ranging from 1-10 in value, The 278 cases from the International Crisis as follows: Behavior (ICB) data set, which can serve as the empirical domain for a new look at the link, include high profile polarity/stability where crises, such as Entry into World War II (1939), and little-known crises in Africa, Asia, Europe, Latin America and the Middle East, such as the Amur River incident (1938) and Tel Mutillah (1951). The data set also includes all of the major inter-state wars of the half-century, from Ethiopia in 1935- 36 to the October-Yom Kippur War of 1973. Quantitative data on each of the 278 cases For any given time period, the Overall have been collected for 50 variables which Severity of crises in progress can be used focus on seven crisis dimensions: Setting; to assess the aggregate level of instability. Breakpoint-Exitpoint; Crisis Management Although it is beyond the scope of this study Technique; Great Power/ to implement Overall Severity as a measure- Activity; International Organization In- ment of instability, one salient aspect should volvement ; Outcome, and Severity. 14 be mentioned. In prior research, each war The data, along with the conviction that involving a major power has been treated as multiple strands of evidence should be tap- an equal source of disruption. This is in- ped in testing competing hypotheses, have appropriate because all wars do not have stimulated the creation of a multivariate identical implications for international struc- measurement of instability with precise tures and processes. By contrast, Overall empirical referents. This index is derived Severity incorporates the varying levels of from the core concept of severity - or inten- intensity across international crises into the sity - of a crisis, with Overall Severity serv- measurement of instability. ing as the composite dependent variable, based upon six indicators: (1) number of 6. Conclusion crisis actors, ranging from one state (with its Several areas of difficulty have been iden- adversary, usually the state which triggers its tified in this review of polarity and stability. crisis) to ’N’ actors, the largest in the ICB The concept of polarity has more than one data set being 21 crisis actors in the Entry dimension to consider; changes in inter-

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 40

national structure may reflect either evolving Kaplan (1957), and Hoffmann (1968) also con- decisional autonomy or redistribution of tended that multipolarity was more stable. 5. Other mixed system structures include: Hanrieder’s power. Stability should be assessed more than in the (1965) hetero-symmetrical bipolarity, essentially a comprehensively past, with crisis synthesis of Kaplan’s (1957) loose bipolar and providing a more inclusive measurement Masters’ (1961) multi-bloc structures; and Hoff- than war of the level of instability. Fur- mann’s (1968) multi-hierarchical system, a very thermore, development of the concept of similar construct to Rosecrance’s bi-multipolarity, and a Overall of an international crisis combining bipolarity, multipolarity poly- Severity centric level. Young too, criticized the allows for (1968), ’weighting’ among disruptive bipolar/multipolar dichotomy as too narrow and events. Research that incorporates the proposed a ’discontinuities’ model which acknowl- renewed concepts of polarity and stability edged both dominant system and subsystem actors and their thus should contribute positively to the relationships. 6. In this context monopoly is not identical to control insoluble debate over structure seemingly over all capability. Rather, it refers to a pre- and conflict.16 ponderance of capability as compared to any other On a final note, it had been suggested at actor or combination of actors. 7. The of can the outset of this that a revival of the concept power is multi-dimensional: it inquiry be subdivided into economic and debate over and have military, political, stability polarity might cultural components, and others as well. for perhaps unanticipated implications other subjects The present focus is, of necessity, on military capa- of interest. In that regard, apart from pro- bility because the research question guiding this viding a new approach to the controversy inquiry is the relationship between structure and conflict. The latter is operationalized others as over structure and conflict, the preceding by war or violence; but in the forthcoming analysis it a new on the analysis suggests perspective will be represented by overt, disruptive interaction. current debate over hegemony. 17 In terms of It is military power which determines the direct the preceding analysis, ’hegemony’ is the capacity to generate such disruptive interaction members of an international conceptual counterpart of unipolarity, more among system. 8. For example, Russett & Starr (1985, p. 112) power and decisional precisely, unipolarity reported the following figures for military expen- as in multipolarity, specified Fig. 1 above. ditures in 1980-81 for the leading powers, in milli- The fundamental difference between the two ons of $U.S. : USA 144.0; USSR 130.0; China 28.0; debates is the focus on economic United Kingdom 26.8; W. Germany 26.7; France hegemony 10.0 versus concentration 26.5; Japan military (polarity). The of the US and the USSR it would be the of preeminent position Although beyond scope as military powers is quite evident from these this inquiry to explore further the issue of rankings. hegemony, it provides at least one example 9. This treatment of stability is not intended as an of the ongoing relevance of the unresolved endorsement of the status quo in world politics; it is that in some areas is more over structure and conflict. acknowledged change controversy desirable than continuity. However, in a system with multiple nuclear powers, extreme forms of instability — possibly leading to global dis-

NOTES - equilibrium are regarded as costly to the inter- 1. This argument is applied more generally by Bueno national community as a whole. de Mesquita (1985) to scholarship on international 10. Some of the analysis of this diagram is based on conflict. Buchanan & Tullock (1962, pp. 63-91). 2. The analysis to follow will focus on the presumed 11. The discussion will be restricted to the dominant impact of systemic polarity on international system in order to facilitate a linkage with the stability. It is recognized, of course, that cause and previous debate. effect also could operate in the opposite direction. 12. In a strictly historical context, multipolarity would For example, the extraordinary disruption repre- enjoy an advantage over polycentrism with regard sented by World War II had a devastating impact on to expected costs from decision making. The larger polarity, with Japan and Germany being removed number of autonomous decisional centers in poly- from the inner circle of great powers. It would be centrism increases the theoretically-possible num- beyond the scope of this investigation to deal with ber of adversarial actor pairs and coalitions which the effects of stability levels on polarity in further tend to generate more disruption (James & Wil- detail. kenfeld 1984). However, the number of decisional 3. The greater stability of a bipolar structure also was centers under polycentrism still falls far short of argued by Riker (1962). 100% of the system, so externalities would not be 4. Wright (1942), Morgenthau (1948), Gulick (1955), that much less than in a multipolar world. Since

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 41

multipolarity and polycentrism may have various coming. Crises in the Twentieth Century Vol I. numbers of decision centers in the abstract, the Handbook of International Crises. Oxford: Per- lower decisional costs of the former are relevant gamon Press. only in historical terms. Buchanan, James & Gordon Tullock 1962. The Calculus 13. There are four possibilities to consider if the time of Consent. Ann Arbor: University of Michigan periods do not show differences in instability: (1) Press. the presumed connection simply does not exist; Bueno de Mesquita, Bruce 1985. ’Toward a Scientific (2) the measurement of polarity is invalid; (3) the Understanding of International Conflict: A Personal measurement of instability is invalid; and (4) both View’, International Studies Quarterly, vol. 29, no. measurements are invalid. 2, pp. 121-136, 151-154. 14. The data set also includes six control variables: Bueno de Mesquita, Bruce 1978. ’Systemic Polarization Geography; Polarity; System Level; Conflict; and the Occurrence and Duration of War’, Journal Power Discrepancy, Involvement by the Powers of Conflict Resolution, vol. 22, no. 2, pp. 241-267. (Brecher & Wilkenfeld 1987). Bueno de Mesquita, Bruce 1975. ’Measuring Systemic 15. The Index of Overall Severity and the underlying Polarity’, Journal of Conflict Resolution, vol. 19, no. derive from a concept of severity long-standing 1, pp. 187-216. interest in tapping the intensity of an international Cannizzo, Cynthia 1978. ’Capability Distribution and crisis and thereby in differentiating any crisis from Major-Power War Experience, 1816-1965’, Orbis, all others. More specifically, the index and concept vol. 21, pp. 947-957. constitute a response to a cluster of interrelated Deutsch, Karl W. & J. David Singer 1964. ’Multipolar questions: What are the (varying) effects of a large Power Systems and International Stability’, World (or small) number of crisis actors on the extent Politics, vol. 16, no. 3, pp. 390-406. of disruptive interaction from the outbreak of an Gulick, Edward V. 1955. Europe’s Classical Balance of international crisis to its termination? What is the Power. New York: W. W. Norton. extent of heterogeneity among the adversaries; that Haas, Michael 1970. ’International Subsystems: Stab- is, do they vary in military capability or are they all ility and Polarity’, American Political Science major (or minor) powers? Are they states with Review, vol. 64, no. 1, pp. 98-123. advanced or premodern economies or do they Hanrieder, Wolfram F. 1965. ’The International exhibit various levels of development? What is the System: Bipolar or Multibloc’, Journal of Conflict extent of superpower involvement? Is it primarily Resolution, vol. 9, no. 3, pp. 299-308. political or economic or military support for one Hart, Jeffrey A. 1985. ’Power and Polarity in the or more of the adversaries, or did the US/USSR International System’, pp. 25-40 in Alan Ned engage in direct military intervention? Is the Sabrosky, ed., Polarity and War. Boulder, Colo.: location of the crisis high or low in geostrategic Westview. salience? What are the issues at stake; are they Hoffmann, Stanley 1968. Gulliver’s Troubles, or the military, political, economic or cultural, or several Setting of American Foreign Policy. New York: of these combined? Is there interstate violence in McGraw-Hill. the crisis; if so, how extensive is it? James, Patrick and Jonathan Wilkenfeld 1984. ’Struc- These questions have played a central role in tural Factors and International Crisis Behavior’, the deductive formulation of the Index of Overall Conflict Management and Peace Science, vol. 7, no. The derivation of the the Severity. latter, including 2, pp. 33-53. allotted to its six indicators, is in weights explained Kaplan, Morton 1957. System and Process in Interna- detail by Brecher & James (1986, part 2). tional Politics. New York: John Wiley. 16. The results of some preliminary testing appear in Keohane, Robert O. 1984. After Hegemony Coop- Michael Brecher and Patrick James, Stab- ’Polarity, eration and Discord in the World Political Economy. ility, Crisis: A New Look at an Unresolved Contro- Princeton: Princeton University Press. in Suffice it to note here versy’, preparation. that, Keohane, Robert O. 1986. in International for 7 of 10 bivariate crisis ’Reciprocity indicators, polycentrism Relations’, International Organization, vol. 40, no. is the most unstable the three clearly among types 1, pp. 1-27. of structure (e.g., number of crises per year, the Keohane, Robert O. & Joseph S Nye 1977. Power and proportion of violent triggers, violence as the pri- Interdependence. Boston: Little, Brown. mary crisis management technique). Levy, Jack S. 1985. ’The Polarity of the System and 17. Some of the contributions include Keo- principal International Stability: An Empirical pp. hane Russett and Snidal Analysis’, (1984, 1986), (1985) 41-66 in Alan Ned Sabrosky, ed., Polarity and War. (1985a, 1985b). Boulder, Colo.: Westview. Masters, Roger D. 1961. ’A Multi-Bloc Model of the International System’, American Political Science REFERENCES Review, vol. 55, pp. 780-798. Morgenthau, Hans J. 1948. Politics Among Nations. Brecher, Michael & Patrick James 1986. Crisis and New York: Knopf, 4th ed., 1967. Change in World Politics. Boulder, Colo.: Morrow, James D. 1985. ’A Continuous Outcome Westview. Expected Utility Theory of War’, Journal of Conflict Michael & Jonathan Wilkenfeld forth- Brecher, 1987, Resolution, vol. 29, pp. 473-502.

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012 42

Nogee, Joseph L. 1974. ’Polarity: An Ambiguous Con- Singer, J. David & Melvin Small 1968. ’Alliance Aggre- cept’, Orbis, vol. 18, no. 4, pp. 1193-1224. gation and the Onset of War’, pp. 247-286 in J. Rapkin, David, William Thompson with Jon David Singer, ed., Quantitative International Christopherson 1979. ’Bipolarity and Bipolarization Politics. New York: Free Press. in the Cold War Era’, Journal of Conflict Resolution, Snidal, Duncan 1985a. ’The Limits of Hegemonic Stab- vol. 23, no. 2, pp. 261-295. ility Theory’, International Organization, vol. 39, Riker, William H. 1962. The Theory of Political no. 4, pp. 579-614. Coalitions. New Haven: Yale University Press. Snidal, Duncan. 1985b. ’Coordination versus Rosecrance, Richard N. 1963. Action and Reaction in Prisoners’ Dilemma: Implications for International World Politics. Boston: Little-Brown. Cooperation and Regimes’, American Political Rosecrance, Richard N. 1966. ’Bipolarity, Multipo- Science Review, vol. 79, no. 4, pp. 923-942. larity, and the Future’, Journal of Conflict Resolu- Thucydides 1930. A History of the Peloponnesian War. tion, vol. 10, no. 3, pp. 314-327. trans. by Benjamin Jowett. Oxford: Ashenden Russett, Bruce M. 1965. Trends in World Politics. New Press. York: Macmillan. Wallace, Michael 1973. ’Alliance Polarization, Cross- Russett, Bruce M. 1985. ’The Mysterious Case of Cutting, and International War, 1815-1964’, Journal Vanishing Hegemony; or, Is Mark Twain Really of Conflict Resolution, vol. 17, no. 4, pp. 575-604. Dead?’, International Organization, vol. 39, no. 2, Waltz, Kenneth N. 1964. ’The Stability of a Bipolar pp. 207-231. World’, Daedalus, vol. 43, pp. 881-909. Russett, Bruce M. & Harvey Starr 1985. World Politics: Wayman, Frank W. 1984. ’Bipolarity and War: The The Menu for Choice. 2nd ed., New York: W. H. Role of Capability Concentration and Alliance Pat- Freeman and Company. terns Among Major Powers, 1816-1965, Journal of Sabrosky, Alan Ned, ed. 1985. Polarity and War. Boul- Peace Research, vol. 21, no. 1, pp. 61-78. der, Colo.: Westview Press. Wright, Quincy 1942. A Study of War. 2 vols., Chicago: Singer, J. David; Stuart Bremer & John Stuckey 1972. University of Chicago Press. ’Capability Distribution, Uncertainty and Major Young, Oran R. 1968. ’Political Discontinuities in the Power War, 1820-1965’, pp. 19-48 in Bruce M. International System, World Politics, vol. 20, no. 3, Russett, ed., Peace, War, and Numbers. Beverly pp. 369-392. Hills: Sage.

Downloaded from jpr.sagepub.com at UNIV OF SOUTHERN CALIFORNIA on October 8, 2012