INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1.The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame. .

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 76-18,058 WESBR00IC, Stephen Dahl, 1948- INTERNATIONAL POWER DISTRIBUTION AND WAR: THE EFFECTS OF CONCENTRATION AND DISPERSION OF POWER ON INTERSTATE WAR. The Ohio State University, Ph.D., 1976 Political Science, international law and relations

Xerox University Microfilmsi Ann Arbor, Michigan 48106

© 1976

STEPHEN DAHL WESBROOK

ALL RIGHTS RESERVED INTERNATIONAL POWER DISTRIBUTION AND WAR*

THE EFFECTS OF CONCENTRATION AND DISPERSION OF POWER

ON INTERSTATE WAR

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School oi The Ohio State University

By

Stephen D. Wesbrook, B.S., M.A.

*****

The Ohio State University

1976

Reading Committee*

Dr. Charles F. Hermann

Dr. Chadwick F. Alger

Dr. James E. Harf Adviser Department of Political Science ACKNOWLEDGMENTS

For their guidance and direction throughout my graduate study, including the preparation of this dissertation, I am indebted to

Professor Chadwick F. Alger, Professor James E, Harf, Professor Charles

F. Hermann, and Professor Allan R. Millett. While I was studying at the Columbus campus of The Ohio State University (September, 19?2 through August, 197*01 they provided me with a personalised program in the area of national security affairs. After I left the campus, they adjusted departmental procedures and their own schedules to allow me to prepare this work and still meet the very considerable demands placed on an active duty officer in the Armed Forces.

The vast majority of this work was written while I was attending the Infantry Officers Advanced Course at Fort Benning, Georgia, in the year following my departure from the Columbus campus. Because I was no longer on a university campus with ready access to libraries, computers, and advice, it would have been impossible to complete this research in the time that it was completed without the help of a friend and classmate at Ohio State— -Lavoy Ledford, III. He provided liaison between myself, my research committee, and the graduate school. Also, he ran the Pearson correlation and GINI index computational programs for me on the com­ puters available at Ohio State. I am also indebted to Dr, Wayne H.

Ferris, Richard Katon Associates, Rockville, Maryland, for supplying me with certain pieces of unpublished data which were crucial to this study.

This saved months of research and computation. ii The final preparation of the manuscript was aided by the efforts

of William J. Gregor and my father, both of whom helped edit the final draft, and by Mrs, Sarabeth Rathbun, who typed the manuscript.

Although this work was not financially supported by or in any way connected with the Department of the Army, this research would not have been possible without its farsighted policy of allowing a few

officers the opportunity to begin their graduate education very early

in their careers.

Finally, this study would have been impossible without the understanding of my wife Sharon, who gave up a great deal to see it completed.

iii VITA

March 2, 19*f6.... Born—-Naha, Okinawa

1970...... B.S., United States Military Academy, West Point, New York

1971-1972...... Infantry Platoon Leader, 101st Airborne Division, Republic of Vietnam

197^...... M.A., The Ohio State University, Columbus, Ohio

1975* ...... Infantry Officers Advanced Course, Fort Benning, Georgia

FIELDS OF STUDY

Major Field: International Relations. Professor Charles F. Hermann

Minor Field: American Military History. Professor Allan R. Millett TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS...... ii

VITA...... iv

LIST OF TABLES...... vi

LIST OF FIGURES...... vii

Chapter

I. INTRODUCTION...... 1

II. THEORIES AND PROPOSITIONS IN INTERNATIONAL RELATIONS...... 7

International Systems Theory...... 11 Historical Interpretation...... 27 Policy Prescription and Formulation...... 35

III. STRATEGY AND CONCEPTS...... 44

Strategy of Inquiry...... 44 Concepts and Operationalization ...... 52

IV. THEORETICAL ANALYSIS...... 87

The Structural Inducement of Aggression 88 The Structural Modification of Aggression... 94 Conclusions and Hypotheses...... 120

V. QUANTITATIVE DESCRIPTION AND ORIGINAL EVIDENCE, 1862-1964...... 126

International Power Distribution, 1862-1964...... 126 Interstate War Data -...»...... 149 Analysis and Evidence ...... 157

VI. CONCLUSION...... 171

BIBLIOGRAPHY...... 176

v LIST 01'' TABLES

5-1. Power Structure of the Total System, 1864 through I960, in Terms of Concentration-Dispersion Indices.. 130

5-2. Pearson Correlation of Concentration-Dispersion Indices within the Total System...... 131

5-3 . Power Structure of the Elite System, 1864 through i9 6 0, in Terms of Concentration-Dispersion Indices...... l^K)

5-4. Pearson Correlation of Concentration-Dispersion Indices within the Elite System...... l4l

5-5. Pearson Correlation of Concentration-Dispersion Indices between the Elite and the Total Systems...... l4l

5-6. Power of the Top Five States as a Proportion of Total System Power...... 144

5-7. Rank-Order List of States Comprising an Elite System of Five States...... 147

5-8. Growth of the Total System in Terms of System Membership, System Power, and System Population...... 152

5-9* Interstate War Data, 1862-1909...... 155

5-10. Interstate War Data, 1910-1964...... 156

5-11. Pearson Correlation of Concentration and the Amount of Vlar in the System, Midpoint Technique...... l6l

5-12. Pearson Correlation of Concentration and the Frequency of War in the System, Midpoint Technique...... 161

5-13. Pearson Correlation of Concentration and the Amount of War in the System, Initial Point Technique...... 3.62

5-l4. Pearson Correlation of Concentration and the Frequency of War in the System, Initial Point Technique..... 162

5-15. Comparison of Concentration and Dispersion in the Amount of War with Concentration and Dispersion in Power Distribution ...... 168

vi LIST OF FIGURES

2-1. International Systems Classified by Power Structure.... 12

3-1. Concepts of Magnitude, Amount, and Frequency of War Graphically Displayed...... 80

5-1. Total System Stratification--The Power Pyramid, 1864... 134

5-2. Total System Stratification Based on the Most Powerful State, 1864...... 134

5-3* Total System Stratification~The Power Pyramid, 1884... 135

5-4. Total System Stratification Based on the Most Powerful State, 1884...... 135

5-5. Total System Stratification— The Power Pyramid, 1909... 136

5-6. Total System Stratification Based on the Most Powerful State, 1909...... 136

5-7. Total System Stratification— The Power Pyramid, 193^*•* 137

5-8. Total System Stratification Based on the Most Powerful State, 193^...... 137

5-9. Total System Stratification— The Power Pyramid, i9 6 0... 138

5-10. Total System Stratification Based on the Most Powerful State, i9 6 0...... 138

5-11. Lorenze Curve Depicting the Power Distribution of the Elite System for 1899 and 1946...... 143

vii CHAPTER I

INTRODUCTION

The central postulate of Harold and Margaret Sprout*s ecological paradigm for the study of international affairs is that*

...the uneven distribution and irregular arrangement of things upon or in association with the earth*s surface significantly affect most if not all important human undertakings, inter­ actions, interrelationships, and other states of affairs.

It does not seem to be an overstatement to suggest that the problem of determining the effects of the "uneven distribution and irregular arrangement of things" upon these "important human undertakings, interactions, and interrelationships" has been one of the central and most perplexing problems of social and political science and philosophy.

Concern with economic, social, and political distributions has occupied scholars since ancient times. The early studies were generally philosophical and dealt with such subjects as the natural­ ness, permanence, and inevitability of various distributions of a variety of things. More recently, economic, social, and political scientists have directed their efforts toward those subjects sus­ ceptible to their methods. Economists have studied the even and uneven distribution of wealth, income, productive capacity, and goods.

Sociologists have been concerned with the distribution of opportunity, living conditions, and class, among others. Political scientists have been primarily concerned with the distribution of power. Although

the concepts and variables may differ slightly, the basic questions

have generally been the same regardless of the field* to what degree

is a particular valued entity concentrated under the control of a few

actors or dispersed under the control of many; and, what are the

effects of this state of affairs?

In international relations a large portion of the philosophical

and scientific attention paid to these questions has involved power as

the "valued entity" and war as the "effects." If the various theories

dealing with this area were simplified and dichotomized, one pole would

present the belief that there is less war when power is dispersed among

many actors in the international system, whereas the other pole would

present the belief that there is less war when power is concentrated

in the hands of a very few actors. This difference of thought is the

general topic of this study.

If this particular controversy does not pre-date most political debates, it certainly has its origins in the very beginnings of modern

political thought. Perhaps the earliest and clearest example in the area of conflict control of contrasting theory on the desirability of power concentration or dispersion is found in the writings of Hobbes and Rousseau. For Hobbes the ideal system for conflict control was one wherein power was totally concentrated in the hands of a single actor. For Rousseau the ideal system for conflict control was one of many small actors. In the minds of both, insecurity led to con­ flict. But where Hobbes saw insecurity as being caused by equality,

Rousseau saw it resulting from inequality. Both believed human greed to be at the roots of war. Yet Hobbes hoped to control greed by a concentration of resources, whereas Rousseau hoped to control it by the

dispersion of resources. Both philosophers recognized the need to con­

trol the damage done by members of the system in intra-systemic conflict.

In Hobbes* system, strife was to be controlled by the threat of a

massive application of force. The certain knowledge of destruction for

the disturber theoretically deterred war. In Rousseau’s ideal system,

strife was to be controlled by numerous small applications of force —

by the invisible hand of pluralism. The weakness of the members of the

system in relation to all others minimized both the desire for strife

and the damage resulting when it occurred.^

As one scans the modern literature which relates systemic power

distribution to violent conflict, one cannot but be struck by two ob­

servations. First of all, in some ways we have not come very far in

terms of theory in this area since Hobbes and Rousseau. The rationale

for preferring one power structure to another has been elaborated upon,

and the circumstances and language of the formulations have changed,

but the basic reasoning has not changed. The second and perhaps more

surprising observation is that very little evidence, historical or

otherwise, has been brought to bear to support either theoretical

position. The purpose of this study is to make a modest contribution

in each of the above two areas.

The first objective is to determine theoretically if one could

reasonably expect the distribution of power in the international system,

in terms of power concentration and dispersion, to affect war in the

system* and if so, in what manner. The second objective is to gather

some original evidence, based on the historical period from 1862 to

1 9 which would tend to either support or refute the theoretical analysis. k

The remainder of this study is devoted to achieving these ob­

jectives. Chapters II and III establish the groundwork from which we

will assault these objectives. Chapter II (Theories and Propositions

in International Relations) presents a large sample of the theories

and propositions relating power distribution to war which have sur­

faced in international relations theory and practice. This chapter

thus serves to familiarize the reader with both the scope of the

problem and the state of disorder that exists in this area. Chapter

III (Strategy and Concepts) contains a discussion of the strategy which

has been used to attack the problem, as well as a detailed analysis of

the independent and dependent variables. Chapter IV (Theoretical

Analysis) Is devoted to the specific attainment of the first objective,

as defined above. We first consider the mechanism whereby power

structure might affect war, and then we analyze the effects which should

result if that mechanism is in fact operating. At the end of the

theoretical analysis, conclusions are presented in the form of explicit

propositions. Chapter V (Quantitative Description and Original Evidence,

1 8 62-I96*f) is devoted to the specific attainment of the second objective, as defined above. We establish quantitatively the distribution of power

in the interstate system from 1862 to 196** and then compare, through a variety of techniques, the fluctuation of power distribution with the fluctuation of war. The historical relationship thus established is then analyzed in light of the propositions with which we concluded the theoretical analysis. Finally, in Chapter VI (Conclusion) we consider the significance of the original evidence developed in the preceding chapter and then end the study with a few comments on the broader im­ plications of its theory and findings. It is a rare occurrence to find a definitive answer to any great political question in a single study. The reader should be aware that this is not such a study. Therefore, he should not ex­ pect the debate between those who advocate either concentration or dispersion of power for the purpose of establishing peace to be ended by this work. The reader should expect to find in this study, however, not only an analysis which combines the seemingly divergent views of many great political analysts into a complementary, unified theory, but also some original evidence which supports that theory. NOTES TO CHAPTER ONE

^Harold and Margaret Sprout, An Ecological Paradigm for the Study of International Affairs; Research Monograph No. 30 (Princeton: Center for International Studies, 19&&), p. 21.

^Melvin Tumin, Social Stratification: The Forms and Functions of Inequality (New York: Prentice-Hall, 1967)* P* 1.

^Stanley Hoffmann, "Rousseau on War and Peace," in his The State of War (New York: Frederick A. Praeger, 196.5)» PP* 56-80; Thomas Hohbes, Leviathan, ed. by Michael Oakeshott (New York: Collier Books, 1962).

6 CHAPTER II

THEORIES AND PROPOSITIONS IN

INTERNATIONAL RELATIONS

While we feel that a valid determination of the relationship

between systemic power distribution and war would have a significant

impact on the study and practice of international relations, we do not expect everyone to initially share this view. Rather we expect that one's opinions as to the significance of this topic will correspond with one's views as to which mechanisms guide world politics. For example, many who deal with world politics on a day-to-day basis might consider the structure of the international system to be a poor topic of study if one wishes to explain interstate behavior. Consequently, in this chapter we hope to establish the significance of our topic by demonstrating that regardless of where one turns for guidance on the operations of world politics, one is likely to find theories and propositions relating the power structure of the international system to that most serious problem of world politics -- war. We will pre­ sent a large sample of theories and propositions from three general topical areas* international systems theory, historical interpretation, and policy prescription and formulation.1 Thus this chapter should serve not only to demonstrate the importance of this topic to those dealing in international relations theory and practice, regardless of a scholar's or a practitioner's particular interests, but also to familiarize the 7 reader with the hulk of the work which has already been done in the

area of our concern.

Before we can present these theories and propositions, however,

a number of concepts that are crucial to this study must be defined.

We do not wish to bog the reader down early in this work with a detailed

analysis of all the key concepts of this study. Nevertheless, a few

definitions are necessary here in order that the reader might be able

to relate and compare the theories and propositions contained in this

chapter with each other and also with the original theory, propositions,

and evidence which is developed later in the study.

The international system is "the ensemble constituted by

political units that maintain regular relations with each other and that p are all capable of being implicated in a generalized war." Thus, the

international system is envisioned as being made up of only that subset

of actors in the broader world system which maintain regular inter­

actions and which are capable of waging war.

The international system may be viewed in terms of its structure

and/or its behavior. The structure of the international system is a

characteristic configuration of some relatively fixed attribute. The

behavior of the international system is a characteristic set of

immediate actions. To focus on the structure of the system is to focus

on relatively static properties; to focus on the behavior of the system

is to focus on dynamic properties.3 of course, few attributes, if any,

are permanently fixed. Moreover, most often it is human action which

causes changes in the arrangement of these attributes. Consequently,

one cannot definitively divide all properties into the categories of structure and behavior. Whether something can be described as structural or behavioral depends on the time perspective one is using

and the specific problem at hand. What is behavioral in one circum­

stance might be seen as structural in another. The primary difference

between structure and behavior is thus one of perspective and emphasis.

One who studies structure focuses on background conditions; one who

studies behavior focuses on the immediate activities of actors.^

If one chooses to describe the international system in terms of structure, then one might choose any combination of a number of attributes as the defining characteristic. These might include diplomatic, legal, economic, and moral attributes, to mention but a few. The choice is usually dictated by one's concern as a scholar. In this study the structure of the international system is defined in terms of power. The justification for viewing power as a structural dimension in this study is that when an actor considers the decision to go to war, the con­ figuration of power in the international system must generally be seen as fixed and relatively unchangeable. Over a long period of time the configuration of power can be changed by human actions. However, in relation to the immediate decision to go to war, the configuration of power is a background condition over which the decision-maker has little control.

As will be discussed in detail in Chapter III, it is extremely difficult to define power in a way which is acceptable to everyone.

The definition that will be used here is that power is the general capacity of an actor to control the behavior of others. It is difficult to say exactly what determines this capacity. Many items have been suggested at one time or another, to include, for example, diplomatic skill, national morale, economic resources, and military assets. 10

Whether power is measured in terms of the actions of actors or the po­

tential which specific resources provide is not important here. What

is important is that power requires a base or source, and that source

is relatively fixed in relation to the decisions of war and peace.

The distribution of power among the actors in the international

system may be characterized in a variety of ways. In this study we are

dealing with distribution in terms of concentration and dispersion. The

concentration of power is a condition such that a high proportion of the

total power of the system Is controlled by a few of the actors in the

system. The converse is dispersion; that is, when power is relatively

equally distributed among many of the actors in the system. "Few" or

"many" may be conceived in terms of percentages or in terms of integers.

Whereas it might seem best to choose one or the other conceptions, this

cannot be done in this study for primarily two reasons. First of all, rarely is the distinction clearly made in the bulk of the theory which will be discussed. To try to sort this literature on the basis of these two ways of viewing concentration and dispersion may be a worth­ while task, but it is one beyond the scope of this study. Secondly, even if these two ways of viewing concentration and dispersion are not integral to the concepts themselves, they are integral to the statistics used to measure concentration and dispersion. All of the statistics commonly used to measure these concepts are influenced by changes in both the percentage distribution and in the number of actors. We will deal with this problem in more detail in Chapter III.

Having defined the above terms, we are now hopefully prepared to turn to those areas where propositions and theories which relate the distribution of power to war are to be found. 11

International Systems Theory

A variety of international systems have been presented at one

time or another as being able to reduce or eliminate war. What is

significant about this fact for us is, as will be shown below, that

the principal distinguishing factor between these systems is their

power structure. Moreover, in each case the rationale whereby the

system is supposed to produce the effect of reducing war is directly

related to the power structure of the particular system. Thus, if we

order these systems according to their power structure, from most

concentrated to most dispersed, we are not only able to relate a

variety of systems to one another but are also able to relate a variety

of reasons whereby each system is supposed to reduce or eliminate war.

Figure 2-1 below presents a sampling of international systems

which have at one time or another been presented as having a significant

capability of reducing or eliminating war. The fourteen specific types

of systems have been grouped into four clusters based on a similarity

of power structure. Within these clusters the specific types of systems

are ranked from most concentrated to most dispersed. The four clusters,

or general categories, are then ranked according to the distribution of

power within the systems. Thus in terms of power structure, the general

categories are ranked in the order of most concentrated to most dis­

persed in the following manners dominance, multibloc, balance of power,

and extreme dispersion. Dominance systems have power substantially

concentrated in the hands of a single actor. Multibloc systems en­

vision power being almost totally controlled by a few actors. Balance

of power systems are composed of many actors, none of whom has enough power to subdue the rest. Extreme dispersion Bystems are composed of an extremely large number of actors, none of whom have a significant

percentage of the world's total power.

Power General Specific Distribution Category Type A. Empire M World Government Dominance Universal Organization Collective Security

Regionalism Multibloc Regional Security Multibloc Polarity

Unipolar Balance Bipolar a of Concert n Power Multipolar

Numerous Equal States Extreme Functionalism Dispersion Anarchism Figure 2-1* International Systems Classified by Power Structure

Not only does this classification scheme group systems accord­ ing to power structure, but it also groups them according to the rationale whereby each has been presented as producing less war. As the power structure changes, so does the rationale. All of the dominance systems rely on a preponderance of power to enforce peace* The multi­ bloc systems are presented as depending on the confrontation of alert,pcw- erful actors to reduce war. Balance of power systems rely on com­ plexity and pluralism to keep the peace. Extreme dispersion systems are presented as producing peace by inducing a des.ir., for cooperation and participation and thus eliminating the motivation for war.

We will discuss in more detail below each of these systems and the reasons why each can be seen as being capable of reducing war. One should note, however, that the specific types of systems are clas­

sified based on a modal power configuration for that type of system.

Thus it is entirely possible that a specific case of a bipolar balance

of power system might be more concentrated than a specific case of a

multibloc polarity system. But here we cannot deal with the entire

theoretical population of each type of system, and in general the multi­

bloc systems are more concentrated than the balance of power systems.

We recognize that our general and specific categories are not totally

mutually exclusive, nor can they always be rank ordered on the con­

tinuum of power concentration and dispersion without some overlap

between extreme specific cases. However, as hopefully will be demon­

strated in the remainder of this section, this scheme allows us to make

some sense out of a large variety of systems, predictions, and rationale.

Dominance Systems

A dominance system is one in which a single actor possesses

power vastly superior to the combined power of the rest of the system.

Although there are no historical examples of systems of world dominance,

certain regional examples, such as the Roman Empire, might be con­

sidered analogous. The dominance system within most nation states also

tends to serve as a substitute for the lack of historical experience

with international dominance systems. Whether based on regional his­

torical memories or theoretical promise, the attractiveness of

dominance systems is a very real force in world politics. This has been

manifested in modem history in the rationale of the Bourbon and

Napoleonic attempts to unify Europe, in the Nazi version of "one world," and in the Marxist image of the political millenium.^ 3A

In the rationale for the advocacy of such systems, anarchy is

a symbol of peril — the peril of uncontrollable disorder.^ The cen­

tralization or concentration of power in one actor is the basis for

the solution to this peril. At the international level, the typical

major themes are that the world is in a state of anarchy, which makes

war inescapable; the elimination of war is a dire necessity; this goal

cannot be reliably achieved by means other than some form of concen­

tration of power in the hands of a responsible actor whose authority

is absolute.? These same themes appear whether it is Dante justifying

empire, Saul Mendlovitz advocating world government, Quincy Wright

justifying a universal organization, or Woodrow Wilson advocating

collective security. The single characteristic which is essential for

maintaining peace in each of these cases is a preponderance of power in

the hands of a single actor. Within the general category of dominance

systems, it is the internal structure of the powerful, authoritative

actor and not the basic mechanism of domination that distinguishes

specific types of dominance solutions. This can be seen as we discuss

each type of dominance system — empire, world government, universal

organization, and collective security.

Periods of empire, although on a regional rather than a world

scale, have been dominant in history. The Roman, Byzantine, Ottoman,

Chinese, and Holy Roman empires are perhaps the most famous examples.

There is a tendency to believe that political pluralism based on a

state system has always been the dominant motif of the international

system. However, this has been a fairly recent development.

One has to go back to earlier writings to find a justification for world empire. It is not, however, a lack of faith in concentrated power "being able to maintain peace that has put it out of vogue, but

rather the desire of self-determination. One of the best known

arguments for world empire can be found in Dante's Monarchia. where

he sees peace as following from world order:

Whence it is manifest that universal peace is the best of all those things which are ordained for our blessedness... If we consider the family, the goal of which is to prepare its members to live well, there must needs be one to guide and rule...If we consider a district, the end of which is helpful co-operation..., one must needs be the guide of the rest, whether he be imposed upon them by another or rise to eminence out of themselves, with the consent of the rest. Else not only do they fail to attain the mutual support they aim at, but sometimes when several strive for pre-eminence, the whole district is brought to ruin...If finally we consider a special kingdom,.there must be one king to rule and govern, also not only do they in the kingdom fail to reach the goal, but the kingdom itself lapses into ruin, accordingly to that saying of the infallible truth, "every kingdom divided against itself shall be laid waste..Therefore there must be one monarch or emperor. Thus it appears that for the well-being of the world there must be a monarchy or empire.0

The principal difference between empire and the three other

specific types of dominance systems lies in the idea that in the

latter three, subordinate elements have some say in the decision

process. However, the governing body still maintains a preponder­

ance of power capable of forcing its decisions on the rest of the

system.

Schemes of world government are also not new to international

politics. Sully's "Grand Design" and Saint-Pierre's scheme for

perpetual peace represent two early m o d e ls .^ More recently Saul

Mendlovitz, believing that it is not a question of whether there will

be world government but of its form, has outlined a number of models.'

But again, whether it is through enforced arbitration or a universal police force, universal government, as with state government, depends

on the effective monoploy of legitimate violence.^ As Theodore Able

has suggested, "no efforts to prevent war can be successful unless they accomplish two things * first, create a situation wherein cal­ culations of chances and consequences will produce negative results; secondly develop means of checking the activities of power groups,

In the case of dominance systems, whatever their internal form, this is to be accomplished by concentration of power in the hands of a central decision-making body. To present it as anything else is a matter of either idealism or deception. One might prefer one form to smother on grounds that it might be "enlightened," However, one should also consider the speed with which an "enlightened" dominance system could turn into a tyrannical one.

The major historical example of a universal organization dominance system is found in Medieval Europe. The state system was not yet fully developed. The Catholic Church was able to dominate within Christendom, in part because of its power to legitimize or condemn specific acts of violence, ^ It has been proposed that the

United Nations could serve as a basis for a universal organization

•1 ti model. This model of supranational authority also rests on some means of sanctions. For instance, even in Quincy Wright's advocacy of a universal organization system, which he sees as capable of solving "the central problem of peace," we find recognition of the necessity of giving the organization a capability to impose "sanctions against aggression" and to "supervise regional and functional organizations and to maintain peace among them."^5 This requires an amount of power superior to that of the member states. Within 17

this framework one often mentioned solution is the creation of an

international police force. Another is exclusive control of nuclear

weapons. But in any case, peace is still a matter of enforcement "by

superior force.

The concept of collective security has accumulated a number

of meanings. Yet in its original version, which may be referred to as the Wilsonian doctrine, the concept involves the creation of an international system in which the danger of aggressive war is to be met by the avowed determination of a security organization to frustrate the aggressor state. Collective security postulates that a prepon­ derance of power would be available for defensive purposes but not for 17 aggressive purposes. Although a collective security system may have numerous actors in relation to most decisions, when the decision to go to war is made, the systeiji (theoretically) has only two actors — the aggressor state and the collective security organization. The strategy of collective security thus explicitly involves a concentration of power.

An early commentator on collective security. Sir Frederick Maurice, makes this very clear:

The efforts of European statesmen (formerly) were directed to securing peace by organizing a balance of military power sufficiently exact to make the risk of attack by the group in one scale upon the group in another prohibitive. This procedure failed...The Covenant of the League of Nations endeavors to find a better guarantee against war by substituting for the small group of Powers a single group so large that its authority cannot be challenged with impunity. The Balance of Power is replaced by the...Concentration of Power. 18

Multibloc Systems

A multibloc system is one which contains only a small number

of actors (usually four or five), each of which is extremely powerful.

There are a variety of schemes which present a multibloc international system as the solution to interstate war. Three such solutions are regionalism, regional security arrangements, and multibloc polarity.

The principal difference between these is the degree of formal co­ hesiveness of the basic actors.

Regionalism implies that regional supranational units will handle the issues of political and military security."^ These units are likely to be formed by the federation of states in a particular geographic region. For instance, the Common Market countries would become a United Europe, the countries of the Organization of African

Unity would form an African unit. Peace will be maintained by both continuous conferences, presumably between powerful equals, and the 20 threat of mutual annihilation. The power of these supranational units would be such that a violent collision would be catastrophic and therefore unlikely.

A regional security system represents a variation of the same theme. Only here, organizations, such as NATO, would serve as the

O'] principal actors. This is a system, along with its deterrence con­ cepts, with which we are currently partially familiar. Again the system relies on the threat of massive destruction, that can be accomplished through the concentration of power, to deter or prevent war.

Another variation is the multibloc polarity system. In this model the regional groups are dominated by a major powerj i.e., an 19

Asian region dominated "by China or a Western Hemisphere dominated "by

the United States. 22

In each of these cases we have a model with a few extremely

powerful actors. Actors confront each other, yet theoretically they

are prevented from going to wax "by the knowledge that there are no

peripheries or weak nations to exploit. There are only powerful blocs

butting against each other. In such a system, the most likely reaction

to aggression is retaliation by at least one extremely powerful neighbor.

Uncertainty is eliminated. The essence of this solution is found in

concentration. It is not concentration in terms of one actor con­ trolling a high percentage of power but rather a concentration produced by reducing many actors to four or five actors.

As with dominance systems, the essential characteristic of power concentration in the multibloc system can be played down. The system can be dressed up with visions of enlightened management of world affairs.

For a less favorable view of the possibilities of such a system than is found in many professional journals, however, one might glance at

George Orwell's Nineteen Eighty-four.23

Balance of Power Systems

A balance of power system is a system of many actors in which power is distributed in such a way that no single actor possesses enough power to dominate all the rest. In the international sphere, this is the system that Gentz had in mind when he warned that "no (jstatej.. .must ever become so powerful as to be able to coerce all the rest put together."

Thus, a balance of power system places no reliance on the controlling direction of a superior agency or actor. A balance of power system, regardless of its specific form, is a decentralized, pluralist system.

Reinhold Niebuhr*s comment that "a balance of power is in fact a kind

of managed anarchy" depicts this.^ There are, of course, distinctions

between the various configurations of balance of power systems. Some

are more concentrated than others. However, the world having ex­

perienced a balance of power system for at least the last three hundred

years, it is very easy to forget that in the universe of possible

international systems the balance of power system is pluralistic and

reflects dispersion of power. A variety of balance of power systems

have been proposed as being capable of reducing the amount of war in

the international system. We have classified these specific types

as either unipolar, bipolar, concert, or multipolar.

A unipolar system is one where there is a clear leader whose power is substantially greater than that of the nearest rival. The

system is not one of dominance, however, because a large coalition of actors would still have power superior to the single powerful actor.

Nevertheless, in the absence of a clear consensus as to the evil of this actor, the system could be described as being in a state of hegemony.

A.F.K. Qrganski presents the rationale for believing that a unipolar system would keep the peace. His position is that*

A preponderance of power on one side...increases the chances for peace, for the greatly stronger side need not fight at all to get what it wants, while the weaker side would be plainly foolish to attempt to battle for what it wants.2®

One must be careful to distinguish, however, between a unipolar system and a condominium. Condominium might be said to occur when most of 27 the powerful states are in a position of agreement. The unipolar advocacy claims that peace results from the regulative strength of 21

a single actor. A condominium model says that peace results when most

of the powerful actors are satisfied and are in consensus — which is

not at all surprising.

A bipolar system is one in which a large portion of power is

concentrated in two powers; that is, where there is little gap in the power of the first two powers, yet a substantial gap between these and the third, fourth, and fifth powers. Kenneth Waltz is a strong advo­ cate of a bipolar world for the purpose of limiting violence. Basically, in his analysis he compares the bipolar model with the multipolar model and finds the former superior for four reasons. First of all, with only two world powers there is constant competition and consequently no peripheries. Thus uncertainty is eliminated. Secondly, the range of factors included in the competition increases as the intensity of the competition increases. Thirdly, there is constant pressure and re- occurring crises in a bipolar world, thus preventing slackness or mis­ calculation of intentions. Finally, Waltz sees bipolar systems as being able to comprehend and absorb revolutionary changes without great dis­ ruption because of the constancy of effort and the preponderant power 28 of the two major contenders.

Essentially this argument is similar to that made for multi­ bloc systems, i.e., the elimination of peripheries reduces uncertain­ ties and prevents miscalculation. Also, because of the concentration of destructive capacity, fear of mutual suicide prevents war. What keeps the peace is "stable deterrence." As Klaus Knorr writes*

A consequence of this pattern is that either of the two super­ powers, when contemplating the use of military power against another state, must reckon with the intervention of the other great power...This prospect naturally inhibits resort 22

to military power unless the purposes axe very compell­ ing. 29

As with the unipolar system and condominium, one must here also

be aware of the tendency and problems associated with viewing short

term behavior as a system defining characteristic. In a bipolar con­

text, the difficulty arises from the fact that almost all wars can be

and usually are envisioned as having two sides. As we will discuss

later, the common attitude that "an enemy of my enemy is my friend"

makes even multipolar conflict appear to be bipolar. One must be care­

ful not to view all periods of conflict as being bipolar based on tem­

porary alliances.

When four or five top powers in a balance of power system

possess relatively equal power, a concert system exists. Richard

Rosecrance, among others, is a strong advocate of this type of system

for limiting violent conflict. He calls it a quintipolar system and

sees it as the optimum point between a bipolar and a multipolar s y s t e m .

It combines, according to Rosecrance, the best of each of these

systems — i.e., sufficient power concentration to insure control and

regulation, and yet sufficient dispersion to insure flexibility and the

ability to maneuver. Usually, however, those who advocate this system

place the emphasis on control or regulation. Joint decisions are arrived at between the great powers either through agreement and con­

sensus or through a balancing of interests. These decisions are then

imposed on the rest of the system. For instance, the concert period begun in Europe after 1815 has been described as "uniting the sum total of states in a federation under the direction of the major powers....

The second-, third-, and fourth-rate states submit in silence and 23

without any previous stipulation to the decisions jointly taken by

the preponderant powers."31

Mention is often made in reference to this type of system

of the beneficial aspects of the equality of the major states. How­ ever, we should be aware of the fact that it is not so much the degree of equality as the number of equal states that serves as the defining characteristic of the system. For instance, a bipolar system posits relative equality between two major powers. The relationship is, of course, that in order to be a major power a state must have approxi­ mate parity of power with the leading powers. We thus see that a system where power is dispersed is generally going to have more major powers than a system where it is not.

A multipolar system is a balance of power system in which there are a large number of major powers, usually more than six. This implies that power is very dispersed. The key points of those who advocate a multipolar system as the solution to war center around the ideas of flexibility and uncertainty. For instance, in Karl Deutsch and J. David

Singer*s well-known argument in favor of a multipolar system, they point out that a multipolar system affords a greater number of interaction opportunities than the other systems we have discussed.-^ These inter­ action opportunities provide the flexibility in the system necessary for it to adjust to change without violence. Interestingly, Deutsch and

Singer believe that an increase in the mere number of states in the system, even if the new states are very weak, and not just in the num­ ber of great powers enhances this effect.

Hans Morgenthau*s position as to the benefits of numbers is similar, except he is concerned with the great powers. In his zk

formulation, greater interaction opportunities increase uncertainty

and uncertainty breeds caution. Consequently, changes are likely to

be incremental and therefore peaceful*

The greater the number of active players, the greater the number of possible combinations and the greater also the uncertainty as to the combinations that will actually oppose each other and as to the role the individual players will actually perform in them...In consequence, the extreme flexibility.. .fnakesj it im­ perative for all players to be cautious in their moves on the chessboard of international politics and, since risks [are*3 hard to calculate,fnations arej compelled.. .to take as small risks as possible.

There is a tendency to equate the multipolar system with

"balance of power", and the pluralist processes that accompany multi­ polar systems with the process of "balance of power." This is re­ grettable because the more concentrated forms of balance of power systems may not exhibit the processes, such as flexibility of alliance formation, that the more dispersed forms (particularly the multi­ polar) do. For the sake of clarity and understanding, it is im­ portant for the reader to note that in this study we do not equate the processes accompaning a multipolar system with the term "balance of power.'*

The advocacy of the multipolar system is extremely old in political literature, and in its original form it rested on the same rationale that it does today. Although the term multipolarity was not used, the desired processes, which were supposedly necessary to

Bolve the problem of war, were to be produced by an equal distribution of power among a large number of leading states. As we have seen, such a distribution would produce the characteristic multipolar system.

This doctrine of'fequal shares" was a topic of a good deal of debate prior to the late 19th century on grounds of both desirability and 25 feasibility. It was felt by some that an equal distribution of power

represented the ideal condition for the functioning of a balance of

power system. For instance, an anonymous eighteenth-century lampoon

of the balance of power defines it as "an equal Distribution of Power

among the Princes of Europe, £ which]) makes it impracticable for the

one to disturb the Repose of the other.In another work of the

same period, the Abbe de Mably ascribed to Queen Elizabeth of England

the assertion that "it is a matter of dividing Europe into states

almost equal in order that, their forces being in balance, they will

fear to offend one another, and hesitate to plan too great d e s i g n s . "35

Others, however, such as Talleyrand denied both the desirability and

the feasibility of this equal shares doctrine.-^ Vattel took a middle

position in his Law of Nations — that it was desirable but not

possible. His solution was to produce the equality by alliances*

The surest means of preserving this balance of power would be to bring it about that no State should be much superior to the others, that all the states or at least the larger part, should be about equal in strength. This idea has been attributed to Henry IV, but it is one that could not be realized without injustice and violence...It is simpler, easier and more just to have recourse to the method...of forming alliances in order to make a stand against a very powerful sovereign and prevent him from dominating.37

Thus one can see that in the advocacy of the multipolar system, as with so many other aspects of human existence, the individuals and the weapons may change but the essence of battle and the battle­ fields themselves remain essentially the same over great periods of time. Extreme Dispersion Systems

At the dispersed end of the power spectrum there are a num­

ber of theories which posit that an extreme dispersion of power is the

solution to the problem of war. These theories envision a spectrum of

worlds whose composition ranges from numerous small states to the ab­

sence of state actors entirely. One might think that this general

category is merely an extension of the balance of power category.

However, in these theories the rationale is very much different. These

theories generally predict that extreme dispersion, and thereby wide­

spread equality, produces a political process which negates the politics

of confrontation totally and substitutes the politics of cooperation.

Peace is not produced because of a flexibility and uncertainty which

restricts the violent ambitions of actors, but rather because the

politics of cooperation eliminate the need for violence.

For example, Silva Brucan, who presents one variation of the

numerous equal states solution, maintains that the origin of politics

is to be found in inequality. He states that "inequality among nations

is at the root of politics in the international arena, "3® Inequality

produces class relations and class struggle. These are at the roots

of national antagonism and conflict. Consequently, if power is dis­

persed among many nations equally, class relations and the correspond­

ing struggle for domination will disappear. By equalizing power one

eliminates the means by which various disparities, gaps, and differences in society are controlled and manipulated.-^ Conse­

quently, a desire for belonging, participation, and cooperation re­

places the politics of confrontation as a means of social control. Similarly, Richard Sterling sees peace as being disrupted

primarily by inequality. War is caused, in his view, by the struggle

for dominance and the fear of being dominated.^ In a world of equals

where there exists neither the strong nor the weak, the desire and

fear of domination (along with the injustice they breed) will be

eliminated. In such a situation mutual trust and respect would out­

weigh mutual fear and contempt. Consequently, violence will be con- 41 trolled, diminished, and eventually eliminated. Both Sterling and

Brucan see the end result of such a system of numerous equal states

as being some sort of supranational unification. But they see this

as a result of a peaceful, cooperative world and not as the cause of

it.

Other solutions which rely on extreme dispersion of power

posit the weakening or dissolution of the state system. The functionalist solution, where state power is eventually dispersed 42 among many international functionalist organizations, is one. The anarchist solution, which calls for the abolition of the state and indeed the rejection of any sovereignty over the individual, repre­ sents another.^ in general, all these schemes which rely on ex­ treme dispersion for the solution to war do so in the belief that such a structure will eliminate the politics of confrontation and thus eliminate or reduce war.

Historical Interpretation

The origin of many of the ideas contained in the last section is unclear. It is fairly obvious, however, that historical analogy plays a major role in the formulation of a great deal of international 28

relations theory. Specific analogies result from an assortment of

explanations, found in a variety of literature, for the occurrence of

historical patterns or events. But in dealing with the explanation

of historical patterns or events, one must be very careful for at

least two reasons. First of all, as F.H. Hinsley has written*

"People often study history less for what they might learn than for

what they want to prove. Secondly, historical interpretation

changes with time and is often influenced by nostalgic hindsight.

For instance, from the terrible violence of the twentieth century

"men look back nostalgically to the mid-nineteenth century as a time

of peace. "^-5 From the terrible violence of the early l6th century,

Italians looked back and "saw the age of Lorenzo the Magnificient

jlast half of the 15th century] bathed in a golden sunlight of serenity and moderation.In fact, the wars of 19th century Europe and of

15th century Italy lacked neither wanton destruction nor a terrible spilling of blood.

Consequently, in this section we do not wish to suggest that the interpretations presented are totally valid. These interpretations are presented merely to show the extent to which propositions about the effects of concentration and dispersion of power have affected the writing of history.Sometimes in the writing of history these propositions are not very explicit. Often they are implicit, such as when a description of the power structure of a period is followed immediately by a description of the pattern of war for that period.

For instance, the very term Pax Romana makes the connection between peace and the dominance of the Mediterranean and European worlds by the Roman Empire. In this section both types of propositions are noted. 29

The Peace of Westphalia in l6*f8, which ended the Thirty

Years War, is often cited as the initiating point of the balance of /tQ power system in Europe, In contrasting the period after 1648 with

that which preceded the Peace of Westphalia, there is a tendency to

give much of the credit for the change in the character of war to the

change in the structure of the system. For instance, Eugene Rostow

writes that:

The first half of the seventeenth century was a time of limit­ less and terjjible warfare. Throughout most of the eighteenth century, on the other hand, Europe existed in relative tran­ quility within the system of order established by the Treaty of Westphalia in 1648.^9

That system of oJder was, of course, a balance of power system.

Richard Rosecrance similarly views the period as one of relative

stability. He gives the credit to the relative dispersion of power:

Despite periods of warfare, the eighteenth century was a generally stable period in international relations. The balance of pober mechanism did not prevent military con­ tests, but it did prevent them from getting out of hand. The limited ajnbitions of European autocrats and cabinets resulted in limited diplomacy and militarism. The out­ comes of international politics did not exceed the bounds of convention and precept.-5°

A number of scholars have specifically identified the distribution

of power during this period as being one of near equality. For example,

F.H. Hinsley writes:

In the long process of rise and fall in the relative strength of states, Europe was attaining from the beginning of the eighteenth century a condition in which there was a greater degree of near-equality between a larger number of leading states than ever before in its history.■53-

Others, such as Stanley Hoffmann, have explicitly made the connection

between this pattern of power distribution and the pattern of war during the period ji 30 The balance operated effectively because the treaties of Westphalia had redistributed territory so as to create a number of major states capable of neutralizing each other and had also removed the poisonous element of religious conflict...The political result was a mechanical balance, frequently disturbed however, either because a state could never be sure in advance whether or when others would try to curb it, or because of individual ambitions. Hence, numerous limited wars occurred* stylized wars of position that only rarely affected the civilian population.32

The eighteenth century in this view was by no means a

peaceful period, but its wars were conservative. On the one hand,

no state was capable of conquering Europe; on the other, Europe

did not collapse into chaos. Wars existed and were numerous, but

they were not total. Because they were limited in objective, they

tended to also be limited in scope and practice. The purpose of

war did not require or justify annihilation, occupation, or the

forceable imposition of alien social institutions onto other

populations.^

The system of order which prevailed in the 18th century

"finally disintegrated in the tumult of the French Revolution."3^

With the Revolution, state power added a new dimension -- the

sociological. In Clausewitz's phraseology, the wars of kings became

the wars of peoples. France was the first country to exploit this

power, and during the course of the Revolution the system became

unipolar. It required the combined forces of Austria, England, Prussia,

and Russia to break the power of f'rance.

It was to be another hundred years until the world would again

experience a total war. Somehow, for nearly a century the leading

nations of Europe managed to conduct the affairs of most of the globe without falling into a total war.'*'* This period was, as Henry Kissinger has suggested, an era of small, controlled wars rather

than of peace.While Europe was enjoying this period of un­

precedented peace, the extra-European world had more than its Bhare

of war. Particularly, the European powers fought a large number of

small wars of conquest. Also, the five great European powers fought

each other a number of times. Yet in these great-power wars we see

remarkable restraint. Prussia fought Austria in 1866 without another

great power intervening. Prussia fought France in 1870-1871 without

another great power intervening. And the Crimean War, where most of

the great powers were involved, was conducted as a limited war with 57 limited means and limited objectives. Despite these wars, security

j- Q and stability, at least in the core area, was achieved.

During most of this concert period, five great powers

(Great Britain, France, Russia, Prussia, and Austria) monopolized the 59 major instruments of violence. Moreover, there was a rough parity of strength between them. The structure of the system is most often described by these characteristics — five major powers with relatively equal strength.^0 However, what exactly accounted of the success of the Concert of Europe in preventing general war is a matter of some speculation. Some analysts, such as E.H. Carr and

A.F.K. Organski, credit British supremacy (Pax Britannia).

However, the dominant view appears to be that stability resulted from the "balance of power" operating between the five great powers.

Of course, as with most explanation other factors are also often considered vital, such as a moral consensus of the European community and some brilliant diplomacy. Many of these other explanatory factors 32

both have and have not been seen as resulting from the structure of

the system.

Interestingly, the existence of a concert system has been

used, either explicitly or implicity, to explain conditions of

political stability in a number of other historic periods. For

example, the Greek City State system originally had an extremely dis­

persed power structure. By the fifth century B.C. it had developed

into a system of a limited number of powerful states — Athens, Sparta,

Corinth, Thebes, Argus, Acragas. After this development the system

experienced frequent wars,yet the first total war (Peloponesian) did

not occur until power became concentrated in the hands of Athens and 62 Sparta after the Greek victories over the Persians.

The Italian city state system after the Peace of Lodi in 1

also took on the concert characteristic of about five major powers of roughly equal strength -- Venice, Milan, Florence, Naples, and the

Papal States. In the forty years that followed, the political pro­ cesses, such as communication patterns and alliance formation, took on characteristics similar to those which occurred in 19th century

Europe.^ Conflicts were numerous yet none were total:

So for forty years, by virtue of the mutual jealousies of its balanced states, by a politics of continuous tension, and by the help of its new diplomatic machinery, Italy did enjoy a kind of uneasy peace. Although scarcely a year was without some sort of crisis or potentially dangerous intrigue, although, at times, the whole system seemed on the brink of disaster, disaster was each time averted. Wars were less destructive than they had been, absorbed leBs of menfs energies, and con­ sumed less of the social income. No major towns were sacked; no desperately bloody fields were fought. And for three years, almost, out of four there was no fighting anywhere in Italy worth a historian's serious attention.

The system disintegrated with the introduction of French and Spanish 33 force at the end of the century. With the introduction of these two

great powers, Northern Italy became the battleground for a series of

devastating wars which destroyed the power and influence of the Italian

city states forever in European affairs.

The Concert of Europe is often seen as breaking down just

prior to World War I. The breakdown was a common British explanation

in the 1920's of the cause of that war.^ The manifestation of this

breakdown most often referred to is the structure of the alliance system.

According to this view, the alliance system had become a two-bloc system

after 1890.^ German miscalculations and French passion are said to

have created inflexibility, thereby making limited objectives and 67 limited war impossible. Because of this alliance system, power became

so concentrated that if war came (as it did) it had to be bipolar in

configuration and gigantic in scope (as it was).^

According to this explanation, World War I was caused by a

concentration of power. However, even if one tends to accept this

view in general, it is unclear whether the war is believed to have re­

sulted from a concentrated structure or from the failure of the major

powers to adhere to the rules or natural processes of a dispersed

system. In fact, the state power structure seems to be one of some dis­

persion. There were more first rate powers at that time than at any

time in the previous century -- Austria, France, Germany, Great

Britian, Italy, Japan, Russia, and the United States.^

The concentration of power produced by a bipolar system has

been used to explain a number of periods of extreme confrontation.

The classic confrontations include Athens vs. Sparta, Rome vs. Carthage, the Empire of Charles V vs. France, England vs. France, 3*f

and the Cold War confrontation of the United States and the Soviet

Union.However, as the World War I example makes clear, it is often

not clear whether the explanation centers on a bipolar state power

structure or on a bipolar alliance structure. In addition, due to

the absence of hard data, many periods of confrontation which have been seen as bipolar can also be seen as unipolar.

Unipolar systems have figured in the explanation of some of the most destructive periods in modem history. Usually the con­ frontation is seen as the struggle for dominance or the drive for hegemony. The explanation rests on the temptation to expand the unipolar system to one of dominance. For instance, in 1805 France was the largest and most powerful state in Europe, but she was not superior to the possible coalitions that could be put against her.'7**- Europe still contained a state system operating under the rules of balance of power, "and Napoleon might still have swung in either of two directions in his over-all strategy — toward balance of power or toward empire.He, of course, committed himself to a grandiose imperial plan for France and Europe; consequently, he committed Europe to a titanic struggle. World War II is often viewed in a similar manner, with Hitler and Germany making the decision to wage a war for total dominance. Of course, as we have mentioned in reference to the Pax Britannia, supposed unipolar structures have also figured in the explanation of some of the most peaceful periods in modem his­ tory. 35 Policy Prescription and Formulation

A third and final area where one can find numerous prop­

ositions relating war to power concentration or dispersion is the

area of policy prescription and formulation. Both the advocacy and

the actual initiation of foreign policy have often been directly re­

lated to beliefs regarding the effects of power distribution. This

is significant because so often those concerned with the day-to-day

operation of international affairs tend to regard systems theory and

historical explanation as being non-essential or trivial. A reader

with that attitude might readily disregard the macro-historical analysis contained in this paper. However, as will be briefly demonstrated, the same types of propositions which turn up in the systems and historical areas, also resurface in the policy area.

To find evidence of this one need only pick up any one of a number of current books on American foreign policy. For instance, picking up Alastair Buchan’s Power and Equilibrium in the 1970*s. one finds an optimistic view of the future based on the continued dispersion of power in the international system. He presents the creation of a pluralist world as the long-term objective of American 7 h. foreign policy. He further states:

I hope I have said enough to impart my personal conviction that a plural world of five, possibly more, centers of major international power and influence is re-emerging. I could offer a string of American quotations from Thomas Jefferson to William James to Henry Kissinger to illustrate that this is not an alien concept to the American view of the world. Why did the United States intervene in the two world wars if not to prevent the domination of Europe by a hegemonial power? What was American pressure for universal decolonization instead of confronting the Soviet Union by means of an American-British- French entente in 19^7~^8 other than a consequence of a con­ viction of the validity of pluralism?75 36 In William 0. Douglas' International Dissent* Six Steps Toward

World Peace, three of his six steps are related to the supposed

beneficial effects of a dispersion of power* the end of all mili­

tary alliances, the elimination of rich-poor cleavages between nations,

and the abolition of all colonies and protectorates.^ Even in

Leonard Beaton's scheme for an international security authority,

The Reform of Power, we find evidence of a faith in pluralism. His

system is designed so that it "should give the whole world, and especially the middle powers, a direct share in their own security" and "should preserve political variety with all the advantages of a 79 plural world."

Much of the argument on aid to developing nations is pitched at the systems level, usually favoring equality and hence dispersion as a method of keeping the peace. Mahdi Elmandjra writes, for instance*

It is because excessive imbalances of power and influence con­ stitute one of the major threats to peace and security that redistribution becomes a key regulatory function within any system.78

Others see different areas resulting from dispersion. Viewing a depolarizing world, Amitai Etzioni sees new dangers as well as new opportunities:

At worst, the more aggressive big powers will make the deals while the US and USSR will seek vainly to make reality con­ form to a bipolar image j at best, the US and USSR — sometimes bi-laterally, hopefully often in cooperation with all powers — will advance a new world order, through neutralized zones, limitations on the spread of nuclear arms, arms reduction and strengthening the UN. To put it more formally, as pluralism increases, the interests of the super-powers, and to a some­ what lesser degree, of all the powers, might well become clad in universal rules, backed up by a growingly effective international machinery aimed above all at limiting the multi­ faceted competition to non-military m e a n s . 79 37 Many also put their faith in concentration. For instance,

Robert Osgood and Robert Tucker have noted the need for an equiva­

lent to the U.S. Cavalry to keep the peace on the international fin frontier. In fact, as Richard Sterling notes, the "law and order"

advocacy based on a concentration of power and the "social justice"

advocacy based on a dispersion of power that we find at the domestic

level of politics is often carried directly to the international level 81 with only slight modification.

In the last few years, the most influential prescription based on power distribution is the much heralded Nixon-Kissinger

"structure of peace." The astute management of relations between five centers of power (US, USSR, Europe, China, Japan) is supposed to bring a generation of peace. In the light of this belief, the

Nixon-Kissinger foreign policy was designed to induce a greater degree of moderation and cooperation. It was to emphasize the 82 laissez-faire arrangements of previous multipolar ages.

Ideas as to the desirability of a particular power structure have been translated into policy not only at the national level, but also at the international level. At least three times since the

Treaty of Westphalia there have been major attempts at restructuring the world based on one structural prescription or another. At the

Congress of Vienna the great powers rearranged the map of Europe so that the five great powers were nearly equal in strength. That they would place their hope on equality and relative dispersion is not surprising in light of their recent experience with a unipolar system.

At Versailles the Allies placed their hope for peace in the concen­ tration of power. The League of Nations did not live up to their expectations, of course, but as a collective security organization it was designed to mass overwhelming power against an aggressor.

That they would opt for power concentration is not surprising if one views the system which failed them in 1$0A in terms of state power rather than alliance configuration. Finally, at San Francisco the decision was back to relative dispersion. The United Nations is, in its Security Council aspects, merely a reformulation of the Con­ cert of Europe. NOTES TO CHAPTER TOO

*^It is often extremely difficult to place a proposition or theory exclusively into one of these three categories. For instance, a theorist might be using an analysis of international systems theory to support the advocacy of a particular policy. We have attempted to classify according to the purpose of the proposition or theory, and such an analysis would be placed under "policy prescription and formulation" rather than under "inter­ national systems theory." However, the important aspect of these propositions and theories is not specifically the category under which they fall but rather the fact that they exist in many areas of international relations.

^Raymond Aron, Peace and War (New York* Doubleday, 1966), p. 94.

3j. David Singer and Melvin Small, "Alliance Aggregation and the Onset of War, 1815-1945," J* David Singer, ed., Quantitative International Politics* Insights and Evidence (New York: The Free Press, 19^8),p. 247. 4 Andrew McFarland, Power and Leadership in Pluralist Systems (Stanford, California* Stanford University Press, 1969), p. 132.

■^Inis Claude, Power and International Relations (New York* Alfred A. Knopf, 1964), p. 2007

6Ibid., p. 212.

"^Raymond Aron, "The Anarchical Order of Power," Daedalus. 95 (Spring, 1966), p. 485; Claude, p. 216. ®As quoted in Edward Vose Gulick, Europe*s Classical Balance of Power (ithica, New York* Cornell University Press, 1955)* PP» 6-7.

%.H. Hinsley, Power and the Pursuit of Peace (Cambridge* Cambridge University Press, 19^3), pp. 26-42.

lOsaul Mendlovitz, "Models of World Order," in Richard B Gray, ed., International Security Systems (Itasca, Illinois* F.Er Peacock, 19&9)” pp* 178-192. 13-Aron, "The Anarchical Order of Power," p, 485. . *

39 *10 o * X6Theodore Abel, "The Element of Decision in the Pattern of War," American Sociological Review. 6 (19^1), p. 859*

"^Gulick, pp.5-6.

l^endlovitz, p. 1 8 5.

15Quincy Wright, A Study of War, 2nd ed. (Chicago* University of Chicago Press, 1965)» P* 1345* ^■^Hans Morgenthau, Politics Among Nations, 5th ed. (New York* Alfred A. Knopf, 1973). P* .

l?Claude, p. 110-113.

l®As quoted in Robert E. Osgood and Robert W. Tucker, Force, Order, and Justice (Baltimore: Johns Hopkins Press, 1968), pp. 108-109.

•^Memllovitz, p. 188.

^Richard J. Barnett, "Regional Security Systems", in Richard B. Gray, ed., International Security Systems (Itasca, Illinois* F.E. Peacock, 1969). pp. 75-76.

22Roger Masters, "A Multi-Bloc Model of the International System," American Political Science Review, 55 (December, 1961), p. 782.

a portion of Nineteen Eighty-four relating power distribution to war can be found in George Lanyi and Wilson McWillians, eds.. Crisis and Continuity in World Politics (New York* Random House, 1 9 6 6), pp. 496-5 0 3, under the title "Triangular Perpetual War."

2^Gulick, p. 9 8. ^5as quoted in Claude, p. 79*

^A.F.K. Grganski, World Politics (New York* Alfred A. Knopf, 1968), p. 293.

^Mendlovitz, p. 1 8 9.

2®Kenneth Waltz, "The Stability of a Bipolar World," Daedalus. 93 (Summer, 1964), pp. 882-886.

29j

33-Morgenthau, p. ^3 8.

-^Karl Deutsch and J, David Singer, "Multipolar Power Systems and International Stability," in James N. Rosenau, ed,, International Politics and Foreign Policy (New York: The Free Press, I9 6 9), pp. 317- JZT.------

-^Morgenthau, p. 3^As quoted in Gulick, p. 3 8.

•^As quoted in Ibid,

36Ibid, pp. 58-59.

^As quoted in Ibid., p. 60. 3®Silva Erucan, The Dissolution of Power: A Sociology of International Relations and Politics (New York: Alfred A. Knopf, 1971), P. 7.

39Ibid., p. 1 1 3. ^Richard W. Sterling, Macropolitics: International Relations in a Global Society (New York: Alfred A, Knopf, 197*0» P* 559.

^Ibid., p. 8 7. ^David Mitrany, A Working Peace System (Chicago: Quad­ rangle Books, 1 9 8 6).

^George Woodcock, Anarchism: A History of Libertarian Ideas and Movements (New York: Meridian Books, I9S2 ), pp. 17,33»

^Hinsley, p. 1 3 .

^■^F.S. Northedge and M.J. Grieve, A Hundred Years of Inter­ national Relations (New York: Frederick A. Praeger, 1972JT P*l« ^Garrett Mattingly, Renaissance Diplomacy (London: Jonathan Cape, 1955)» P. 99.

^ W e have attempted to rely totally on the interpretations of others in this section and not add our own thoughts. However, in piecing together these interpretations from a variety of sources we have had to make a number of implicit judgments as to which in­ terpretations might be related to each other. 42

^®Winfried Franke, "The Italian City-State System as an Inter­ national System," in Morton Kaplan, ed., New Approaches to Inter­ national Relations (New York: The Macmillan Company, 1970), p. 426.

^Eugene Rostow, Law. Power, and the Pursuit of Peace (New York* Harper and Row, 19&8), p. 18, -

^ORichard Rosecrance, Action and Reaction in World Politics* International Systems in Perspective (Boston* Little, Brown, 1963), P. 30.

■^Hinsley, p. 1 7 6,

52Stanley Hoffmann, "International Systems and International Law," in his The State of War (New York* Frederick A. Praeger, 1966), p. 101. •53k.J. Holsti, International Politics* A Framework for Analysis (Englewood Cliffs, New Jersey* Prentice-Hall, 1 9 6 7)", p. 6 3. cZf “^Rostow, p. 18.

-5-5ibid., p. 26.

•5^Henry A. Kissinger, Nuclear Weapons and Foreign Policy (New York* Harper Brothers, 1957), p. I**2-

•57Hinsley, pp. 226-227.

3®Hoffmann, "International Systems and International Law," p. 107. ■59Rosecrance, Action and Reaction in World Politics, p. 2 8 7 .

^°Ibid.; Gulick, p. 4-; Northedge and Grieve, p. 14.

^Organski, p. 291; Claude, p. 71*

62Holsti, pp. 44-45.

63ibid., pp. 54, 92-93j Franke, pp. 448-449.

^Mattingly, p. 9 6. ^According to Rosecrance, Action and Reaction in World Politics. p. 181. 66Waltz, Man. the State, and War, p. 218. ^ 3

^Morton Kaplan, "Intervention in Internal War* Some Systemic Sources," in James Rosenau, ed., International Aspects of Civil Strife (Princeton* Princeton University^Press, 1964), p. 97.

Rosecrance, International Relations, p. 118.

^Morgenthau, p. 3^1 • ?ty]artin Wright, "The Balance of Power and International Order," in Alan James, ed., The Bases of International Order (London* Oxford University Press, 19^3)* P» 1£T7

^■Rosecrance, International Relations, p. 120.

?2Gulick, p. 25* . 73c>rganski, p. 2 9 1 .

^Alastair Buchan, Power and Equilibrium in the 19701 s (New York* Frederick A. Praeger, 197377 P* 45.

75ibid., p. ¥f.

^^William 0. Douglas, International Dissent* Six Steps Toward World Peace (New York* Random House, 197i)» PP* 30-31» 48, 72.

^Leomard Beaton, The Reform of Power * A Proposal for an International Security System (New York* The Viking Press, 1972j» pp. 156-157. ?®Mahdi Elmandjra, The United Nations System * An Analysis (Hamden, Connecticut* Archon Books, 1973)» P» 300. 79Amitai Etzioni, "Strategic Models for a De-polarizing World," in Clagett G. Smith, ed., Conflict Resolution* Contributions of the Behavioral Sciences (Notre Dame, Indiana* University of Notre Dame Press, 1971), p. 546. ®°0sgood and Tucker, p. 125*

^Sterling, p. 86. ®2William A. Hazleton, "Kissinger on Globalism, Regionalism, and the Latin American Community," Paper delivered at the 1974 Annual Midwestern Regional Meeting of the International Studies Association, Chicago, November 21-23, 197^» CHAPTER III

STRATEGY AND CONCEPTS

Having presented in the last chapter a large sample of the

thought (much of it seemingly contradictory) on how power distribution affects war, we are now prepared to outline our strategy of inquiry for determining if, how, and why the distribution of power affects inter­ state war. It is also necessary at this point to go into some detail concerning the key concepts which are related to our independent and dependent variables. The first section of this chapter is devoted to the strategy of inquiry; the second to the concepts.

Strategy of Inquiry

Although we have drawn insight and evidence from a variety of disciplines and levels of analysis, our principal focus is political and international. There are, as J. David Singer writes*

...always several ways in which the phenomena under study may be sorted and arranged for the purpose of analysis. Whether in the physical or social sciences, the observer may choose to focus upon the parts or upon the whole, upon the components or upon the sys­ tem.-* This study focuses upon the whole, and the entire international system constitutes the basic level of analysis. The perspective may be termed systemic or macroscopic.2 This study is not, for instance, concerned with determining which states go to war, but rather with determining general patterns of war. It is not concerned with the power of M 4 45 specific actors in the system, but rather with the power distri­

bution of the system as a whole.

Given this systemic perspective, the key elements in the

strategy of inquiry are the strategy whereby the theory is formulated

and the strategy which is used to gather original evidence.

Strategy of Theory Formulation

The strategy for the formulation of theory used in this study

can best be described as eclectic and interdisciplinary. This study is

both inductive and deductive, and it relies on elements of both the

empirical, and the philosophical approach to theory building. Moreover,

it draws on the theory and findings of a number of disciplines.

This study is based on the belief that there is a need for the

theorist to oscillate between induction and deduction in formulating

his theory. In the social sciences, at least in their present state,

there are too many problems associated with each method for a theorist

to rely totally on one or the other. Consequently, there is a need to

pursue discovery by reflection and discovery by investigation at the

same time. In addition, this study uses elements of both the empirical

and the philosophical approach to theory building. That is, it

attempts to use theoretical propositions whose validity has been

"tested by reference to empirical events" if available. However, these

are often not available. Also, the methodological problems of many of

those that are available make their validity less than certain. Con­

sequently, they have been supplemented with propositions that, although

"untested", seem analytically sound. These are based on some type of data, although usually it is not readily quantifiable. Historical 46

judgment falls into this category, for instance. Fortunately, while

"both types of propositions have problems, in many ways (at least in

relation to this problem) they tend to support each other.

Whereas we have used both types of propositions for the theory

formation, we have leaned slightly toward the empirical approach in

the actual expression of our theory. In particular, we hope to have

expressed our propositions in a form so as to allow precision of

statement. This permits quantification and thus allows for the

presentation of quantified evidence that has not previously been

brought to bear on this problem. We do not wish to imply, however,

that one form of matching reality against theory is totally superior

to another. As will later be emphasized, we view all forms of evi­

dence -- to include at least historical, quantitative, and transfers

from analogous situations — as desirable and necessary. Moreover, no

single form (however significant) is sufficient in itself.

The essence of the relationship with which we are concerned

at the international level has been dealt with in many disciplines and

at many levels of analysis. Because of this, there are a number of

analytic schemes and various types of evidence available on the subject

in these disciplines. The major disciplinary sources of theory and

evidence available in the area of our concern are history* both

ideographic and nomotheticj various social and behavioral sciences

other than international relations? and international relations itself, both traditional and behavioralist. From within these various disci­ plines the specific works used have come from a variety of levels in a variety of systems. At the small group level a number of small group studies from

the social, political, management, and communications perspective have

been helpful. The principal problem in using these studies is that the

condition of isomorphism is sometimes rather tenuous between the low-

level abstractions found in these studies and the more complex relation­

ships found at the international level. For instance, one may be hesi­

tant to equate peer power distribution in small groups of individuals to

distribution of power among states.^ At the sub-national level sub­

stantial guidance was found in both the community power and the

community stratification studies. Unfortunately, only a few of these

valuable studies are concerned with how the general welfare of the

community is affected, if at all, by power distribution. The concern h. has been principally descriptive. Only recently has the shift been

away from a concern with "Who Governs?" to one of "With What Effect?

The debate between the pluralists and elitists at the community level

has also been extended to the national level. In addition, at the national level literature dealing with theories of violence, revolution, and economic and resource distribution were helpful. At this level the analogies become easier to make; however, the trade-off comes in the loss of the clarity and precision that we find in the lower-level studies. At the international level, works that deal with our topic are found either in diplomatic history or international relations literature. Most of these works were discussed in Chapter II.

When all of these levels are combined, the researcher realizes that there is a substantial body of literature concerning itself with how power is distributed — whether it be in a committee or in the international system — and with what effects. The controversies, 1*8

terminology, and interest in specific effects vary with the level of

politics under analysis. Nevertheless^ all of these levels of analysis

contribute to some degree to the understanding of what occurs in the

international system when power is concentrated in the hands of a few

or is dispersed among many.

But this use of interdisciplinary sources presents a major

problem; to what extent can concepts and findings from one discipline be transferred to another? Por instance, can findings as to the effects

of the centralization of power in the hands of a few individuals in a small group on conflict within that group be considered evidence con« ceming the effects of the concentration of power in the hands of a few states in the international system on conflict within that system?

There have been many warnings against the use of analogy to transfer knowledge from one discipline to another. Clearly, various transfers of knowledge between disciplines and systems are a source of possible error in any study of this sort. However, we should be aware that such transfers are made every day in many areas. Often one has little choice because there is not enough work done in a single discipline or at a single level of analysis to draw reliable conclusions.

The position taken in this study is that evidence or theory that has been transferred from one discipline to another, or from one level to another, cannot be conclusive; but on the other hand, it cannot be ignored, because every piece of available evidence and theory should be incorporated into an analysis. In this study, supposedly analogous theory is treated skeptically and supposedly analogous results are treated as just another piece of evidence, in itself inconclusive. Hopefully, such a position is not likely to lead a researcher far astray before anomalies cause him to he redirected.

Strategy for Gathering Original Evidence

After one develops a theory based on both deduction and

existing evidence, the next step in the preparation of this type of

study is the gathering of original evidence to either support or re­

fute that theory. Commonly this second step is referred to as "testing"

the theory, but we have little faith in our ability to conduct a true

"test" to prove or disprove. At the present time the methods, con­

cepts, assumptions, and supposed interrelationships that must be relied

upon are all just too uncertain to allow one to believe that he can

adequately test propositions of the sort dealt with here. Rarely does

one find crucial experiments with clear findings in international re­

lations. Even when these tests appear to exist, they must be suspect

because such tests, at least at the international level, require a

tremendous number of assumptions. Rarely are there adequate grounds

for accepting these assumptions and thereby relating the results to

reality. Even if it is based on adequate theory and perfectly con­

ducted, no single study is capable of proving or disproving a theory.

At best the study can probe it. The study can provide evidence either

for or against, but it cannot prove or disprove.

What we have done in this study to gather original evidence

is to secure quantitative data on both state power and war at various points in time between 1862 and 1964. This has been converted into specific measures of the independent and dependent variables in order that the relationship between the two can be ascertained. This approach then is both quantitative and comparative. As such, this 50 study has not escaped the problems that have plagued other quan­

titative, comparative studies. These problems include the necessity

of the presumption of "all other things being equal" between two or

more cases except for the variables, the difficulty of obtaining

accurate aggregate data, the problem of the comparability of data

between different actors, and the necessity of sacrificing detail for

the sake of comparability.

Specifically, this approach is diachronic-nomotheticj that

is, a time series is involved and we are concerned with general

propositions.^ This leads to its own set of difficulties. First of

all, comparisons are being made across roughly 100 years. Conse­

quently, the "all other things being equal" proposition is very shaky.

The interstate system has changed radically over that period; and a

number of people have questioned--with grounds— how valid comparisons

are across such a period.'* Secondly, not only can the researcher not

be sure that if conditions in one period lead to certain effects then

parallel conditions in another period would lead to similar effects, but he can also not be sure even if conditions are roughly parallel.

For instance, one particular operationalization of this work's de­ pendent variable is battle deaths. Over the 100 year period one would expect the number of battle deaths to vary for a variety of reasons totally unrelated to power distribution. The operationalization in this case suffers from inflation produced by both a continued growth in the lethality of weapons and a growth in the world's population. The operationalization can be adjusted to minimize this distortion, but this introduces its own distortion. In addition, the representativeness of operationalizations varies over time. Further, the extent to which 51 this measure accurately represents power is going to vary depending

on the circumstances to which power was applied during the period.

We do not wish to gloss over these problems. Fortunately,

however, we think that we have found solutions to at least minimize

the distortion produced by them. These specific solutions are dis­

cussed in Chapter V, where the original evidence which has been

gathered in this paper is presented.

As the reader might recall from Chapter II, there has been a

great deal, of speculation as to the impact of system power structure

on war. But while a great deal of explanation can be found as to how

and why the system power structure should affect war, very little can

be found as to how much it should effect it. This is a question that

must be addressed here theoretically in order to be able to interpret

the evidence that we will later bring to bear to show how and why

power distribution affects war.

How much war should one hope or expect to be able to explain

by studying the structure of the international system? Kenneth Waltz

identifies three causes of war — man, the state, and the state sys- Q tern. Were the researcher able to conclude that these three factors were equal, then he might hope for 335^ explanation. But of course he cannot conclude this. Rousseau, for instance, thought that the major cause of war lay in the state system.^ Others might attribute little influence to the system. They might say that war is begun in the minds of men. Even if these minds are not unique, or if motivation for war is not arbitrary, they might conclude that these minds are influenced by the immediate social context and not by some larger 10 analytic entity such as the international system. We suspect that the truth probably lies somewhere in between. Those who stress

systemic causality have a tendency "to impute to the structure of the

international system controlling influence over the behavior of

actors,"11 to lead the observer into a position "which exaggerates T2 the impact of the system upon the national actor." On the other

hand, those who stress the individuality and free will of the actors

probably fail to consider the impact of the environment on the set of

choices available to the actors.

What we have then is another version of the determinism --

free will controversy. To us it seems unreasonable to expect either

no impact or an extraordinarily large impact from structure alone.

Most human actions are a mixture of free will and determinism, of

individual taste and a relatively fixed menu from which to choose. Or, as Karl Marx wrote: "Men make their own history, but they do not make it under circumstances chosen by themselves.1,13 Consequently, we would expect to be able to explain some behavior by reference to structure alone, but certainly not all behavior. As a result, we do not expect the relationships which we will posit to be extremely strong nor the evidence which we will present to be self-evident. However, we do anticipate that the relationships do exist and that the evidence will be convincing.

Concepts and Operationalization

Having discussed the strategy of inquiry in the first section of this chapter, we are now prepared to outline and discuss in detail the concepts which are at the heart of this study — the structural independent variable (power distribution) and the behavioral dependent 53 variable (war). Also, integrated into this discussion of concepts is

a discussion of the decisions which were made to operationalize these

variables in order to examine their relationship quantitatively.

The Independent Variable

There are four items that need to be discussed in detail which relate to the independent variable. These are the subunit of analysis within the system, the nature of the concept of power, the nature of the concepts of concentration and dispersion, and the distinction be­ tween and the composition of the elite and the total systems.

Subunit of analysis within the system

The power structure of the international system can be con­ ceived of and computed in two principal ways— either in terms of state power or in terms of alliance power. In other words, the power structure may change either through the gradual growth and decay of states, or it may change based on the shifting pattern of alliances.

Consequently, a choice must be made as to which of these subunits will be used to compute power distribution.

The reader might take offense at this limitation to two basic actors. There is certainly a great deal of literature demonstrating the importance of non-state centered actors. However, the significance of these types of actors to the relationship that we are examining has not been dealt with in detail either theoretically or empirically.

One should not a priori rule out the significance of this relationship.

There is just not the theory or data to make the use of these subunits a viable alternative for this study. & In this study we have chosen to depict the system power

structure in terms of state power and not alliance power. There are

two basic reasons for this. First of all, for at least the last 300

years "national actors...have "been the distinctive actors in the in­

ternational system.""^ World politics have centered on the inter­

actions among states, at least in the area of conflict behavior. The

second reason why we have chosen to use state power is that alliances

rarely act on significant issues as single unified actors, except per­

haps out of the dominance of a single state actor. An alliance rarely

ever attempts to transform but a small fraction of the total interests 15 of member states into common policies. ^ Indeed, often alliances are

* 1 i f created for essentially different purposes by national actors. Usually alliance members are united only in the determination of a

common enemy and are far from agreement in relation to the objectives

to be sought or the policies to be pursued. Consequently, they rarely

act as unified actors, even when they become semi-permanent.

Those studies that have tried to use alliances as the subunit

of analysis in order to determine the effects of power distribution on

war have run into a number of difficulties, in terms of both measure­

ments and concepts. The general topic of these studies has been

whether bipolar or multipolar systems have the greater amount of war.

The concept of polarity with which these studies generally deal has two

elements — the numerical clustering of states into groups and the dis­

tribution of power between groups.^ However, it is extremely diffi­

cult to measure the power of groups. For instance, how does one de­

termine how much power each state contributes to an alliance? Or, how does one distribute the state power of a nation that is involved in 55 four or five alliances? Consequently, while the theory is heavily

dependent on the distribution of power, the measurements have generally *| o been restricted to numerical clustering.

Problems which are centered more on conceptualisation than

measurement also arise when dealing with alliances as the subunit of

analysis. Michael Haas' study, "International Subsystems: Stability

and Polarity," provides an example of one such problem. The study

attempts to discover "whether the power structure of an international

subsystem in fact is a concomitant of one or another type of instability

in the behavior of members of the subsystem.However, the deter­

mination of power structure is based on historical judgment. In his

study the following time periods are seen as bipolar: 1649-1713,

1790-1814, 1823-1847, 1891-1918, 1919-1945, and 1946-1963. Accordingly,

the Wars of Louis XIV, the Napoleonic and French Revolutionary Wars,

World War I, and World War II all fall into bipolar periods. This

superficially tends to condemn bipolarity and consequently power con­ centration. However, there is no clear evidence that these periods were in fact bipolar. Two of the periods can be seen as unipolar

(1649-1713, 1790-1814) and three as multipolar (1823-1847, 1891-1918,

1919-1943) in terms of state power distribution. It appears that they are seen as bipolar in terms of alliances, not because of any unified goals or broad cooperation in objectives, and certainly not because of any real concentration of power in the hands of the alliances, but rather because of a basic theorem of human nature. This theorem is that?"The friend of my friend is my friend, the friend of my enemy is on * • my enemy, the enemy of my enemy is my friend." This explains why authors of The Wages of War could break all wars between 1816 and

1965 down into two sides and why Wayne Perris could study the power

capabilities between the "two sides" in all wars between 1865

and 1965^ despite the fact that a great deal of conflict is multi­ lateral. If one looks at temporary coalitions, one will almost

always see a pattern of bipolarity in all systems during conflict

periods. This may be a relevant finding, but it is the conflict which

causes the bipolarity and not vice versa.

While we believe that the state is the most appropriate unit

to use in dealing both theoretically and empirically with the relation­

ship between the distribution of power and war, we do not suggest that

there is not a relationship between alliance patterns and war or that the search for it should be halted. We merely wish to provide the reader with our reasons for not using the alliance subunit and to distinguish those works that do use it from the work done here.

The nature of the concent of power

The concept of power has been identified and used inter­ changeably with force, forces, influence, dominance, ability, and authority. We all seem to know intuitively that power exists and is important. Even with our common cultural definitions, we conjure up a sense of its centrality in understanding human behavior and organi­ zation. Yet there is a big step between feeling its importance, or proclaiming its importance, and using it as an analytical tool to demonstrate something about the operation of human behavior and organization. At the very least, to use it we must be able to measure it; before we can measure it, we must define it. 57 It is extremely difficult to come up with a definition that

everyone will accept. The more people attempt to define power, the

more complex the concept "becomes. It is one of those concepts that,

like time, everyone knows perfectly well what it is — until someone 22 asks them. Because it cannot be explained fully, perhaps, everyone

insists on their own definition — all of which have their own

limitations. Frankly, we think that most definitions grasp the

majority of the concept in one way or another. Most everyone is in the

ball park, at least to the degree that we can use the concept with a

confidence that measurement errors are at least as great as conceptual

differences. The best way to make this point is perhaps to present a

variety of definitions:

When the acts of an agent can actually or potentially modify the behavior of a person, or group of persons, the agent has power over that person or group of persons.23 Barry E. Collins and Harold Guetzkow

The power of a system is its influence potential, i.e., its poten­ tial to. lnduoe its will on other systems.2^" Johan Galtung

Power is a capacity to overcome part or all of the resistance, to introduce changes in the face of opposition,25 Amitai Etzioni

...power is latent force...2^ R. Bierstedt

A_ has power over B^ to the extent that he can get B to do something that B. would not otherwise do.2? Robert Dahl

Power, put simply and crudely, is the ability to prevail in con­ flict and to overcome obstacles.2® Karl Deutsch

Specifically, the definition that will be used in this study is that

power is the general capacity of an actor to control the behavior of

others. But the general characteristic of all these definitions is

the idea of a capacity, or capability, or ability of an actor to impose its will on another. 58

There are a variety of ways of splitting up the power di­

mension, but usually it is done on the basis of the source of the power.

What is it that an influencer uses to exercise influence? In terms of

relatively visible sources, this type of analysis produces such

classifications as economic, financial, industrial, political, and

military power. In terms of more abstract sources, there are a variety

of classifications. Weber's classification included legitimate, legal, 29 traditional, and charismatic power. Raven and French talk about re­

ward, coercive, referent, legitimate, and expert power.3^ Organski

identifies three methods of exercising power ~ rewards, punishments,

and force.Etzioni posits another three-fold classification — coer- 32 cive, utilitarian, and persuasive power.

These latter classifications in general recognize that the

concept of power in its most useful form is a mixture of coercion and

persuasion, i.e., of negative and positive influence. However, in

international relations there is a tendency to stress the coercive

aspect of power. This study does not. Rather is sees power as a means

of influence, a way of getting things done. It is neither good nor bad

in itself; it may be used or abused. Perhaps something in the nature

of power causes it to tend to be abused; perhaps "power c o r r u p t s . "33

Perhaps something in the nature of man, as Hobbes argued, requires coercion in order to produce a common action. But this should be determined after and not before this study. In this study we are not talking about coercive power any more than we are talking about utilitarian power. Moreover, while military forces are an important indicator of state power, they are not the only indicator or even the dominant indicator. It is difficult to overcome prejudices that deal 59 with "both the nature of man and the dominant processes of the inter­

national system. But it is essential to try. We envision power as

neutral, and we ask the reader to do the same.

Methods of measuring power

Power as such is never directly observable. What is viewed

and analyzed is either power as it is exercised (power as action) or

power in terms of resources (power as potential). There has been

quite a bit of debate as to which method is more proper. Actually,

most decision-makers appear to use both, particularly in the analysis

of the power of states.

Of course, in a basic sense all power may be considered to be potential, as a capability requiring motivation to be activated. We use the term "power as potential" because it is commonly associated with the technique of inferring power from resources. The justifi­ cation for judging power based on resources is that A (which possessed certain resources in an amount superior to 2.) has in the past been able to control the actions of 2.. Therefore, as long as A continues to possess a superior amount of those resources, it is likely to be more powerful than B at some future time. Resources become important in the estimation of power because they serve as a power base.

Before an actor can exert influence, it must have access to resources upon which it can draw. Specific resources that are often used to in­ fer state power include population, gross national product, military forces, urban population, industrial capacity, natural resources of a large variety, and territoryThese are commonly referred to as

"power capabilities."36 Again, however, too often power capabilities 6o

axe defined in terms of sanctions, threats, and destruction. Because

in any society certain things are widely valued, certain resources have

high applicability for influence through positive or utilitarian means.

It is a decided mistake to consider resources to he only "war potential." oQ The power base is in itself inert and passive. Its existence does not dictate its use. The analogy between money and power made by Talcott

Parsons and others, is particularly applicable in this sensed— both are based on resources and both can be used for good or evil.

The principal advantage of analyzing power in terms of resources

is that the resources are readily observable and easily measured. Thus, this tectinique of evaluating power lends itself to comparative analysis of both the intuitive and quantitative type. Its disadvantages stem from the difficulty of applying the measurement, that is from conceptual difficulties. In particular, we know little of the actual process of converting resources into the overt actions that influence behavior.

The conversion of resources into military force is one of the few areas that we do know something about — again serving to distort one's view by forcing one to see power in terms of dominance through coercion.

As has been mentioned, there is also a school which feels that power should be measured as it is exercised. This school believes that *0 behavior should be studied directly to determine who has power over whom.

Conceptually, as Robert Dahl, Nelson Polsby, and others have demon­ strated, this technique is appealing. A number of methods have been tried. One method is to equate the power of a social group with the ex-

/l t tent to which political decisions favor its interests. Other methods involve attempts to record and measure interactions or relationship ho patterns. These have been mildly popular with those dealing with power 6 1

concentration. S.J. Brams has attempted to deal with power concen­

tration in terms of the movement of government o fficials.^ Bruce

Russett has considered the number of decision-makers influenced by a

given decision-maker.*^ L.S. Shapley and Martin Shubik have used a

games-theoretic approach to evaluate the distribution of pcwer between

Congress and the President.^

This method of dealing with power eliminates some of the

problems of "power as potential," However, it creates a number of its

own. One of the greatest problems is the difficulty involved in the

selection of issues, as variable results have been shown to stem LA directly from issue selection. Similarly, there are few criteria

for differentiating rigorously between relevant and irrelevant inter- 47 actions. Also, though this technique has proven successful at a low

level of analysis under controlled conditions, its application to

interstate interactions in the international environment is extremely difficult. Finally, this technique generally does not lend itself to

comparison between a large variety or number of units.

The question of which of the two basic techniques to use is not totally academic. Scholars and laymen at all levels characterize people and states by class, and they make decisions based on those classifications. At the international level the frequency with which the terms "superpower," "major power," "middle-range power," and

"great power" (in addition to less flattering terms such aB "banana republic") come up in explaining policy decisions testifies to the significance of power rankings. Great powers are treated differently than middle powers. Despite formidable obstacles to measurement, the leaders of states must include an assessment of the power of numerous states in almost all their international dealings; and errors in

policy may result from errors in power analysis.

Often these classifications are little more than semi-

intuitive hunches.**'® The source of the classifications is difficult to identify. It appears, however, that both techniques of power estimation are used. Inference is drawn from past performance and resources. For instance, a major power almost always is required to have a substantial resource base. However, many states with large resource bases are not recognized as major powers until they have defeated another power.

We believe that a complete power analysis requires the use of both techniques, and in this study we have used indicators of both power as potential and power as action. Unfortunately, at the international level the latter do not lend themselves to quantitative techniques, and consequently we have used them only in discussion and evaluation. The bulk of our quantitative analysis is dependent upon power-as-potential techniques.

Measuring power from resources

For the reasons discussed, most of the studies that have attempted to quantify and compare state power have taken the power- as-potential route. Much of the theoretical work regarding which re­ sources or bases constitute "determinants" or "elements" of power was done in the 1950’5• Klaus Knorr, in broad terms, identified three determinants of state poweri "The Will to Fight, usually called

'morale'; Administrative Capacity; and Economic Capacity."*^ Raymond

Aron, again being general, singled out three also* milieu (space occupied), resources (available materials), and capacity for

collective action.Hans Morgenthau, being more specific, identi­

fied geography, natural resources, industrial capacity, military pre­

paredness, population, national character, national morale, quality

of diplomacy, and quality of government.'^' A.F.K. Organski selected

geography, resources, population, economic development, political

structure, and national morale. More recently scholars have tried to operationalize these

determinants, and they have had to make a series of trade-offs between

theoretical import and data availability. For instance, in his study

of power concentration, Bruce Russett used an index based solely on

population. Although he recognized that it is not a particularly good

indicator, his choice was made on the basis of the accuracy and avail­

ability of the data.-^ In a concentration study done by Singer, Bremer,

and Stuckey, six indicators were used: total population, urban pop­

ulation, energy consumption, iron and steel production, military ex­

penditure, and size of armed forces.-^ Others have also opted for a

multiple indicator index. Wayne Ferris initially used six: armed

forces personnel, defense expenditure, population, government revenue,

trade value, and area.'*'’ Michael Wallace in his status studies used five'**’ — all of which have been mentioned above — as did North and

Choucri in their study of the outbreak of World Weir 1 . ^ In Cox and Jacobson's stratification study five indicators were also used: gross

national product, gross national product per capita, population, nuclear capability, and prestige.-^®

It is clear that the researcher has a variety of indicators from which to choose. It is also clear that he has to decide whether (A

to use single or multiple indicators to construct his power index.

Most have chosen to use a multiple indicator index. The assumption

of the multiple indicator index is that the validity of the index is

increased "by the combination of several indicators. The multiple

indicators when combined are believed to cancel out distortions that

are bound to occur with a single indicator. Moreover, several power

theorists have warned against attributing too much importance to one

factor in estimating power.59

However, there are advantages of using a single indicator

index. It is easier to standardize, scale, and interpret than a z n multiple indicator index. Also, the combination of multiple indi­

cators, unless they can be weighted in accordance with their importance,

introduces a different type of distortion. The effects of the most important indicators are dampened by the less important.

We have chosen to use a two-indicator index and have attempted to select the two indicators which are most important in determining state power. Most modern analyses of the demographic foundations of national power have settled on gross national product as being the most important.^ Not only is it a homogeneous measure in itself, thereby facilitating comparison, but it is the grand result of all the productive forces at a state's command. However, gross national pro­ duct data is very limited historically, and for a good reason. A nation's total production was not considered a paramount indicator of 62 power by contemporary strategists until the middle of World War I.

Consequently, no one kept the necessary figures. They kept what they considered to be important, and one such figure was total government revenue. This is an indirect measure of productive capability, but 65 it also has the elements of both motivation and capacity for

collective action. In one respect, it is not the total production

that is important but that portion available for collective action.

For these reasons we have selected government revenue as one of our

two equally important indicators.

The second indicator that we have selected is defense ex­

penditure, The significance and theoretical importance of some

measure of military strength is so obvious that it needs little ex­

planation, Moreover, we feel that this military indicator is superior

to others that are available, such as troop strength, because it ad­

justs for national preference (naval forces vs. ground divisions,

for example) and facilitates comparison.

Although these two indicators can be combined into an index

that has the desirable qualities of being scalar, standardized, and reliable, there are certain difficulties associated with this index.

This is particularly true because the index.must be comparable not only between nations but over a span of 100 years. We are certain that there is some error. However, we are assuming that it is not appreciable enough to obscure the relationship that we are hoping to detect.

Other problems are involved which are more conceptual in nature. For instance, in our quantitative analysis we do not crank in a qualitative adjustment. Certainly some states make better use of their defense dollars than others. Also, we have made no effort to adjust for geography, to develop a loss of strength gradient. We have made no adjustment for the variation in the domain, range, or 66

scope of power. In addition, while we recognize that the signifi­

cance of an indicator varies with the purposes of the wielder of

power, with the specifics of the application, we do not attempt to

adjust for this. In all these areas, at least in the quantitative

analysis, we rely on randomization to minimize the error involved.

Measuring power from actions

As we have mentioned, power should not he conceived of totally

in terms of resources despite the fact that it is most easily

measured in that manner. At the least one should attempt to evaluate,

in terms of power as action, the believability of the relationships computed on the basis of power measured from resources. At the inter­ national level, the most visible form of power as action tends to be the results of war. Indeed, some people consider war results to be the most precise indicator of state power. ^ The action perspective thus tends to look at the application of power, which often surfaces as the use of force. Violence or force has in fact been defined by some "as the most direct form of power"^ or as "manifest power."^-5

However, except in cases of total war, it is very difficult to de­ termine how much of a state's actual power was devoted to a conflict.

One would be hesitant to state that North Vietnam ranked as a greater power than the United States on the basis of the final collapse of the South Vietnamese. Also, in the cases of world war, where this problem becomes less acute, other problems surface. For instance, world wars are fought by alliances, and dividing up each member's contribution to victory or responsibility for defeat is extremely difficult. The historical debate as to which ally was crucial in the defeat of Germany and its allies in World Wars I and II attest

to this difficulty.

The very visibility of war results probably causes it to be

overemphasized, however. War is not really a common phenomenon.

Most problems in the international arena are solved without recourse

to war. And these solutions have winners and losers. Consequently,

power as action might also be viewed with reference to peaceful or non-violent conflictual interactions. Unfortunately, the results of these interactions are difficult to grasp and comprehend. Perhaps the best approach is to observe how other states reacted to the state in question. If one does not consider this aspect, one's view of power will be unduely weighted in terms of the ability to get things accomplished with violence. For instance, prior to World War I, and perhaps to a Somewhat lesser degree prior to World War II, Germany was extremely limited in its ability to accomplish its goals through non-violent means in comparison to France or England. Yet someone concentrating on war capability and results might see Germany as the most powerful state in the world at this time. What is a factual determination? Certainly Germany was more powerful than its non­ violent diplomatic record would indicate, yet less powerful than its performance during the brief spasms of uncontrolled violence would indicate. The two perspectives thus need to be combined. Obviously, there is no easy solution to power evaluation. 68

The nature of the concents of concentration and dispersion

In many ways, the concepts of concentration and dispersion

are as hazy as that of power. Scholars feel that they know what it

is, yet it is very difficult to get an acceptable definition. The

basic element of concentration is minority control of some value or

valued entity. The basic element of dispersion is relatively equal

control. Yet the concepts are used in a variety of ways, and again

the best means of getting at the essential characteristics is once

again to present a variety of those uses.

For instance, S.J. Brams writes: "Clearly, what we mean by

use of the term 'concentrated' influence relationship is one

characterized by the exercise of minority control — that is, when

the 'few' influence the 'many'." Amos Hawley defines concentration

in a similar manner:

Any given amount may be in some instances concentrated in a small sector of the system or in other instances distributed more or less uniformly over all sectors or subsystems. ?

Collins and Guetzkow write, in reference to small group studies:

The distribution of power among individuals in decision-making groups...may vary from a concentration within one or two per­ sons to a wide spread of influence among many members of the group.68

Michael Haas states that: "Structures are concentrated... if they are located in a small geographic area within a larger system."^9

Whereas K.J. Holsti remarks that:

Sometimes a system's structure is typified by power being con­ centrated in one state...; in other eras, power may be diffused equally among a large number of states....?0

Gideon Rosenbluth refers to economic concentration as "the extent to which a small number of firms account for a large proportion of an industry*s output."71

It should he clear that concentration and dispersion, like

power, mean many things. However, a number of regularities appear.

First of all, we can view concentration as a continuum and refer to

more or less concentration rather than to concentration and dis­

persion. But most authors choose to refer to a low level of con­

centration as dispersion or equality. We have done the same.

Secondly, the essential element of minority control over a large

proportion of something is evident in almost all the usages. Finally,

the key distinction dividing the usages is whether minority is con­

ceived of in terms of numbers or percentages. One method of clarify­

ing the concept would be to define minority-in-terms-of-numbers as

concentration and minority-in-terms-of-percentage as inequality. Thus

inequality and concentration would be related but not identical.

Given a fixed number of actors, concentration would increase with

inequality; and given a degree of inequality, concentration would de- 72 crease with an increase in the number of actors. However, this

latter distinction is rarely made, particularly in international re­

lations literature. Most see the effects as being similar. Most

theorists conceive of concentration in both ways. In fact, most of •

the statistics for reducing a distribution to an index are in­ fluenced by both changes in the number of actors and changes in the

percentage distribution. Separating theoretical literature on this conceptual basis would be a gargantuan task. We have chosen to follow the majority and recognize the dual nature of the concentration- dispersion dimension. 70

If we were to compare systems with fixed numbers of units we

would have no problem. Unfortunately, both percentage distribution

and the number of units vary in the historical international system.

Comparison in such circumstances is extremely difficult conceptually.

Fortunately, this problem is eased slightly by the statistical in­

dicators, whose fluctuations are at leant predictable.

In addition, there are a variety of concepts besides in­

equality and equality which are related to concentration and dis­ persion. The most important of these theoretically are stratification

("difference between topdogs, middledogs, and u n d e r d o g s "73)F which is often classified as hierarchical or equalitarian; and centralization- decentralization ("centralization involves consolidation of power wielded within a relatively small number of stable organizational positions'^). These are particularly important because much of what is said about them can be and has been transferred to our dimension.

Often, in fact, these terms tend to be used interchangeably. For instance, Melvin Tumin writes that for all practical purposes stratification and inequality are synonymous. 7-5 Also, in their article, "Indices for Comparing Inequality," Alker and Russett write that "...the student of politics must be concerned with political 76 inequality, that is, with concentrated, unequally distributed power.".

There are a myriad of indices available to measure concentration and dispersion. It would not be difficult for any imaginative in- 77 vestigator to develop additional measures of his own. Fortunately, under most conditions most of these indices appear to be highly correlated with each other.7® 71 Three indices were chosen for the quantitative analysis

section in this work* concentration ratios (C), the Herfindahle-

Hirschmam index (HH), and the GINI index. There were two basic

criteria for selecting these. First of all, the indices had to he

in wide use. This eases the problems of interpretation and communi­

cation of results. Secondly, the different measures needed to

reflect a slightly different theoretical emphasis. For instance,

concentration ratios seem to measure monopoly power quite well, that

is, the amount of power that a small group of actors can bring to bear

on the rest of the system. The GINI index is sensitive to changes at

the bottom of the power spectrum as well as at the top. The HH index

is less sensitive to changes at the bottom and more sensitive to

changes at the top of the spectrum than the GINI. It is also less

sensitive to changes in the total number of actors than the GINI.

Consequently, if correlations changed substantially from index to in­

dex, one would be better able to determine the mechanism causing the

change, thereby facilitating the interpretation of results.

Concentration ratios are fixed point indices and can be criticized on the grounds that the index may remain unchanged even though the distribution (in terms of some sort of curve or plot) may change substantially. The concentration ratio is widely used in economic studies of monopoly power, however, because it gives a good measure of what a small group can bring to bear against the rest of the industry or market.^ a concentration ratio is a simple index — the ratio of the amount of some resource controlled by a fixed number of actors to the total system supply of that resource. But this very simplicity is of great value in interpreting results.

The other two measures are summary measures, that is, statis­

tically they take into account the slope of the entire distribution

curve. The first measure, the GINI, needs little explanation. The

index has "been widely used and publicized in economic and political

studies. The GINI index is based on the well-known Lorenze curve,

which measures the cumulative percentages of an attribute (such as

industrial output, income, and power) accounted for by various per­

centages of actors. The GINI index is a function of the area between

the Lorenze curve and the diagonal line the curve would follow if all 80 actors had an equal amount of the attribute. In extreme equality

or dispersion, the index will approach zero; in extreme concentration

it will approach 1-l/n. The advantages of the index have been widely

publicized. The greatest is perhaps that it seems to fit one's in­

tuitive idea of inequality or concentration. It measures what percent

of the actors control what percent of the resources. However, it has

some significant disadvantages as well. It is relatively insensitive

to changes in the distribution of the entity being measured. More

importantly to this work, it is rather sensitive to changes in the

number of actors. Because it is dependent on the percentage of actors,

a large increase or decrease in the number of actors, even if these

actors wield little power, will significantly affect the measure of

concentration.

Our second summary measure, the HH index, minimizes the im­ portance of the smaller actors. It is defined as the sum of the 81 squared percentages of all actors 73 Obviously, therefore, this measure is going to be more sensitive to changes in the larger nations than it is to changes in the smaller.

It is in many ways an elitist index.

Quite frankly, none of these measures is without advantages and disadvantages, When one tries to shift the number of actors and the shares of power in a game-like manner in order to compare changes in an index to one's intuitive or conceptual thoughts as to what is occurring, one realizes this limitation. Again, however, we are relying on one measure to offset, or at least alert us to, dis­ tortions in another.

The composition of the elite and total systems

It is not always clear in those theories which relate power distribution to war involvement whether the theorist is concerned with power distribution in an elite system of some sort or in the total system. For example, theorists who advocate the concert system do so on the basis of the advantages of equality among an elite group of four or five states, not of a general equality. In fact, some advocate a great disparity between those four or five states at the top and the rest of the system. Yet some theorists, particularly those who advocate great dispersion, will take the same theoretical benefits of equality and apply them to the total system. They are therefore critical of the elitist aspects of a concert system. In another area, some people feel that small states are reactors, and as such have little real influence on what occurs in the system.

Others, however, see the whole system as being important. They see small states as also exerting influence. Therefore, it is possible that two individuals could totally agree on the abstract effects of

power distribution on war and yet advocate entirely different policy

because one sees the total system as important whereas the other sees

an elite group as important. In order not to make a premature judgment,

we will analyze both the elite and the total systems.

The total system

The identification of the shifting and generally expanding

membership of the international system centers on determining which

political units constitute independent states at any one point in

time. Anyone familiar with the variety of political units that existed

prior to the world-wide movement toward independence following World

War II will appreciate this difficulty. Fortunately, Singer and Small

have solved this problem for us. Based on hard criteria, population

and diplomatic recognition, as well as some intuitive judgment, they have identified the composition of the state system from 1815 to 1970*

This information is contained in two studies — "The Composition and

Status Ordering of the International System i 1815-19^*0" and "The

Diplomatic Importance of States, 1816-1970." We will use their results.

The elite system

At all levels of theorizing there has been a tendency to con­ centrate not on the entire system structure but rather on the internal structure of some elite group. The rationale is that the few always govern or control the many, that powerful elites shape and control op virtually all societies. Those who hold this point of view believe that the crucial distinction between societies is the structure of those who shape and control. Although their ideal internal structures 75 are different in most cases, we can turn to Mosca, Michels, Marx*and

others for examples of the belief that there have always been and will

continue to be a few small sets of dominant, ruling elites.

This is certainly a dominant belief in international relations theory as well. For instance, Raymond Aron writes* "The structure of international systems is always oligopolistic. In each period the

Q£i principal actors have determined the system...."0^ Or, as Kingsley'

Davis expresses it*

Just as big trees control the ecological balance in a forest, so the powerful nations determine the pattern and trend of relations among all countries.®5

Identifying the possible candidates for the elite system is a much easier task than trying to identify the members of the total system. It is usually fairly clear who might be considered powerful enough to be included. The difficulty comes in making and justifying the selection. There are at least three ways to separate out the elite actors from the candidates. The first method involves the use of fixed number indicators. One can take an arbitrary figure, or one can choose this number on the basis of a natural break or separation point.

One might question the validity of this technique when trying to com­ pare systems with varying numbers of actors. We feel that it is valid, however, because there is a good deal of theoretical and empirical support for the proposition that the size of the elite remains rela- 86 tively fixed even as the system size fluctuates. In international relations literature, a number of authors have suggested that even though the number of states has risen from about thirty in 186.5 to 120 by the mid-1960*s, the number of great powers has remained roughly the same ~ somewhere about five. A second method is the use of a fixed percentage indicator.

One has two choices here. One could rank-order nations and then select

all those states whose combined power equals a certain percentage of

the total power, for example 50%« Or one could base the cutoff per­

centage on the power of the leading state; for example, all those

states whose power is 30% of that of the leading state. Both of these

techniques would of course vary the size of the system from period to

period to some degree.

A final method is to select the elite system based on his­

torical judgment. Singer, Bremer, and Stuckey chose this technique

in their study of power distribution and war involvement. There are

certain advantages to this technique, principally that often the his­

torian can add judgmental factors for which the dry quantitative

measurements do not account. The disadvantages are obvious, par­

ticularly the subjective nature of the determination.

We have chosen to use the fixed number technique, selecting

what appears to be the natural break point of five. This affords the

principal advantage of keeping the indices stable in terms of n size.

This eliminates a host of problems regarding their interpretation.

This choice does not imply that only five states are important. Five

is not a magical number. To understand what is occurring in the system,

one must also look at the internal indicators. For instance, in cases where the historian might see two great powers, the indices within the • elite system will show it to be highly concentrated; where the his­ torian might see eight powers, the internal indices will show the elite system to be dispersed. We will not disregard historical judgment in our quantitative analysis; but in keeping with the rest 77 of this study, we will use it principally to check our input and

interpret our results.

The Dependent Variable

Until now we have been admittedly somewhat vague about the

dependent variable of this study. Primarily this has been because

we have been presenting the theories of others when discussing war;

and to be quite sincere without being critical, in most of these

theories relating system power structure to war the exact nature of

the war predicted has been vague. Reference is usually made to such

terms as "peace" or "stability."

The concept of war appears to be relatively straightforward*

"War is organized violence carried on by political units against each

other.As the political units with which we are concerned are

states, we are interested in interstate war. The problem arises when

one attempts to measure or describe for purposes of comparison the degree of organized violence between these states. Among those who are attempting to be analytical, there seem to be two general ways of describing the degree of violence. If one is attempting to be analytical in the traditional school, one might talk about small wars or global wars. If one is attempting to be analytical in the behavior- alist school, one might talk in terms of the amount of war, fre­ quency of war, or magnitude of individual wars. There is a good reason for this. The traditional analyst talks of small and global wars because through those terms he is able to capture a concept or event and relate it to other events without having to quantify it and then adjust it statistically to make comparisons across time. He can classify wax on the basis of what people thought at the time, how

they reacted, what effect it had on the social, system, and what con­ temporaries compared it to at the time. Certainly there are stand­ ardization problems here, but they are no more perplexing than those produced by behavioralist methods. Whereas the traditionalist probably has a historical pattern of events in mind when he classi­ fies wars as small or global, the behavioralist most likely has a numerical indicator in mind (such as nation months at war or battle deaths) when he speaks of magnitude or amount. When the traditionalist compares wars he does it by interaction patterns which must be adjusted for the peculiarities of periods. When the behavioralist compares wars, he does it by statistics which must be adjusted by some quan­ titative factor. In both cases there is an element of judgment. Each method has its advantages and disadvantages, and we will use each where it is most appropriate.

However, in our theoretical and quantitative analysis we will principally be concerned with three measures of interstate wari magnitude, amount, and frequency. When we speak of magnitude we are referring to the size of an individual war. Magnitude is operation­ alized quantitatively in terms of two indicatorsj battle deaths inflicted in a particular war and nation months at war during a particular war. Thus, the average magnitude of the wars begun during a period would be the sum of the magnitudes of each war divided by the number of wars during the period.

By amount of war we mean the sum of the magnitudes of all the wars begun in the system during a particular period of time. This is what we refer to when we describe a period as producing more or less 79 war than another period. Amount of war is operationalized as either

the total battle deaths produced during a period or the total nation

months of war produced during a period. Amount of war and magnitude

are thus related, but they are very different measures. When we refer

to magnitude, we refer to an individual war; when we refer to amount,

we refer to the stun of the magnitudes of the wars that began during a

period of time.

By frequency of war we mean the total number of wars which were -produced during a period of time. To determine frequency one need only add up the number of wars that begem during the period of concern. The reader should note that it is possible for one period to have more wars (frequency) and yet less war (amount).

For instance, let us imagine two equal time periods, as depicted in Figure 3-1 below. The first period produced five wars, each of which involved 50,000 battle deaths. The second period pro­ duced two wars, one involving 10,000 battle deaths and the other in­ volving 2^0,000 battle deaths. The amount of war (in terms of this one indicator) in each period would be the same, i.e., 250,000 battle deaths. However, the first period produced more wars, i.e., it had a frequency of five wars while the second period had a frequency of two wars. Moreover, the magnitude of the individual wars in the first period would tend to be less than, in the second — the average magnitude being 5 0 ,00 0 battle deaths in the first vs. 125,000 battle deaths in the second. 80

Battle Period I Period II Deaths (ED) Magnitude of Magnitude of (Thousands) Wars 1-5 = 50,000 War 1 « 10,000 BD BD each War 2 = 240,000 BD 2 5 Amount of War Amount of War « 250,000 BD = 250,000 BD 2D Frequency of Frequency of War = 5 • War = 2 150-

100. War War War War War 12 3^5 50. War 1 0

Figure 3-1* Concepts of Magnitude, Amount, and Frequency of War Graphically Displayed

It is important that the reader have these concepts clearly in mind as we begin the next chapter, in which we will theoretically analyze the effects of concentration and dispersion of power on inter­ state war. NOTES TO CHAPTER THREE

Singer, "The Level-of-Analysis Problem in International Relations," World Politics, 14 (October, 1961), p. 77* o Bruce M. Russett, "A Macroscopic View of International Politics," in James Rosenau, Vincent Davis, and Maurice East, eds., The Analysis of International Politics (New Yorki The Free Press, 1972), p. 109.

-^Ronald A. Feldman, "Power Distribution, Integration, and Conformity in Small Groups," American Journal of Sociology. 79 (1973)» pp. 639-664•

^Michael Aiken, "The Distribution of Community Power1 Structural Bases and Social Consequences," in Paul E. Mott and Michael Aiken, eds. The Structure of Community Power (New Yorkj Random House, 1970), p. 488.

■^Terry N. Clark, Community Power and Policy Outputs 1 A Review of Urban Research (Beverly Hills: Sage Publications, 1973)» P« 53*

^Johan Galtung, "The Social Sciences: An Essay on Polarisation and Integration," in Klaus Knorr and James N. Rosenau, eds., Contending Approaches to International Politics (Princeton: Princeton University Press, 196977 P» 277. ^Robert Rothstein, "Alignment, Nonalignment, and Small Powers: 1945-1965t" International Organization. 20 (1966), p. 400; Richard Rosecrance, Action and Reaction in World Politics, pp. 6-7.

^Kenneth Waltz, Man, the State^and War: A Theoretical Analysis (New York: Columbia University Press, 195977” P« 12.

^Ibid., pp. 6-7.

James E. Harf, Bruce E. Moon, and John E. Thompson, "The Relevance of the International System for Explaining Foreign Policy," Paper delivered at the Midwest Political Science Association Annual Meeting, Chicago, April 26, 1974, p. 7«

HGeorge Liska, "Continuity and Change in International Systems," World Politics. 16 (October, 1963)» P* 128.

l^J.D. Singer, "The Level-of-Analysis Problem in International Relations," p. 80. 81 82

13as quoted in C. Wright Mills, "The Structure of Power in American Society," in Marvin Olsen, ed.. Power in Societies (New York: The Macmillan Company, 1970), p. 254.

l**Morton Kaplan, System and Process in International Relations (New York: John Wiley and Sons, 19^7), p. 5^« l^Morgenthau, p, I8 5.

l6Holsti, p. 116.

^^Wolfram Hanrieder, "The International System: Bipolar or Multibloc?" Journal of Conflict Resolution. 9 (September, 196.5)* pp. 30^-305; Osgood and Tucker, p. 170.

3*%or two studies which deal with numerical clustering see Singer and Small, "Alliance Aggregation and the Onset of War, 1815-1945," and Michael Wallace, "Alliance Polarization, Cross-Cutting, and Inter­ national. War, 1815-1964," Journal of Conflict Resolution, 17 (December, 1973)* PP. 578-581. -^Michael Haas, "International Subsystems: Stability and Polarity." American Political Science Review. 64 (March, 1970), P* 100. 20 Johan Galtung, "Small Group Theory and the Theory of Inter­ national Relations: A Study in Isomorphism," in Morton Kaplan, ed., New Approaches to International Relations (New York: St. Martin's Press, 19^8), p. 281; Brian Healy and Arthur Stein, "The Balance of Power in International History: Theory and Reality," Journal of Conflict Resolution. 17 (March, 1973)* p. 49! also see Andrew McFarland, Power and Leadership in Pi viral is t Systems, p. I89 for a discussion of the tendency to see dualist conflict in most circumstances. 21J,David Singer and Melvin Small, The Wages of War, 1816-19651 A Statistical Handbook (New York: John Wiley and Sons, 1972)* pp. 59- 75* Wayne H. Ferris, The Power Capabilities of Nation-States: Inter- national Conflict and War"(Lexington, Massachusetts: Lexington Books, 1973), P. 52. ^R. Bierstedt, "An Analysis of Social Power," American Sociological Review, 15 (December, 1950), P* 730.

2^Barry E. Collins and Harold Guetzkow, A Social Psychology of Group Processes for Decision Making (New York: John Wiley and Sons, 1964), p. 121 . ^Quoted in Berenice A. Carroll, "Peace Research: The Cult of Power." Journal of Conflict Resolution, 16 (December, 1972), p. 586. 25Amitai Etzioni, "Power as a Societal Force," in Marvin Olsen, e<*»» Power in Societies (New York: The Macmillan Company, 1970), P* 18. ^Bierstedt, p. 733* 2?Robert A. Dahl, "The Concept of Power," Behavioral Science, 2 (July, 1957), 202-203. 2®Karl Deutsch, "On the Concepts of Politics and Power," in James N. Rosenau, ed., International Politics and Foreign Policy* A Reader in Research and Theory (New York* The Free Press, 1969), p. 257.

2 9fiax Weber, "The Type of Authority and Imperative Coor­ dination," in Marvin Olsen, ed., Power in Societies (New York; The Macmillan Company, 1970), P* 36-38.

3°Referenced in Dorwin Cartwright, "Influence, Leadership, Control," in Roderick Bell, ed., Political Power (New Yorki The Free Press, 1969), p. 148; Ronald Feldman, p. 640.

31Organski, pp. 104-105. 32j2tzioni, "Power as a Societal Force," p. 24.

-^Robert S. Lynd, "Power in Society as Resource and Problem," in Marvin Olsen, ed., Power in Societies (New York* The Macmillan Company, 1970), p. 380.

R o b e r t Dahl, p. 202.

35carroll, p. 587.

3^Ferris, p. 6.

3?William A. Gamson, "Reputation and Resources in Community Politics," in Paul E, Mott and Michael Aiken, eds.. The Structure of Community Power (New York* Random House, 1970), p. 2&f.

38Dahl, p. 202 . 39Talcott Parsons, "The Monopoly of Force and the 'Power Bank', in Marvin Olsen, ed., Power in Societies (New York* The Macmillan Company, 1970) , p. 55* ^®Gamson, "Reputation and Resources in Community Politics," p. 267. ^Tord Hoivik, "Social Inequality — The Main Issues," Journal of Peace Research. No. 2 (1971), p. 134. ^Roderick Bell, "Political. Power* The Problem of Measurement, in Roderick Bell, ed., Political Power (New York* The Free Press, 1969), p. 19. 84

Brams, "Measuring the Concentration of Power in Political Systems," American Political Science Review. 62 (1968), p. 470. 44 Bruce M. Russett, "Probabilism and the Number of Units Affected: Measuring Influence Concentration," American Political Science Review. 62 (1968), p. 476.

^L.S. Shapley and Martin Shubik, "A Method for Evaluating the Distribution of Power in a Committee System," in Roderick Bell, ed., Political Power (New York: The Free Press, 1969), p. 210.

^Paul E. Mott and Michael Aiken, "Locating Centers of Power," in Michael Aiken and Paul E. Mott, eds., The Structure of Community Power (New York: Random House, 1970), p. 197.

^Bell, "Political Power: The Problem of Measurement," p. 19.

^ a v i d Vital, The Survival of Small States: Studies in Small Power/Great Power Conflict (London: Oxford University Press, 197l), p. 6.

^KLaus Knorr, The War Potential of Nations (Princeton: Princeton University Press, 1956), p.3*

■5°Aron, Peace and War, p. 577.

5lMorgenthau, p. 112.

-52Qrganski, pp. 116, 148.

53firuce M. Russett, "Is There a Long-Run Trend Toward Con­ centration in the International System?, " Comjiarative Political Studies, 1 (April, 1968), p. 104.

5^J.David Singer, Stuart Bremer, and John Stuckey, "Capability Distribution, Uncertainty, and Major Power War, 1820-1865," in Bruce M. Russett, ed., Peace. War, and Numbers (Beverly Hills, California: Sage Publications, 1972), pp. 19-48. ■55perris, p.

■^^Michael Wallace, War and Rank Among Nations (Lexington, Massachusetts: Lexington Books, 1973), P» 33« 57r.c. North and N. Choucri, "Background Conditions to the Outbreak of the First World War," Peace Research Society Papers. 9 (1968), p. 127.

■^Robert Cox and Harold R. Jacobson, "The Stratification of Power," in Robert Cox and Harold Jacobson, eds., The Anatomy of In­ fluence (New Haven: Yale University Press, 1973), pp. 437^+307

■59organski, p. 196; Morgenthau, p. 158, 85 6pBernard Barber, Social Stratification (New York* Harcourt, Brace and Company, 1957)» P* 171*

^•Kingsley Davis, "The Demographic Foundations of National Power," in Monroe Berger, T, Able, and C.H. Page, eds,, Freedom y d Control in Modem Society (New York* Van Nostrand, 1954),p. 20&; Knorr, Military Power and Potential, p. 4j?i Norman Alcock and Alan Newcombej "The Perception of National Power," Journal of Conflict Resolution. 14 (September, 1970), p. 342. ^Rosecrance, Action and Reaction in World Politics, pp. 146, 161-162 . ^3Knorr, The War Potential of Nations, p. 28.

64 h .L. Nieburg, "Violence, Law, and the Social Process," American Behavioral Scientist. 11 (March-April, 1968), p. 18.

65Peter Bachrach and Morton Baratz, "Decisions and Non­ decisions j An Analytical Framework," in Roderick Bell, ed., Political Power (New York; The Free Press, 1969), P* 103*

^Brams, p. 63. 67Amos H, Hawley, "Community Power and Urban Renewal Success," in Paxil E. Mott and Michael Aiken, eds., The Structure of Community Power (New York: Random House, 1970), p. 46,5.

£ 0 Collins and Guetzkow, p. 153*

69Michael Haas, "Societal Asymmetries and World Peace," Pro­ ceedings of the International Peace Research Association Conference. 2 (1969), p.“53- 70Holsti, p. 28. 7lGideaon Rosenbluth, "Measures of Concentration." Business Coneentration and Price Policy. National Bureau of Economic Research (Princeton* Princeton University Press, 1955), P* 57*

72rbid., pp. 61-62.

73fiosecrance, International Relations. p. 119.

7*4larvin Olsen, Power in Societies (New York* The Macmillan Company, 1970)» p. 229. 75rumin, p. 12.

^^Hayward Alker and Bruce Russett, "Indices for Comparing Inequality," in Richard Merritt and Stein Rokkan, eds., Comparing Nations * The Use of Quantitative Data in Cross-National Research

( 86

(New Haven: Yale University Press, 1966), p. 3^9*

??Rosenbluth, p. 58, ?®Alker and Russett, p. 3&7*

79John Perry Miller, "Measures of Monopoly Power and Concen­ tration: Their Economic Significance," in Business Concentration and Price Policy. National Bureau of Economic Research"(Princeton. Princeton University, 1955)* PP* 130-131*

®^Martin Bronfenbrenner, Income Distribution Theory (Chicago: Aldine-Atherton, 1971), pp. ^7“^8*

®^M.A. Adelman, "Comment on the 'Hf Concentration Measure," Review of Economics and Statistics. 51 (February, 19^9)* pp* 99-100.

^Kenneth Prewitt, "From the Many Are Chosen the Few," American Behavioral Scientist. 13 (November-December, 1969)* p. I6 9.

^^Bruce Mayhew, "System Size and Ruling Elites," American Sociological Review. 38 (August, 1973)* P* ^68; Olsen, p. 106. S^Aron, Peace and War, pp. 9^-95*

®-T)avis, p. 206 .

®^Mayhew, p. *+68.

®?Hedley Bull, "War and International Order," in Alan James, ed.. The Bases of International Order (London: Oxford University Press, 1973)* p.“llS7 CHAPTER IV

THEORETICAL ANALYSIS

The vast array of various types of acts which lead directly to

war may be grouped under the general heading of aggressive stimuli.

War may begin in the minds and emotions of men, as the popular cliche

goes, but those minds and emotions are influenced by specific events.

Also, the attitudes contained in those minds must be converted into

specific events before war can occur. Often it seems that there cire

few generalizations that can be made about these aggressive stimuli

which are transmitted between actors, except for the fact that they

are a necessary condition for war. Wars, after all, do not occur for

nothing; there must be some sort of dispute that leads to aggression

by at least one party. The sources of these disputes are many and varied, but these disputes seem to be always present in international

politics.

If we start with the assumption that these aggressive stimuli are the material from which wars are made, then there are two principal ways that the power structure might influence war. First of all, the power structure might affect the actual production of these aggressive stimuli. This will be referred to as the structural inducement of aggression. Of course, only a small number of these stimuli are con­ verted into wars. Therefore, the second major way whereby power 87 88

structure can significantly influence war is by affecting the manner

in which the aggressive stimuli are converted into actual wars. This

will be referred to as the structural modification of aggression. We

will first analyze the structural inducement of aggression and then

the structural modification of aggression.

The Structural Inducement of Aggression

It is possible to view the system structure as actually being

the source of, and hence inducing, certain types of aggressive stimuli.

This view generally presupposes a theory of motivation in which the

structure of the system significantly contributes to certain types of

motivation, which produce certain types of aggressive action.

Those who would advocate a dispersed power structure might tend

to see aggression as being motivated by concentration of power. Some might see this motivation based on the well-kncwn discontent-aggression mechanism, which suggests that the inequalities which are produced by concentration cause frustration and tend to lead low-rank and middle- rank actors to solutions through violence. Others might see concen­ tration of power as producing war because of a mechanism which suggests that in concentrated systems the most powerful actors are continually confronted with the temptation to attempt to eliminate each other and establish a condition of dominance. We will briefly look at each argument.

It is perhaps one of the basic beliefs of political and social science that frustration, despair, deprivation, lack-of-access to decision-makers, or any combination thereof, is at the root of social and political violence. This belief may be loosely termed the discontent-aggression hypothesis. At least four variations of this

hypothesis are currently receiving some attention by scholars. The

frustration-aggression hypothesis, the first variation, states that*

"Frustration imposed by external sanctions or the physical world,

generates a cumulative rage that, at some point, breaks through in

violent behavior. This hypothesis gets translated into the realm of

systemic analysis primarily through the assumption that frustration is most likely, to occur when the system is concentrated, that is, when

there is a disparity of some entity or value such as wealth or power.

The despair-aggression hypothesis, is a second variation. The classic formulation is the Marxist theory of revolution and its exhortation that the lower classes have nothing to lose but their chains. Sooner or later, out of despair, this class will rise up and abolish the in­ equalities of the system. Although this theory was originally con­ cerned with intrastate class struggle, the Marxist view of the origins 2 of interstate war is very similar. The third offshoot is the depri- vation-aggression hypothesis, which posits that violence is likely to occur when a group becomes aware of the fact that they are being deprived of something which they could obtain if the system were structured differently. The variations of this hypothesis--total 3 ^ deprivation-aggression, relative deprivation-aggression, and the 5 discrepancy theories — differ primarily in the location at which the aggression will occur. The final variation of the discontent-aggression hypothesis is the lack-of-access-aggression hypothesis, which rests on the proposition that the "less access to a remedy for its grievances a group has, the more violent it tends to become in demanding 90 such access."^ This hypothesis becomes important to this analysis

because of the assumption that the greater the concentration of power,

the less access will those at the lower end of the power scale have to

decision-makers•

Whereas the above hypothses posit that concentration of power would tend to produce a high incidence of aggression in low-rank and middle-rank actors, a struggle-for-dominance view of world politics would posit that concentration tends to produce a motivation for war among the great powers. That motivation is basically one of temptation/

In a period of high concentration, a victory by one powerful state over another powerful state would put the victorious state in a position of world dominance or near dominance. From this position the victorious state could end (supposedly) its fears and insecurity and could im­ pose some future good (in its view) on the rest of the system. In a dispersed period this temptation does not exist. If one state were to conquer another, the victorious state would not experience a period of security and fulfillment. It might become more powerful, but that very power would attract new enemies. Not only does the prospect and fear of dominance provide a rationale for initiating war, it also provides the motivation for more devastating wars. In wars for dominance, those who contemplate defeat see the fight as a last chance because they tend to believe that no counter is available in the future in cases of failure. Those who contemplate victory can exhort excessive sacrifice by the promise of future good. More sacrifice now for the sake of less sacrifice in the future is accepted (however foolishly) as reasonable and rational. 91 On the other hand, those who would advocate a concentrated power structure might tend to see aggression as being motivated by a dispersion of powers. Some might see this motivation based on a social control perspective, which holds that when power is dispersed it is impossible to establish clear norms and laws, to build habits or compliance, or to punish those who violate the norms and laws. The lack of law and law enforcement thus encourages aggression. Others might see dispersion producing war* because of the necessity of power­ ful actors to confront each other to satisfy their demands, rather than being able to rely on subordinate states. Again, we will briefly look at each argument.

All orderly societies require a body of understandings about proper behavior. Without such a conceptual network, there are no guidelines against which a desired course of action can be checked for O its acceptability. Initially these understandings take the form of 9 norms; eventually they may be codefied into laws. One might argue that norms and laws are less likely to become effective in a dispersed sys­ tem because of the difficulties involved in getting a large number of actors with very diverse interests to agree on anything. With such a group, the habits of behaving, cooperating, and at least respecting some decisions as binding, axe not likely to develop voluntarily. More- 10 over, these habits are the essence of order. Order also requires en­ forcement,'*''*' and a dispersed system does not allow for such en­ forcement because of the difficulty of organizing an adequate police force. Thus, because of a lack of norms or laws, compliance habits, and an enforcement agency, those who would disrupt the system for their own benefit axe encouraged to do so.

One might also argue that a dispersed system effects the pro­

duction of aggressive stimuli because in such a system the major states

are forced to turn against each other, rather than to weaker states, to 12 satisfy their external demands. If a major state requires something

from its external environment, then it can get that something from one

of two places— either from a peer or a subordinate. If the needy state

demands it from a peer, the needy state is likely to have to fight for

it; if that state demands it from a subordinate, the state will probably

be able to get it peacefully. All normative judgments aside, if major

states can make their demands and fulfill their wants at the expense

of smaller states, there will be a minimum of violence in the system.

If major states must make their demands upon each other, there will be the risk of world war. Thus, only in terms of reducing war, it is in the interests of the major states and the woxld as a whole that major states cooperate with each other at the expense of smaller states. This sort of cooperation is possible in a concentrated system, but in a dis­ persed system the number of equals at the top becomes too large for such collusion and the resource base at the bottom becomes inadequate to meet the demands of the major states. In a dispersed system the major states have nowhere to go but to each other.

The reason for presenting'the above four arguments— two claiming concentration produces motivation toward war and two claiming dispersion produces motivation toward war— is not to demonstrate or challenge the validity of any of them. There is not enough analytical work done on any of them to do this. Only the area of discontent-aggression has re­ ceived significant attention. The point is merely to suggest that there 93 axe grounds for assuming that all might be valid.

In fact, one need not view the above arguments as contrasting.

Rather they might be viewed as complementary. It is one thing to say

that one state of affairs (concentration or dispersion of power) pro­ duces a motivation for war and quite another thing to say that one state

of affairs produces more motivation than another. It could be argued that each state of affairs produces equally intense motivation for war; however, as the power structure of the system changes, so does the cause of the motivation. When the system is concentrated, the moti­ vation is produced by discontent and by the struggle for dominance. As the system becomes less concentrated, these factors become less im­ portant, and the motivation for war is produced by a lack of social control and a lack of the ability for collusion between the great powers.

But as has been mentioned, until more work has been done in each of the areas discussed above, this cannot be demonstrated. How­ ever, a determination of the effects of power on the production of aggressive stimuli is crucial to the remainder of the analysis in this study. Therefore, an assumption must be made. That assumption is that the power structure of the international system is the source of aggressive stimuli, but one power distribution does not produce more motivation for war than another. That is, the location at which aggression originates may change with changes in power structure, as might the motivation which produced the aggression, but overall the effect is the same.

This is, of course, an extremely important assumption, and for this reason we have gone into some length in explaining the thought process which led to it. One cannot consider the effects of power

distribution on war without considering human motivation. On the other

hand, scholars do not know enough as of yet about human motivation in

this area to demonstrate those effects conclusively. Therefore, the above assumption was made. Having made it, we are now able to proceed to the second area where one might see power distribution as affecting war, that is, to the processes and patterns of interaction whereby aggressive stimuli from all sources are converted into actual wars.

The Structural Modification of Aggression

The relationship with which we are concerned is this: power structure (l) modifies interaction patterns and processes (2) and these in turn have some effect on war (3 ),

We have already dealt with the concepts of power structure and war. The term "interaction patterns" is used to denote merely the configuration of interactions. Specifically, we are concerned with the possibilities for interaction and with the location within the system at which certain types of interaction are most likely to occur. The term "interaction processes" refers to a regular sequence of activity which manifests itself repeatedly in connection with a certain function.

Such processes include, for example, the formation of coalitions, the resolution of demands, the development of collective action, and the response to change.

This is not the first work to examine power structure, patterns and processes of interaction, and war. However, in this study a con­ clusion is reached which is substantially different than those conclusions arrived at in most previous work. The reasons for this

are two. First of all, rarely has the entire 1 2-^3 connection

"been made. There is a good deal of work connecting structure to inter­

action patterns and processes (l ~ > 2). The connection between inter­

action patterns and processes and war (2 -) 3) has also been made not

infrequently. On occasion some relatively normative works have even

jumped directly from structure to war (l -4 3)* But here we have

striven for a complete analysis (l -•y 2 3)r and subsequently have

reached a conclusion which is different than those reached in many other

works. Secondly, because of a felt need for a solution to war, for a

structural utopia, the dialectic method of looking at all aspects of a

problem has not often been used. Here we have looked in each instance

at the possibility of a tendency both toward and away from war. When both of the above procedures are followed— that is, a dialectic analysis across the entire 1 2 3 spectrum— a vast array of conflicting theory and data jells into a single, unified, internally consistent theory which seems to conform with the empirical world.

This conclusion is basically that the pattern of war slowly evolves with the structure of power. There is no structural utopia; the power structure of the international system does not tend to favor either more or less war. It does tend to favor certain patterns of war, but not more or.less war. Thus the system itself never wins, that is, the system as a whole does not benefit more from one power structure than another.

In the rest of this chapter we will go through the specific analysis which led to these beliefs. We will also present specific propositions as to the effects of structure on the amount and pattern

of war in the international system. Our sources are many and varied.

However, once we strip away the prescriptive aspects from each, they

tend to support each other almost totally. Those who have been looking

at one part of the problem have been generally correct in what they

have seen. But by seeing only some pieces, they have each been con­

vinced that they were right and others were wrong. However, when all

the pieces are viewed together, it is possible to combine them to

develop a consolidated theory. Indeed, one of the greatest reasons

that one might accept,for the most part, the theory presented here as

being correct is that it takes seemingly conflicting theories and pro­

duces a unified theory.

It is of course impossible to identify, discuss, and compare

all of the interaction patterns and processes of the international

system which would vary as the system fluctuates between conditions

of power concentration and dispersion. What we have done, therefore, is to identify roughly a dozen topical areas, such as the frequency of interaction, dependency-interdependency relationships, response to social change, etc., which have been connected most frequently to changes in both power structure and interstate war. For purposes of coherence these have been combined into three broad categories* general characteristics, politics of confrontation, and system regulation. In each area we will first center the analysis on the effects of power structure on processes and interaction patterns and then follow with discussion centered on the influence of these processes and patterns on war. Thus the remainder of this section will contain six subsections. 97 It is, however, very difficult to impose any classification scheme over a collection of thoughts on how the interstate system functions. This is true primarily "because all aspects of the system are interrelated. In our minds we are capable of seeing these re­ lationships in many dimensions at one time; we can see the web of interrelationships. Yet on paper we must impose some sort of linear order. We therefore expect the reader to perhaps disagree with our placement of certain patterns or processes under these categories.

We hope that there is less disagreement with what we say about the processes and patterns themselves.

General Characteristics: Complexity vs. Simplicity

A system in which power is dispersed is a complex system. A system in which power is concentrated is a simplistic system. This relationship between complexity and simplicity and the distribution of power has received wide publicity through two related concepts-- entropy and pluralism.

Entropy, originally a mechanical concept, has been adopted by social scientists to reference conditions of social order and disorder,Synonyms for high entropy would be "disorder,"

"messiness," or "distribution with maximum unpredictability." Low entropy could be expressed as "order," or "distribution with maximum predictability.

As entropy began as a mechanical concept, pluralism began as a decision-making concept. A low degree of pluralism in a society is characterized, according to Nelson Polsby, by high degrees of overlap in decision-making personnel among issue-areas, regularity in the

procedures of decision-making, and institutionalized "bases of power.

High degrees of pluralism are characterized, on the other hand, by

wide sharing of powers among leaders, constraints upon decision­

making imposed by elites and non-elites, conditions of all kinds

imposed by impersonal outside forces, and uncertainty about the dis­

tribution of pay-offs of political actions.^ The decision process

in pluralist systems tends to be "decentralized, fragmented, con­

strained, reversible, and relatively uncoordinated...."^ Because

decisions can be seen as power in action, it is not conceptually

difficult to convert these decision-making concepts for use in power

structure analysis. Andrew McFarland has done a superb job of this

in his book, Power and Leadership in Pluralist Systems. Pluralism

and elitism become almost synonymous with power dispersion and con­

centration. For instance, McFarland uses "pluralism" to refer to the

"dispersal of power among many elites... "^and he finds that those who see "concentrated power" see "power elites.

The concepts are not always synonymous. For instance, one might argue that although state power might be concentrated, the system might be highly pluralistic because of numerous non-state actors 1 9 or because of the presence of other authority besides state authority.

But in this study most of these exceptions can be put aside in light of the initial assumptions and definitions. Moreover, as Andrew

McFarland writes: "...one can find complexity and pluralism almost 20 anywhere, if one is sufficiently obstinate." The same can be said of elxtism. 99 Interestingly, in both the concepts of entropy and pluralism

we again run into the dual influence of numbers and percentage shares.

In regard to entropy, when we increase the number of actors, we in­

crease the "messiness" of the system. Similar effects occur when we

take concentrated, and thus "predictable" power and redistribute it.

To a high degree, pluralism also seems dependent upon both a large 21 number of component units and wide dispersal of power.

In general, however, as dispersion decreases, pluralism and

entropy decrease; as dispersion increases, pluralism and entropy in­

crease, In all three concepts (dispersion, pluralism, and entropy)

complexity and simplicity vary in the same direction, and for the same

reason. Moreover, the manifestations of this complexity or sim­

plicity that tend to be identified as significant are the same in most

systems. In international relations theory, those manifestations

most often cited are frequency of interaction, uncertainty or cer­

tainty in behavior, flexibility or inflexibility in behavior, and

alliance patterns. We will discuss each in detail.

Interaction frequency

If one envisions power dispersion resulting from increasing the number of actors, then it is clear that concentrated power systems have fewer interaction opportunities than dispersed systems. An in­ crease or decrease in the number of actors, increases or decreases the number of possible pairs of dyads in the system. The standard 22 formula for computing possible dyadic interactions is (n)(n-l)/2.

Thus, the number of interactions changes almost exponentially as the number of actors changes arithmetically. 100

Even if we keep the number of actors the same, however, one

can envision interaction complexity as varying with concentration and dispersion. As the system moves from concentration to dispersion, the

number of interactions between equals at the top of the spectrum in­ creases. This is perhaps far more important, for reasons which we will develop later, than the number of total interactions. Thus, if there are two equal actors at the top of the spectrum, such as in a bipolar system, there is only one dyad or pair possible. As power becomes more dispersed, such that there are perhaps five equal actors at the top (as in a concert system), there are ten possible pairings.

Certainty-Uncertainty

Uncertainty in its broadest sense refers to a situation in which a set of alternative outcomes is not fully predictable, either because of a lack of information or a lack of understanding. Cer­ tainty refers to a condition where outcomes are predictable. As concentration increases, certainty increases; as dispersion increases, uncertainty increases. This is a result of human incapacity in two areas. First of all, the ability of an actor to follow the action varies inversely with the complexity. Secondly, even if the action can be followed, the ability to predict the results also varies in­ versely with complexity.

This certainty-uncertainty dimension can be viewed in the first sense in terms of fluxuation in interaction possibilities. For instance,

Hans Morgenthau writes that;

The greater the number of active players, the greater the num­ ber of possible combinations and the greater also the uncertainty as to the combinations that will actually oppose each other and as 101 23 to the role the individual players will actually perform in them.

If one is speaking of concentration or dispersion strictly in terms of

shifting shares (i.e.r holding the number of actors constant), then one

has two choices. One can again view the interactions of equals at the

top of the spectrum and observe increased interaction opportunities with

dispersion of power. Or one can fall back to the concept of entropy to

understand changes in certainty. Where power is concentrated in a few

actors, it is easier to identify and follow. One can follow its move­

ment and predict possible collisions, at least between the largest .

actors. These calculations become less possible when that power is

divided among numerous bodies.

Others choose to view uncertainty in the second manner. For

instancej concentrating on the importance of alignments among major powers, Richard Rosecrance also sees increased uncertainty with dis­

persion, When power is concentrated (bipolar in his analysis) the

effects of a change in power alignment are relatively simple to predict.

Other actors are able to understand its effect and adjust for it. In dispersed systems, however, (multipolar in his analysis) a change in one of the major state*s position may be recognized, but its signifi- 2/f cance may be difficult to determine and adjustment may not be made.

Rigidity-Flexibility

The terms "flexibility" and "rigidity!' refer respectively to a capacity and incapacity for rapid modification of plans or actions. As the power structure moves from concentration to dispersion, the characteristics of interaction processes and patterns evidence less rigidity and more flexibility. In general terms, this can be envisioned 102

in two ways. First of all, in a high entropy, dispersed system 25 there are "no clear fronts in the total system." J Because of the

absence of clear threats and opportunities, strategies must change rapidly with rapidly changing circumstances. Strategies will also shift on the basis of short term rather than long term interests,

26 * because in such a situation long term interests are unclear. As power becomes concentrated, fronts become clear and strategies must be more rigid. Moreover, it is then possible to base strategies on long term interests. Secondly, one can envision the system in terms of the presence or absence of numerous cross-pressures which produce 27 overlapping interests and loyalties. As the system moves from concentration to dispersion, actors become less self-sufficient and dimensions of interaction develop. Thus, the more dispersed the system becomes, the greater the number of dimensions of interaction. As the dimensions of interaction increase, so do the alternatives available to accomplish goals. One might expect that rapid changes in tactics or strategy will be more common in situations with many alternative means than in situations with limited means. One is less likely to be un­ yielding when one's options are many than when they are few.

Alliance patterns

The discussion of flexibility and inflexibility often centers on alliance formation. There is some agreement that when power is concen- 28 trated, alliance structures tend to be rigid. In such a system, nations lose their freedom of action and are not able to shift from one alliance to another.^ Certain alignments become almost mandatory 30 and operate without regard to short term interests. 103 In dispersed systems, however, alliances become quite flexible,

primarily because they are derived from specific needs and short term 31 interests. The multiple interaction opportunities and uncertainty of

strategies make more permanent, rigid alliances unlikely. Alliances do

not become rigid because the interests underlying them tend to shift

quickly in a complex system. Historically we have seen evidence of this

in the descriptions of the flexibility of alignment which occurs in

"balance of power" systems, a term which usually is used to refer to the

quite dispersed multipolar system.^ To suggest, however, as many do,

that flexibility is essential for balance of power is to get the cart

before the horse. The process results from the structure and not vice

versa.

Influence of Complexity and Simplicity on War

A dispersed power structure will produce a set of complex in­

teraction patterns and processes which will, in turn, tend to produce

a complex pattern of war. A concentrated power structure will produce

simplistic interaction patterns and processes which will, in turn, tend

to produce a simplistic pattern of war.

As the number of interaction opportunities moves from few to

many, so does the possibility of conflict--of collisions between actors.

However, each collision is likely to produce less energy, to be less ^ destructive. A large number of interacting parties means that there is reduced danger of mutually supporting and reinforcing antagonisms between states, thereby limiting the size of conflicts. However, it also means that antagonisms are likely to be occurring somewhere in the system almost constantly. Conversely, a small number of interactions 104 limits the number of contests which can occur; however, when such contests occur the results are likely to be on a larger scale.

Similarly, there are no grounds to support the position that certainty or uncertainty influences the amount of war at the systemic level. Some might argue that a high level of certainty is at the root of war because outcomes are not so much in doubt. However, if outcomes are not in doubt, then one would expect the weaker actor to make necessary concessions rather than be destroyed. If calculations are more precise, then one would expect fewer mistakes turning into wars.

Qf course, while wars would be fewer, the consequences of each would be greater because it takes increased concentration of power to produce that certainty. The position of others might be that it is uncertainty that makes for war, because war is the result of misjudgment or erroneous perceptions. However, as we have seen, systemic uncertainty is the product of a dispersed power system. Mistakes may occur more often, but the consequences of each mistake are less. Thus no matter what position one takes, the systemic pattern is the same--concentration tends to produce certainty which tends to produce fewer wars; dispersion tends to produce uncertainty which produces more wars. However, the wars in a concentrated system tend to be of larger magnitude than those in a dispersed system.

In general, in a concentrated system certainty of measurement is crucial yet relatively easy to achieve. In a dispersed system certainty of measurement is less easy to achieve but also less im­ portant. The end result is not more ac less war, but a different pattern of war. 105 Finally, there is a good deal of evidence to make one doubt, despite a goodly number of propositions to the contrary, that an increase or decrease in flexibility (unless it is divorced from the power structure that should produce it) has any overall effect on the amount of war in the system. Theories relating power structure to flexibility, particularly of coalitions, abound in social science. Yet only in international relations do scholars connect flexibility or in­ flexibility to violence without regard to circumstances. Although there is little evidence to suggest that one pattern of coalitions produces more or less violence, there is a good deal of evidence to suggest that coalition patterns are tied to power structure. For instance, theories of structural balance predict flexibility or rigidity of coalitions in terms of power structure alone, without regard to outcome or goals de­ sired. Predictions made by Caplow, Gamson, Vinacke-Arkoff, and "game theory" all agree that any coalition is possible in a triad when the 33 power of A t= b = C. This, of course, is the condition of perfect power dispersion. Once power becomes less dispersed, this flexibility decreases.

A clear advantage or disadvantage of either flexibility or rigidity turns up only when they are separated from the conditions which produce them. In other words, this occurs only if we make the

2 — ^ 3 connection without regard to the 1 --^2 connection. For instance, flexibility in a dispersed system definitely allows cer­ tain conflicts to be by-passed. But the system is not any better off, because the same complexity which allowed the flexibility also intro­ duced more possible conflicts to_ be by-passed. One can conjure up 106

an advantage in one’s mind only if one takes the natural flexibility

of a complex system and theoretically transfers it to a simplistic

system. In a simplistic system, where there are fewer opportunities

for conflict, this ability to by-pass those few conflicts that do occur

appears to be an advantage. This is not valid, however. One cannot

transfer the flexibility selectively. If one is going to transfer the

flexibility to by-pass conflicts that one finds in a complex system to

a simplistic system, then one must also transfer the increased number

of conflicts that one finds in a complex system. If one does not, the

theoretical results are enticing; unfortunately they are invalid. Not

only are they invalid theoretically, but it would not work empirically

even if it could be done. Empirically, a flexible alliance pattern in

a system which is otherwise simplistic represents a discontinuity and

indicates that something is wrong with the system. As Morton Kaplan

recognized, in a concentrated system flexibility of alignment indi-

cates that considerable tension exists in the system. This indi­ cates instability rather than stability; war rather than peace. For

instance, if without an accompanying dispersal of power the alliance

system of the Cold War had begun to break up in the mid-19,50's > analysts would quite probably have predicted that something about the system was unnatural, that it was unstable. More concretely, the shift from a flexible to a rigid'alliance structure prior to World

War I (a period of dispersed power) has been identified as being one of the major causes of that war.^ This can be seen as an indicator that the system was in disequilibrium, for suddenly a dispersed power system had generated a rigid alliance pattern. The system in this case was, 107

of course, getting ready to go to war. The irritant causing the dis­

ruption was Germany, which through bad diplomacy and outright blunders

had managed to threaten almost every state in Europe. The cause was not

the rigid alliance pattern, but rattier an aggressive state acting as

though it had more power than it did. Rigidity was a symptom of an

unnatural state of affairs. Yet in a concentrated system, rigidity is

both natural and desirable. It makes calculations more accurate, which

is so essential when catastrophic war can be the result of miscal­

culation. Similarly, in a dispersed system rigidity is unnatural and undesirable.

Patterns and Processes of Influence: Demand and Response

Almost by definition, the units which make up any social system have certain areas of overlapping interests. Because of these interests, the actions of one actor affect the actions of another. Consequently, actors are constantly attempting to impose their will on other actors.

The characteristics of the patterns and processes of these influence attempts and the responses vary with the power structure of the system.

When power is perfectly dispersed in a system, all dyadic in­ teractions are between equals. As power becomes more concentrated, the number of interactions involving power discrepancies increases. This has some very definite effects on the pattern of demand and response which accompanies influence attempts. These effects are in addition to the fact that, as was developed earlier, the number of interactions is likely to increase as power moves from concentration to dispersion. All other things being equal, therefore, the number of demands is likely to increase. But the effects with which we are now concerned can best be 108

seen in the analysis of two aspects of demand and response— -compliance

vs. non-compliance and command vs. "bargaining.

Com-pliance vs. non-compliance

If interests are totally incompatible and a demand or influence attempt is made, the actor on the receiving end has basically two choices— compliance or non-compliance. Often these choices get translated into more normative terms, such as submission or defiance.3?

Certainly this decision is a result of many factors and calculations; but all other things being equal, the ability to non-comply requires that the actor on the receiving end of the influence attempt be roughly equal in strength to the actor initiating the attempt. This is cer­ tainly one of the basic propositions of international politics, particularly when it is formulated in terms of the use of force (an extreme attempt to influence). Quincy VJright states, for instance, that the "comparative equality of power would tend to augment the capacity of oQ each state to defend itself. Kenneth Walts comments that "the equality of states gives each of them the ability to fend for itself."39

Osgood and Tucker write that "the restraint of force requires an equilibrium of military power. ' This characteristic also extends to lesser interests and forms of influence. This relationship was even noted by Thucydides, who wrote that "into the discussion of human affairs the question of justice only enters where there is equal power to enforce it.

At the other end of the spectrum, it is reasonably clear that in most instances it is in the interests of the weak to avoid the application of superior force by the strong. Many potential wars have 109 "been avoided because weaker states have chosen the course of com- ij, 2 pliance.

Equality gives the capability to resist demands. Therefore,

in a dispersed power structure, where many interactions are between

equals, successful resistance is more possible than in a concentrated

system where fewer interactions are between equals.

Economic pressures follow the same pattern as the use of force.

In other words, asymmetrical relationships tend to outnumber symmentri-

cal relationships as systemic concentration increases. This leads to

patterns of dependence in concentrated systems and patterns of inter­

dependence in dispersed systems. When equal states trade, the relation­

ship tends to be one of interdependence, each state having equal

leverage on the other. When a small state enters into exchange agree­

ments with a large state, however, the small state is usually more vulnerable to and affected by a termination of those arrangements.

The small state is generally much more vulnerable to economic in­ fluence than the larger. Again, equality is necessary for resistance, all other things being equal.

We do not wish to oversimplify this relationship, because obviously small powers do resist larger powers. But at the same level of threat, each side being willing to devote the same relative resources into the fray, the smaller power tends to comply. For instance, one will find equals fighting over exchange agreements, commerce, and a variety of fairly small issues. Those instances where small states resist substantially larger states usually involve the national sur­ vival of the smaller state. On the less significant issues the smaller 110

states will tend to comply*

' Command vs* bargaining

Not all national interests are totally incompatible, requiring

a direct choice between compliance and non-compliance. In fact very

few are, although these few tend to be very dramatic. Most demands and

responses are characterized by a process the spectrum of which is

represented by bargaining at one end and command or authority at the

other. In concentrated systems, the process of command or authority

tends to dominate. The command may be subtle or blatent; it may allow

flexibility of response, or it may not. But it originates within the

powerful state and is transmitted to the less powerful. Whereas command

is a form of control by dominance, bargaining is a form of control by

reciprocity. Bargaining tends to be the dominant mode of demand and

response in pluralist, dispersed systems. Of course, there is a bar­

gaining dimension even in concentrated, hierarchical systems; just as

there is the element of authority in pluralist, dispersed systems. But

because of interdependence and the ability to resist authority, bar­

gaining becomes the dominant form of exchange in dispersed systems.

Whereas concentrated systems are characterized by politics of authority

and obedience, dispersed systems are characterized by politics of corn- 45 promise and balance.

Influence of Demand and Response on War

Here again it is easy to find a solution to the problem of war

if one is willing to extract one pattern or process from the totality

of the system which produced it. In general terms, for instance, many Ill

people see that numerous conflicts axe avoided in pluralist systems

by the give-and-take process of bargaining. They, therefore, conclude

that if they could transplant the bargaining process to a more concen­

trated system, it would serve to eliminate conflicts there as well.

However, they forget that the conditions which produced the tendency to

bargain, most particularly the frequency of non-compliance responses,

just do not exist in the concentrated system. Again, when we observe

the bargaining process in a concentrated system, it is usually an in­

dication that the system is breaking down. For instance, when a mili­

tary unit must bargain out a solution within its ranks, it is invariably

an indication of the ineffectiveness of that unit.

A hierarchical, concentrated system eliminates conflict through

an authority structure. In such a system voluntary habits of com- 46 pliance develop along with a feeling of legitimacy. Minor demands,

therefore, are not even viewed with the option of non-compliance in

mind. This is the model of most domestic law, which is obeyed in most

cases because of habits of compliance and not because of fear of en­

forcement. This observation leads many who see the international system as pluralist, with its many divisive confrontations, to suggest that

the more concentrated the system the more peaceful it becomes. Their reasoning is that when there is a strong central authority, many po­ tential conflicts will be avoidecl. Again their solution is invalid, because it separates the process from the conditions which produced it.

Empirically we find, in fact, that attempts to "restore order" to pluralist systems through the imposition of strong central power usually end in extreme failure. 112

As the power structure changes, the styles of influence also

change; as the styles change, the system must accept the had with the

good. The bargaining process does provide channels and machinery for

adjustment of interestsit does tend to introduce restraint and

moderation;**® it may in fact tend t o w a r d * * ^ an "exchange solution" rather

than a "threat solution," but it also has greater need for this adjust­

ment, moderation, and exchange. The capabilities of non-compliance and

the tendency to non-comply over relatively minor issues increase the

possibilities of violent conflict. Perhaps the number of violent con­

flicts that are avoided will be greater in a dispersed than a concentra­

ted system, but so will the number that are not avoided. These con­

flicts will be more frequent and will be fought over less significant

issues. For those same reasons, they will be of lesser magnitude than

those in a concentrated system.

The command process, which tends to accompany concentrated

systems, tends to reduce the number of conflicts by raising the level

of demands over which actors will go to war and by developing authority,

legitimacy, and habitual compliance. However, the equal nations who

have the ability to non-comply (at least those at the top of the

spectrum) tend to be more powerful in concentrated systems. Whereas

instances of non-compliance are fewer, the results are likely to be in­ dividually of greater magnitude.’ This effect is reinforced by the fact

that probability of conflict is inversely proportional to the expected 50 consequences. Thus the frequency of non-compliance in concentrated systems is less because of mechanical reasons and increased consequences; however, when that non-compliance does occur, the results are likely to 113 "be more catastrophic than in dispersed systems

System Regulation: Invisible Hand vs. Human Design

In the language of systems theory, regulation is "the process

by means of which a system attempts to maintain and preserve its

identity over time as it adapts to changing conditions."-^ More sim­

ply. regulation is the process whereby the system keeps itself operating

smoothly. When power is dispersed, systemic regulation is accomplished

by the more or less independent actions of numerous actors.-^ In a com­

petitive economic market this process is referred to as the "invisible

hand,"-^ In policy science the concept of "partisan mutual adjustment"

has been used to refLect the characteristics of the regulatory process in tjZi pluralist systems. In international relations, the "balance of power

process" (the "disorganized counterpoint, a clutter of independent pieces only loosely coordinated"--the "managed anarchy" where the "play

of chance is as marked as the designs of states")-’-' depicts this pro­ cess. Just as the "invisible hand” of the market place ideally requires the perfect competition of numerous, independent actors, so the

"balance of power" process ideally requires numerous, equal states.

Just as the "invisible hand" cannot function in a market where power is concentrated in the hands of a few firms, the "balance of power" process cannot operate effectively in a system where power is concen­ trated in the hand of a few states. In a concentrated economic system 56 regulation is handled through collusive agreements, and it is little different in a concentrated international power system. As power be­ comes more concentrated, more and more activities are influenced or 114

controlled by a small set of actors.^ In economics this state of

affairs may be referred to as monopoly or oligopoly; in political

and policy science the term "synoptic ideal" has gained some usage

in international relations "great power dominance" is most often used.

There is, of course, no tangible regulatory process that ob­

servers can directly see. Rather, we see manifestations of it, and

from these manifestations we tend to make judgments as to the de­

sirability or undesirability of a particular regulatory process. Two

such manifestations tend to be prominent in most theory-capacity for

collective action and response to change.

Capacity for collective action

In a simplistic sense, the greater the concentration of a sys­

tem, the greater its ability to mobilize collective goods. For

instance, in a bipolar system a global problem can probably be dealt

with if the two major powers agree to solve it. In all systems, many

things are possible when power is concentrated that would not be if it were dispersed.While power concentration is not a sufficient con­

dition for successful collective action, its absence tends to make

collective action extremely difficult at the global level. When power

is dispersed, some see a severe weakening of leadership.^ Actors tend 61 to resist each other, and the system may be effectively paralyzed.

On the other hand, a dispersed system appears to be able to

organize collective action on a wider range of interests. More issues

are likely to be dealt with, but they will receive attention at a lower level.^ Where power is dispersed, local or regional problems which would not reach global proportions will receive attention, but fewer 115

resources will "be available for their solution.

In community power studies there has been a debate over whether

concentrated or dispersed systems are more efficient in mobilizing

collective goods. Behavioralist studies have produced no clear results

either way. At the least the results are issue dependent. The same

sort of debate occurs in international relations, as we will discuss in

the next subsection dealing with regulation and war. What is not taken

into account in these debates is that each system mobilizes collective

goods differently. Concentrated systems have a greater capacity to

organize collective action if an issue reaches a global level of aware­

ness, and few issues will. Dispersed systems mobilize fewer resources 64 per issue but deal with more issues. In one case, a few issues re­

ceive effective attention while most go neglected; in the other, many

issues receive attention, while none of them are likely to be totally resolved. This can be conceived in terms of access to power. In a

concentrated system a great deal can be accomplished if an issue gains access to central power. In a dispersed system access to power is broadened, but effectiveness on each individual issue is reduced.^

Response to social change

In the discussion of collective action, we envisioned the system responding to some perceived need in a manner which would some­ how eliminate that need. In the discussion of social change, we con­ ceived of the system merely adapting itself to changes in the en­ vironment without trying to change that environment.

There is a feeling among many that dispersed, pluralist systems 66 have a quality of adaptability. Bruce Russett, for instance, has 1 1 6 written about future international politics that: "We must maintain a pluralist approach so as to be able to adapt to the unexpected...." {

This belief is often central to those who advocate the "balance of 68 power" or multipolar system. It is true that dispersed systems are more capable of quickly adapting to changing circumstances. However, such systems are also more subject to change; that is, they are in a constant state of flux, reacting to each small environmental change.

Whereas in a society of equals each minor disturbance is likely to ripple through them all, in a hierarchical society dis­ turbances are likely to be dampened. Concentrated systems are less subject to constant flux than dispersed systems. In dispersed systems, change is constant and comes in minor wavelets; in concentrated systems change is likely to be step-level and come in tidal waves. Neither is more capable of adapting than the other, they just do it differently.

Influence of Regulation on War

There is generally much more agreement about the effects of power structure on the processes of regulation than on whether those effects are good or bad. As one might suspect, international relations scholars tend to align themselves at one of two poles. Some see the dispersed system as more capable of regulating conflict, others the concentrated system. What both-sides fail to see is that the regulation process is different but in general no better or worse. Richard

Rosecrance, in particular, seems to recognize this. He sees each pole of power distribution possessing exactly the opposite advantages and 69 disadvantages in regard to conflict regulation. This is a position 117

that has been also realized by findings in both community power studies 70 and studies in organizational behavior. The advantages of each

structure reach a maximum under opposite sets of conditions. This can best be seen by developing the regulation debate in each of two areas— conflict solving capabilities and conflict absorption capabilities.

Some writers will observe an instance where major problems or crises were not resolved because a dispersed system could not get to­ gether and come up with clear, decisive action. He will, therefore, tend to praise the effects of concentration. Another scholar will observe crises occurring in concentrated systems because the great powers were unconcerned until a problem grew and turned into a major crisis. Such scholars tend to praise the virtues of a dispersed system. But as we have pointed out repeatedly, one cannot hope to solve the failures of one type of system by superimposing the processes which work well in another. Such a superimposition would be artificial and unnatural; and it would be rejected by the system itself. For instance, some scholars have held that the preference of the United

States and the Soviet Union to handle international matters solely between themselves— thus excluding the rest of the world’s states and 71 the United Nations— is unhealthy. They see this as collusion to the detriment of the interests of the rest of the world. They contend that the great powers do not address themselves to the real problems of the world, such as hunger, poverty, and development, because such problems do not concern them. The usual proposed solution is the greater use of the United Nations and its developing pluralist structure. Yet the system when concentrated will not accept this superimposition of 118

pluralist processes of problem solving. As Quincy Vlright remarked:

The doctrine of the politcal equality of states, attributing to each state equal weight in international conferences, has in the past proved a serious obstacle to international legislation.. .A state with a population of one hundred million will not recognize a state with a population of one million as entitled to equal political influence.72

The same sort of incompatability applies to those who attempt to use

the "authoritarian" or "police power" solutions to conflict resolution

in pluralist systems. The fact that these solutions work well in con­

centrated systems does not mean that they will work at all in dispersed

systems.

In fact, the form of the problem-solving body will fluctuate

with structure. In the bipolar world of the early 1950's summitry- gained popularity— two actors getting together to solve the world's problems. In the more dispersed world of the 19th century, congresses were used. Neither appears to have been more effective in problem solving. A concentrated system is more capable of dealing with world emergencies or crises, because it can mobilize great power rapidly. A dispersed system may be able to act only with great difficulty. However a dispersed system is likely to have a lesser number of major crises to deal with, because it is more able to recognize and deal with conflict issues at their origin when they are still small. Dispersed or pluralist systems, while less able to handle large crises, will see fewer of them.

Most crises are dealt with before they get out of hand and reach global proportions.

In the area of response to change, we find again predictions of more or less war depending on system structure. The debate is 119 generally centered on the question of social order, which is usually defined more or less in terms of the amount of conflict in the system.

One school holds that there is greater order in a dispersed system be­ cause the system can adjust to change. Minor adjustments can be made by all actors, and peaceful solutions can be found to problems. This school contends that in a concentrated system minor adjustments cannot be made. Change is resisted until an issue explodes with global con­ sequences. On the other side, scholars claim that concentrated sys­ tems have the ability to absorb change within the structure of the system.^ The system is not thus marked by constant fluctuation.

Constant fluctuation is seen as producing constant conflicts of interest which in turn produce constant war.

Moth schools are partially correct. Dispersed systems are marked by constant fluctuation and constant war. But the wars tend to be of lesser magnitude. Every change in environment is felt and some sort of adjustment must be made. Consequently, the likelihood of dis­ pute increases. But the disputes are likely to themselves be small- fought by relatively smaller nations for more limited goals than in concentrated systems. The concentrated system has fewer wars, but when they come they are of greater magnitude. The system can absorb change without adjusting to it, thus providing for long periods of peace.

Change tends to be step-level, and when a system takes the step, the conflict that results is likely to be between relatively larger nations for relatively greater goals. 120

Conclusions and Hypotheses

One might now feel that this analysis has "been much too long

in light of the simplicity of the conclusion that the system as a whole does not benefit more from one power structure than another.

Overall, the amount of war in the system is not dependent upon the concentration or dispersion of power. The pattern of war is so de­ pendent, however. As concentration in the system increases, the frequency of war will decrease, but the magnitude of individual con­ flicts will tend to increase. As concentration decreases, the fre­ quency of war will increase. but the magnitude of individual con­ flicts will tend to decrease.

While these conclusions seem logical and simplistic, it is evident that they have not been recognized by many people concerned with international relations. This is true primarily not because of a lack of clarity of thought but rather because of the search for a solution to the problem of war in terms of power distribution that just does not exist.

We do not mean to assert that the total sum of war in the world is at all times constant. The amount of war is subject to, at the very least, random distribution as well as human free will. Our conclusions refer only to the structural factor of power distribution, i and our conclusions are, of course, formulated in terms of probability*

Consequently, the comparison of our theory to empirical data is not likely to be clear-cut. We will concern ourselves with this comparison in the next chapter. NOTES TO CHAPTER POUR

^H.L. Nieburg, Political Violence: The Behavioral Process (New Yorks St. Martin's Press, 19^8), pp. 40»3Fl.

^Karel Kara, "On the Marxist Theory of War and Peace," Joumal of Peace Research, no. 1 (1968), pp. 2,4,

^Galtung, "A Structural Theory of Aggression," J oumal of Peace Research, no. 2 (1964), p. 9 8. Galtung discusses a number of theories which focus on the complete underdog. ^Derek Birrell, "Relative Deprivation as a Factor in Conflict in Northern Ireland," The Sociological Review. 20 (August, 1972), p. 317; Ted R. Gurr, "Psychological Factors of Civil Violence," in Ivo K. Feierabend et al., eds., Anger, Violence, and Politics (Englewood Cliffs, New Jersey: Prentice Hall, 1972), pp. 37-38. c Ivo K. Feierabend and Rosalind L. Feierabend, "Systemic Conditions of Political. Aggression: An Application of Frustration Aggression Theory," in Ivo K. Feierabend et al,, eds., Anger, Violence, and Politics (Englewood Cliffs, New Jersey: Prentice Hall, 197^), p. 137; Wallace, War and Rank Among Nations, pp. 6-7.

^Nieburg, Political Violence, p. 40.

^This is a view held by many of the earliest balance-of-power theorists (see Gulick, pp. 38-60) as well as by modern theorists (see Sterling, pp. 87, 559) who advocate dispersed systems. D Alan James, "law and Order in International Society," in Alan James, ed., The Bases of International Order (London: Oxford University Press, 1973)* P-"~^5-

%elio Jaquaribe, "World Order, Rationality, and Socioeconomic Development," Daedalus, 95 (Spring, 1966), p. 607,

•^^Deutsch, "On the Concepts of Politics and Power," p. 256.

^Ibid.j Osgood and Tucker, p. 125 , 12por a real-life picture of this view in operation, see the discussion between Stalin and Churchill recorded in Forrest Pogue's George C. Marshall: Organizer of Victory (New York: The Viking Press, 1974)7 P. 525. 121 •^^Nieburg, Political Violence, p. 3 3. -*Ajohan Galtung, "Entropy and the General Theory of Peace," Paper delivered at the second IPRA General Conference, Sweden, summer 1967, p. 2.

15Nelson Polsby, Community Power and Political Theory (New Haven: Yale University Press, 1963), pp. 119, 13^*

l6Ibid., p. 132 .

■^McFarland, p. 1.

^®Ibid., p. 2k»

■^Edward Buehrig, "The International Pattern of Authority," World Politics. 17 (April, 1965), p. 370.

^°McFarland, p. 53*

21Ibid., pp. 18-19.

22Deutsch and Singer, p. 318.

^^Morgenthau, p. 3^3* 2if Rosecrance, International Relations, pp. 117-118.

2^Johan Galtung, "On the Future of the International System," Journal of Peace Research, no. U (1967), P» 3°7* 26 Kaplan, "Intervention in Internal War: Some Systemic Sources p. 95. ^McFarland, p. 190.

2®Alan Dowty, "Conflict in War-Potential Politics: An Approach to Historical Macroanalysis," Peace Research Society Papers. (1969), p.

2^Louis Rene Beres, "Bipolarity, Multipolarity, and the Tragedy of the Commons," Western Political Quarterly, 26 (December, 1973), p. 632j Organski, p. 31^*

■^Richard A. Falk, "Zone II as a World Order Construct," in James Rosenau, Vincent Davis, and Maurice East, eds., The Analysis of International Politics (New York: The Free Press, 1972), p. 189.

•^Holsti, pp. 92, 110. 123

Kenneth Waltz, "International Structure, National Force, and the Balance of Power," Journal of International Affairs. 21 (1967), p. 216; Martin Wright, p. 104. 33jer0me M. Ghertkoff, "Sociopsychological Theories and Re­ search on Coalition Formation," in Sven Groennings, E.W, Kelly, and Michael Leiserson, eds., The Study of Coalition Behavior (New York* Holt, Rinehart, and Winston, 1970)* P» 298. ^Kaplan, System and Process in International Relations, p. 35*

■^Rosecrance, Action and Reaction in WorId Politics, p. 181; Waltz, Man, the State^ and War, p. 218.

^ G a l t u n g , "A Structural Theory of Aggression," p. 102. ^Kenneth E. Boulding, "Toward a Pure Theory of Threat Systems," in Roderick Bell, ed., Political Power (New York* The Free Press, 1969), pp. 287”288 .

-^Quincy Wright, p. 755*

^Waltz, "International Structure, National Force, and the Balance of Power," p. 228.

^°0sgood and Tucker, p. 30.

^Quoted in Waltz, Man. the State, and War, p. 211. ^Vital, The Survival of Small States, p. 3; Ferris, p. xxi.

vital. The Inequality of States, pp. 88-89* 117*

^^McFarland, pp. 40, **2-4-3.

^■^William E. Connolly, "The Challenge to Pluralist Theory," in William E. Connolly, ed., The Bias of Pluralism (New York* Atherton Press, 1969)» P* 13* ^Deutsch, "On the Concepts of Politics and Power," p. 255*

^Quincy Wright, pp. 850-85I; Nieburg, Political Violence, p. 133. ^Buchan, p. 58• ^9l3oulding, "Toward a Pure Theory of Threat Systems," p. 285. ^Rosecrance, International Relations, p. 118.

■^Kaplan, System and Process in International Relations, p. 89.

^McFarland, p. 219. ■^Olsen, p. 186. Clf. ^Charles Lindblom, The Intelligence of Democracy t Decision Making Through Mutual Adjustment (New York: The Free Press, 1965) * p.3* 55 Claude, Power and International Relations, p. 79*

-^Michael Gort, "Analysis of Stability and Change in Market Shares." Journal of Political Economy, 71 (February, 1963)* p. 55•

Olsen, p. 228, •5®Lindblom, The Intelligence of Democracy, p. 1*H.

^Edward Banfield, "Concerting Action by Influence," in Roderic Dell, ed., Political Power (New York: The Free Press, 1989)* p« 333* William A. Gamson, Power and Discontent (Homewood, Illinois: The Dorsey Press, 1968), p. 13*

^°Arthur Kornhauser, "'Power Elite' or 'Veto Groups'?" in Marvin Olsen, ed., Power in Societies (New York: The Macmillan Company, 1970), p. 288.

Olsen, p. 186; Etzioni, "Power as a Societal Force," p. 22. 62 Ralph Dakin, "Variations in Power Structures and Organizing Efficiency:. A Comparative Study in Four Areas," Sociological Quarterly. (July, 1962), p. 231 .

^Aiken, "The Distribution of Community Power: Structural Bases and Social Consequences," pp. 506-508. 64- E.J. deKadt, "Conflict and Power in Society," International Social Science Journal. 17 (1965)* P* **58. 65 Paul E. Mott, "Configurations of Power," in Michael Aiken and Paul E. Mott, eds., The Structure of Community Power (New York: Random House, 1970), p. 8 6. 66 . ,, Neiburg, Political Violence, p. 51* PP* 160-166.

^Bruce M. Hussett, "The Ecology of Future International Poli­ tics," International Studies Quarterly, 11 (March, 1967)* PP* 30-31*

^Deutsch and Singer, p. 322. ^Rosecrance, International Relations, p. 116.

^°Komhauser, p. 289; Daniel Katz and Robert L. Kahn, The Social Psychology of Organizations (New York: John Wiley and Sons, p. ^Chadwick Alger, "The United States in the U.N.," International Organization, 27 (Winter, 1973)* P» 3»

"^Quincy Wright, p. 9^6,

^^V/altz, "The Stability of a Bipolar World," p. 886; Osgood and Tucker, p. 170. CHAPTER V

QUANTITATIVE DESCRIPTION AND ORIGINAL EVIDENCE,

1862- 1964

At the end of Chapter I, we stated that this work had two ob­

jectives. The last chapter was devoted to one of them— the building

of a theory which explains the effects of the power structure of the international system on interstate war. This chapter is devoted to the other objective— the gathering of original historical evidence based on the period from 1862 to 1964-. A biproduct of this gathering of evidence is a quantitative description of the international power structure for roughly a century. Host of the description is presented in the first section, which deals with the independent variable— the system power structure. The second section provides the reader the data pertaining to the dependent variable— interstate war. The original evidence is presented in the third and final section, where we analyze the relationship between the independent and dependent variables.

International Power Distribution. 1862-1964-

The principal source of data for determining the international 1 power structure is the Arthur Bank's Cross-Polity Time-Series Data.

Data taken directly from this collection accounted for 87% of the power index in terms of value and 77% of the total number of pieces

126 2 of data. In order to fill in some crucial data missing from the

Bank's collection, particularly the figures on defense expenditure and government revenue for some of the larger states, the data set collected by Wayne Ferris for his book The Power Capabilities of

Nation-States was used. This source accounted for 11% of the power index in terms of value and 5% of the total number of pieces of data.3

After combining these two sources, we were left without hard data for 2% of the estimated overall power and for 18% of the number it of pieces of data. However, we were able to compute data for these initially missing cases on the basis of J. David Singer and Melvin

Small's previously mentioned studies—I’The Composition and Status

Ordering of the International System" and "The Diplomatic Importance of States, 1816-1970". Singer and Small rank states in these studies according to indicators of status or perceived power. Based on these rankings we have given a state with missing data the average of the available hard data (Banks and Ferris) for the two states above and below it. This procedure was slightly modified in some cases where a distortion would obviously have been produced,^ Also, if the value of one indicator was missing and the other available, the missing indicator was given a value equivalent to that of the available in­ dicator. As a result of these procedures, we were left with no missing data. These procedures may have introduced some distortion, but we feel that it is far less than would have occurred if the data had been left missing.

We chose to measure the structure of the system at twenty points in time which were at roughly five year intervals, beginning in 186** and ending in i960. There were two deviations. Because we 128

did not want to take a measurement in the middle of World War XI, a

measurement was taken at 19^*6 instead of 19*t4. To preclude only a

three year interval between 19^6 and the next regular measurement

point, 19^9, we took the next measurement at 1950 and then continued

the five year intervals to i960. This spacing of measurement points

also allowed us to follow rather closely the works by Singer and

Small which defined the membership of the system.

Occasionally data was not available for the year at which we

took the measurements. We desired, however, to avoid the distortion

produced by missing data. Therefore, if data was available within two

years of the desired measurement year, it was accepted without modi­

fication. For instance, no data was available for the year 191^-. There

was, however, data available for 1913* Therefore, the 1913 data was

accepted and applied to the 191^ system of states as determined by

Singer and Small. Likewise, at some of the measurement pointb we did not have the exact system membership. Specifically, we desired to use the years 1919t 1924-, 1929» 193^» and 1939 as measurement years.

However, Singer and Small developed the membership of the state system at the years 1920, 1925» 1930, 1935* and 19^0. Therefore the member­ ship of the system as determined by Singer and Small was accepted as being valid for the preceding year as well. Also, Singer and Small did not develop the composition of the state system for either 19^5 or 19*t6. Therefore, to establish the 19*^6 membership we took the 1950 membership and eliminated those states which came into existence be­ tween 19^7 and 1950. 129 Having gathered the "basic data for each of the measurement

points, the next task was to convert the raw data into a standard

power index so that the various measurement points could "be compared.

We decided to base our index on 1000 points. We also chose to weigh

equally each of our two indicators of state power. Consequently, we

totalled separately the values of both defense expenditure and government revenue for each measurement year. The specific value of

each state was then divided by the total value for that year and the result multiplied by 50°• The two scores for each state were then 6 added together to measure each state's relative power. The sum of all the states for any measurement year was obviously 1000 points.

Thus the structure of one measurement year could be directly compared to the structure of the other years. For instance, one can compare the power of the most powerful state in 1864 (157 points or 15*7^ of the world total) with the top state in i960 (430 points or 43. C$ of the world total) without having to guess at the significance of growth factors which make the raw data difficult to interpret. Because of this ease of comparability, one can follow the changes in the system power structure relatively easily. In the next two subsections of this chapter, we will present these changes for both the total and the elite system. To do so we will use combined indices, such as the GINI, as well as more graphic forms of display.

Total System Power Structure

Table 5-1 on the next page presents a- description of the total system power structure in terms of three indices of concentration- 130

Table 5“1 J Power Structure of the Total System, 1864- through i960, in Terms of Concentration-Dispersion Indices

Year GIN I IIH Index Index Inf?ex

1864 D* .7779 D .0992 D .6415 1869 D .7329 D .0908 D .6055 18?4 'D .7113 D .0889 D .5682 1879 D .7399 D .0977 D .6071 1884 D .7275 D .0925 D .5937 1889 D .7374 D .0858 I) • 5756 1894 D .7618 D .0966 D .6269 1899 D .7935 D .0999 C .6550 1904 C .8009 G .1006 D .6394 1909 D .7797 D .0902 D .5968 1914 D .7794 D .0930 D .5932 1919 G .8913 C .1932 G .8093 1924 C .8313 G .1170 G .6616 1929 C .8425 C .1426 C .6762 1934 C .8835 G .1508 C .7194 1939 G .8771 C .1378 G .7331 1946 G .9354 G .3109 C .8802 1950 C .8963 C .2152 G .8119 1955 C .9017 G .2398 G .7905 i960 C .8804 C .238 O G .7832

«

*C or D to the left of a number indicates that when the en­ tire period was dichotomized according to that index, the particular year was in the more Concentrated (C) or more dispersed (D) half• 131 dispersion: the GINI index, the HH index, and the concentration ratio of the power of the top five states to the total system power (C5). We have also dichotomized the twenty measurement points into the ten most concentrated and the ten most dispersed, and in­ dicated this classification "by placing either a C or a D at the left of the index. The reader should look at this table before continuing.

To one who is not extremely familiar with the indices, Table

5-1 probably seems very dry. Hopefully, the structure of the system will become more vivid in the stratification drawings which will be presented shortly. However, before presenting these, we need to note some of the significant aspects of the system structure which this table portrays. To begin, one might notice that all three indices are highly correlated. Table 5-2 below lists the Pearson correlation co­ efficients for the three indices.

Table 5-2: Pearson Correlation of Concentration-Dispersion Indices within the Total System

GIIII HH C5 0 0 « GINI 1.000 .872 \

HH .872 1.000 .9^2

°5 .950 .9^2 1.000 The lowest correlation (.872), between the GINI and the HH indices,

is obviously quite high. The degree to which all of the indices are

correlated can also be seen by looking at the dichotomization

(Table 5-1). The ten periods of concentration and the ten of dis­

persion are exactly alike based on the GINI and HH indices. Based on the Cjj index there is only one change. Given the theoretical difference between the indices, we are somewhat surprised to find such a high correlation between them. In Chapter IV we presented the GINI as being the egalitarian measure because the small state gets represented proportionally to its power. We presented the HH index as being elitist because the percentages are squared and consequently the small state is barely considered. The Ctj measure was presented as being a measure of monopoly power. Consequently, we would expect that the least correlation would be between the GINI and the HH indices, which is the case. But we would not expect this correlation to be that high. The explanation for the high correlation among all of the in­ dices probably rests on the fact that during the entire period the majority of the total power of the system remained substantially con­ centrated in the hands of a few actors. The lowest GINI measure is a substantial .711* At the most dispersed point, based on the concen­ tration ratio, the top five states controlled a whopping 5^.£^ of the t world's power.

Moreover, in terms of the overall system, there was not a great deal of change across the entire period. The GINI fluctuates across a range of only .224- and the C^ fluctuates only .312. Where the real variance does occur is in the elite system, among the top few states. 133 This will "be clearer later when we describe the elite system. Be­

cause so little power is controlled by those states at the bottom of

the spectrum, they impact very little on any of the measures of con-

centration-dispersion. Thus the egalitarian GINI, the elitist HH,

and the monopoly C*j all correlate highly.

This fact, that the total system has been highly concentrated during the entire period, can be easily seen in the stratification drawings on the next five pages. We have selected five evenly spaced measurement points (186**, 1884> 1909» 193^» and 19^0 ) and have drawn two figures for each. In the first figure of each year the drawing depicts the number of nations which, when rank-ordered, are required to make up each one-fifth of the total system power. For example, in l86*f (see Figure 5-1) all of the greatest state's power plus 29# of the second state's power was required to make up 20# of the world's power.

However, it took the last 31*28 states, when rank-ordered, to make up another 20#. The second figure of each year depicts the number of states whose power was greater than 80#, 60#, 40#, 20#, and less than

2Q# of the power of the greatest state. For example, in 186*1 (see figure 5-2) the top state had a score of 156.76 points or 1 5*7# of the world's power. Four states (including the top state) had power scores greater than 125.*KL, in other words, greater than 80# (or within 2Q#) of the top state. Thirty states had scores of less than 31*25* in other words, less than 20# of the top power.

These figures are not as comparable as Eire the indices listed in Table 5-1, and there are only five measurement points depicted and not twenty; but the figures do hopefully leave a more vivid impression Number % of Total of States System Power 100

29 N =* 39 80 States 1.'44 60 1.72 40 3.2? 20 31.28 0

Figure 5-1: Total System Stratification— The Power Pyramid, 1864

State Power Number as a % of the of States Greatest State 4 Z 80 0 Z 60, ^80 2 > 40; <60 £ 20, < 4 0 3

30 <20

Figure 5*2 1 Total System Stratification Based on the Most J'owerful State, 1864 135

Number % of Total of States System Power

100 1.1? States 80 1 .6 5 60 2.26 40

3-55 20

26.37

Figure 5-3: Total System Stratification— The Power Pyramid, 1884

State Power Number as a % of the of States Greatest State

Z 80 2 6of <80

7 40, *6 0

2 20, <40

4.20 27

Figure 5-4: Total System Stratification Based on the Most Powerful State, 1884 136

Number % of Total of States System Power

100 1.39 80 States 1.75 60 1.90 *f0

20

0

Figure 5“5* Total System Stratification— -The Power Pyramid, 1909

State Power Number as a % of the of States Greatest State

> 80 1 £j. 2_ 60, <80

y 4o, <6o > 20, <40

<20 3^

Figure 5-6* Total System Stratification Based on the Most Powerful State, 1909 Number of % of Total States System Power 100 N « 66 States .85

1.71

3.9 6 20 58.82

Figure 5-7: Total System Stratification— The Power Pyramid, 1934

State Power Number as a % of the of States Greatest State

> 80

> 60, <80

> 4o, <60

> 20, <40

61 <20

Figure 5-8* Total System Stratification Based on the Most Powerful State, 1934 138

Number % of Total of States System Power 100 80

States 60

4o 3.81 20

Figure 5“9* Total System Stratification— The Power Pyramid, i960

State Power Number as a fo of the of States Greatest State

2 80

> 60, t 80

> 40, *60

7 20, <40

87 < 2 0

Figure 5-10* Total System Stratification Based on the Most Powerful State, i960 of the system structure than the indices. The information is, of course, the same* the form is merely different. Again we see the highly concentrated nature of the total system throughout the period and the low variance except among the top few states. For instance, in 1864 (see Figure there were four states in the top bracket

(having at least 8C$ of the power of the top state). In i960 there were only two states in the top four brackets. Yet at both points we have the vast majority of states at the bottom of the spectrum with less than 2Q# of the power of the top state. The system has changed, but the change has occurred at the top and not at the bottom.

The vast majority of states are at the bottom in both cases. Yet in the elite system we have in one case a high degree of dispersion while in the other we have extreme concentration. In the next subsection we will look more closely at the top of the spectrum, for it is here that the most significant changes have occurred.

Elite System Power Structure

Table 5“3 on the next page presents, for the period of our concern, the structure of an elite system composed of the top five states at each measurement year. Again the power structure is de­ scribed in terms of three indicesj the GINI index, the HH index, and the concentration ratio of the ppwer of the top state to the total power of all five states (G^).

Again we notice that all three indices are highly correlated.

As Table on page l4l shows, the lowest correlation is .973* Also, if we compare the dichotomization (Table 5-3)» we find it substantially the Bame based on all three indices. The GINI and HH indices produce Table 5-3* Power Structure of the Elite System, 1864 Through I960, in Terms of Concentration-Dispersion Indices

*

! Year GINI HH Cl Index Index Index #--■ 1864 D* .0959 D .2095 D .2444 1869 D .O858 D .2073 D .2412 1874- D .1116 D .2106 D .2380 1879 G .1711 G .2232 G • 3133 1884 D .1469 D .2173 D .2979 1889 D .0699 D .2041 D .2417 1894 D .1066 D .2111 D .2480 1899 D .0657 D .2034 D .2401 1904 D .1209 D .2113 D .2793 190 9 D .0685 D .2043 D .2373 1914 D .1406 D .2183 G .3154 1919 C .3375 C .2921 G .3684 1924 G .2553 C .2324 G .3391 1929 C .3299 C .2979 G .4633 1934 C .3087 G .2805 G .4172 1939 G .2018 G .2336 D .3070 1946 C .4793 C .4003 G .5708 1950 C .3821 C .3245 G .4297 1955 C .4366 G .3808 G .5311 i960 G .4538 G .3848 C .5487

■#(2 or D to the left of a number indicates that when the entire period was dichotomized according to that index, the particular year was in the more concentrated (C) or more dispersed (D) half. 141

Table 5-4: Pearson Correlation of Concentration-Dispersion Indices within the Elite System

GINI HH Cl

GDI I 1.000 .976 .973

HH .976 1.000 .978

Cl .973 .978 1.000

Table 5-5: Pearson Correlation of Concentration-Dispersion Indices between the Elite and Total Systems

■

Total System

GINI HH °5 T GINI .869 -936 .884 Elite

System HH .837 .968 .873

.816 .916 .812 ci 142

identical results. The concentration ratio differs only in the

classification of two points. Moreover, if we compare the di-

chotomizatio n based on the elite system indices to that based on the

total system indices, we also find them almost identical. Similarly, the Pearson correlations between the two systems are high, as Table

5-5 on page 1^1 depicts.

Whereas the total system showed slight variance, particularly in terms of the GINI index, we see in Table 5“3 substantial variance in the elite system. At the most dispersed point (1899) the GIWI in­ dex is ,0657, which is remarkably close to perfect equality. Figure

5-11 on page 1^3 depicts this distribution. We see that the Lorenze curve for 1899 almost touches the 4$° line which represents perfect equality. At the most concentrated point (19^) the GINI index for the elite system is .4793* This obviously indicates that there was not the equality in 19^6 at the top of the power spectrum that we found in 1899*

Figure 5“H also presents the Lorenze curve for 1946.

Table 5-6 on page 144 presents the power indices for the en­ tire elite system for each of the twenty measurement points. The num­ bers depicted represent each state's relative power in terms of the total system. For example, in 1919 the most powerful state had a score of 298.19 and thus controlled 2 9 .8# of the world's power. We have deliberately avoided placing in the names of each state because here it is the distribution and not the historical interpretation that interests us. We will provide the names of these states in the next section, however. As with the stratification drawings of the total system structure (Figures 5-1 through 5“10)*we lose some comparability I*f3

100

^5° line of Perfect Percent Equality of 1899 System Distribution Power

19*16 Distribution 20

20 80

Percent of System Membership

Figure 5-11* Lorenze Curve Depicting the Power Distribution of the Elite System for 1899 and 19^6 Table 5-61 Power of the Top Five States as a Proportion of 144 Total ^ystem Power

1864 1869 1874 1879 1884

156.76 146.02 146.58 190.23 176.86 148.42 133.92 129.17 142.86 138.23 130.25 129.70 126.64 111.70 103.06 97.^6 129.93 116.77 83.48 81.19 - 76.14 79.08 82.36 81.12 78.08

1889 1894 1899 1904 1909 153.68 139.1^ 155.49 157-24 178.59 124.28 145.35 138.94 143.28 117.79 112.91 135.02 132.18 119.10 112.94 109.30 120.18 119.29 108.65 110.58 90.00 70.91 107.33 89.81 101.85

1914 1919 1924 1929 193^

187.07 298.19 224.35 313.32 300.13 194.58 117.30 247.17 172.76 137.31 108.32 197-84 156.02 114.05 90.49 94.16 38.95 59.09 57.75 67.14 67.04 86.31 2 7 .18 49.33 53.81

i960 1939 1946 1950 1955

231.23 502.46 348.89 419.83 429.71 234.78 218.55 197.92 284.84 210.37 120.79 131.61 76.37 61.02 59.89 43.48 96.47 24.35 53.81 38.73 8 6 .10 2 3 .8 6 47.99 36.17 39i70 1^5

in such a display. However, hopefully the reader can more clearly

see what changes have occurred during this period through this tahle,

particularly if he choses to match the complete distribution with the

more comparable indices contained in Table 5“3»

Historical Interpretation

If we view the measuring points less as abstract points and

more as historical dates, we can draw a number of conclusions. For

instance, historical interpretation seems to agree with our results,

thus substantiating their validity. We can clearly see, for example,

^ and we would predict based merely on the abstract figures, the end of

the "Concert of Europe" and the end of the pluralist "balance of power"

after World Wax I. From lf36*f to 191^ power had been fairly well dis­

persed among the elite states. After World War I, however, the system

entered a period of concentration. This concentration became most

pronounced after World War II. When the system became concentrated

the pluralist processes could no longer go on operating. Of course,

here we are speaking only about the relationship between the major

powers, for as was demonstrated previously, the total system was very

concentrated throughout the period.

Moreover, we feel that these results fit nicely with a number

of macro-historical behavioralist studies produced by the Correlates

of War Project which have noticed that the international system be­

haved quite differently in the 19th century than it does in the 20th 7 century. Our data shows that there is a significant difference in

the degree of power concentration and dispersion between substantial 1H6 portions of these centuries. If the theory contained in the last

chapter dealing with effects of power structure on the patterns and processes of interaction is correct, then much of the change in systemic behavior between the 19th and 20th centuries could be ex­ plained by these changes in power structure.

Not only are our results consistent with historical inter­ pretation at the general level, we also feel that they are consistent with more specific analysis. Table 5-7 or* page 1^7 rank-orders the top five states for each measurement year. This table becomes more meaningful for historical analysis if one will refer back to Table 5-6, where the actual power scores are presented. While each reader familiar with world history will undoubtably make personal judgments as to the validity of these rankings, we feel that these results are extremely valid. We feel that our results give an excellent measure of which nations were capable of exerting the most influence in the international system. For instance, we see in the post World War II era the United States, Soviet Union, England, France, and China ranked in an order and possessing power indices that seem consistent with their influence during that period. We see in our data the dominant position of France in the period between world wars, and we can under­ stand the complacency of the Allies in the face of a militant Germany.

The temporary elimination of Germany and Russia from world politics after World War I is depicted in our data. Moreover, the United States does not emerge in the 20th century as a major world power until after

World War I. Studies based on other indicators of power have seen the

United States as a world power much earlier, long before it began to 1 47 Table 5“7* Rank-Order List of States Comprising an Elite System of Five States -i------1864 1869 1874 1879 1884

U.S. England France France France France France England England England Russia Russia Russia Russia Russia I England U.S. Austria Germany Austria Austria I Austria U.S. Austria Italy

1889 1894 1899 1904 1909 1

France Austria Russia Russia Russia 1 England France England England Germany 1 Russia Russia U.S. Austria England 1 Austria England France France Austria 1 Germany U.S. Austria U.S. France 1

1914 1919 1924 1929 193^

Russia England France France France 1 England U.S. England U.S. U.S. France France U.S. England England Germany Italy Germany Germany Germany Austria China China Belgium Russia

1939 1946 1950 1955 i960

• France U.S. U.S. U.S. U.S. Russia Russia Russia Russia Russia England England England England England U.S. China France France France Germany Canada China China China 1^8

act like one. Our study, however, shows the United StateB in the

elite system after the Civil War, around the turn of the century

(Spanish American War), and from World War I to i960. This indicates

that our power index achieved the balance of potential and motivation

that we hoped it would. The rise of German power in the decade prior

to World War I is clear in our data, as is the dominance of Russia,

France, and England during the l8?0's*l880's, and 1890*s.

Some readers will probably question the low scores of Germany

during most of the period. But as was mentioned earlier, we feel that

because of Germany's war-like behavior (and because of many people's

incorrect total reliance on war capability as a measure of power) its

power has been generally overrated. If we look at the total diplomatic

picture, our results gain support. For example, we can see not only

why the German General Staff was so hesitant about moving against the

Allies in 1939 "but also why they were able to sustain the war after

initially eliminating France and then converting French strength for

their own use. Moreover, in the period before World War I the alliance

system did not solidify against Germany until after 190*f. This too

makes sense if we observe the growth in German strength that appears

in our data by 1909.

In addition, our data might tend to clarify at least one

apparent historical anomaly. After it became clear that France would be defeated in the Franco-Prussian War, the other major powers did not attempt to help her. Normally, when one major power even temporarily subdues another, the rest of the system usually turns against the victor to keep that power from moving to a position of 1^9

system dominance. In I87I this did not occur. The reason becomes

clear in our data. Prussia had not reached a power position of such

a magnitude that it threatened the rest of the system, despite its

quick military success. The rest of the.system did not turn against

Germany, despite its constant militancy, until after the turn of the

century, when Germany's industrial might caught up with its military

skill.

We wish to make clear, however, that in the last few para­

graphs we were not attempting to revise historical interpretation. Our

data is certainly not sufficient for such a task. We were merely

demonstrating the validity and usefulness of our power index. If one

is offended at our having dealt only with the powers at the top, the

stratification drawings (Figures 5-1 through 5~10) should.once again

make the reason clear. When one realizes that in a system of 89

states in i960 it took the combined power of slightly over 83 of them

to equal 2Q% of the world's total power, then one can understand why

so little has been written about the influence of the smaller states.

The system structure may have changed since then, but up to that point

one could hardly expect these states to have much effect on world

politics except as tools of one of the great powers.

Interstate War Data

Our source for this data is the well-known work by J. David

Singer and Melvin Small, The Wages of War, 1816-1965* A Statistical

Handbook. The information from this work with which we are most con­

cerned is the year during which particular wars began, the number of battle-connected military deaths in those wars (battle deaths), and 150 the sum of all the participating nations* separate months of active

involvement (nation months). We have used this basic data to compute

our three measures of war— magnitude, amount, and frequency— as Q described in Chapter III.

Given this data, there are two problems associated with

measuring our dependent variable. The first comes from our definitions

of war and the system. In discussing war in the system, we are talking

about war between member states. However, during the period covered

in our quantitative analysis there were a series of wars where one member of the state system engaged in conflict against a political unit which would not be classified as a state. These conflicts are most often called colonial wars. One might believe that they should be in­ cluded in our analysis. We have chosen to exclude them. Theoretically we have not dealt with extra-systemic actions and reactions, and, therefore, we cannot consider them now empirically. Also, we have chosen not to include civil wars.

The second problem is adjustment. For instance, one might argue that battle deaths in I865 cannot be compared directly in many cases to battle deaths in i960 without adjustment. The techniques of destruction, as well as the world population have increased tre­ mendously during this period. Similarly, the number of wars and nation months at war in a system’with 39 states may not be directly comparable with nation months at war in a system with 89 states. In order to aid comparison, we have attempted to adjust these figures by appropriate dimensions to allow for the continued growth of the system across the period of our concern. 151 Table $-8 on the next page details the growth of the system in

terms of three factors— the number of nations in the system, total

system power, and population. We used the growth in the number of

nations in the system to adjust two of our indicators— frequency and

nation months. The number of nations in the system was determined as

previously discussed, based on Singer and Small's two studies of

system membership. We adjusted the frequency by dividing the number

of wars begun in a period by the number of nations in the system at the

time. We adjusted the nation months indicator by dividing the number of nation months of war begun during a period by the number of nations in the system at the time.

We used the growth in total system power to adjust the battle deaths indicator. To do this we established 1864 as the base year for both defense expenditures and government revenue (the same data used to determine the distribution of power). We then divided the total dollar figure of each succeeding year for both defense expenditure and government revenue by the 1864- figures. To weigh each indicator of power equally, we tool: the average of the two resultants as our ad­ justment factor. We then adjusted each year's battle deaths by dividing by the adjustment factor.

If one refers to Table $-8, one will notice that the growth * in system power is rather phenomenal. There is no single factor to which one can refer in order to substantiate this growth or to put it in more lifelike terms. However, if one looks at weapons technology, at man's ability to kill each other, across the period, one will see that the adjustment factor is at least reasonable. In 1864 men were Table 5-8: Growth of the Total System in Terms of System Membership, System Power, and System Population

r ■■■ — Number Power Popula­ Date of (based tion Nations on 1864) (based on 1864)

1864 39 1.000 1.000 1869 34 .993 1.055 1874 31 1.142 1.069 1879 34 1.218 1 .1 10 1884 35 1.236 1.151 1889 38 1.400 1.217 1894 38 1.608 I.258 1899 41 2.049 1.321 1904 42 2.336 1.373 1909 43 2.858 1.445 1914 43 3.320 1.499 1919 61 5.150 1.535 1924 62 6.980 1.617 1929 63 9.413 1.706 1934 66 10.590 1.803 1939 62 16.606 1.818 1946 65 29.984 1.779 1950 75 43.363 2.728 1955 84 84.694 3.227 i960 89 108.295 3.571 153 still using rifled muskets and smoothbore cannons to kill each other.

By World War I they were using machine guns and rapid fire artillery.

Our factor indicates a growth of roughly three times across this period.

Between World War I and World War II our factor increases five times,

which is not unreasonable considering developments such at the stra­

tegic bomber. Between World War II and i960 we see a growth of seven

times. Considering the introduction of nuclear weapons, this too seems

to be an accurate adjustment. We do not wish to imply that we en­

vision growth only in terms of weapons capabilities. However, our

simplistic reference to weapons technology does hopefully impart some

realism to our adjustment factor, as well as provide at least an

elementary check on its validity.

We also considered population growth as a means of adjusting battle deaths, and we have, therefore, listed the growth in terms of

1864 population in Table 5-8• The population was computed by adding together the population figures contained in the Arthur Bank's Cross-

Polity Time-Series Data for those states previously identified as being members of the international system. However, the population of the system grew only 1.779 times between 1864 and 1946. Using such a fac­ tor we would be saying that a war producing one million battle deaths in 1864 was equivalent in magnitude to a war producing 1,7 million battle deaths in 19^6. Considering the growth in destructive capa­ bility between the two points in time, we could not accept this.

Therefore, we decided to control only for the dollar growth in de­ fens e expenditures and government revenue, which seemed to provide a better representation of reality. 154

Tables 5-9 and 5-10 list the interstate wars that occurred

"between 1862 and 1964, as well as the data on each in terms of nation

months of war, adjusted nation months, battle deaths, adjusted battle

deaths, frequency, and adjusted frequency. The last column of the

tables indicates whether the war occurred during a concentrated (C)

or a dispersed (d ) period based on a dichotomization of the entire

length of time using the GINI and HH index for the total system. An

asterisk to the left of a war indicates that the war might be considered

a major war in that it involved at least two major powers fighting each

other for substantial stakes.

This designation of major wars in terms of the power of the

o pponents and their intentions is certainly judgmental. In most cases

it is relatively clear, however. One might question the inclusion of

the Russo-Turkish War of 1877 and the Italo-Turkish War of 1911 on the

basis that Turkey was not a major power. But here we are dealing with

the Ottoman Empire and not just the Turkey of the post World War I

period. Similarly, one might question the inclusion of Spain as a major power in I898. However, one should remember that Spain still had a substantial colonial empire until it was stripped away during that war. We chose not to include the Sino-Indian War of 1962 and the Russo-Japanese War of 1939 in the category of major wars, not because of the power ranking of the states involved but because we are not sure that either was a war in the normal sense. Both apjjear to have been border incidents. It appears that the designs of none of the states participating involved any significant impact on world politics. Moreover, the latter of these incidents was fought 155 Table 5-91 Interstate VIar Data, 1862-1909

c Nation Nation Battle Battle Freq. Freq. AdjugJ- or War Months Months Deaths Deaths AdjugJ- (1000* s) Adjust^ D

Franco-Mexican, 115.4 2.96 20.0 20.0 1 .0256 D 1862 D Eduadorian- 1.0 .03 1.0 1.0 1 .0256 Columbian, 1863 .0256 D 2d Schleswig- 10.8 .28 4.5 4.5 1 Holstein, 1864 D Spanish-Ghilean, 1 6 .8 .43 1.0 1.0 1 .0256 1865 .0256 D *Seven Weeks, 14.5 .37 36.1 36.1 1 1866 .0294 D *Franc o-Prus sian, 2 7 .0 .79 187-5 188.4 1 18?0 .0294 D ^Russo-Turkish, * 17.6 .52 285.0 234.0 1 18 77 .0294 D Pacific, 170.3 5.01 14.0 11.5 1 1879 .0286 D *Sino-French, 2 3 .6 .67 1 2 .1 9.£ 1 1884 .0286 D Central American, 1.2 .03 1.0 .£ 1 1885 D *Sino-Japanese, 1 6 .0 .42 1 5 .0 9-: 1 .0263 1894 .0244 D Greco-Turkish, 6.2 .15 2.0 1.0 1 1897 .0244 D *Spanish-American, 7.4 .18 10.0 4.9 1 1898 1 .0238 C *Russ0-Japanese, 38.4 .92 1 3 0 .0 55? 1904 .0238 C Central American, 5.4 • .13 1.0 .4 1 1906 D Central American, 6.3 .15 1.0 .3 1 .0233 1907 D Spanish-Moroccan 17.0 .39 10.0 3.5 1 .0233 1909

1 ^Indicates major war. Table 5~10i Interstate War Data, 1910-1964 156

1 Ration Battle Battle Freq. Freq. C Ration Months Deaths Deaths AdjusJ- or War Months (AdjugJ- AdjugJ- (1000's D

*Italo-Turkish, 25*4 .59 20.0 7.0 1 .0233 D 1911 First Balkan, 20.4 .47 82.0 24.7 1 .0233 D 1912 Second Balkan, 4.2 .10 6 0 .5 18.2 1 .0233 D 1913 *World War I, 607.8 14.14 9000.0 2710.8 1 .0233 D 1914 Hungarian-Allies, 10.8 .18 11.0 2.1 1 .0164 C 1919 Greco-Turkish, 82.6 1.35 50 .0 9.7 1 .0164 C 1919 •^Manchurian, 33.2 .53 60.0 6.4 1 .0159 C 1931 Chaco, 7 1 .8 1.09 13.7 1.3 1 .0152 C 1932 Italo-Ethiopian, 14.4 .22 20.0 1.9 1 .0152 C 1935 Russo-Japanese, 12.6 .20 19.0 1.1 1 .0161 C 1939 *World War II, 981.8 15.84 16000.0 963.5 1 .0161 C 1939 Russo-Finnish, 6.8 ,11 90 .0 5.4 1 .0116 C 1939 Palestine, 19.4 .26 8.0 .2 1 .0133 C 1948 ^Korean, 514.0 6 .8 5 2000.0 46.1 1 .0133 C 1950 Russo-Hungarian, 1.6 ■ .02 32.0 .4 1 .0120 C 1956 Sinai, 1.0 .01 3.2 .0 1 .0120 C 1956 Sino-Indian, 2.2 .03 1.0 .0 1 .0112 C 1962

1 indicates major war; 157 officially by client troops and not the forces of Japan and Russia.

Tables 5"9 and 5**10 thus contain the data that is the basis

for our dependent variable# The significance of the information

contained in these tables is probably not readily apparent to the

reader at this time. Hopefully, the information becomes more meaning­

ful to the reader in the next section, when we connect the data re­

lating to the independent and dependent variables and thus provide

the original evidence that was promised earlier.

Analysis and Evidence

We have chosen two means of establishing quantitatively the

relationship between our variables. We will first of all use

correlational analysis. Then we will dichotomize the entire period

of concern into a concentrated and a dispersed period and compare the

characteristics of one with those of the other.

Correlational Analysis

There are two basic techniques than can be used to correlate

our independent variable at the twenty measurement points between

186*f and i960 with our dependent variable. One technique is to correlate the independent variable data with the war data for the five-year period beginning at the measurement year. In other words, the war data for 186^ to 186G would be correlated against the power structure for 1864-. We call this the initial point technique. A second technique, which we call the midpoint technique, is to corre­ late our independent variable data with the war data during the five- year period for which our measurement year is the middle year. For 158 example, "the wax data fox 1862 to 1866 would be correlated against

the structure for 1864.

Each technique has advantages and disadvantages. The mid­

point technique probably gives the best estimate of the structure

during the period because the measurement year and the extreme years

of the period are only two years apart. However, with the midpoint

technique we are considering wars that occurred before the measure­

ment year. Thus one might raise a question of the validity of any

supposed causation. The initial point technique eliminates this

problem, but the extreme year is separated by four years from the

measurement year. Thus the actual structure of that extreme year

will most likely be closer to the next measurement year than the

measurement year for which it is counted. In other words, using the

initial point technique the actual structure of the year 1868 is

probably closer to the measured 1869 power structure than the

measured 1864 power structure. For this reason we feel that the mid­

point technique is probably the more valid. However, for the corre­ lational analysis we have used both.

We will correlate six indicators of system power distribution

(see Tables 5-1 and 5-3) with four indicators of the amount of wax in the system (the summation of the magnitudes of all the wars that occurred during the five-year period) and two indicators of the frequency of wax in the system (see Tables 5"9 and 5”10)» Based on the theory developed in Chapter IV, we would predict that there would

be no correlation between the amount of wax in the system and the

distribution of power in the system. The amount of war in the system 159

is not dependent upon the overall power structure of the system in

terms of concentration and dispersion. What is dependent upon the

power structure of the system is the pattern of war that occurs. A

concentrated system should have fewer wars, "but individually the wars should tend to be of a greater magnitude than those of a dispersed system. A dispersed system should have more wars, but they should tend to be of lesser magnitude individually than those of a concen­ trated system. As concentration increases, frequency should decreasej as concentration decreases, frequency should increase.

It should be noted that we are making no direct correlational analysis between power distribution and the average magnitude of wars, even though the theory posits a relationship. The reason for this is that there are so few cases of war in each five-year period (the average number of cases is 1.7)• Thus, the presence or absence of even one war, however minor, in a five-year period produces extreme differences in average magnitude. We do analyze this relationship directly later when we dichotomize the total period and are considering fifty-year periods and hence a greater number of cases in each period

(nineteen in one and fifteen in another). We also consider the re­ lationship here indirectly. Because the amount of war is a function of frequency of war and the magnitude of individual wars, if the amount

• of war is not affected as power becomes concentrated and the frequency is reduced, it follows that the average magnitude should increase.

Thus, in summary, evidence which would tend to support our theory would be no correlation between concentration and the indicators of the amount of war in the system and a negative correlation between 160

concentration and the indicators of the frequency of war in the sys­

tem. Tables $-11 through 5”!^ depict the results of the correlation.

The results are fairly clear; therefore, there is no reason

to restate them in detail. If one looks at the average correlation

for the measures of system power concentration for both the elite and

total systems, one can see that the results are what we expected. For

both the midpoint and the initial point techniques we do in fact have

almost no correlation between most of our indicators of concentration

and the amount of war. Only the adjusted battle deaths indicator of

the amount of war shows any degree of correlation. This indicator is

also the deviant case when we dichotomize, and we will discuss the reasons for this deviation when we examine the dichotomization. The frequency indicators for both techniques do in fact have a negative correlation, and they are the most substantial correlations that we find in our analysis. The inverse relationship between concentration and frequency of war becomes most clear when we adjust for the growth in the number of nations in the system, reaching a correlation as high as -.^f2 (midpoint technique, HH index, elite system).

Thus we can conlude that the correlational evidence generally supports the theory. The existence of the relationship which was posited in the theory becomes easier to visualize, however, when we dichotomize the twenty measurement periods into the ten most concen­ trated and the ten most dispersed, and then compare the war data for each. -161 Table 5-11t Pearson Correlation of Concentration and the Amount of War in the System, Midpoint Technique

Amount of War Nation Battle Nation Months Battle Deaths Months Adjusted Deaths Adjusted

GIN I .204 .088 .156 -.0 8 0 Total HH -.015 -.1 0 8 -.0 6 2 -.189 System .158 .088 -.174 c5 .033

Average .116 .004 .061 -.148

Elite GIN I '-.009 -.094 -.0 6 6 -.1 6 1

System HH -.0 9 2 -.1 7 0 -.1 2 9 -.194

Ci -.0 6 5 -.1 2 2 -.0 9 0 -.117

Average -.055 -.129 -.095 -.157

Table 5-12i Pearson Correlation of Concentration and the Frequency of War in the System, Midpoint Technique

Frequency Frequency Adjusted Total GIN I .004 -.264 1 to System HH • -.394

C5 -.053 -.287 CO 0 1 • Average -.315

Elite GIN I -.2 3 1 -.441

System HH -.2 5 1 -.442

-.240 -.424 C1

Average -.241 -.436 .. Table 5-13* Pearson Correlation of Concentration and the Amount 162 of War in the System, Initial Point Technique

Amount of War Battle Nation Nation Battle Deaths Months Months Deaths Adjusted Adjusted

GIN I .200 .069 .156 -.091 Total HH -.011 -.1 2 3 -.063 -.195 System -.186 C5 .156 .012 .087 H 1 Average .115 -.01*1 .060 • H 1 Elite GIN I -.0 0 7 -.110 -.067 .

System HH -.086 -.182 -.129 -.201

-.132 -.092 -.130 c5 -.059 MO CO iH 1 Average -.0 5 1 -.1*4-1 -.096 •

Table 5-l*4-i Pearson Correlation of Concentration and the Frequency of War in the System, Initial Point Technique

Frequency Frequency Adjusted Total GIN I .00*4- -.3 0 6

System HH -.115 -.359

C -.031 -.309 5

Average -.0*4-7 -.325

Elite GIN I -.1 6 0 -.*4-20

System HH -.157 -.*10*4- .=}■ O t Cl -.158 • 1 0 £ Average -.158 • Analysis through Dichotomization

For this analysis we have chosen to dichotomize based on the

GINI and HH indices for the total system (see Tables 5-1» 5-9» aJid 5-10).^ To match the war data to the power distribution data we have

chosen to use the midpoint technique. Thus, when we dichotomize we

find the dispersed period stretching from 1862 to 1901 and from 1907

to 1916. The concentrated period stretches from 1902 to 1906 and then

from 1917 to 1962. If one refers to Tables 5-9 and 5-10, which con­

tain the data on all the interstate wars which occurred during these

periods, one will discover a substantial amount of evidence, most of

which tends to support our theory.

To begin, if one is concerned only with the major wars, one

finds that eight major wars occurred during the dispersed period and

only four occurred during the concentrated period. During the dis­

persed period major powers fought each other in earnest eight times,

but fortunately only one of these wars exploded into a world war

(World War i). During the concentrated period major powers fought

each other in earnest only four times, and one (World War II) and

perhaps another (Korean War) exploded into world wars. Further, we

should not forget that the world came very close to its first global

nuclear world war in 1962 (Cuban Missile Crisis) during a period of

extreme concentration, and the mere nearness of such an experience further supports our theory. Thus the system behaved exactly as our

theory suggested it should. During the dispersed period more wars did occur, but the system appeared better able to contain them. It is on this fact that most of the praise of the pluralist processes 164

(associated by many with "balance of power") rests. The concen­

trated period did have fewer wars, but when they came they were

extremely destructive. The concentrated period also produced long

periods of relative peace.

Moreover, the evidence is just as positive if we examine the

frequency of all interstate wars. The dispersed period produced

nineteen wars; the concentrated period produced fifteen wars (thirteen

if we do not count the two questionable wars discussed earlier).^

This may not appear to be a substantial difference. However, one

should note that the average number of states in the dispersed period

was 37*6, while it was 66.9 during the concentrated period. Thus we

find that if we add the adjusted frequency measures for each period,

the dispersed period had over twice the frequency of war as did the

concentrated period (.4883 vs* .2368 wars per nation).

The evidence relating the amount of war to system power

structure is almost as positive in support of our theory, which states,

of course, that neither a concentrated nor a dispersed system will

produce more or less war. If we add the nation months of war for

both periods, we find the dispersed period producing 1108.9 nation months of war and the concentrated period producing 1796.2 nation months of war. When we add the adjusted statistics, we find the dis­

persed period producing 2 7 .7 nation months of war per nation and the concentrated period producing an identical 2 7 .7 nation months of war per nation. Thus the costs of war in each period in terms of nation months of war is exactly the same when adjusted. 165 When we consider battle deaths we get a slightly different

result. The total battle deaths for the dispersed period are 9*76

million. For the concentrated period they are 18.44 million. When we

adjust for system growth in terms of total power based on 1864, we

find the dispersed system producing an 1864 equivalent of 3*27 million

deaths and the concentrated system producing an 1864 equivalent of

1 .0 9 million deaths. If our theory is correct and the costs of each

system are the same, then we would in fact expect that the unadjusted

figures would be greater in this particular concentrated period than

in this dispersed period because of growth (the concentrated period

falls mostly in the twentieth century). This is what occurred with

the nation monlhs indicator of the amount of war, and it adjusted to

perfect equality. When we adjust battle deaths, however, the con­

centrated period appears to produce less war.

There could be a number of explanations for this anomaly.

Our theory could of course be wrong. But this is one of the few

pieces of contradictory evidence. Almost everything else fits per­ fectly. Obviously, therefore, we do not think that this is the case.

We may have over-adjusted in our attempt to equate 1939 or i960 battle deathB with 1864 battle deaths. Perhaps dividing by factors

of 16 .6 (1939) or IO8 .3 (i960) was just too great. The concentrated period was historically the 20 th century and, therefore, was perhaps overly affected by these high adjustment factors. Or, perhaps the numerous other factors which affect the amount of war in the system made our results a little less than perfect. For instance, World War

I seems to be somewhat of an anomaly in itself. It is the classic example of men getting carried away and losing touch with whether the

costs justify the ends. Nationalism undoubtably buried the normal

ends-means relationships which govern most human action and pushed

the costs of that war past what we would normally expect. Nationalism

was a historical factor for which we could not control. Also, again

we should mention the near catastrophe in 1962. which would have

certainly balanced the casualty figures had it developed. As we are

dealing with probabilities in our theory, this near occurrence is

relevant. Finally, our battle death statistics did not include civilian

casualties because during most of the period the killing of civilians

was an unfortunate biproduct of war and not the result of the de­

liberate decisions of war. Yet beginning in the latter part of the

concentrated period (1939-19^5) civil populations became deliberate

targets of war. Perhaps civilian casualties should have begun to be

counted at this point. Had the bombs dropped on England and Germany been dumped on troop concentrations instead, the military battle deaths would have certainly been much higher. Had we counted these civilian casualties resulting from deliberate decisions, our adjusted battle deaths figures would have matched more closely and thus lent more support to our theory (estimates of civilian deaths in World War II range from 20 to 35 million11). But as we mentioned earlier, the statistician's ability to adjust is no better than the historian's, it is just more explicit. Perhaps in this case we are slightly in error.

The data also tends to support the belief that the wars of a concentrated system, while fewer, tend to be of greater magnitude 167

than the wars of a dispersed system. The average magnitude of the

wars during the concentrated period was 120 in terms of nation months

and 1,229 thousand in terms of battle deaths. The average magnitude of

the wars during the dispersed period was in terms of nation months

and 51^ thousand in terms of battle deaths. The adjusted average nation

month figures are I.8 5 nation months per nation in the concentrated

period and 1.46 nation months per nation in the dispersed period. Again,

for the reasons we have mentioned above, the adjusted battle death

figures do not conform with the rest of the data. In terms of adjusted

battle deaths, the average magnitude of the wars during the concentrated

period was 73 thousand battle deaths, and during the dispersed period it was 173 thousand battle deaths.

As our final bit of original evidence, we have used the same

concentration-dispersion indices on the war data as we did on the power structure data. We would expect that the more concentrated the period, the greater the indices should be when applied to the amount of war data; the more dispersed the period, the smaller the indices should be when applied to the amount of war data. Table 5“15 on the next page shows the results.

In this case the single piece of contradictory evidence is the battle deaths indicator using the HH index. In every other instance, the concentrated period produces* a more concentrated pattern of war.

The dispersed period produces a more dispersed pattern of war. This in itself is a demonstration of the influence of structure on human be­ havior. However, as. we. demonstrated* earli'er, while the pattern of be­ havior changes, the basic nature of the behavior in terms of the amount of war does not change. 168

Table 5“15* Comparison of Concentration and Dispersion in the Amount of War with Concentration and Dispersion in Power Dis- tribution

Indicators of the Amour:t of War Power Nation Battle Distribution Nation Months Battle Deaths Months Adjusted Deaths Adjusted Dispersed Period .7^2 GINI .7^8 .919 .895

INDEX Concentrated Period .891 .886 .976 .963

Dispersed Period .338 .308 .851 .6 8 9 HH

INDEX Concentrated .385 .393 .765 .780 Period notes to chaiter five

^■Arthur S. Banks, Cross-Polity Time-Series Data (Cambridge, Massachusetts: The MIT Press, 1971)• 2 The principal sources that Banks used to compile the revenue and expenditure data which we used were The Statesman’s Yearbook (for 1860-1939) a-nd the U.K. Statistical Yearbooklfor 19*R^-1966). Banks indicates that the greatest problem with the reliability of this data stems from the difficulty of conversion to a common currency. How­ ever, in cases where certain factors precluded the assembly of reasonably reliable data, that data was omitted. Banks feels that a margin of error in the vicinity of 5% must be allowed for even the most rigorous of his aggregate statistics and that the range of error from 10-20J2 is seldom exceeded in his collection. For a more de­ tailed discussion of sources and reliability, one should see the Introduction of Cross-Polity Time-Series Data.

^For a discussion of the sources drawn upon by Ferris, the procedures used in gathering data, and the methods used for estimating missing data see Wayne H. Ferris, Nation-State Power and Economic Variable Data 1850-1966 (Lexington, Mass.: D.C. Heath and Co, forth- coming). k This difference between the percentage of missing items and missing overall value is accounted for because of the fact that most of the missing data was for extremely small states such as Yemen, Albania, Mongolia, and Ethiopia.

■^For instance, in 1929 no data was available for Ethiopia, Afghanistan, and Mongolia, all of which were ranked near each other in the Singer and Small data. Had we used the averaging method, their power index would have been inflated because Australia, Canada, Ireland, and South Africa would have figured in the averaging. There­ fore, these three states were given a value compiled based on states ranked near them, such as Ecuador, the Dominican Republic, and Albania, which seemed more equivalent than Canada and Australia.

^For the top five states, which controlled an average of 6Qfc the world’s power during the entire period, the Pearson correlation between the two scores was .8 3 6. One would expect that the corre­ lation between defense expenditures and government revenue would be reasonably high. This is not high enough in our opinion, however, to justify using only one indicator and dropping the other.

169 170 7 See, for instance, Singer and Small, "Alliance Aggregation and the Onset of War, I8l5-19^\5»" P» 283. Q We have made two modifications to this data. First of all, the Chaco War (1932-1935) between Paraguay and Bolivia is listed as having produced 1 3 0 ,0 0 0 battle deaths to military combatants even though there were only seven thousand soldiers on both sides at the beginning of the war. This ratio of almost 19 soldiers killed for every one in the armed forces is roughly ten times that of the next most severe war, World War II (2 battle deaths per soldier in service at the beginning of the war). We have therefore reduced the magni­ tude of this war (in battle deaths) such that it has a severity (battle deaths per armed forces) equivalent to World War II. We feel that this reduced magnitude more accurately portrays the international importance of this war. Secondly, the original data lists the Sino-Japanese War which began in 1937 as a separate war until 19^1, at which time it becomes part of World War II. We have chosen to include it in World War II. World War II was a grouping of numerous wars that began at different times and ended at different times. We see no grounds for separating this particular phase out. We have chosen to follow the Singer and Small disposition of two questionable cases of war, but with some reservations. We are referring to the hostilities between the Soviet Union and Japan that occurred along the Mongolian border for four months in 1939 and the hostilities between China and India along their border for one month in 1964. Singer and Small list these as interstate wars. However, they seem more to be mere border disputes, which in that part of the world are a distinctly different form of politics than war. For instance, the fighting in recent years along the Chinese and Russian border does not seem to be a war in the usual sense. We are not sure that these two are either, and we will keep these reservations in mind during the analysis of results. a 7We have chosen to dichotomise based on the total system be­ cause both the GINI and HH indices for the total system give iden­ tical results. However, because of the high correlation between all of our indicators of power distribution, we would expect results similar to those presented here regardless of which indicators were • used to dichotomize the period.

"^The Russo-Japanese dispute of 1939 and the Sino-Indian dispute of 1962. 11 Singer and Small, The WageB of War, 1816-1 9 6 5, p. ^8. CHAPTER V I

CONCLUSION

All the theory contained in this study was developed and on paper long before the quantitative work of the last chapter was done.

The results brought no revision of the theory, for they supported it almost totally. Frankly, considering the vast number of other fac­ tors which influence the amount and pattern of war, we were surprised at the minimal amount of contradictory evidence. Of course, while almost all of our figures point us in the direction predicted, the magnitude of our relationship is not spectacular. But, we never ex­ pected it to be. As we developed in Chapter III, we expected the power structure in terms of concentration and dispersion to have some effect on war, but certainly not a dominant effect. This is exactly what our quantitative analysis shows. The effect is not dominant, but it is there. And what is more important, it is there in almost the exact manner that we predicted it would be. Even if one is un­ comfortable with, unfamiliar with, or just wishes to ignore the in­ dices and statistics, the evidence is still there; particularly in the fact that during the dispersed period the major powers fought each other often and with little chance of precipitating a global war.

In the concentrated period this was not so.

171 172

There are, however, valid grounds for not accepting this

evidence. One could very reasonably assert that the world is just

too complex to be dealt with by bivariate analysis. Also, one could

question the significance of results derived from so short a time

period. But we recognize these difficulties and present the quan­

titative results of the last chapter not as proof of our theory but

merely as an additional piece of evidence to supplement the non-

quantitative evidence which was integrated into our theory in the

first place.

If this evidence does not close the case on the effects of

concentration and dispersion on interstate war, then where can one go

for additional evidence? We can think of a number of places. At the

international level a researcher might try to piece together the

structure of the international system back into the 18th and 1 7th

centuries, and then run the same sort of experiment that we have run.

The structural data would be difficult to obtain, however. Also, at

the international level one might examine our predictions regarding process change. This evidence would be somewhat indirect, but at least the data is probably available to empirically determine the effects of concentration and dispersion on alliance patterns and activity, capacity for collective action, response to change, and crisis frequency and management, just to mention a few areas. At the national level, one might study cross-nationally the effects of domestic power concentration and dispersion on the amount and pattern of domestic violence. At the small group level, controlled ex­ periments might be designed directly to measure the relationships we 1 7 3 have posited. So while we have not closed the case or exhausted all possible sources of evidence, we feel that we have formulated the theory and demonstrated its probable validity at least to a degree that other researchers can examine the relationships posited with a high expectation of positive results.

In this study we rather ambitiously set out to shed some light on a relationship which has attracted the interests of scholars for many centuries. While this work will not end the debate over whether concentrated or dispersed power systems are more capable of solving the problems of war and peace, we do feel that we have made the relationship between power structure and interstate war somewhat more clear than it has been previously. Our theoretical technique was to reduce the relationship to its most basic concepts and then to read as broadly as possible in as many disciplines as have worked with these basic concepts. Hopefully, therefore, we would be able to gather all the pieces to the puzzle, and not just those few p. ces available in any one discipline. Having gathered the pieces, we then tried to put the puzzle together. Somewhere in this process the pieces did fit together, and a vast amount of outwardly conflicting evidence and theory began to make sense. It made sense not as contrasting theories to be "tested" against each other, but as a unified theory.

Not only did it make sense in terms of the thoughts and evidence available, but it proved to make sense in terms of the original evi­ dence which we gathered in this study.

Perhaps this study is too elaborate in light of the simple truth that it seems to have reaffirmed— that rarely if ever does one I?1*

get something for nothing. Every benefit has a cost. Sometimes the

cost is hidden, but it is there. Those who see only the benefits of

any structure may lead the rest joyfully onward for a while, but

sooner or later the cost catches up and someone pays. More often

than not in the international arena, the cost is paid in part

immediately by hundreds of thousands of noble souls who sacrifice

themselves for the sake of a future good, a future utopia, which

never comes.

The conclusions of this work are no different than the theory.

There is no structural utopia, at least in terms of power structure.

Not only do the results of this study show this, but elsewhere one finds a great number of obvious pieces of evidence which might tell

one that this is true. At the national level, for instance, we have

every manner and type of power distribution from pluralist democracy to dictatorship. Yet in none of these types of systems has violent con­ flict ever been eliminated. It may take different forms, but it is still there. The state government may do the killing during periodic purges in one case, while in another the citizens are allowed to con­ tinuously slaughter each other on the streets, The violence Has not stopped in any system for structural reasons at the national level, and we see no structural reasons why it should stop at the inter­ national level. Perhaps these people tend to see a structural utopia because they fix their minds on the benefits and ignore the costs of one particular type of system or another.

Our conclusions are not really as grim as some might take them to be. While we say that there is no structural solution to war in terms of power distribution, we do not say that there is no solution.

While the magic structural formula does not appear to exist, perhaps a formula involving the thoughts and attitudes of men does in fact exist. But that formula is not, of course, within the scope of this work. This work was undertaken to study and establish what we con­ sider to be an extremely important relationship as objectively and scientifically as possible. Hopefully this has been done successfully. BIBLIOGRAPHY

Abel, Theodore. "The Element of Decision in the Pattern of War." American Sociological Review. 6 (1941), 853-58*

Adelman, M.A. "Comment on the *H' Concentration Measure.1' Review of Economics and Statistics^ 51 (February, 1 9 6 9), 99-101.

Aiken, Michael. "The Distribution of Community Power* Structural Bases and Social Consequences." in Paul E. Mott and Michael Aiken, eds. The Structure of Community Power. New York* Random House, 1970» PP* 4Q7-5 2 5.

Alcock, Norman and Alan Newcombe. "The Perception of National Power." Journal of Conflict Resolution^ 14 (September, 1970), 335-43.

Alger, Chadwick. "The United States in the U.N." International Organization. 27 (VJinter, 1973)» 1-24.

Alker, Hayward R. and Bruce M. Russett. "Indices for Comparing In­ equality." in Richard Merritt and Stein Rokkan, eds. Comparing Nations* The Use of Quantitative Data in Cross-National Re­ search. Hew Haven* Yale University Press7 1966, pp. 349-72.

Aron, Raymond. Peace and War. New York* Doubleday, 1 9 6 6,

. "The Anarchical Order of Power." Daedalus. 95 (Spring, 1936J, 479-502.

Bachrach, Peter and Morton Earatz. "Two Faces of Power." American Political Science Review. 58 (1962), 947-52.

. "Decisions and Non-decisions* An Analytical Framework^' in Roderick Bell, ed. Political Power* A Reader in Theory and Research. New York* The Free Press, 1 9 6 9, pp. 100-09. r

Bain, Joe S, "Changes in Concentration in Manufacturing Industries in the United States, 1954-1966* Trends and Relationships to the Levels of 1955 Concentration," Review of Economics and Sta­ tistics. 52 (November, 1970), 411-16.

Banfield, Edward. "Concerting Action by Influence." in Roderick Bell, ed. Political Power. New York* The Free Press, 1 9 6 9, pp. 325-34.

176 177 Banks, Arthur S. Cross-Polity Time-Series Data. Cambridge, Massachusetts: The MIT Press, 1971.

Barter, Bernard. Social Stratification. New York: Harcourt, Brace and Company, 1957*

Barnet, Richard J. "Regional Security Systems." in Richard B. Gray, ed. International Security Systems: Concepts and Models of World Order. Itasca, Illinois* P.E. Peacock, 19f>9, PP* 75“92.

Beaton, Leomard. The Reform of Power: A Proposal for an Inter­ national Security System. New York: The Viking Press, 1972.

Bell, Roderick. "Political. Power: The Problem of Measurement." in Roderick Bell, ed. Political Power. New York: The Free Press, I9 6 9, pp. 13-27. Berelson, Bernard and Gary A. Steiner, Human Behavior: An Inventory of Scientific Findings, New York: Harcourt, Brace, and World, 1961k

Beres, Louis Rene. "Bipolarity, Multipolarity, and the Tragedy of the Commons." Western Political Quarterly. 26 (December, 1973)t 649-58.

______. "Bipolarity, Multipolarity, and the Reliability of Alliance Commitments." Western Political Quarterly. 25 (Decem­ ber, 1972), 702 -1 0 .

Berkowitz, Leonard. "The Study of Urban Violence: Some Implications of Laboratory Studies of Frustration and Aggression." American Behavioral Scientist, 11 (March-April, 1 9 6 8), 14-17.

Berlin, Isaiah "History and Theory: The Concept of Scientific His­ tory." History and Theory. 1 (i960), 1-31.

Bierstedt, R. "An Analysis of Social Power." American Sociological Review. 1 5 (December, 1950), 730-38. f Birrell, Derek. "Relative Deprivation as a Factor in Conflict in Northern Ireland." The Sociological Review, 20 (August, 1972), 317-^3. Bloomfield, Lincoln P. and Amelia C, Leiss. Controlling Small Wars* A Strategy for the 1970's. New York: Alfred A. Knopf, I9 6 9.

Boulding, Kenneth E. "Toward a Pure Theory of Threat Systems." in Roderick Bell, ed. Political Power, New York: The Free Press, 1969, pp. 285-92.

. "Equality and Conflict." Annals of the American Academy of Political and Social Science. "409"(September. 1-8. 178

Brams, S.J. "Measuring the Concentration of Power in Political Sys­ tems." American Political Science Review. 62 (19 68), *f6l-75.

Bronfenbrenner, Martin. Income Distribution Theory. Chicago* Aldine Atherton, 1971•

Brucan, Silva. The Dissolution of Power* A Sociology of Inter­ national Relations and Politics. New York* Alfred A. Knopf, 1971. Buchan, Alastair, Power and Equilibrium in the 1970's. New York* Frederick A. Fraeger, 1973*

Buehrig, Edward. "The International Pattern of Authority." World Politics. 17 (April, 1 9 6 5), 369-8 5.

Bull, Hedley. "War and International Order." in Alan James, ed. The Bases of International Order. London* Oxford University Press, 1973» pp. 116-3 2 . Carroll, Berenice A. "Peace Research* The Cult of Power." Journal of Conflict Resolution, 16 (December, 1972), 585“6l6.

Cartwright, Dorwin. "Influence, Leadership, Control." in Roderick Bell, ed. Political Power. New York* The Free Press, 1969, PP. 123- 65. Chertkoff, Jerome M. "Sociopsychological Theories and Research on Coalition Formation." in Sven Groennings, E.W. Kelly, and Michael Leiserson, eds. The Study of Coalition Behavior. New York* Holt, Rinehart, and Winston, 1970, PP« 279-322.

Clark, Terry N. Community Power and Policy Outputs * A Review of Urban Research. Beverly Hills* Sage Publications, 1973*

Claude, Inis L. Power and International Relations. 3^ edition. New York* Alfred A. Knopf7 19*&.

Collins, Barry E. and Harold Guetzkow. A Socii Psychology of Group Processes for Decision Making. New York* John Wiley and Sons, 19W. Connolly, William E, "The Challenge to Pluralist Theory." in William E. Connolly, ed. The Bias of Pluralism. New York* Atherton Press, 1969, pp. 3-3^« Cox, Robert and Harold R. Jacobson. "The Stratification of Power." In Robert Cox and Harold Jacobson, eds. The Anatomy of Influence. New Haven* Yale University Press, 1973, pp. 437-^3*

r 179 Dahl, Robert A. "The Concept of Power." Behavioral Science, 2 (1957), 201 -1 5.

Dakin, Ralph. "Variations in Power Structures and Organizing Efficiency: A Comparative Study of Pour Areas." Sociological Quarterly, (July, 1 9 6 2), 228-50.

Davies, J.C. "Toward a Theory of Revolution." American Sociological Review, 2? (1962), 5-1 9.

Davis, Kingsley. "The Demographic Foundations of Rational Power." in Monroe Berger, T. Able, and C.H. Page, eds. Freedom and Con­ trol in Modern Society. New York: Van Nostrand, 195^* PP- 206- W ~

Denton, Frank H. "Some Regularities in International Conflict, 1820- 19^9." Background. 9 (1966), 283 -9 6.

______and Warren Phillips. "Some Patterns in the History of Violence." Journal of Conflict Resolution, 12 (June, 1 9 6 8), 182-95. Deutsch, Karl. The Analysis of International Relations. Englewood Cliffs, New Jersey: Prentice-Mall, 1 9 6 6.

. "On the Concepts of Politics and Power." in James N. Rosenau, ed. International Politics and Foreign Policy: A Reader in Research and Theory. New York; The Free Press, 1 9 6 9, pp. 255-^0 ."

and Manfred Kochen. "Toward a Rational Theory of De­ centralization: Some Implications of a Mathematical Approach." American Political Science Review, 63 (September, 1969), 73^- W* and J. David Singer. "Multipolar Power Systems and Inter­ national Stability." in James N. Rosenau, ed. International Politics and Foreign Policy. New York: The Free Press, 1 9 6 9, PP* 315-2^. Douglas, William 0. International Dissent: Six Steps Toward World Peace. New York: Random House, 1971*

Dowty, Alan. "Conflict in War-Potential Politics: An Approach to His­ torical Macroanalysis." Peace Research Society Papers. (1 9 6 9), 85-95. East, Maurice A. "Status Discrepancy and Violence in the International System: An Empirical Analysis." in James Rosenau, Vincent Davis, and Maurice East, eds. The Analysis of International Politics. New York: The Free Press, 1972, pp. 299-319. 180

Eckstein, Harry. ”0n the Etiology of Internal Wars." in Ivo K. Feierabend, Rosalind L. Feierabend, and Ted R. Gurr, eds. Anger. Violence, and Politics; Theories and Research. Englewood Cliffs, New Jersey; Prentice Hall, 1972, pp. 9-30i

Ellis, Desmond, "The Hobbesian Problem of Order; A Critical Appraisal of the Normative Solution." American Sociological Review. 36 (August, 1971), 692-703. .

Elmandjra, Mahdi. The United Nations System; An Analysis. Hamden, Connecticut: Archon Books, 1973*

Emerson, Richard M. "Power-Dependence Relations." American Socio­ logical Review, 27 (February, 1962), 31-41.

Etzioni, Amitai. "Power as a Societal Force." in Marvin Olsen, ed. Power in Societies. New York; The Macmillan Company, 1970, pp. 1 8-2 7 .

. "Strategic Models for a De-polarizing World." in Clagett G. Smith, ed. Conflict Resolution; Contributions of the Behavioral Sciences^ Notre Dame, Indiana: University of Notre Dame Press, 1971» PP» 542-50.

Fain, H. "History as Science." History and Theory. 9 (1970), 154-73*

Falk, Richard A. "Law in Future International Systems." in Richard B, Gray, ed. International Security Systems. Itasca, Illinois: F.E. Peacock, 1 9 6 9, pp7 163-77*

. "Zone II as a World Order Construct." in James Rosenau, Vincent Davis, and Maurice East, eds. The Analysis of International Politics. New York: The Free Press, 1972, pp. 187-20^.

Feierabend, Ivo K. and Rosalind L. "Systemic Conditions of Political Aggression: An Application of Frustration Aggression Theory." in Ivo K. Feierabend, Rosalind L. Feierabend and Ted R. Gurr, eds. Anger. Violence, and Politics. Englewood Cliffs, New Jersey: Prentice Hall, 1972, pp. 1 3 6-8 3.

______and Ted Robert Gurr. "Theories of Revolution." in Ivo K. Feierabend, Rosalind L, Feierabend, and Ted R. Gurr, eds. Anger, Violence, and Politics. Englewood Cliffs, New Jersey: Prentice Hall, 1972, pp. 1-67

Feldman, Ronald A. "Power Distribution, Integration, and Conformity in Small Groups." American Journal of Sociology, 79 (1973)» 639-64.

Ferris, Wayne H. The Power Capabilities of Nation-States: Inter­ national Conflict and War. Lexington, Massachusetts: Lexington Books, 1973* 181

Fisherr Roger. "Fractionating Conflict." in Clagett G. Smith, ed. Conflict Resolution: Contributions of the Behavioral Sciences. Notre Dame, Indiana: University of Notre Dame Press, 1971, pp. 157-69.

Franke, Winfried. "The Italian City-State System as an International System." in Morton Kaplan, ed. New Approaches to International Relations. New York: St. Martin's Press, 1968, pp. 426-58.

Galtung, Johan. "A Structural Theory of Aggression." Journal of Peace Research, no. 2 (1964-), 95-119.

_ . "Entropy and the General Theory of Peace," Paper de­ livered at the second IPRA General Conference, Sweden, summer, 1967. . "On the Future of the International System." Journal of Peace Research, no. 4 (1 9 6 7)* 305-3** •

_ . "Small Group Theory and the Theory of International Relations: A Study in Isomorphism." in Morton Kaplan, ed. New Approaches to International Relations. New York: St. Martin's Press, 1968, pp. 270 -3 0 1.

. "The Social Sciences: An Essay on Polarization and Integration.” in Klaus Knorr and James N. Rosenau, eds. Con­ tending Approaches to International Politics. Princeton: Princeton University Press, 1969, pp. 243 -8 5.

Gamson, William A. Power and Discontent. Homewood, Illinois: The Dorsey Press, 1 9 6 8.

. "Reputation and Resources in Community Politics." In Paul E. Mott and Michael Aiken, eds. The Structure of Community Power. New York: Random House, 1970, pp. 235-75*

Gantzel, Klaus J. "Dependency Structures as the Dominant Pattern in World Society." Journal of Peace Research, no. 3 (1963), 203-15. Gort, Michael. "Analysis of Stability and Change in Market Shares." Journal of Political Economy. 71 (February, 1 9 6 3)* 51-63#

Grossack, Irvin M. "The Concept and Measurement of Permanent Industrial Concentration." Journal of Political Economy. 80 (July/August, 1972), 7^5-60.

Gulick, Edward Vose. Europe's Classical Balance of Power. Ithica, New' York: Cornell University Press, 1955* 182

Gurr, Ted R. "Urban Disorder: Perspectives from the Comparative Study of Civil Strife." American Behavioral Scientist. 11 (March- April, 1968), 50-55.

"Psychological Factors of Civil Violence." in Ivo K. Feierabend, Rosalind L. Feierabend, and Ted R. Gurr, eds. Anger. Violence, and Politics. Englewood Cliffs, New Jersey: Prentice Hall, 1972, pp. 31-57.

Haas, Ernst. "The Balance of Power; Prescription, Concept, or Propoganda." World Politics. 5 (1953J* 44-2-77•

Haas, Michael. "Societal Asymmetries and World Peace." Proceedings of the International Peace Research Association Conference, 2 TT9 3 5 7, 70-1 1 0 .

"International Subsystems; Stability and Polarity." American Political Science Review. 64 (March, 1970), 98-1 2 3 .

Hanrieder, Wolfram. "The International System: Bipolar or Multibloc?" Journal of Conflict Resolution. 9 (September, 1965)* 299-30 8.

Harf, James E., Bruce E. Moon, and John E. Thompson. "The Relevance of the International System for Explaining Foreign Policy." Paper delivered at the Midwest Political Science Association Annual Meeting, Chicago, April 26, 1974.

Harsanyi, John C. "Measurement of Social Power, Opportunity Costs, and the Theory of Two-Person Bargaining Games." in Marvin Olsen, ed. Power in Societies. New York; The Macmillan Company, 1970* PP* 226 -36.

Hawley, Amos H, "Community Power and Urban Renewal Success." in Paul E, Mott and Michael Aiken, eds. The Structure of Community Power. New York; Random House, 1970, pp. 464-73*

Hazleton, William A. "Kissinger on Globalism, Regionalism, and the Latin American Community." Paper delivered at the 1974 Annual Midwestern Regional Meeting of the International Studies Association, Chicago, November 21-23, 1974.

Healy, Brian and Arthur Stein. "The Balance of Power in International History: Theory and Reality." Journal of Conflict Resolution, 17 (March, 1973), 33-62.

Hinsley, F.H. Power and the Pursuit of Peace. Cambridge: Cambridge University Press, 1 9 6 3.

"On the Present State and the Future Development of the International System." World Politics, 20 (July, 1 9 6 8), 705-19.

r 183

Hobbes, Thomas. Leviathan, edited by Michael Oakeshott. New Yorkt Collier Books, 19f>2.

Hoeming, K.H. "Power and Social Stratification." Sociological Quarterly. 12 (Winter, 1971), 3-1^.

Hoffmann, Stanley. "Restraints and Choices in U.S. Policy.*' in his The State of Wari Essays on the Theory and Practice of Inter­ national Politics. New York: Frederick A. Praeger, 1966, pp. 160-97• . "International Systems and International Law." in his The State of War. New York: Frederick A. Praeger, 1966, pp. 88-122.

, "Rousseau on War and Peace." in his The State of War. Hew York: Frederick A. Praeger, 1965* pp. 5^-&7»

Hoivik, Tard. "Social Inequality— The Main Issues." Journal of Peace Research, no. 2 (1971)* 119-^2.

Holsti, K.J. International Politics: A Framework for Analysis. Englewood Cliffs, New Jersey: Prentice-Hall, 19^7*

Hopkins, Raymond F. and Richard W. Mansbach. "The Structural Bases of System Transformation." in their Structure and Process in International Politics. New York: Harper and Row, 1973* pp. 107-32. Hopkins, Terence. The Exercise of Influence in Small Groups, Totowa, New Jersey: The Bedminster Press, 19£?f,

Horvath, William J, and Caxton C. Foster. "Stochastic Models of War Alliances." Journal of Conflict Resolution.7 (1963), 110-16.

Jaguaribe, Helio. "World Order, Rationality, and Socioeconomic De­ velopment." Daedalus. 95 (Spring, 1966), 607-26.

James, Alan. "Law and Order in International Society." in Alan James, ed. The Bases of Internationa]. Order. London: Oxford University Press, 1973* pp. 6l-8*f. deKadt, E.J. "Conflict and Power in Society." International Social Science Journal. 17 (1 9 6 5), ^5^-71.

Kaplan, Morton. "Intervention in Internal War: Some Systemic SourceB." in James N. Rosenau, ed. International Aspects of Civil Strife. Princeton: Princeton University Press, 196*f, pp. 92-121.

______. System and Process in International Relations. New Yorkt John Wiley and Sons, 1967. Kara, Karel. "On the Marxist Theory of War and Peace." Journal of Peace Research, no. 1 (1 9 6 8;, 1-27.

Karlsson, Georg. "Some Aspects of Power in Small Groups," in Joan Criswell, Herbert Solomon, and Patrick-Suppes, eds. Mathematical Methods in Small Group Processes. Stanford* Stanford University Press, 1962, pp. 193-202.

Katz, Daniel and Robert L. Kahn. The Social Psychology of Organi­ zations. New York* John Wiley and Sons, i9 6 0,

Kaufman, Arnold S. "Human Nature and Participatory Democracy." in William E. Connolly, ed. The Bias of Pluralism. New York* Atherton Press, 19&9, PP* l?8-200.

Kissinger, Henry A. Nuclear Weapons and Foreign Policy, New York* Harper Brothers, 1957*

Knorr, Klaus, The War Potential of Nations. Princeton* Princeton University Press, 1956.

_ . Military Power and Potential. Lexington, Massachusetts* Heath Lexington Books, 1970*

Komhauser, Arthur. "'Power Elite* or 'Veto Groups'?" in Marvin Olsen, ed. Power in Societies. New York* The Macmillan Company, 1970, pp. 2&2-94.

Lagos, Gustavo. International Stratification and Underdeveloped Countries. Chapel Hill, North Carolina* University of North Carolina Press, 1963*

Lasswell, Harold and Abraham Kaplan. Power and Society* A Framework for Political Inquiry. New Haven* Yale University Press, 1950*

Leiserson, Michael. "Power and Ideology in Coalition Behavior* An Experimental Study." in Sven Groennings, E.W. Kelly, and Michael Leiserson, eds. The Study of Coalition Behavior. New York* Holt, Rinehart, and Winston, 1970r PP* 323-35*

Lindblom, Charles, "The Science of 'Muddling Through'." Public Ad­ ministration Review, 19 (Spring, 1959)» 79-88.

. The Intelligence of Democracy* Decision Making Through Mutual Adjustment. New York* The Free Press, I9 6 5.

Liska, George. "Continuity and Change in International Systems." World Politics, 16 (October, 1983)* 118-36.

Lynd, Robert S. "Power in Society as Resource and Problem." in Marvin Olsen, ed. Power in Societies. New York* The Macmillan Company, 1970, pp. 380-8S. 1 8 5 Masters, Roger. "A Multi-Bloc Model of the International System," American Political Science Review. 55 (December, 196l), 780-98.

Mattingly, Garrett. Renaissance Diplomacy. London* Jonathan Gape, 1955. Maurseth, Per. "Balance-of^-Power Thinking from the Renaissance to the French Revolution." Journal of Peace Research, no. 2 (1964), 120-33.

Mayhew, Bruce. "System Size and Ruling Elites." American Sociological Review, 38 (August, 1973)» 468-75*

Mayer, Lawrence C. Comparative Political Inquiry: A Methodological Survey. Homewood, Illinois: Dorsey Press, 1972.

McFarland, Andrew. Power and Leadership in Pluralist Systems. Stanford, California* Stanford University Press, 1969.

Mc-Gowan, Patrick J. and Howard B. Shapiro. The Comparative Study of Foreign Policy* A Survey of Scientific Findings. Beverly Hills, California* Sage Publications, 1973*

Mendlovitz, Saul H. "Models of World Order." in Richard B. Gray* ed. International Security Systems. Itasca, Illinois* F.E, Peacock, 1969* pp. 178-92.

Miller, John Perry. "Measures of Monopoly Power and Concentration* Their Economic Significance." in Business Concentration and Price Policy. National Bureau of Economic Research, Princeton* Princeton University Press, 1955. PP* 119-40.

Mills, C. Wright. "The Structure of Power in American Society." in Marvin Olsen, ed. Power in Societies. New York* The Macmillan Company, 1970, pp." 254-62.

Mitrany, David, A Working Peace System. Chicago* Quadrangle Books, 1966. • Modelski, George.. "War and the Great Powers." Peace Research Sooiety Papers. 1-3 (1972), 45-60.

Morgenthau, Hans. Politics Among Nations. 5th edition. New York* Alfred A. Knopf, 1973*

Mott, Paul E. "Power, Authority, and Influence." in Michael Aiken and Paul E. Mott, eds. The Structure of Community Power. New York* Random House, 1970, pp. 3“1^» 186 « "Configurations of Power," in Michael Aiken and Paul E, Mott, eds. The Structure of Community Power. New York* Random House, 1970, pp. 85-IOO.

______and Michael Aiken. "Locating Centers of Power." in Michael Aiken and Paul E. Mott, eds. The Structure of Community Power. New York* Random House, 1970, pp. 193-202,

Nagel, Jack. "Inequality and Discontent* A Nonlinear Hypothesis." World Politics. 26 (July, 197*0, ^53-^72.

Nieburg, H.L. Political Violence* The Behavioral Process. New York* St. Martin's Press, 1968.

. "Violence, Law, and the Social Process." American Behavioral Scientist. 11 (March-April, 1968), 17-19.

North, R.C. and N, Choucri. "Background Conditions to the Outbreak of the First World War." Peace Research Society Papers. 9 (1968), 12>137.

Northedge, F.S, "Order and Change in International Society." in Alan James, ed. The Bases of International Order. London* Oxford University Press, 1973, pp.' 1-2^.

______and M.J. Grieve. A Hundred Years of International Relations. New York* Frederick A, Praeger, 1972.

Olsen, Marvin, ed. Power in Societies. New York* The Macmillan Company, 1970*

Grganski, A.F.K, World Politics. New York* Alfred A. Knopf, 1968.

Orwell, George. "Triangular Perpetual War." in George Lanyi and Wilson McWilliams, eds. Crisis and. Continuity in World Politics. New York* Random House, 19*66, pp. 496-503.

Osgood, Robert E. and Robert W, Tucker. Force, Order, and Justice. Baltimore* Johns Hopkins Press, 1968.

Paige, Jeffery M. "Inequality and Insurgency in Vietnam* A Re­ analysis." World Politics. 23 (October, 1970), 2*1-37•

Parenti, Michael. "Power and Pluralism* A View from the Bottom." The Journal of Politics, (August, 1970), 5°1“30.

Parsons, Talcott. "The Monopoly of Force and the'Power Bank1." in Marvin Olsen, ed. Power in Societies. New York* The Macmillan Company, 1970, pp. 5*K58._

Platig, E. Raymond. International Relations Research. Santa Barbara, California* Cilo Press, 19 67. 187 Polsby, Nelson. Community Power and Political Theory. New Haven* Yale University Press, 1 9 6 3.

Pogue, Forrest, George C. Marshall* Organizer of Victory. New York* The Viking Press, 197V.

Prewitt, Kenneth. "From the Many Are Chosen the Few." American Behavioral Scientist, 13 (November-December, I9 6 9), 16 9 -8 7 •

Pryor, Frederic L. "An International Comparison of Concentration Ratios." Review of Economics and Statistics, 5^ (May, 1972), 130-40.

Reinton, Per Olav. "Inequality in International Systems of Nations." Peace Research Society Papers, (1 9 6 8), 47-55*

Rogers, Everett. "Social Structure and Social Change." American Be­ havioral Scientist. 14 (May-June, 1971), 767-8 2 .

RoBecrance, Richard N. Action and Reaction in World Politics* Inter­ national Systems in Perspective. Boston* Little, Brown, 1963*

. "Bipolarity, Multipolarity, and the Future." Journal of Conflict Resolution. 10 (September, 1 9 6 6), 314-27,

______. International Relations* Peace or War? New York* McGraw-Hill, 1973*

Rosenbluth, Gideon. "Measures of Concentration." Business Concen­ tration and Price Policy. National Bureau of Economic Re­ search. Princeton* Princeton University Press, 1955* pp. 57-99. Rossi, Peter H. "Power and Community Structure." in Michael Aiken and Paul E. Mott, eds. The Structure of Community Power. New York* Random House, 1970, PP* 110-1(5.

Rostow, Eugene. Law. Power, and the Pursuit of Peace. New York* Harper and Row, 1968.

Rothstein, Robert. "Alignment, Nonalignment, and Small Powers* 1945- 1 9 6 5." International Organization. 20 (1 9 6 6), 397“4l8.

Russett, Bruce M. "Inequality and Instability* The Relation of Land Tenure to Politics," World Politics. 1 6 (April, 1964), 442-54.

. "The Ecology of Future International Politics." International Studies Quarterly. 11 (March, 1 9 6 7), 12-32. 188

______. "Is There a Long-Run Trend Toward Concentration in the International System?" Comparative Political Studies, 1 (April, 1968), 103-122.

______. "Probabilism and the Number of Units Affected: Measuring Influence Concentration." American Political Science Review. 62 (1968), 476-80.

. "A Macroscopic View of International Politics." in James Rosenau, Vincent Davis, and Maurice East, eds. The Analysis of International Politics. New York* The Free Press, 1972, pp. .

______. "The Rich Fifth and the Poor Half: Some Spec­ ulations about International Politics in 2000 A.D." in his Power and Community in World Politics. San Francisco* W.H. Freeman, 1974, pp. 155-70.

Scitovsky, Tibor. "Economic Theory and the Measurement of Concen­ tration." in National Bureau of Economic Research. Business Concentration and Price Policy. Princeton* Princeton University Press, 1955*

Shapley, L.S. and Martin Shubik. "A Method for Evaluating the Dis­ tribution of Power in a Committee System." in Roderick Bell, ed. Political Power. New York* The Free Press, 1969, pp. 209-13.

Sibley, Elbridge. "Some Demographic Clues to Stratification." American Sociological Review. 7 (June, 1942), 322-30.

Simon, Herbert. "Notes on the Observation and Measurement of Power." in Roderick Bell, ed. Political Power. New York* The Free , Press, 1969, pp. 69-7 8.

Singer, J. David. "Inter-Nation Influence* A Formal Model." American Political Science Review, 57 (June, 1 9 6 3), 420-31.

. "The Relevance of the Behavioral Sciences to the Study of International Relations." Behavioral Science, 6 (1961), 324-35. ______. "The Level-of-Analysis Problem in International Relations." World Politics. 14 (October, 196l), 77-92.

______. "The Incompleat Theorist* Insight Without Evi- dence.ll in Klaus Knorr and James N. Rosenau, eds. Contending Approaches to International Politics. Princeton* Princeton University Press, 1969, PP. 62-86.

' ______, Stuart Bremer, and John Stuckey. "Capability Dis­ tribution, Uncertainty, and Major Power War, 1820 -1 9 6 5." in Bruce M. Russett, ed. Peace. War, and Numbers. Beverly Hills, California* Sage Publications, 1972, pp. 1948, 189 ______and Melvin Small. "The Composition and Status Ordering of the International System* 181,5-19^0." World Politics. 18 (January, 1966), 236 -8 2 .

. "Alliance Aggregation and the Onset of War, 1815“19^5«" in J. David Singer, ed. Quantitative Inter­ national Politics: Insights and Evidence. New York* The Free Press, I96S, pp. 24-7-86.

, The Wages of War. 1816-1965* A Statistical Handbook. New York: John Wiley and Sons, 1972.

Small, Melvin and J. David Singer. "Patterns in International Warfare, 1816-1965.1 Annals of the American Academy of Political and Social Sciences, 391 (September, 1970), 145-35*

. "The Diplomatic Importance of States, 1816-1970: ArT Extension and Refinement of the In­ dicator." World Politics. 25 (July, 1973)» 577-99.

Sprout, Harold and Margaret. An Ecological Paradigm for the Study of International Affairs; Research Monograph No. 30. Princeton * Center for International Studies, 19°8 .

Sterling, Richard W. Macropolitics * International Relations in a Global Society. New York: Alfred A. Knopf,197^K

Teune, H. and S. Synnestvedt. "Measuring International Alignment." Orbis. 9 (Spring, 1 9 6 5), 171-89.

Tumin, Melvin. Social Stratification: The Forms and Functions of Inequality. New York* Prentice-Hall, 1967*

. "On Inequality." in Melvin Tumin, ed. Readings on Social Stratification. Englewood Cliffs, New Jersey* Prentice-Hall, 1970, pp. *100-08.

Vanhanen, Tatu. 'Distribution of Power in the 1960's," Comparative Political Studies. 4 (January, 1972), 387-^5.

Vital, David. The Inequality of States * A Study of the Small Power in International Relations. Oxford* Clarendon Press, 1 9 6 7.

. The Survival of Small States^* Studies in Small Power/ Great Power Conflict. London* Oxford University Press, 1971*

Wallace, Michael. "Power, Status, and International War," Journal of Peace Research, (1971)* 23-35* 190

. "Status, Formal Organization, and Arms Levels as Factors Leading to the Onset of War, 1820-196*4-.M in Bruce M. Russett, ed. Peace, War, and Numbers. Beverly Hills, California: Sage Publications, 1972p pp* *49-6 9.

______. War and Rank Among Nations. Lexington, Massachusetts: Lexington Books, 1973*

. "Alliance Polarization, Cross-Cutting, and Inter­ national War, 1815-196*4." Journal of Conflict Resolution^ 1? (December, 1973)» 575-6

Walter, E.V. "Power and Violence." American Political Science Review. 58 (June, 196*4), 350-60.

Waltz, Kenneth If. Man, the State, and War: A Theoretical Analysis. New York: Columbia University Press, 1959*

______. "The Stability of a Bipolar World." Daedalus, 93 (Summer, 196*4), 081-909*

______. "International Structure, National Force, and the Balance of Power." Journal of International Affairs, 21 (1967), 215-31.

Weber, Max. "The Type of Authority and Imperative Co-ordination." in Marvin Olsen, ed. Power in Societies. New York: The Macmillan Company, 1970t PP* 35"39*

Woodcock, George. Anarchism: A History of Libertarian Ideas and Movements. New York: Meridian Books, 19o2^

V/right, Martin. "The Balance of Power and International Order." in Alan James, ed. The Bases of International Order. London: Oxford University Press, 19^3 * PP* 85-11 6 .

Wright, Quincy, A Study of War, 2d edition. Chicago: University of Chicago Press, 1955*

Zinnes, Dina. "An Analytical Study of the Balance of Power Theories." Journal of Peace Research, no. 3 (1967)* PP* 270-88.

"Coalition Theories arid the Balance of Power." in Sven Groennings, E.W, Kelly, and Michael Leiserson, eds. The Study of Coalition Behavior. Hew York: Holt, Rinehart, and Winston, 1970, pp. 351-68.

Zoppo, Cdro Elliott. "Nuclear Technology, Multipolarity, and Inter­ national Stability." World Politics, 18 (July, 1 9 66), 579-606.