War in the Modern Great Power System

1495–1975

Jack S. Levy

• Publisher: The University Press of Kentucky
• Year: 2021
• ISBN:
  9780813182070
• Topic: Politics & International Relations
• Subtopic: International Relations
• Index: Politics & International Relations

The apparently accelerating arms race between the United States and the Soviet Union and the precarious political conditions existing in many parts of the world have given rise to new anxiety about the possibility of military confrontation between the superpowers. Despite the fateful nature of the risk, we have little knowledge, as Jack S. Levy has pointed out, "of the conditions, processes, and events which might combine to generate such a calamity." No empirically confirmed theory of the causes of war exists, and the hypotheses—often contradictory—that have been proposed remain untested.

As a step toward the formulation of a theory of the causes of war that can be tested against historical experience, Levy has developed a unique data base that will serve as an invaluable resource for students of international conflict in coming years. War in the Modern Great Power System provides a much-needed perspective on the major wars of the past. In this thorough and systematic study, Levy carefully defines the Great Power concept and identifies the Great Powers and their international wars since the late fifteenth century. The resulting compilation of war data is unique because of its five-century span and its focus on a well-defined set of Great Powers.

Turning to a quantitative analysis of the characteristics, patterns, and trends in war, Levy demonstrates that although wars between the Great Powers have become increasingly serious in every respect but duration over the last five hundred years, their frequency has diminished. He rejects the popular view that the twentieth century has been the most warlike on record, and he demonstrates that it instead constitutes a return to the historical norm after the exceptionally peaceful nineteenth century. Applying his data to the question whether war is "contagious, " he finds that the likelihood of war is indeed highest when another war is under way, but that this contagious effect disappears after the first war is over. Contrary to the popular "war-weariness" theory, he finds no evidence that war generates an aversion to subsequent war.

This study, extending the scientific analysis of war back over five centuries of international history, constitutes a major contribution to our knowledge of international conflict.

1

Introduction: The Empirical Study of War

War has been a pervasive and persistent phenomenon throughout history. It is a major distinguishing characteristic of international politics and probably the most destructive form of human behavior. War is often perceived as a useful and sometimes necessary instrument of policy for the achievement of state political objectives, yet the avoidance of war without the sacrifice of other core values is a primary foreign policy objective of nearly all states. The buildup of national military capabilities is a constant preoccupation of statesmen but one that diverts significant resources from more constructive social pursuits, often contributes little to the security toward which it is directed, and may be a leading cause of the war it aims to avoid. For these and other reasons, the study of war has been a central concern of philosophers, historians, and social scientists since the beginning of recorded history.

In spite of the importance of war and the enormous human and economic resources devoted to its many aspects, our understanding of war remains at an elementary level. No widely accepted theory of the causes of war exists and little agreement has emerged on the methodology through which these causes might be discovered. Instead, the literature is characterized by a proliferation of competing and often contradictory theories. There have been relatively few efforts to subject these theories to rigorous and systematic empirical testing in an attempt to resolve their contradictions and contribute to the accumulation of scientific knowledge about international war. More emphasis has been placed on generating plausible theoretical arguments concerning why certain relationships ought to be true than on verifying empirically whether in fact they are true. An assumption underlying this study is that further progress toward understanding international conflict requires not only the careful specification of theoretical relationships but also the rigorous and systematic testing of these theories to ascertain their validity.

In the last two decades, however, there has been a proliferation of empirically based studies of international conflict.¹ Among the most rigorous and systematic are those based on the war data compiled by Quincy Wright, Lewis F. Richardson, and J. David Singer and Melvin Small.² The Singer-Small data and the associated Correlates of War Project have had an especially profound impact on recent conflict research. The data set published in The Wages of War, 1816-1965, is unparalleled and has served as the foundation for the statistical analysis of a wide range of important hypotheses regarding international conflict.³ The studies by Singer and Small as well as those based on Richardson’s data, however, are limited in that most of them focus on war behavior in general and fail to differentiate the Great Powers from the hundred or so other states in the international system.⁴ Consequently, their findings are not necessarily applicable to the Great Powers.

Few would doubt the importance of war involving the Great Powers. The Great Powers have traditionally been distinguished from other states and viewed as the dominant actors in international politics, particularly with respect to security-related issues. Statesmen have always been most concerned with security threats deriving from the Powers and have devoted disproportionate attention and resources to dealing with these threats. The uniqueness of the Great Powers has also been recognized in international law. They were differentiated from the lesser powers at the Congress of Vienna, for example, and were given a distinct set of rights and responsibilities.⁵ Scholars also have recognized the leading role of the great powers. Leopold Von Ranke conceived of the international history of Europe as a history of Great Power relations, and A. J. P. Taylor argues that “the relations of the Great Powers have determined the history of Europe.”⁶ Many of our theories of international politics are essentially theories of Great Power behavior. Balance-of-power theory, for example, was for years the dominant paradigm in international relations. Its central assumptions of anarchy, the absence of an external authority, and self-help are less valid when applied to lesser states living in the shadows of the Powers. Its concern with balances, the avoidance of war, and the independence of states clearly refers to the Great Powers rather than to states in general.⁷ Kenneth N. Waltz argues that any general theory of international politics must necessarily be based on the Great Powers, for in any system the leading actors essentially define the context for others as well as for themselves.⁸

Most important for the purposes of this study is the fact that the Great Powers have participated in a disproportionately high percentage of history’s wars. Quincy Wright finds that during the last five centuries four states have each participated in over 20 percent of European wars and that France alone participated in about 47 percent of the 2,600 battles involving European states.⁹ Even more compelling is the fact that approximately 70 percent of the wars Wright examines involve at least one of the Great Powers. Moreover, these states had a higher rate of war involvement during the years when they were Great Powers than when they were not.¹⁰ Similarly, Frederick Adams Woods and Alexander Baltzly, focusing on the proportion of years states have been at war since 1700, find that the strongest nations have devoted the most time to war.¹¹ The same pattern is true for more recent times, though to a slightly lesser extent. In Singer and Small’s compilation of all wars since 1816, 60 percent of the interstate wars and 75 percent of the “extra-systemic” wars involve a Great Power.¹² The wars of the Great Powers also account for most of the losses of life from war, with a significant fraction of these losses occurring in a small number of “general wars,” defined by the participation of nearly all of the Great Powers. These general wars have been major turning points in international history, marking the rise and fall of hegemonial powers and serving as the primary vehicle for fundamental transformations of the international system. Perhaps George Modelski exaggerates when he concludes that “war is primarily a Great Power activity,” that war occurs “because there are Great Powers,” and that “the Great Power system exists because there is war,”¹³ but the importance of the Great Powers for war and war for the Great Powers is undeniable.

Wars in which the Great Powers participate should be analyzed apart from wars in general because of the importance of the Great Powers and the distinctiveness of their behavior, including their war behavior. If Great Power wars are not analyzed separately, significant patterns of Great Power behavior may be obscured by noise generated by smaller states operating in more restricted regional systems. But analyses of the Great Power wars selected from the Singer-Small compilation would fail to produce meaningful statistical results because the compilation includes fewer than thirty interstate wars involving the Great Powers and fewer than ten wars between the Powers. To test hypotheses of Great Power war behavior by a methodology of aggregate data analysis, it is necessary to consider a longer time span.

Extension of the temporal domain is necessary for other reasons as well. Although the 1816-1980 period is adequate for the testing of many important hypotheses, its utility is limited for application to other theoretical questions. Some important variables—including structural characteristics of the international system that are central to balance-of-power theory and theories of systemic transformation—have changed little over the last century and a half; a meaningful analysis of the impact of these variables on war requires an examination of historical periods in which they fluctuated more widely. The nineteenth and twentieth centuries alone do not provide the variety and richness of data that can be obtained from a more extended period. For example, an analysis of the relative stability of bipolar and multipolar systems should not be restricted to an era characterized by only one period of bipolarity, especially when that period coincides with the development of nuclear technology, the confounding effects of which preclude the generation of any causal inferences. Similarly, an analysis of the impact of technology on the incidence of warfare is best served by incorporating preindustrial periods characterized by less advanced technology and lower rates of change. In addition to increasing both the number of cases and the range of the variables, the extension of the time span facilitates scientific analysis by incorporating additional control variables and by increasing the randomization of extraneous variables and hence minimizing their impact on the relationship under consideration.

Two other sets of questions also call for study of a more extended period. One is the question of long-term trends in war. Not only is the last century and a half too short for effective analysis, but the fact that the nineteenth century was the most peaceful of modern history introduces an upward bias into the secular trends. A more important set of theoretical questions is raised by the recent world system paradigms. Both the capitalist world economy paradigm of Immanuel Wallerstein and Christopher Chase-Dunn and the global political economy of George Modelski and William R. Thompson challenge the traditional realist theories based on power politics.¹⁴ These hypotheses involve processes spanning the entire five centuries of the modern period and can only be tested against realist propositions by using a more extended temporal domain than that employed by Singer and Small.

War data covering a broader historical scope have been compiled by Wright, Pitirim A. Sorokin, and Woods and Baltzly.¹⁵ Each of these compilations has serious limitations. None is specifically a study of the Great Powers, for all include some lesser states. Nor are any of their data bases alone sufficient for the purposes of this study. Neither Sorokin nor Woods and Baltzly provides explicit and rigorous criteria for the inclusion and exclusion of wars, and Wright’s criteria are excessively legalistic. These problems, which will be further discussed in Chapter 2, preclude a truly systematic study of the nature, causes, and consequences of war.

This brief survey of existing compilations of war data is not meant to detract from the enormous intellectual accomplishments that these works represent. The studies by Sorokin, Wright, Richardson, and Singer and Small are all critical landmarks in the evolution of modern scientific analysis of war. Each has made a significant contribution to our understanding of international conflict and the methodology by which we analyze it, and each has generated sophisticated theoretical and methodological approaches. But none of these studies deals adequately with the specific question of Great Power war behavior. For this purpose a separate analysis, based on a new set of data and drawing upon the efforts of these earlier scholars, is needed.

Such an analysis cannot be undertaken in the absence of a conceptual and historical examination of the Great Powers. Although the central role of the Great Powers has been widely recognized in the traditional literature, the concept has never been refined for systematic empirical research. As we shall see in Chapter 2, there have been several thorough analytical treatments of the Great Power concept and a few attempts to determine the identity of the Powers in historical systems.¹⁶ The latter, however, are seldom related to the former. Attempts to identify the Powers historically are generally conducted in the absence of any nominal or operational definition of the Great Power concept, which raises serious questions about the validity of these systems of Powers and their utility for systematic empirical research. In addition, there have been few attempts to specify the theoretical assumptions underlying the concept of a Great Power system. These tasks must be completed if theoretical generalizations about the Great Powers and their wars are to be possible.

The basic aims of this study, then, are to define and identify the modern Great Powers; to define, identify, and measure their wars; and to analyze the characteristics, patterns, and historical trends in these wars. In Chapter 2 the assumptions underlying the concept of the modern Great Power system are specified. After an extensive review of previous analytical treatments of the Great Powers, that concept is defined and the relevant empirical referents of Great Power rank are suggested. The origins of the Great Power system are established, and a historical analysis to determine the identity of the Great Powers over time is undertaken. In Chapter 3 we turn to the definition and identification of war. Considerable attention is given to the problem of inclusion and exclusion—the criteria used either to include specific wars or exclude them from the compilation. On the basis of these criteria, a list of interstate wars involving the Great Powers is generated and then compared with previous compilations. After conceptualizing war on a variety of levels, dimensions, and units of analysis, wars are measured in Chapter 4. The war data are presented graphically and analyzed statistically beginning in Chapter 5. The characteristics of the individual wars and of the yearly amount of war are examined. Empirical relationships among the various dimensions of war are determined and interpreted in light of a number of hypotheses in the literature. In Chapter 6 linear and cyclical trends in war are examined and a comparison of different historical eras is undertaken. Finally, the relationship between the amounts of war in successive periods is considered in Chapter 6 in order to deal with the question of whether war is “contagious.” Finally, a summary and conclusions are presented.

This is a univariate study of war in the modern Great Power system, and for this reason no effort will be made to analyze the causes of war or its consequences (other than its impact on subsequent war). The premise is that the Great Powers and their wars must be defined, identified, and measured before causes and consequences can be systematically analyzed. This study is designed to lay the groundwork for subsequent testing of key theoretical propositions relating to the causes and consequences of war in the modern Great Power system.

2

The Modern Great Power System

Before the Great Powers can be defined and identified it is necessary to specify the assumptions underlying the concept of a Great Power system. The Great Power framework and the realist paradigm from which it derives is only one of several possible approaches to the study of international relations. Other frameworks are based on a different set of assumptions and concepts, identify another set of leading actors, and offer alternative explanations for war, change, and other phenomena in world politics.¹ If these competing paradigms are to be compared and their constrasting propositions subjected to a critical test, the assumptions upon which they are based must be clearly specified.

Assumptions of the Great Power Framework

The Great Power framework shares the basic assumptions of the realist paradigm of international politics but focuses explicitly on the small number of leading actors in the system.² It is assumed that in any anarchic international system there exists a hierarchy of actors determined on the basis of power. In the modern system since 1500 the dominant actors have been dynastic/territorial states and nation-states; in the international system of ancient Greece and Renaissance Italy the dominant actors were city-states. The more powerful states —the Great Powers—determine the structure, major processes, and general evolution of the system.³ Therefore, the actions and interactions of the Great Powers are of primary interest. Secondary states and other actors have an impact on the system largely to the extent that they affect the behavior of the Great Powers. This hierarchy of actors is intimately related to a hierarchy of issues dominated by military security. It is assumed that issues overlap and that the currency of military power is applicable to and effective in the...

Table of contents

• Cover
• Half title
• Title
• Copyright
• Contents
• Preface
• 1. Introduction: The Empirical Study of War
• 2. The Modern Great Power System
• 3. Definition and Identification of the Wars
• 4. Measurement of the Wars
• 5. Quantitative Description of the Wars
• 6. Historical Trends in War
• 7. War Contagion
• 8. Conclusion: A Base for Further Investigation