Causes of War E-Book

Jack S. Levy

• Herausgeber: John Wiley & Sons
• Kategorie: Geisteswissenschaft
• Sprache: Englisch

Seitenzahl: 621

Veröffentlichungsjahr: 2011

Beschreibung

Written by leading scholars in the field, Causes of War provides the first comprehensive analysis of the leading theories relating to the origins of both interstate and civil wars.

• Utilizes historical examples to illustrate individual theories throughout
• Includes an analysis of theories of civil wars as well as interstate wars -- one of the only texts to do both
• Written by two former International Studies Association Presidents

Leseprobe

Contents

Acknowledgments

1. Introduction to the Study of War

What is War?

The Changing Nature of Warfare18

The Levels-of-Analysis Framework

Other Conceptual Issues in the Analysis of War

2. System-Level Theories

Realist Theories

Balance of Power Theory

Hegemonic Theories

3. The Dyadic Interactions of States

International Rivalries

The “Steps-to-War” Model

The “Bargaining Model of War”

Theories of Economic Interdependence and Conflict

4. The State and Societal Level

Marxist–Leninist Theories of Imperialism and War

Coalitional Models

The Diversionary Theory of War

The Democratic Peace

Culture and War: The “Clash of Civilizations” Thesis

5. Decision-Making: The Individual Level

Rational Models of Decision-Making

Psychological Models of International Conflict

The Poliheuristic Theory of Decision-Making

Crisis Decision-Making

6. Decision-Making: The Organizational Level

The Bureaucratic Politics and Organizational Process Models

The Causal Links to War

The Small-Group Level of Analysis

7. Civil War

Focusing on Opportunity

Systemic and Dyadic Influences

Back to Motivation?

Ethnicity and the Individual Level

Conceptual Issues

8. Conclusion: Reflections on Levels, Causes, and War

Reflections on the Levels of Analysis

Theoretical Generalization and Historical Interpretation

References

Index

This edition first published 2010

© 2010 Jack S. Levy and William R. Thompson

Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell.

Registered Office

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ,

United Kingdom

Editorial Offices

350 Main Street, Malden, MA 02148–5020, USA

9600 Garsington Road, Oxford, OX4 2DQ, UK

The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell.

The right of Jack S. Levy and William R. Thompson to be identified as the authors of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Library of Congress Cataloging-in-Publication Data

Levy, Jack S., 1948–

Causes of war/Jack S. Levy and William R. Thompson.

p. cm.

Includes bibliographical references and index.

ISBN 978-1-4051-7560-9 (hardcover: alk. paper)–ISBN 978-1-4051-7559-3 (pbk.: alk. paper)

1.       War–Causes. 2. International relations. I. Thompson, William R. II. Title.

JZ6385.L48 2010

355.02′7–dc22

2009030166

To students of war everywhere

Acknowledgments

A number of people provided valuable feedback at various phases of this project, and we want to acknowledge our appreciation. Five anonymous reviewers provided helpful comments on our initial proposal. Tim Knievel read the entire manuscript and offered detailed suggestions for improvements in both substance and style. Kim Marten gave us insightful comments on chapter 1 and provided additional feedback on questions arising in later chapters.

We also want to thank the people at Wiley-Blackwell for making this book a reality. Nick Bellorini, the Senior Commissioning Editor for Philosophy and Politics, provided the initial encouragement. Liz Cremona, the Senior Production Editor, supervised all phases of the project and kept the process running smoothly. Paul Stringer did a nice job of copyediting, and then flawlessly incorporated our last-minute changes into the final product. We thank them all for their patience in dealing with us throughout the process.

The authors would also like to thank each other for tolerating each other’s idiosyncrasies without escalating to warfare.

1

Introduction to the Study of War

War has been a persistent pattern of interaction between and within states and other political units for millennia. In its many varieties, it is probably the most destructive form of human behavior. War kills people, destroys resources, retards economic development, ruins environments, spreads disease, expands governments, militarizes societies, reshapes cultures, disrupts families, and traumatizes people. Preparation for war, whether for conquest or for protection, diverts valued resources from more constructive social activities, and it often undermines security rather than enhances it.

War also has a profound impact on the evolution of world politics and the behavior of states. Over the years it has been one of the primary mechanisms for change in the world system, through its impact on both the distribution of military power and wealth and the structure of the world economy. War also has a profound impact on the institutional structures and cultures of states, and it has played a key role in the birth and death of many states. We cannot understand the development of the modern nation-state system four or five centuries ago, or of earlier or more recent states, in the absence of patterns of warfare. As Tilly (1975:42) argued, “war made the state, and the state made war.”

It is hard to imagine what life would have been like in the late twentieth century in the absence of World War I and World War II, which had such profound effects on the global system and on domestic societies. The same can be said for the Cold War. For nearly a half century it shaped both international and domestic politics and cultures, not only in the United States and the Soviet Union but also in Western Europe and the Third World (Weart, 1989). The development of new states in the contemporary era continues to be influenced by warfare and preparations for war. With the proliferation of nuclear weapons, and with the threat of the acquisition of nuclear weapons by terrorist groups and “rogue states,” new threats to the security of even the most powerful states in the system have emerged.

The proliferation of civil wars and conflicts involving “non-state” actors has changed life throughout the developing world. A better understanding of the causes of war is a necessary first step if we are to have any hope of reducing the occurrence of war and perhaps mitigating its severity and consequences.

The unquestioned importance of war as a social phenomenon has led scholars, journalists, and others to devote enormous amounts of intellectual energy in attempt to better understand the nature of war and its causes. Ever since Thucydides (1996) wrote his History of the Peloponnesian War over 2,400 years ago in an attempt to explain the great war between Athens and Sparta (431–404 BCE), scholars from a wide range of disciplines–philosophy, history, political science, theology, anthropology, sociology, psychology, economics, mathematics, biology, literature, and others–have engaged the questions of what causes war and how humankind might eliminate war or at least bring it under greater control. Their efforts have led to a proliferation of theories but to no consensus as to the causes of war or of other forms of social violence.

Scholars disagree not only on the specific causes of war, but also on how to approach the study of war. It is not surprising that there are divisions between scholars in different countries (Wæver, 1998) and in different disciplines–that psychologists generally emphasize psychological factors, that economists emphasize economic factors, that anthropologists emphasize cultural factors, and so on. We also find enormous differences within each discipline. Scholars debate not only what the causes of war are, but also what theoretical approaches and methodologies are best suited to identifying those causes. The only consensus that seems to be emerging is that the question of the causes of war is enormously complex, although a minority of scholars question even that. Scholarly debate goes on, but the scourge of war continues.

The complexity of the question of the causes of war is compounded if we consider the many different forms of war. Most of the scholarly research on war since the time of Thucydides has focused on wars between states. Interstate wars dominated the study of war in political science until the last couple of decades, even though civil wars have actually been more frequent than interstate wars during most periods (Levy and Thompson, 2010b), and dramatically so in the last half century (Sarkees, Wayman, and Singer, 2003; Human Security Centre, 2005; Hewitt, Wilkenfeld, and Gurr, 2008). If we broaden our focus from interstate war to include civil war, colonial war, ethnic war, tribal war, and other forms of warfare, the question of the causes of war becomes even more complex. Although each of these forms of warfare shares some common elements (for example, the use of military force is usually seen as a strategy for advancing group interests), there are important differences as well.

Differences across types of war are particularly clear in the scholarly literature on interstate war and civil war. As we demonstrate later in this book, theories of interstate war emphasize fundamentally different factors than do theories of civil wars. To take just one example, the emphasis on the distribution of military power in the international system that is so common in discussions of the causes of interstate war, particularly great power wars, is given relatively little emphasis in theories of civil war. Similarly, the key variables of levels of economic and social welfare, which are critical in much of the literature on civil war, are given much less attention by scholars who study interstate war. As a result, most of the contemporary literature on war focuses either on interstate war or on civil war, but not on both.

With the changing nature of warfare, we believe that no general book on war is complete without some treatment of both interstate war and civil war. For that reason we break with the scholarly norm and include discussions of both. Still, we give most of our attention to interstate war, for a variety of reasons. More than any other form of warfare, interstate wars have shaped the evolution of the modern international system. This has made them the central focus of scholarly attention and debate for several centuries. Thus the literature on the causes of interstate war is intimately tied to the literature on international relations theory that has developed during the past 60 years. It is only recently that international relations theorists have engaged the question of the causes and consequences of civil war. Prior to that, students of comparative politics had a monopoly on the study of civil war, and for many years their approach was more descriptive than theoretical. The fact that little consensus has emerged on the causes of interstate war is another strong argument for continuing to study it.

There are other considerations as well. Though it has been declining in frequency, interstate war continues to have a profound effect on the contemporary world. The United States has already fought two interstate wars in the first decade of the new century–against the Taliban government of Afghanistan in 2001 and against the Iraqi government in 2003–and each war evolved into an internationalized civil war in which the United States was deeply involved. The Iraq war contributed to enormous human costs, a significant decline in US prestige around the world, political divisions at home, and economic costs that contributed to its declining economic fortunes. As we write in summer 2009, analysts debate whether it will be possible for the US to win the ongoing internationalized civil war in Afghanistan. Elsewhere, the Russian–Georgian war of fall 2008 signaled a renewed Russian assertiveness in international politics and sent shock waves through the West.

In addition, a brief survey of the world suggests a number of “flash points” that could trigger an interstate war, and some of these carry a significant risk of escalation to a broader conflict. We have almost certainly not seen the end of Palestinian–Israeli conflicts, which recently led to short wars in Lebanon involving Hezbollah in 2006 and in Gaza involving Hamas in 2008–09, and the potential for one of these conflicts to draw in other Arab states cannot be discounted.

The prospective proliferation of nuclear weapons involves other possible flash points. When Israel suspected that Syria was in the early stages of developing a nuclear program, it launched a limited preventive strike against a Syrian facility in September 2007.1 In response to the development of Iran’s nuclear program, which most observers believe is within a few years of becoming operational, and to Iranian President Ahmadinejad’s open call for the destruction of Israel, Israel has threatened to launch a preventive strike against Iran. The United States strongly prefers non-military means of preventing Iran from acquiring nuclear weapons, but it has thus far refused to take the military option “off the table.”

One also thinks of the Indo–Pakistani rivalry, which has already led to three major wars in the past 60 years (1948, 1965, 1971) and which is increasingly dangerous because both sides have nuclear weapons and because of domestic instability within Pakistan. The rivalry led to a war over Kargil in 1999 and to high levels of tensions at other times, including after the deadly terrorist attack on Mumbai in 2008. It is known that Pakistani citizens led the attack, and India charges that Pakistani security forces trained and equipped the terrorists. Other possible danger points are located in the Far East. One is the Korean Peninsula, with a nuclear-armed North Korea often acting in unpredictable ways. Still another danger is the dispute between China and Taiwan over the political status of the latter, which has enormous implications for US–China relations, particularly in the context of the possibility of a “power transition” involving the ascendancy of China over the United States within a few decades.

Thus while interstate war is not likely to be the most frequent form of warfare in the upcoming years, it has the potential to be the most destructive in human and economic terms. A war involving advanced nuclear states could be the most catastrophic war in history and fundamentally change human life as we have known it. Thus we devote most of our attention to interstate war, while reserving some attention to the phenomenon of civil war, which continues to occur on a regular basis.

We proceed as follows. In the rest of this chapter we provide a theoretical and historical context for our study of the causes of war. We define war and identify some of its primary characteristics. We then attempt to describe the changing nature of war over time, in order to put our extensive treatment of interstate war and briefer discussion of civil war in a broader historical context. Next we summarize the levels-of-analysis framework that we use for organizing our survey of the causes of war. Then in subsequent chapters we examine some of the leading theories of interstate wars. Our aim is not to present our own theory of war, but rather to survey some of the most influential theories advanced by scholars over the years and to point out some of the limitations of each of those theories. We give particular attention to the theories developed by international relations scholars in political science, but we also include important theoretical work from other disciplines as well.

What is War?

If our aim is to explain the causes of war, we must begin with a brief definition of the subject of our inquiry. We define war broadly as sustained, coordinated violence between political organizations.2 Such a definition includes great power wars like World War I, colonial wars like those fought by the European great powers in Africa and Asia from the eighteenth century to early twentieth century, civil wars like those in the United States in the nineteenth century or in the Congo or in Yugoslavia in the 1990s,3 organized insurgencies like the one against American forces in the Iraq War, tribal wars among pre-modern societies, and a wide variety of other forms of violence. This definition has several component parts, and it would be useful to examine each of them individually.

First, and most obviously, war is violent. It involves the use of force to kill and injure people and destroy military and economic resources. The German military theorist Carl von Clausewitz ([1832]1976:89) ended the first chapter of his famous book On War by identifying “primordial violence” as the first element of a “trinity” of “dominant tendencies” of warfare.4 That violence has the potential to be quite extreme. Earlier in the same chapter Clausewitz argued (p. 77) that “war is an act of force, and there is no logical limit to the application of that force.”

The element of violence in warfare separates it from other forms of intergroup and interstate conflict. Conflicts of interests–over power, territory, resources, and more symbolic issues–are common in world politics. Rivalries involving sustained and hostile competitions between actors are also common, as are threats of force by actors in an attempt to resolve disputes in their own favor.5 But conflicts of interests, rivalries, disputes, and threats of force do not become a war unless they involve sustained violence. The “Cold War” between the United States and the Soviet Union was a rivalry, not a war.6 Indeed, one of the distinctive features of the Cold War was the fact that the US–Soviet rivalry, unlike most previous rivalries between the leading states in the system, did not escalate to war. This is something that many scholars have spent a great deal of time trying to explain, with little agreement (Gaddis, 1987).

To take another example, the Arab–Israeli conflict goes back to the founding of the state of Israel in 1948 and beyond. Yet we would not describe it as a continuous war. Rather, it is a conflict or rivalry that has involved frequent low-level military activity, including armed incursions across borders and subsequent retaliations, but that has also been punctuated by a number of well-defined wars. The most prominent of these are the wars of 1948, 1956, 1967, 1973, and 1982, though we would probably also include as wars the Israeli conflicts in Lebanon in 2006 and in Gaza in 2008–09. The point is that conflicts of interests and rivalries are fairly common, whereas wars are not. Explaining why some rivalries, conflicts, or disputes lead to war while others do not is an important question. This makes it all the more important to define war as a separate concept, distinct from conflict or rivalry.

Another component of our definition of war involves the apparently innocuous word that follows violence in our definition–“between.” Yet this element of the definition is far from trivial. It indicates that violence must be reciprocated for it to qualify as war. A war is between two political organizations. If the target of the initial violence does not fight back, we do not normally call it a war. The Hungarian army forcibly resisted the Soviet invasion in 1956, and consequently scholars refer to the violent struggle that followed as the Russo–Hungarian War (Singer and Small, 1972). The Czechoslovakian army did not forcibly resist the Soviet invasion in 1968, and consequently we describe this as the Soviet invasion of (or intervention in) Czechoslovakia, but not as a war. To take another example, in 1981 Israel bombed an Iraqi nuclear reactor, with the aim of destroying the facility before it could become operational. Iraq did not respond militarily, in part because it was already engaged in a war with Iran. For that reason, scholars refer to the Israeli action as a preventive strike but not as a war.

Thus we treat war as the joint outcome of the behavior of two or more actors. In an alternative use of the concept, scholars sometimes talk about war as a strategy rather than as an outcome (Vasquez, 1993: chap. 1). Here the question is why a state or other political organization adopts a strategy involving the substantial use of military force rather than some other strategy. In speaking of war as a strategy, it is generally assumed that military action will be resisted. If it is not resisted, however, most scholars would not refer to the outcome as a war.

This brings us to the actors who engage in war. The actors are organizations, not individuals. Individuals do the actual fighting, but they fight on behalf of a larger collective political unit, under the direction and coordination of political and/or military leaders, to advance the goals of the collectivity, or at least of its leadership. An individual who acts on his own to kill a border guard, or who crosses a border to kill citizens of another political system, is not engaging in war. But if that individual is part of a political system’s formal military organization, and that military organization engages in a sustained campaign of violence against the military organization of another state or another organized group, we would call it a war.

Most books on the history of war in the modern era (which historians date from about 1500 on) focus on interstate wars, with particular attention to interstate wars between the great powers, the most powerful states in the system.7 These wars were the primary focus of Clausewitz ([1832]1976), who wrote after the experience of the French Revolutionary and Napoleonic Wars (1792–1815) and who emphasized the importance of major battles between the armies of the leading states in the system.8 Interstate wars, however, constitute only one manifestation of the wide variety of sustained, coordinated violence that we observe over the millennia.

In addition to fighting other states in interstate wars, states fight domestic challengers in internal or civil wars for the control of the state or for secession from the state. Those domestic challengers may fight each other. States may also fight non-state entities in their external environments, as illustrated by the current US wars against al Qaeda and against the Taliban insurgency in Afghanistan, and by the frequent armed conflicts between the state of Israel and the Palestinian authority and other non-state actors such as Hezbollah and Hamas. Wars may involve many of these elements simultaneously. The Iraq War started out as an interstate war (between the United States and the Iraqi government of Saddam Hussein) but then involved a domestic insurgency against an external state (the US), a civil war (between Shia and Sunni) for the control of Iraq, a war for secession from–or at least independence within–Iraq (by the Kurds), and international intervention in the civil war by state and non-state actors (the United States, Iran, and al Qaeda).

We must also remember that the nation-state, or even the broader category of the territorial state, is a relatively modern phenomenon. Before the rise of the state in early modern Europe, life was organized around kings and nobles, before that around “city-states,” and long before that around looser forms of social organization, including agricultural communities and groups of hunter-gatherers.9 During each of these periods organized violence between groups was fairly frequent. It differed in many respects from organized violence in later eras, but one thing that much of that violence had in common was that it involved the sustained, coordinated use of armed force by one political organization against another.10 We define war broadly enough to include those phenomena.

Thus far we have said nothing about the purpose of violence. Although political leaders’ motivations are not technically part of our definition of war, implicit in our discussion is the idea that violence is usually driven by a purpose. The political organization, as represented by its authoritative leadership, has goals, and one of the strategies they sometimes adopt in pursuit of those goals is the use of force. The purposeful nature of violence was most famously captured by Clausewitz ([1832]1976: 87), who repeatedly emphasized that war is a “political instrument, a continuation of political activity by other means. … The political object is the goal, war is the means of reaching it, and means can never be considered in isolation from their purpose.”11

A good example of an appreciation of the Clausewitzian view of the fundamentally political nature of war is an exchange between an American colonel and his North Vietnamese counterpart a couple of years after the end of the Vietnam War, which was widely regarded as a major defeat for the United States. The American colonel stated that, “You know you never defeated us on the battlefield.” The North Vietnamese colonel replied, “That may be so, but it is also irrelevant” (Summers, 1984:21).

It is the diplomatic and political outcomes of war that are important, and they are not always congruent with military outcomes on the battlefield. Egypt was in a stronger diplomatic position after the 1973 Arab–Israeli War than it was before the war, even though it was on the verge of a major military defeat at the end of the war until the United States forced Israel to withdraw its forces rather than crush the Egyptian army that it had surrounded. Egypt was militarily defeated but politically successful in the 1973 war.12

When political actors resort to military force, the goal is usually to influence the adversary’s behavior in ways that advance their own interests. As Clausewitz ([1832]1976:75) emphasized on the first page of On War, “War is thus an act of force to compel our enemy to do our will.” War is fundamentally coercive, driven by the aim of influencing the behavior of other actors. The Greek historian Polybius recognized this nearly two millennia before Clausewitz wrote, when he stated in his Histories (second century BCE) that “It is not the object of war to annihilate those who have given provocation, but to cause them to mend their ways.”

Sometimes the immediate goal of the use of force is not to influence the enemy’s behavior directly but instead to destroy or weaken his military forces or economic resources.13 This is usually an instrumental strategy, however, since weakening the adversary militarily and economically reduces its future battlefield performance and therefore its coercive bargaining leverage. In their use of force and conduct of war, state leaders aim to change the adversary’s expectations of the outcome of the war if the war were to continue, and presumably to make the adversary more willing to make extensive concessions to avoid that outcome.

In most cases, of course, political leaders would prefer to achieve their goals through non-forceful means, including diplomacy and economic pressure, which are generally less costly and less risky than the use of military force. Political leaders may use threats of force to reinforce their demands, but they generally prefer that their adversaries comply with those threats and concede what is demanded, so that the actual use of military force is unnecessary. In fact, the most effective uses of military power are often found in those situations in which military force is not actually used but where the mere threat of force is sufficient to change the adversary’s behavior. Deterrence, or the dissuasion of an adversary from taking an action that would be harmful to one’s own interests, is a good example. If the adversary is unwilling to make sufficient concessions, however, and if political leaders are convinced both that they can achieve more through military force than through negotiation and that they have no other option that would work as well, then the use of force often becomes an attractive option.

It is sometimes argued that diplomacy stops when war begins, that diplomacy and military force are two alternative strategies for preserving or advancing state interests. That view is quite misleading. The use as well as the threat of force is often an integral part of an actor’s bargaining strategy. It is a highly coercive activity, aimed at influencing the cost–benefit calculus of the adversary and persuading the adversary to change its behavior. The goal is to convince the adversary that the costs of continuing the war will be sufficiently great that it is preferable to make concessions now through a negotiated settlement. Referring to the subtitle of a recent book, this is “bargaining with bullets” (Sisk, 2009). The American use of the atomic bomb against Japan in 1945, for example, was driven by the goal of coercing Japan to end the war quickly, by sending a signal that additional violence would follow if Japan did not surrender. US leaders wanted to avoid the casualties that would be involved in the prolonged warfare that would otherwise be necessary to defeat the Japanese army. Thus diplomacy and force are often inseparable. As Frederick the Great of Prussia is widely reputed to have said, “Diplomacy without force is like music without instruments.”

This argument about the coercive nature of military force applies to nearly all political organizations, including terrorist groups. Terrorism against Israel is almost always motivated by the goal of imposing high enough costs on Israeli society to convince Israeli leaders that the benefits of occupying Arab territories are exceeded by the costs of doing so and that Israel would be better off by changing its policies. In initiating attacks against US military barracks and naval vessels overseas and against the World Trade Center in New York City, al Qaeda had many political goals, including using the threat of further terrorist attacks to try to persuade the US to remove its troops from Saudi Arabia and to reduce its support for other conservative Arab regimes (Pape, 2005b).14

Although we have emphasized that the use of military force is generally purposeful, we have not formally incorporated that into our definition of war. This contrasts with the approach of scholars like Malinowski ([1941]1968:523), an anthropologist who defined war as an “armed contest between two independent political units, by means of organized military force, in the pursuit of a tribal or national policy.” Our argument is that cases of sustained, coordinated violence between political organizations that are not driven by a clear sense of the political interests of the organization, but instead by personal or domestic political interests or perhaps by an insubordinate military leader, still qualify as wars. Our definition of war is based on the behavior of two adversarial political organizations, not on their motivations. In most cases, we believe that the use of military force is purposeful, but that is ultimately an empirical question rather than a definitional one.15 Identifying the motivations behind the use of force is a key task in explaining the causes of a particular war.

Finally, let us turn to the “sustained” element of the definition. Our aim is to differentiate war from organized violence that is more limited in its magnitude or impact. A minor border incident involving opposing armies may result in casualties on one or both sides, but we want to preserve the term “war” for those incidents that escalate and cross a certain threshold of violence. Border clashes between Chinese and Indian forces in 1962 continued to escalate and involved sustained fighting, and we refer to the “Sino–Indian War.” Border clashes between Chinese and Soviet forces over disputed areas around the Ussuri River occurred in March 1969 and then again six months later, but successful crisis management soon ended the crisis without further escalation. Thus we generally refer to that conflict as a “border clash” rather than a war (A. Cohen, 1991). States can mass armies on their borders for weeks or months, as each side attempts to demonstrate its resolve while at the same time seeking some formula for de-escalating tensions. The Indo–Pakistani “Brasstacks crisis” in 1986–7 is a good example (Ganguly 2002:85–8). Unless such an incident involves the sustained use of violent force, however, it does not constitute a war.

The question is what threshold of violence to use. Some scholars use the criterion proposed by the “Correlates of War Project” (Singer and Small, 1972). The “COW” project requires at least 1,000 battle-related deaths among all participating states and an annual average of 1,000 battle deaths for wars lasting more than a year. That criterion is quite reasonable for COW’s purposes of analyzing wars during the last two centuries. It is less useful for earlier periods when populations and armies were much smaller and when fewer battle deaths reflected a larger relative proportion of the army or of the population. Since we want our definition to apply to the organized violence between much earlier political systems as well as contemporary ones, we prefer a different criterion than battle deaths.16

Note that a precise (or “operational”) threshold is particularly important if the analyst is compiling lists of wars, which requires that s/he has explicit and replicable criteria for determining whether a violent conflict gets included in or excluded from a list of wars. We are not compiling a data set on wars, however, so more general criteria will suffice for our purposes. The main point is that our analysis of the causes of war is limited to those violent conflicts that cross some kind of threshold of magnitude. The fight-ing must be sustained rather than sporadic in order to differentiate war from “lesser” uses of military force. By sustained we mean not only duration but magnitude. There must be a fairly regular use of force of a certain magnitude during the period of the war.17

The Changing Nature of Warfare18

Human warfare has changed significantly over time. There is substantial evidence of warfare going back roughly ten thousand years to the beginning of agricultural societies (Keeley, 1996; Haas, 1999; Cioffi-Revilla, 2000), and growing evidence of war between hunter-gatherer groups before that (Gat, 2006), though archaeological evidence about warfare is more plentiful for the last 5,000 years (Ferrill, 1997). By that time there is evidence of full-fledged armies equipped with armor and organized into formations. Gradually, these armies became larger in size and more lethal in weaponry, and war became increasingly deadly. If we examine major battles, which admittedly are not representative of all wars, deaths per war more than doubled between the fifth century BCE and the fourteenth century CE, more than doubled again between the fourteenth and early nineteenth centuries CE, and then increased by as much as a factor of 10 between the early nineteenth and twentieth centuries (Levy and Thompson, 2010b).

This enormous increase in the severity of war, defined in terms of battle -related deaths, is countered by another trend, at least for the great powers over the past five centuries. There has been a steady decline in the frequency of great power war during this period, from about 22 in the sixteenth century to five in the nineteenth century and five or six in the twentieth century, depending on one’s precise definitions.19

These opposite trends for the last five centuries are probably related in a causal sense: the increasing destructiveness of great power wars has reduced the incentives of great powers to fight them. This may help to explain another interesting pattern: the world has experienced no great power war in the last half century. This is by far the longest period of peace between the great powers in the last five centuries of the modern era. Many scholars trace this absence of great power war to the development of nuclear weapons and their deterrent effects (Jervis, 1989), but other arguments have also been advanced (Gaddis, 1987; Kegley, 1991).

The absence of great power war for over half a century have led some to refer to this period as “the long peace” (Gaddis, 1987). This is quite misleading, since the period since World War II has witnessed a proliferation of smaller wars and other forms of armed conflict.20 Interstate wars have continued to occur, initially at about the same rate as in the period prior to 1945 (Levy and Thompson, 2010b), though in the last two decades the frequency of interstate war has begun to decline (Hewitt, Wilkenfeld, and Gurr, 2008).

One noticeable change in interstate war, however, is where these wars are fought. We do not have a perfectly reliable database on global wars, but what evidence we have suggests that for most of the last five centuries of the modern era a disproportionate number of interstate wars were fought in Europe (Wright, 1965:641–51).21 The global system was centered in Europe, and the world’s leading powers were all located in Europe until the beginning of the twentieth century. Those great powers fought each other, expanded by fighting weaker European states, and engaged in colonial wars throughout the world. Since 1945, however, we have witnessed a dramatic shift in warfare (both interstate and civil) away from Europe to other parts of the world (Singer, 1991). The Yugoslav wars of the 1990s were the first in Europe since 1945.

Another significant trend is a significant increase in the frequency of civil wars and other forms of intrastate conflict (K. Holsti, 1996).22 The ratio of internal to external wars increased from about two to one before 1945 to nearly five to one after 1945 (Levy and Thompson, 2010b). There was a particularly strong increase in the number of civil wars beginning in the 1970s after the period of decolonization, and civil wars continued at a relatively high frequency until the late 1990s. After that, there has been a decline in the frequency of civil wars. This pattern may be surprising given the constant images of warfare seen on the television and elsewhere, but it is well-documented (Hewitt, Wilkenfeld, and Gurr, 2008; Human Security Centre, 2005; Harbom and Sundberg, 2008). Whether this decline in civil wars and other forms of armed conflict is likely to continue is a source of considerable debate (Gleditsch, 2008).

These patterns suggest that there has been a shift in the nature of warfare over time–away from the great powers, away from Europe, and, increasingly, away from state-to-state conflict and toward civil war, insurgency, and other forms of intrastate and trans-state warfare. It is a kind of warfare that differs in many respects from the wars that have dominated the past five centuries of the modern international system. The wars of most interest to scholars have been interstate wars that were “symmetric” in the sense that the two sides were of roughly equal strength and fought with similar types of weapons. The primary actors were states that possessed a monopoly of legitimate force within their borders, a description that characterized most of the leading states of Europe by the mid-seventeenth century. This was the basis for Clausewitz’s ([1832]1976) image of war as militarized conflict between state armies, directed by state leaders on behalf of state interests, and resolved by decisive battles.23

With the increasing shift from interstate war to civil wars, and with the changing character of civil wars themselves, scholars have begun to question whether the conventional “Westphalian” model of warfare continues to be relevant for the contemporary era.24 Fewer and fewer wars involve conventional clashes of two opposing armies. The Russian–Georgian war of 2008 is a recent exception, though it was a highly asymmetric conflict.

Civil wars themselves have changed. Unlike the American civil war of the nineteenth century, the army of the state often faces not a single rebel army but instead a coalition of rebel soldiers representing different groups with different interests (Horowitz, 1985). Many of these central players in civil wars are ethnic or religious groups, and the wars are sometimes referred to as “ethnic wars” or “identity wars.” Some question, however, whether most of these wars are primarily about ethnicity or identity, or whether ethnicity and identity mask underlying conflicts that are driven primarily by security goals, economic resources, political power, or private interest (Gagnon, 2004).25 Warlords, aiming to protect or advance their own parochial interests, play a key role in many of these conflicts (Marten, 2006/07). Globalized criminal networks have also come to play a significant role in the funding of civil wars and insurgencies, and wars are often sustained by illicit black markets (Mueller, 2004; Andreas, 2008). Armies have increasingly “outsourced” many of their traditional functions, and in many respects wars have become more privatized (Avant, 2005).

Strategy and tactics have also changed, along with the norms of warfare. Warfare is increasingly “asymmetric.” Rebel groups are often outmatched by the state in organization and military technology, and they respond by adopting strategies of guerrilla warfare, insurgency, and terrorism. Tactics increasingly include the direct targeting of civilians, and massacre and ethnic cleansing have become more common. This has led some to talk of the increasing “barbarization of warfare” (Kassimeris, 2006). Most of this behavior is purposeful, driven by the aim of persuading and coercing people to shift their political loyalties by demonstrating that the state is unable to protect its citizens. Contemporary civil wars are rarely settled by decisive battles, but instead by protracted struggles.

This is the image of the “new wars,” which are often contrasted with the “old wars” of the Westphalian era (van Creveld, 1991; Kaldor, 1999; Münkler, 2004). A major debate has emerged, however, as to whether the “new wars” are really new, or whether elements of the new wars can be found in past historical periods (Kalyvas, 2001; Duyvesteyn and Angstrom, 2005; Malešević, 2008).26 That debate focuses more on how war is conducted than on the varied causes of war, which is our primary concern in this book, with primary attention to interstate war but with some attention to civil war, which we address in chapter 7.

The Levels-of-Analysis Framework

Any survey of the causes of war needs an organizing framework that helps to make sense of the many varied causes of war. We need a typology that groups similar causes together. One framework that many international relations theorists have found useful for the analysis of war and of foreign policy behavior is the “levels-of-analysis” framework. This framework goes back to Kenneth Waltz’s book Man, the State, and War (1959), which identified three “images” of war. These images referred to sources of causation associated with individuals, the nation-state, and the international system, respectively. Following Singer (1961), scholars began to refer to these images as “levels” of analysis. The levels-of-analysis framework is not a theory of war but instead a typology of the causes of war. More accurately and more generally, it is a framework for classifying the different causal factors influencing the policies and actions of states and of other actors.27

The individual level of analysis aims to explain the foreign policy decisions made by the political leaders of the state (or other political unit). It includes characteristics shared by all individuals, such as “human nature” and its hypothesized predispositions toward aggression. The individual level also includes factors that vary across individuals, including belief systems, personalities, psychological processes, political socialization, lessons learned from history, management styles, and similar variables. The presumption of individual-level theories is that the particular individual or individuals in power have an important causal impact. The implication is that if another individual had been in power the outcome might have been different. Many interpretations of World War II, for example, focus on German Chancellor Adolf Hitler, and argue that if Hitler had not come to power the war might have been avoided (Mueller, 1989).

The national level or nation-state level of analysis includes both factors associated with the government and factors associated with society. The former include variables like the institutional structure of the political system and the nature of the policy-making process, and the latter include variables like the structure of the economic system, the influence of economic and noneconomic interest groups, the role of public opinion, and political culture and ideology. At the national level, for example, there is considerable evidence that because of democratic institutions and political cultures, democracies behave differently than authoritarian regimes with respect to war. At a minimum, democracies rarely if ever go to war with each other (Doyle, 1983). At the societal level, one hypothesis is that some political cultures are more warlike than others, although many scholars have concluded that there is not much evidence to support this argument (Wright, 1965). Diversionary theory suggests that political leaders sometimes decide on war when they anticipate that war against an external adversary will increase their domestic political support by generating a “rally round the flag” effect. Certain governmental bureaucracies may push for higher military budgets as part of a strategy to increase their power and influence within the government, or domestic economic groups may push for more aggressive foreign policies because it serves their own parochial interests. Each of these factors would be encompassed by the nation-state level of analysis.

System-level causes include the anarchic structure of the international system,28 the number of major powers in the system, the distribution of military and economic power among them, the pattern of alliances, and other factors that are closely related to the distribution of power, including the structure of the system’s political economy.29 Most realist theories, including balance of power theory, are system-level theories, as are theories of hegemonic order and power transitions. The system level also includes other factors in the external environment common to all states, including the structure of international institutions, the nature of international norms, or system-wide ideologies or cultures.30

Waltz’s (1959) conception of three images or levels has been extremely influential in the study of international relations and foreign policy. While many scholars adopt Waltz’s three-level framework, others modify it. Following Rosenau (1966), some scholars disaggregate state (or governmental) and societal sources of causation into two separate levels. Jervis (1976: chap. 1) modifies Waltz’s framework by distinguishing the levels of decision-making, the bureaucracy, the state and domestic politics, and the international environment. Others simplify Waltz’s framework and identify two levels of causation, one internal to the state (which Waltz (1979) labels “unit level”) and one external.31

There is no single “correct” number of levels. Levels-of-analysis frameworks are analytic constructions to help us make sense of the world, and they are best evaluated in terms of their theoretical utility rather than seen as a direct reflection of “reality.” For the purposes of summarizing theories of the causes of war, we find it most useful to distinguish among theories that emphasize sources of causation at the system, state and societal, and decision-making levels of analysis, and to divide the latter into individual and organizational levels.32

We also introduce an additional level, commonly referred to as the “dyadic” or “interactional” level, which reflects the bilateral interactions between pairs of states. The past history of interactions between two states would be included in this category, as would territorial conflicts and bilateral bargaining between states. Some scholars include this factor in their system-level category, and thus define the system level broadly to include everything in a state’s external environment. We find it more useful to distinguish between causal variables that reflect the entire international system (polarity, for example) and those that reflect the relationship and interactions within a particular pair of states within that system.33 It is also useful, for certain questions, to distinguish between the international system and various regional systems nested within it. The Middle East system and the South American system have different characteristics and dynamics, though both exist within a single global system. Crisis dynamics between two states can be influenced by the structure of power in the global system, by the structure and culture of the regional system within which they interact, and by the characteristics and history of the dyad itself.

The levels-of-analysis question has important normative implications, particularly in terms of evaluating moral responsibility. If the primary causal factors leading to war or a state’s decision for war arise from systemic or dyadic-level threats to the national interest, so that any reasonable state or individual in that situation would have responded in roughly the same way, we would not ordinarily attribute moral responsibility for the war to that state or its leaders. Political leaders understand this, of course, and we often hear political leaders say, whether they had a choice or not, that “I had no choice.” Assessing the causal weight of various factors is an important step in evaluating blame, and differences in assessments of causality complicate efforts to attribute blame.

After World War I, for example, the victorious Western allies forced a defeated Germany to sign a “war guilt” clause in the Treaty of Versailles (1919). This may have just been victors’ justice, however, as within a decade many historians began to shift to the view that the primary causes of the war were more systemic, based on the system of power politics and secret alliances (Fay, 1928) and not on the actions of particular governments or states. After Fritz Fischer’s publication of Germany’s Aims in the First World War in 1961 and the English translation six years later, and Fischer’s argument that Germany’s aims went beyond security to world power, opinion on responsibility for the war shifted back.34 Political scientists are often less interested in questions of “war guilt,” or moral responsibility for war, than are historians (Schroeder, 2001), but it is clear that any such evaluation rests on an empirical analysis of the causes of the war.

It would be useful to illustrate our levels-of-analysis framework with respect to various explanations that analysts have proposed to explain the US decision to invade Iraq in 2003. Although some explanations emphasize causal variables at a single level of analysis, others combine variables from several different levels. Some argue that the US intervention was the product of President George W. Bush’s worldviews and religious beliefs, his determination to finish the job begun by his father in the 1990–91 Gulf War, or Bush’s confidence in the correctness of his beliefs or his disregard for information running contrary to his beliefs and policy preferences. These are all individual-level causal factors, which we discuss in chapter 5 on decision -making at the individual-level. The implication of these theoretical arguments is that if someone else besides Bush had been president, the probability of US military action would have been different.

Others attribute the US decision to the nature of the American political system and society. They emphasize the traditional US commitment to democracy and the promotion of democracy abroad, the impact of the September 11 attacks on American political culture and on public opinion (which created a permissive environment for an aggressive policy toward Iraq), the hesitancy of members of Congress to argue or vote against the war for fear of possible political repercussions,35 and the influence of the US oil industry or perhaps of the “Israeli lobby” (Mearsheimer and Walt, 2007) on US policy. Still others focus on decision-making at the bureaucratic/organizational level, and emphasize the influence of neoconservatives on the policy-making process, the influence of Vice-President Cheney and Secretary of Defense Rumsfeld and the political marginalization of Secretary of State Colin Powell from the inner circle of decision-making, and the flaws in an intelligence process (including the “politicization of intelligence”) that generated grossly misleading estimates about Iraqi weapons of mass destruction.36

Another set of interpretations argue that the US intervention was driven primarily by system-level threats and opportunities (at the regional as well as global level) and the calculations about the national interest related to them. They point to the George W. Bush Administration’s aim to destroy what they perceived as Iraq’s existing or developing weapons of mass destruction, which was the administration’s primary public rationale for the war. Other system-level causes include the impact of the 9/11 terrorist attacks on Americans’ perceptions of their vulnerability and on the assumed link between al Qaeda and Saddam Hussein; the aim of bringing democracy to Iraq or perhaps to the Middle East as a whole, both as an end in itself and as a means of enhancing US security by creating like-minded regimes; or the permissive conditions created by the collapse of Soviet power and the end of the Cold War over a decade earlier.

The levels-of-analysis framework is normally applied to states and to interstate relations. It can also be applied to the behavior of non-state actors, where it leads us to ask similar questions about the sources of causality. Does a particular ethnic group act the way it does because of external threats and opportunities in its environment, because of internal politics among various subgroups within it, or because of the particular beliefs and charisma of an individual leader? Similarly, are the actions of a terrorist group aimed primarily to advance the interests of the group as a whole, or are they the product of infighting between competing factions within it or of the beliefs and preferences of a particular leader?

The levels-of-analysis framework is useful for organizing the varied sources of conflict into categories that help simplify and impose some structure on the way we think about war and foreign policy more generally. Although we prefer this framework to others as an organizing device, we should acknowledge some of its limitations. Ideally, a typology should have categories that are both exclusive and exhaustive–causal factors should fit into one and only one category, and there should be some category for all causal factors. The levels-of-analysis framework (like most typologies) falls short of this ideal-type standard. The important factor of misperceptions, for example, can result from system-level uncertainty or adversary strategic deception, national-level ideologies that predispose leaders to interpret the behavior of others in certain ways, and individual-level personalities that contribute to further distortions in incoming information. Economic factors include both national economic interests such as the stability of a society’s economic system and the influence of private economic groups (e.g., arms manufacturers) on state foreign policies.

We should also note that the levels-of-analysis framework is better for classifying causal variables than for classifying theories. Although some theories incorporate variables from a single level of analysis (most psychological theories, for example), most theories combine variables from multiple levels of analysis. In these cases, we classify the theory based on the level of its variable of greatest causal weight. For example, although neoclassical realist theory incorporates domestic and individual-level variables, it gives primacy to the international system, and we classify it accordingly. Multiple-level theories sometimes complicate the use of the level-of-analysis framework. At the same time, however, by distinguishing among different levels the framework is useful in identifying the different kinds of factors that operate within a particular theory and how they interact with each other in the processes leading to war and peace.

Although causal variables at any level of analysis can be used to help explain individual beliefs, state behavior, and dyadic or systemic outcomes, we need to point out a potential logical problem associated with certain types of explanations for war. One concerns individual-level explanations for war. Although analysts often trace the outbreak of a particular war to the beliefs or personalities of a single individual (Hitler and World War II, for example), that does not constitute a logically complete explanation of war.

We have defined war in terms of the behavior of political organizations, whether the state, a rebel group, or a terrorist organization. War is institutionalized, not individual. This means that to explain a state’s decision to adopt a strategy of war, we need to know more than the preferences, beliefs, and personality of the leader. It is incumbent upon us to explain how the leader’s preferences, along with the preferences of other decision-makers, are translated into a foreign policy decision for the state. Since war is made by states, not individuals, an individual-level theory of war has to be linked to a broader theory of foreign policy.

It might seem that in a dictatorship such additional variables might not be necessary for a complete explanation for war, but in fact the dictatorial structure of the regime is part of the explanation. The politically centralized nature of the regime helps to explain how a leader who wants war actually gets war implemented as policy.37 In more decentralized regimes, including democratic regimes (especially parliamentary regimes), sometimes political leaders who want war are prevented from implementing that strategy by domestic constituencies or by the cabinet. Alternatively (but less frequently), political leaders who believe that war is contrary to the national interest and who prefer to avoid war are sometimes pushed into war by a xenophobic public opinion or by powerful domestic groups. US president William McKinley hoped to avoid war with Spain in 1898, but because of domestic pressures McKinley “led his country unhesitatingly toward a war which he did not want for a cause in which he did not believe” (May, 1961:189).

There is a second logical problem. War involves violence between political organizations. A theory of war must explain how both sides get to the brink of war in the first place and why both are willing to fight. Since war is a dyadic or system-level outcome resulting from the joint actions of two or more states, understanding the causes of war requires an explanation of the strategic interaction of the two (or more) adversaries. For this reason individual -, societal -, and state-level causal factors cannot by themselves provide a logically complete explanation for the outbreak of war. That is, they are not jointly sufficient for war. We need to include dyadic or system -level causal variables (a theory of bargaining, for example) for a complete explanation. This does not necessarily mean that dyadic and system-level variables have a greater causal influence than do individual or domestic variables, only that the former cannot be logically excluded from the analysis.

An example will help illustrate the point. Consider the hypothesis that the primary cause of a particular war is the existence of a state or political leader with particularly aggressive intentions. This hypothesis is not a logically complete explanation for war. Nazi Germany, for example, behaved quite aggressively in the mid-1930s. It violated international treaties by rearming, remilitarizing the Rhineland, annexing Austria, and demanding the incorporation into Germany of the Sudetenland. For several years, however, none of these actions led to war, because the West chose to pursue a policy of appeasement rather than stand up to Germany and thereby risk war. Many historians conclude that Hitler actually wanted war over Czechoslovakia, and was forced to abandon that goal, at least temporarily, when the West responded with enough concessions to make war politically infeasible (Taylor, 1961). War eventually came, of course, but only after the West changed course after further aggression by Hitler. A complete explanation for World War II, and for any war, requires that we not restrict the analysis to the behavior of a single individual or state.

Similarly, it is not enough to explain a peaceful outcome by showing than one side pursues a conciliatory policy. A strategy of extensive concessions often leads to a peaceful outcome, but it is also possible that it might create the image of weakness and lead the adversary to increase its demands in the hope of coercing further concessions. British and French appeasement of Hitler in the 1930s illustrates this point as well. Britain and France each sought peace, but peace was not the outcome because Hitler responded to their concessions with further aggressive moves and demands for further concessions. In fact, some argue that appeasement actually made war more likely, though that proposition makes the problematic assumption that a more confrontational policy would have deterred Hitler. Again, the broader theoretical argument is that theory of war requires a theory of bargaining or strategic interaction that explains how states respond to each other’s actions and how they act in anticipation of each other’s responses.

Other Conceptual Issues in the Analysis of War

Another conceptual issue relates to our earlier discussion of the changes in war over time, which reminds us that war is a variable, not a constant. War varies in terms of who fights, where they fight, how often they fight, with what intensity, and so on. Thus the primary phenomenon that most international relations scholars want to explain is variations in war and peace over time and space. Why does war occur between some states rather than other states, at some times rather than other times, under some conditions rather than other conditions, by some political leaders rather than other leaders. As Bremer (2000) asked, “Who Fights Whom, Where, When, and Why?”

The variable nature of war and peace has important but often-neglected implications for the study of war. Any theory that predicts that war is a constant must be rejected, or at least modified to include additional variables that explain variations in war. An “independent variable” that is constant cannot explain a “dependent variable” that varies.