Democracies at War

DEMOCRACIES AT WAR

DEMOCRACIES AT WAR

Dan Reiter and Allan C. Stam

PRINCETON UNIVERSITY PRESS

PRINCETON AND OXFORD

COPYRIGHT ©2002 BY PRINCETON UNIVERSITY PRESS

PUBLISHED BY PRINCETON UNIVERSITY PRESS, 41 WILLIAM STREET,

PRINCETON, NEW JERSEY 08540

IN THE UNITED KINGDOM: PRINCETON UNIVERSITY PRESS, 3 MARKET PLACE,

WOODSTOCK, OXFORDSHIRE OX20 1SY

ALL RIGHTS RESERVED

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

REITER, DAN, 1967–

DEMOCRACIES AT WAR / DAN REITER AND ALLAN C. STAM.

P. CM.

INCLUDES BIBLIOGRAPHICAL REFERENCES AND INDEX.

ISBN 0-691-08948-5 (ALK. PAPER) — ISBN 0-691-08949-3 (PBK. : ALK. PAPER)

1. DEMOCRACY. 2. WAR. I. STAM, ALLAN C. II. TITLE.

JC421.R4185 2002

355.02—DC21

2001019859

THIS BOOK HAS BEEN COMPOSED IN GALLIARD

PRINTED ON ACID-FREE PAPER. ∞

WWW.PUP.PRINCETON.EDU

PRINTED IN THE UNITED STATES OF AMERICA

1  3  5  7  9  10  8  6  4  2

1  3  5  7  9  10  8  6  4  2

(PBK.)

We dedicate this book with love to our wives,

Carolann Reiter and Cyndi Stam

CONTENTS

TABLES AND FIGURES

ACKNOWLEDGMENTS

ONE

Democracy’s Fourth Virtue

TWO

Democracy, War Initiation, and Victory

THREE

Democracy and Battlefield Success

FOUR

Balancers or Bystanders? The Lack of Fraternal

Democratic Assistance during War

FIVE

Winning Wars on Factory Floors? The Myth of the

Democratic Arsenals of Victory

SIX

Democracy, Consent, and the Path to War

SEVEN

The Declining Advantages of Democracy:

When Consent Erodes

EIGHT

Why Democracies Win Wars

NOTES

BIBLIOGRAPHY

INDEX

TABLES AND FIGURES

Tables

2.1. Winning percentage for war initiators and targets by state type

2.2. Probit models of war outcomes

2.3. Probit analysis of determinants of victory for targets, 1816–1990

2.4. Models of dispute and war initiation, 1904–1986

3.1. Ordered probit analysis of determinants of battle outcomes, 1800–1982

3.2. Regression analysis of effects of democracy on battlefield military effectiveness, 1800–1982, limited sample

3.3. Regression analysis of effects of democracy on battlefield military effectiveness, 1800–1982, complete sample

4.1. Cross tabulation of target’s democracy and defender intervention in wars, 1816–1992

4.2. Probit analysis of decision to join a target during wars, 1816–1992

5.1. Regression analysis of determinants of economic capability among all war participants, 1816–1990

5.2. Regression analysis of determinants of economic capability among war initiators, 1816–1990

5.3. Time series analysis of determinants of military mobilization, 1816–1992

5.4. Regression analysis of democratic military mobilization for war belligerents, 1816–1990

5.5. Negative binomial regression of democracy and the quantity of tanks, artillery pieces, and air sorties in battles, 1800–1982

7.1. Coefficient estimates, multinomial logit model of war outcomes and duration

7.2. Block variable tests for variables with time interactions

7.3. Predicted vs. actual outcomes

7.4. Predicted probabilities of war outcomes at different stages, 1816–1990

Figures

2.1. Probability of victory by level of democracy among target states

2.2. Probability of victory by level of democracy among initiator states

7.1. Changes in initiator’s chances of victory over time

7.2. Probability war continues by time and initiator democracy

7.3. Probability of democracies and nondemocracies accepting a draw over time

ACKNOWLEDGMENTS

DURING THIS BOOK’S gestation, we received tremendously helpful comments from a number of different sources. At various stages of the manuscript’s development, we received much advice and many useful suggestions. Early on, Scott Bennett, Bruce Bueno de Mesquita, Scott Gartner, Ali Ghobadi, Joshua Goldstein, Shoon Murray, Eric Reinhardt, and Chris Zorn provided important insights and criticisms. As we completed our initial draft, a number of individuals were kind enough to read the entire manuscript and provide extensive comments. They include Robert Bartlett, Erik Gartzke, Hein Goemans, Paul Huth, Peter Katzenstein, John McCormick, Steve Rosen, Bruce Russett, Ian Shapiro, and Alastair Smith. The resulting book is far better than it would have been absent their help, though we remain exclusively responsible for all opinions and any remaining errors.

We thank Scott Bennett for permitting us to use in chapter 7 portions of an article he coauthored with Stam (D. Scott Bennett and Allan C. Stam, The Declining Advantages of Democracy: A Combined Model of War Outcomes and Duration, Journal of Conflict Resolution 42 (June 1998): 344–366). Earlier versions of chapters 2 and 3 appeared as Dan Reiter and Allan C. Stam III, Democracy, War Initiation, and Victory, American Political Science Review 92 (June 1998): 377–389, and Dan Reiter and Allan C. Stam III, Democracy and Battlefield Military Effectiveness, Journal of Conflict Resolution 42 (June 1998): 259–277, respectively. Jay Kim and Asit Gosar provided much needed research assistance. Chuck Myers at Princeton University Press provided unflagging support for the project for which we are very grateful. For financial support, we thank at Emory University the Institute for Comparative and International Studies and the University Research Committee. Yale University provided substantial resources in both time and research support.

We presented early versions of various portions of this project at different institutions, including the John M. Olin Institute at Harvard University, the Program on International Security Policy at the University of Chicago, the Center for International Security and Cooperation at Stanford University, the Center for Political Studies at the University of Michigan, the Rockefeller Center at Dartmouth College, the Research Program in International Security at Princeton University, and the departments of political science and government at the University of California-Riverside, the University of California-San Diego, Duke University, Rutgers University, the University of Washington, the University of Colorado, Arizona State University, and Yale University. We received useful feedback from those attending our talks. We thank the assembled attendees for their indulgence and considered criticism.

Gazing on back lets us extol; book-writing exhausts, and thankfully others supported us. With loving gratitude, we thank our wives for their endless patience and support of the more personal kind. Our book is dedicated to them, Carolann Reiter and Cyndi Stam.

DEMOCRACIES AT WAR

ONE

DEMOCRACY’S FOURTH VIRTUE

The world crisis has given new urgency to the question of the meaning of democracy. If democracy is indeed to be the hope of the future, we know now that we must have its lineaments clearly in mind, so that we the more surely recognize it and the more responsibly act upon it.

—Arthur Schlesinger

THE TWENTIETH CENTURY ended with near consensus among leaders, populations, and academics alike on the virtues of democracy. Successive waves of democratization crashed upon the world with unexpected rapidity and completeness: in all regions of the world, autocratic regimes have been swept from power to be replaced by new, more democratic forms of government. Even states such as the Soviet Union, Nicaragua, South Korea, and Chile that had seemed in the middle 1980s to be paragons of authoritarian stability were by the early 1990s fledgling democracies. This most recent wave of democratization has renewed democrats’ faith in their political system: as the American President William Jefferson Clinton proclaimed in his 1994 State of the Union address, Ultimately, the best strategy to ensure our security and to build a durable peace is to support the advance of democracy elsewhere. Democracies don’t attack each other, they make better trading partners and partners in diplomacy.¹ Many politicians and most political scientists base their devotion to democracy on the belief that liberal democracy brings with it at least three important virtues: freedom, prosperity, and peace.

While some may contest these beliefs, of late, dissenting voices sound out less frequently and with diminished fervor. For instance, regarding democracy’s first virtue, freedom, most critics have abandoned the old Marxist canards about the sham of voting and supposed fascist norms that identified the individual with the state. Instead, demands for free and fair elections and the protection of individual liberties in the formerly Communist states of Eastern Europe and the Soviet Union led to what some refer to today as the Velvet Revolution.²

Belief in the second virtue, prosperity, has been a central driving force behind the new wave of democratization. Citizens in Communist bloc nations found it increasingly difficult to ignore their societies’ economic stagnation, particularly when compared to the material prosperity of the democratic, capitalist West. As technological advances in communication made it increasingly difficult for Communist leaders to maintain the veil of ignorance pulled over their societies, citizens in those countries demanded a rapid combination of democratic and market reforms that they believed would allow them the ability to emulate the heavenly cities of the West, known as the European Union and American consumer markets.

Third, successful democratization appears to bring peace to otherwise potentially warring nations. Resurrecting an empirical observation first speculated on by the political philosopher Immanuel Kant, a consensus formed in the academic community during the early 1990s that democracies almost never fight each other.³ This belief spread to the American presidency, leading the Clinton administration to emphasize democratization in its foreign policy as it concluded that the best way to stabilize traditionally dangerous regions like Eastern Europe was to foster the spread of liberal democratic institutions.

Together, these three virtues seem to offer elegant and just solutions to the human condition, the perfect recipe for the organization of society, and even, in the words of one observer, the end of history.⁴ But what if, in their dealings with other nations, democracies prove vulnerable to predation? Is democracy a luxury that states can afford only during times of peace? Or are the attributes associated with democratic institutions, those that provide for the personal liberty, freedom of expression, and collective material growth of common citizens, also the same attributes that, in the worst of times, allow states to provide for their national security as well? Largely underappreciated by scholars and political observers has been a fourth virtue of democracy: democracies win wars. Since 1815, democracies have won more than three quarters of the wars in which they have participated. This is cause for cheer among democrats. It would appear that democratic nations not only might enjoy the good life of peace, prosperity, and freedom; they can also defend themselves against outside threats from tyrants and despots.

The martial effectiveness of democracies comes as a surprise to some. Critics of the democratic experiment pessimistically believed that the obverse side to a liberal political culture’s fostering of prosperity and commerce would be a corresponding inability to muster the military discipline and spirit necessary to conduct an effective foreign policy, particularly in times of war. The American Founding Fathers saw that threats to national security might require the sacrifice of liberty, as freedom hampers the ability of the state to conduct war. That great and perceptive observer of America, Alexis de Tocqueville, agreed, stating his qualms frankly: I have no hesitation in saying that in the control of society’s foreign affairs democratic governments do appear decidedly inferior to others. Decades later, disheartened after the Union Army’s debacle at Bull Run, General William Tecumseh Sherman put the point somewhat differently: I doubt if our democratic form of government admits of that organization and discipline without which an army is but a mob. Indeed, doubts about the ability of democratic government to defend itself have plagued American presidents as diverse as Abraham Lincoln and John F. Kennedy. Some pessimists recommended that elected governments would have to subvert democracy in order to make effective foreign policy; others doubted that the democratic experiment would even survive.⁵

In this book, we explore why democracies win wars. In doing so, we will try to show why events repeatedly prove the pessimists wrong, ironically among whom rank some of democracy’s greatest leaders. We also discuss how it is that the nation-states most capable of safeguarding freedom also exhibit prowess on the battlefield, paradoxically to some, by putting governance in the hands of the people. By addressing these puzzles, we hope to say something about both democracies and wars.

About democracies, we want to find out what characteristics enable them to prevail on the battlefield more often than not. Democracies are complicated creatures, exhibiting many anatomical and behavioral qualities that distinguish them from other political systems such as dictatorships or monarchies. We explore which of these differences explain why they win wars, and in turn, we aim to say something about which of these differences are more important than others to students of pluralism, democratic institutions, and international relations.

About wars, we aim to improve our understanding of the tragedy of war by exploring why states win them. Some readers may find our detailed study of the process by which states prosecute war as somewhat repellant, given war’s obvious horrors. Indeed, the modern academic study of conflict by political scientists has focused almost myopically on the causes of war. Perhaps this is because the study of war outcomes seemed either pointless in the shadow of nuclear weapons or politically incorrect in the wake of the Vietnam War.⁶ However, like it or not, the history most commonly taught is that which is written by the victors in war; understanding the course of history and the lessons we might divine from it requires, in part, tracing the steps to victory on the battlefield. Perhaps most importantly, thinking about how countries win wars is a necessary step to understanding how wars begin, since leaders think about whether or not they will win wars before they start them. Fleshing out our understanding of how states win wars will help us, in turn, to understand how wars begin.

Our central argument is that democracies win wars because of the offshoots of public consent and leaders’ accountability to the voters. Regardless of the particular permutation, at the core of democracy is the notion that those who govern are accountable in some way to the consent of the people. In democracies, leaders who act without the consent of the voters do so at considerable political risk of removal from office. This commitment to consent, contrary to the negative declarations of observers such as Tocqueville, George Kennan, and Walter Lippmann, offers democracies a set of peculiar advantages that enable them to prevail in war.

We outline two specific advantages that flow from the democratic commitment to consent of the governed. First, being vulnerable to the will of the people restrains democratic leaders and helps prevent them from initiating foolhardy or risky wars. Democratic leaders know that there are few greater political disasters than wasting the lives of their citizens in a losing cause. The explicit threat of electoral punishment and the need to generate consent of the governed at the time of action pushes democratic leaders to be particularly cautious when starting wars and, typically, to start only those wars that they will go on to win. Though Tocqueville and others feared this caution would paralyze democratic leaders, we demonstrate the opposite, that democrats are prepared to use military force but are unwilling to risk decisive defeat when compared to their autocratic counterparts. We present and test this argument in chapter 2. In the same vein, democratic leaders are also quite fearful of fighting wars that may drag on for too long, as public support for war steadily and inevitably erodes as casualties mount. As a result, democracies also tend to fight wars that are both short and victorious, or they willingly compromise and accept bargain outcomes short of outright victory; we develop this point in chapter 7.

A second advantage that emerges from consent occurs on the battlefield itself. What kinds of soldiers might we expect a society based on popular consent to produce? Sherman worried that the soldiers of liberal societies stand to be beaten, as most individuals, if given the choice, will resist the rigors of military discipline necessary for victory on the battlefield. We turn Sherman’s worry inside out: the soldiers produced by consent-based societies will in fact enjoy certain advantages. Specifically, the emphasis on individuals and their concomitant rights and privileges in democratic societies produces better leaders and soldiers more willing to take the initiative on the battlefield. Rather than empowering the individual at the expense of the collective, democratic institutions are associated with states filled with individuals more capable of serving the state’s needs in times of duress. We develop this argument and test it in chapter 3.

To approach the multifaceted question of why and how democracies prevail in war, we present four different theoretical perspectives on the nature and behavior of democracies. Each perspective offers a slightly different answer to the questions, what is a democracy, and what are the links between these political institutions and war? Though we might agree on the general conception that democracies offer the people greater control over their leaders, beyond this, disagreements emerge over what it means to be a democracy.⁷ Our goal here is to ask which theoretical perspectives explain best the specific phenomenon of why democracies win. Each of these perspectives offers an alternative vision of why democracies might appear powerful and contains different implications for the relations between the governed and those that represent them.

Each perspective generates a set of hypotheses as to why democracies win wars. We develop and test these hypotheses in a series of chapters that compose the bulk of this book. To this end, we show that the historical record supports some hypotheses but not others. We also use these findings to gain purchase on two different questions: how do democracies generate the consent needed to initiate a war, and when do democracies seek to end wars? These we explore in chapters 6 and 7. In the concluding chapter, we discuss the implications of our findings for democracy, war, and the future of the international order.

Perspective 1: The Skeleton of

Democracy—Political Structures

There is universal agreement that political institutions create the essence of democracy: some set of rules or law that provides for the direct or indirect control of a state’s leaders by the citizens of that state. In order for us to consider a state a democracy, its leaders must be, at some level, answerable to the people. Typically, in democratic states, voting in regular, fair, and competitive elections is one means by which citizens hold leaders accountable for potentially reckless behavior. More specifically, this accountability can be of three basic forms, either in combination or singularly. In one form, known as retrospective voting, voters may use their franchise to punish or reward leaders’ past behavior. In another conception of accountability, known as prospective voting, voters will select leaders believed to be the ones most competent to deal with the challenges ahead in the foreseeable future—those leaders offering the best prospects, expectations of this competence perhaps being formed on the basis of past behavior. Last, political institutions may explicitly or implicitly require leaders to generate popular consent for a policy at the time of its enactment, what we will refer to as a contemporaneous consent model.

The distinguishing characteristics of prospective versus retrospective voting models hinge on assumptions about where the voters’ and their leaders’ uncertainty lies. Voters may focus on the future and may expect the future to be quite different from the past. If this were the case, citizens might well vote out of office officials who performed well in the past because they nevertheless believed other potential leaders might have some special competence to deal with new challenges in the future, regardless of the current leader’s prior performance. Voters would not necessarily punish failure in the past either—rather, in the prospective view, voters would search for other possible leaders, sometimes changing course, sometimes staying the course, relying on the financial investment maxim that past performance is no guarantee of future returns. Retrospective voting is a somewhat less sophisticated notion of voting, that voters simply reward success and punish failure, focusing myopically, and perhaps a bit irrationally, on the past. Rather than developing a sophisticated gauge of leader competence in the face of an uncertain future, retrospective voters implicitly assume the future will be like the past and the past performance is a fair indicator of future performance. Our view is that voters and leaders alike tend to focus on the matter at hand. Leaders in liberal democracies seek out contemporaneous approval for political choices. Voters then punish leaders not so much for particular failure or success, but instead for failing to heed the more popular sentiments at the time the leaders settle on a particular policy. We discuss these distinctions in more detail in chapters 6 and 8.

Beyond the basic vote, there are of course a myriad of democratic forms of checks and balances: presidential versus parliamentary systems, representative versus participatory systems, variations in the separation of powers and protection of individual rights, and so forth. How does the existence of the vote and other systems of checks and balances affect foreign policy in general, and democracies’ proclivity for victory in particular? A number of observers going back to Kant and forward to modern scholars of international relations have argued that the vote acts to constrain democratic leaders from engaging in any and all military action. According to this logic, the people who ultimately pay the price of war in higher taxes and bloodshed would oust any leader who recklessly threw their nation into war.

The insight that democratic political structures provide foreign policy constraints makes an interesting prediction for democracies’ tendencies to victory. The people certainly do not want to suffer costly or meaningless wars; correspondingly, they also do not want to fight losing wars. We argue that this means that democratic leaders pick only winnable wars; that is, when they do start wars, they will be especially likely to win. Conversely, autocratic leaders know that in all but the worst of conditions, their power is secure, even following defeat in war. This political insulation leads them to start wars they know they may have little chance to win but where the prize at hand might be particularly enticing.

We also consider an important corollary to this insight: what happens when democrats are unable to gather popular consent and instead carry out policies beyond the shadow of public consent, bypassing constraining checks and balances? Specifically, what happens when the government carries out foreign policy in secret, out of the popular view? We argue that when carrying out such covert actions, democracies begin to act more like other kinds of states; they take violent actions against other democracies, engage in doomed foreign policy ventures, and violate the human rights of their opponents in ways that democratic citizens would likely find repugnant. In chapter 6, we explore patterns of democratic behavior regarding covert action.

Perspective 2: The Spirit of Democracy—Political Culture

Political culture has been at the center of the modern study of politics. Do the societies of different political regimes exhibit different values and norms that would either emerge on the battlefield or indirectly lead to democratic elites selecting out of certain types of wars for normative reasons? Is a state’s proclivity to develop or depend on a violent political culture (both at home and abroad) a cause or effect of its regime type? What exactly are the foreign policies associated with different forms of political culture?

Democracy has been at the center of the debate over political culture, presenting a number of central questions. Does democracy require a certain political culture to thrive? Does democracy promote changes in political culture? Do citizens in democracies exhibit different norms and values from citizens of other societies? We argue that differences in political culture help democracies to win wars, but not in the ways most political scientists and military historians might suspect. Rather than weakening the whole to empower the individual, we argue the opposite, that democratic institutions that empower the individual in the end empower the whole as well. Specifically, in chapter 3 we argue that democracies’ emphasis on the prerogative of the individual translates into soldiers and leaders that perform better and exhibit stronger initiative on the battlefield.

Perspective 3: The Family of Democracy—International Community

When Kant thought about the connections between domestic politics and international relations, he saw an opportunity for democracies to form an international community and transform the nature of global politics. The idea that democracies see themselves as a group advocating common interests was revived by Woodrow Wilson at the close of World War I, as he sketched out a new world order based on peaceful relations among democratic states with his famous Fourteen Point manifesto. Though that effort failed, some observers interpreted the post–World War II era as one characterized by the emergence of a democratic community of states, a community that has persevered and prospered even past the end of the Cold War.⁸

What are the observable manifestations of this sense of democratic community, aside from lofty sounding speeches from presidents and prime ministers? Some attribute the near absence of war between democracies to a powerful sense of democratic community. Under the logic of this argument, when one democracy comes under attack, others will come to its rescue, that is, democracies heretofore on the sidelines will join the cause and help one of their own. We explore and test this proposition in chapter 4.

Perspective 4: The Power of Democracy—Economic Might

A central part of war is the clash of military and industrial power. Interstate war is the ultimate test in world politics by which one state uses its human and industrial capital to impose its will on another. An important, though not complete, determinant of war outcomes is the relative balance of military-industrial power—victory often going to the stronger. What, therefore, are the relationships among democracy, industrial power, and war?

A common explanation of democratic victory is that democracies win wars by amassing more material and industrial power.⁹ We consider two different ways that democracies might be able to muster more power in time of war. First, democracies in general might be more prosperous than other kinds of states and therefore able to assemble more massive and better-equipped armies than their opponents. This was the faith leaders on both sides of the Atlantic put in the United States in both world wars, that its industrial might would tip the balances in favor of victory, moving Franklin Roosevelt to call on his country to be the arsenal of democracy. Second, democracies might be able to muster greater collective material sacrifices from society than other kinds of states.¹⁰ The greater popularity of democratic systems might inspire their people to be more willing to make the sacrifices to assure victory, perhaps by allowing deeper cuts in civilian consumption and thereby providing relatively greater resources for the state’s military forces. Alternatively, democracies might emerge more powerful by sending more brothers and sons to join the armed forces. Therefore, whereas the first general point proposes that the democratic pie of power might be larger, this second point posits that from that pie, democracies are willing to cut a larger slice—in terms of either material resources or manpower—to commit to the demands of war.

Our Perspectives Reduced: Why Do Democracies Win Wars?

In the chapters that follow, we show that, in fact, democracies do not win wars because of some sense of international democratic community. Nor do they win because they are generally richer or typically better able to extract resources from their economies. Instead, as we shall see, the power of democracies lies not in the leaders or political elite, but instead in the people themselves—ultimately, power lies in the governed, not in the governors. Democratic war initiators are especially likely to win. In fact, just as some have claimed that democracies have almost never fought each other, we show in chapter 2 that a democracy has almost never started a war it went on to lose. We will show that this is a direct result of the constraining power of political consent granted to the leaders and the people’s ability to withdraw it. We also find support for the hypothesis that soldiers fight better for democracies than for other kinds of states, as they exhibit qualities of better initiative and leadership. In short, we find the skeleton and spirit of democracy to explain best why democracies win wars, and the power and family of democracy to be less useful.

Primarily, we use statistical methods of analysis to test our hypotheses. We do this for two main reasons. First, it permits the simultaneous and rigorous testing of a large array of cases. By looking at the entirety of war in the last two centuries, we can be more confident in generalizing our results to war in general, as opposed to, for example, restricting our analysis to one or two wars. Happily, there is plentiful data in quantified form that permits this analysis. Second, statistical analysis permits the comparative testing of competing hypotheses. This allows us to assess alternative explanations for our hypotheses of interest, improving our ability accurately to assign causality.

We recognize, however, that the interest in the statistical details will vary from reader to reader. To make the book accessible to a wide variety of readers, we have minimized the statistical context of the texts of each chapter, with references only to percentages and line graphs. However, each empirical chapter (chapters 2–5 and 7) contains an extensive appendix that describes in detail the research design and methodology used to produce the findings summarized in the chapter itself. In the next chapter, we take up the issues of democracy, war initiation, and victory.

TWO

DEMOCRACY, WAR INITIATION, AND VICTORY

I am thinking of how best to win this war with a least possible amount of casualties and in the quickest time.

—United States Secretary of War Harold L. Ickes, February 1943

WARS DO NOT occur by chance. They are not the result of some fanciful alignment of the planets and moons; rather, states choose to start them. States deliberately select themselves into the population of war participants by attacking other states. Armies clash on battlefields not because of mechanical accidents, flights from reason, or the whimsy of the gods, but because at least one national government or leader prefers war to peace. Belligerents in war are closer to Oedipus, meeting his peculiar fate because of his personal tragic flaws and fatal choices, than to Job, a pure innocent, made to suffer by divine action rather than his own failings.¹

We assume that states pick their fights: they start wars when the stakes are high enough, and when they are confident they will win. Note that this latter point directly connects the causes of wars to their outcomes. If states think about whether or not they will win a war when they make the decision to initiate the use of force, then understanding who wins wars sheds light on the decision to attack. To push this point a bit further, we might expect that the more conservative, risk-avoidant states will win more of the wars they start, as they will attack only when they are very confident they will win. One could think of such states as one might a boxing champion who fights only weak opponents to safeguard his grip on his title.

In this chapter, we construct an argument explaining war outcomes that integrates the initiation and the fighting of wars. Our aim is to explain how both the initial decision to take up arms and the eventual conduct of military operations on the battlefield can affect who wins and loses wars. This will then help us tackle our specific puzzle, understanding why democracies win wars. The story that follows produces two general answers to this puzzle. First, democracies win wars because they start wars only when they are very confident they will go on to win. Democratic leaders, compared to tyrants, are more likely to lose political office if they fight a losing war. This possibility makes liberal democrats especially cautious when starting wars, fighting only when they are very confident they will win.

We call this proposition the selection effects explanation. The process leaders use to select which wars to start and which wars to avoid leads directly to the effect we observe, democratic victory. While the term selection effects smacks of social science jargon, it gets precisely to the point. Who wins and who loses wars is, in part, a direct consequence of the choices state leaders make before the wars ever begin. Typically, when we think of the kinds of prewar choices that might affect war outcomes, we think of strategy choice, decisions about mobilization or perhaps choices of which weapons to acquire. Here we extend the intuition to its logical extreme. If we assume that leaders can choose to start a fight, selecting their country into a war, then they can also choose not to fight, selecting out of war. We are interested in the systematic effects on winning and losing when leaders choose not to fight, or select their states out of wars based on how they expect the wars to turn out in the end, if they were actually to happen. We develop this argument in detail below and then present tests using historical data that provide strong support for it.

The second general answer to the puzzle is that democracies win wars because they are more effective at fighting wars once hostilities begin. This is true whether the democracy itself chose to start the war or if it was the target of attack. We call this second proposition the warfighting explanation and find much support in the historical record for it as well. In later chapters, we delve into both explanations more deeply, considering in chapters 3–5 specific arguments as to why liberal democracies fight more effectively. We investigate arguments that democracies’ soldiers fight harder, that democracies ally together during wartime, and that they can muster more economic resources. In chapters 6 and 7, we explore more deeply the political assumptions underlying the selection effects explanation, that democratic foreign policy is driven by the motivation of elected leaders to maintain public consent. In these chapters, we explore how the factor of consent affects the democratic decisions to initiate and terminate war.

A Theory of War Initiation and Outcomes

If wars result from choices rather than chance, how do leaders make these crucially important decisions, selecting paths that may bring their nations to the heights of greatness or the depths of annihilation? We propose that leaders think about war the way they do about any other policy issue, seeking to capture benefits while minimizing costs. This is not to say that leaders do not make mistakes and sometimes launch wars that they go on to lose, but merely that they act on the basis of reason rather than fancy. In particular, leaders prefer winning to losing, and to retain the offices of leadership.

We first note that war begins when at least one state chooses to attack another. In social science terms, we mean that war initiators select themselves into the population of war participants.² We also assume that before attacking they make some guess as to whether or not they will