Unit 2 reflection

88

Above: States may go to war over territory that they consider integral to their nation. Indian soldiers (above) and Pakistani soldiers are stationed along the heavily militarized Line of Control that divides the contested territory of Kashmir between the two countries, and troops occasionally exchange cross- border fire. Could tensions over the Kashmir region lead to war?

3 Why Are There Wars? T HE PU Z Z L E War is an extremely costly way for states to settle their disputes. Given the

human and material costs of military conflict, why do states sometimes wage war rather than resolving their disputes through negotiations?

89

In August 1914, the major countries of Europe embarked on a war the likes of which the world had never before seen. Convinced that the war would be over by Christmas, European leaders sent a generation of young men into a fight that would last four years and claim more than 15 million lives. The fighting was so intense that in one battle the British army lost 20,000 soldiers in a single day, as wave after wave of attacking infantrymen were cut down by German machine guns. At the time, it was called the Great War. Those who could never imagine another such horrific event dubbed it the “war to end all wars.” Today, we know this event as the First World War, or World War I, because 20 years later the countries of Europe were at it once again. World War II (1939–45) claimed 30–50 million lives.

There is no puzzle in the study of international poli- tics more pressing and important than the question of why states go to war. It is the most tragic and costly phenom- enon that we observe in social and political life. The costs of war can be counted on a number of dimensions. The most obvious cost is the loss of human life. By one esti- mate, wars among states in the twentieth century led to 40 million deaths directly from combat, plus tens of millions

more deaths owing to war-related hardships. 1 In addition, wars have left untold millions injured, displaced from their homes and countries, impoverished, and diseased.

War has economic and material costs as well. Between 2001 and 2017, the United States spent $1.8 trillion fighting wars in Iraq and Afghanistan. 2 Large sums of money are also spent every year to prepare for the possibility of war. In 2016, military expenditures by all countries amounted to about $1.7 trillion—a sum that represented about $227 per person. 3 Wars can also disrupt the international economy. Conflicts in the oil- rich Middle East, including the Iran-Iraq War (1980–88),

1. Bethany Ann Lacina and Nils Petter Gleditsch, “Monitoring Trends in Global Combat: A New Dataset of Battle Deaths,” European Journal of Population 21 (2005): 145–66.

2. Neta C. Crawford, “U.S. Budgetary Costs of Wars through 2016,” Watson Institute for International and Public Affairs, Boston University, September 2016, http://watson.brown.edu/costsofwar/ files/cow/imce/papers/2016/Costs%20of%20War%20through%20 2016%20FINAL%20final%20v2.pdf.

3. Nan Tian, Aude Fleurant, Pieter D. Wezeman, and Siemon T. Wezeman, “Trends in World Military Expenditure, 2016,” Stockholm Interna- tional Peace Research Institute, April 24, 2017, https://www.sipri.org/ sites/default/files/Trends-world-military-expenditure-2016.pdf.

90

Thinking Analytically about Why Wars Happen Most disputes between states are settled without the parties resorting to war. Although wars tend to capture our attention, it is important to remem- ber that war is an exceedingly rare phenomenon; most countries are at peace with each other most of the time. Figure 3.1 shows the number of states involved in interstate wars in each year from 1820 to 2017, expressed as a percentage of the total number of states in existence at the time. As the figure indi- cates, war is a recurrent feature of international pol- itics in the sense that it fluctuates in frequency but never disappears completely. Yet war is the exception rather than the rule: in most years, the percentage of states involved in war is quite low. All this peace can- not be explained by an absence of issues to fight over.

Hence, when seeking to explain war, we need to ask not only, “What are they fighting over?” but also “Why are they fighting?” 4 In terms of the framework laid out in Chapter 2, answering the first question requires that we understand how states’ interests can give rise to conflicts over things like territory, poli- cies, and the composition or character of each other’s government. The answer to the second question lies in the strategic interactions that determine whether or how these conflicts are resolved. As we saw, the international system lacks institutions—such as leg- islatures, courts, or international police forces—that can resolve conflicts between states through legal, judicial, or electoral mechanisms. As a result, inter- state conflicts have to be settled through bargaining. Understanding why wars occur requires that we iden- tify the factors that sometimes prevent states from settling their conflicts through peaceful bargains that would permit them to avoid the costs of war.

the Persian Gulf War (1990–91), and the Iraq War (2003–10), contributed to spikes in world oil prices (see Figure 4.1 on p. 91). In short, as U.S. Civil War general William Sherman famously declared, “War is hell.”

But if everyone recognizes that war is hell, why do wars happen? The very costs that make the puzzle of war so pressing also make the phenomenon so puzzling. Given the enormous costs associated with war, why would states sometimes choose this course?

At first glance, the answer might seem straightfor- ward: states fight wars because they have conflicting interests over important issues. Often, two states desire the same piece of territory. Nazi Germany wanted to expand into central Europe; World War II started when the Poles, who did not want to give up their territory, fought back. In 1980, Iraq invaded Iran, in part, to seize the latter’s southern oil fields, leading to the Iran-Iraq War. Alternatively, one state might object to the policies or ideology of another. World War I grew out of Austria- Hungary’s demand that Serbia end nationalist agitation that threatened to tear the multiethnic empire apart. The war between the United States and Afghanistan in 2001 occurred because the United States wanted Afghanistan to hand over Osama bin Laden and dismantle terror- ist training camps on its territory, something that the Afghans refused to do. Clearly, part of the explanation for any war requires that we identify the conflicting interests that motivated the combatants.

Although such explanations are correct, they are also incomplete. By identifying the object or issue over which a war was fought, they neglect the key question of why war was the strategy that states resorted to in order to resolve their dispute. In each case, the conflicts were disastrously costly to at least one and, in some cases, all the states involved. In addition to the millions of dead men- tioned earlier, World War I led to the ouster of three of the leaders who brought their countries into the war, and it hastened the breakup of the Austro-Hungarian, Russian, and Ottoman empires. World War II brought about the defeat and occupation of its main instigators, Germany and Japan. The Nazi leader Adolf Hitler committed suicide, and his Italian ally, Benito Mussolini, was hung by his own people. Iran and Iraq fought to a stalemate for eight years, causing 1–2 million casualties and leav- ing Iraq on the brink of economic collapse. The Afghan government’s refusal to hand over bin Laden led to its removal from power. Given these grave consequences, it makes sense to wonder whether all participants would have been better off if they could have come to a

settlement that would have enabled them to avoid the costs of war. Explaining war thus requires us to explain why the participants failed to reach such an agreement.

4. The theory of war developed in this chapter relies extensively on James D. Fearon, “Rationalist Explanations for War,” International Organization 49 (Summer 1995): 379–414.

91What Is the Purpose of War?

What Is the Purpose of War? A war is an event involving the organized use of military force by at least two par- ties that reaches a minimum threshold of severity.5 All the components of this definition are important. The requirement that force be organized rules out spon- taneous, disorganized violence, such as large-scale rioting. The requirement that force be used by at least two parties distinguishes war from mass killings perpe- trated by a government against some group that does not fight back. The minimum threshold—scholars often require that a war cause at least 1,000 battle deaths— excludes cases in which military force is used at low levels, such as brief skirmishes or minor clashes. If the main adversaries in the conflict are states, then we refer to the event as an interstate war; if the main parties to the conflict are actors within a state—such as a government fighting a rebel group—then the event is a civil war.6 In this chapter, the discussion focuses on understanding interstate wars. Chapter 6 covers conflicts involving nonstate actors, including civil wars.

Given that scholars have been trying to understand war for millennia, the num- ber of theories and explanations that have been proposed over the years is natu- rally quite large, and the bargaining model that we lay out in this chapter is only one approach. At least three other broad schools of thought have been influential in modern scholarship. While these approaches contribute important insights, they also have gaps that are filled by the bargaining approach.

18401820 1860 1880 1900 1920 1940 1960 1980 2000 2017 0 5

10 15 20 25

35

45 50

40

30

Year

Pe rc

en ta

ge o

f S ta

te s

In vo

lv ed

in W

ar

FIGURE 3.1 The Percentage of States Involved in Interstate War per Year, 1820–2017

Figure source: Meredith Reid Sarkees and Frank Wayman, Resort to War: 1816–2007 (Washington, DC: CQ Press, 2010). Updated to 2017 by author.

5. See, for example, J. David Singer and Melvin Small, The Wages of War, 1816–1965: A Statistical Handbook (New York: Wiley, 1972).

6. There are also cases that have elements of both kinds of wars. In the Vietnam War, South Vietnam, with the support of the United States, fought a civil war against a communist insurgency while at the same time fighting an interstate war against communist North Vietnam, which was supporting the rebellion. See Chapter 6 for more on the international dimensions of civil wars.

war An event involving the organized use of military force by at least two parties that reaches a minimum threshold of severity.

interstate war A war in which the main participants are states.

civil war A war in which the main participants are within the same state, such as the government and a rebel group.

92 Chapter 3 Why Are There Wars?

The first, described in this book’s Introduction as realism, argues that war is the inevitable result of international anarchy (pp. xxxi–xxxiv). The absence of a central authority capable of policing interstate relations means that wars can happen because there is nothing to stop states from using force to get their way. 7 Moreover, anarchy creates insecurity and a competition for power. In this view, states fight wars either to increase their own power (such as by enlarging their territory) or to counter the power of others (such as by destroying adversaries and their allies). For this reason, realism emphasizes two primary dynamics that can lead to war. One is a preventive motive: the desire to fight in order to prevent an enemy from becoming relatively more powerful.8 The second is a phenomenon known as the security dilemma. This dilemma arises when efforts that states make to defend themselves, such as acquiring a bigger military, make other states fear that they will be attacked. If threatened states arm them- selves in response, the result is a spiral of fear and insecurity that may end in war.9

As we will see, the bargaining model of war shares realism’s presumption that anarchy leads to a world in which military force is often threatened or used to further state interests, and conflicts are addressed through bargaining rather than institutions like courts. The bargaining model also anticipates prevention and the fear of attack as two of several mechanisms that can lead to war. The main depar- ture from realism is in recognizing that the use of military power imposes costs on states, so even if threatening force is useful to get a better deal, states are generally better off if they do not have to actually use force—creating the puzzle that is at the heart of this chapter.

A second alternative approach emphasizes the role of misperceptions or mistakes. Starting from the observation that the costs of war often far exceed any potential benefits, scholars in this tradition conclude that wars must occur because decision makers inaccurately estimate their chances of winning or the costs that will have to be paid. When European leaders sent their armies to war in August 1914, most were convinced that the troops would be home by Christmas—a pre- diction that turned out to be off by four years. Both sides also expected to win the war, meaning that at least one side was being overly optimistic. This kind of argu- ment rests either on research in cognitive psychology showing that people are bad at weighing risks and often fall prey to wishful thinking, or on organizational approaches that emphasize how the ideology and interests of political and military elites can lead to incorrect and overly optimistic assessments of war.10

A concern about such theories is that while perceptual pitfalls are universal, war is quite rare, and these theories have a hard time explaining why wars happen

7. Kenneth N. Waltz, Theory of International Politics (Long Grove, IL: Waveland Press, Inc.).

8. See, for example, Dale C. Copeland, The Origins of Major War (Ithaca, NY: Cornell University Press, 2000); and Stephen Van Evera, Causes of War (Ithaca, NY: Cornell University Press, 2001).

9. Robert Jervis, “Cooperation under the Security Dilemma,” World Politics 30 (January 1978): 167–214.

10. The classic work applying cognitive psychology to international relations is Robert Jervis, Perception and Misperception in International Politics (Princeton, NJ: Princeton University Press, 1976). For an argument that military ideologies can lead to incorrect and overly optimistic assessments, see Stephen Van Evera, “The Cult of the Offensive and the Origins of the First World War,” International Security 9, no. 1 (Summer 1984): 58–107. The effect of military biases may depend on domestic institutions that influence how much civilians control the military; this argument is taken up in Chapter 4.

security dilemma A dilemma that arises when efforts that states make to defend themselves cause other states to feel less secure; can lead to arms races and war because of the fear of being attacked.

93What Is the Purpose of War?

at some points in time but not others. As we will see, uncertainty about the likely outcome and costs of war plays an important role in the approach developed here, but we emphasize not the psychological or organizational origins of potential mistakes, but rather the difficulties of gathering the necessary information in the strategic context of bargaining.

Finally, a long tradition of scholarship argues that wars are fought not because they serve the interests of states, but because they serve the interests of influential groups within the state, such as corporations, arms merchants, and the military.11 In this view, the puzzle of war has an easy answer: Wars are fought in spite of their costs because those costs do not fall on the actors who call the shots. Although this chapter introduces the bargaining model by treating states as the main actors, without internal politics, we will see in Chapter 4 that the potential role of domestic interests can be incorporated quite easily.

Interests at War: What Do States Fight Over? At the root of every war lies a conflict over something that states value. The pur- pose of warfare is not to fight but rather to obtain, through fighting or the threat of fighting, something the state wants. Hence, we should think about the problem of war as a problem of bargaining over objects or issues that are of value to more than one state. Using the framework developed in Chapter 2, we focus on situations in which states’ interests conflict, giving rise to a strategic interaction that involves bargaining over the distribution of whatever is in dispute. The analysis thus starts by assuming that there is an object of value—what we will sometimes refer to as a “good”—and that each state prefers more of the good over less.

What kinds of goods do states fight over? Territory has historically been the most common source of trouble. Indeed, a study of 155 wars over the last three centuries found that over half (83) involved conflicts over territory—more than any other single issue.12 States come into conflict if more than one wants the same piece of territory. There are a number of reasons why a piece of territory may be valuable to more than one state. First, it might contribute to the wealth of the state, particularly if it contains valuable resources such as oil, natural gas, or minerals. Iran and Iraq fought a lengthy war, from 1980 to 1988, in part because Iraq coveted Iran’s southern oil fields. Territory can also be economi- cally valuable simply by adding to the industrial or agricultural resources at the state’s disposal.

A second reason that territory can cause conflict between two states is its military or strategic value. For example, the Golan Heights, on the border between Israel and Syria, has a commanding position over northern Israel from which it is possible to launch devastating attacks on the towns below. Israel seized the Golan Heights from Syria in the 1967 Six Day War, and this territory has been a source of conflict between the two states ever since.

11. See citations to this literature in Chapter 4.

12. Kalevi J. Holsti, Peace and War: Armed Conflicts and International Order, 1648–1989 (Cambridge: Cambridge University Press, 1991).

94 Chapter 3 Why Are There Wars?

Finally, a piece of territory might be valuable for ethnic, cultural, or historical reasons. The long-standing conflict between Israel and its Arab neighbors stems from the latter’s resistance to the creation of a Jewish state on a land where many Arabs lived. In this case, the dispute goes beyond simply the location of a border, as some Arab states have refused to recognize Israel’s right to exist.13

Similarly, the long-running conflict between India and Pakistan is driven by the fact that both have historical and ethnic claims to the region of Kashmir. When the two countries gained their independence from Britain in 1947, India pressured Kashmir’s Hindu leader to join with India, which is predominantly Hindu. How- ever, because the people of Kashmir, like the people of Pakistan, are predominantly Muslim, Pakistan has claimed the territory on the basis of religious ties. Map 3.1 illustrates the current boundaries of the disputed area, which are further compli- cated by China’s claims to neighboring portions of Tibet. In territorial disputes, states threaten or use military force to compel concessions and/or to seize and hold disputed land.

Wars can also arise out of conflicts over states’ policies. Such conflicts come about when one state enacts a policy that benefits it but harms the interests of another. The conflict that led to the Iraq War centered on Iraq’s alleged pursuit of weapons of mass destruction (WMD), which the United States saw as threatening

13. Egypt recognized Israel in 1979, and Jordan did likewise in 1994.

claimed by India ,

Map note: The Line of Control, which originated as a cease-fire line after the First Kashmir War (1947–48), divides the region into Indian- and Pakistani-administered areas.

MAP 3.1 Territorial Claims in the Kashmir Region

95What Is the Purpose of War?

to its broader interests in the region. Ongoing U.S. conflicts with North Korea and Iran have similar sources. Conflict between Russia and Ukraine that started in 2014 revolved around the latter’s decision to pursue closer economic ties to the West and policies that were alleged to harm the interests of ethnic Russians living there. A state’s mistreatment of its own citizens can also spark international conflict. For example, the United States and its allies waged war against Serbia in 1999 for its repressive treatment of civilians in the Kosovo region. Similarly, the United States launched cruise missiles against Syria in 2017 after the government used chemical weapons against its own people.

When states have policy disputes, war, or the threat of war, may be a mecha- nism for compelling policy change. In the case of Kosovo, two months of bombing by the United States and its allies led the Serb government to end its military cam- paign against the Kosovars; in the case of Syria, U.S. threats led to a deal in which that regime gave up its chemical arsenal. Alternatively, war may be used to replace the offending regime with a friendlier one that will pursue different policies. This was the U.S. plan in Iraq and Afghanistan, though the effort to install more pliant regimes in those countries embroiled American forces in civil wars that lasted much longer than the initial operations to oust the former, hostile regimes (see Chapter 6).

The possibility of using military force to change regimes suggests a third kind of conflict between states: conflicts over regime type, or the composition of another country’s government. In its conflict with Ukraine, Russia sought reforms of the Ukrainian political system that would give the Russian minority there greater

States sometimes make threats or use military force to influence the policies of foreign governments. The Syrian government attacked rebel-held villages with chemical weapons in August 2013 (pictured here), and April 2017. After the second attack, the United States carried out an air strike on a Syrian air base that it believed to be the source.

96 Chapter 3 Why Are There Wars?

autonomy and influence. During the Cold War, the United States saw communist regimes as natural allies of the Soviet Union, and hence it sought to prevent the establishment of such regimes. U.S. involvement in the Vietnam War was driven by the desire to protect the pro-American South Vietnamese government from internal and external enemies bent on overthrowing it. During this period, both superpowers intervened regularly in other states to prop up friendly governments or to remove unfriendly ones. The role of outside powers in civil wars, such as in Vietnam, will be addressed in Chapter 6.

As this last example suggests, conflicts over territory, policy, and/or regime may spring from deeper conflicts that give rise to concerns about relative power. The specific conflict that started World War II was a territorial dispute between Germany and Poland over a small strip of territory that lay between them. How- ever, Britain and France were concerned that a victory over Poland would fur- ther strengthen and embolden Germany, making it a more formidable foe in their ongoing struggle for influence and territory in Europe. Hence, the German-Polish territorial dispute impinged on the interests of other states because of its potential impact on their relative power vis-à-vis Germany. The Cold War rivalry between the United States and the Soviet Union similarly imbued many local conflicts with global importance because of their perceived impact on the relative strength of the superpowers.

Bargaining and War Conflicting interests are clearly necessary for wars to happen, but they are not suf- ficient to explain why wars actually do happen. To understand why some conflicts become wars and others do not, we have to think about the strategic interactions that states engage in when they seek to settle their disputes. In a well-functioning domestic political system, the kinds of disputes that lead to wars are often settled through institutional mechanisms. Property disputes can be resolved by courts backed by effective police powers. If one person engages in actions that harm another, the latter may turn to the legal system to solve the problem. Within states, policy disagreements and conflicting ideas over who should govern can be settled by elections. As noted in Chapter 2, however, the international system lacks reliable legal, judicial, and electoral institutions. For this reason, states must generally try to settle conflicts with one another through bargaining.14

Bargaining describes interactions in which actors try to resolve disputes over the allocation of a good. They may bargain over the distribution of a disputed ter- ritory to determine whether there is a division acceptable to both sides. Or they may bargain over each other’s policies so that objectionable ones might be modified or eliminated. Although we often think of bargaining as entailing compromise or

14. Although some interstate disputes have been adjudicated through institutions like the International Court of Justice (ICJ), these institutions lack strong enforcement mechanisms to guarantee compliance with their rulings, and disputants often engage in bargaining after an ICJ ruling to determine whether and how its terms will be implemented. Hence, these rulings are a part of, rather than a substitute for, the bargaining process. See, for example, Cole Paulson, “Compliance with Final Judgments of the International Court of Justice since 1987,” American Journal of International Law 98 (2004): 434–61.

97What Is the Purpose of War?

give-and-take, the bargaining process does not always imply that differences will be split. Indeed, in many cases, states assume all-or-nothing bargaining positions. For example, when President Bush demanded in October 2001 that the Afghan gov- ernment hand over Al Qaeda leaders responsible for the 9/11 terrorist attacks, he declared that “these demands are not open to negotiation or discussion.”15

A crisis occurs when at least one state seeks to influence the outcome of bar- gaining by threatening to use military force in the event that it does not get what it wants. At this point, we enter the domain of coercive bargaining, in which the con- sequences of not reaching an agreement can involve the use of force, including war. We sometimes refer to bargaining under the threat of war as crisis bargaining or coercive diplomacy. In all such interactions, at least one state sends the message “Satisfy my demands, or else”—where the “or else” involves imposing costs on the other side through military action.

In some cases, this message takes the form of an explicit ultimatum, such as Bush’s March 2003 pronouncement that Iraqi leader Saddam Hussein had 48 hours to leave the country or face an invasion. In other cases, the threat is con- veyed implicitly, through menacing actions such as mobilization of troops or mil- itary maneuvers. The Russian military incursion into Ukraine in August 2014 was preceded by the massing of troops on the border and military exercises, but not by an explicit ultimatum or set of demands (at least not publicly). Whether explicit or implicit, the purpose of such actions is clear: they seek to wrest concessions from the other side by making the alternative seem unacceptably costly.

The costs and likely outcome of a war determine which deals each side will consider acceptable in crisis bargaining. We can generally assume that the best pos- sible outcome for a state in a crisis is to get the entire good without having to fight. If the other side gives in, the state gets its most preferred settlement of the underly- ing issue and avoids paying the costs associated with war. It is quite likely, though, that a state would also accept something less than its most preferred settlement, given that the alternative of fighting is costly.

For example, imagine a conflict over a piece of territory worth $100 million. Assume a state believes that in the event of a war, it is certain to win the territory; however, the costs of war, if put in monetary terms, would amount to $30 million. In that case, the expected value of going to war for that state is $100 million – $30 million = $70 million. Hence, the state should be willing to accept any deal that gives it at least $70 million worth of the territory. Since a state has the option to wage war if it determines that doing so is in its interests, the state will accept a bargain only if that bargain gives it at least as much as it can expect to get from war. And for any deal to prevent a war, it must satisfy all sides in this way: each state must decide that it prefers the deal over fighting a war. Hence, in our simple exam- ple, war can be averted only if the other state is willing to settle for the remaining $30 million worth of territory or less.

15. Elizabeth Bumiller, “Bush Pledges Attack on Afghanistan Unless It Surrenders bin Laden Now,” New York Times, September 21, 2001, A1.

crisis bargaining A bargaining interaction in which at least one actor threatens to use force in the event that its demands are not met.

coercive diplomacy The use of threats to advance specific demands in a bargaining interaction.

98 Chapter 3 Why Are There Wars?

The discussion at the outset of this chapter implies a very simple proposition: Because war is costly, a settlement that all sides prefer over war generally exists.16 The graphic on pages 100–101 illustrates the simple idea behind this proposition. Assume that two states—call them State A and State B—have conflicting interests over the division of a particular good, represented by the green bar in each diagram on the left-hand page. To make this concrete, think of this good as a piece of terri- tory, as in the example on the right-hand page. Any line dividing the bar, such as the dotted line labeled Possible deal in panel 1, represents a possible division of the good, such as a border drawn through the territory. State A receives the share of the good to the left of that line, and State B receives the share of the good to the right. Since both states prefer more of the good over less, State A wants to get a deal as far to the right as possible, and State B wants to get a deal as far to the left as possible. Put another way, State A’s most preferred outcome, or ideal point, is at the far right of the bar, which would mean State A gets all the good, and State B’s ideal point is on the far left, which would mean State B gets all the good. (Hence, this horizontal bar is identical to the diagonal line used to illustrate bargaining in Figure 2.2 on p. 56).

Now consider what happens if the two states go to war. Moving to panel 2, the line labeled War outcome indicates the actual, or expected, outcome of a violent conflict. In the case of a conflict over territory, we can think of this line as indicating how much territory each state would control after a war. The farther that line is to the right, the better State A is expected to fare in the event of war; the farther it is to the left, the better State B is expected to fare.

Crucially, fighting entails costs. These costs diminish the value of the expected war outcome to each state. As shown in panel 3, the value of war to each state is the share of the good it expects to win, minus the costs it expects to incur. Notice that the value of war to State A (which corresponds to the portion of the bar to the left of the dotted red line) is smaller than the share of the good that it expects to get from war (the portion of the bar to the left of the War outcome line). Indeed, as the figure is constructed, State A expects to win more of the good in a war than it would get from the possible deal in the first panel, but once the costs of war are taken into account, the possible deal is actually preferable to war.

To understand how the costs affect the value of war for each state, we have to think about those costs as they relate to the value of the good. Looking at the cor- responding example on the right-hand page of the graphic, imagine that State A expects to win 70 percent of the territory, at the cost of 5,000 lives. State A’s war value hinges on how much State A values those lost lives relative to the value of the territory. If State A thinks that losing those soldiers is equivalent to losing 30 percent of the territory, then its value for war is 70 – 30 = 40 percent of the value of the territory. State B would make a similar calculation. For both states, the costs of war would effectively go up if either the number of lives lost went up or the value attached to the territory went down.

As panel 4 shows, State A prefers any deal over war that gives it more than its value for war (deals that fall to the right of the dotted red line). Deals in this range

16. See Fearon, “Rationalist Explanations for War.”

99What Is the Purpose of War?

give State A more than it expects to get through war, once the costs are taken into account. Similarly, State B prefers any deal over war that gives it more than its value for war (deals that fall to the left of the dotted blue line). Notice that the set of deals that State A prefers to war overlaps with the set of deals that State B prefers to war, creating a region called the bargaining range. Any division of the good in this bar- gaining range gives both states more than they expect to get from fighting a war. As this exercise demonstrates, the fact that war imposes costs on both sides means that such a range of deals always exists. Hence, in theory, there are bargains that both sides would prefer over war.

Although the preceding exercise treated the good in dispute as a piece of ter- ritory, the general framework and insight can be applied to any kind of issue over which states have conflicting interests. For example, consider the conflict between North Korea and the United States over the former’s nuclear program. In this case, the United States’ ideal outcome, which we could put at one end of the bar, is for North Korea to disarm completely and become a democratic state. North Korea’s ideal outcome, which would be located at the opposite end of the bar, would be for the United States to accept North Korea’s status as a nuclear power and recognize its existing regime.

In between are various compromise alternatives that neither considers ideal, including the current status quo, in which North Korea has a small number of nuclear devices but the United States refuses to recognize its right to them. This point is close to North Korea’s ideal point. Another point on the bar, closer to the U.S. ideal point than to the status quo, would be for North Korea to freeze its nuclear program and admit international inspectors. Though the substantive meaning of any location on the bar depends on the exact issue in dispute, the basic argument for the existence of a bargaining range holds across issues. (See “Controversy” on p. 104 for an extended discussion of negotiations between the United States and North Korea.)

This simple model is useful not because it is right in the sense that it correctly describes the complexity of real-world bargaining interactions, but rather because it forces us to think about all the ways in which bargaining could go wrong. Under- standing why wars happen even though a peaceful bargain theoretically exists for virtually any conflict is the main purpose of this chapter.

Compellence and Deterrence: Varieties of Coercive Bargaining The model is also useful for thinking about the conditions under which states might have an interest in initiating a crisis in the first place. The pre-crisis distribution, or status quo, can be represented as a line dividing the bar. Where the status quo is located relative to the states’ values for war determines which state, if any, might have an interest in changing the status quo through a threat of force. If a state is already getting from the status quo at least as much as it expects to get through war, then it generally cannot gain by threatening war to change the status quo. However, if a state expects to get more through war than it has in the status quo, then it has an interest in making a challenge.

bargaining range The set of deals that both parties in a bargaining interaction prefer over the reversion outcome. When the reversion outcome is war, the bargaining range is the set of deals that both sides prefer over war.

100 Chapter 3 Why Are There Wars?

VALUE OF WAR TO B

1. A possible deal: Two states, A and B, have conflicting interests over a good (represented by the green bar). Both states want as much of the good as possible. The dotted line represents the distribution of the good based on one possible deal. Any deal determines A's and B's shares.

2. The expected outcome of war: As they bargain, each state evaluates how much it can expect to get as a result of war. The dotted line shows the expected outcome of war. Here, A would get more by going to war than by accepting the above deal.

3. Costs of war: However, war also involves costs, and these costs must be subtracted from what each state hopes to gain. So, although war would give A more of the good than it would get under the deal in panel 1, the costs incurred reduce the value of war to A.

4. The bargaining range: Each state prefers any deal that gives it more than its expected value of war, once costs are subtracted. Because war inflicts costs on both sides, there is a range of possible deals that both sides would prefer over war.

The Model: Bargaining and War

58 Chapter 3 Why Are There Wars?

STATE A’S SHARE STATE A’S IDEAL OUTCOME

STATE B’S SHARE

POSSIBLE DEAL

A’S SHARE AFTER WAR

B’S SHARE AFTER WAR

WAR OUTCOME

WAR OUTCOME

COSTS TO A

COSTS TO B

VALUE OF WAR TO B

VALUE OF WAR TO A

STATE B’S IDEAL OUTCOME

DEALS B PREFERS TO WAR

VALUE OF WAR TO A DEALS A PREFERS TO WAR

A B

A B

A B

A B

WAR OUTCOME

BARGAINING RANGE

101What Is the Purpose of War?

WAR OUTCOME

1. A conflict over territory: Imagine two states have a conflict of interest over a piece of territory. State A would like to have the whole territory, as would State B. A deal is proposed to divide the territory at the dotted line.

2. The expected outcome of war: Based on the strength of their respective armies and other resources, a war over the territory is expected to result in State A winning the portion of the territory to the left of the dotted line.

3. Costs of war: Each side also considers the costs of war. Imagine that State A values the territory it would win as a result of war at $100 million but values the costs of war (both financial and human) at $30 million. These costs would reduce the value of war to A by 30 percent.

4. The bargaining range: Each side should prefer any deal that gives it more territory than what it expects from war, once the costs of war are subtracted from the territory it wins. Any border drawn through the shaded range divides the territory into shares that both sides prefer over war.

An Example: Bargaining over Territory

59Wh at Is t he P u r p ose of Wa r?

WAR OUTCOME

POSSIBLE DEAL

A BA B

A B

COSTS OF WAR

WAR OUTCOME

BARGAINING RANGE

102 Chapter 3 Why Are There Wars?

The top panel of Figure 3.2 depicts such a situation: in the status quo, State A’s share of the good is less than the value of war to A, even when the costs of war are taken into account. This does not mean that there will be a war, since, after all, a bargaining range still exists. But in this situation, the dissatisfied state can profit by threatening war in order to get a better deal. As the lower panel in Figure 3.2 shows, this logic divides the bar into three segments. If the status quo is to the left of the red dotted line indicating State A’s value for war, then State A has an interest in threatening war to try to get a better outcome. Similarly, if the status quo is to the right of the blue dotted line indicating State B’s value for war, then State B prefers war over the status quo and thus has an interest in sparking a crisis. Finally, if the status quo is in the bargaining range, then both prefer the current situation over war, and neither can expect to gain by waging war.

We often classify threats according to whether they are intended to preserve or change the existing relationship between states. An effort to change the status quo through the threat of force is called compellence. A compellent threat is intended to coerce the target state into making a concession or changing a current policy. Compellent threats take the form of “Give me y, or else” (where y is something that the threatener values) or “Stop doing x, or else” (where x is an objectionable policy).

A

STATUS QUO

A’S SHARE

VALUE OF WAR TO A COSTS TO A

B’S SHARE

COSTS TO B VALUE OF WAR TO B

B

WAR OUTCOME

BARGAINING RANGE

STATUS QUOS NEITHER WILL CHALLENGE

WAR OUTCOME

B

STATUS QUOS THAT A WILL CHALLENGE

STATUS QUOS THAT B WILL CHALLENGE

A

compellence An effort to change the status quo through the threat of force.

FIGURE 3.2 Bargaining and the Status Quo Here, the status quo clearly gives State A less than war would (the value of war equals the war outcome minus costs).

State A has an incentive to challenge any status quo that gives it less than war would, because it could do better by going to war. The same holds for State B. If the status quo is within the bargaining range, however, neither State A nor State B could improve its outcome by going to war, so neither state will challenge the status quo.

103Do Wars Happen by Mistake? War from Incomplete Information

The U.S. demand that Afghanistan hand over Osama bin Laden and stop harboring the Al Qaeda terror- ist network after the 9/11 attacks is an example of compellence.

Deterrence, by contrast, is used to preserve the status quo by threatening the other side with unaccept- able costs if it seeks to alter the current relationship. A deterrent threat takes the form of “Don’t do x, or else” (where x is some possible future action that the threat- ener finds objectionable). The most common deterrent threat is one that all states make implicitly all the time: “Don’t attack me, or I’ll fight back.” The effort to deter attacks on one’s own country is referred to as general deterrence, and it is an activity that states are constantly engaged in. Another form of deterrence occurs when a state seeks to protect a friend. In this case, the deter- rent message takes the form of “Don’t attack my ally x, or else.” This kind of threat is generally referred to as extended deterrence because in this case the threat- ener attempts to extend protection to another state. Extended deterrence is crucial in the context of alli- ances, a subject that we will consider in Chapter 5.

In general, crises may involve a combination of deterrent and compellent threats. When one side tries to compel another, the target may issue deterrent threats, or some third state may issue a deterrent threat on the target’s behalf. Cri- ses often involve such threats and counterthreats, as each side tries to improve its bargaining position by bringing to bear its capability to harm the other. In many cases, threats alone will succeed in bringing about an outcome that both sides find acceptable. Indeed, the most effective threats never need to be carried out, since they coerce the target into making the desired concessions or refraining from objec- tionable actions. When this contest of threats fails to generate an outcome that both sides prefer over fighting is when we observe the descent into war. This brings us back to the core puzzle: Given the costs associated with war, why does crisis bar- gaining sometimes fail to achieve a peaceful solution? The following sections elaborate several possible answers.

Do Wars Happen by Mistake? War from Incomplete Information In July 1990, Iraq was engaged in coercive diplomacy with its small neighbor to the south, Kuwait. Two years earlier, Iraq had emerged from a disastrous, eight-year war with Iran, and Iraqi leader Saddam Hussein desperately needed to rebuild his

In a 2012 speech to the United Nations, Israeli prime minister Benjamin Netanyahu declared that Israel would not permit Iran to reach the final stage of acquiring a nuclear weapon. He argued that a credible threat of military action was needed to deter Iran from crossing that “red line.”

deterrence An effort to preserve the status quo through the threat of force.

104

CON T ROV E R S Y

Can We Negotiate with North Korea? North Korea has one of the most oppressive regimes on the planet. It has been governed by a dictatorship since 1948, passed down across three generations. The regime, now led by Kim Jong-Un, has murdered, tortured, and starved its people into submission. It has sealed them off from the outside world, denying virtually all contact with relatives in South Korea and with other foreigners. In foreign affairs, North Korea has waged war on its southern neighbor, spewed threats at others in the region, sold missile technology to other “rogue” regimes, and kidnapped Japanese citizens.

North Korea has also spent the last three decades developing nuclear weapons. In 2006 it test-detonated a nuclear device, and the country is now believed to have somewhere between 20 and 60 warheads. In 2017, North Korea tested a ballistic missile capable of hitting not just its immediate neighbors, but also anywhere in the United States. As a result, the United States and other countries have sought to persuade the regime to end its nuclear program and give up its weapons.

Although bargaining pro- vides a way for states to settle disputes and avoid the costs of war, the question of whether to negotiate with North Korea over its nuclear ambitions has been fraught with controversy. Disagreements revolve around whether the interests moti-

vating North Korea’s quest for these weapons make nego- tiations viable and whether, with its autocratic domestic institutions, North Korea is likely to live up to any agree- ment it signs. There is also disagreement about the role that military force should play in this interaction.

The argument against negotiating hinges on the view that North Korea neither wants a deal nor can be trusted to honor one. From this perspective, the regime’s interest in nuclear weapons stems from a desire to reunify the Korean peninsula by force. Nuclear weapons are useful because they would deter the United States and Japan from coming to South Korea’s aid in the event of war. Thus, negotiations are unlikely to succeed, and they serve only

to give North Korea time to advance its program. In addi- tion, North Korea’s closed, highly autocratic regime pres- ents a key impediment, since it can easily cheat. Because it can advance its nuclear program in secrecy, away from the prying eyes of satellites and inspectors, it could break a deal without being detected.

Doubts about negotiating are also based on a series of broken agreements between the United States and North Korea over the last three decades. After nearing the brink of war in 1994, the United States and North Korea signed the Agreed Framework, under which North Korea pledged to freeze construction and operation of all nuclear reactors. In return, the United States promised oil, two proliferation- resistant nuclear reactors, and to work toward a normal- ization of trade and diplomatic relations, which have been suspended since the Korean War (1950–53).

While this agreement froze North Korea’s plutonium production, the deal collapsed in 2002 after the United States accused North Korea of circumventing the agree- ment by engaging in uranium enrichment. Subsequent deals in 2007 and 2011 also unraveled within a year or two of signing. The end result of decades of negotiation is a nuclear-armed adversary. In light of this situation, some argue, the ultimate goal of U.S. policy should be to change the regime, not to negotiate with it.

The alternative argument is that North Korea’s inter- est in nuclear weapons stems not from ambition, but from fear: the regime is built on suspicion of outsiders and wor- ries about being overthrown. This insecurity is not unrea- sonable, since the United States has, in both rhetoric and practice, shown its willingness to oust hostile dictators. In

U.S. State Department officials return from a diplomatic mission in North Korea with boxes of information on the country’s nuclear operations.

Applying the Concepts

105

2003 the United States invaded Iraq and removed Saddam Hussein on the basis of inflated claims about that country’s WMD capabilities; in 2011 the United States and its allies intervened in a civil war against Libyan leader Mu’ammar Qaddafi, who had earlier agreed to give up his nuclear and chemical weapons programs. Watching these events could have easily convinced Kim that the possession of nuclear weapons would guard against a similar fate.

In this view, negotiations might work, but only if Washington can credibly commit not to seek regime change if North Korea lowers its defenses. Such a commitment would involve economic assistance to help bolster the regime and promote development. In past negotiations, North Korea also sought security guarantees, includ- ing a formal promise that the United States would not attack it and a reduction of the U.S. military presence in South Korea and Japan. Whether a comprehensive deal along these lines could have worked when North Korea’s program was in its infancy, now that Kim actually has a nuclear arsenal and means of delivery it is unclear what concessions would convince him to give them up.

In early 2018, U.S. president Donald Trump signaled his willingness to hold direct talks with the North Korean leader. At the time of this writing, it was unclear whether such negotiations would take place or what would come out of them. Although a meeting between the heads of state would be unprecedented, failed talks or an agree- ment that soon collapses would not.

If a lasting deal is not reached, the question becomes, What role will the threat of military force play in this inter- action? Key advisers to Trump have argued that, given North Korea’s aggressive intentions and barbaric regime, the United States cannot tolerate that country’s capability to hit the United States with a nuclear weapon. In this view, the United States and its allies have to apply maximum pressure to induce disarmament, including either a lim- ited military strike to convince Kim of the costs of defiance or a broader attack designed to eliminate his arsenal.

Those who believe that the regime is motivated by fear worry that such a strategy is both unnecessary and counterproductive. If Kim is interested in his own survival,

then the threat of retaliation by the United States can deter North Korea from using nuclear weapons. If so, there is no urgency to eliminate the program by force—a prospect that carries enormous risks and costs.

An attempt to disarm North Korea by force might fail to destroy its arsenal, creating the risk of a nuclear attack on South Korea, Japan, or the United States. Even if North Korea did not use its nuclear weapons, its enormous conventional capabilities, including thousands of artil- lery pieces within range of Seoul, could inflict horrendous losses on South Korea and the U.S. troops stationed there. The United States would likely prevail in a war, but at a cost that could run to hundreds of thousands of lives. Ultimately, the price of war may mean that there is no alternative but for the two sides to talk about a way to manage this dangerous rivalry.

North Korean dictator Kim Jong-Un controls the country’s military and nuclear arsenal.

Thinking Analytically 1. How might Kim Jong-Un respond to a limited

military strike by the United States? How do different assumptions about his interests lead to different expectations about his response?

2. Should countries that value human rights negotiate with regimes that brutalize their own people? What are the costs and benefits of engaging a regime like North Korea?

106 Chapter 3 Why Are There Wars?

shattered economy. Given that Iraq was sitting on oil deposits containing an esti- mated 112 billion barrels of oil, it was not hard to imagine where money for recon- struction would come from. But Hussein was not content simply to pump and sell Iraq’s own oil, and he quickly turned his gun sights on Kuwait.

Kuwait’s 95 billion barrels of recoverable oil reserves made a tempting tar- get, and moreover, Hussein felt that his neighbor was standing in the way of his plans for economic recovery. For one thing, Kuwait was pumping more oil than it had previously agreed to. This extra supply meant that the price of oil was lower than it would otherwise be, depriving Iraq of needed revenue. Iraq also charged that Kuwait was stealing oil from oil fields that straddled the countries’ shared border. Finally, Kuwait had lent Iraq substantial sums of money during the war, and Saddam Hussein hoped to get this debt forgiven. When Kuwait refused these demands, Iraq flexed its military muscle. Beginning in mid-July, Iraq started mov- ing its forces closer to the border with Kuwait, at one point moving an entire divi- sion per day. By the end of the month, 100,000 Iraqi troops, supported by thousands of tanks, were in position near the border.

With the help of spy satellites, American officials watched the buildup, and they passed on their intelligence to the Kuwaitis. Despite their concern, most offi- cials in the U.S. government did not anticipate that Iraq would invade, nor did the Kuwaitis bow to Hussein’s demands. To many observers, the Iraqi moves looked like an effort to intimidate, not a prelude to invasion. According to reporter Bob Woodward’s account, “Everything Saddam had to do to prepare for an invasion was exactly what he also had to do if his intention was simply to scare the Kuwaitis. There was no way to distinguish the two.”17 On August 2, with Kuwait still holding out, Hussein revealed that he had not simply been bluffing: Iraqi forces swept into Kuwait and fully occupied the country in a matter of hours.

In retrospect, it is clear that Iraq was willing and able to wage war against Kuwait if its demands were not met. At the time, however, key decision makers in Kuwait and Washington were not sure of Hussein’s intentions. Would he really risk the wrath of the world by gobbling up his small neighbor? Would he be willing to put his military and his country through another war so soon after the previous one? The invasion of Kuwait happened in part because, unsure of the answers to these questions, the Kuwaitis decided that calling Hussein’s bluff was preferable to giving in to his demands. When it became apparent that the threat was not, in fact, a bluff, war was already upon them.

This episode illustrates one reason that bargaining can fail to resolve disputes and avert war. When states have poor and incomplete information about each other’s willingness and ability to wage war, two mistakes are possible, both of which can lead to war. First, a state confronted by demands may mistakenly yield too little or not at all—just as Kuwait failed to budge in the face of Hussein’s threats. In these cases, bargaining can break down because at least one state feels that it can achieve more through fighting (in this case, Iraq) than the other is willing to offer in the

17. Bob Woodward, The Commanders (New York: Pocket Star Books, 1991), 200.

107Do Wars Happen by Mistake? War from Incomplete Information

negotiations (Kuwait). The second, related danger is that a state may demand too much (Iraq in this example) under the mistaken belief that the other side will cave in (Kuwait, and its protector, the United States). In this event, the state may not realize its mistake until war is already upon it. In either case, even though there might be a settlement of the issue that both sides would prefer to war, uncertainty about each other’s willingness to wage war can prevent such a settlement from being reached.

Where does this uncertainty come from? Recall that the main issue in crisis bargaining is how each state evaluates its prospects in a war. How likely is it that the state will be able to win the war? What will the human, financial, and political costs be? These assessments are important because each state’s value for war determines what bargains it prefers over fighting. If one state is uncertain of how much its adversary values war, then it is also uncertain of how much it must concede in order to prevent a war. This uncertainty will arise whenever a state lacks information about any of the myriad factors that determine its adversary’s evaluation of war.

A poker analogy is useful here. In poker, the fact that at least some cards are hidden from view means that each player knows more about the strength of her own hand than do her opponents. The hidden cards are what might be called pri- vate information: important facts that are known only to the player who observes those cards. Because no player sees all the cards, the game is played under a condi- tion known as incomplete information. All players lack information about their opponents’ hands and have private access to information about their own hands.

Incomplete information arises in crisis bargaining when states cannot read- ily observe or measure the key political and military factors that determine their adversaries’ expected value for war. The hidden cards in this context can be many and varied, and we typically differentiate between two broad classes of unknowns: capabilities and resolve. Capabilities are the state’s physical ability to prevail in war: the number of troops it can mobilize, the number and quality of its armaments, the economic resources it has to sustain the war effort. We might also include in this list the quality of the country’s military leadership and mili- tary strategies. In addition, since third parties sometimes join wars on one or both sides, any uncertainty about what those third parties will do can lead to uncer- tainty about the capabilities that each side will bring to bear in the event of war (see Chapter 5).

Resolve, a more abstract concept, refers to a state’s willingness to bear the costs of fighting and how much the state values the object of the dispute relative to those costs. How many people is the country willing to lose in order to obtain, say, a given piece of territory? How much is it willing to pay in blood and money in order to win policy concessions or change another country’s regime? Resolve not only determines whether a state is willing to fight, but also how much of the state’s potential capabilities would actually be mobilized in the event of war. We frequently make the distinction between total wars (in which states mobilize their entire military and economic resources) and limited wars (in which states fight with something less than their full potential, often because their aims are limited or of relatively low value).

incomplete information A situation in which actors in a strategic interaction lack information about other actors’ interests and/or capabilities.

resolve The willingness of an actor to endure costs in order to acquire a particular good.

108 Chapter 3 Why Are There Wars?

How a state evaluates the stakes of conflict determines where on this contin- uum its effort will be. World War II, for example, was seen as a war of national survival, and during the height of the war (1943–45) the United States spent about 37 percent of its gross domestic product (GDP) on defense, and other belligerents spent half of their GDP or more. By contrast, total U.S. defense spending in 2008, in the midst of wars in Iraq and Afghanistan, amounted to about 4 percent of GDP, and only a portion of that paid for those wars. Resolve is obviously a difficult quality to measure, as it hinges on a variety of political, ideological, and psychological factors. Indeed, it may be difficult for a leader to accurately assess his or her own country’s resolve, much less the resolve of its adversary.

How could such uncertainty lead to war? When states have incomplete infor- mation about the capabilities and/or resolve of their opponents, bargaining over goods that they both desire may fail to achieve peaceful settlements. A central dynamic of bargaining under this kind of uncertainty is a phenomenon known as a risk-return trade-off: essentially, there is a trade-off between trying to get a good deal and trying to minimize the possibility that war will break out.

At one extreme, a state can generally ensure peace by capitulating to all of its adversary’s demands. “Peace at any price” might not be a very attractive outcome, however. Kuwait, for example, could have given in to all of Iraq’s demands and would likely have avoided war. At the other extreme, a state can hold firm and yield nothing to its adversary. This strategy promises a good deal if it works, but doing so runs a risk that the adversary will decide to fight rather than settle for nothing—as Iraq did when Kuwait refused to give in. Between these extremes, a state will gen- erally find that it can reduce the risk of war only by making more generous offers, moving farther from its ideal outcome. Put another way, a state can improve the bargaining outcome for itself only by embracing a higher risk of war. Although war is costly and regrettable in retrospect, bargaining strategies that entail a risk of war can be perfectly rational, given the uncertainty that states face.

In September 2016, Russia initiated massive military drills, showcasing the country’s ships, cruise missiles, and antiaircraft system, off the coast of Crimea. How much of a country’s capabilities it is willing to mobilize for war can be a source of uncertainty in bargaining.

risk-return trade-off In crisis bargaining, the trade-off between trying to get a better deal and trying to avoid a war.

109Do Wars Happen by Mistake? War from Incomplete Information

Incentives to Misrepresent and the Problem of Credibility Given that incomplete information can lead to war, why can’t states simply tell each other how capable and resolved they are and thereby avoid war? Actually, a large part of what goes on in a crisis consists precisely of such efforts at communication. Crises are generally characterized by diplomatic exchanges, threats and counter- threats, mobilization of forces, movement of troops, and positioning of weaponry. These actions, in part, have a military purpose: one cannot wage war, after all, without first mobilizing the necessary forces and putting them in place. But these actions also have a political purpose: they are the language of coercive diplomacy, the vocabulary that states use to convince one another that they are willing to back their bargaining positions with the threat of force.

The problem that arises in this context is that, as much as states may have an interest in communicating their hidden information, they may not always be able to do so effectively. A crucial question that arises in crisis bargaining is whether the messages a state sends have credibility. A credible threat is a threat that the target believes will be carried out. We say a threat lacks credibility if its target has reason to doubt that the threat will be carried out. The credibility of a threat refers not only to the belief that the threatener will start a war; the target also has to believe that the threatener is willing to fight long enough and hard enough that giving in to its demands is a better option. In the case of the U.S. war against Afghanistan, the Taliban government probably had little doubt that President Bush would implement his threat to invade, but it may have doubted that the United States would be willing and able to bear the costs of a long war. In other words, the threat to start a war was credible, but the threat to remove Taliban leaders from power was, in their eyes, not.

Note that the credibility of the threat refers to the target’s beliefs, not the actual intentions of the state issuing the threat. A state may fully intend to carry through on a threat, but it may have a hard time convincing its adversaries. Saddam Hussein genuinely intended to invade Kuwait if his demands were not met; observers in the United States and Kuwait, however, did not see the threat conveyed by his mobilization as credible. Similarly, a state may have no intention of carrying out its threat, but the target may mistakenly believe otherwise. In such cases, the bluff can succeed.

Why is credibility hard to achieve? There are two interrelated reasons. First, carrying through on threats is costly. A state may say that it will wage war if its demands are not met, but the costs of war might be such that it would not make sense to fulfill this threat if called on to do so. This concern about credibility was particularly pronounced during the Cold War between the United States and the Soviet Union. With both sides in possession of large arsenals of nuclear weapons, it was well understood that war could quickly escalate into total annihilation. Given this situation, officials in the United States worried about how they could deter the Soviet Union from attacking Western Europe. Would the United States really risk New York to save London or Paris? If the Soviets believed the answer to

credibility Believability. A credible threat is a threat that the recipient believes will be carried out. A credible commitment is a commitment or promise that the recipient believes will be honored.

110 Chapter 3 Why Are There Wars?

this question was no, then the U.S. commitment to defend Western Europe would lack credibility. It was precisely these kinds of concerns that led Great Britain and France to develop their own nuclear capabilities in order to deter the Soviets; after all, it was much more credible that France would risk Paris to save Paris.

Even without the prospect of nuclear annihilation, threats may lack credibil- ity because their targets appreciate the costs of carrying them out. In the midst of the crisis between Iraq and Kuwait in 1990, the United States announced joint naval exercises with the United Arab Emirates. On the same day, the State Department spokesperson issued an extended deterrent threat by reaffirming the U.S. commit- ment to protect its friends in the Persian Gulf. Iraq’s response was contemptuous. Hussein called the U.S. ambassador to his office the next day and told her that he was not scared by American threats. After all, he reportedly said, “Yours is a society which cannot accept 10,000 dead in one battle.”18 Thus, the U.S. deterrent threat had little credibility in Hussein’s eyes because he believed that the United States would be unwilling to bear the costs of war.

The second reason that credibility is hard to achieve stems from the conflict- ing interests at the heart of the bargaining interaction. Even though states have a common interest in avoiding war, each also wants the best possible deal for itself, so they have incentives to hide or misrepresent their information.

In some cases, this incentive means that states will conceal information about their true strength. After Iraq invaded Kuwait, the United States massed a large force in Saudi Arabia and threatened war unless Iraq retreated. It was widely assumed that if war came, U.S. forces would attack Iraqi positions in Kuwait head-on. Such a strategy would have provided the most direct route to the objec- tive, but there were clear costs involved: the Iraqi forces in Kuwait were dug in behind strong defenses, including trenches filled with oil that could be lit on fire as soon as U.S. troops tried to cross.

Saddam Hussein’s resistance to U.S. pressure stemmed in part from his belief that his defenses would make the liberation of Kuwait very costly. In fact, war plan- ners in the United States decided early on that they would not attack directly at the strength of the Iraqi positions. Instead, they secretly shifted the bulk of the U.S. force into the desert west of the Kuwaiti border. The military plan called for a “left hook”: U.S. tanks would enter Iraq on the western flank of Iraqi forces in Kuwait and then swoop around behind them, thereby outflanking the enemy’s fortifica- tions. This tactical decision meant that the United States expected to incur lower casualties than Iraq expected to be able to inflict.19

Theoretically, if the United States could have communicated these expectations to Iraq, its threat would have been more credible and perhaps Hussein would have decided to back down. But it is easy to see why the United States could not say to

18. A transcript of this meeting was published in the New York Times, on September 22, 1990. Hussein may also have been encouraged to ignore U.S. threats because American ambassador April Glaspie seemed to indicate that the United States had no strong interest in the outcome of the Iraq-Kuwait dispute.

19. For a discussion of Persian Gulf War strategy, see Lawrence Freedman and Efraim Karsh, “How Kuwait Was Won: Strategy in the Gulf War,” International Security 16 (Autumn 1991): 5–41.

111Do Wars Happen by Mistake? War from Incomplete Information

Hussein: “You think that war will be too costly for us, but you are mistaken. Rather than attack your forces head-on, we will go around them on the western flank.” Had the United States sent such a message, Iraq could have taken measures to counter the tactic, such as by repositioning its forces to the west. Hence, any bargaining advantage that the United States might have reaped by revealing its strength would have been nullified. In the strategic context of the crisis, it made sense for the United States to hide its strongest cards.

In other cases, states misrepresent in order to hide their weakest cards. Any- one who has played poker knows that it sometimes makes sense to bluff—that is, to act as if one has a strong hand in the hopes that others will fold. A similar incentive exists in international crises. In this context, a bluff is a threat to use force that the sender does not intend to carry out. In a crisis, a successful bluff could reap large rewards. In 1936, Germany marched its military forces into a region on its border with France known as the Rhineland—a region that, by the 1919 treaty that ended World War I, Germany was required to keep demilitarized. Hitler sent his forces in anyway, daring the Western powers to stop him. Though alarmed by this move, both France and Great Britain chose not to risk a full-scale war over the issue, and the remilitarization of the Rhineland took place unopposed. Interestingly, there is good reason to believe that Hitler’s move was a bluff. Although this point is somewhat controversial, there is evidence that German troops were under orders to retreat if confronted.20 If so, then one of the key moments in the lead-up to World War II was a well-executed bluff.

This observation raises a dilemma: How can states credibly convey their infor- mation in order to diminish the risk of war due to uncertainty? Given a strategic environment that sometimes rewards misrepresentation, how can a genuine threat be made believable?

Communicating Resolve: The Language of Coercion To help us answer these questions, another example will be helpful. On June 25, 1950, without any warning, North Korea invaded South Korea. The Korean peninsula had been split in two after World War II, divided at the 38th parallel between the communist North and the non-communist and pro-Western South. North Korea’s attack was a bold attempt to reunify the country under communist rule, and the United States quickly joined the South in repelling the attack. After three months, the U.S. efforts were largely successful, and North Korean forces began to retreat to their side of the 38th parallel. At this point, the United States decided to press the attack, cross into North Korea, and topple the communist regime there.

This possibility raised grave concerns in neighboring China, which had only the previous year been taken over by a communist government. On October 3,

20. James Thomas Emmerson, The Rhineland Crisis, 7 March 1936 (Ames: Iowa State University Press, 1977), 98–100.

112 Chapter 3 Why Are There Wars?

1950, Chinese diplomats conveyed a message through the Indian ambassador that a move across the 38th parallel would trigger Chinese intervention.21 Nonetheless, the threat went unheeded. The U.S. operation had been planned and authorized under the assumption that the Chinese would not intervene, and the October 3 warning did nothing to change any minds.22 U.S. forces crossed into North Korea on October 7 and advanced rapidly. In response, 600,000 Chinese troops poured into the Korean peninsula, leading to three more years of fighting and a costly stalemate.

Why was the Chinese threat dismissed, in the words of Secretary of State Dean Acheson, as “a Chinese Communist bluff ”? An October 4 memorandum describes Acheson’s rationale:

The Secretary pointed out that the Chinese Communists were themselves taking no risk

in as much as their private talks to the Indian Ambassador could be disavowed. . . . if they

wanted to take part in the “poker game” they would have to put more on the table than

they had up to the present.23

Acheson’s reasoning for downplaying the Chinese threat is instructive. The Chinese government was making an extended deterrent threat: “Don’t invade North Korea, or we will intervene to defend it.” From the U.S. perspective, it was possible that China would actually make good on this threat, but it was also possi- ble that China was simply trying to bluff the United States into staying out of North Korea. Regardless of which possibility was true, the message conveyed through the Indian ambassador was cheap and easy to send. There was nothing in the mes- sage or the way it was sent that would give American decision makers reasons to think that China was not simply bluffing. Unless the Chinese were willing to pay some costs—“to put more on the table”—there was little reason to take their threat seriously.

This example suggests a more general insight: For threats to be credible, they have to be costly in such a way that the sender would make the threat only if it really intended to carry the threat out. Consider the problem that the United States faced in October 1950 in these, admittedly oversimplified, terms:

It is possible that the Chinese government is resolved to intervene if we attack North Korea, and it is possible that it is not so resolved. How can we know if we are facing a “resolute” China or an “irresolute” China? What would we look for to distinguish these two “types” of adversaries? The answer is this: We would want to look for actions that a resolute China would be willing to take but an irresolute China would be unwilling (or, at least, less likely) to take. If we see such actions, then we are more

21. The United States did not recognize the People’s Republic of China as a legitimate government, so the two countries did not have direct diplomatic contacts.

22. See, for example, William Stueck, Rethinking the Korean War (Princeton, NJ: Princeton University Press, 2002), chap. 4.

23. U.S. Department of State, Foreign Relations of the United States, 1950, vol. 7, Korea (Washington, DC: U.S. Government Printing Office, 1976), 868–69.

113Do Wars Happen by Mistake? War from Incomplete Information

likely to be facing a resolute China, and we have to take its threat seriously. If we do not see such actions, then there may be reason to doubt China’s resolve.

The message that the Chinese actually sent did not have much credibility because an irresolute China could just as easily have made the same claim.

What kinds of actions can help an opponent distinguish whether its adversary is resolved? In general, the literature has identified three mechanisms that states use to make their threats credible: brinksmanship, tying hands, and paying for power.

Brinksmanship: The “Slippery Slope” Some of the earliest scholarship on the ques- tion of how to make threats credible took place in the 1950s, when policy makers and academics were preoccupied by the credibility of threats in the nuclear age. If everyone understood that nuclear war would bring total annihilation to each side, then under what conditions could threats between nuclear powers ever be credible? Since no state would ever intentionally “pull the trigger” and bring about Armaged- don, the threat to do so was not credible. This observation raised the question of whether nuclear weapons had any utility at all in the emerging Cold War between the United States and the Soviet Union.

The most important insight into this issue came from Thomas Schelling, an early theorist of the strategy of crisis bargaining. In Schelling’s view, although it was understood that no state would intentionally bring about its own destruction by starting a total nuclear war, these weapons could nonetheless be wielded for diplomatic effect through a strategy known as brinksmanship. The basic idea was that states could signal their resolve in the crisis by approaching the “brink” of war through provocative actions. Schelling describes this concept:

The brink is not, in this view, the sharp edge of a cliff where one can stand firmly, look

down, and decide whether or not to plunge. The brink is a curved slope that one can stand

on with some risk of slipping, the slope gets steeper and the risk of slipping greater as one

moves toward the chasm.24

The costs of war are such that, if faced with a simple choice of whether to jump or not, no sane decision maker would jump. But rational leaders might decide to step out onto the “slippery slope” and thereby increase the risk that war would start inadvertently. Schelling famously referred to such an act as a “threat that leaves something to chance.”25 The willingness to take such a chance separates resolute from irresolute adversaries. After all, the less the state values the good in dispute, and the more it fears a war over that issue, the less willing it will be to step onto the slope and embrace a risk of war.

In a brinksmanship crisis, each side bids up the risk of war—moving further and further down the slippery slope—until either one side decides to give in or they fall together into the precipice (as in the game of Chicken discussed in the

24. Thomas C. Schelling, The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960), 199.

25. Schelling, Strategy of Conflict, 187.

brinksmanship A strategy in which adversaries take actions that increase the risk of accidental war, with the hope that the other will “blink” (lose its nerve) first and make concessions.

114 Chapter 3 Why Are There Wars?

“Special Topic” appendix to Chapter 2, on pp. 84–85). Exactly how a war might start “inadvertently” is not always clear; fortunately, in the case of nuclear war, we do not know the answer to that question. Absent a computer glitch, it still takes a human hand to pull the trigger.

The general idea was that as tensions rose in an international crisis, the risk of accidents would increase. A limited skirmish between forces could inadvertently escalate if a nervous local commander thought his position was about to be overrun and decided to launch his tactical nuclear weapons in order not to lose them to the enemy. Or, in the midst of a tense crisis, a flock of geese might be mistaken on the radar for incoming bombers (as actually happened in the 1950s), leading to a deci- sion to launch rather than risk being disarmed by a first strike (which, fortunately, did not happen in this incident). Alternatively, the tension of a nerve-wracking cri- sis might cause leaders to give in to passion and fury and lose their cool, rational heads. In any event, it was precisely the willingness to court this risk that would credibly separate the genuinely resolved opponents from the bluffers.

Tying Hands A second way in which states can send credible signals of their will- ingness to fight is by making threats in ways that would make backing down dif- ficult. For example, after Iraq invaded Kuwait in August 1990, President Bush repeatedly and publicly stated that the conquest “will not stand.” He made this commitment first on August 5, 1990, and reiterated it throughout the crisis, includ- ing during his State of the Union address on January 29, 1991. These words were also matched by deeds, particularly the deployment of over 500,000 U.S. troops to the region and an extensive diplomatic effort to build international consensus for an attack. Unlike the case with the Chinese threat, which Acheson felt could be

Leaders may use brinksmanship to make the threat of nuclear war credible. In the 1962 Cuban missile crisis, American president John F. Kennedy took steps toward nuclear war—for example, putting missile crews on alert—in order to pressure the Soviets to dismantle nuclear missile sites in Cuba.

115Do Wars Happen by Mistake? War from Incomplete Information

disavowed, Bush was clearly on record as asserting that the policy of his administra- tion was to reverse the Iraqi invasion.

By taking such clear, public statements and actions, Bush put his reputation, and that of the country, on the line. It was not unreasonable to expect that in doing so, he had made it quite costly to retreat from this position—to decide, in the face of Iraqi resistance, that the invasion would be allowed to stand after all. Doing so would have been embarrassing for him as a leader and for the country as a whole. It would have called into question the credibility of future U.S. threats, and Bush could expect his political opponents to use such a retreat against him at the next election.

The general insight here is that under some conditions, threats can generate what are known as audience costs—that is, negative repercussions that arise if the leader does not follow through on the threat.26 Two audiences might impose these costs. The first is the international audience—other states that might doubt future threats made by the president or by the country. Such international audience costs could be felt through an inability to convince future adversaries of one’s resolve. For example, one of the reasons that U.S. decision makers were skeptical of the Chinese threat to intervene in the Korean War was that it came in the wake of a series of unfulfilled threats over a different issue: Taiwan. As early as March 1949, the Chinese communist government began issuing threats to “liberate” Taiwan from the Nationalist Chinese forces that had fled there. These threats, repeated several times over the course of the next year and a half, combined with China’s failure to carry them out, led some analysts in the United States to discount the threats over Korea as similarly empty bluster.27 International-audience costs might also arise if allies come to doubt the trustworthiness of the country’s threats, which may lead them to seek other protectors (see Chapter 5).

A second audience that might punish a leader for backing down from a threat is in the leader’s own country: voters and political opponents who might seek to repri- mand a president who has tarnished the country’s honor and reputation by mak- ing empty threats. For example, in 2012 U.S. president Barack Obama declared that any use of chemical weapons by the Syrian regime against its people would cross a “red line” and compel a forceful response. His subsequent decision not to strike Syria after it used the banned weapons the following year was widely criticized by domestic opponents and used to foster the perception that Obama was weak on foreign policy.

In either event, if threats expose state leaders to audience costs, they can have the effect of “tying their hands.” This phrase comes from a famous scene in Homer’s Odyssey in which the main character, Odysseus, asks to have his hands tied to the mast of his ship as they sailed past the sirens. The call of the sirens was so beauti- ful that sailors who heard it were bewitched into steering their ship into the rocks.

26. James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88 (September 1994): 577–92. For recent arguments that audience costs do not play a major role in crises, see Marc Trachtenberg, “Audience Costs: An Historical Analysis,” Security Studies 21 (2012): 3–42; and Jack Snyder and Erica D. Borghard, “The Cost of Empty Threats: A Penny, Not a Pound,” American Political Science Review 105 (August 2011): 437–56.

27. Anne Sartori, Deterrence by Diplomacy (Princeton, NJ: Princeton University Press, 2005), 33–39.

audience costs Negative repercussions for failing to follow through on a threat or to honor a commitment.

116 Chapter 3 Why Are There Wars?

By tying his hands to the mast, Odysseus hoped to experience the sirens’ call with- out succumbing to this enchantment. State leaders in a crisis tie their hands for similar reasons: because threats are costly to carry out, they know that they might choose not to follow through if faced with that decision. By exposing themselves to audience costs, they blunt the temptation to back down from their threats and thereby tie their own hands.

In the process of tying their hands, of course, state leaders also send a powerful message to their opponents: “I cannot back down; hence, my threat is completely credible.” As with acts of brinksmanship, engaging in actions that incur audience costs separates the resolute from the irresolute. Those most likely to back down from a threat are less willing to take steps that will make backing down costly.

Paying for Power A final mechanism that states use to signal their resolve in a crisis involves taking costly steps to increase their capabilities, such as by mobilizing and deploying a large military force, increasing military manpower, and/or spending large sums of money. For example, during a 1961 crisis between the United States and the Soviet Union over Berlin, President Kennedy asked Congress to increase the army’s strength by 350,000 troops, with much of the growth coming from an increase in the number of people drafted and a call-up of reserve forces. This last step was particularly costly, in a political sense, because reservists are people who have finished their mili- tary service and have families and regular jobs. As a result, calling up 150,000 reservists led to political resistance. There was also a concrete monetary cost to these actions.

As already noted, military mobilizations play an important role in crises because they give the states the material capabilities needed to back up their threats. The most credible threat is not going to be effective unless the action threatened is very costly to the adversary. Thus, one purpose of military mobilization is to put force behind diplomatic maneuvers. At the same time, costly actions of this sort can also affect the opponent’s estimate of the state’s resolve, through several mechanisms.28 First, by increasing the state’s military capabilities, these actions may decrease the costs associated with carrying out the threat. Since the costs of carrying out a threat are what call credibility into question in the first place, visible steps taken to reduce those costs can make it more believable that the threatener will do what it says. Second, the willingness to pay the costs associated with these actions can signal that this issue is one that the threatening state cares a great deal about. By calling up the reserves, Kennedy was trying to send the message that he cared so much about the fate of Berlin that he was willing to pay the political costs associated with this action.

We started this section by noting that a condition of incomplete information is dangerous because states might miscalculate in bargaining, such as by mistakenly resisting threats that turn out to be genuine. Such mistakes played a role in the Iraqi invasion of Kuwait in 1990 and in the onset of war between the United States and China in 1950. Brinksmanship, tying hands, and paying for power are strategies for

28. Branislav Slantchev, Military Threats: The Costs of Coercion and the Price of Peace (Cambridge: Cambridge University Press, 2011).

117Do Wars Happen by Mistake? War from Incomplete Information

communicating a willingness to fight; they are mechanisms that help states figure out which threats are genuine and which are not. Interestingly, though, these cures for incomplete information entail risks that can be just as dangerous as the underlying problem they address.

This danger is clearest in the case of brinksmanship strategies, through which states bid up the risk of accidental war in order to prove their resolve. The hope in doing so is that the adversary will blink, and thus war will be averted. But there is no guarantee that the states will not lose control and fall into the precipice. Ironi- cally, then, to avoid a war driven by uncertainty, states have to embrace some risk of accidental war. Hand-tying strategies can have similar risks. Although actions that raise the costs of backing down can convince the other side to give in, there is also a risk that both sides will tie their hands in a contest of threats and counterthreats. Both sides can then become locked into intransigent and incompatible bargaining positions from which it is too costly to retreat. Once each side has eliminated its ability to compromise, war may be inevitable—even if the initial uncertainty that led to the crisis has been removed. Finally, while military mobilizations might per- suade the adversary to yield, they can also provoke the other side to strike first—a possibility that we will revisit shortly.

In the case of the Persian Gulf War, it is quite possible that President Bush’s hand-tying actions between August 1990 and January 1991 eventually convinced Saddam Hussein that the United States was willing to fight, in spite of his original belief to the contrary. He may have resisted nonetheless because his own pattern of defiance and counterthreats in those months had served to tie his own hands as well. Hussein may have feared domestic repercussions for a retreat in the face of American threats, or he may have been concerned that such a retreat would embolden Iraq’s neighbor and long-standing enemy, Iran. By January 15, 1991, neither side was willing to compromise, and the U.S.-led war to liberate Kuwait began. We thus see that incomplete information can cause war both directly, through miscalculation, and indirectly, by forcing states to communicate their resolve in ways that can foreclose successful bargaining.

This discussion generates several predictions about the conditions that make war more or less likely. In general, the harder it is for states to learn about each other’s capabilities and resolve, the more severe the problem of incomplete infor- mation will be. When states are relatively opaque, in the sense that it is hard for outsiders to observe their military capabilities or their political decision-making processes, there is likely to be greater scope for uncertainties of this kind to arise and bedevil the quest for negotiated settlements. The strategic situation might also influence the degree of uncertainty that states face. For example, as the number of states that might potentially get involved in a particular crisis increases, the number and importance of the “hidden cards” grows dramatically: in the event of a war, who will join and who will not? Finally, this discussion sensitizes us to the question of whether states can find ways to signal their intentions in a credible manner. Prob- lems of incomplete information are more likely to be overcome when states can find costly ways to signal their intentions and thereby convince their adversaries to make concessions.

118 Chapter 3 Why Are There Wars?

Can an Adversary Be Trusted to Honor a Deal? War from Commitment Problems Incomplete information provides one compelling answer to the puzzle of war: uncertainty about capabilities or resolve can make it difficult for states to agree on a settlement that all sides prefer over war. The ability to identify such a settlement, however, does not always guarantee that war will be avoided. What happens if the states do not trust one another to abide by that settlement in the future?

In this section we develop a second set of explanations for why bargaining may fail. The causes of war considered here all arise from a common underlying chal- lenge: the difficulty that states can have making credible promises not to use force to revise the settlement at a later date. In this context, credibility has the same meaning as before, but we use it here to describe not a threat to use force but rather a promise not to. A credible commitment to abide by a deal is a commitment by one state assuring the other side that it will not threaten force to revise the terms of the deal. A commitment problem arises when a state cannot make such a promise in a credible manner.

The Prisoner’s Dilemma introduced in Chapter 2 is a quintessential exam- ple of such a problem. Although the prisoners in this game would like to commit to cooperating with each other, their incentives are such that a commitment to do so is not credible: when given the choice, they will prefer to defect. Commitment problems are particularly common in the absence of any enforcement mechanism, such as a court, that can hold people to their commitments. In the international sys- tem, external enforcement of commitments can be difficult to arrange (though not impossible, as we will see in the concluding section of this chapter). Next we con- sider three ways in which an inability to make credible commitments to a bargain might undermine the search for a peaceful settlement of international disputes.29

Bargaining over Goods That Are a Source of Future Bargaining Power The clearest place to see the role of commitment problems is in disputes over goods that can serve as a source of future bargaining power. The best examples of such goods are strategically important pieces of territory and weapons programs. States bargain over territory all the time, but in some cases the piece of territory in question has military significance, perhaps because it contains high ground from which one could launch an effective attack or islands that sit astride strategically

29. The ideas are introduced in Fearon, “Rationalist Explanations for War,” and are further elaborated in Robert Powell, “War as a Commitment Problem,” International Organization 60 (2006): 169–203.

119Can an Adversar y Be Trusted to Honor a Deal? War from Commitment Problems

important sea routes. For example, China and Japan have in recent years been locked in dispute over eight uninhabited islands in the East China Sea, known as the Senkaku to Japan and the Diaoyu to China. Not only are the surrounding waters rich in fish and in oil and gas deposits, but the islands are also close to stra- tegically important sea routes that are vital for naval movements in the region. As a result, control over the islands potentially impacts the countries’ relative military power.

Bargaining over weapons programs has a similar quality. In recent years, the United States has sought to pressure several states—including Iraq (prior to 2003), Libya, Iran, and North Korea—into abandon- ing the development of WMD. These efforts have met with varying levels of success. Libya agreed to dismantle its weapons programs in December 2003. North Korea agreed several times to freeze its nuclear program, but these deals unraveled, and North Korea effectively entered the nuclear club in 2006. In 2015, the United States and its partners reached a deal with Iran to freeze its nuclear program for at least a decade. As with strategically important territory, a deal on this matter does not simply resolve a dispute; it also directly affects the military capabil- ities of the participants. A country that agrees to give up a weapons program makes itself weaker by doing so.

The difficulty in bargaining over such objects is that a state will be reluctant to render itself more vulnerable to attack without credible promises that the other side will not exploit that vulnerability in the future. The state may be able to avoid war now by making concessions, but doing so entails a risk that its adversary, made stronger by the deal, will then press new claims. Unless there is some way for the other side to credibly commit not to use its newfound power, a threatened state may decide that it would rather fight today than face a future in which it is considerably weaker. Thus, even if there is some deal that is preferable to war now, if this deal will lead to a change in capabilities that can be exploited later, the state that would be rendered weaker may decide to forgo that deal and gamble on a war.

This strategic dilemma has presented an important obstacle in U.S. efforts to convince countries like North Korea and Iran to give up their nuclear programs peacefully. It is important to remember that the United States had hostile relations with these countries prior to their seeking nuclear weapons. As we already saw, the United States fought a war against North Korea from 1950 to 1953. That war ended with a cease-fire but not a peace treaty, and ever since, more than 30,000 U.S. troops have been stationed in South Korea. U.S. hostility toward the North is driven not only by the lingering issue of Korea’s division but also by the nature of its regime, which is one of the most repressive systems in the world. Similarly,

Territory with military-strategic value can be hard to bargain over because control over the good affects the states’ relative military power. The uninhabited Senkaku/Diaoyu Islands have been a source of conflict between Japan and China partly because they sit astride strategically important sea routes.

120 Chapter 3 Why Are There Wars?

hostility between the United States and Iran dates back to the 1979 revolution, which toppled Iran’s pro-American leader and installed an Islamic fundamentalist government.

To the extent that these states see their nuclear programs as a counterweight against American (and, in the case of Iran, also Israeli) power, they would be reluc- tant to give up those programs if by doing so, they rendered themselves more vulnerable to U.S. demands on other issues. The experience of Libya is instructive in this respect. As noted earlier, Libya agreed in December 2003 to end its chem- ical and nuclear weapons programs and to submit to international inspections. In return, the United States and Britain promised to normalize relations with the country and not to press for fundamental changes in the regime of Libyan leader Mu’ammar Qaddafi—even though he ran a brutal dictatorship. In March 2011, how- ever, Qaddafi faced a rebellion against his rule, and his attacks on civilians triggered military action by the United States, Britain, and other European countries. Qaddafi was ousted and then killed by rebel forces seven months later.

Whether or not this intervention would have been deterred if Libya had con- tinued its weapons programs, the lesson was clear: a promise by the United States to forswear regime change is not necessarily a credible or lasting guarantee. The success of any effort to end these programs peacefully requires the United States to find a way to commit credibly not to exploit the power shift brought about by disarmament.

Prevention: War in Response to Changing Power A second, related problem arises if the balance of military capabilities is antici- pated to change because of factors external to the bargaining process. A common source of this kind of power shift is different rates of economic growth. As noted in Chapter 1, uneven economic development has led to the relative rise and decline of states over time. The growth of Germany’s power in the late nineteenth and early twentieth centuries had a dramatic impact on that country’s ability to challenge its neighbors. Similarly, as we will discuss in Chapter 14, China’s impressive economic growth since 1980 has greatly increased its influence in international politics. If a state is growing much more rapidly than its adversary, then the military capabilities that it can bring to bear in future disputes will be greater than those it can bring to bear today. A second important source of large shifts in military capabilities is the development and acquisition of new technologies, such as nuclear weapons. The acquisition of nuclear weapons can cause an abrupt and profound shift in a state’s capacity to impose costs on its adversaries.

Regardless of their exact source, anticipated changes in military capabilities can present an insurmountable dilemma in crisis bargaining. To see this, we revisit the bargaining model introduced earlier. Figure 3.3 illustrates what happens to the bargaining interaction when the power of one state, in this case State A, is expected to increase. In the top panel, the expected outcome of the war is initially close to State B’s ideal point (or the far left of the bar, where State B would get all of the

121Can an Adversar y Be Trusted to Honor a Deal? War from Commitment Problems

good). Let’s assume, however, that State A’s power is expected to grow, so that, at some time in the future, the new war outcome will be closer to State A’s ideal point (or the far right of the bar), as in the lower panel.

In the initial period, the states can agree to some distribution inside the bargaining range. But both sides can anticipate that, in the future, State A will no longer be satisfied with this deal and will demand a new deal in the new bargaining range. As this example is constructed, State B prefers the war outcome it could obtain under the initial power distribution (labeled in blue in the top panel of Figure 3.3) over any outcome that falls in the future bargaining range. Even the best deal that State B can hope to get in the future (the red dotted line in the lower panel), gives State B less than it expects to get from war today. Hence, State B would rather fight a war now than face worse terms in the future.

WAR OUTCOME TOMORROW

COSTS TO A

COSTS TO B

VALUE OF WAR TO B

VALUE OF WAR TO A

A BBARGAINING RANGE

WAR OUTCOME TODAY

COSTS TO A

COSTS TO B

VALUE OF WAR TO B

VALUE OF WAR TO A

A BBARGAINING RANGE

FIGURE 3.3 Bargaining and Shifting Power Initial power distribution: Today, State B is more powerful than State A, so fighting a war favors State B, as indicated by the fact that the war outcome is close to State B’s ideal point.

Future power distribution: Tomorrow, after State A’s power increases, fighting a war will favor State A, as indicated by the fact that the war outcome is now close to State A’s ideal point. The best possible deal that State B can get after the power shift is indicated by the red dotted line. Because this deal gives State B less than it can expect to get from war today (the blue dotted line in the top panel), State B has an incentive to wage war now in order to prevent the power shift.

122 Chapter 3 Why Are There Wars?

Thus, even if the states can locate a deal that they prefer over war today, the state that is getting stronger will face a strong temptation to use its future power to try to revise the deal later. Unless there is some way for the growing state to make a credible commitment not to do that, then its adversary may decide that it is bet- ter to gamble on war today in order to stop or slow the anticipated shift. A war that is fought with the intention of preventing an adversary from becoming relatively stronger in the future is a preventive war.

Notice that this logic is compelling only if it is believed that war will halt, or significantly delay, the anticipated change in power, as might happen if war could successfully disarm the other state or cripple its economic growth. If the shift in power will happen anyway, then there is nothing to be gained by fighting now, only to have the outcome revised later in the rising state’s favor. The need to destroy the source of the adversary’s rise means that wars fought in the context of shifting power tend to be especially long and costly (see “How Do We Know?” on p. 124).

The U.S. war against Iraq in 2003 had a preventive logic, even if much of the motivating intelligence turned out to be flawed. Saddam Hussein’s regime was believed to have some minimal capability—and a demonstrated intention—to develop WMD. The preventive argument for attacking was that it would be easier to oust Hussein before he had fully developed these capabilities than to do so after he had succeeded in deploying them in his arsenal. Of course, the uncertainties sur- rounding Iraq’s weapons program illustrate a major risk of engaging in preventive war, since the rationale for doing so is only as strong as the evidence that a disad- vantageous shift in relative capabilities is coming.

Preemption: War in Response to Fear of Attack A final commitment problem that can prevent states from reaching negotiated settlements of their disputes arises from fear of attack by an opponent with a first- strike advantage. A first-strike advantage exists when there is a considerable benefit to being the first to launch an attack. It arises when military technology, strategies, and/or geography give an advantage to offensive actions over defensive ones. When defense is relatively effective, a state can afford to wait to see whether its opponent is going to attack because a first strike can be defeated or absorbed. When offense is relatively effective, a first strike is potentially devastating, and fear of attack creates incentives to abandon the bargaining table and rush to the battlefield.30

For example, if one state could launch its nuclear missiles and destroy all of its adversary’s missiles on the ground before they could be launched, then that state would enjoy a first-strike advantage. A state that can land a disarming blow may be tempted to do so, and a state that is vulnerable to such a blow may feel a “use it or lose it” imperative to strike first rather than be disarmed.

30. There is a considerable literature on the “offense-defense balance” and its effect on the likelihood of war; see, for example, Keir Lieber, “Grasping the Technological Peace: The Offense-Defense Balance and International Security,” International Security 25 (Summer 2000): 71–104. A related literature emphasizes that what matters is not the actual offense-defense balance but rather the balance that is perceived by military and political leaders; see Van Evera, “Cult of the Offensive.”

preventive war A war fought with the intention of preventing an adversary from becoming stronger in the future. Preventive wars arise because a state whose power is increasing cannot commit not to exploit that power in future bargaining interactions.

first-strike advantage The situation that arises when military technology, military strategies, and/or geography give a significant advantage to whichever state attacks first in a war.

123Can an Adversar y Be Trusted to Honor a Deal? War from Commitment Problems

First-strike advantages can create a potentially insurmountable commitment problem: unless each state can credibly promise the other not to attack first, there is a danger that bargaining will break down as each side hurries to get in the first blow. Each side may be confident that if it manages to strike first, it can do better in war than if it accepts the deal currently on offer. Indeed, there may be no deal that both sides prefer over a war that they start.

An example of this situation is depicted in Figure 3.4. Here, we assume that there are two different war outcomes, depending on which state lands the first blow. The expected outcome of a war started by State A is depicted in the top panel, and the expected outcome of a war started by State B is depicted in the lower panel. The first-strike advantage is captured by the assumption that each side expects to do better in a war that it starts than in a war started by the adversary (that is, the war outcome in the top panel is closer to State A’s ideal point than the war outcome

WAR OUTCOME

COSTS TO A

COSTS TO B

VALUE OF WAR TO B

VALUE OF WAR TO A

A B

SET OF BARGAINS PREFERRED TO WAR STARTED BY STATE A

COSTS TO A

COSTS TO B

VALUE OF WAR TO B

VALUE OF WAR TO A

A B

WAR OUTCOME

SET OF BARGAINS PREFERRED TO WAR STARTED BY STATE B

FIGURE 3.4 Bargaining and First-Strike Advantages State A attacks first: State A enjoys a significant advantage in war if it attacks first, and the war outcome will be close to its ideal point.

State B attacks first: Here, the war outcome heavily favors State B. Because the bargaining ranges in the top and lower panels do not overlap, there is no deal that both sides prefer over a war that they initiate.

The duration of interstate wars varies a good deal. Figure A shows the distribution of war durations for 125 interstate wars over the last two centuries. About 10 wars have lasted three years or more, and they account for most of the com- bat deaths from wars in this period. However, most wars are relatively brief.

What determines how long a war will last? A number of factors might explain why some wars are longer than others, including the power of the adversaries and what they are fighting over. But, at a more general level, how long a war lasts is related to the strategic problem that caused the war in the first place. Once the adversaries go to war, the incentive to reach a deal that spares each side the costs of fighting does not disappear. Indeed, relatively few wars are fought until one of the states is completely occupied or unable to mount further operations; most end with some kind of bargain. This observation raises a puz- zle: How does war make it possible for states to reach a settlement when they could not do so beforehand?

In a recent book, Alex Weisiger argues that if bargain- ing fails because of incomplete information about relative capabilities and resolve, then war can bring about a deal by revealing the hidden information.a As states compete on the battlefield, features that were hard to observe before- hand—the power of each state, the effectiveness of mil- itary technologies and strategies, the willingness of the populace to bear the burdens of war—become observable. In this case, war lasts as long as is necessary for uncer- tainties to be resolved.

Alternatively, if the war came about because of a commitment problem, then the war can end only when the source of that problem is removed. A preventive war fought in response to the rising power of an opponent can end when either the declining power is unable to keep fighting or the rising power is crippled to the point that its rise has been forestalled.

To test these propositions, Weisiger collected data on 103 cases of interstate war since 1816 and estimated the effect of different factors on how long each war lasted. He found that one of the strongest predictors of long and severe wars was whether the war was preceded by a large shift in the adversaries’ relative military capabilities, as measured not only by the size of their militaries, but also by their population size and level of economic develop- ment. In particular, the typical duration of a war preceded by a large power shift is three times that of war preceded by little or no such shift. Thus, as expected, the longest wars tend to be those fought between declining and rising states, when preventive motivations loom largest. On the other hand, among wars that were not preceded by a shift in power—and were therefore more likely to be rooted in an information problem—those that experienced rel- atively frequent and intense battles tended to end more quickly. This finding is consistent with the idea that bat- tles serve to resolve uncertainty, permitting the adversar- ies to find a settlement that both prefer over continued bloodshed.

Bargaining and the Duration of War HOW DO W E K NOW ?

FIGURE A The Duration of Interstate Wars, 1816–2014

Number of Wars

Du ra

tio n

(M on

th s)

0 0 6

12 18 24 30 36 42 48 54 60 66 72 78 84 90 96

102 108 114 120

3 6 9 12 15

Source: Meredith Reid Sarkees and Frank Wayman, Resort to War: 1816–2007 (Washington, DC: Congressional Quarterly Press, 2010).

a. Alex Weisiger, Logics of War: Explanations for Limited and Unlimited Conflicts (Ithaca, NY: Cornell University Press, 2013). For another important study along these lines, see Branislav Slantchev, “How Initiators End Their Wars: The Duration of Warfare and the Terms of Peace,” American Political Science Review 48 (September 2004): 813–29.

125Can an Adversar y Be Trusted to Honor a Deal? War from Commitment Problems

in the lower panel, and the opposite is true for State B). As a result, although there is a set of deals that both states prefer over a war started by either State A or State B, no deal is preferable to both of these possible wars. A deal within the lower bargain- ing range would satisfy State B but give State A less than it expects to get by attack- ing first (the area labeled in red in the top panel). Similarly, a deal within the upper bargaining range would satisfy State A but give State B less than it expects to get by striking first (the area labeled in blue in the lower panel).

Under these conditions, neither state will make concessions to the other at the bargaining table; instead, both will likely rush to the exits, each trying to beat the other to the punch. Negotiations in this context may be seen as nothing more than a ploy to delay the other side from mobilizing. A war that arises in this way is a preemptive war.31

The 1967 Six Day War between Israel and four Arab states (Egypt, Syria, Jordan, and Iraq) is a classic case of a war that started this way. In May 1967, Egypt responded to border skirmishes between Israel and Syria by massing troops on the Israeli border and imposing a partial blockade. On June 5, fearing that war was likely, Israel launched a surprise, preemptive attack on the Egyptian air force, destroying 300 aircraft while they were still on the ground. With Egypt’s main offensive threat crippled, the ensuing war lasted six days and left Israel in control of large swaths of formerly Arab-held territory.

This dynamic forms an important component of what we referred to earlier as the security dilemma. When military technology favors offensive action, a state that builds arms, even out of purely defensive motives, inevitably makes other states worry about a possible attack. If threatened states arm in response, the original state will, in turn, feel less secure, feeding a self-reinforcing cycle of fear. In addition to causing arms races and hampering international cooperation, the commitment problem created by this fear fuels a risk of war.

Both preemption and prevention arise from the difficulty states can have in mak- ing credible commitments not to use their military power. The difference between the two concepts revolves around timing. Preemption is a response to an imminent threat when there is an existing first-strike advantage. Prevention is a response to antici- pated changes in the distribution of power that might result in an increased threat sometime in the future. (For a discussion of how such considerations contributed to the outbreak of World War I, see “What Shaped Our World?” on p. 126.)

We have seen, then, that even if states can locate a deal that both prefer over war at the moment, concerns about their willingness to abide by the deal in the future can cause bargaining failure and war. Although these concerns can arise for different reasons, all the paths to war discussed in this section have at their root a common commitment problem: the difficulty of committing not to use one’s power to one’s advantage in the future.

31. It should be emphasized that this example assumes a very large first-strike advantage relative to the costs of war, so it is possible that this kind of situation arises only rarely. For a historical overview of preemptive war, which finds that very few wars start this way, see Dan Reiter, “Exploding the Powder Keg Myth: Preemptive Wars Almost Never Happen,” International Security 20, no. 2 (Autumn 1995): 5–34.

preemptive war A war fought with the anticipation that an attack by the other side is imminent.

126

World War I redrew the maps of Europe, the Middle East, and Africa. It decimated a generation of young people, and the postwar settlement contributed to World War II 20 years later, as well as to more recent conflicts.

For all its far-reaching consequences, the First World War had relatively modest origins. On June 28, 1914, Archduke Franz Ferdinand, heir to the throne of the Austro- Hungarian Empire, was assassinated during a visit to neigh- boring Serbia. How could this event lead to a war that would ultimately involve 32 states fighting on three continents?

Interests Europe in 1914 was rife with conflicting inter- ests. Germany’s growing power brought it into compe- tition with Britain and provoked fear in its neighbors. France hoped to reclaim territory from Germany, which in turn had designs on territory in Russia and overseas. Austria-Hungary and Russia were competing for influence in the Balkan region. The Ottoman Empire was in decline, and outsiders sought to expand at its expense.

Institutions States with compatible interests often form alliances, institutions that help them to cooperate militar- ily (see Chapter 5). By 1914, the major powers were divided into two hostile camps: Germany, Austria-Hungary, and Italy formed the Triple Alliance against Great Britain, France, and Russia in the Triple Entente. This alliance sys- tem meant that any conflict had the potential to expand throughout Europe. It also posed a problem for German military planners, who knew that, in the event of a war, they would face foes to the east and west.

Interactions In this combustible atmosphere, the assas- sination of the archduke by Serb terrorists created a dan- gerous spark. Austria-Hungary had a substantial Serb minority, and agitation by Serb nationalists threatened to tear the multiethnic empire apart. Austria-Hungary sought to compel Serbia to end its support for these mili- tants and threatened war. This threat brought a deterrent response from the Russians. Germany then promised to protect the Austrians. Though, in principle, the conflict could have been defused by a negotiated bargain, preven- tive and preemptive considerations loomed large.

Although German leaders did not relish a war with Russia, they feared Russia’s growing power. In the

decades prior to the war, Russia had made great economic strides, building its heavy industries and railroad network. Germany’s leaders watched these developments with con- cern and came to believe that they had a short “window of opportunity” in which to prevent future Russian domination.

In contemplating war with Russia, however, there was a significant obstacle: France. To deal with the pos- sibility of a two-front war, the Germans came up with an audacious solution. The Schlieffen Plan took advantage of the fact that Russia’s vast size made its military machine very slow. After the Russians started mobilizing, it would take six weeks before their forces could join the fight. The Germans hoped to exploit this delay to invade and quickly defeat France. German troops could then be shifted to the east, in time to meet the Russian advance.

Planners believed that the strategic situation gave an advantage to whoever struck first. By acting quickly, Germany could seize key bridges and tunnels in Belgium, roll onward into Paris, and then turn their attention to Russia. If they waited, the Belgians could fortify or destroy the bridges, causing the German army to bog down in the west, while the Russian “steamroller” bore down from the east. Under these conditions, negotiations threatened an intolerable delay. Preemptive incentives also meant that military mobilizations were not a useful instrument of cri- sis diplomacy. When the Russians mobilized on July 30, they hoped that signaling their resolve would lead the Austrians to reduce their demands. Instead, Russia’s mobilization spurred the adversaries to the battlefield.

By August 3, Europe was at war. When the Russian army mobilized, Germany invaded Belgium and France. Great Britain joined the fray. Other states were tempted by one side or the other to pile on, and the battlefield expanded across and beyond Europe.

Prevention and Preemption in World War I W H AT SH A PE D OUR WOR L D ?

127Is Compromise Always Possible? War from Indivisibilit y

In addition to shedding light on the general problem of war, this discussion generates a number of predictions about the conditions under which war is more or less likely to occur. First, war is more likely to occur when the good in dispute is a source of power to those who possess it. For example, it is harder to strike bargains over strategically important territory than over territory that is valuable for eco- nomic reasons.32 Second, preventive incentives arise when there are relatively rapid and dramatic changes in the military balance between two countries. Hence, war is more likely when such changes are anticipated or under way. Finally, bargaining failures are more common when the military-strategic situation creates substantial advantages for striking first. These advantages generally arise from the nature of military technology, which sometimes imparts large advantages to the actor who goes on the offensive first—although, as we saw in the case of World War I, particular military strategies can also create pressure to act preemptively.

Is Compromise Always Possible? War from Indivisibility Finally, we consider a third kind of problem that can prevent states from reaching mutually beneficial settlements of their disputes because the disputed good can- not be divided. A good is divisible if there are ways to split it into smaller shares; an indivisible good cannot be divided without destroying its value. Imagine, for exam- ple, the difference between having 100 pennies and having one dollar bill. Although the amount of money available is the same in both cases, the pennies can be divided up between two people in many different ways, while the dollar bill cannot be split without ruining it. When the good in question is indivisible, compromise solutions are impossible to reach, and the bargaining becomes “all or nothing.”

It is easy to see how indivisible goods could create an insurmountable obstacle in crisis bargaining. Consider a situation in which each state would prefer to fight a war rather than get none of the good. Even though there might be deals that both sides prefer over war—such as, say, a 50-50 split—an inability to divide the good into the necessary shares renders such a deal unattainable. In an all-or-nothing bargain- ing situation, one state must get nothing. And if both states prefer war over getting nothing, then war becomes inevitable.

Although the logic of indivisibility is quite clear, what is less clear is whether or how often indivisibility is actually a problem in international politics. What goods are truly indivisible? A key point is that indivisibility is not a physical property of a good, but rather the way in which that good is valued. This point is made most dra- matically by the biblical story in which King Solomon, confronted by two women claiming to be the mother of the same baby, decides that the only fair solution in

32. Paul K. Huth, Standing Your Ground (Ann Arbor: University of Michigan Press, 1996).

indivisible good A good that cannot be divided without diminishing its value.

128 Chapter 3 Why Are There Wars?

the face of these incompatible claims is to cut the baby in half. This solution seems odd not because it cannot be physically implemented, but because the process of doing so would kill the baby. (Fortunately, in King Solomon’s case, the decision did not have to be carried out, since one woman insisted that she would rather let the other woman have her child than see it killed— whereupon Solomon decided that this woman must be the true mother.) Similarly, in a dispute over a valuable painting, one could slice the painting in two, but doing so would destroy the object’s value.

Hence, when we say that a good cannot be divided, we generally do not mean this literally. Rather, we mean that the good loses much, if not all, of its value when it is divided.33 This can be the case with core values that cannot be compromised or with divisible goods that are closely linked to such core values.

A commonly cited example of an indivisible good in international relations is the city of Jerusalem.34 It is a city that contains some of the holiest sites of Christianity, Islam, and Judaism and has historical, cultural, and reli- gious significance unlike any other piece of territory in the world. As a result, the status of Jerusalem is a major stumbling block in efforts to bring about peace in the Middle East.

For many Jews, Jerusalem is the focal point of the desire—sometimes symbolic, sometimes literal—to return to the promised land of Zion. Every year at Passover, Jews all over the world utter the words “Next year in Jerusalem,” underscoring the role of that city in forging a connection between the land of Israel and the Jewish identity. The centrality of this city has been given political expression as well: a Basic Law passed by Israel in 1980 declares that “Jerusalem, complete and united, is the capital of Israel.”

For Muslims, East Jerusalem, or Al-Quds, is the third-holiest city (after Mecca and Medina in Saudi Arabia) and the site of the Dome of the Rock, where the prophet Mohammed is believed to have ascended to Paradise.

Palestinians claim East Jerusalem, which Israel annexed after the 1967 Six Day War, as the capital of an eventual Palestinian state. This claim clearly clashes with the Israeli position that the city is indivisible and the capital of the Jewish state. Moreover, the holiest sites for both faiths sit literally on top of one another. The Dome of the Rock and Al-Aqsa Mosque were built (1,300 years ago) directly on the Temple Mount—the site of the original Jewish temple—whose Western Wall is an

33. H. Peyton Young, “Dividing the Indivisible,” American Behavioral Scientist 38, no. 6 (1995): 904–20.

34. See, for example, Cecilia Albin, “Negotiating Indivisible Goods: The Case of Jerusalem,” Jerusalem Journal of International Relations 13, no. 1 (1991): 45–76.

In Jerusalem, the Dome of the Rock, one of Islam’s holiest shrines, sits on top of the Temple Mount, the site of the Jewish temple until 70 C.E. The concentration of holy sites in Jerusalem complicates negotiations between Israelis and Palestinians over whether and how to divide or share the city, which both sides claim as their capital.

129Is Compromise Always Possible? War from Indivisibilit y

important place of worship for Jews. The question of whether or how to divide the city, and how to ensure that people of all faiths have access to their holy sites, is one that has so far defied resolution.

It is important, however, not to place undue emphasis on indivisibility as a source of bargaining failures. First, some of the difficulty of dealing with indivisi- ble goods arises not from indivisibility per se, but from weak enforcement mecha- nisms. King Solomon’s quandary is one that divorcing parents routinely face when each wants custody of their children. These disputes are settled not by dividing the children physically, but by dividing the time they spend with each parent. This kind of alternating possession would generally fail in international politics, how- ever, because a state would not trust its adversary to hand over the good when the time came to do so. In child-custody cases, the deal is enforced by courts and police, which are generally lacking in the international context. Thus, the main obstacle to striking such a deal lies not in the nature of the good, but in the difficulty of making a credible commitment to share that good over time.35

A second reason to be skeptical of claims of indivisibility is that states may have strategic incentives to claim that they cannot compromise on a particu- lar issue, even if they actually could. Recall that one of the strategies that states employ in crisis bargaining is to tie their hands through public pronouncements from which it would be costly to back down. It is quite possible that claims of indivisibility—such as the Israeli Basic Law pronouncing Jerusalem the capital of Israel—have a strategic quality: they represent an effort to tie the government’s hands so that it will find compromise difficult, if not impossible. The hope in making such claims is that the other side will realize that it has no choice but to give in entirely. In this sense, objects may take on the appearance of indivis- ibility in the course of the bargaining process through the public positions that states take.36

The point of this discussion is not to suggest that indivisible goods do not exist in international politics. Rather, it is to suggest that we be appropriately skeptical when the participants in a dispute claim that the good in question is indivisible and hence no compromise is possible. Such a claim may reflect a bargaining position adopted for strategic reasons rather than a true description of the fundamental nature of the good in question.

In any event, there may be ways to allocate apparently indivisible goods that do not involve physical division. One possible mechanism is shared control. It has been proposed, for example, that Israelis and Palestinians jointly control certain portions of Jerusalem to ensure that all people have access to the sacred sites. Indeed, this already happens to some degree at the Temple Mount: though the site falls under Israeli sovereignty, the area on top, including the Dome of the Rock and the Al-Aqsa Mosque, is administered by Islamic authorities.

35. See Powell, “War as a Commitment Problem,” for a discussion of how indivisibility leads to commitment problems.

36. See Stacie E. Goddard, “Uncommon Ground: Indivisible Territory and the Politics of Legitimacy,” International Organization 60 (2006): 35–68.

130 Chapter 3 Why Are There Wars?

A second mechanism for dealing with indivisible goods is compensation on another issue. Although a rare painting cannot be physically divided, a dispute over it can be resolved by having one party compensate the other in exchange for the good. In this case, the object is made divisible by adding a new dimension to the deal: rather than arguing about who gets the painting, the issue becomes how much money one is willing to pay the other to get it. Since money is generally divisible, adding this new dimension creates the possibility for compromise where none previously existed. Hence, disputants may be able to find a second issue dimension on which the loser in the main issue can be compensated. The strategy of making one dispute easier to solve by bringing in a second issue is known as linkage, as we saw in Chapter 2.

Has War Become Obsolete? War arises from a variety of factors that can prevent states from reaching peace- ful settlements of their disputes. Of course, the reason we study war is not just to understand why it happens, but also to identify factors that might reduce or elim- inate its occurrence. On this front, there is some room for optimism. One striking observation from Figure 3.1 is that the incidence of interstate war seems to have dropped since 1950. Moreover, there has not been a war between two great powers since the Korean War ended in 1953—the longest such streak in modern history.37

The apparent decline in interstate war over the last six decades has been the sub- ject of considerable attention among political scientists.38 However, we must approach this trend carefully. As Figure 3.1 shows, earlier periods in history experienced dips in the frequency of war that turned out to be temporary. And as this book was going to press in 2018, numerous crises involving the United States, Russia, China, North Korea, and Iran were threatening to escalate. Depending on what happens in the world, a section on the obsolescence of war might not be appropriate for the next edi- tion of this book, a few years from now. Still, the observation that the past six decades have seen a decline in interstate war is striking and important enough to motivate the following questions: Have there been changes in world politics that could explain a decline in the incidence of interstate war since World War II? To what extent can the framework offered in this chapter guide our thinking about this development?

The logic developed in this chapter has three main components: (1) war becomes possible when states have conflicting interests over a contentious issue, (2) the costs of war ensure that there exists a peaceful deal that both states prefer

37. Scholars typically identify the following states as great powers in the post–World War II period: the United States, the Soviet Union/Russia, Great Britain, France, China (after 1950), Germany (after 1991), and Japan (after 1991).

38. See, for example, John E. Mueller, Retreat from Doomsday: The Obsolescence of Major War (New York: Basic Books, 1989); Joshua S. Goldstein, Winning the War on War: The Decline of Armed Conflict Worldwide (New York: Dutton, 2011); Steven Pinker, The Better Angels of Our Nature: Why Violence Has Declined (New York: Viking, 2011). For a skeptical view, see Tanisha M. Fazal, “Dead Wrong? Battle Deaths, Military Medicine, and Exaggerated Reports of War’s Demise,” International Security 39 (Summer 2014): 95–125; and Bear F. Braumoeller, “Systemic Trends in War and Peace.” In The Causes of Peace: What We Now Know–Nobel Symposium 161, edited by Olav Njølstad. (Oslo, Norway: The Norwegian Nobel Institute, 2017.)

131Has War Become Obsolete?

over fighting, but (3) bargaining may fail because of information or commitment problems or because of an inability to divide the good at stake. It follows that a decline in war could arise from (1) changing interests that lead to a decrease in the value of goods that have historically driven conflict, (2) an increase in the costs of war that leads to changing interactions, or (3) the growth of institutions that help states solve the information or commitment problems associated with uncertainty and changing power, or the problem of the indivisibility of a good. There is reason to think that all three of these factors have changed in ways that have made war less likely overall.

Changing Interests: Declining Conflict over Territory A striking feature of world politics since 1945 is the declining role of territory in driving interstate conflicts. Not only has there been a drop in the frequency of interstate wars over territory, but there have also been fewer instances of success- ful conquest or annexation of territory since World War II.39 Explanations for this observation abound, but any understanding has to start with changes in the value of possessing territory in the modern era. Whereas control of territory and popu- lation was historically an important source of state power, technological changes have weakened this link. In an age of nuclear weapons, precision-guided missiles, and increasing reliance on unmanned aerial vehicles (popularly known as drones), military power no longer depends on how large an army a state can field. Moreover, the growth of international trade and investment (see Chapters 7 and 8) means that it is generally easier to acquire resources through markets than through conquest.40 At the same time, the spread of nationalism has made it harder for foreign states to control conquered populations, thereby raising the costs of taking territory. Thus, some of the interests that drove states to contest territory have become less potent.

The interests of key states also changed after World War II. Whereas previ- ously the rival territorial ambitions of the European great powers had cast that continent into frequent warfare, leadership of the international system after 1945 passed to two states—the United States and the Soviet Union—with little interest in further territorial expansion. The United States had satisfied all of its ambitions by the early twentieth century, with the end of westward expansion against the Native American tribes and the final settlement of the Canadian boundary in 1902. The Soviet Union used its military success in World War II to restore its dominance over lands that it had historically controlled. Thus, the most powerful states in the system had an interest in stabilizing the territorial order rather than changing it. The United States, in particular, played a role in advancing and defending a norm of “territorial integrity”—the idea that borders should not be changed through

39. See, for example, Mark W. Zacher, “The Territorial Integrity Norm: International Boundaries and the Use of Force,” International Organization 55, no. 2 (2001): 215–50; Tanisha M. Fazal, State Death: The Politics and Geography of Conquest, Occupation, and Annexation (Princeton, NJ: Princeton University Press, 2007); and Gary Goertz, Paul F. Diehl, and Alexandru Balas, The Puzzle of Peace: The Evolution of Peace in the International System (New York: Oxford University Press, 2016).

40. Stephen G. Brooks, Producing Security: Multinational Corporations, Globalization, and the Changing Calculus of Conflict (Princeton, NJ: Princeton University Press, 2007).

132 Chapter 3 Why Are There Wars?

force—a norm that became enshrined in international institutions like the United Nations (for more on the role of norms, see Chapter 11).41

Many newly decolonized countries also embraced this norm, even though their borders often made little sense, having been drawn in European capitals by imperial powers with scant knowledge of local conditions. Leaders in these new states decided that it was generally in their interests to respect the inherited borders and consolidate power within them, rather than open up a Pandora’s box by calling all of the colonial borders into question.42 In sum, changes in the international system after 1945 reduced states’ interest in fighting over territory— reducing the value of territory as a good—which has historically been a major cause of interstate war.

Changing Interactions: The Rising Costs of War One of the main disincentives for engaging in war is the human, economic, material, and psychological costs it imposes. Indeed, the costs of war generally ensure that states can do better by finding a negotiated settlement of their disputes. As war becomes less attractive, states are more willing to make compromises in order to avoid it. In addition, increasing the costs of war increases the range of status quo distributions that neither state has an incentive to challenge (see Figure 3.2).

At least two major developments since 1945 have increased the expected costs of war. One is the advent of nuclear weapons, which gave states the abil- ity to completely obliterate each other. As a result, the expected costs of a war that involve nuclear exchange are so large that they swamp the value of what- ever good is at stake. The threat of “mutually assured destruction” in the event of war engenders caution among nuclear-armed states, reducing the attractive- ness of taking risks in a crisis. Indeed, it is striking that despite five decades of intense hostility between the United States and the Soviet Union during the Cold War, the two superpowers never waged war directly against one another. Although there are many theories about the cause of this so-called long peace, both states’ possession of nuclear weapons certainly played a stabilizing role.43 We will return to this topic in Chapter 14, when we consider the implications of the spread of nuclear weapons.

The second major development since 1945 is the explosive growth of interna- tional trade and financial transactions (see Figure 14.2 on p. 589). Scholars have long thought that, as countries became more economically interdependent, the costs of war between them would grow. The more two countries value one another as trad- ing partners, the more incentive they have to avoid conflicts that could disrupt those

41. Zacher, “Territorial Integrity Norm”; Fazal, State Death, 47–52.

42. Jeffrey Ira Herbst, States and Power in Africa: Comparative Lessons in Authority and Control (Princeton, NJ: Princeton University Press, 2014).

43. See, for example, John Lewis Gaddis, “The Long Peace: Elements of Stability in the Postwar International System,” International Security 10 (Spring 1986): 99–142.

133Has War Become Obsolete?

profitable exchanges.44 Whether this theory is right has been a subject of vigorous academic debate.45

Part of the challenge is that while it is plausi- ble that increased trade makes countries less likely to fight one another, the reverse could also be true: peaceful relations create the conditions for greater interdependence through trade. So, although coun- tries that trade more do fight less, figuring out how much of that relationship is due to trade causing peace—rather than peace causing trade—is difficult. On balance, though, most scholars consider grow- ing commercial and financial interdependence since 1945 as a contributor to peaceful relations, particu- larly among the advanced industrial countries.46

Changing Institutions: Democracy and International Organizations Finally, the decline in interstate conflict may be due to the growth of institutions that have helped states overcome information and commitment problems that can cause bargaining to fail. Two kinds of institutions could play a role in this regard: domestic and international.

On the domestic front, the major trend since World War II has been the expan- sion in the number of countries that have democratic political systems (see Figure 4.6 on p. 170). This development is striking when paired with the observation that there are few, if any, clear cases of war between democratic states. This means that the spread of democratic institutions might account for some of the overall decline in interstate war in the past six decades. Why democracy might have this effect is the subject of a larger discussion in the next chapter, but prominent explanations center on three factors that are relevant to the discussion here. First, democracy may make the leaders of states more sensitive to the costs of war. Second, democracy may make states more transparent to outsiders, thereby reducing uncertainty about their capa- bilities and resolve. Third, democratic states may bargain with one another differ- ently than they do with other states, avoiding threats and emphasizing compromise.

The post-1945 world has also seen a dramatic expansion in the number and activity of international organizations (see Figure 14.1 on p. 587). This period wit- nessed the emergence and growing assertiveness of the United Nations, as well as the

44. For a good discussion of some of the intellectual origins of this view, see Michael W. Doyle, Ways of War and Peace: Realism, Liberalism, and Socialism (New York: Norton, 1997), chap. 7.

45. For a recent review, see Erik Gartzke and Jiakun Jack Zhang, “Trade and War,” in The Oxford Handbook of the Political Economy of International Trade, ed. Lisa L. Martin (Oxford: Oxford University Press, 2015), www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199981755.001.0001/oxfordhb-9780199981755-e-27.

46. See, for example, Håvard Hegre, John R. Oneal, and Bruce Russett, “Trade Does Promote Peace: New Simultaneous Estimates of the Reciprocal Effects of Trade and Conflict,” Journal of Peace Research 47 (November 2010): 763–74.

Although recent decades have been relatively free of interstate war, disputes among nuclear- armed states, such as China, persist. The high expected costs of nuclear war have contributed to peace among these states, since the threat of “mutually assured destruction” makes them cautious to escalate conflicts.

134 Chapter 3 Why Are There Wars?

development of numerous regional organizations devoted to promoting peace and security.47 A fuller examination of the role and track record of these organizations will come in Chapter 5, but we can anticipate some of the ways these institutions may have improved states’ abilities to solve informational and commitment problems.

First, international organizations may enhance transparency by providing neutral observers of a state’s military activities. The International Atomic Energy Agency (IAEA), for example, conducts inspections of members’ nuclear energy programs in order to determine whether nuclear fuel has been diverted to produce weapons. Similarly, the United Nations often provides monitors to help observe whether parties to a conflict are fulfilling pledges to withdraw troops or disarm, thereby reducing uncertainty among adversaries about what the other side is up to.48

International organizations may also make it easier to resolve commitment problems. As discussed earlier, a promise not to use one’s power to one’s advantage is a very difficult commitment to make in a credible fashion, largely because states cannot generally rely on repeated interaction to hold one another to such a prom- ise. Repeated interaction can help make promises credible if the prospect of future dealings leads a state to forgo the temptation to break a promise today because it fears retaliation by the other state tomorrow. Under conditions of shifting power, however, the state that is becoming more powerful has less to fear from the other side’s retaliation. Furthermore, if a state can use its increased power to destroy the other state, then there may be no “shadow of the future” to stay the growing state’s hand. The same holds in situations involving first-strike advantages: the temptation to deliver a crippling blow is hard to blunt if the other side’s retaliatory capability will be rendered ineffective in the process.

Because adversaries generally cannot solve this commitment problem on their own, any solutions are most likely to come from third parties—that is, a state or group of states, including international organizations, that are not directly involved in the bargaining game between the opponents. In some cases, such outsiders may be able to help build credibility for the opposing states’ commit- ments not to exploit power or first-strike advantages. Third parties can play this role by monitoring and enforcing agreements, by providing security guarantees to one or both sides, and sometimes by interposing their forces directly between two potential combatants—all roles that the United Nations performs, as we will see in Chapter 5.

In sum, changes in interests, interactions, and institutions since 1945 have contributed to conditions that make peace more likely in international disputes. It is too strong to declare war obsolete or a thing of the past. But even incremental reductions in the risk of war are reason for optimism and opportunities for greater understanding of this phenomenon.

47. Goldstein, Winning the War on War.

48. See, for example, Virginia Page Fortna, Peace Time: Cease-Fire Agreements and the Durability of Peace (Princeton, NJ: Princeton University Press, 2004).

135Conclusion: Why War?

Conclusion: Why War? We started this chapter with a puzzle: Why do states fight wars in spite of the enor- mous costs associated with fighting? The answer involves two ingredients: a con- flict in states’ interests, and some factor or factors that prevent states from reaching a peaceful solution to that conflict. States’ underlying interests in power, security, wealth, and/or national identities can give rise to disputes over territories, policies, and the composition of one another’s governments. In the absence of authoritative institutions that resolve interstate disputes, states engage in a bargaining interac- tion over these issues, sometimes invoking the threat of military force to enhance their leverage. War occurs when features of this strategic interaction prevent states from reaching a settlement that both prefer over war, with all its uncertainties and costs.

Three kinds of problems can prevent states from settling their disputes in ways that enable them to avoid the costs of war: incomplete information, difficulty in committing to honor a deal, and goods that are hard to divide. Since these obstacles occur in different degrees and in different combinations from one crisis to the next, there is no single answer to the puzzle at the heart of this chapter. Instead, the dis- cussion has sought to identify and explore a set of mechanisms that can stymie the effort to resolve disputes peacefully. Having done so, we are in a better position to understand and interpret behavior and outcomes in international crises. The con- cepts introduced here highlight some of the key factors that determine whether a dispute can be settled without resorting to war.

In particular, we need to pay attention to (1) what the adversaries in a dispute believe about one another’s willingness and ability to wage war and how uncer- tain those beliefs are; (2) how each side seeks to communicate its resolve, whether those efforts are credible, and to what extent they either entail a danger of acci- dental war or “lock in” incompatible bargaining positions; (3) whether the good in the dispute is a source of future bargaining power; (4) whether the distribution of power between the adversaries is expected to change as a result of different eco- nomic growth rates or technological progress; (5) whether the military technolo- gies and strategies of the adversaries generate sizable first-strike advantages; and (6) whether the good in question is indivisible because of its close connection to core values, such as religious identity.

In spite of the many things that can go wrong and prevent a peaceful solution, it is important to remember this point: Wars are rare and appear to have become rarer over time. Even though information is often incomplete, credible commitments are difficult to make, and core values are hard to compromise, most states most of the time are at peace with one another. The costs of war mean that not every disagree- ment is worth fighting over or even threatening to fight over. Hence, the very costs that make war hell also ensure that war is not the most common way that states settle their disputes.