CHAPTER TWO
The battlefield is the most honest place on Earth.
—Retired U.S. Lieutenant Colonel Bo Gritz
WAR IS FUNDAMENTALLY POLITICAL. It is launched, fought, and ended in pursuit of political goals. This was the fundamental insight of nineteenth century Prussian thinker Carl von Clausewitz. He opposed the more traditional view that war is an exercise in military engineering, a fulfillment of one nineteenth-century German general’s wish that, “The politician should fall silent the moment that mobilization begins.”1
War is about politics, and politics, especially in this context, is essentially about the allocation of scarce goods. Goods are phenomena valued by political actors. Goods are scarce if there is not an optimal or infinite supply of the good, meaning that all actors cannot simultaneously consume or possess an optimal or infinite amount of the good. Territory, natural resources, and the composition of a national government are all examples of phenomena viewed by international actors as scarce goods. Even nonmaterial phenomena like reputation can be viewed as scarce goods, as in a confrontation where a state best proves its toughness by coercing the other side to retreat and concede its weakness.
A useful metaphor for thinking about how states interact with each other and attempt to maintain or expand their possessions of scarce goods is bargaining, the process by which two actors strive to divide a disputed good.2 Buyers and sellers bargain over price, labor and management bargain over wages, and nations bargain over goods like the placement of an international border. Scholars have also used bargaining models to shed light on how wars start, focusing in particular on the role of information and uncertainty. This chapter advances the application of the bargaining metaphor, going beyond using it to understand how wars start, and applying it to also understand how wars end. It presents the first half of this book’s theory.
A WAR INITIATION PUZZLE
Imagine that two states dispute the allocation of a scarce good, such as the placement of a territorial border, and war looms. Assume that war is costly for both sides, since both the winner and loser must pay war costs in blood and treasure. If both sides knew who would win the war, and how the disputed good would be (re)divided at war’s end, wouldn’t each side be better off not fighting and just dividing the good to reflect how the good would have been divided if the war had occurred? Each would get the same amount of the good it would have gotten had the war occurred, and neither has to pay the unavoidable war costs of money wasted and lives lost. This insight is also demonstrated by a “divide the dollar” game, where two players pay a referee five cents each to roll a one hundred-sided die, and the number turned up by the die roll determines how a dollar is divided between the two players (for example, a roll of seventy-five gives one player seventy-five cents, and one player twenty five cents), and the payoff is the die roll minus the nickel fee (seventy-five cents minus five cents equals seventy cents). But, if each player knew ahead of time what number the die would turn up, the two of them could just divide the dollar according to what that die roll would have been, each saving the five cents that the referee would have collected as a fee for rolling the die. So the puzzle is: Why would these two players ever roll the die? Why do states ever go to war?
Importantly, the assumption that wars are always on balance costly for each side is not uncontroversial. Some feminist approaches contend that states may fight for the sake of fighting, as wars serve patriarchy by reinforcing gender identity.3 A more mainstream critique is that leaders go to war for domestic political reasons, such that a war-avoiding bargain might not be reachable even when both sides knew who would win, as fighting itself provides domestic political benefits from a war to both winner and loser.4 Under some conditions, especially if a state is undergoing democratization or if a national leader is experiencing domestic political problems such as unrest or economic downturn, a state may see war as a way to rally the public around the leader and stave off domestic political challenges.5 The proposition that leaders go to war when facing domestic difficulties is often called the “diversionary” hypothesis.
However, the evidence that leaders choose war to solve internal political problems is thin. The underlying assumption is that going to war engenders a rally round the flag effect that boosts the popularity of leaders, but leaders reap this benefit only under very narrow conditions (which often cannot be controlled by the attacking state), and even the biggest rallies are short-lived.6 Importantly, there is almost no smoking gun historical evidence of a leader launching a war primarily as a means of solving domestic political problems. At most, politicians have occasionally speculated about diversionary action, such as Secretary of State William Seward’s (ignored) April 1861 suggestion to President Abraham Lincoln that the United States provoke crises with European powers as a means of staving off civil war between the Union and the seceding southern states.7 A Russian minister is famously thought to have declared just after the outbreak of the 1904–05 Russo-Japanese War that, “We need a little, victorious war to stem the tide of revolution,” but the story is likely too good to be true.8 Leaders sometimes see indirect relationships between starting war and reaping domestic political benefits, such as the possibility that Lyndon Johnson escalated the Vietnam War in 1965 to protect his Great Society program from domestic political attack.9 Some quantitative studies have found that the presence of internal problems like declining economic growth, rising inflation, partial democratization, or declining leader popularity are correlated with an (often slightly) increased likelihood in the use of force. However, these relationships are often limited in scope, occurring only under certain economic or political conditions.10 Any possible diversionary effects might in turn be moderated by the tendency of states to avoid provoking other states that might have diversionary incentives.11
AN INFORMATION-BASED SOLUTION TO THE WAR INITIATION PUZZLE
One long-standing solution to the war initiation puzzle of “Why do states ever fight?” is to relax the assumption that the two states agree on what the outcome of the war would be, and allow for the two sides to disagree. For example, if each side thinks it will win, that constitutes disagreement about how the war will end. If each side thinks it will win, then prior to war each will demand a victor’s share of the disputed good as the price for avoiding war. However, the scarcity of the good makes such an arrangement impossible, as one cannot give simultaneously to each side the majority of a finite good. At least one state will prefer to take its chances with war rather than take an unacceptably low share of the good, encouraged by confidence in its own war-making abilities and faced with an adversary who is unwilling to make a sufficient concession because it in turn also thinks it will win.
This argument received formal treatment in a famous 1995 article by James Fearon.12 He developed a formal model of two states bargaining over a disputed good, culminating with this same puzzle that when two states agree on how a war will unfold, they should be able to avoid war and (re)divide the good to reflect how the good would be divided if the war were fought. Fearon developed three answers to the puzzle of why war happens. First, war may occur if the two sides do not agree on how the war will unfold on the battlefield.13 Second, war may occur if the good under dispute does not lend itself to division. This argument is discussed at the end of chapter 3. Third, war may occur if at least one side fears the willingness of the other side to comply with a war-ending bargain. This argument is closely related to the commitment propositions developed in chapter 3.
The focus in this chapter is on the first solution to the puzzle, disagreement between states about how the war will unfold. More specifically, there may be war-causing disagreement between states over five different phenomena, each of which affects war outcomes. The first is aggregate military power, which is a combination of size of standing army, the industrial capacity to produce munitions, the size of the population that can be mobilized into military service, and the political capacity to mobilize. Aggregate military power is one of the single biggest factors determining war outcomes, and if states disagree about the comparative ability of each side to field and arm troops, they may not agree on who will ultimately win, and will not be able to reach a war-avoiding agreement.14
More concretely, many wars have begun because of disagreement over aggregate military power. The core flaw behind Germany’s ultimately doomed decision to invade the Soviet Union in June 1941 was the overestimation of its military strength in relation to Soviet military power.15 Commenting in August 1941 on General Heinz Guderian’s 1937 estimates of Soviet tank strength, Hitler told Guderian, “If I had known that the figures for Russian tank strength which you gave in your book were in fact the true ones, I would not—I believe—ever have started this war.”16 Conversely, when the asymmetry in power is unavoidably apparent, both sides may agree on the likely outcome of conflict, and avoid war. For instance, when the USSR made severe demands on Estonia in late September 1939 following the German invasion of Poland, the Estonians considered putting up a fight, but decided instead to make concessions, recognizing that their military of 15,000 would be no match for even the limited detachment of 160,000 Soviet soldiers deployed to the Soviet–Estonian border earlier that month.17 Interestingly, however, there are some examples of leaders of clearly overmatched states nonetheless deciding to fight (examples include Belgian resistance to Germany in 1914 and 1940), often motivated by a variety of reasons including honor, domestic politics, and international reputation.18
The second area of potential war-causing disagreement is military technology, defined as machines or tools that contribute to aggregate firepower, accuracy of firepower, mobility, logistics, force protection, command and control, and/or intelligence gathering. There may be disagreement either about what technology each side has available to it, or how untested technology will play out during war. For example, in 1990–91 Iraq’s military strategy rested on the assumption that it would be able to impose significant casualties on American and coalition forces, and that American and other democratic publics would be unwilling to absorb casualties to liberate Kuwait. However, Iraq underestimated the ability of American technology to protect coalition forces, specifically American air superiority and the longer firing range of American tanks, the latter of which permitted American tanks to destroy Iraqi tanks before entering the firing range of Iraqi tanks.19 Serbia made a similar miscalculation in 1999, resting its strategy in the conflict with NATO over Kosovo on the hope it would be able to shoot down NATO aircraft. However, NATO forces, such as the American B-2 bomber, were able to execute bombing sorties from altitudes high enough to be out of range of some Serb air defense assets, thereby providing perfect force protection (no NATO personnel were lost during the war), and undermining Serb strategy.20 A last example is the 1973 Yom Kippur War. Egypt secretly acquired advanced surface-to-air missiles (SAMs) from the Soviet Union after the debacle of the 1967 Six-Day War, which enabled Egyptian forces (at least initially) to improve the balance of power by neutralizing Israeli air assets.21 Their confidence in their secret weapon helped encourage Egypt to launch the attack.
Third, there may be disagreement about the comparative interactions and effectiveness of the two sides’ military strategies.22 As with military technology, strategy—the plans for the employment of military forces during war—may be unknown and/or untested. In 1940, Germany and France disagreed about what military strategy Germany would employ in the event of a German invasion. France expected Germany to repeat its World War I strategy and move through Belgium en route to Paris. Germany instead planned a feint at Belgium to draw Allied forces forward, making their real attack through the Ardennes forest. Because France misunderstood Germany’s strategy, France substantially overestimated its chances of victory, the “Phony War” of autumn 1939–spring 1940 blew up into a real war in Western Europe in May, and Germany conquered France, Belgium, Luxembourg, and the Netherlands with shocking speed.23
Secret strategy also played a role in the outbreak of the 1967 Six-Day War. Egypt maintained a substantial air force to bolster its ground power. Israel devised a strategy to neutralize the entire Egyptian air force before it could be used in battle. Israeli defense planners knew that essentially the entire Egyptian air force was on the ground at a particular time every morning, making it vulnerable to Israeli airstrikes. At the outset of the Six-Day War, the Israeli aircraft struck Egyptian aircraft on the ground at just this moment in the morning, destroying the bulk of the Egyptian air force and swinging the balance of power decisively in Israel’s favor.24
Fourth, there may be disagreement about resolve, which is how a state values the stakes of the war in relation to the costs it must pay to win. Levels of resolve are an integral part of war outcomes, as the decision to continue fighting is a function of the costs of continued fighting as well as what benefits a side expects to get by continued fighting. Even if both belligerents knew how many casualties each would suffer if the war continued, they might disagree on the relative willingness of the two sides to suffer casualties. More generally, if both sides agreed on who would back down first and when, they could avoid war and settle on the outcome that would have happened had they fought.25 If the two sides disagree about the balance of resolve and who would back down first and when, then war may ensue or endure if already begun. Iraq invaded Kuwait in 1990 falsely confident that the United States did not have the will to go to war to liberate Kuwait. Saddam Hussein blustered to an American diplomat before the invasion of Kuwait that America “is a society which cannot accept ten thousand dead in one battle.”26 Conversely, America underestimated North Vietnamese resolve in the 1960s. The United States assumed that North Vietnam had a “breaking point,” and that all it needed to do was wield its considerable military superiority sufficiently to impose (or threaten to impose) enough losses on North Vietnam to pass that breaking point, and North Vietnam would make concessions to end the war. Secretary of State Henry Kissinger often remarked that, “North Vietnam could not be the only country in the world without a breaking point.”27 This strategy was half successful, in that the U.S. did impose staggering losses on North Vietnam during the war, in the range of 500,000 to 600,000 deaths, amounting to almost as much as 3 percent of its prewar population. And yet, North Vietnam never reached its breaking point in terms of civilian deaths. It made concessions only after American airpower had inflicted significant damage on North Vietnamese military assets in 1972.28 As former Secretary of State Dean Rusk commented in 1971, “I personally underestimated the resistance and the determination of the North Vietnamese. They’ve taken over 700,000 killed which in relation to population is almost the equivalent of—what? Ten million Americans? And they continue to come.”29
Fifth, there may be disagreement about the likelihood and/or impact of third-party intervention. This point can be related to the first four points since disagreement over third-party intervention means disagreement over the capabilities, technology, strategy, and/or resolve of the third party. An otherwise overmatched side may prefer war to concession if it thinks a third party may rescue it.30 The Southern states of the U.S. were very confident up to 1861 that European dependence on the import of Southern cotton would force European intervention on the Southern side in the event of war. This “King Cotton” theory encouraged the Southern provocation and obstinacy that culminated in the outbreak of the American Civil War.31
Belief that a third party will not intervene to rescue its ally may inspire an aggressor to attack. Hitler ordered the September 1939 German attack on Poland confident that neither Britain nor France would uphold their treaty obligations to defend Poland.32 Third-party intervention was also a critical factor in the escalation of the Korean War. In early September 1950, President Truman committed the United States to conquering North Korea as well as liberating the entirety of South Korea, “provided that at the time of such operations there has been no entry into [that area of]...major Soviet or Communist Chinese forces, no announcement of intended entry, nor a threat to counter our operations [there] militarily.”33 This assumption of Chinese neutrality proved, of course, to be wrong, and Chinese intervention that autumn turned the course of the war dramatically against South Korea and the United States.34
Disagreement between states about these five factors is endemic to the international system, and is created and maintained by a variety of conditions. There is, of course, no central clearinghouse of information on warmaking capacity or resolve. Making matters worse, states have incentives to misrepresent their capabilities and resolve to each other. Many states maintain their military and even industrial capacities as state secrets, forcing their adversaries to make rough estimates of actual fighting ability. Moreover, states are likely to be especially secretive about military strategy and technology since the power offered by military strategy and technology can be substantially reduced if it loses its surprise value. The German strategy of attacking through the Ardennes in 1940 worked only because France guessed wrong and placed the bulk of its military forces elsewhere. The Israeli strategy of launching a preventive airstrike against the Egyptian air force at the outset of the 1967 Six-Day War worked only because Egypt did not know the timing of the attack, and hence most of its aircraft were caught on the ground. Egyptian SAMs in 1973 made an impact initially because Israel was unprepared. Indeed, war becomes especially likely when states mobilize their forces secretly, hoping to capture a surprise attack advantage.35
INFORMATION AND WAR TERMINATION
The focus in this book is on war termination rather than war initiation. Some existing bargaining model scholarship often has a very simple, essentially mechanical, view of war as a “costly lottery.”36 That is, once war starts, its outcome is determined by some fixed probability, such as one side having a 70 percent chance of winning, and the other side having a 30 percent chance of winning.
A more nuanced approach rejects the assumption that war is an apolitical mechanical process, instead allowing for political actors to change their beliefs during wartime and bargain towards reaching a war-ending deal. This approach sees fighting as having a very specific political function: to reduce the war-causing disagreement between the two sides.37 The point was perhaps first made in 1904 by the prominent German sociologist Georg Simmel, who wrote that “the most effective presupposition for preventing struggle, the exact knowledge of the comparative strength of the two parties, is very often only to be attained by the actual fighting out of the conflict.” This idea was occasionally restated in the literature on war termination from the 1950s through the 1970s, perhaps most famously by Geoffrey Blainey in his classic 1973 book on war.38
In recent years, scholars have applied formal analysis towards understanding the relationship between combat, information, and war. There is an array of different models of bargaining during war, but they often describe a fundamentally similar dynamic, that fighting battles provides information to the belligerents, which in turn affects war-termination decisions.39 This current of scholarship describes two states in disagreement over the distribution of a particular good, such as the placement of a border. There is a fixed balance of power between the two states, and more narrowly a fixed probability that one side will win any given battle between them.40 At least one side does not know the true balance of power, and instead can only speculate about the true balance.41 The offers to divide the good reflect each side’s assessment of the balance of power, such that a side demands more of the good as it becomes more confident in its own strength. War can be avoided when the two sides agree on the balance of power, as the side seeing itself as stronger demands more, and the side seeing itself as weaker demands less. However, if each side sees itself as stronger, then the two sides will make ambitious, ultimately irreconcilable offers (such as each demanding seventy-five cents of a single dollar, or both France and Germany each demanding exclusive control of Alsace and Lorraine). In the face of irreconcilable offers, one side may choose to attack as a means of proving its military power, pushing the other side to change its understanding of the balance of power, and demand less of the good. If one side attacks, war erupts, and battles occur. The belligerents observe battle outcomes, taking note of things like territory gained or lost, and casualties suffered by each side. On the basis of the observed battle outcomes, belligerents update their beliefs about the likelihood of winning future battles—in other words, the true balance of power.
Critically, each side’s war-termination offer is inextricably tied to its views of the balance of power: more confident belligerents demand more as a condition of ending the war, and less confident belligerents demand less. After a state wins a battle and updates its estimate of the balance of power, it changes its war-termination offer, that is, it demands more of the disputed good as a condition of ending the war. Therefore, following a combat success, a belligerent will raise its war-termination offer (demand more); however, following a combat defeat, a belligerent will lower its war-termination offer (demand less). I refer to this as the “information proposition.” Note that it bears directly on Clausewitz’s ideas about the relationship between “real” or limited war and absolute war. Clausewitz proposed that a war is “real” or limited if it ends short of the utter defeat of one side’s military, and is absolute if it ends with the complete defeat of one side’s military. Clausewitz proposed that wars often end in limited outcomes because, by fighting, belligerents learn about who would win if the war was fought to the finish. Better informed, the belligerents are able to end the war short of an absolute outcome and curb the costs of fighting.42
The preceding proposition begs the question, what is combat success? As discussed in greater detail in chapter 4, combat success is a highly contextual concept, as sometimes belligerents want to gain territory irrespective of casualties, sometimes they want to inflict casualties irrespective of territory, and so on. Success needs to be defined on a case by case basis, in the context of the political-military strategy of the belligerent, rather than using a one-size-fits-all rule like territorial loss or gain.
The information proposition emerges from a more purely mathematical vision of this updating process, such that even if a belligerent has a strong belief in its relative power, a battlefield victory will encourage the belligerent to increase incrementally its estimate of the balance of power, and then increase incrementally its demands of the other side. In practice, the process may not be so precise. A belligerent may expect to win battles, and then when it does win those battles it will see this as evidence supporting its general beliefs about the balance of power, thereby encouraging the belligerent to maintain (rather than increase) its (high) war-termination demands. This possibility produces a slight modification of the information proposition, that changes in war-termination demands are most likely when battle outcomes are surprising.
Besides the capability of the armed forces of the primary belligerents, another important factor affecting the balance of power is the potential contribution of third-party interveners. If another country decides to intervene in the war, this will change each side’s estimate of the balance of power and its eventual likelihood of winning. Therefore, another form of the more general information proposition is that if a third party intervenes on behalf of a belligerent, that belligerent becomes less likely to lower its war aims and more likely to raise its war aims. If a third party intervenes on behalf of a belligerent’s opponent, that belligerent becomes more likely to lower its war aims and less likely to raise its war aims.
The focus so far has been on how factors more or less outside the control of the belligerents, such as battle outcomes and the decisions of third party interveners, affect the war-termination process.43 However, wartermination offers may be both a cause and effect in this process, since if a belligerent’s adversary makes a new war-termination offer, the offer itself sends information about the true balance of power or balance of resolve, which in turn may cause the belligerent to change its war-termination offer.44 Specifically, if a belligerent’s adversary sends a new offer that demands fewer concessions, this may be seen by the belligerent as a sign of weakness, and may in turn cause the belligerent to stiffen its own negotiating position. Conversely, demanding more concessions can send a signal of resolve and power. Further, even if a bargainer makes no changes in its settlement offer, the very act of deciding to continue the costly conflict without making concessions sends a credible signal of resolve.45
INFORMATION AND CHANGING EXPECTATIONS
The discussion thus far has assumed a certain evenness in war. Belligerents’ beliefs about the likely costs and outcome change across the war, but the assumption is that the likely casualties and outcome probability of all battles is the same for all battles across the war. As states fight, they observe battle outcomes, and eventually both come to agree sufficiently on the true and unchanging balance of power and end the war.
This assumption that casualties and outcomes are determined by a constant probability across the span of a war is not trivial. Consider instead the possibility that belligerents recognize that in the future the costs of fighting will change, and/or the likelihood of winning any battle will change. This could occur for several reasons. Future combat may take place on terrain more favorable to the adversary. A third party may be on the verge of intervening on behalf of the defender. Conversely, one’s own ally may be close to exiting the war. Third parties may threaten to impose economic sanctions or other diplomatic costs if the war continues. One’s own military power may soon collapse if one’s standing stocks of munitions approach exhaustion.
Figure 2.1 displays these two separate assumptions, that the per battle costs of war (or, the likelihood of victory in any given battle) remain constant throughout the war, and that the per battle costs of war promise to escalate midway through the war. The x axis reflects time, starting at the beginning of the war. The y axis reflects expected costs per unit of time (that is, not cumulative costs). The solid line reflects the assumption that per-battle costs are constant for all battles throughout the war. The dotted line shows a different assumption, that at time A, the per-battle costs of fighting jump up significantly, and remain at a high level for the rest of the war.
Copyrighted image removed by Publisher
Relaxing the assumption of constant probabilities of combat outcomes affects our forecasts for war-termination behavior. If a belligerent is doing well in combat, then the information propositions developed in this chapter would forecast that that belligerent should prefer to keep fighting and would not be likely to lower its war-termination demands, and might even raise them. However, a belligerent may believe that despite past and present successes, future combat might bear significantly higher costs, and/or the likelihood of success in future combat might be significantly lower. In such cases, even a winning belligerent may move to end the war before costs escalate and/or achieving combat success becomes much more difficult, accepting a limited outcome which might seem excessively generous given its past combat successes. Therefore, a belligerent is more likely to make concessions if it believes that the costs of war promise to escalate significantly, and/or if the likelihood of combat success promises to drop significantly. Chapter 3 will return to the topic of interactive effects of changing expectations about costs and victory.
WAR TERMINATION AND DOMESTIC POLITICS
The discussion thus far has black-boxed the state as a single actor. In a major work of scholarship, Hein Goemans developed a war-termination theory that fused domestic politics with war-termination bargaining.46 He (almost incidentally) laid out the core bargaining dynamics of the information proposition, that battles serve to reduce uncertainty, and that if a state wins a battle it will demand more of an adversary, and if it loses a battle it will offer concessions. Goemans’ real objective, however, was to present an entirely novel theory of war termination based on domestic politics. He proposed that leaders in more repressive states were less likely to lose power following defeat in war than leaders in less repressive or moderately repressive states. He further proposed that leaders in more repressive and moderately repressive states had a greater fear of losing political power than leaders of less repressive states, because more repressive and moderately repressive leaders were likely to suffer severe political punishment, namely prison, exile, or death. He combined these two propositions to hypothesize that leaders of moderately repressive states, that is states with mixed political institutions, are especially fearful of suffering even moderate defeat. In contrast, highly repressive leaders are less fearful of facing moderate defeat because they can effectively use the tools of repression to stay in power. Democratic leaders are also less fearful of facing moderate defeat because a democratic leader thrown from power does not suffer severe personal consequences. As a result, when such mixed regimes are losing wars, Goemans predicted that rather than offer concessions, as the information dynamic would forecast, such regimes should instead raise their war aims, hoping that doing so will make available greater spoils of war, which can be distributed to the regime supporters, thereby keeping the regime in power. Leaders in such circumstances may “gamble for resurrection,” adopting risky military strategies that increase the chances of ultimate victory while simultaneously increasing the chances of exposing the state to disastrous military defeat. Goemans presented impressive case studies of belligerents in World War I to test his arguments, as well as some quantitative analyses that tested more indirect implications of his theory. In this book, Goemans’ theory receives empirical attention in chapters 9 and 10.
Other scholars have also described unusual decision-making behaviors for regimes with both autocratic and democratic qualities, so-called mixed regimes. Some have argued that such regimes are prone to over expansion, as the tendency for such regimes to be ruled by cartelized oligarchies makes pro-imperial log-rolling politics more likely. Further, such regimes are especially prone to imperial myth-making, falling victim to false confidence in one’s own power and being unable to analyze objectively the likelihood of victory in a potential war. Although this theory has been applied to the initiation of war, it could also be applied to the termination of war. The core propositions might be that such mixed regimes are especially averse to making concessions during war, and also that such regimes are especially unlikely to assess accurately the information provided by battle outcomes.
Our last domestic politics perspective on war termination focuses on public opinion.47 The basic idea is that democratic leaders are more politically sensitive to casualties than non-democratic leaders. This notion emerges from the democratic peace proposition that democracies are especially unlikely to fight (each other) because democratic leaders must beware popular discontent, which is likely to grow as casualties mount. However, there is no scholarly consensus as to whether and how the accumulation of American casualties in war has eroded popular support for presidents.48 Some studies, however, propose that democracies fight shorter wars because, as casualties mount, democratic governments become more likely to offer concessions in reaction to the sliding popularity of the war.49 This book engages this casualty-sensitivity argument empirically, exploring in its empirical chapters whether democratic leaders were especially likely to make concessions as casualties mounted.
THE TRUTH OF COMBAT?
Otto von Bismarck once declared that, “People never lie so as much as after a hunt, during a war, or before an election.”50 However, the information proposition declares that, public statements aside, combat bears unavoidable truth. Although some see the battlefield as the most “honest place on earth” because it mercilessly reveals the character of the soldier, it also bears this moniker because it provides unmanipulable information about the true balance of power between two belligerents. As horrible as combat is, it does serve the critical function of providing real information to the belligerents. That information may be the critical resource belligerents need to reach an agreement ending the war. The reliability of such war-ending agreements, and the effect of (un)reliability on war-termination behavior is the subject of the next chapter, to which we now turn.
Returning to the simple costs/benefits framework presented at the beginning of this chapter, the information environment affects a belligerent’s belief about the shape of the cost curve. Poor battlefield performance will cause a belligerent to raise its estimates of the costs of continued conflict, making the belligerent more likely to consider crafting concessions to seek an end to the war. If the enemy’s military is faring well, it will take longer and/or be more costly to destroy the enemy’s willingness to resist, or take longer to convince the adversary that eventually its military will be destroyed if the war continues to a fight to the finish. If the belligerent receives information that a third party is likely to intervene on behalf of the adversary, then this also raises the belligerent’s estimates of the costs of continuing conflict, making the belligerent more likely to consider concessions.