Introduction
Time to Talk
On August 6, 1905, representatives from Russia and Japan arrived in Portsmouth, New Hampshire. For Russian delegation leader Sergei Witte and his team, the occasion in this unfamiliar city was a bitter one. For head Japanese delegate Baron Komura and his diplomats, it was a slightly more auspicious affair. The two empires had spent the last eighteen months fighting a war over their competing ambitions in East Asia, and hostilities had slowly but surely laid bare Japan’s military superiority. Tsar Nicholas II had persistently held out hope for victory despite a relentless sequence of battlefield losses. But sixteen days of fighting between February 22 and March 9, 1905, in the Manchurian city of Mukden marked a moment of reckoning. In what would be one of the biggest land battles to be fought prior to World War I, Russia and Japan suffered a collective total of over 150,000 casualties. Japan, however, emerged from that battle with full control of southern Manchuria, one of the key disputed territories between both expanding empires, and as the dust settled at Mukden, the reality of Japan’s advantage became harder for the Russian government to ignore. One final naval disaster in the battle of Tsushima in late May sounded the death knell for Russia’s cause. By June, Tsar Nicholas II quietly decided to pursue talks with the Japanese,1 accepting a mediation offer from President Theodore Roosevelt of the United States, who had gotten diplomatically involved in March once a Japanese envoy requested his assistance. It was through Roosevelt’s efforts that the two delegations came to sit around a long mahogany table half a world away to discuss peace.
Over twelve sessions spanning August 9 to August 30, the opposing diplomats debated the terms of a potential settlement. Disagreements arose about whether Japan would demand an indemnity from Russia, but many of the other outstanding issues were resolved in short order. Japan would eventually drop its indemnity clause but claim southern Manchuria, the Korean peninsula, and the southern half of Sakhalin (Russia’s largest island, dozens of miles away from the coast of northern Japan).2 The resulting Treaty of Portsmouth was signed on September 5, 1905, ending the war.
The conflict transformed both sides’ futures. Russian humiliation would feed the flames of the First Russian Revolution, Japan’s military would gain increasing prominence in its government, and Western powers would no longer underestimate the Japanese Empire. Of particular note, however, is when diplomatic interactions started and how long they took. Diplomacy during the war was delayed in starting but moved quickly once it began. The first and only round of negotiations took place 95 percent of the way into the entire conflict. Talks, which came to pass over the remaining 5 percent, promptly reached a settlement.
Diplomatic interactions looked radically different during the Falklands War several decades later. On April 2, 1982, Argentinian troops invaded and occupied the Falkland Islands, which were three hundred miles east of Argentina and had been the subject of a long-standing dispute involving the United Kingdom. A mere three days in, US secretary of state Alexander Haig sprang into action as a mediator between the two sides. Through much of April, Argentina and the United Kingdom exchanged proposals with little room for compromise.3 The death of one thousand Argentinian soldiers due to the sinking of the ship ARA General Belgrano on May 3 brought both sides back to talks, this time led by United Nations secretary-general Javier Pérez de Cuéllar. Discussions once again faltered as Argentina rejected the United Kingdom’s terms, which essentially described the prewar status quo. By late May, the languishing discussions became irrelevant. British reinforcements had traversed the Atlantic while the talks were happening and swiftly overwhelmed the Argentinian occupying troops. On June 14, General Mario Menéndez, commander of the Argentinian garrison at the capital town of Stanley, surrendered to Major General Jeremy Moore.4 The two sides discussed, drafted, and signed a document of surrender the following day, bringing the conflict to a close.
Over the course of the war, two rounds of talks came and went before the third codified British victory. The first attempt began almost immediately once violence erupted—only 5 percent of the way into the war that would come to pass. The second took place at approximately the halfway point. These two efforts together spanned over 50 percent of the conflict’s entire duration. The third negotiation, lasting only one day, consolidated the United Kingdom’s victory over Argentina.
The Russo-Japanese War and the Falklands War exemplify two very different stories regarding the timing, frequency, and consequences of negotiations that occur during armed conflict. Their contrasts are indicative of the wide range of variation we see in wartime diplomacy across time and space. As this book demonstrates, talks occurred almost immediately in some conflicts, later on in others, and not at all in some. Several negotiations took almost the entire duration of a given war, others a handful of days. And only some talks resulted in peace, whereas others only prolonged, or even exacerbated, the conflict. What explains the differences in the timing, frequency, and consequences of wartime negotiations? When and why do belligerents decide to start negotiating again? What conditions help dictate whether or not these talks result in peace? What effects, if any, can negotiations have on war beyond codifying its potential settlement? To what extent does this line of thought help to promote or revise our understanding of diplomacy—either during conflict or more generally?
This book addresses these questions by developing and substantiating a theory of wartime negotiations. My theory explains what conditions promote the occurrence of negotiations, as well as what impact they have on the subsequent trajectory of the war. I show that negotiations are a far more strategic, complex, and potentially cynical activity than most scholars and practitioners have generally recognized. A central aspect of my argument is that talking while fighting can serve more than one purpose and that we can understand when each purpose is more likely to be prioritized over the other. This claim, as innocuous as it may seem, leads us to recognize what some might consider a “dark side” of diplomacy. Ultimately, negotiations not only are used to help settle conflicts but can also be harnessed to help delay, fight, and potentially win them. Diplomacy does not intrinsically promote peace. Depending on the circumstances, it may promote a belligerent’s war aims and be worse than staging no talks at all. If negotiations can indeed worsen conflict, it is crucial to understand the conditions under which negotiations are more likely to do so and how to minimize the risks of diplomacy being exploited to perpetuate violence.
What We Know about Negotiations in War
Up to this point, very few sources of data allowed us to see this variation in a more systematic manner. New negotiation data for ninety-two interstate wars between 1823 and 2003 offer one of our first chances to see this variation across two centuries of violent international conflict. This variation is captured in figure 0.1.
Figure 0.1 shows how dramatically wars vary in the relative timing of their negotiations. We must first note that a fair number of conflicts proceed without a single round of talks between belligerents. The invasions of Afghanistan and Iraq at the turn of the twenty-first century are contemporary examples; the La Plata War between Argentina and an alliance led by Brazil, which began in July 1851 and ended with a Brazilian military victory in February 1852, is a case from the more distant past. For wars that do feature negotiations, some talks occurred almost immediately once hostilities erupted, as was the case in the Falklands War; others started shortly before the conflict came to an end, as in the Russo-Japanese War; and many others began somewhere in the middle. The 1999 Kargil War between India and Pakistan is an instance of this last scenario: the first talks began on June 12, thirty-five days after the war began and thirty-five days before it ended.
Figure 0.1 also demonstrates how many talks failed to produce peace and instead ended up marking what would become the early stages of the war. If all negotiations quickly led to settlement, the plots would feature one short bar on the right-hand side of each conflict. The negotiation data indicate that 49 out of 189 total negotiation efforts across these wars, or about 25 percent, directly resulted in the termination of hostilities. Figure 0.1 also highlights a remarkable degree of variation in how much time warring states chose to communicate with each other while fighting. In a majority of conflicts, belligerents do not afford much time to sit at the bargaining table. Yet in a sizable number of cases, official state representatives spend between 20 percent and 80 percent of the war’s overall duration talking to one another.
The reality of wartime negotiation thus appears contradictory. States frequently hesitate to negotiate, but when they do negotiate, talks are intermittent and regularly evince apathy or antipathy toward peace. For all the intricate differences we see across these conflicts, contemporary scholarship presents few explanations for why or how diplomatic interactions take place during war. But before I advance a new theory to tackle these questions, it is worth addressing what answers have already been offered.
Rational Reasons for War
Since the 1990s, scholars of international relations have broadly scrutinized war using a “rational choice” framework, which conceives of war as a form of bargaining.5 We imagine that two or more actors are in a zero-sum dispute over how to divide a valuable good, such as territory or policy, between themselves. The actors either could split the resource peacefully by striking a mutually acceptable bargain or may attempt to use force to wrest a larger share. Each side’s preferences regarding whether to take a pacific or bellicose approach are shaped by beliefs about the distribution of power between them, as well as the potential costs they would be willing to withstand from fighting. These two concepts are commonly referred to as capabilities and resolve. All of these judgments are made under a shroud of uncertainty, where the other side’s true capabilities, resolve, and intentions are not fully known.6
In an agenda-setting article, James Fearon produces a game-theoretical model reflecting this logic and demonstrates that wars can be understood as a rational process where belligerents use violence to determine their minimally acceptable bargaining offers, which in turn reveals where a mutually acceptable zone of potential agreements—in political science parlance, a bargaining range—may exist. Using this framework, Fearon underscores two mechanisms that could trigger war: incomplete (or asymmetric) information and commitment problems.7 In an incomplete information dynamic, actors may have incentives to overstate their capabilities and resolve—that is, to bluff. By doing so, one party hopes to convince the other that a war over the disputed good would be unwise, which in turn would produce a more generous diplomatic bargain.8 The escalatory moves each side makes to demonstrate the credibility of its words can eventually push both sides over the precipice and into an unintended armed conflict.
With respect to commitment problems, an actor that is already or will become powerful cannot reassure anyone that it will not use its power to change the status quo; it would be fully rational for the actor to take advantage of its own growing strength. Other actors may worry that, if left unchecked, the ascending actor will push the bargaining range dramatically in its own favor and be more likely to realize its revisionist goals. These fears could lead concerned parties to launch a war with the goal of precluding such a disadvantageous shift in the first place.9
These two mechanisms highlight why actors may engage in war even though there is likely a peaceful and less painful deal that could have been made in the abstract. For whatever way in which actors divide a good, they would both prefer to make the split without fighting, as violence inevitably creates human, economic, and psychological suffering that negates at least some of the gains made. The fact that wars occur despite this fact is commonly called the inefficiency puzzle.10 War, in this view, is a mechanism through which actors impose and absorb real costs in order—or, rather, because they failed—to demonstrate the validity of their own words.
Three Assumptions in Current Knowledge
The rationalist perspective on war has usefully illuminated the role of information in motivating conflict. Yet while works following this tradition seem intuitive and internally consistent, they harbor three interrelated assumptions regarding wartime diplomacy that are worth identifying and questioning. These involve the timing, cost, and purpose of negotiations.
Timing
Early rationalist models of war were focused on explaining war initiation and did not directly address the question of what happens once wars erupt, either on the battlefield or at the bargaining table. The outcome of the entire war is viewed as a coin flip weighted by each side’s military power. Such models are often called costly lottery models. Most quantitative studies of war adopt a similar methodology in which entire conflicts are taken as the unit of analysis; only fixed or initial characteristics at the outset of hostilities can shape the overall war, which itself is usually understood in terms of the victor or the duration of conflict. These methods preclude the study of how events that occur during the war itself, including negotiations, can reshape the war’s trajectory once hostilities have already commenced.11
Subsequent bargaining models acknowledged that a comprehensive theory of war should explain not only its initiation but also its duration and termination.12 These game-theoretic works explore intrawar dynamics and, importantly, acknowledge that diplomacy plays a part in ending many conflicts. Many models of this variety share a roughly similar structure. War is depicted as an alternating process between the exchange of a take-it-or-leave-it bargaining offer and a potentially war-ending battle if the offer is rejected.13 Bargaining offers made near the outset of conflict may be more extreme. These inordinate proposals could reflect a genuine demand made in the midst of uncertainty, or they may even be intended to strategically manipulate the information environment and signal apparent strength. As the war progresses, each battle fought reveals new and less manipulable information about each side’s relative capabilities, and costs incurred from these clashes further test the parties’ resolve. Fighting is learning. Through the conduct of war, whatever proposals were initially made at the bargaining table based on less information are pushed aside. Belligerents are forced to update their beliefs about future expectations and thus revise their bargaining positions to be more closely aligned with the harsh truth revealed from hostilities.14 Wars end either when belligerents reach a common set of expectations such that one side accepts the other’s revised offer or when one side is totally defeated through a fatal battle. Critically, this generation of bargaining models explicitly clarified that wars need not end in complete victory or defeat, nor do they even require the belief that either total outcome is possible. War is an attempt by at least one side to improve on the status quo, and not necessarily to vanquish an opponent.15
Scholars disagree on the exact mechanism through which actors translate what they learn from fighting into behavior that could settle the conflict. These divergences lead to contrasting implications about the frequency and timing of negotiations. In most cases, models portray talks as a convergence process where belligerents slowly make more generous offers in order to screen out weak actors and to identify genuinely strong actors that are willing to fight longer to obtain more. These models imply that negotiations occur constantly throughout the entirety of a war.16 A smaller but notable contingent of research instead argues that belligerents’ bargaining process is better reflected as a war of attrition, where actors obstinately refuse to change their bargaining positions through most of the conflict before relenting and swiftly reaching an agreement.17 These works suggest that negotiation does not occur for much of the conflict. Regardless of the specific mechanism, both representations of diplomacy assume that verbal bargaining is essentially instantaneous, allowing belligerents to exchange countless offers while fighting.18
More contemporary studies, predominantly focused on civil war, have challenged both views by suggesting that negotiations are a distinct stage of a broader conflict resolution process that is motivated by numerous political and strategic factors.19 These works follow in the tradition of I. William Zartman’s work on ripeness theory.20 In brief, ripeness theory identifies two conditions that prompt belligerents to seek a negotiated settlement. First, the parties must feel as if they are trapped in a costly deadlock with no victory in sight; second, the belligerents must perceive a way to break out of the deadlock through the possibility of a negotiated and mutually acceptable settlement. When these two conditions are met, warring parties may thus be “ripe” for negotiations. Ripeness theory has for many years been the leading framework to understand the onset of negotiations during hostilities and will be discussed in greater detail in the next chapter.
Costs
Many formal models of war suggest that the act of negotiating during hostilities, separate from the substantive concessions made, is costless. The idea that negotiations occur in between every round of fighting attests to this notion. Such an assumption is a natural consequence of the costly signaling tradition, which treats most diplomatic communication as “cheap talk.” An extensive literature argues that verbal communication is not particularly costly and is thus not credible on its own, because it is hard to verify and easy to take back.21 If diplomacy that occurs in the midst of crises can be cheap, then diplomacy during war seems even cheaper and less meaningful to understanding active conflict, especially when extraordinarily costly violence is communicating far more credible information.
Oriana Skylar Mastro’s work challenges this view.22 According to her costly conversations thesis (CCT), belligerents generally adopt a “closed” diplomatic posture when they fear the opponent may infer that an interest in diplomacy reveals vulnerability. Mastro refers to this as the “adverse inference of weakness.” CCT proposes that belligerents are likelier to adopt open diplomatic postures when they believe they have demonstrated enough strength to overcome any adverse inference and when their opponent has relatively low strategic capacity to exploit any signs of weakness. Impressively researched studies of several wars support these claims.
Purpose
Across many formal models, the bargaining positions taken or any concessions made by a belligerent are a mechanical reflection of the battlefield. The more one actor dominates in recent clashes, the more that bargaining positions drift in that actor’s favor. The longer one actor continues suffering costs without relenting, the more that bargaining offers will be more generous to that actor. When parties do choose to negotiate, it is because their bargaining positions have become sufficiently similar and their desire to end the conflict supersedes their interest in further bloodshed. Negotiations are a formal manner through which belligerents codify the outcomes of hostilities.
This last element overlaps with conventional notions of negotiation, which assume participants are willing to discuss items in good faith with the sole purpose of reaching an agreement, even if they may ultimately fail. In the bestselling book Getting to Yes, the cofounders of the Harvard Negotiation Project, Roger Fisher and William Ury, define negotiations as “back and forth communication designed to reach an agreement.”23 An article reporting on a laboratory experiment by the psychologist Peter Carnevale and the sociologist Edward Lawler describes negotiation as “a form of symbolic communication that involves two or more people attempting to reach agreement on issues where there are perceived differences of interest.”24 Indeed, experimental studies of negotiations often adopt designs that reward participants for reaching a mutually acceptable agreement.25 In the policymaking realm, former secretary of state Dean Acheson also observes that “negotiation, in the classic diplomatic sense, assumes parties [are] more anxious to agree than to disagree; parties who are, therefore, willing to make concessions in determining what shall be agreed upon.”26
The Puzzling Reality of Negotiations
The general thrust of conflict research therefore portrays wartime diplomacy as an ever-present, costless, mechanical, and earnest endeavor. Yet this logically consistent and intuitive depiction faces a problem: the empirical record does not support it. As figure 0.1 shows, patterns of negotiations during conflict are highly sporadic, vary widely by conflict, and do not always respond to outcomes in fighting.
Histories of war suggest that leaders frequently feel reticent to talk while fighting and consider the decision to engage in diplomacy a serious one. During the Crimean War of 1853–56, Tsar Nicholas I of Russia refused to negotiate with the Allies (Great Britain, France, and the Ottoman Empire) despite a series of battlefield victories by the Allies in mid and late 1854. It was only once Nicholas I died and was replaced by Alexander II that the belligerents met in Vienna in March 1855. But after several days of talks, all parties realized that nobody was prepared to make significant concessions.27
The Vienna talks during the Crimean War also illustrate that the onset of negotiations is not equivalent to the settlement of conflict. According to the negotiation data that I describe later in this book, approximately three-quarters of all periods of negotiation during the last two centuries of interstate war are not associated with the cessation of hostilities. About half of all interstate wars in the same period feature more than one round of negotiation. Each of these statistics suggests that belligerents fluctuate in their willingness to talk, sometimes pulling back from diplomacy after making the initial choice to participate. Current theories of conflict are not well equipped to explain these shifts.28
Many negotiations do not simply fail to reach an agreement but seem to reveal complete antipathy toward any form of compromise. Several months into the Second Sino-Japanese War of 1931–33, diplomats at the League of Nations hurriedly attempted to spearhead mediation efforts to stem the explosion of violence. The most notable efforts were formulated by the US and British ministers to China, Nelson T. Johnson and Sir Miles Lampson. The two arrived in Shanghai (the next target of Japan’s expansion into China) on behalf of the league on February 12, 1932, with hopes of establishing a neutral zone. The visit went nowhere. By February 18, Japanese military representatives issued the mayor of Shanghai terms of an ultimatum with a deadline of February 20 that, in Lampson’s words, were “in so arrogant a tone and dictatorial a form that it was hardly possible to expect their acceptance by the Chinese.”29 The mayor rejected the ultimatum, and Japanese forces that had arrived on February 18 renewed their assault. As such, the idea that belligerents constantly drift toward agreement and that negotiations signify a nearly irreversible shift toward peace is not justified.
One potential interpretation of this multifaceted negotiation behavior is that diplomacy is little more than “noise.” Patterns of negotiation could perhaps be motivated by extraneous factors that are disconnected from the conflict itself. If so, then negotiations reveal very little about war, the observed variation is not worth further scrutiny, and conflict scholars have rightly treated diplomacy in the manner I have described thus far. But another interpretation is that we have severely undervalued the highly strategic nature of wartime negotiations and what they reveal about conflict. A lack of detailed data about conflict diplomacy and a paucity of theoretical tools to understand it may have created a vicious cycle that precludes a serious investigation of the topic.
The Argument
How can we understand when, why, and how diplomatic talks are used during war? I offer a straightforward overarching argument: negotiating in the midst of conflict involves costs and benefits, and belligerents continuously weigh these competing forces in order to choose whether to negotiate and how to negotiate. The simplicity of this thesis is somewhat deceptive, as the scholarly literature has not fully appreciated or synthesized the numerous upsides and downsides associated with diplomacy during war.
To develop this argument, I distinguish between two forms of negotiation, which I call sincere and insincere. Sincere negotiations refer to good-faith attempts by belligerents to reach a mutually acceptable agreement. Sincerity does not guarantee successful compromise but underscores the participants’ genuine preference to reach an agreement rather than not. As I will discuss in the next chapter, both our common and scholarly understandings of negotiations overwhelmingly adopt this view. One certainly cannot deny that negotiations are necessary to end conflicts short of complete military victory or defeat. But to assume that this is diplomacy’s only function embodies an unsophisticated view of what negotiation can accomplish and reveal during conflicts. My theory therefore emphasizes the insincere mode of negotiation, which refers to interactions where at least one party seeks to use talks for some purpose unrelated or antithetical to agreement. In the wartime context, this means that a belligerent exploits the time spent talking in order to realize some other objective that serves to promote policy aims that aid its war effort. Two significant goals motivate insincere negotiations: deflecting political blame for participating in armed conflict and buying time to regroup and remobilize one’s military forces to fight more effectively in the near future.
I am certainly not the first to suggest that negotiations can be orchestrated in bad faith. Numerous scholars and practitioners have identified this phenomenon using terms such as “side effects,”30 “devious objectives,”31 “posturing,”32 “duplicitous negotiations,”33 “avoidance bargaining,”34 “phony bargaining,”35 “striving for no,”36 “false negotiations,”37 “false readiness,”38 “insincere bargaining,”39 and “insincere negotiations.”40 These works, however, merely note the existence and possibility of bad-faith diplomatic bargaining without providing an explanation for when it should be more prone to occurring. My argument provides one of the first systematic explanations of the conditions in which negotiations would more likely be sincere or insincere.
I establish a theory of wartime negotiations that contends that belligerents’ decisions to negotiate and whether to negotiate sincerely or insincerely are dictated by two primary factors. The first is what I call latent external pressure. This reflects the amount of influence that third parties can activate in order to compel belligerents to engage in diplomacy. The descriptive term “latent” captures the notion that the overall degree and quality of pressure that outside parties can place on belligerents in any particular conflict are partially moderated by the broader existence of institutions, norms, and structures that facilitate such behavior. The second factor, already widely accepted in extant research, is information obtained from the battlefield. The interactions of these two factors account for the relative costs and benefits of negotiating during war.
In the absence of any external pressures to negotiate, belligerents perceive enormously high costs to negotiating with the enemy. Looking too eager to talk can be interpreted as a sign of weakness, which may encourage the enemy to fight harder because it thinks victory is at hand or discourage one’s own people from fighting because they believe defeat is imminent.41 The simple act of negotiation, regardless of what positions belligerents espouse when talking, can therefore worsen a belligerent’s prospects in a war if talks do not produce a diplomatic settlement. Consequently, when latent external pressures are low, we should expect negotiations to happen infrequently, and most would-be attempts to negotiate insincerely should be deterred. The likelihood of talks will increase as the results of fighting trend heavily in one side’s favor, and any talks that do occur have a high probability of being sincere attempts to cease hostilities and minimize one’s costs by avoiding needless bloodshed.
The talks in Portsmouth during the Russo-Japanese War are a paradigmatic example of sincere negotiations, which took place in an environment with relatively low latent external pressures for peace and a battlefield solidly trending in one side’s favor. Tsar Nicholas II very reluctantly acceded to negotiations because Russia’s weakness had been made apparent and few other alternatives remained. Both sides came to the table of their own accord and were therefore able to quickly settle their affairs.
Incentives to engage in diplomacy change considerably when latent external pressures to negotiate are high. Constant attempts by third parties to initiate talks, as well as the pervasiveness of norms or institutions aimed at peace, lower the costs to negotiation. Being willing to talk is interpreted more as a submission to the international community, and belligerents do not have to initiate negotiation efforts on their own. These cost-reducing effects exist regardless of whether third parties wield sufficient leverage to actually alter the prospects of a lasting settlement.42 On the one hand, this may enable some wars to end more quickly because warring parties can start talking more expeditiously (or at all) without worrying about as many negative repercussions of doing so. On the other hand, this permissive environment also enables belligerents to negotiate more freely regardless of the state of the battlefield. Actors thus have greater liberty to try negotiating insincerely in hopes of improving their ability to prosecute the war effort when talks fall apart by design. While diplomatic communications are relatively common in this high-pressure environment, the battlefield plays an important role in dictating whether talks are sincere or insincere. When the battlefield provides limited or indeterminate information, negotiations that come to pass are probably insincere in nature, exploited to further at least one side’s cause. As fighting tilts heavily in one side’s favor, negotiations are more likely to be sincere and to bring the war to a diplomatic conclusion.
The two fruitless negotiations that punctuated the Falklands War exemplify insincere negotiations. Both were products of enormous international pressure that was immediately activated once hostilities erupted, occurring before fighting had taken place in earnest. Mediated diplomatic communication moved forward even though neither side had any real desire to reach a settlement. But negotiations were not only unsuccessful in reaching peace. The United Kingdom used the period of diplomatic bargaining to deploy its forces to Argentina, and both sides exploited the apparent diplomatic deadlock to animate their domestic audiences in favor of more conflict. The only sincere talks of the war lasted one day and wrote down the terms of Argentina’s military surrender.
The Russo-Japanese War and the Falklands War are two archetypal cases that illustrate my theory in stark terms. But as I will establish in the chapters to come, my argument effectively explains the causes and consequences of wartime diplomacy over two centuries of global conflict. The seemingly arbitrary nature of negotiations in conflict is, I show, not arbitrary at all.
Contributions
The core evidence for my theory relies on new quantitative data of battles and negotiations over two centuries of interstate conflict. These data are based on a wide-scale review of several hundred primary documents, reference materials, and periodicals. By combining these two resources, I am able to track the ebbs and flows of wartime fighting and talking at the daily level. My data afford the opportunity to study wars at a level of breadth and granularity that is unmatched by current studies of war. Figure 0.1, as simple as it looks, reveals variation that cannot be readily explained using extant theories of conflict. This suggests that there is a great deal about negotiations that the fields of international relations, diplomacy, and conflict resolution do not fully appreciate, much less understand.
My battle and negotiation data not only enable a systematic assessment of my theory but may also help to reinvigorate the quantitative study of war. Inspired by Fearon’s seminal article, a surge of research around the turn of the twenty-first century used game-theoretic models to further flesh out the notion of war as a bargaining process, but the endeavor has since become a more dormant enterprise. This is likely not a coincidence. Very few resources have been created to directly validate, challenge, or refine the premises or conclusions of these bargaining models—many of which contradict each other—on a larger scale.43 Debates regarding whether wars are resolved through a convergence or attrition process are one prominent instance where conclusions differ based on the structure, payoffs, and presumptions of the model. My new data represent a sizable step in this direction and may help motivate others to gather more quantitative measures of intrawar activity. Doing so would help further crack open what Scott Gartner memorably called the “black box of war,”44 breaking the cycle where a lack of data on wartime negotiations precludes an exploration or appreciation of wartime negotiations, which in turn inhibits the collection of data on them.
On the academic front, this endeavor challenges the penchant of international relations scholarship to shortchange the importance of diplomacy during war. It is natural to think of wartime negotiations as being the converse to armed violence; belligerents shed blood on the battlefield and resolve their differences at the bargaining table. But even though all diplomatic settlements require negotiation, not all negotiations lead to diplomatic settlements. The fact that these “unsuccessful” negotiations can have political implications, be harnessed to moderate conflict, alleviate third-party pressures, and reshape battlefield outcomes poses a nontrivial challenge to extant scholarship that does not seriously contemplate intrawar diplomacy or that solely views it as a manner to grant concessions. As such, my theory indicates that the rationalist tradition of dismissing diplomacy as mere “cheap talk,” especially during war, is overly reductive and deleterious to our understanding of international conflict. Negotiations during war are not an alternative to war but an essential tool to navigate it.
In pointing out how the bargaining model of war harbors an oversimplified conception of negotiations, I do not mean to suggest that the bargaining model’s broader conception of war is fundamentally wrong. My argument adopts a rational approach where actors weigh the balance of costs and benefits when determining their optimal course of action moving forward; concerns about information, capabilities, resolve, and intentions all matter greatly in these calculations. Instead, I emphasize the fact that the occurrence of negotiations—regardless of the specific proposals exchanged during them—can produce new consequences that influence each side’s leverage and thus its prospects in the violent bargaining process of war.
While scholars in the realm of international relations have deeply scrutinized the strategic underpinnings of war without as much emphasis on negotiations that could occur at those times, scholars focused on negotiation studies and diplomatic studies have taken the matter of negotiations more seriously. But they frequently do not consider them in a wartime context. My theory thus represents a useful coalescence of insights from, and contributes to, related research tracks on war, conflict resolution, diplomacy, and negotiations.
This book also offers significant insights on conflict resolution when it comes to policymaking. Since most wars involve and end through diplomacy, understanding the incentives that belligerents have with respect to negotiations is vital for producing thoughtful policies that promote peace rather than undermine it. For instance, my theoretical framework and empirical findings advance thinking on the seminal policy debate about the merits of outside intervention in wars. Reflecting on the United Nations’ uneven impact in the Bosnian War of 1992–95, Edward Luttwak memorably argued that international actors should stay out of conflicts and “give war a chance,” as the only manner to truly end hostilities is to allow belligerents to fight to mutual exhaustion or complete defeat.45 Such a proposal insinuates that the international institutions we have built since World War II to maintain peace and stability are actually undermining their professed goals. This book indicates that constant efforts to spur diplomacy, even if well meaning, could ultimately be counterproductive. Yet it also identifies moments when diplomacy facilitates peace. By systematically pinpointing the specific conditions of when those moments may (or may not) occur, this book thus furthers our understanding of when and which conflict resolution policies might be effective.
Additionally, my theory highlights the potentially unscrupulous side of diplomacy. US policymakers often treat negotiation as a good-faith activity and do not see the cynical ends that negotiation can facilitate.46 In a memoir, Dean Acheson lamented many policymakers’ naive view of the relationship between negotiation and conflict: “I have heard people who should know better, including a head of government, say happily, ‘As long as we keep them talking, they’re not fighting.’ Nothing could be more untrue: they are fighting… . To our minds international conferences and negotiations are so completely means for ending conflict that we are blind to the fact that they may be and, in the hands of experts, are equally adapted to continuing it.”47
Strong normative beliefs conceive of diplomacy as an inherently genuine process,48 which perhaps explains why policymakers around the world do not fully appreciate the power of duplicitous negotiations as a weapon of war. This book calls for a serious revision of this unsophisticated view of diplomacy.
The Book to Come
Chapter 1 lays out my full theoretical framework regarding wartime negotiations. To that end, I first define key concepts of the book with reference to current literatures on war, negotiations, and their intersection. The remainder of the chapter develops the central components of my theory and explains the conditions in which we are most likely to see negotiations that may productively resolve hostilities, or when negotiations may be more easily exploited for instrumental ends. I place particular emphasis on insincere negotiations, as their existence and effects introduce the implications that stand in starkest contrast with existing research on conflict and its resolution.
This discussion leads to a series of testable implications that I assess in subsequent chapters. Since my theory features many moving parts, no single test can fully substantiate it in its entirety. I rely instead on several forms of empirical inquiry to assess different aspects of my argument at various levels of analysis, and I have selected case studies to show the validity of the argument across a wide span of time and space.
Chapter 2 discusses the empirical cornerstone of this book: data on battles and negotiations in almost all interstate wars over the last two centuries. I describe the raw materials, definitions, and procedures used to convert hundreds of qualitative historical sources into two quantitative datasets on wartime negotiations and battles across ninety-two conflicts. I demonstrate the utility and plausibility of these two datasets by illustrating patterns of wartime negotiations and battlefield trends within wars. Even these purely descriptive statistics reveal insights about war that would be infeasible with extant resources.
Chapter 3 is the first to directly assess my theory of wartime negotiations. I address the impact that latent external pressures and battlefield outcomes have on the initiation of negotiations, as well as their relationship with the termination of hostilities. While my battle data allow me to create straightforward measures of battlefield outcomes, it was necessary to develop a method to measure latent external pressures on belligerents to negotiate. To that end, I leverage the dramatic change in the international environment before and after 1945. This decision is based on the idea, backed by substantial scholarship, that the post–World War II era witnessed an immense growth of institutions, actors, and norms that promote peace and stability. I argue that these innovations have led to a substantial and systemic decrease in the costs of negotiating, as it becomes easier for third parties to activate pressure on belligerents to talk to one another. I establish that the 1945 line is indeed associated with a dramatic upward shift in the rate of wartime negotiations.
Furthermore, I show that wars before 1945—that is, wars with relatively low latent external pressures—indeed have less frequent negotiations, but that talks become more frequent and promptly terminate wars when the battlefield firmly moves in one side’s favor. Meanwhile, in wars after 1945, negotiations occur quite frequently and bear little relationship with the state of the battlefield. Many talks come and go without any steps toward peace. A battlefield where one side has a solid advantage, however, does help to predict whether negotiations that do occur will lead to settlement. A number of supplementary analyses provide more context for why the pre-1945 and post-1945 diplomatic environments look so different, as well as what factors may account for the actual deployment of diplomatic pressure by third parties. I further substantiate my argument regarding latent external pressures by leveraging whether wars involved any major power belligerents, who I expect to be more immune to external diplomatic pressures. In support of my expectations, I find that wars involving only minor powers feature higher rates of reluctant negotiation because the belligerents are more dependent on external material and political support.
Any quantitative study of conflict will inevitably sand away details and nuances of individual events in exchange for evidence of broader trends.49 As such, the second part of chapter 3 features a qualitative comparison of two conflicts that feature the same main belligerents: the Greco-Turkish War of 1897 and the Turkish invasion of Cyprus in 1974. The case studies show the importance of latent external pressures in influencing negotiation behavior. The 1897 war featured relatively few external sources of diplomatic pressure; the only round of talks that took place codified the Ottoman Empire’s victory. In contrast, the 1974 conflict was woven into a broader fabric of international politics. A two-week diplomatic effort spearheaded by the United Kingdom and the United States produced no peace and instead allowed the belligerents to mobilize their political and military assets so that they could fight more effectively once talks ended.
Whereas chapter 3 analyzes how battlefield activity can influence negotiation behavior, chapter 4 explores how negotiations can reshape the battlefield. This tests my claim that insincere negotiations not only fail to end hostilities but can be exploited to help reshape the subsequent trajectory of conflict. I return to my quantitative data to substantiate two implications of my theory. First, I confirm that periods of negotiation are associated with lower levels of active hostilities on the battlefield. These moments of relative quiet on the battlefield are natural for belligerents who want to sincerely stop fighting, but they are also necessary for belligerents that merely want to abuse negotiations to improve their political and military position. Second, I demonstrate that negotiations that do not result in peace are followed by battlefield trends that systematically favor war targets and that are more consistent with prewar expectations. This result is particularly noteworthy and underscores the importance of viewing negotiations as a very real extension of war.
The chapter then uses a qualitative approach to walk through the relationship between fighting and bargaining and to show the deep military and political impacts of negotiation in two distinct wars: the War of the Roman Republic in 1849 and the Cenepa Valley War of 1995. The former is a valuable example of how insincere negotiations can occur in the absence of explicit external pressures. The latter, a South American conflict, is the most recent war for which there is adequately accessible documentation of public and private diplomatic activity.
Chapter 5 offers an in-depth qualitative study of the Arab-Israeli War of 1948–49, a historical clash in the Middle East that is also one of the first cases to erupt in an environment of high latent external pressures after 1945. This conflict illustrates how the dynamics that I examine separately in chapters 3 and 4 can coalesce into a more complex and strategic story. I show that constant mediation efforts by the United Nations, while motivated by optimistic intentions, enabled both Israel and the Arab states to exploit phases of negotiation to rearm themselves and to frame their opponent as the wayward party that was unwilling to consider peace. I argue that Israel’s victory and survival in early 1949 is largely the product of how the Ben-Gurion government exploited insincere negotiations to give Israel valuable time to buttress its unprepared military forces.
Up to this point, much of my empirical evidence treats entire periods of negotiation as being sincere or insincere. Chapter 6 explores whether my theory can also explain variation in negotiation behavior within individual rounds of talks. I do so by analyzing the Korean War of 1950–53. The war is notable in that it is still technically active today and in that it also involves major powers on either side. I conduct a hybrid analysis that includes both a qualitative review of the conflict and a highly granular quantitative analysis of battlefield activity and negotiation behavior. Using a combination of text-as-data and machine learning methods, I convert thousands of pages of declassified archival documents that track battlefield movement, casualties, and negotiation behavior to daily-level quantitative measures. These data allow me to directly test how small changes on the battlefield influence a belligerent’s proclivity to be more sincere or insincere within longer bouts of negotiation, all while keeping the level of latent external pressures constant. They reveal patterns that are consistent with my theory of wartime negotiations. When the battlefront provided new information through troop movement or losses, both delegations engaged in more substantive and seemingly sincere discussions that were related to settling the conflict. Conversely, when little was learned from fighting, the delegation’s negotiation behavior became more overtly hostile, obstinate, and insincere.
The conclusion wraps up the book with directions for future lines of inquiry and policy recommendations regarding the conduct of wartime diplomacy. I emphasize the complicated balancing act that my theory and findings uncover regarding the role that the international community can play in steering the course of conflicts. In some cases, external actors may succeed in their professed goal of preventing wars from causing more suffering. But in other instances, similar diplomatic endeavors not only may fail to strike a peace but may even undermine that objective. When wars cannot end through military means, diplomacy becomes an olive branch in one hand and a sword in the other. This book provides the structure to understand when and why each hand is extended.