The greatest film ever made about nuclear war was the 1964 black comedy Dr. Strangelove Or: How I Learned to Stop Worrying and Love the Bomb, directed by Stanley Kubrick. In it, the title character, an eccentric presidential adviser with an affected manner and heavy German accent, described a doomsday machine. This device, when triggered, would detonate a series of nuclear explosions that would release enough radiation to destroy humanity. The doomsday machine is intended to deter the enemy from attacking. It would be activated automatically by a computer rather than a human, on signal of enemy attack. Dr. Strangelove explains to the president, “Deterrence is the art of producing in the mind of the enemy the fear to attack. And so, because of the automated and irrevocable decision-making process which rules out human meddling, the doomsday machine is terrifying, and simple to understand, and completely credible and convincing.”Footnote ¹

When working on the Dr. Strangelove script, Kubrick needed an adviser on nuclear weapons and recruited a Harvard economist named Thomas Schelling. The logic of the doomsday machine was something that Schelling had been considering for years, as an academic with deep concerns about foreign policy. America believed that it needed to threaten nuclear war to deter Communist aggression around the world. But nuclear war would also destroy America, meaning that the threat to start a nuclear war would not be believable, it would not be credible. The question for Schelling and others was, how can the US or any actor make an incredible commitment or threat more credible?

Schelling saw the (hypothetical) doomsday machine as part of a cluster of solutions: The US, or any country for that matter, could make its commitments to go to war more credible by tying its hands. The heart of the problem was choice, that if a government committed to go to war under certain conditions, when the time came it would choose not to follow through on its commitment, because doing so would mean the mutual assured destruction of nuclear war. The commitment lacks credibility because governments would choose surrender over self-destruction.

Tying hands was seen as a solution to this credibility problem. If leaders cannot be trusted to choose to follow through on a commitment, then they can solve this problem by altering the terms of choice. Tying hands echoes the old aphorism, “I have no choice, my hands are tied,” distantly related to Odysseus in ancient Greek myth tying himself to the mast of his ship, so he can’t crash his ship upon hearing the bewitching songs of the Sirens. Altering choice can mean making it impossible not to follow through on the commitment, as Dr. Strangelove’s doomsday machine promised to do, or by making not following through on the commitment prohibitively costly.

Schelling and others used tying hands to describe what the US did during the Cold War to make its commitments more credible. They outlined a series of different tying hands foreign policy tools states implemented to make their commitments more credible, including threatening inadvertent escalation, signing alliance treaties, deploying tripwire forces abroad as human hostages, and giving computers or mad leaders the authority to call down the furies of war. The tying hands perspective became widely accepted as an account of how American foreign policy has operated since 1945 and as a wide-ranging description of the dynamics of international relations and other elements of politics.

The tying hands perspective seemed to solve a number of puzzles of Cold War history. How did the US get the Soviets to back down in the 1962 Cuban Missile Crisis? By creating a risk of inadvertent escalation to nuclear war, causing the Soviets to “blink” first. Why did America sign alliances with dozens of countries around the world after 1945? To stake its reputation on the defense of many countries, risking the destruction of its prestige if it did not intervene to stave off aggression against an ally. Why did America deploy a militarily trivial number of troops to West Berlin? So that those troops would serve as human hostages, a tripwire whose deaths would force American intervention. Why did President Nixon pretend to be crazy in negotiations with North Vietnam? To frighten the Communists into making concessions at peace negotiations. Why did the Soviet Union create the “Dead Hand” doomsday machine system giving a computer the ability to launch a nuclear war? To cut human common sense out of the decision to launch suicidal retaliation.

The tying hands perspective was not merely a Cold War concept quickly forgotten after the Berlin Wall fell. To the contrary, the idea became one of the most important ideas in the study of international relations, and to political science more broadly. It was critical in helping create enthusiasm for using rational choice theory to understand international relations. It has also framed foreign policy debates into the 2020s, on issues as wide ranging as American policies toward NATO, the containment of North Korea, China, and Russia, foreign troop deployments, the role of artificial intelligence in foreign policy, and many others. It is now an article of faith that states and leaders tie hands as a means of making their commitments more credible.

This book challenges this conventional wisdom. The tying hands perspective is a false image. States and leaders avoid tying hands, during the Cold War and beyond. States, leaders, and their societies understand that tying hands means sacrificing the flexibility to avoid getting dragged into an undesirable war, the wrong war. Actors are deeply averse to forgoing this flexibility, because the costs of the wrong war are so spectacularly high, especially in the nuclear age. They prefer their bridges unburned, their exit doors open, and their hands untied, as described in the graphic illustration on “Dr. Strangebrain” located just after this book’s title page. As a result, they make flexible rather than binding commitments, giving them the ability to avoid war. Understanding that states and leaders avoid tying hands provides an alternative perspective on how states make threats and other commitments, as well as on crisis stability, alliances, deterrence and coercion, domestic politics and foreign policy, nuclear weapons, foreign troop deployments, and other critical areas of international relations.

This chapter provides a brief survey of the development of the credible commitment puzzle – that threats between states might not be believable – and the emergence of tying hands as a solution. It then summarizes the central theoretical and empirical arguments of the book, before concluding with a road map for the chapters to come.

America and the Foreign Policy Problem of Credible Commitments

Across its first 170 years, the United States mostly enjoyed detachment from world affairs outside of the Americas. The Atlantic and Pacific Oceans were tremendous moats offering continental safety, and the US saw little need to establish global interests. It had essentially no alliances and few global territorial possessions, and rarely even asked the question, how can we make credible international commitments?

World War II woke America to the unavoidable reality that the vast oceans could no longer protect the homeland. New military technologies including long-range bombers, missiles, and atomic weapons dissolved America’s sense of removed invulnerability. World War II also demonstrated that wars were now conflicts of complete economic mobilization, and going forward the balance of power and the outcomes of wars would be determined by who controlled global military–industrial capability. George Kennan in the late 1940s argued that the five centers of military–industrial power around the world were the US, Great Britain, Germany and central Europe, the Soviet Union, and Japan, and war-winning power was determined fundamentally by how many of these centers a superpower controlled.Footnote ² Others described more expansive needs, that the US also required access to areas in Latin America, Africa, and Asia to procure the raw materials necessary for modern economic and military power.Footnote ³

Sharpening America’s sense of vulnerability was the perception of the global Communist threat emanating from Moscow. America now faced a great power that appeared bent on expanding its global reach. Because Communist control of the sources of world power would pose a direct threat to American national security, Washington recognized that the US must contain Soviet expansion, as Kennan famously laid out in his 1946 “Long Telegram.” Events in the late 1940s reinforced belief in the growing Soviet and Communist threat, including the solidification of Soviet control of Eastern Europe, the emergence of Communist movements around the world, the 1948 Berlin Crisis, the 1949 Soviet nuclear test, the 1949 Chinese Communist revolution, and the 1950 North Korean invasion of South Korea.

The central question for US foreign policymakers during these and the following years was, how can the US deter Communist aggression? A straightforward solution would be to maintain a World War II–sized military, to deter attacks on the American homeland and foreign areas important to the US. However, the American public was not willing to maintain that scale of mobilization after 1945, especially one large enough to confront equally the massive Soviet and Communist Chinese armies. Indeed, President Eisenhower’s New Look foreign policy in the 1950s recognized that extensive economic mobilization was unpopular and would require the undesired expansion of the American state and the subversion of American democracy.Footnote ⁴

Another possible solution was collective security, to build a universal organization of states that would make war illegal, and resolve conflicts between states without violence. However, deep skepticism of collective security, grounded in the 1930s failure of the League of Nations to prevent World War II, became wired into the structure of the newly formed United Nations. Neither the Secretary General, International Court of Justice, nor General Assembly of the UN had much actual power, and the veto power of the five permanent members of the Security Council ensured frequent deadlock in that body. The absence of effective collective security meant that states could not be forced to comply with treaty commitments or other promises. Consequently, others could not be certain that they would follow through. Without enforcement, commitments lacked credibility.Footnote ⁵

Beyond the absence of effective world government, there were two other reasons why observers feared that international commitments might not be credible. The first was democracy. Kennan and others blamed the outbreak of World War II on democracy itself. For them, the enduring isolationism of the American public prevented President Franklin Roosevelt from taking necessary steps in the 1930s to balance against rising fascist aggression. They also blamed democracy-empowered pacifism of the 1930s for pushing Britain and France to cave to Benito Mussolini and Adolf Hitler. After 1945, they feared that democratic publics might similarly balk at enduring the human and financial costs of fighting new wars, encouraging American presidents and other elected leaders to abandon international commitments rather than go to war to honor them.

A second factor undermining the credibility of commitments to go to war was the nuclear revolution. The atomic bombings of Japan demonstrated that humanity had developed a weapon that could easily destroy civilization. The scale of horror expanded after 1945, as the American nuclear arsenal continued to grow, the Soviet Union acquired nuclear weapons much earlier than expected, and both sides in the 1950s developed hydrogen bombs, each of which was hundreds of times more powerful than the atomic (fission) bombs used against Japan. As ghastly as World War II was, a nuclear World War III would be much worse, slaughtering hundreds of millions instead of tens of millions, in days rather than years in an annihilation that no country could escape. A national leader, especially an elected leader such as an American president, might very well abandon a commitment and accede to foreign aggression, rather than go to war to abide by a commitment and pay the awful price of nuclear suicide.

This was the fundamental dilemma America faced after 1945. The US saw it as essential to prevent Communist aggression, to protect growing American national security interests and maintain the global balance of power. The best way to do this would be to threaten war in response to aggression. Yet no world government could enforce American commitments, the US was politically unable to support massive standing armies abroad, American voters might reject a militant internationalist foreign policy, and the prospect of nuclear suicide would provide any president with more than a little pause in complying with a commitment to go to war to fight Communist aggression abroad. It was this dilemma that became the focus of analysis of American foreign policy and international relations starting in the late 1940s, continuing to the present day. Where might a solution to this dilemma be found? How might foreign policy commitments be made credible, to maintain peace and keep adversaries at bay?

Tying Hands to Make Commitments More Credible

During the first several years of the Cold War, a set of solutions to the credible commitment problem appeared in the United States. A conceptual structure for thinking about this problem had begun to emerge in the so-called dismal science of economics. The 1940s and early 1950s saw the beginnings of the economics revolution of game theory, a new development appearing first in mathematics. This corner of economics sought to understand how individuals and firms make economic choices, and more generally how actors behave strategically, making choices in anticipation of the actions of others. Game theory’s fundamentals lent themselves well toward tackling the credible commitments problem. Game theory sought to understand actors who were looking out for their own interests, when other actors’ plans could only be guessed at and whose verbal declarations could not be accepted at face value, and when actors’ choices could not be controlled by a higher authority, all elements of the international relations credible commitment problem. Game theory’s origins lay in trying to understand games like poker, and the bluffing intrinsic to poker lies at the heart of the credible commitments problem.

In the 1950s and 1960s, Schelling did more than any other individual to export game theory to the study of foreign policy and international relations.Footnote ⁶ After working for the Marshall Plan in the late 1940s, Schelling received his PhD in economics from Harvard in 1951 and began to examine how game theory might inform economics, politics, and international relations. His first major work was a landmark 1956 article in the American Economic Review.Footnote ⁷ This essay was an early effort at interdisciplinary thinking, drawing ideas from one field of study to illuminate another one, in this case applying ideas from economics (and math) to international relations.

Schelling’s starting point was the economic concept of bargaining, of two actors negotiating a price one actor will pay another in exchange for a good or service. One of the key insights from the bargaining perspective connects exactly to the problem of credible commitments. Economic actors have an incentive to lie, to not follow through on verbal commitments made during bargaining. Indeed, lying is an essential part of bargaining, as bargaining power is essentially about manipulating the other actor’s beliefs in an environment of uncertainty, that it is “the ability to fool the other person … to set the best price for yourself and fool the other [person] into thinking this was your maximum offer.”Footnote ⁸ This incentive to bluff is a commitment problem. If a buyer could credibly commit that a certain price was indeed her maximum offer, then bluffing would be impossible. However, such a commitment is not credible, and the seller might suspect that this is a bluff and the buyer might be willing to pay more.

Schelling connected this economics perspective to American foreign policy. He evaluated the Eisenhower administration foreign policy of “massive retaliation,” announced by Secretary of State John Foster Dulles in 1954, that America would launch a general nuclear war against the Soviet Union in reaction to even minor acts of aggression against American allies.Footnote ⁹ Schelling echoed the concerns of some that the massive retaliation policy faced problems of credibility, that such a policy amounted to America forcing itself to choose between suicide (executing the threat, and inviting Soviet nuclear retaliation against the US) and surrender (failing to execute the threat, and abandoning the ally). If America’s foes concluded that it might prefer surrender of an ally’s territory to suicide, then the commitment to go to war on behalf of an embattled ally would lose credibility.

For Schelling, a conceptual solution to the economic bargaining credible commitment problem could also be used by foreign policy decision-makers. The actor making the commitment to go to war could tie its hands, by making it costly or impossible for that actor to avoid carrying out the threat. He wrote, “It is essential, therefore, for maximum credibility to leave as little room as possible for judgment or discretion in carrying out the threat.”Footnote ¹⁰

To be sure, the idea of tying hands was not in the 1950s new to international relations. As far back as the fifth century BCE, ancient Greek city-states would exchange oaths as a means of making promises more credible, given the perceived social, diplomatic, and perhaps divine costs of breaking an oath. Later that century, during the Peloponnesian Wars, oaths were sometimes not enough, as belligerents would exchange hostages to bolster the credibility of their commitments to a peace agreement, as violating the agreement would cause the adversary to kill the hostages.

Schelling fleshed out these initial ideas especially in his two classic books, Strategy of Conflict (1960) and Arms and Influence (1966). He developed a conceptual foundation to help understand tying hands, describing what he saw the US and other states as doing to make credible commitments, and placing that description within a broader framework. Schelling proved to be especially influential in advancing these ideas because he wrote so clearly and engagingly, weaving together in beautiful prose economic theory, history, contemporary policy concerns, and even the everyday, children manipulating parents the way states manipulate each other.

Schelling described several tools which could be used to tie one’s own hands, to make incredible foreign policy commitments more credible by making it logistically impossible or prohibitively costly for a state to violate its commitments. Schelling saw these tools as being commonly used by the US and in international relations more broadly. This book focuses on five of these tying hands tools that have received extensive scholarly and policy discussion over time as potential solutions to the credible commitment problem.

The first tying hands tool is brinkmanship, creating risks of inadvertent escalation to war. It would not be credible to threaten to start a nuclear war to get the other side to back down and make concessions, as it would never be rational to commit suicide to accomplish virtually any foreign policy goal. However, it might be possible to create a possibility of inadvertent escalation to war, by for example mobilizing forces to forward areas near the adversary. That would ameliorate the credibility problem, because escalation and implementation of the commitment could happen accidentally, out of the direct control of the national leadership making the threat, no longer requiring one side to choose suicide for the threat to be implemented. The fear of the possibility of inadvertent escalation might in turn be enough to cause the other side to back down. Schelling used his conception of brinkmanship to frame his understanding of the Taiwan Straits, Berlin, and Cuban Missile Crises.

The second set of tools includes alliance treaties and verbal commitments to use force. If a state signs a formal treaty or makes a verbal statement committing itself to defend an ally, then there are political costs, or audience costs, of not rescuing an embattled ally, potentially both domestically and internationally, in terms of damaged reputation. These costs are even higher if the state has a network of alliances or commitments, meaning that failure to protect one ally might damage a state’s reputation across a number of commitments. This was the lens through which Schelling saw America’s extensive post-1945 alliance network.

The third is deployment of small contingents of forces abroad, to serve as “tripwires.” These contingents would be inadequate to shift the local balance of power and defeat or even slow down an attacker, but their deaths would incur national honor and reputation, raise the costs of failing to aid those fallen troops, and make broader intervention more likely. Anticipation that American deaths would require broader American military intervention would make the American commitment to defend its allies more credible, deterring attack. Schelling was influenced by the apparent ability of a small contingent of deployed US troops to deter a Soviet attack on West Berlin.

The fourth tool is the rationality of appearing irrational, or “mad leader theory.” The entire logical structure of the bargaining approach rests on the assumption that when two actors bargain, each is rational and views the other as rational, and this is what makes suicidal threats not credible. But what if one leader, including a sane leader attempting to feign madness, could convince the adversary that she actually was irrational, that she would be willing to execute an apparently suicidal threat? Under these conditions, suicidal threats might actually become credible. Here, Schelling was speaking more hypothetically, but this idea gained more salience in the years to come, especially in the Nixon and Trump administrations, and in American images of foreign leaders such as Muammar Qaddafi, Saddam Hussein, and Kim Jong Un.

The fifth tool is making commitments automatic. Commitments are not credible because a normal person would not take an action that would lead to her own death, nor perhaps the mass genocide of enemy civilians. But what if these human frailties could be circumvented, what if the decision to execute a threat was taken out of human hands and given to a machine? Perhaps that might make a commitment more credible. These ideas entered the cultural mainstream with films such as Dr. Strangelove (parodied in the graphic illustration after the title page) and the 1983 thriller WarGames.

Why Examine Tying Hands?

It would be difficult to exaggerate the significance that the tying hands insight has had for the study of international relations and political science more broadly. The post-1945 study of international relations marked a giant turn away from the interwar legalistic orientation and toward a fundamentally economic perspective, embracing two assumptions: states and other political actors look after their own interests, and there is no international actor that can control state behavior. When self-interested actors are not punished for lying or breaking commitments, the core questions become: How do states interact? How do they communicate with each other? Why not lie all the time? How can states tell if others are not lying?

Schelling’s framework for addressing these questions along with his other theoretical contributions made him perhaps the most influential international relations scholar of the post-1945 era, and the only international relations scholar to win a Nobel Prize (in economics). No other scholar was as important in moving game theory from economics to international relations, alongside the seminal efforts to introduce game theory to American politics by Anthony Downs, and to comparative politics by Kenneth Arrow and William Riker. Schelling demonstrated how abstract concepts like bargaining and noncooperative games could be applied to very real international relations problems. His ideas laid the groundwork for signaling games, perhaps the most important theoretical models in international relations.

Tying hands insights have infiltrated and dominated the three biggest post-1945 debates in international relations: Why do states fight each other, how do domestic politics affect international relations, and why do international institutions matter? Regarding the first, making incredible threats credible was the essential focus of rational deterrence theory, rational deterrence in turn being for decades the central theory explaining the causes of war and peace between states. Scholars developed signaling games, using tying hands tools such as brinkmanship, to explain how states attempt to communicate credibly with each other, to try to avoid war while protecting their interests. Others used tying hands logic to help explain why alliance agreements are not mere “scraps of paper,” and why alliances help deter aggression.Footnote ¹¹ Though scholars allowed for other means of making commitments more credible, including sinking costs (see below), many emphasized the significance of tying hands. James Fearon, perhaps the most influential international relations scholar since 1990, wrote: “Leaders should generally prefer tying hands to sinking costs when the former is possible, despite the fact that doing so tends to ‘lock in’ the leader and creates greater risks of war.”Footnote ¹²

Tying hands also greatly informed the field’s understanding of the connections between domestic and international politics. In the 1970s, 1980s, and early 1990s, empirical studies demonstrated that democracies were unlikely to fight each other, the so-called democratic peace, but did fight other kinds of regimes such as dictatorships. However, the theoretical explanations for these patterns were unsatisfying, as simple democratic pacifism wouldn’t explain why democracies fight nondemocracies.

In 1994, Fearon published an article that provided two gigantically important and integrated scholarly insights into the democratic peace puzzle and the escalation of crises to war, building on the tying hands framework. He presented a model of crises and war that advanced the signaling games of the previous decade, demonstrating that states can signal more credibly if they tie their hands, because they suffer the costs of backing down from a commitment to use force. He presented the term “audience costs,” the political costs a leader would pay for not following through on threats. He then applied the model to the democratic peace, showing that because democracies pay higher costs for not following through on a threat, this enables them to make more credible threats, communicate more efficiently with each other, and avoid conflict with each other though not with other kinds of regimes.Footnote ¹³ This tying hands insight as applied to domestic politics and commitment credibility, foreshadowed in Schelling’s 1956 article,Footnote ¹⁴ pushed forward massive waves of research over the decades that followed. The term audience costs, a manifestation of the tying hands argument, engulfed the field of international relations, and spread to other subfields of political science as well.

The concept of tying hands also became critical in the modern understanding of why international institutions matter. The interwar legalistic vision of international institutions, proposing that states comply with institutions simply because they are the law, was discredited by the collapse of the League of Nations. Robert Keohane’s landmark 1984 book After Hegemony rebooted interest in international institutions, providing a rationalist, economic perspective. The central proposition was that international institutions can foster cooperation between self-interested states even under international anarchy. One critical way they do this is tying states’ hands by creating the prospect of damaged state reputations if states violate the institution’s provisions. Since then, there has been a mountain of research on the topic of how international institutions exert influence by tying hands, in areas such as international lending, human rights, and others.Footnote ¹⁵

These three main areas are only some of the ways in which the tying hands insight infiltrated and engulfed international relations and political science more broadly. Our understanding of the emergence and effectiveness of the modern European state, in which seventeenth-century governments created domestic institutions to make their commitments to self-restraint credible, is essentially a tying hands story.Footnote ¹⁶ Indeed, the modern vision of all political institutions, domestic and international, is fundamentally about tying hands, about writing rules describing expected behavior, with those rules tying hands by making it impossible or costly to break them. The contemporary understanding of authoritarian politics focuses on efforts authoritarian leaders take to tie their hands, as autocrats craft power-sharing deals with potential regime challengers that make it costly for autocrats to break the terms of the deal, in turn dissuading challenges. Indeed, one of the leading 2010s books on Chinese politics was titled, Tying the Autocrat’s Hands.Footnote ¹⁷ Our modern understanding of the array of democratic advantages, including why democracies can more easily secure financial credit and enjoy higher levels of cooperation, stresses the ability of democracies to tie hands more effectively than dictatorships.Footnote ¹⁸

Tying hands has been applied to many other areas of political science, as well. The deeply influential 1988 theory of “win-sets” linking domestic politics to international negotiating drew directly on Schelling’s tying hands insight, and the long string of research building on the win-set insight routinely framed it as a tying hands argument.Footnote ¹⁹ Fixed exchange rate monetary policies are framed as hands-tying.Footnote ²⁰ Tying hands inform how we think about tools to resolve violent conflicts between and within states, such as belligerents agreeing to binding arbitration or peacekeepers raising the audience costs of breaking a commitment to a peace deal.Footnote ²¹ Tying hands have been used to explain how governments can build public support for policies fighting climate change, use branches of government such as judiciaries to check each other within separation of powers systems, issue broadcast licenses, and self-impose fiscal discipline through balanced budget laws, to name a few.Footnote ²² Put simply, how to make credible commitments is one of the most fundamental puzzles in international relations and indeed all of political science, and tying hands has been and continues to be one of the most salient and fertile theoretical solutions to this puzzle.

Outside of academic scholarship, hands-tying tools and ideas remain essential in contemporary foreign policy debates. Into the 2020s, the term “brinkmanship” is frequently tossed around in media and government circles, usually suggesting that a particular leader is engaging in actions creating risks of inadvertent escalation to war in order to push the other side to back down. Russian President Vladimir Putin has been called the czar of brinkmanship, and his tactics in the winter 2021–2022 crisis preceding the February 2022 Russian invasion of Ukraine were described by leaders and the media as brinkmanship. Chinese leader Xi Jinping’s actions have been explained as brinkmanship, regarding Taiwan and trade policy.Footnote ²³ US and Chinese actions especially regarding the South China Sea are frequently framed as brinkmanship,Footnote ²⁴ as are US–North Korean interactions.Footnote ²⁵

Many continue to embrace one of the core assumptions of brinkmanship, that crises create real risks of inadvertent escalation to war. Any time a crisis arises involving a nuclear-armed state, a journalist will shriek about the imminence of inadvertent escalation to nuclear war and describe that crisis as the most dangerous episode since the Cuban Missile Crisis, whether it is the 1987–1988 Brasstacks crisis between India and Pakistan, the 2017 US–North Korea crisis, the 2022 Ukraine War, or some other flare-up. As the tying hands brinkmanship tool logic presumes, crises are seen as powder kegs set to explode at the merest spark. Typical is New York Times columnist Thomas Friedman’s 2023 grim warning that, “The smallest misstep by either [Taiwan or China] could ignite a U.S.-China war that would make Ukraine look like a neighborhood dustup.”Footnote ²⁶

The other hands-tying tools continue to dominate policy debates and discussions, as well. Contemporary leaders such as Donald Trump and Kim Jong-Un have cultivated images of madness to acquire coercive bargaining leverage. After Russia’s February 2022 invasion of Ukraine, the American intelligence community debated whether Putin might be pretending to be mad, or actually be mad. Militaries are beginning to deploy autonomous weapons systems, and debates continue as to whether computers should be allowed to initiate the use of force, without a human entering the decision loop, perhaps making use of artificial intelligence.

Regarding alliances, policy debates continue to be infused with tying hands perspectives. Republicans critique the withdrawal from Afghanistan in 2021 as damaging American reputation, leading directly to the February 2022 Russian invasion of Ukraine, an argument linked to the “interdependence of commitments” perspective that underlies the assumption that alliance agreements tie hands. Discussion of US–Taiwan relations centers on the degree to which the US has tied its hands with alliance treaties and public statements. Debates about the admissions of Finland, Sweden, and Ukraine to NATO following Russia’s February 2022 invasion of Ukraine have centered around to what degree admission would tie the hands of NATO members, forcing them to intervene in future wars.

There are also ongoing debates about tripwire troop deployments. On one hand, the isolationism of Donald Trump focused on a desire to reduce the costs of foreign troop deployments, perhaps in the direction of cutting back troop contingents to tripwire numbers. On the other hand, growing belligerence from Russia and China has renewed a sense of the need to deter attacks on a range of territorial areas, including the Senkaku/Diaoyu islands, Taiwan, and NATO members, and debates have focused on whether or not deployment of small contingents of forces are sufficient to deter aggression through tripwire effects.Footnote ²⁷

This Book’s Central Argument

Tying hands means a tradeoff, of making a commitment more credible in exchange for reducing the flexibility to stay out of an undesired war, the wrong war. This book proposes that states, leaders, and societies are averse to accepting this tradeoff, they prefer to preserve the flexibility to avoid war rather than make their commitments more credible by tying hands. This is because the costs and risks of war can be very high, especially under the wrong political and military conditions. Tying hands means reducing the ability of leaders and states to keep themselves out of the wrong war, and because the stakes are so high, they prefer to retain flexibility, even at the cost of forgoing an opportunity to make a commitment more credible. Put differently, states are deeply averse to playing Russian Roulette, to giving up their ability to prevent escalation to national or even global destruction, as illustrated in the graphic comic at the front of this book. Leaders and states will tie hands only under narrow circumstances, when they are willing to reduce flexibility to bolster commitment credibility. Further, even when they do tie hands, they tie them as little as possible, endeavoring to make the threat just credible enough to work while still maintaining flexibility to perhaps permit escape from the commitment if necessary. Leaders and states are more likely to attempt to bolster threat effectiveness by “sinking costs,” costly actions states take before the commitment is challenged, than by tying hands. Sinking costs are more attractive than tying hands because they bolster the credibility of the commitment without reducing flexibility to avoid the wrong war.

This argument may seem striking to a reader well-versed in international relations scholarship. So much theoretical and empirical scholarship is devoted to exploring tying hands, so how can it be that states are not actually tying hands? The empirical argument here would seem to be flawed, at its face.

This book demonstrates that several hands-tying tools are never actually implemented, meaning, states do not really tie hands, appearances notwithstanding. Though states sometimes make threats, mobilize troops, and deploy forces close to the adversary, these actions do not create risks of inadvertent escalation to war. They do not automate commitments, and they do not install in power truly mad leaders. They sign alliance treaties, but these treaties do not tie hands because they always include flexible language that allows signatories to stay out of wars involving allies without violating the treaty. Regarding the other tying hands tools, feigning madness and deploying small troop contingents, states sometimes engage in these actions, but they do so knowing that these actions do not truly tie their hands. That is, apparent hands-tying is not actual hands-tying.

There are two arguments this book does not make. The book does not argue that all tying hands tools cannot work. To the contrary, it is because at least some tying hands tools if implemented properly would make reneging on a commitment impossible or costly that states and leaders avoid using them. That said, some tying hands strategies are unlikely to work because of credibility problems, as states are unlikely to believe another leader’s claim that they are mad, and they are often unlikely to believe that the deaths of small contingents of troops could trigger a revenge effect.

Additionally, though the book proposes that states and leaders are averse to tying hands, it is not making the absolutist argument that states and leaders never tie hands. It proposes that when states and leaders elect to tie hands they do so only under narrow circumstances, and when they do tie hands they proceed in a measured way, tying their hands no more than they believe to be necessary, to afford themselves as much flexibility as possible to help them stay out of the wrong war. The book is also not arguing that political actors necessarily choose flexibility over tied hands in areas outside of commitments to war, such as in drafting political constitutions or writing criminal sentencing guidelines. In any area in which an actor has the choice to tie hands, it will weigh the benefit of a more credible commitment versus the advantages of flexibility. As this balance tilts toward commitment credibility over flexibility, actors will be more willing to tie hands. Because the risks of inflexibility are so high in decisions to commit to war, we should expect that that is an area in which we are unlikely to see actors tie hands.

This book’s argument provides a new lens with which to understand several important areas of foreign policy. Acts of alleged brinkmanship never escalate inadvertently to war because states are not creating actual risks of inadvertent escalation. So-called mad leaders never seem to take intentionally suicidal foreign policy actions, because they are not actually mad; further, observers usually slough off attempts at feigned madness. States understand the deep folly in handing over to computers the ability to launch war. States are deeply hesitant to expand alliance commitments unless they understand that those commitments contain loophole language that provide the flexibility to stay out of wars involving embattled allies. When states deploy small troop contingents abroad, they are not establishing hands-tying tripwires, but are motivated by other foreign policy goals and pursuing other strategies.

Empirical Approach

The structure in this book is a bit different from that in some other books. Some books develop a hypothesis from a theory, the hypothesis then proposing that movement in some independent variable causes change in some dependent variable. Empirical analysis engages in hypothesis-testing, assessing if the available evidence demonstrates that changes in the independent variable cause the dependent variable to change.

This book makes a simpler claim, that across-the-board states and their leaders prefer not to tie hands when they make commitments to go to war. That is, the claim is not so much declaring that certain conditions make states less likely to tie hands, but more comprehensively that states and their leaders usually do not tie their hands.

The book’s primary empirical claims, discussed in chapters to follow, include:

• – inadvertent escalation to war almost never happens, especially after 1945 (Chapter 3)

• – states never deliberately engage in brinkmanship, intentionally creating risks of inadvertent escalation for coercive advantage (Chapters 3 and 4)

• – fear of inadvertent escalation, the desired product of a brinkmanship strategy, has almost never translated into coercion success (Chapter 4)

• – all post-1945 alliances include flexible language, allowing signatories leeway in interpreting whether nonintervention in the event of war would violate the alliance (Chapter 5)

• – there are no clear violations of alliance treaties after 1945, accounting for flexibility language (Chapter 5)

• – states do not deploy small contingents of troops abroad intending to tie their hands (Chapter 7)

• – mad leaders do not come to power, and feigning madness does not tie hands (Chapter 8)

• – states do not attempt to make commitments more credible by giving computers the ability to start a war (Chapter 8).

The empirical focus is on the post-1945 time period. This period is most appropriate because the advent of nuclear weapons makes the credible commitment problem even more acute, especially for deterring attacks on other states. Brinkmanship is a much more practicable strategy in the nuclear age. Mad leader tactics are most appropriate in the nuclear age, because of the potentially suicidal costs of war. There are other developments that make the pre-1945 time period qualitatively different from the post-1945 time period with regard to studying tying hands. Advances in aerial reconnaissance and electronic communications after 1945 substantially reduced the likelihood of inadvertent escalation, meaning that the likelihood of inadvertent escalation (and the effectiveness of brinkmanship strategy) in the pre-1945 age does not compare closely with the post-1945 age. Making commitments to go to war automatic is only practicable in the computer age. Deploying tripwire troops makes the most sense in the post-1945 era of global alliance commitments. Regarding alliances, international audience costs and the interdependence of commitments applies more heavily after 1945, as only in this era did the major powers sign alliance treaties with dozens of other countries.

The book also presents extensive qualitative empirical analysis, with a heavy focus on Cold War crises. Cold War crises are essential for assessing the tying hands view of international relations, as these crises have been used to both develop and test tying hands theory. We need to ask, are the tying hands historical interpretations of this era accurate? Another reason to focus on Cold War events is that these are now very well-documented episodes, thanks to the release of primary documents from American, Soviet, Chinese, and other archives. We can now assess relatively carefully what states and leaders were doing during these events, and why they were doing them, essential for testing the theory.

This qualitative analysis offers several perspectives about Cold War events, some in striking contrast to conventional interpretations. They include:

• – All of the major Cold War crises, including the Cuban Missile Crisis and Berlin Crises, were much less dangerous than is commonly believed.

• – The Soviet Union did not attempt to create risks of inadvertent escalation in the 1948 Berlin Blockade Crisis by flying its planes dangerously close to Western planes.

• – Brinkmanship did not prevent China from invading Taiwan in 1954 or 1958.

• – The US and the Soviet Union during the Cold War designed their alliances to be flexible rather than binding.

• – Neither Truman, Eisenhower, or Kennedy sought to tie hands through tripwire effects by deploying small contingents of American troops to Europe.

• – American decision-makers mostly did not believe that alliances tied their hands tightly, requiring them to intervene on the behalf of an embattled ally; Johnson’s belief in 1964 that an alliance tied America’s hands to defend South Vietnam was the exception rather than the rule.

• – Had John F. Kennedy lived, he likely would have withdrawn rather than escalated the American troop presence in South Vietnam, despite America’s alliance ties with South Vietnam.

• – Attempts to feign madness by Mao, Khrushchev, and Nixon all failed, and did not tie hands.

• – The Soviet Perimeter system did not give computers the ability to start nuclear war.

In sum, the evidence points to substantial reinterpretations of the study and practice of international relations. When we observe states engaging in apparently provocative moves during international crises, they are not seeking to create risks of escalation to war, but are serving other goals, such as raising short-term readiness for war. International crises carry trivial risks of inadvertent escalation to war. Attempts to feign madness are routinely dismissed as just so much theater. Alliances do matter, not so much because they tie hands by risking reputation, but rather because they permit peacetime military coordination, boosting aggregate military power. States do not tie hands with troop deployments.

Outline of the Book

The chapters to follow present this book’s general argument, and then apply the argument to the five tying hands tools. Chapter 2 develops the general theoretical argument of the book, that tying hands means sacrificing flexibility to avoid the wrong war to gain commitment credibility. Leaders, states, and societies prefer to retain that flexibility, and as a result very rarely tie their hands, because they do not want to lose the flexibility to avoid fighting the wrong war.

Chapters 3 and 4 examine brinkmanship. Chapter 3 demonstrates that states do not create risks of inadvertent escalation to war. Crises are actually much less dangerous than is commonly thought, and crises escalate to war when one side wants war to happen, but not out of accident or inadvertence.

Some might propose that even if inadvertent escalation isn’t actually happening, states might be creating real risks of inadvertent escalation and thereby tying hands, but escalation doesn’t occur because the targets of these moves are making concessions before inadvertent escalation has the opportunity to occur. This would imply that brinkmanship is occurring and succeeding. Chapter 4 explores this speculation. It examines two data sets of all attempts at coercion involving nuclear-armed states, asking, are there any examples of one side achieving coercion success through brinkmanship? With the partial exception of the Cuban Missile Crisis there are not, providing further evidence that states are not creating actual risks of inadvertent escalation during crises.

Chapter 5 examines alliances and to some degree verbal statements as potential tying hands tools. This chapter proposes that no alliance treaties since 1945 tie hands, as they all give signatories the ability to stay out of wars with reduced risk of incurring audience costs. The treaties do this by including flexible language which gives signatories the ability to argue that nonintervention in the event of war does not violate the treaty. However, on rare occasions signatories do wish to tie hands more tightly to make their commitments to defend their allies more credible. They do this by making verbal statements clarifying their understandings of their alliance commitments. However, when states do this, they tighten their hands very carefully and often minimally, trying to leave themselves as much flexibility as possible.

Chapter 6 applies these theoretical insights to three of America’s most important Cold War alliances, NATO, the US–Taiwan alliance, and the Manila Pact, showing that they all demonstrate these dynamics. In particular, the theory sheds new light on American alliance commitments and America’s decision to intervene in the Vietnam War, illustrating that President Johnson understood that the Manila Pact was flexible and provided many avenues to justify American noninvolvement in the Vietnam War, and Johnson and other presidents frequently used this flexibility to avoid involvement in several conflicts involving states covered by the Manila Pact. American involvement in the Vietnam War under Johnson then becomes an intriguing outlier rather than demonstrative that America tied its hands with alliances during the Cold War.

Chapter 7 explores tripwire deployments. The chapter argues that states do not dispatch small contingents of troops abroad to tie hands. It demonstrates that neither Truman, Eisenhower, or Kennedy deployed troops to Europe in order to tie their hands.

Chapter 8 examines the mad leader strategy, and the approach of making commitments automatic. It first explains the logic of the two variants of mad leader strategy, of an actually mad person becoming national leader, and a sane leader feigning madness to make commitments more credible. It argues that mad individuals do not become national leaders, and that sane leaders cannot credibly signal madness and achieve bargaining leverage. The chapter shows that there are no instances of suicidally mad leaders coming to power, or of a leader feigning madness and achieving coercion success by conveying madness. The chapter then evaluates the possibility of making commitments automatic by giving a computer the authority to implement a commitment by launching a war, including a nuclear war. Though richly described in fiction, states do not actually do this, they do not give a computer the autonomy to commit mass genocide or suicide to make commitments more credible.

Chapter 9 concludes, drawing out implications of the book’s argument and findings for international relations theory, applying the book’s argument to the 2022 Ukraine War, and discussing why the tying hands proposition has endured over time.