I recently read the 2020 book “The Myth of the Nuclear Revolution: Power Politics in the Atomic Age” by Keir A. Lieber and Daryl G. Press. It is the most insightful nuclear security book I have read to date, and while I disagree with some of the book’s outlook and conclusions, it is interesting and well-written. The book is also very accessible and fairly short (180 pages). In sum, I believe more people interested in nuclear security would benefit from reading the book.
In this post, I share some of my copied-and-pasted notes from the book that were of particular interest to me. These notes cannot do justice to the book's extended discussion of the covered issues. Still, I hope some people may find my notes valuable and perhaps become inspired to read the book.
To let you decide whether to keep on reading, here is a summary of the book’s core argument (taken from here):
In The Myth of the Nuclear Revolution, Keir A. Lieber and Daryl G. Press tackle the central puzzle of the nuclear age: the persistence of intense geopolitical competition in the shadow of nuclear weapons. They explain why the Cold War superpowers raced so feverishly against each other; why the creation of "mutual assured destruction" does not ensure peace; and why the rapid technological changes of the 21st century will weaken deterrence in critical hotspots around the world.
By explaining how the nuclear revolution falls short, Lieber and Press discover answers to the most pressing questions about deterrence in the coming decades: how much capability is required for a reliable nuclear deterrent, how conventional conflicts may become nuclear wars, and how great care is required now to prevent new technology from ushering in an age of nuclear instability.
Throughout this post, my highlights are in bold.
Introduction: The Nuclear Puzzle
Nuclear peace
The link between nuclear weapons and peace among the great powers is not mere coincidence. (...)
First, nuclear weapons have dramatically raised the potential costs of war. As terrible as conflict was in the prenuclear era, the devastation from nuclear war could be far worse. (...)
Second, and more important, nuclear weapons appear to make victory in war impossible. Throughout human history the costs of war—though often staggeringly high—were borne disproportionately by the losing side. Leaders plunged their societies into terrible conflicts because if they won, war could pay. But in the shadow of nuclear weapons, both aggressors and victims face the same fate: devastation. By making war both enormously costly and seemingly futile, nuclear weapons have greatly reduced the incentives for major conflicts.
Nuclear stalemate
Stalemate exists when neither adversary in a war can escape destruction and emerge victorious in any practical sense of the term. If nuclear weapons make victory in war impossible, then such wars cannot serve a rational purpose. (...) Strategic stalemate should not only deter war but also allows countries to abandon a wide range of costly and risky policies designed to increase their security.
The nuclear puzzle
[Nuclear] weapons are the most effective instrument of deterrence ever created. But therein lies the key puzzle of the nuclear era. If these weapons are so good at preventing war, (...) why is there so much geopolitical competition among the major powers? If nuclear-armed countries are fundamentally secure from attack, why don’t they act accordingly? (...) major power rivals have continued to compete for strategic advantage by building and improving their nuclear and conventional military forces, enlarging alliances, expanding into new strategic territory, and all the while eyeing warily any changes in the balance of power.
Explaining the nuclear puzzle
We argue that the discrepancy between theory and decades of reality has a simpler explanation: the theory is wrong. We argue that nuclear weapons are the most effective instruments of deterrence ever created, but the logic linking nuclear stalemate to pacified international politics is flawed. (...) [I]ntense security competition persists in the shadow of nuclear weapons for at least three reasons:
[First,] Creating stalemate among nuclear-armed adversaries is not as simple as many scholars have assumed. Building nuclear weapons is relatively easy, if one can acquire the necessary materials, but building an arsenal that can reliably survive an enemy, disarming attack and be used in retaliation is arduous work (...)
[Second,] Nuclear stalemate does not lie at the end of a one-way street. Even countries that build survivable arsenals need to keep working, since they will be afraid that an enemy might escape stalemate. They may also be tempted to try to escape stalemate themselves, by building their own war-winning options. (...)
[Third,] Nuclear stalemate does not automatically solve a country’s deterrence requirements. If countries need to use nuclear threats to deter conventional attacks (not just nuclear ones), then merely building a survivable nuclear arsenal may not be sufficient. Those countries will feel compelled to build large, flexible, and resilient nuclear forces to make the threat of escalation credible.
Chapter 1: Power Politics in the Nuclear Age
Logic of nuclear deterrence
Although the immense destructiveness of nuclear weapons undoubtedly gives leaders pause in a way no military innovation ever has, the fundamental difference between nuclear weapons and all other instruments of war is rooted in their propensity to cause stalemate. (...) Nuclear weapons are uniquely deterring because they appear to make victory in war impossible. They are the ultimate tools of stalemate. In the past, war was hell—but primarily for the defeated. (...)
Leaders in the prenuclear world could rationally launch wars, even when they knew the consequences of losing would be terrible, because they had a chance to win, and thus might avoid the worst of war’s consequences. What is unique about the nuclear age is not that war can now be devastating. Rather, it is that both sides—strong and weak, victor and vanquished—can be destroyed if war occurs. (...) Specifically, if both adversaries have deliverable nuclear weapons, even stunning battlefield successes would still leave the victor exposed to devastating retaliation.
Causes of nuclear stalemate
The stalemating qualities of nuclear weapons stem from three characteristics.
First, nuclear weapons are small. A modern nuclear bomb or missile warhead is only a few feet long. Even a warhead on a road mobile missile launcher is only the size of a tractor trailer. Because of their small size, nuclear weapons are difficult to locate and destroy with a disarming strike.
Second, nuclear weapons are unique in the amount of explosive power they pack per weapon. As a result, a disarming strike would need to destroy nearly every target to succeed. An attack that left the victim with “only” a dozen functioning weapons would leave him with the capability to inflict terrible damage in response.
Third, nuclear weapons are relatively easy to deliver. Because nuclear weapons are small and relatively lightweight, they can be delivered by a wide range of methods, including ballistic missiles, cruise missiles, aircraft, and artillery.
Nuclear peace
Most importantly, great power wars—the most destructive types of conflict—disappeared just after nuclear weapons were created. The timing of the disappearance is striking: the world suffered through at least five great power wars (roughly one every twenty years) during the century leading up to the nuclear age, but none in over seven decades since. The abrupt halt to major war is even more telling because the Cold War contained many catalysts for conflict: the superpowers had major ideological differences, held competing universalist visions for the world, demonized each other, and prepared intensely for conflict. Yet peace prevailed—or, to be more precise, major war was prevented. Nuclear deterrence deserves substantial credit for this outcome.
Anomaly: limited attacks on nuclear armed states
In 1969, Russian and Chinese ground forces fought over disputed territory, killing hundreds of soldiers. In 1982, Argentina seized the Falkland Islands, a British possession, triggering a significant naval clash in the South Atlantic. During the 1991 Persian Gulf War, Iraq launched forty-two conventionally armed Scud missiles at Israeli cities. (...) In each of those cases, the stakes for the nuclear-armed power were relatively modest, and hence the attacks were unlikely to trigger nuclear escalation.
Nuclear stalemate predictions: (1) Reduced relative gains concerns
One would expect the invention of nuclear weapons to greatly mitigate relative-gains concerns. If nuclear weapons make war unlikely and conquest nearly impossible, nuclear-armed adversaries no longer need to worry about shifts in the military or economic balance of power, because one side’s “gain” cannot undermine the fundamental security of the other. (...) The persistence of relative-gains concerns in the nuclear age is a puzzle.
Nuclear stalemate predictions: (2) Reduced arms racing
If nuclear weapons make those who possess them secure, then there is little reason for countries to engage in costly competitions for greater military capability. (...) Nuclear weapons should diminish the need for conventional arms races as well. If a nuclear deterrent provides the ultimate protection against defeat in war, then that arsenal should deter, whether the threat comes in the form of nuclear or conventional attack. (...) The history of the nuclear age does not support these predictions.
Nuclear stalemate predictions: (3) Reduced value of alliances
Before nuclear weapons, international alliances were crucial, as they allowed states that faced common threats to pool their resources. Although relying on allies entailed other risks (for example, the risks of entrapment or abandonment), confronting threats without allies was often far costlier and dangerous. With nuclear weapons, countries can seemingly ensure their own security without taking on commitments to others or depending on pledges of assistance from abroad. (...)
In reality, alliances appear to be just as relevant in the nuclear era as before. (...) The logic behind those alliances was clear. Conventional war was still possible, so alliances were still necessary for all the reasons they always were— to share costs, pool resources, and provide access to strategically crucial territory
Nuclear stalemate predictions: (4) Reduced competition for strategic territory
Groups have always competed for territory, to control it for their own advantage and deny it to their adversaries. Some lands were valued for what lay within: fertile soil or valuable minerals. Other territory was deemed “strategic” because it sat astride key transport routes, or because of its proximity to enemies or suitability for hosting military forces. By creating stalemate, nuclear weapons seem to greatly reduce the value of strategic territory, because the gain or loss of such territory (or resources tied to geography) has little effect on the balance of power. In an era in which nuclear weapons make countries fundamentally secure, countries no longer need to control the natural resources that permit them to sustain large military forces. (...)
The nuclear era does not resemble this predicted world. Much of the actual tension and conflict in the Cold War was about competition for control of territory.
Assumptions about nuclear stalemate
the ameliorating effect of nuclear weapons on international security competition depends on three assumptions about the nature of nuclear stalemate: (...)
[(1) stalemate is easy to produce:] if robust deterrence requires assured-retaliation capabilities, nuclear-armed states can create a condition of stable nuclear stalemate. But if stalemate requires robust, survivable forces, nuclear-armed states should be expected to engage in arms races, pay attention to relative gains, and maintain interest in allies (to share costs or provide cover during their period of nuclear vulnerability). (...)
[(2) stalemate is nearly irreversible:] survivability may be a two-way street. Perhaps investments in counterforce capabilities can pay off, and efforts to locate and destroy enemy nuclear forces can succeed. If so, arsenals that are safe from attack today may become vulnerable in the future. (...) [If] the nuclear balance can meaningfully shift, then competition in the nuclear age will resemble prenuclear geopolitics. (...)
[(3) stalemate effectively deters not just nuclear conflict but conventional war:] using nuclear threats to deter conventional war poses a conundrum: How can a victim of conventional attack credibly threaten to employ nuclear weapons if the attacker can respond in kind? That is, if escalating a conventional war will merely trigger a nuclear response, then rational nuclear-armed states would never do so. (...) an arsenal optimized for deterring conventional attacks must be highly survivable—not survivable only in peacetime, but still survivable after extended periods of conventional conflict have degraded nuclear forces, communications, and command and control.
Chapter 2: Getting to Stalemate: How Much Is Enough?
Effects of early Soviet nuclear arsenals
First, even a small and vulnerable Soviet nuclear arsenal likely bolstered peacetime deterrence during the first decade of the Cold War. Even when the Soviet nuclear force was highly vulnerable to a U.S. disarming strike during the 1950s, the United States demonstrated notable restraint toward its rival. (...)
Second, the Soviet Union’s initial deployment of a nuclear arsenal did not make it safe and secure. In fact, by fielding a vulnerable nuclear force, the Soviets energized U.S. efforts to build weapons and war plans to rapidly destroy the Soviet arsenal if war erupted. The small, vulnerable Soviet nuclear force was a double-edged sword: it benefited the Soviets in peacetime by making U.S. aggression less likely, but it vastly increased the damage the Soviets would suffer in the event of war.
Four views on deterrence: (1) Existential deterrence
fear of nuclear war is so great that the mere possibility of nuclear retaliation is enough to deter any act that could trigger escalation. The implication of this view is that a nuclear arsenal need not be large, survivable, or even easily deliverable to terrify and thus deter; its mere existence is enough. (...) existential deterrence holds that nuclear deterrence is easy to achieve. As soon as a country develops nuclear weapons, other states will be very unlikely to initiate major attacks, even if the new nuclear power has no clear ability to deliver its warheads. (...) the relative size of a country’s nuclear arsenal compared to that of its adversary or the vulnerability of a country’s arsenal to attack are essentially irrelevant for generating deterrence.
Four views on deterrence: (2) Minimum deterrence
Deterrence will be robust whenever retaliation in the wake of an attack is plausible, instead of merely possible. To create the plausibility of retaliation, a country needs to be capable of delivering nuclear weapons against an adversary. A small deliverable nuclear arsenal—even if it is not certain to survive a disarming strike—will still deter attack because the chance of retaliation will be sufficiently terrifying. (...) a sufficient nuclear retaliatory force is simple to build, cheap to maintain, easy to hide, and mostly unaffected by adversary nuclear force levels, deployment policies, and strategies.
Four views on deterrence: (3) Assured retaliation
This school views deterrence as robust only when the defender’s ability to retaliate after suffering a disarming strike is nearly certain. The threat of retaliation need not be society ending, however; the prospect of absorbing “just a few” nuclear strikes will deter any aggressor. (...) The goal is to convince potential attackers that even a flawlessly executed disarming strike, carried out under ideal circumstances, would still fail—and result in the destruction of several of the attacker’s cities. (...) The ingredients of a survivable nuclear force typically include some combination of the following: moderate (or large) force size; diverse delivery systems; survivable basing modes; redundant command and control systems; and well-trained personnel.
Four views on deterrence: (4) Assured destruction
truly robust deterrence requires that retaliation be both assured and massive (...) When the potential victim possesses a very large and invulnerable retaliatory capability, a nuclear attack would be tantamount to national suicide—and no remotely rational leader would invite that outcome. (...) nuclear arsenals need to do more than deter typical leaders in normal circumstances. A robust nuclear posture needs to terrify and frustrate the goals of even highly risk-acceptant leaders during high-stakes crises. (...) Only the guaranteed and utter destruction of the attacker’s population, industry, and leadership suffices against the most motivated foes.
Chapter 3: Escaping Stalemate: The New Era of Counterforce
Reversibility of stalemate
For countries to relax and abandon the geopolitical strategies of the prenuclear age (...) nuclear- armed countries would need to be confident that the condition of stalemate is irreversible. In reality, however, countries can escape from stalemate. Therefore, even well-armed nuclear countries continue to engage in military competition. They compete to prevent adversaries from developing effective disarming strike capabilities, and in some cases they compete to build first strike capabilities themselves.
The reversibility of stalemate explains much of the competition in the last decades of the Cold War—as well as an increase in great power competition today. During the Cold War, the Soviet development of a survivable retaliatory force in the 1960s did not end or even dampen the nuclear arms race; instead, it triggered decades of intense U.S. efforts to track and target Soviet nuclear forces, to give the United States a war-winning option if deterrence failed. Those highly-classified U.S. efforts were more effective than most analysts have appreciated, and they explain much of the U.S. and Soviet arms racing behavior until the end of the Cold War. Today new technologies are increasing the vulnerability of nuclear forces around the world—unleashing renewed competition among the major nuclear powers.
Ways to increase the survivability of nuclear arsenals
Military planners have employed three basic approaches to protect their countries’ nuclear forces from attack: hardening, concealment, and redundancy.
“Hardening” refers to the deployment of nuclear forces (such as delivery systems, warheads, and command sites) in reinforced structures that are difficult to destroy (...) Hardening is attractive, but it comes at the price of concealment—for example, it is difficult to hide the major construction entailed in building a nuclear silo. Also, hardened sites are not mobile; once discovered, they remain so. Similarly, concealment comes at the price of hardening. If mobile forces are discovered, they tend to be easy to destroy. (...)
concealment, by which we mean efforts to prevent adversaries from identifying or locating one’s forces (such as through the use of camouflage, decoys, and especially mobility). Concealment is the foundation of survivability for mobile delivery systems, such as ballistic missile submarines (SSBNs) or mobile missile launchers (...)
redundancy is used to bolster every aspect of the nuclear mission, especially force survivability. Most nuclear-armed states use multiple types of delivery systems and warheads to complicate enemy strike plans and protect against warhead design flaws. They spread their forces and warheads across multiple bases.
Effects of technology improvements on survivability
Major technological trends are directly undermining these strategies of survivability. (...) Leaps in weapons accuracy threaten nuclear forces that rely on hardening (...) [and] an unfolding revolution in remote sensing threatens nuclear forces that depend on concealment. (...) Deterrence will weaken as arsenals become more vulnerable.
Delivery system accuracy
Technological improvements chipped away at the sources of inaccuracy, however. Leaps in navigation and guidance, including advanced inertial sensors with stellar updates, improved the ability of missiles to precisely determine their positions in flight and guide themselves, as needed, back onto course.
Accuracy solves the fratricide problem, enabling counterforce strikes
The accuracy revolution, however, also offers a solution to the fratricide problem, opening the door to assigning multiple warheads against a single target, and thus paving the way to disarming counterforce strikes. (...)
Destroying hard targets typically requires low-altitude detonations (so-called groundbursts), which vaporize material on the ground. When the debris begins to cool, six to eight seconds after the detonation, it solidifies and forms a dust cloud that envelops the target. Even small dust particles can be lethal to incoming warheads speeding through the cloud to the target. Particles in the debris cloud take approximately twenty minutes to settle back to ground (...)
Only those [warheads] that miss—that is, those that travel to the target area and detonate outside the [lethal radius]—will create a dust cloud that shields the target from other incoming weapons. In short, leaps in accuracy are essentially reducing the set of three outcomes (hit, fail, or miss) to just two: hit or fail. The “miss” category, the key cause of fratricide, has virtually disappeared.
Greater accuracy makes low-casualty counterforce attacks plausible
Counterforce was tantamount to mass casualties throughout the nuclear age, but the accuracy revolution is severing that link. In the past, the main impediment to low-casualty nuclear counterforce strikes has been radioactive fallout. (...)
If weapons are detonated at high altitude (above the “fallout threshold”), very little debris from the ground will be drawn up into the fireball, greatly reducing fallout. In practice, however, this targeting strategy has never been feasible against hardened sites. The problem is that any high-yield weapon that detonates low enough to destroy a hardened target will also be low enough to create fallout. Low-yield weapons could do the job and remain above the fallout threshold, but that has always been impractical because low-yield weapons would need to be delivered with great precision to destroy hardened sites, which was previously impossible. (...)
As accuracy continues to improve, the effectiveness of conventional attacks on hard targets will continue to increase. Today, low-yield nuclear weapons can destroy targets that once required very large-yield detonations. In the future, many of those targets will be vulnerable to conventional attacks.
Overcoming concealment
Finding concealed forces, particularly mobile ones, remains a major challenge. Trends in technology, however, are eroding the security that mobility once provided.
Five trends are ushering in an age of unprecedented transparency. First, sensor platforms have become more diverse. (...) Second, sensors are collecting a widening array of signals for analysis using a growing list of techniques. (...) Third, remote-sensing platforms increasingly provide persistent observation. (...) The fourth factor in the ongoing remote sensing revolution is the steady improvement in sensor resolution. (...) The fifth key trend is the huge increase in data transmission speed.
Countermeasures?
Countries will surely address the growing vulnerability of their nuclear arsenals by trying to develop countermeasures to thwart advanced sensor and strike systems. They will seek to deploy radar jammers, antisatellite weapons, and decoys. (...) Yet, there are good reasons to expect that the net result of these efforts will leave nuclear-delivery systems more vulnerable than they have been in the recent past.
First, hunters are poised to do well in the back-and-forth battle of countermeasures. Counterforce is the domain of the powerful; those that are seeking to track enemy nuclear forces typically have greater resources than their rivals. (...) many countermeasures reduce one vulnerability at the cost of exacerbating others. For example, limiting communications between mobile missiles or submarines and their command authorities reduces vulnerability to signals intercepts, but it increases vulnerability to attacks designed to sever (or simulate) their command and control. (...)
Second, the potential targets of disarming strikes cannot respond to just a single counterforce technology; they must respond to a daunting list of them. (...)
Third, some vulnerabilities are difficult to fix. In the late 1960s, the Soviet Union learned that its SSBNs were being tracked by the United States, but it took more than a decade to counter this U.S. capability.
Impact of emerging technologies on nuclear arsenal survivability
although the impact of emerging technologies on nuclear command, control, and communications (NC3) and early warning (EW) systems is difficult to predict, there are good reasons to expect that this will be an area of growing vulnerability for countries facing sophisticated adversaries. (...) Today, all nuclear-armed countries need to be concerned that advances in quantum computing, artificial intelligence, and cyber could undermine the ability to detect a nuclear first strike or order and deliver a retaliatory response.
Policy implications of improvements to counterforce capabilities
First, if nuclear forces are becoming increasingly vulnerable to counterforce, then states need to improve their retaliatory arsenals just to maintain the same level of deterrence. (...)
Second, the growing threat to nuclear arsenals (from nuclear strikes, conventional attacks, missile defenses, ASW, and cyber operations) raises major questions about the wisdom of cutting the size of nuclear arsenals. (...)
Finally, leaps in accuracy and remote sensing should reopen debates in the United States about the wisdom of pursuing effective counterforce systems.
Chapter 4: Deterrence under Stalemate: Conventional War and Nuclear Escalation
Argument
In this chapter we make two key arguments. First, we show that a strategy of deliberate nuclear escalation in a conventional conflict can be rational— despite nuclear stalemate—if the side that wishes to escalate has flexible, limited nuclear options. (...) Second, we show that nuclear-armed countries seeking to deter major conventional attacks with nuclear threats act as if substantial nuclear capabilities are required to make that mission credible.
The stability-instability paradox
In short, if initiating nuclear war under the condition of stalemate is virtually tantamount to suicide, no country would do so in response to limited aggression. But if nuclear escalation is therefore not in the cards as a response to a limited conventional attack, then deterrent threats based on escalation will not be credible. (...) The heart of Snyder’s paradox is that nuclear stability could lead to conventional instability. (...)
Moreover, nuclear stalemate might actually make a country more likely to initiate a conventional war if it is highly confident that its attacks will not trigger nuclear escalation. In such a case, the nuclear shield allows one to strike with the conventional sword.
Four views on the stability-instability paradox
[1. Existential deterrence:] the mere possibility of catastrophic escalation is sufficient to produce conventional deterrence. (...) No potential benefit of launching a conventional attack could be worth the potential costs of fighting a nuclear war. Even if an attacker chooses limited conventional objectives so as not to trigger retaliation, it can never rule out the possibility of catastrophic escalation stemming from misperception, miscalculation, unauthorized use, or some other unintentional cause or irrational decision (…)
[2. Manipulation of risk:] countries can take simple steps during crises to increase the risk of escalation, thereby coercing potential attackers to back down. (...) The point of manipulating danger (...) is to generate unbearable risk in the minds of potential attackers, not to change the actual military balance in the conventional or nuclear realm. (...)
[3. Coercive escalation:] agrees with optimists who contend that nuclear-armed countries can manipulate risk in order to deter limited attacks, but argues that this entails greater effort and attention to military capabilities than the optimists allow. (...) This strategy calls for building flexible and limited nuclear use options, and especially nuclear capabilities that would be survivable through extended periods of conventional conflict during which the adversary might substantially degrade delivery systems, command sites, and communications links.
[4. Escalation dominance:] conventional deterrence requires the capabilities to win at every stage of nuclear conflict. A core belief of this strategy is that military victory is possible, even in nuclear wars. (...) [It] entails the ability to match or outmatch the potential attacker in strictly military terms at every possible rung of the escalation ladder. Only this ability, proponents argue, renders nuclear options truly “usable” and thus credible. In short, a potential attacker needs to understand clearly that it cannot win or will be defeated under any military contingency.
Rationality of nuclear escalation
Nuclear escalation may seem cold-blooded, but it is neither irrational nor far-fetched. (...) Coercive nuclear escalation during a conventional war would constitute an enormous gamble—a long shot at producing a cease-fire. But escalation can be a rational choice even if it offers only a small chance of success. Leaders on the verge of conventional defeat may rationally accept enormous risks to generate a cease-fire if the outcome of not escalating is sufficiently terrible. In fact, even if escalation worsens the expected consequences for the leaders on the losing side, they may still rationally choose this strategy if it increases the variance in the distribution of potential outcomes. (...)
Japan’s attack on Pearl Harbor is illustrative of the gamble. Japan was being strangled by U.S. economic sanctions; specifically, it was denied needed oil reserves by Washington’s oil embargo. Japan’s leaders understood that a surprise attack on a country with ten times the GDP would likely end badly, but they believed that the path they were on would inevitably lead to defeat. So they gambled, taking a leap to create the possibility of success, even though they anticipated—correctly—that escalating the dispute with the United States would worsen their odds of success.
Nuclear arsenal requirements for conventional deterrence
For nuclear weapons to function as the greatest tool of deterrence, they must be usable in response to a major conventional attack. Yet, to make nuclear weapons usable, countries need to build larger, more flexible, and more survivable arsenals than what might otherwise be necessary or wise under conditions of nuclear stalemate. The ultimate result is the persistence of intense security competition, which nuclear weapons seemingly should have banished. (...)
deterring conventional attacks using nuclear weapons is a difficult mission. (...) countries that face powerful adversaries are not satisfied with modest and simple nuclear forces. Instead, they seek to build arsenals optimized for wartime coercive nuclear escalation—that is, arsenals that can be employed flexibly and can ride out conventional and nuclear-disarming strikes.
Strategies of coercive nuclear escalation (CNE)
[CNE] cover[s] a wide range of potential actions. At the low end of the escalatory ladder, employment could be merely issuing threats or visibly alerting forces. Climbing a few rungs, employment could include nuclear demonstrations—such as wartime weapons tests, or detonations over international waters. At the high end of the ladder, employment can mean strikes on military or civilian targets.
Force requirements for coercive nuclear escalation (CNE) strategies
First, all things being equal, the more survivable a nuclear arsenal is the better suited it is for CNE. (...) If a country with a vulnerable arsenal were to attempt CNE, the adversary has another option besides being coerced—it could respond with a counterforce strike to prevent any future escalatory use (...)
The second implication is about the ideal size and diversity of a nuclear arsenal for the CNE mission. In short, more is better. (...) a large arsenal would permit planners to devise a coercive campaign with several rungs in the escalatory ladder (...) Furthermore, a large arsenal would facilitate CNE by contributing to resilient survivability, deterring the target of coercion from responding to coercion with a disarming strike. Finally, diversity of nuclear systems would allow a coercer to modulate the amount of pain it inflicts on its enemies.
Conclusion: Solving the Nuclear Puzzle
Summary of the book's argument
Stalemate emerges through a highly competitive process in which countries must work hard, and often against the determined efforts of their adversaries, to build truly survivable retaliatory arsenals. Countries continue to compete while they are locked in stalemate, because stalemate is reversible. And even when stalemate exists, countries seek flexible capabilities—conventional and nuclear—to deter or respond to nonexistential threats. (...) [Nuclear weapons] are, indeed, the best instruments of deterrence ever created. They have not, however, solved the problems of international anarchy or revolutionized international politics. (...) International security competition is alive and well.
Policy implications
First, the future of nuclear arms reductions looks bleak. (...) Historically, potential adversaries have agreed to nuclear arms cuts when they did not undermine strategic stability—that is, when the cuts did not raise the danger that either side could launch a disarming first strike. (...) Whether right or wrong in the past, however, the assumption that arms cuts and deterrence stability go hand in hand is increasingly dubious today (...) The problem is stark: arms-control agreements that cut nuclear weapons reduce the number of targets that must be destroyed in a disarming strike, and all the while, the nonnuclear forces that aim at those targets grow in number and capability
A second major policy implication of our analysis—the flip side of the difficulty of nuclear arms reductions—is that one should expect to see serious arms racing among nuclear adversaries. In some cases, countries will work hard to create truly survivable retaliatory forces, while their rivals will strive to hone counterforce capabilities to keep those retaliatory forces vulnerable.
A third major policy implication pertains to the process of nuclear proliferation and efforts to stop such proliferation. The decision to acquire nuclear weapons is fraught with danger because adversaries have incentives to strike militarily before countries can join the nuclear club (...) and the main benefit of joining the nuclear club—deterrence—does not accrue instantaneously or automatically. As we saw with the early Soviet arsenal, the initial deployment of nuclear weapons can make a country more vulnerable than before if it causes an adversary to ramp up its targeting and attack plans. (...) Our research shows that the threshold at which a nuclear arsenal truly stops being a “target” and starts to be only a “deterrent” appears far higher than many analysts assume. (...) In short, would-be nuclear proliferators face a long road to security, paved with the dangers of preventive attack, preemptive strike, conventional war, and a relentless struggle to maintain a survivable arsenal in an age of rapid, dramatic, and often unpredictable technological change.
Nuclear disarmament
abolition is a deeply misguided goal. In fact, even if the core of the abolitionist argument were correct—that the dangers of a nuclear-armed world outweigh the deterrent benefits those weapons offer—the effort to abolish nuclear weapons would likely have terrible unintended consequences.
First, a successful abolition campaign would not only lead to more conventional wars; it would also likely trigger dangerous nuclear rearmament races during crises and conflicts. (...) In essence, abolitionists present a false choice: between the world of conventional conflict that existed before 1945 and the world of nuclear dangers we inhabit today. But abolition, if successful, would create a third, more dangerous situation: a world of conventional wars and nuclear know-how, in which adversaries teetered on the brink of a rearmament race and faced real incentives to preempt once they reacquired nuclear weapons during a crisis. (...)
Second, the likely result of the abolition movement is not global elimination but rather incremental progress toward disarmament, a result that would create dangerous deterrence dynamics. All else being equal, smaller nuclear arsenals mean more vulnerable nuclear arsenals. And given long-term trends in technology that are making nuclear forces increasingly vulnerable to conventional attacks, deep multilateral nuclear cuts are likely to create mutual vulnerability to disarming strikes (...) In short, no country should aspire to have “just barely enough” nuclear deterrence. (...) Chipping away at that certainty [in the minds of aggressors that victory is impossible] through arms cuts is potentially dangerous and counterproductive.
Impacts of technology on nuclear arsenal survivability
developments rooted in the computer revolution are making nuclear forces around the world far more vulnerable than before. The superpower nuclear arsenals seemed robustly survivable for the latter half of the Cold War, when each side deployed large and dispersed forces, but the foundations of stalemate are being eroded today by vast improvements in weapons accuracy, sensing technology, data processing, communication, and artificial intelligence. The task of securing nuclear arsenals against attack is becoming more challenging.
I am grateful to Will Aldred for encouraging me to turn my reading notes into this post.
This is a really good summary. I think the main area I have remaining uncertainty about these types of arguments is around if tech decisively favors finders or not. I believe there's been a lot of analysis that implies that modern nuclear submarines are essentially not feasible to track and destroy, and there are others that argue AI-enabled modeling and simulation could be applied as a "fog of war" machine to simulate an opponent's sensor system and optimize concealment of nuclear forces.
Nevertheless, without more detail these sorts of counterarguments seem highly contingent to me: they depend on the state of technology that actually gets deployed, and there may be decisive counter-counter arguments based on other military capabilities.
It would be interesting to see more EAs "grok" these sorts of arguments and to think through corresponding nuclear modernization and arms control strategies that would be implied by trying to assure a desirable long-term future. I've tended to think more redundant and accurate arsenals of low-yield weapons like neutron bombs would tend to strengthen deterrence while eliminating most of the risk of nuclear winter and long-term radiation, but it's easy to imagine there might be better directions yet to ratchet competition and negotiation since that kind of proposal could also be very destabilizing!