The Great American Gamble - Deterrence Theory and Practice from the Cold War to the Twenty-First Century by Keith B Payne
The Great American Gamble is a book analyzes how (ownership of) nuclear weapons was used and can be used in the future, to achieve important goals, like preventing nuclear war. It looks at the actual policies, but also the theory behind those policies. Were these policies and theories wise or were they MAD? Did the US avoid nuclear war by skill or by luck? Can these same policies keep the US and its allies safe from new threats, like North-Korea, China or terrorists?
Theory - Schelling
The book contrasts two strategies to deter a nuclear attack, based on the ideas of two theorists, Thomas Schelling and Herman Kahn. Schelling argues that it is possible to achieve a stable balance of terror, where it is irrational for either side to launch a nuclear first strike, because the retaliatory strike will result in unacceptable damage. So no rational nation will then use nuclear weapons on the US or their allies, purely for reasons of survival. The biggest threat to this strategy is a loss of confidence in the retaliatory strike being sufficiently damaging to be a deterrent. This can be the result of first strike weapons that can destroy a large number of nuclear weapons of the opponent, as well as a partially effective ballistic missile defense system (BMD). In both cases, a first strike may be able to knock out the ability of the other side to retaliate with sufficient force, either because enough of their nuclear weapons have been destroyed, or because enough of their remaining missiles will be intercepted by the BMD. This can make a first strike a rational choice, because it removes the threat of destruction to your own nation.
Since a fully reliable BMD is (currently) impossible, Schelling argues that the goal should be to achieve a stable situation of mutual assured destruction, minimizing the fear on each side that their retaliation will not work, through unilateral and/or cooperative measures. Trying to gain an advantage by seeking first-strike capability or building a BMD is going to result in a dangerous arms race, which is Schelling calls an action-reaction dynamic, where the other nation will respond with increased offensive capability. So the rational choice is to avoid that situation, in favor of a stable situation of mutually assured destruction.
However, the question remains how the opponent is deterred from a smaller or conventional attack? Is the American nuclear arsenal a deterrent to such an attack? Is it not preferable to be ruled by the Russians, rather than to die in a mutual nuclear exchange? The conventional military ability of USSR started to greatly exceed the capability of the West, so the Russians could also not be held back in Europe merely with conventional weapons. Schelling's answer to this was that the deterrent effect was not in the certainty of nuclear war, but in the uncertainty. It is not necessary to threaten with certain retaliation, but to make the credible claim that nuclear may happen due to reasons outside of the control of the one making the threat, like chance, accident, third-party influence, imperfect in the decision making machine or processes that we don't entirely understand.
To make this more credible, one can introduce complicating factors, like the presence of an American force in Germany. Furthermore, it is possible to dissuade opponents from making choices that may lead to nuclear war by "rocking the boat." This can be a bluff, but also a minor escalation. To be able to act this way, it must be possible to engage in partial attacks or retaliation, rather than merely through a full nuclear strike. Then any rational opponent would back down. Yet this is inconsistent, because it asserts that it is rational for the opponent to back down when doing otherwise would seriously increase the risk of nuclear war, but for oneself to escalate the situation, to dissuade the enemy. It ignores the possibility of a scenario where both sides consider their own demands to be reasonable, but the demands of the other side to be unreasonable, and expect the other side to back down if they escalate. Furthermore, it is also inconsistent for Schelling to recognize that nations may consider it very plausible for imperfect decision making to result in outcomes that the leaders didn't want, and be dissuaded by that possibility from taking actions that bring the nations close to nuclear war, yet not to see imperfect decision making as a way in which deterrence may fail.
Schelling gives the following analogy to explain why we don't have to be afraid while we depend on mutual deterrence, or have to hate our enemy to use it: "People regularly stand at the curb watching trucks, buses, and car hurtle past at speeds that guarantee injury and threaten death if they so much as attempt to cross against the traffic. They are absolutely deterred. But there is no fear. They just know better." However, in reality, pedestrians do walk into traffic, because they are drunk, distracted by their phone or because they are suicidal. Nor do drivers always stay on the road, by accident or on purpose. According to the WHO, hundreds of thousands of pedestrians die every year, with many more injured. Car manufacturers consider pedestrian safety when they design their cars, even including active safety systems that merely benefit the pedestrian, for example, by raising part of the hood when an impact is imminent, to cushion the blow.
Theory - Kahn
A competing theory is offered by Herman Kahn, who argues that a self-destructive threat on behalf of others is not credible. The Soviets cannot be expected to believe that the US will initiate a nuclear war to defend NATO allies, when the US will not survive the expected retaliation. Instead, the Russians will likely see it as a bluff. So the US can only credibly threaten nuclear escalation by being able to limit the damage from a retaliatory strike. He argues that being able to limit the loss to 10% of the US population might be sufficient to make the threat credible. Only the US needs to have the ability to credibly threaten with nuclear war, because the Soviets have superior conventional forces. Such survivability could be created by relocating people and industry, creating shelters, air defenses and a BMD.
Maximizing the credibility of the threats that are necessary to defend NATO allies requires a threat that leaves little to chance and that doesn't limit itself to merely targeting cities and civilian industry, as Schelling suggests. Instead, it should also be capable of striking at (hardened) military targets.
Kahn doesn't believe that deterrence is 100% reliable and that the suggested defenses that he suggests are not merely helpful to discourage Russian aggression, but also to reduce the damage if nuclear war does happen. Perhaps enough of society will remain standing to allow it to continue, rather than to end up in the Mad Max universe.
Rationality?
Both Schelling's and Kahn's theories are ultimately based on the idea that there are universal goals, in particular societal survival, that motivate the leaders of nations. The problem these theorists are primarily concerned with, is the plausibility of the threat. Are nations sufficiently worried about their society being destroyed, to prevent them from launching a nuclear strike on the US, attacking Western allies that the US really doesn't want to be taken over by the Soviets, or take risks whose outcome they cannot oversee?
Yet neither theorist examines history to see if their definition of rationality is truly universal. Japan made a choice to go to war based on the belief that giving in to the American demands, or accepting the American boycott, would result in being relegated to a nation of minor importance. The advocates argued that this was an unacceptable loss, so a chance of victory was better than sure defeat. A prince of the Imperial family wrote that if Japan lost: "there will be nothing to regret because she is doomed to collapse even without war." And at the end of the war, Japan preferred the enormous death and destruction of civilian resistance to an American invasion, with no hope of victory, over surrender, because they deemed a honorable loss more important than preserving life. Similarly, Nazi Germany preferred to fight a losing battle over surrender. Several of the top leaders were under no illusion about their fate upon surrender and preferred suicide. Would they have refused to use nuclear weapons out of fear of retaliation?
Both Mao and Fidel Castro expected an American invasion and one point, and suggested to Soviets to respond with a nuclear first strike from their territory (using Soviet nukes), rather than fight a conventional war. Both accepted an US nuclear retaliation on their nation and weren't willing to first see whether a win or stalemate could be achieved with conventional arms; or whether the goal of the invasion was limited.
However, it is not just the enemies of the US or those with a very different culture who may accept enormous destruction and death because they consider other values to be more important. The British decision to declare war on Nazi-Germany after the Polish invasion was assumed to result in mass bombings by the Germans on British cities, leading to hundreds of thousands or even millions of casualties in a few weeks. A post-war prime minister who served in the Churchill administration during the war, wrote that they considered the expected consequences of declaring war on Germany to be in the same category as later strategists considered the consequences of nuclear war. Yet they did declare war.
So it is highly doubtful that self-preservation or avoiding mass death and destruction is or will be the most important goal of all leaders who have nuclear capability. This is true in particular for leadership that has values or beliefs that are relatively alien to many in the West, like honor, destiny, martyrdom, reincarnation or a belief in an Utopia that is worth immense sacrifice to achieve.
Schelling recognizes the possibility that a leader may make a decision that he considers irrationally suicidal, but argues that those orders will not be followed. However, this ignores that orders that may seem irrational to Schelling, may not seem rational to people in another culture or with another ideology. Or, the military leadership may not be willing to question or countermand decisions.
Deterrence during the Cold War
The actual policies followed by the US during the cold war were far more based on Schelling's theory than Kahn's. One of the strongest proponents and key figures in establishing US nuclear strategy for the next decades was Robert McNamara, who in 1962 said that he wanted the Soviets to have confidence that they could retaliate effectively. This is the logic of Schelling, where being vulnerable to the opponent and them being vulnerable to you, creates a stable situation.
The Pentagon calculated that offensive spending was far more cost effective than defensive spending. Reducing casualties would cost 3-4 times as much as the cost of new weapon capabilities to increase the casualties back to their previous levels. It was calculated that maintainable levels of spending would ensure at most a survivability of 80% of population and industry, which was considered insufficient. A loss of 20% of people and industry would destroy US society and thus would be utterly unacceptable. A senior White House official in the Kennedy and Johnson administrations argued that "10 million casualties are no different than 100 million casualties." By this he didn't literally mean that there was no difference, but rather, that a reduction from 100 million casualties to 10 million was not considered worthwhile. The National Security Advisor to Kennedy and Johnson even argued that only perfection was good enough for missile defense: "[nuclear weapons] make it necessary to defeat them all. Anything less is not good enough for safety."
McNamara assumed that the Soviets would have similar beliefs and being able to kill 20-30% of the Soviet population in retaliation would be sufficient to deter a Soviet attack on the US and its allies. He based the size of the nuclear missiles arsenal on that number. However, McNamara and other politicians seemed to diverge from Schelling's theory on the point of uncertainty. They generally seemed to commit to a retaliatory strike against the Russians and sought to minimize chance, for example, by trying to maximize order-following by missile operators.
Successive administrations and officials largely followed the same lines as McNamara, although probably because it was the most politically advantageous one. It was hard to sell large spending on defense, when those defenses didn't provide any obvious benefits. Promising 100% safety was also more appealing than Kahn's belief that safety could not be assured. The advocates of Schelling's balance of terror portrayed Kahn as a warmonger, who sought to start nuclear war. By Schelling's logic, his policies were likely going to lead to nuclear war, or at least increase the risk, but obviously that was not true according to Kahn's theory. However, it seems more likely that Kahn's insistence that we think through the consequences of nuclear war, is beyond most people. The horror is so immense, that it is preferable to find a way to prevent it. A theory that claims to provide it, is far preferable to one that doesn't, even if significant criticisms can be made of the former.
However, merely recalling the Cuban missile crisis and the statement by one of the involved that the chance of nuclear war was about a third to even, should be sufficient to dismiss the idea that deterrence is highly reliable. And the decision by the Russians to 'destabilize' the situation by stationing nukes so close to the US, which makes a first strike more feasible, undermines the theory that the US can restrain others by restraining itself. The belief that one's own actions determine the actions of the opponent seems extremely narcissistic. While Schelling's theory asserts that there is a symmetric action-reaction dynamic, in practice the American politicians and advisers seem to always assume that the Americans act and the Russians react. The result is a strong inward-looking perspective, where the US is assumed to be safe if it manages to restrain itself. Fairly little recognition exists that the opponent may deal with the same difficulty to restrain itself (or have no desire to do so) and that this scenario requires a strategy as well.
The evidence strongly suggests that the Russians rejected the idea of stable deterrence, with no side having the upper hand. At least two studies analyzed American and Russian decision making and found no evidence that American choices to not develop certain nuclear weapons or BMD, restrained the Soviets. The Anti-Ballistic Missile Treaty, signed in 1972, states in it's preamble that restricting BMD would be a "substantial factor in curbing the race in strategic offensive arms." Yet it only caused Russian development of offensive nuclear weapons to accelerate, as the Russians could focus more on those. This outcome disproves Schelling's theory that rejecting BMD is (always) stabilizing.
While the Americans intentionally refused to build nuclear missiles that were both powerful and accurate, which could be used to attack and destroy hardened missile silo's, the Russians developed these and build a large number. It was only thanks to nuclear submarines that the US maintained survivable retaliatory capability. The Russians also built BMDs to defend Moscow, while the Americans didn't build one for their capital. Russia consistently appealed to the American theory by objecting to any American BMD that could defend against (some) Russian nukes as being destabilizing and a threat to peace, using language that is persuasive to adherents of Schelling's theory, yet they themselves didn't act as if they believed the theory. When Americans discussed their deterrence beliefs with the Russians, they rather consistently rejected the theory of the Americans.
The build-up of nuclear stockpiles peaked in 1986, shortly after Perestroika, Gorbachev's attempt to restructure the Soviet Union to solve their poor economic growth. The stockpiles declined rapid after that. The START I treaty that put limits on the stockpile, was only signed in 1991 after both sides had already begun to significantly reduce their stockpiles (between 1986 and 1991, the Soviets went from 40,000 to 29,000 nukes). Schelling had great faith in the ability for arms control through treaties, but it in reality, it seems that these 'arms control' treaties merely tend to affirm choices already made.
Deterrence after the Cold War
As the cold war progressed, deterrence strategy became more and more dogmatic, probably by the same mechanism that we see in many fields. The innovators think things through deeply and investigate different possible strategies, but later generations mimic their ancestors, but often without understanding. So they keep applying the same strategies even when circumstances change and invalidate the premises underlying the strategy. Furthermore, nuances get removed from the strategy that those earlier generations did advocate. For example, McNamara didn't advocate the same strategy against China as against the USSR, because he deemed it possible to build a BMD that could defend the US sufficiently from the limited nuclear arsenal of the Chinese, especially given their poor economy at the time. Even though it was just as true for China that evading American missile defense was much cheaper than building those defenses, China was so poor and technically incompetent that they couldn't maintain such an arms race. Yet after the cold war, a lot of strategists and politicians argued that BMD was useless and dangerous against any enemy, even the most weak ones.
In the post-war reality, there is a need for a recognition of the differences between opponents, including what kind of deterrence works on them (if any). For example, terrorists that value martyrdom are not deterred well by mass retaliation, if that is even possible, but seem to be deterred by being prevented from dying a 'true' martyr, by dying without being able to inflict casualties. So targeted strikes may deter others from choosing this path, or may persuade them to abandon it. On other opponents, nuclear deterrence may work well. Ultimately, it is important to not apply the same hammer to every nail, but to analyze each major opponent and choose a course of action that is likely to work best on them. It is also important to recognize that biological and chemical weapons are also WMD, whose use should be deterred.
Deterrence may require a different strategy or different weapons against different components. For example, a common occurrence during the 20th century was that opponents of the US or its allies, felt encouraged by supposed humanitarian 'softness' that would make democratic nations unwilling to accept the deaths that were needed to stop their aggression. Examples of this are Hitler, Hideki Tojo, Mao, Saddam Hussein and Slobodan Milošević. This often led to lose-lose scenario's where those politicians made choices that they expected would not result in American/Western aggression and that they probably would not have made if they anticipated this. The more difficult to achieve solution to this, is to change the perception of humanitarian 'softness'. However, US aggression can also seem more plausible if weapons can be used that are more surgical. The nukes that were developed for the cold war were high yield, low accuracy, perfect for destroying cities. However, it may be more deterring to many opponents if the US can use low yield, high accuracy weapons. For example, these weapons could then plausibly be used to threaten the lives of the leadership of those countries, even if they hide in well-protected bunkers, with little harm to civilians. Then the US can afford to be seen as humanitarian and very wary of collateral damage, while still deterring 'minor' transgressions, like the invasion of allied countries.
As the goals of the US and it's allies go beyond preventing nuclear war, those objectives should be considered as well. The book identifies four main methods for which nuclear capabilities can be used: deterrence, assurance, dissuasion and damage limitation.
Various allies have indicated that they place great trust in the US, including the 'nuclear umbrella.' Both Japan and Germany have indicated that they may acquire nuclear weapons if this trust is lost. So preventing nuclear proliferation among allies (which in turn may cause their opponents to acquire nukes), requires preserving this trust, in particular that a nuclear or existential attack on those countries will result in nuclear retaliation by the US. The reasonably aligned goals means that the allies can be asked about this, where their answer is likely to be much more trustworthy than when questioning opponents. For example, they may indicate that having American nukes stationed on their territory provides assures them that the US nukes will keep them safe. American nukes were and still are stationed in European nations, which may have contributed to Soviet reluctance to attack Western Europe, but also to confidence by Western nations in that their alliance with the US was worthwhile and that they didn't have to acquire nukes. Preventing nuclear proliferation among allies may prevent a cascading effect, where their (local) enemies also seek to acquire nukes. Preventing such a scenario may be one of the best ways for the US and their allies to reduce the risk of nuclear war, or of nuclear terrorism. After all, nations regularly descend into chaos, experiences coups or such, with government property falling in the hands of those who may not use it responsibly.
Dissuasion can also be used to prevent nuclear proliferation. For example, a BMD system that provides good defense against the number of nukes that a smaller, poorer or otherwise less capable nation can acquire, may convince those nations to not invest in a nuclear program or even to abandon such a capability. Another goal of dissuasion can be to discourage China from getting into a nuclear arms race with the US.
Damage limitation is particularly useful against nations or actors with limited nuclear ability. BMD can be quite effective against a limited attack by nations like North Korea, terrorists, or rogue elements within powerful nations. In the case of a successful attack, many lives can be saved with relatively modest investments, for example, by stockpiling iodine pills.
Verdict
I would definitely recommend the book. The public debate on this topic often appears dogmatic and simplistic, if it is not absent altogether. For example, advocates and critics of nuclear disarmament rarely bring up the possibility that nuclear proliferation may actually accelerate if allies lose their trust in the American nuclear umbrella, even though this seems very plausible. The value of nuclear weapons against contemporary enemies is also rarely debated, even though it is very important for people to understand the purpose of these weapons, lest they be abandoned without suitable alternative ways to achieve the same goals, or are misused. Trump allegedly asked an policy expert three times why the US couldn't use nuclear weapons, which suggests that the expert might not have been able to articulate this clearly. The reporting on this story seemed limited to attacking Trump for his foolishness, rather than actually explaining why they thought his question was wrong. Yet in itself, the question was fair. Government spending should be spent on things with a purpose and the purpose should be made clear to both politicians and the populace. I feel that the media are not doing their job on this topic. By reading this book, you can inform yourself reasonably well, or at least, as well as can be expected from one well-articulated and -sourced book from a specific point of view.
Yet that doesn't mean that the book is perfect. The author seems to occasionally exaggerate his sources. Insinuating that Fidel Castro wanted to unilaterally attack the US with nukes, rather than in response to an invasion, is misleading, for example. However, most of the book seems well-supported even if there are some exaggerations, because many of the claims in the book are fairly modest. The book primarily seeks to rebut the certainty of others and suggests instead to tailor the nuclear strategy to specific opponents (and allies), rather than assuming that the same strategy works for every opponent/ally and in any circumstance. Suggestions are offered, but never claimed to work with certainty, so readers are expected to draw their own conclusions.
What I found frustrating about the book is its structure. Instead of exhaustively exploring a topic and then moving on to the next, the book keeps revisiting subjects, where they keep getting expanded upon. Instead of continuing where it left off, a lot of the previous material is summarized. The book already extensively quotes politicians and advisers, even if they have largely the same beliefs as others, which does make it clear how nuclear strategy (didn't) develop over time, yet also results in a feeling of repetitiveness. Combined with the tendency to summarize earlier material, the books feels very padded. Fortunately, there is an index, so the book is still usable as a reference, as the structure makes it very hard to find anything otherwise.