Defence CommitteeWritten evidence from Ward Wilson, Senior Fellow, British American Security Information Council (BASIC)

Executive Summary

1. Nuclear deterrence has clear limits that are not well understood.

2. A careful review of facts from the Cold War seems to point to numerous nuclear deterrence failures. The evidence from neuroscience raises doubts about how sensible it is to rely on rational though in a crisis.

3. Nuclear deterrence does not appear to be reliable or safe over the long run.

Introduction

4. I am a Senior Fellow at the British American Security Information Council (BASIC) where I direct the Rethinking Nuclear Weapons Project. Trained as a historian, I have thirty years of experience with nuclear weapons issues. My recent book, Five Myths About Nuclear Weapons poses pragmatic challenges to fundamental ideas about nuclear weapons. BASIC is a think tank based in London and Washington focused on nuclear weapons issues. Its most recent work includes the Trident Commission. I have presented before the All Party Parliamentary Group on Weapons and Protection of Civilians, the Top Level Group, and recently debated Sir Laurence Freedman at Chatham House.

Commentary

5. Deterrence is persuading an adversary (usually by threat) not to take action.

6. Deterrence is different from defense, no matter what some theorists may say about “deterrence by denial.” Defence is placing a physical barrier between your adversary and his goal. A castle wall defends. A shield defends. A threat to burn down your adversary’s village deters. The difference between defense and deterrence is that defence is physical while deterrence is mental. For deterrence to work you have to get inside your adversary’s head.

7. This is why deterrence fails against madmen and people overwhelmed by emotion. A wall can stop a madman. A threat to burn down the madman’s village may not. Throughout history leaders have preferred defence over deterrence, because defence works against all comers, while deterrence only works against some. Deterrence is limited in its application: it only works with those who stop and consider rationally the costs of what they’re about to do. This means that from the outset, deterrence cannot be expected to cover all situations.

8. And this limitation is an important one, since one of the defining features of war is that during the course of war people—including leaders—often feel overwhelming emotion. Deterrence does not work reliably against adversaries overwhelmed by emotion. Therefore, deterrence cannot be expected to work reliably in war.

9. In addition there is considerable confusion about why we currently rely on and talk so much about nuclear deterrence. Most experts assert that nuclear deterrence is necessary because of the characteristics of nuclear weapons. This is false. We currently rely so much on deterrence because of the characteristics of missiles, not nuclear weapons. Missiles are relatively unstoppable, which means your only option is to deter them.

10. Consider the following imaginary world in which nuclear weapons do not exist but missiles do. In such a world it is still possible to launch a bolt from the blue attack against your adversary’s capitol and leadership. Such an attack would require more missiles, it would be more expensive, but it would still be possible. Since missiles are relatively unstoppable, the only way you have to prevent an adversary from launching such an attack is to threaten a counterattack of some sort. To deter, in other words.

11. Now imagine the opposite world. A world in which nuclear weapons exist but missiles do not. Or perhaps nuclear weapons are so large that they cannot be carried by missiles. They must be carried to coastal targets by giant freighters or carried inland targets by tractor-trailer trucks. In such a world it would only necessary to defend against freighters and tractor-trailer trucks. To construct barriers in your harbors or coastal waters. Freighters and tractor-trailer trucks (unlike missiles) can be stopped. Nuclear weapons could still be used to harm your troops along the front lines, but they could not be used to destroy cities far removed from the front lines.

12. It would be possible to build physical barriers against the delivery systems of nuclear weapons and thus defend against them. In a world without missiles, defence would be the preferred method of stopping nuclear weapons.

13. It is the unstoppability of missiles that forces us to rely on deterrence, not the size of the explosions the warheads produce. And we rely on deterrence reluctantly: it is always the second best choice.

14. One of the crucial differences between nuclear deterrence and ordinary deterrence is that the bar for effective nuclear deterrence is set much higher. A harsh law that deters some murderers, can be said to be working. It can be counted as a deterrence success. But since any failure of nuclear deterrence runs the risk of catastrophic nuclear war, even one failure of nuclear deterrence is too many. You could say that for nuclear deterrence, “failure is not an option.”

15. Because the consequences of failure are so high, nuclear deterrence must be perfect. That is a difficult standard to meet in perpetuity for a method that is so dependent on something so subject to human frailty—the mind.

16. Nuclear weapons proponents often say that nuclear deterrence worked perfectly during the Cold War. And it is certainly true that no Cold War crisis ended in nuclear war. But the claim that nuclear deterrence worked perfectly—or at least demonstrated an admirable reliability—can only be sustained if you pick and choose the evidence selectively.

17. Take the Berlin crisis of 1948. Proponents of nuclear weapons talk about Truman’s shifting B-29 bombers to the United Kingdom. They claim that this implicit nuclear threat was the reason that Stalin did not escalate the crisis. They claim it, in other words, as a success for nuclear deterrence. But they do not ask how Stalin was able to begin the crisis in the first place. In 1948 the United States had a monopoly on nuclear weapons. Any crisis that brought the military forces of the Soviet Union and the United States, United Kingdom, and France into close proximity risked escalating to war and potentially nuclear war. If nuclear deterrence means that a leader sees the danger of nuclear war and pulls back, why wasn’t Stalin deterred from initiating the crisis in the first place?

18. Take the Korean War. Proponents of nuclear weapons point out that Truman again shifted B-29 bombers close to the point of conflict—in this case to the island of Guam. They claim this implicit nuclear threat deterred the Soviet Union from becoming involved in the Korean War. They don’t ask, however, why this threat of nuclear war failed to deter the Chinese from becoming involved in the Korean War.

19. Take the Cuban missile crisis. Proponents of nuclear weapons claim this as the classic instance of nuclear deterrence succeeding. After all, the Soviets put the missiles into Cuba, there was a risk of nuclear war, and they took them out. What clearer demonstration of the power of nuclear deterrence could you ask for? But they don’t ask, “What about President Kennedy?” Kennedy knew that if he blockaded Cuba he ran the risk of nuclear war. In the week of secret deliberations during which US strategy was mapped out, they mentioned the risk of nuclear war 60 times. (And in the event they were right to be concerned. The United States and Soviet Union came within a hairsbreadth of nuclear war at least three separate times during the crisis. They were preserved from disaster by luck, not the perfect functioning of nuclear deterrence. See Michael Dobbs, One Minute to Midnight.) If nuclear deterrence means that a leader sees a risk of nuclear war and then pulls back from risky and aggressive action, how can we explain Kennedy’s action?

20. Consider the Middle East war of 1973. Proponents of nuclear weapons point to Henry Kissinger putting US nuclear forces on alert in order to warn the Russians not to send troops to Egypt. And the Russians didn’t send troops to the Middle East. “See?” they say. “Nuclear deterrence worked.” But this overlooks a far more important failure of nuclear deterrence. It doesn’t address the question, “What were Sadat and Assad thinking?” Everyone knew Israel had nuclear weapons, it had been reported in the New York Times. If nuclear deterrence stops leaders from taking risky and aggressive action how do we explain the fact that Egypt and Syria’s leaders made war against the Israelis?

21. Consider the Falkland Islands war of 1982. Why didn’t the United Kingdom’s nuclear weapons deter Argentina’s leaders from launching a risky and aggressive war?

22. Finally, consider the Gulf War. General Kevin Chilton, at one time commander of all US nuclear forces, in an article in strategic studies quarterly, claims that the Gulf War is proof that nuclear deterrence works. Secretary of State James Baker delivered a letter to the Iraqis which said that if they used chemical or biological weapons the United States would “make the strongest response possible.” And the Iraqis didn’t use chemical or biological weapons. Chilton points to this as evidence that deterrence works. But if you read the letter closely, it actually draws three red lines in the sand: don’t use chemical or biological weapons, don’t set the oil wells on fire, and don’t make terroristic attacks against our friends and allies (in other words, Israel). And as we all know, the Iraqis crossed two of those red lines. They set the oil wells on fire, and they fired SCUD missiles at Israeli civilians. Does this mean that nuclear deterrence only works one third of the time?

23. Sir Laurence Freedman claims that “everyone knew” that this was not a nuclear threat. If you examine the transcripts of the meeting, he says, it is clear that this was not a nuclear threat. I’m not sure I’m persuaded. But even if Freedman is right, it’s a useful commentary on how the “success” record of nuclear deterrence has been assembled. Gen. Chilton claims this episode as proof that deterrence works. If the rest of the proof that deterrence works is as reliable as this piece of evidence, it is time to make a wholesale reassessment of the historical record of nuclear deterrence.

4. Some scholars claim that deterrence ought to be easy, after all, you can deter a dog from getting up on your favorite chair. If you can teach a dumb animal not to do something, why wouldn’t deterrence work? The problem here is that two different types of brain function are at work: fast and slow thinking. (See Daniel Kahneman, Thinking Fast and Slow.) Fast thinking happens so fast it seems instinctive. It is our subconscious taking care of tasks for us. Information about hot stoves, for example, gets stored in our brains. Past experiences are tucked away for quick retrieval. When our hands get near a hot stove, it isn’t necessary for us to think about the danger, to make a rational assessment of the costs and benefits. Our subconscious reacts—pulling our hand back—before our conscious mind has even considered the danger.

25. The mental process involved with nuclear war and deterrence, however, is quite different. The hot stove response is built out of experience. Our fast brains know to be careful of hot stoves because our fingers have been burned in the past.

26. Nuclear deterrence, however, cannot be based on experience. No one has ever experienced a nuclear war (although hibakusha have experienced nuclear attack). Nuclear deterrence cannot rely on fast thinking. It must rely on slow thinking—the thinking that goes on when we deliberate. This is the kind of thinking that is especially prone to being short-circuited by emotion, prejudice, and other unnoticed influences. And a further difficulty arises. Since no one has experienced a nuclear war, nuclear deterrence must rely on imagination. In order to be deterred, an adversary has to be able to imagine a nuclear war.

27. Nuclear deterrence has three vulnerabilities. 1) It relies on deliberative thinking. People who are out of their minds or emotionally overwhelmed cannot be deterred. 2) Deliberative thinking can be easily be undermined or influenced by emotion. 3) Your adversary must be able to imagine a nuclear war.

28. In this connection, the failure of capital punishment to deter murder is not encouraging. Capital punishment relies on the same sort of event as nuclear war. Capital punishment requires the person you are trying to deter to imagine his own death—a horrible state he has never experienced. This is just like nuclear deterrence: you must imagine a horrible state you’ve never experienced. People sometimes say that nuclear deterrence is sure to work because the idea of nuclear war is so horrible. It would be, many people say, “The end of everything.” Capital punishment threatens an equally final result. It is, literally, the end of everything for the individual who is put to death. Yet capital punishment regularly fails. This is worth examining more closely. We can expect that people who tend toward wishful thinking—that is, whose imagination is strongly shaped by their desires—will not be strongly affected by nuclear deterrence.

29. Nuclear deterrence calls on unique characteristics and mental faculties. It is not like training a dog. It involved imagination, forethought, rational deliberation, and clarity. Conventional deterrence threatens an action that your adversary is familiar with, may even have experienced personally. Conventional deterrence has a much higher chance of working reliably than nuclear deterrence.

Conclusions

30. The evidence seems quite persuasive that nuclear deterrence is not very reliable. Other forms of deterrence (criminal deterrence, for example) fail regularly. Since even a single failure could result in a catastrophic nuclear war, it makes sense to demand a somewhat higher standard of effectiveness from nuclear deterrence.

31. Cold War crisis review. In order to insure that nuclear deterrence is very reliable—as reliable as some advocates for the weapons claim, it seems prudent to re-examine the record of the Cold War quite closely. Did leaders clearly discern a danger of nuclear war and yet press ahead with reckless and aggressive actions?

32. Criminal deterrence review. What is the evidence from capital punishment? Does it show that nuclear deterrence is likely to be reliable? A careful review of various kinds of criminal deterrence should be conducted, with implications for nuclear deterrence highlighted.

33. Neuroscience review. The latest science about how we think has interesting and important evidence to offer about the potential reliability of rational thought during periods of high stress. A full study of the latest neuroscience research should be conducted, keeping in mind its implications for nuclear deterrence.

34. Relying for security on nuclear deterrence appears to be a strategy with a high level of risk.

September 2013