PART I
WHY THE ARMS RACE DOESN'T MATTER

Everything about the atomic bomb is overshadowed by the twin facts that it exists and that its destructive power is fantastically great.
—Bernard Brodie

Chapter One
Why the Arms Race Doesn't Matter

A tired old Bear bomber, and a maneuvering hypersonic re-entry vehicle … [either one] can destroy New York just as well.
—Admiral Noel Gayler, former commander in chief of U.S. forces in the Pacific

The Fallacy of Weaponitis

The existence of nuclear weapons is a fundamental feature of the modern world. Their possession—by the human race generally or by any particular nation—definitively matters. They have completely changed the consequences of full-scale war between the dominant world powers, posing a threat to the very survival of civilization and the natural order. Correspondingly, international nuclear disarmament, or even unilateral denuclearization of individual states, would be singular historical events.

The fallacy of weaponitis lies in attributing great significance to the size and technical characteristics of the superpowers' nuclear stockpiles, and especially to the margins of each arsenal—incremental additions to or subtractions from the immense current force, such as building MX missiles or removing Pershing II and cruise missiles from Europe.^[1]

With conventional military technology, such concerns about weapons make sense. From the ancient discovery of the club, reenacted in 2001: A Space Odyssey, to the horrors of modern saturation bombing, the types and quantities of conventional weaponry have undeniably affected military and political power. Conventional arms races matter because conventional wars are processes of attrition. The guns, tanks, and planes of one side oppose and try to neutralize the weapons and fighters of the other. One side's forces must deplete those of the enemy before a threat of destruction can be posed to the enemy's inner society.

Because no single weapon or small arsenal of weapons determines the result, the quality and size of the overall fighting forces matter.

The side with more or better weapons does not always win, of course, because technical factors must share the military stage with psychological, social, economic, and political ones. In modern guerrilla warfare, for example, primitively armed local organizations sometimes defeat huge, highly advanced military powers. But the military balance has determined much of human history. Firearms helped European settlers conquer Native Americans. Germany's buildup of naval power prior to World War I increased the military threat to Britain. Large, technically advanced interventionary forces supported American power in Korea and other Third World conflicts after World War II.

At the beginning of the nuclear age, too, the weapons paradigm made sense. The atomic bomb was a new weapon, and it revolutionized war and politics. The reason was the immense power of an individual atomic weapon, especially the later hydrogen weapon—so powerful that a single warhead could destroy a city. As Bernard Brodie wrote in 1959: "People often speak of atomic explosives as the most portentous military invention 'since gunpowder.' But such a comparison inflates the importance of even so epoch-making an event as the introduction of gunpowder."^[2]

Consider what a single large warhead could do to Chicago:

One twenty-megaton nuclear bomb explodes just above ground level, at the corner of LaSalle and Adams. In less than one millionth of a second the temperature rises to 150,000,000 degrees Fahrenheit, four times the temperature of the center of the sun. A roar follows but no one is alive to hear it.

Chicago has disappeared. The crater is 600 feet deep, one-and-one-half miles in diameter. Within a five-mile radius, skyscrapers, apartment buildings, roads, bridges, trains, subways, planes, hospitals, ambulances, automobiles, gas mains, trees, earth, animals, people—all have vanished….

The fireball is hotter than five thousand suns. The firestorm roars out in all directions, absorbing all available oxygen, thereby suffocating or incinerating all the living in its path. Before it burns out it will devastate 1,400,000 acres and most of the people on them.

The firestorm is followed by the shockwave, the latter at close to the speed of sound. Then the mushroom cloud, reaching twenty miles in height, and the beginning of lethal radioactive fallout.^[3]

All weapons are subject to diminishing returns, but with weapons this powerful the point of saturation—when increasing the number or quality of weapons adds little to military potential or risks—was reached very soon, perhaps as early as 1955 and no later than the early

1960s, although the date is unimportant now. Both sides had by then acquired so much destructive power that only secondary importance would attach to any further quantitative or qualitative improvements in the leading weapons of the day. The same was true for even large reductions in weapon stockpiles. The weapons paradigm was already obsolete.

For with nuclear weapons, a nation's armed forces no longer must be defeated, or even seriously confronted, before its inner society can be destroyed, because the penetration of so few warheads is needed to accomplish the task.^[4] General war would no longer be a drawn-out process of attrition but an orgy of mutual devastation. Additional weapons on one side could do little to inflict greater damage on the other.

Even modern air attack with huge conventional bombs does not dispense with the task of defeating the enemy's armed forces. For example, the Allied bombing of German cities during World War II killed hundreds of thousands of people but did not make a decisive difference in the war. In more recent memory, the heavy bombardment of cities such as Hanoi and Beirut caused unimaginable human horror, but even in combination with the extensive bombing and shelling of other cities and villages, it did not completely destroy Vietnamese or (at least so far) Lebanese society.

In contrast, as McGeorge Bundy writes, "a decision that would bring even one hydrogen bomb on one city of one's own country would be recognized in advance as a catastrophic blunder; ten bombs on ten cities would be a disaster beyond history; and a hundred bombs on a hundred cities are unthinkable." Herbert York, another former high-level American official, concurs: "From one to ten are enough whenever the course of events is being rationally determined." Yet we urgently debate the composition of nuclear arsenals that now number not in the tens but in the tens of thousands. As the military historian Michael Howard notes, the amount of damage to be expected from a war that employs such weapons is so insensitive to the sizes of the nuclear arsenals held by the opposing powers that "the nuclear arsenals of the superpowers could be reduced by a factor of a hundred without affecting their capacity to destroy each other, and probably the rest of the world as well."^[5]

The Existential Balance of Terror

Overkill in itself did not make the arms race irrelevant. The arms race would matter had it produced either (1) a way to defend society against

nuclear weapons or (2) a way to control nuclear war and prevent escalation to all-out carnage. In either case, leaders might believe that their nations could survive or even win a nuclear war and hence might be more inclined to fight one. But because of the great power of the individual hydrogen weapon, no arsenal can provide an effective defense or a reliable means to prevent escalation.

Brodie predicted, almost from the beginning, that an effective defense would be impossible. To those who said that every new weapon always brought forth a suitable defense, he retorted that "after five centuries of the use of hand arms with fire-propelled missiles, the large number of men killed by comparable arms in the recent war indicates that no adequate answer has yet [been] found for the bullet." When a single warhead can demolish a city, a meaningful defense of society must be nearly perfect—a goal that no military technology, including President Reagan's Star Wars plan, can achieve (see Chapter 5). As former secretary of defense James Schlesinger—hardly a nuclear dove—said without qualification: "There is no realistic hope that we shall ever again be able to protect American cities."^[6] If we could build weapons to protect cities, these weapons would certainly matter. Since we cannot, MX missiles, orbiting battle stations, and arms control treaties are irrelevant; everything will probably be destroyed with or without them once the missiles are launched—unless, that is, nuclear war could be kept limited.

If a nation could build weapons or other devices to reliably control escalation, then even without a defense, leaders might imagine surviving or even winning a limited nuclear war. But the hydrogen weapon is too powerful to be controlled by any known means. Through blast, heat, radiation, and electromagnetic pulse, even a few nuclear weapons could destroy the leaders, organizations, and communication devices needed to control nuclear war. As we shall see, no one has devised a weapons system, a command mechanism, or an arms control treaty that could give leaders reliable control over events after a nuclear war erupted or one that could reliably prevent escalation once leaders lost control.^[7]

Noted strategic analyst John Steinbruner observes that "regardless of the flexibility embodied in individual force components, the precariousness of command channels probably means that nuclear war would be uncontrollable, as a practical matter, shortly after the first tens of weapons are launched, regardless of what calculations political leaders might make at the time." Desmond Ball's thorough study agrees that "control of a nuclear exchange would become very difficult to maintain after several tens of strategic nuclear weapons had been used." The conservative

strategic theorist Colin Gray regards the effort to wage nuclear war in controlled steps as little more than "suicide on the instalment plan."^[8]

Once central control of a nuclear war is lost, the fate of the earth will be left to dozens, perhaps hundreds, of isolated commanders with nuclear weapons at their disposal. No one can say what they will do. As Michael Howard observes: "The length of the war and the destruction it causes would be determined not by the number of weapons available but by the readiness of the belligerents to endure punishment in the hope of ultimate victory. A few missiles directed against carefully selected targets might cause the moral collapse of one belligerent, or indeed both; or they might fight on as grimly among the radioactive ruins as did the Russians in the ruins of Stalingrad or the Germans in the ruins of Berlin."^[9]

Leaders, then, cannot know just what will happen if nuclear war breaks out, regardless of the state of the arms race. They know only that they will probably not be in control and that mutual annihilation may well occur. That risk, that uncertainty, is what frightens and deters them. It may or may not be enough to prevent a holocaust. But pending the abolition of nuclear weapons, an effective defense against them, or a way to curb the risk of escalation, further arming or disarming cannot affect the terrible uncertainty that produces the balance of terror. That terror, as Brodie noted, is not at all delicate; it is existential, inherent in the existence of the "absolute weapon."^[10]

The nuclear warhead is the ultimate blunt instrument of human violence whose effects cannot be calculated in advance and whose use always risks the destruction of far more than is originally intended. After four decades of effort and hundreds of billions of dollars of investment, no one has found a way to sharpen the weapon to eliminate or even greatly reduce this risk. That much is well known. In a 1984 survey 90 percent of Americans agreed that "we and the Soviets now have enough nuclear weapons to blow up each other several times over"; 89 percent said that "there can be no winner in an all-out nuclear war; both the U.S. and the Soviet Union would be completely destroyed"; 71 percent believed that "there is no defense in a nuclear war"; and 83 percent thought that "a limited nuclear war is nonsense."^[11]

McGeorge Bundy coined the term "existential deterrence" to describe the implications of these basic nuclear realities. He observes that the "terrible and unavoidable uncertainties in any recourse to nuclear war" invalidate all strategies based on specific weapons and scenarios. ^{[ 12]}

All that matters is the possibility of uncontrolled escalation once the nuclear shooting begins. No one knows the likelihood of escalation or how to prevent it. So real political leaders, as opposed to abstract models of nuclear strategists, concern themselves only with the gross fact that catastrophic escalation could occur—that is all that they really know.

The whole complex labyrinth of nuclear hardware and the doctrines governing its use thus collapse to a single easily understood fact: each side has "large numbers of thermonuclear weapons that could be used against the opponent, even after the strongest possible preemptive attack." This reality is "essentially unaffected by any changes [in weapon deployments] except those that might truly challenge the overall survivability of the forces on one side or the other."^[13] The remote likelihood of such a challenge (see Chapter 2) makes deterrence inherent in the existence of the weapons.

If the strength of deterrence—the main force preventing nuclear war—is existential and thus relatively independent of the arms race, then logically the risk of nuclear war should be also. Like most nuclear doves, however, Bundy is well known for his opposition to "the competition in weapons systems," which, he dramatically writes, "is now itself becoming the largest single threat to peace." Similarly, George Kennan writes that "today we have achieved, we and the Russians together, in the creation of these devices and their means of delivery, levels of redundancy of such grotesque dimensions as to defy rational understanding"; that "the nuclear bomb is the most useless weapon ever invented"; and that "the relative size of the arsenals has no serious meaning." But Kennan too looks to these "redundant" and "useless" weapons for solutions to the nuclear problem, most dramatically with his well-known advocacy of a mutual 50 percent across-the-board cut in superpower nuclear arsenals.^[14] Yet if this "modest proposal" were ever carried out, could Kennan claim that deterrence and the risks of war had been fundamentally altered—rooted as they are in a "redundancy of grotesque dimensions" that even a 50 percent cut would hardly begin to undo?

The right, the center, and the left seem equally misguided in attaching great significance to which nuclear weapons are deployed or destroyed by the superpowers. Because no realistic changes in the pace, balance, or even direction of the arms race can alter the basic conditions of our nuclear existence, they should make little difference to the incentives to start nuclear war, the damage we can expect should a war occur, or the division of international political power. Even major steps by the superpowers to rearm (including Star Wars) or to disarm (including the nuclear

freeze or large cuts in nuclear arsenals) would leave the nuclear problem essentially unchanged, as we argue in more detail in succeeding chapters. The common tendency to identify the problem of nuclear war with the nuclear arms race is a logical fallacy that dangerously distorts nuclear politics by promoting technical fixes to what is overwhelmingly a political problem.

Chapter Two
What About First Strike?

What do Soviet leaders think U.S. submarine crews are going to do if they learn that the United States has been destroyed? Go to Tahiti and retire?
—Frank von Hippel

First Strike, You're Out

Many people who understand the uncertainty and terror of any nuclear war continue to worry about the arms race, fearing that "firststrike" weapons may undermine the mutual vulnerability on which stability depends. If one side could strike hard and fast enough to effectively disarm the enemy, then it could wage nuclear war without inviting its own destruction as a nation. Hawks worry about Soviet first strike weapons, doves about American ones. The argument is the same. Only the names change.

This striking parallelism came dramatically to light in 1985 with the publication of two articles—one in a conservative journal, the other in a progressive one. They appeared at the same time (July 1985) with nearly the same title ("First Strike, You're Out" and a slight variation), and they mirrored each other's arguments down to details. Daniel O. Graham and Gregory A. Fossedal warned in the American Spectator that "nuclear weapons are not unusable, not if you have enough of them, with enough accuracy, plus the capacity to find the necessary military targets and defend against the handful that survive." They worried that "it is now at least plausible that the Soviets will have a firststrike capability by 1990, and it is likely by 1995—unless there is a change in U.S. nuclear strategy and weapons programs." Daniel Ellsberg warned in the Progressive that "with a combination of Trident II and MX and Pershing II, plus our anti-submarine warfare, one side—namely,

the United States—would … have total coverage of the other side's retaliatory capability." Similarly, Howard Moreland wrote in the Nation several months earlier of "a U.S. first-strike capability" that "would upset the nuclear balance and could provoke the Russians to launch a pre-emptive strike." "The only way to avoid that situation," he said, "is to cancel the MX, Trident D-5, and Midgetman missile programs, and negotiate a mutual, easily verifiable moratorium on all ballistic missile flight tests."^[1]

You Can Run but You Cannot Hide

Despite all the worry, the fact remains that hydrogen warheads are too powerful, nuclear weapons platforms too diverse and well-defended, and societies too fragile for one side to rob the other of retaliatory capability. A successful first strike would require an overwhelming, near-simultaneous threat to almost all nuclear delivery vehicles or, even less plausibly, to warheads after launch. Even one surviving ballistic-missile submarine or a few dozen ICBMs or strategic bombers would sustain the essential risk—destruction of the attacking nation. First-strike weapons may add something to abstract military capabilities and even betray evil intent, but they cannot change the balance of terror.

First, neither side can find and destroy enough enemy weapons to even begin to prevent retaliation. Nuclear weapons platforms are diverse and highly protected: missile silos are very "hard," bombers can fly off on warning into the sanctuary of the air, and submarines, while "soft," are so difficult to locate in the open oceans that they are "immune to surprise attack."^[2] In theory, of course, future advances in technology could dangerously increase the vulnerability of submarines and other strategic weapons. But no such advances are in sight, and should they develop, countermeasures will probably be found to further hide and protect nuclear forces.

Even if one side could theoretically destroy all the enemy's strategic weapons, massive uncertainty—the key feature of deterrence—would still accompany any nuclear first strike. Except for the dropping of two atomic bombs on nearly defenseless Japanese cities over forty years ago, there is no operational experience with nuclear attack. Yet a successful first strike depends on the correctness of dozens of assumptions, many of which are, as MIT physicist Kosta Tsipis describes, "uncertain," "questionable," and "untestable." Everything must go according to plan the very first time, with nearly 100 percent effectiveness. Tsipis

asks, "If we cannot predict the performance of one missile against one silo with the kind of precision the formula for calculation implies, what if anything can we say ahead of time about the outcome of an attack against, say, 1,000 silos with warheads carried by several hundred missiles?"^[3]

Vital but untested human and political calculations must also play out as expected. The missile operators of the attacking nation must all faithfully execute the launch orders, knowing what the consequences will be for millions of people, and they must do so with perfect timing. And the enemy, after detecting the strike, must not launch its weapons on warning, which would doom even a technically perfect attack.

Mathias Rust, a West German student, dramatically demonstrated how fictional the supposed Soviet first-strike threat is by safely flying not an advanced B-1 or Stealth bomber but a defenseless private plane, with no cloaking devices, all the way to Red Square. He could have leveled Moscow if his plane had carried only one nuclear bomb. As the most conservative nuclear strategist we interviewed readily admitted: "Neither side can ever take away the other side's ability to destroy its cities and industries. If you have to you can do that with bombs in suitcases."^[4]

A Soviet first strike against the United States could succeed only if whoever survives to command U.S. forces decides not to destroy the Soviet Union and if the many lower-level U.S. commanders with control over nuclear weapons that can strike the Soviet Union decide the same thing. As the Soviets know, the chances of that are remote (and have little to do with the technical characteristics of the weapons anyway). A major study recently concluded that "a large-scale [first-strike] attack on strategic forces would cause so many civilian casualties that it would be difficult to distinguish from a deliberate attack on the population."^[5] Hence, vengeful retaliation is almost certain.

Thomas Schelling, perhaps the dean of strategic analysts, observes that "over a protracted period, both sides in a strategic nuclear war would preserve a more than sufficient capacity to do horrendous damage, and neither could hope to gain even a 'bargaining advantage' with any confidence." Most fears of a Soviet first strike, he points out, derive from nothing more than the existence of theoretically vulnerable, fixed land-based missiles in the United States. The remainder of the U.S. deterrent force has acquired a bizarre "vulnerability by association." Schelling knows there is no real first-strike danger to worry about, but

he proposes the elimination, unilateral if necessary, of land-based missiles in order to "clean the atmosphere psychologically."

If we unilaterally abolished our land-based missiles, we would instantly deprive a large part of the Soviet land-based missile force for [sic ] its raison d'être. It might look to them as if they had much less to preempt. They actually would not, because the U.S. missiles they might have preempted were redundant in the first place. Looking over a seascape inhabited by U.S. submarines and at bombers likely to be launched on warning, they would see, without the smoke and ruins, what would have been left over after they preempted.^[6]

In other words, the fear of a Soviet first strike has become so irrational that reducing the American deterrent force would help calm it!

Those who warn of an American first-strike threat are as misleading as those who warn of a Soviet one, and for the same reasons. When the new generation of U.S. strategic weapons is fully deployed, the United States may in theory pose a greater threat to the Soviet deterrent than the Soviets can pose in return, particularly considering U.S. advantages in antisubmarine warfare. But no matter how many MX and D-5 ballistic missiles, B-1 and Stealth bombers, air-launched cruise missiles. and other weapons the United States builds, striking first still means almost guaranteeing America's own destruction in return.

A major study by analysts at the Brookings Institution and the Lawrence Livermore National Laboratory recently confirmed that neither superpower has even the remotest ability to achieve a disarming first strike against the other, and that neither is in any danger of acquiring such an ability. Based on the superpower arsenals of early 1986, the study was designed to overestimate the first-strike capability of each side by a wide margin. The calculations unrealistically assume, for example, that the attacker's forces are 100 percent reliable and are all on full alert for war; that none of the victim's forces are at an unusual state of readiness; and that the attacker succeeds in destroying all of the enemy missile silos it attacks.^[7]

Nevertheless, the study found that "either side as the aggressor would have to face retaliation by at least roughly 3,000 surviving warheads." That is enough to utterly destroy either society many times over. In fact it is doubtful, as the study notes, that either side even has 3,000 targets its opponent's military would consider worth retaliating against; beyond 1,500–2,000 targets, "the marginal contribution to the overall effect would be so small that a prudent military command would presumably

choose to preserve any remaining weapons rather than to expend them in redundant and unnecessary retaliation." An aggressor would also have to assume that after detecting the attack, the victim would launch at least some missiles before they were destroyed, an action that would increase the retaliatory threat even further. No numerical change in superpower forces or technological development now on the horizon would greatly change the threat of cataclysmic retaliation following a first-strike attempt. And as we will see in Chapter 9, even reductions in the superpower arsenals far greater than any now contemplated likewise would change little about the feasibility of a first strike. As George Kennan writes, "The fact is that there are today, in this threatened world of ours, no 'windows' of vulnerability that could be opened or closed. We are vulnerable—totally vulnerable. There is no way that could be changed."^[8]

Though the public debate focuses on so-called first-strike weapons, specialists warn of other technical factors that supposedly produce additional "incentives" for launching nuclear weapons first. The current fashion in strategic analysis is to warn of the dangers of "decapitation"—that is, attacking the enemy's nuclear command and control network. Though as we have seen, neither side can even begin to destroy enough of the other's missiles and bombers to significantly affect its military capabilities, either side could quickly devastate the other's system for commanding and controlling those forces. A few high-altitude nuclear bursts, for example, could knock out radar and communications through electromagnetic pulse and other effects. Shortly thereafter direct attacks on ground-based radar and communications facilities could finish the job. Attacks on Washington and Moscow could kill national leaders and destroy command centers such as the Pentagon. Most other command centers are equally vulnerable, and though both sides have alternative command arrangements, including airborne command posts and hardened communications devices, none would likely survive long in the event of a large-scale strategic war.^[9]

Many scholars argue that in a crisis the fear of sudden decapitation could give either superpower a reason to launch first in an effort to destroy the enemy's command system while its own is still intact. John Steinbruner, for example, writes that the vulnerability of command and control "is probably the single most important and dangerous … incentive" to launch first, and that modernizing the command system could reduce that incentive.^[10] Though not strictly an instance of weaponitis (since command and control systems are not technically weapons),

that view is simply another version of the fallacy that technological factors can significantly stabilize or destabilize the nuclear stalemate.

Obviously, trying to decapitate the enemy is no less foolishly suicidal than trying to disarm it. As seventeen highly regarded strategic analysts (including Steinbruner) write, even after a completely successful surprise decapitation attack

there is every reason to believe that there would be retaliation…. The U.S. government has often stated that provisions are always in place that guarantee large-scale retaliation under the most dire of circumstances. Given the vast forces (primarily those at sea) that would survive any attack on command, the possibility of predelegated contingency arrangements, the possible existence of high-level command elements and communications links of which the attacker has no knowledge, and the multitude of communication channels that are likely to survive or that could be established, these official statements should be taken at face value by a potential attacker.^[11]

Even if no command channels survived, at least some of the many commanders with nuclear weapons at their disposal would almost certainly launch their weapons on their own authority. As Frank von Hippel asks, "What do Soviet leaders think U.S. submarine crews are going to do if they learn that the United States has been destroyed? Go to Tahiti and retire?" In fact, isolated commanders acting chaotically, without central coordination, might well inflict far greater damage on an attacker than an intact central command would order. Retaliation following a decapitating first strike could "actually result in the 'assured destruction' of Soviet society—in contrast to the less apocalyptic retaliatory options that are likely to receive serious consideration in a 'deliberate' centrally coordinated response following full assessment of an attack." The same applies in reverse; Soviet leaders, like their American counterparts, need not fear that an American decapitation attack would rob them of the ability to retaliate in kind.^[12] No plausible development in the arms race could make decapitation or indeed any first strike other than what it is today: suicide.

Suicide for Fear of Death

Still, specialists claim that in at least one extreme circumstance a first strike might logically be considered a rational move: when one side feels that nuclear war is inevitable no matter what steps it takes. If a nuclear war cannot be avoided, the argument goes, then preemption—striking before the enemy does—might be the best way to fight it. Richard K.

Betts argues that "there are few plausible circumstances in which striking first could seem to make sense—at least for the superpowers. … when one believes the enemy is about to strike … is the only situation in which the initiator would have reason to believe that starting a nuclear war could cost less than waiting to try other options."^[13]

But as Stansfield Turner writes, "I cannot imagine a Director of Central Intelligence [Turner's former position] ever having anything approaching 100 percent confidence in his prediction that the Russians were truly going to attack. The President would be faced with a choice between the total probability of nuclear destruction" if he launched a preemptive strike that brought down the almost inevitable Soviet response and "some lesser probability" if he waited. Richard Ned Lebow, in an unusually sensible discussion of first strike, agrees that "the judgment that the other superpower is about to strike can never be made with full certainty. … the side that strikes first risks making its fear of nuclear war unnecessarily self-fulfilling. … preemption is an altogether irrational act."^[14]

Preemption is, in Bismarck's phrase, "suicide for fear of death." No arms control treaties are needed to ensure that preemption is suicidal and totally irrational, and this reality cannot be altered by any plausible development in the arms race. It is an existential fact of life in the nuclear age."^[15]

Itchy Fingers on the Button

Finally, we must consider perhaps the most horrifying scenario: One side receives evidence that its enemy has already launched a nuclear attack. Satellites report missile launchings, radars warn of missiles en route, and—in the worst case—leaders receive reports of actual nuclear explosions. In that event, if in no other, many observers contend, the hardware deployed by each superpower could prove fateful—strongly influencing whether the side receiving the warning of attack would make a hasty decision to launch its own strategic nuclear weapons.

The seventeen strategic analysts quoted earlier, for example, state that because of the extreme vulnerability of nuclear command and control, the United States has been "forced" to seriously consider "a prompt launch" of its strategic forces (very "prompt," within several minutes) should it receive warning of a Soviet attack in progress. Why? Because waiting until the attack is over "runs the risk that the command system may not have sufficient coherence … to execute a coordinated counterattack."^[16] That is, even if national leaders survived, they would

probably not be able to communicate with and control their forces well enough to determine precisely what kind of retaliation took place—for example, how many weapons were fired and where they were aimed. Devastating retaliation would, as we saw, almost certainly take place; but it would not be under firm central control. To ensure that retaliation occurs under central control, the analysts contend, it should perhaps be ordered almost immediately after warning of an enemy attack.

The seventeen analysts recognize that if leaders decide to launch hastily, they would probably feel strong pressures to do so massively, with "thousands of nuclear warheads." Leaders would not receive reliable estimates of the extent and nature of the attack in the few minutes available for decision before they are killed or control otherwise disintegrates; "too much would be at stake to give the attacker the benefit of the doubt." Furthermore, because of "the limited endurance of the backup command network, decision makers could not expect to employ withheld forces in a coordinated fashion. In all likelihood, the selection of a small [attack] option would mean that all other strategic options would be relinquished forever … [which] would encourage selection of a major attack option."^[17]

In discussing prompt launch, like so many other questions, so-called experts dispense dangerously misleading analyses that fly in the face of common sense. The danger of losing central control over nuclear forces is hardly a rational reason to launch the missiles hastily, virtually guaranteeing devastation on a planetary scale on the basis of warning of an enemy attack. Such a warning, of course, can always be mistaken or misleading, even after indications that nuclear explosions have actually occurred. Conceivably, technical malfunctions could produce a warning of attack when no attack was in fact occurring. A limited attack—perhaps even accidental—could also be misinterpreted as a full-scale first strike in the few frenzied minutes available for analysis. As Lebow points out, prompt launch demands

an organizational capacity to make an almost instantaneous decision to respond to attack coupled with an entirely foolproof mechanism to safeguard against false alerts. Both requirements are unrealistic. There is no way of guaranteeing that a complex system of sensors, computers, related software, and human operators can function error-free all of the time…. A hair-trigger—and this is a fair description of quick-launch options—would multiply the probability of error by several orders of magnitude [that is, a hundred times].^[18]

To fire thousands of strategic nuclear weapons suddenly because one thinks the enemy has launched a major attack would be the greatest war

crime in history, and perhaps the last one. It would also be totally irrational, even from the most amoral, self-interested point of view, particularly since, as we saw, devastating retaliation is virtually assured no matter what leaders decide.

The complex analyses of first strike by the right, the left, and the professional experts have only obscured the simple truth. No state ever has the slightest rational incentive to launch strategic nuclear weapons first or in haste, and certainly not before confirming beyond any doubt the explosion of many nuclear warheads on its territory. Even in that instance elementary morality would dictate that retaliation be withheld, since it would be little more than the mass murder of innocents—on a scale that would make the Nazi slaughter seem modest by comparison—with no prospect of accomplishing any legitimate goal. Of course the threat of retaliation may be important to mutual deterrence and the prevention of nuclear war; but that threat, as we emphasize throughout this book, is existential, inherent in the existence of the weapons. The threat to launch first or to launch hastily is not important to deterrence. Does it matter if an attacker is destroyed minutes, hours, days, or even weeks after it launches its missiles? Morton Halperin gives much wiser advice: avoid launching nuclear weapons suddenly, perhaps by making it physically impossible, while maintaining the option of retaliating after considerable time for reflection and perhaps negotiation.^[19]

The best way to reduce the danger of a first strike is to encourage wider recognition that launching nuclear missiles first or launching them precipitously is insane regardless of the hardware on either side. In addition, our overwhelming concern should be to avoid the extreme political conditions in which Soviet or American leaders could actually believe that their counterparts had decided to launch World War III—when leaders, seduced by the prevailing nuclear nonsense, might make a split-second decision that could terminate human civilization. In the nuclear age, the desperate feeling that the ax is about to fall is the key factor—not the size, speed, or accuracy of the ax.

Chapter Three
What about Credibility and the Firebreak?

SENATOR GLENN :

I get lost in what is credible and not credible. This whole thing gets so incredible when you consider wiping out whole nations, it is difficult to establish credibility.

SECRETARY [OF DEFENSE] BROWN :

That is why we sound a little crazy when we talk about it.
—Senate Committee on Foreign Relations,
Hearing on Presidential Directive 59,
September 16, 1980

Nuclear Credibility:
Do We Need or Want It?

During the brief American nuclear monopoly after World War II, many believed that the United States might use nuclear weapons again in a war or political crisis. No country could retaliate, and the 1945 nuclear attacks on two defenseless Japanese cities as well as the earlier killing of hundreds of thousands of German and Japanese city dwellers with conventional weapons showed that moral considerations might not inhibit American actions.

As the Soviets developed their own nuclear arsenal, however, American strategists worried that the fear of retaliation could reduce the United States to a hobbled nuclear giant. They became obsessed with what was politely termed the "credibility" problem: how to make the world believe that American nuclear threats might actually be carried out even though the result could be the destruction of the United States. The belief that special weapons are required to solve this problem is one of the major driving forces of, and rationales for, the nuclear arms race.

In most cases, however, credibility is not problematic and does not in fact require special weapons. Should any country launch a general nuclear attack on the United States, few have ever doubted that the U.S. military would do its best to return fire in kind, regardless of the weapons available. There would be little left to lose. Here the credibility problem reduces to the technical matter of whether retaliation would be

possible in the rubble and chaos. It would (as we argued in Chapter 2), if only through the vengeance of surviving nuclear weapons commanders on land or at sea. Even a limited Soviet first strike on the United States would probably kill millions of Americans; the United States would surely retaliate and would have many credible options for doing so regardless of the details of its nuclear arsenal.

No great credibility problem attends the defense of Western Europe, either. The specter of a Soviet blitzkrieg is by now, certainly, largely mythical, and in any case no new or special weapons are necessary to deter it. The Soviet Union has little to gain and everything to lose by invading its powerful trade and financial partners in Western Europe. The Soviets have enough trouble controlling their Eastern European satellites, such as Poland, and could not even conquer Afghanistan. NATO has consistently outspent the Warsaw Pact, enjoys a large technological lead, and would occupy the favorable position of defender in the event of a Soviet invasion. The Warsaw Pact's much ballyhooed numerical superiority in tanks and troops means little when readiness, technology, and alliance loyalty are considered.^[1]

Besides, invading Soviet generals would face more than one thousand European-owned nuclear warheads, which are slated for major expansion and improvement over the next few years. Many in the United States dismiss the comparatively small French and British nuclear arsenals. That is pure weaponitis. No country actually faced with the possibility of a nuclear attack by France or Britain (or any other nuclear power) would take the prospect lightly. And, as Earl Ravenal observes, "even the whiff of American nuclear retaliation is probably enough to keep the Soviet Union from invading Western Europe."^[2]

The United States, however, wants its nuclear weapons not only to defend itself and its European allies but also to project American power into the Third World. One example, among the many discussed in Chapter 6, is the 1973 U.S. nuclear alert during the Arab-Israeli war. Nuclear threats in the Third World may indeed pose credibility problems. But should we seek to make such threats more believable? Obviously, even a noble goal, such as defending a small country against aggression, does not justify endangering the planet. And, as we will argue, U.S. motives have often been less than noble—as in 1973, when Henry Kissinger created a crisis by giving Israel permission to violate the cease-fire he had just negotiated in Moscow.

Special efforts to enhance nuclear credibility seem to us either unnecessary

or illegitimate. But for the purpose of argument let us grant that there is a real credibility problem and see if special weapons can solve it.

Rationalizing the Irrational

Allan S. Krass and Matthew Goodman write that for nuclear threats to be credible, "the capability and the will must exist for carrying them out…. Nuclear threats cannot be a rational tool of policy unless nuclear war itself is also rational" or at least appears to be. "This defines the problem that nuclear strategists have been trying to solve since 1945: can nuclear war be made rational? " Colin Gray, among others, argues that only when "victory is possible" will American nuclear threats appear rational and believable, because unless the U.S. homeland is defensible, carrying out the threats would amount to suicide. With the right weapons (strategic defenses and missiles aimed at the Soviet state apparatus), he contends, the United States could survive all-out nuclear war and hence credibly threaten to use nuclear weapons.^[3]

Other observers, who recognize that victory is in fact impossible, seek only, as Earl Ravenal explains, to "limit damage to 'tolerable' levels of casualties and destruction … so an American president can persuade others that he would risk an attack on the U.S. homeland, or that he could face down a threat to attack that homeland."^[4] But of course the damage cannot be made tolerable. No matter what weapons each side fields, large-scale nuclear war will probably destroy both superpowers and much more; hence threats to unleash it are altogether irrational and, in all but the most dire circumstances, incredible. Other strategists thus contend that the only way to make nuclear threats believable is to threaten action that is not suicidal—that is, to threaten a very limited nuclear war and to suggest that it can be kept limited. In the words of Robert Osgood, credibility "requires that the means of deterrence be proportionate to the objectives at stake."^[5]

One approach to enhancing "limited nuclear options," as former secretary of defense James Schlesinger called the doctrine, is to reduce the yields of the weapons so they do less damage. Another approach is to deliver them more accurately, which minimizes collateral damage and permits the use of smaller-yield warheads. One can redesign warheads to reduce collateral damage—the essential purpose of the neutron bomb. One can aim weapons away from population centers, and so on. But because no one knows how to keep nuclear war limited, credibility

remains as problematic as ever. There is, however, another way to achieve it.

The Doomsday Machine

In the black-comedy film Dr. Strangelove, a Soviet "doomsday machine" automatically destroys the planet when American nuclear bombers, launched by a deranged general, attack the Soviet Union. That neatly solves the credibility problem. Of course, as far as we know, there is no isolated contraption consisting of nuclear weapons wired to an automatic tripping device. But according to Paul Bracken and others, current nuclear command and control procedures may have created, in effect, a doomsday machine.

In Europe, Bracken contends, "the NATO strategy of relying on nuclear weapons is politically and militarily credible because the governing command structure is so unstable and accident-prone that national leaders would exercise little practical control over it in wartime." That is, the nuclear command structure and the weapons it controls are not so much "designed to gain battlefield advantage through attrition of enemy forces" as to "enforce deterrence by necessitating that any war be nuclear." Nuclear weapons are thoroughly integrated into NATO (and Warsaw Pact) conventional forces, including, by one estimate, several hundred nuclear land mines, over one thousand nuclear artillery rounds, several thousand surface-to-surface missile warheads, several thousand more nuclear bombs and missiles for delivery by aircraft, hundreds of surface-to-air missiles and aircraft-launched antisubmarine weapons, and, for a while, Pershing II and cruise missiles.^[6] U.S. naval forces in the area are assumed to carry enormous tactical nuclear arsenals as well. Many individual fighting units, which could easily become isolated in wartime, are equipped to fire nuclear weapons.

In peacetime most nuclear warheads are stored at special sites to avoid accidental or unauthorized use. But because the stored weapons are vulnerable to preemptive destruction, during the development of a crisis "there are likely to be strong pressures within NATO for a general release of weapons to NATO military forces assigned the task of using these weapons," spreading six thousand nuclear warheads "from the North Sea to the plateaus of eastern Turkey." Since the Soviets could paralyze the command system, thus preventing the transmission of the codes needed to fire NATO's weapons, Bracken believes that "a strong pressure exists to release any needed codes at the same time that the

weapons are dispersed from their storage sites."^[7] Field commanders could then launch nuclear weapons on their own initiative early in a conflict. The same is presumably true of Soviet nuclear forces.

Compounding the problem, a number of nations in addition to the United States could fire NATO nuclear weapons in their possession without the U.S. approval theoretically required. Indeed, during the 1974 Turkish invasion of Cyprus, the Pentagon prepared U.S. marines for a helicopter assault on nuclear warhead storage facilities in Turkey, if necessary, to ensure U.S. physical control over the weapons. The warheads normally stored aboard quick-alert aircraft in Greece and Turkey were removed to safe storage. In wartime the United States might be unable to hold on to nuclear weapons deployed on allied territory if local forces were determined to seize them. And if the warheads were intentionally dispersed, it might be impossible to get them back where they belonged.

Under mutual high-alert conditions, as Bracken warns, "a single Turkish pilot could trigger World War III, as could a Soviet naval officer in charge of a Shaddock cruise missile." What we have is "a strategy of deterrence by massive duplication of a nuclear hair trigger."^[8]

Others besides Bracken have noticed the doomsday threat, and it is by no means limited to Europe. In a 1985 review of "Who Could Start Nuclear War," the widely respected Center for Defense Information (CDI), composed largely of former high-ranking American military officers, observes:

There is a wide discrepancy between those people, primarily government officials, who have the authority to order a nuclear attack, and those persons, primarily military officers, who actually have the physical capability to fire nuclear weapons…. During a military crisis, many people in the U.S., U.S.S.R., Britain, France, and China will likely have the authority to order a nuclear attack or the physical capability to start a nuclear war. In a surprise attack, government officials in the U.S. and U.S.S.R. may not have time to decide whether to order a retaliatory nuclear strike before their own nuclear forces are hit and they themselves are killed. Even if the order to retaliate is given by the political leadership, destruction of communications channels might impede transfer of the launch order to the appropriate military commanders. Prudent military doctrine suggests, therefore, that each nuclear nation has developed procedures for the transfer of authority to release nuclear weapons to others below the highest government officials.

As CDI notes, "commanders would have the tendency to issue [the codes needed to unlock nuclear weapons] early, before an actual attack against civilian and military officials occurred. Once unlocked, the

weapons would be ready for use by lower level military personnel." Moreover, "there are no [electronic locks on] any of the U.S. sea-launched nuclear weapons," which can be carried by 85 percent of U.S. ships. "A captain and several officers on any of the U.S.'s 37 ballistic missile submarines could launch their nuclear weapons at any time without receiving permission from the President or NCA [National Command Authority]."^[9] So could the crew of other nuclear-armed U.S. Navy ships, including the aircraft carriers that are routinely rushed to Third World crises and war zones.

Another group of respected mainstream analysts notes that "the U.S. nuclear weapons deployed in densely populated Europe are numerous, diverse (both in warheads and delivery systems), and integrated into conventional units." They conclude that "negative control is therefore complex and difficult"—where "negative control" means "preventing the unauthorized or accidental use of nuclear weapons." They add that what Bracken terms the "regional doomsday machine" in Europe is really a planetary doomsday machine: "Because the loss of C³ I systems in the theater would undoubtedly affect strategic operations on both sides—and because theater war could easily spill over onto Soviet territory—rapid escalation to strategic conflict must be considered a very likely possibility if deterrence were to fail."^[10]

To cite only one more study, Barry Posen suggests another reason to suppose that a doomsday machine of some kind exists. He observes that Soviet strategic nuclear forces, both on land and at sea, would probably be attacked, or would justifiably fear attack, by NATO forces during a large-scale conventional war in Europe, regardless of whether NATO intended such attacks. For example, whereas U.S. strategic-missile submarines hide in the open sea for protection, two-thirds of the Soviet Union's strategic submarines are believed to be stationed in a defended sanctuary in the Barents Sea, protected by two-thirds of the Soviet nuclear-powered attack submarines as well as by their best surface vessels. In the event of a conventional war, NATO would almost certainly hunt down Soviet attack submarines there to protect vital sea-lanes. Soviet attack subs are so difficult to distinguish from Soviet strategic nuclear missile subs that some of both types would probably be sunk by NATO forces. The U.S. Navy might deliberately attack strategic subs in any case since, in addition to long-range ballistic missiles, they carry torpedoes and mines capable of sinking NATO attack subs and surface ships. Also, the chance to degrade Soviet nuclear capabilities under the cover of conventional operations might be too great to resist. As the Navy's director

of Command, Control, and Communications once acknowledged, "In a conventional war all submarines are submarines. They are all fair game." Posen comments: "The Soviet Union could see … sinkings [of nuclear missile submarines] as a deliberate attempt to degrade the Soviet Union's nuclear retaliatory capability rather than as 'accidents' to be accepted with equanimity."^[11] In the context of a growing conventional threat to their other nuclear forces as well, the Soviets might be tempted to use nuclear weapons in a desperate attempt to stop the war and save their nuclear deterrent.

According to Posen, part of the doomsday problem results from reckless NATO conventional war plans, which grant little or no attention to the avoidance of inadvertent nuclear escalation. Those plans are perhaps one more element of a deliberate strategy to ensure that any war must be nuclear, thereby shoring up the credibility of the nuclear deterrent. As Posen writes, "The 'threat to lose control' is an important element of NATO's flexible response strategy."^[12]

The Firebreak

Before examining whether the arms race really matters much to nuclear credibility, we must mention the nuclear "firebreak," the conceptual prism through which much of the arms control community and the peace movement view the issues under discussion. Michael Klare writes:

The only existing barrier to … escalation [from conventional to nuclear war] is a moral and psychological firebreak—the widely shared perception that nuclear weapons are different from all other weapons, and that their use could unleash a chain reaction of strikes and counterstrikes leading to total world destruction. So long as this firebreak remains wide and secure, so long as the distinction between nuclear and conventional arms remains sharp and unambiguous, potential combatants will retain an incentive to stay on the non-nuclear side of the divide, no matter what their prospects are on the conventional battlefield. But if that distinction were to fade or disappear, the inhibition against nuclear escalation would decrease and the risk of global annihilation would skyrocket.

The conventional/nuclear firebreak is crucial, because as Alain Enthoven observed in 1965, "There does not appear to be another easily recognizable limitation on weapons—no other obvious 'firebreak'—all the way up the destructive spectrum to large-scale thermonuclear war."^[13]

Klare contends that the superpowers are eroding the firebreak from both ends. Near-nuclear conventional weapons such as precision-guided

munitions, cluster bombs, and new explosive technologies "could possess a destructive potential comparable to that of low-yield nuclear munitions." Low-yield tactical nuclear weapons are simultaneously reducing the destructiveness of nuclear warheads to near-conventional levels. The neutron bomb behaves even more conventionally, killing soldiers but minimizing collateral damage. Dual-capable weapons span the firebreak by permitting the same weapon to shift easily between conventional and nuclear ammunition. For example, tactical aircraft (such as F-4, F-15, F-16, A-6, A-7) can carry both kinds of bombs, tactical missiles (such as Lance, Terrier, ASROC, Tomahawk, Nike-Hercules) can fire both kinds of warheads, and artillery pieces (such as 155-millimeter and 8-inch guns) can lob both kinds of shells. As U.S. Army General Louis Wagner told Congress in 1980, "We use the same troops and we have the same cannons available to do the conventional and nuclear job." Klare believes that the growing deployment of such weapons "diminishes the problems involved in moving from the conventional to the nuclear realm—thus narrowing the 'pause' or discontinuity that separates one from the other."^[14] Similar trends can be observed in Soviet forces.

By now almost all types of conventional weapons—from the giant strategic bomber to the lowly land mine—have nuclear counterparts. The military can wipe out tanks and soldiers, sink ships, shoot down planes, interdict supplies, disrupt command, and do almost everything else with nuclear weapons. Tactical nuclear weapons systems are generally deployed with or made available to U.S. conventional military forces wherever they are based on land or at sea, effectively rendering the entire U.S. military dual-capable.

The Two Sides of the Coin

Credibility and the firebreak both come down to the same question: Can nuclear war be made to resemble conventional war? Conventional war, even with modern weapons, is limited; it is unlikely to end up in the complete physical destruction of either side.^[15] The question is whether nuclear war can be limited in this way, whether the gulf that separates it from conventional war can be narrowed.

If so, then nuclear threats, like conventional ones, can be credible, for they are not threats to commit suicide. And the firebreak will be narrow, because nuclear and conventional war will not seem so different. Conversely, if the gulf between the two kinds of war stays wide, then nuclear

threats lack plausibility unless a state's survival is threatened, and the firebreak will be vast and clear. To put it another way, narrowing the firebreak enhances nuclear credibility. Widening it undercuts nuclear threats. The nuclear weapons systems that some advocate to boost credibility are precisely the same ones that others criticize for eroding the firebreak (e.g., the neutron bomb).

The debate about such matters, then, does not involve fundamentally different ways of looking at nuclear war. The main point of contention is values . Those concerned about credibility want to have a real nonsuicidal option for starting nuclear war so they can plausibly threaten it not only to strengthen deterrence but also to advance American military and political goals. Those concerned about the firebreak want to make it as difficult as possible for anyone to start a nuclear war and do not want states threatening such a war for political purposes. Both groups are discussing the same thing: the width of the gulf between conventional and nuclear war. One group wants it narrow, the other wide.

Suicide on the Installment Plan

But the gulf between conventional and nuclear war cannot in fact be made much narrower or wider than it already is. Hence, neither a continued arms race nor further arms control—with an important exception—can much affect the level of nuclear credibility or the width of the nuclear firebreak. Credibility is inherently questionable and the firebreak inherently wide because nuclear war is different from conventional war. The gulf between the two is intrinsically huge since any extensive use of nuclear weapons would be, as McGeorge Bundy said, "a disaster beyond history."^[16] Even a very limited nuclear attack would inescapably risk escalation to total annihilation, whereas the use of conventional weapons does not.

As we have said, no one has devised a way to control a nuclear war and thereby confidently prevent escalation to large-scale nuclear exchanges, and no one has found a way to defend society against such exchanges.^[17] If the arms race or arms control ever do produce a way to reliably control escalation or defend society—both remote prospects—then the firebreak might narrow and the credibility of nuclear threats might increase (although probably not by very much, given the uncertainties and the destructiveness of nuclear weapons).

We consider the prospects for a defense against nuclear weapons in Chapter 5. As for controlling escalation, "is it realistic to expect that a

nuclear war could be limited to the detonation of tens or even hundreds of nuclear weapons, even though each side would have tens of thousands of weapons remaining available for use?" As we have seen, "The answer is clearly no," because, as former secretary of defense Robert McNamara explains, expecting nuclear war to remain limited

requires the assumption that even though the initial strikes would have inflicted large-scale casualties and damage to both sides, one or the other—feeling disadvantaged—would give in. But under such circumstances, leaders on both sides would be under unimaginable pressure to avenge their losses and secure the interests being challenged. And each would fear that the opponent might launch a larger attack at any moment. Moreover, they would both be operating with only partial information because of the disruption to communications caused by the chaos of the battlefield (to say nothing of possible strikes against communications facilities). Under such conditions, it is highly likely that rather than surrender, each side would launch a larger attack, hoping this step would bring the action to a halt by causing the opponent to capitulate…. It is inconceivable to me, as it has been to others who have studied the matter, that "limited" nuclear wars would remain limited—any decision to use nuclear weapons would imply a high probability of the same cataclysmic consequences as a total nuclear exchange.^[18]

Klaus Knorr agrees that "a large risk of continuous escalation, intentional or inadvertent, cannot be excluded once nuclear exchanges begin. Initiation means accepting that risk of unwanted escalation." As Leon Wieseltier writes, "The worst may not happen first, but it may happen fast." General David C. Jones, former chairman of the Joint Chiefs of Staff, concurs: "I don't see much of a chance of nuclear war being limited or protracted."^[19]

Unless someone finds a defense or a means of control, nuclear war cannot be made to resemble conventional war. It cannot be made rational in any sense. No amount of weaponry can allow a nation to win a nuclear war, to fight one like a conventional war, or to limit damage. It is possible, as the flexibility advocates argue, to start a nuclear war without immediately destroying everything. But that ability has always been available to any substantial nuclear power. One always has the option of firing fewer or more weapons, and relatively low yield weapons have been available from the beginning. The bomb that destroyed Hiroshima yielded only about 12–15 kilotons, very little by today's standards. During the 1950s the United States deployed warheads with yields in the low kiloton and even subkiloton range, easily small enough to provide limited and flexible nuclear options.^[20]

And the real issue is not how the war begins but how it ultimately will end . As Desmond Ball concludes,

Given the impossibility of developing capabilities for controlling a nuclear exchange through to favourable termination, or of removing the residual uncertainties relating to controlling the large-scale use of nuclear weapons, it is likely that decision-makers would be deterred from initiating nuclear strikes no matter how limited or selective the options available to them . The use of nuclear weapons for controlled escalation is therefore no less difficult to envisage than the use of nuclear weapons for massive retaliation.

Although, as we shall see, decision makers may in fact initiate nuclear strikes, Ball is surely correct to insist that for practical purposes limited nuclear options differ little from unlimited ones and are only marginally more credible. The fundamental problem that has dogged credibility—the suicidal nature of any threat to start nuclear war—remains. In brief, "the problem of making nuclear war rational has no solution at all."^[21]

Those worried about the firebreak also exaggerate the consequences of changes in weapons. With or without fuel-air munitions and neutron weapons, any decision to go to nuclear war is unmistakably a decision to court mutual suicide. True, Ronald Reagan once proclaimed the neutron bomb "conventional." But, as we discussed in Chapter 4, that does not necessarily mean he would underestimate the consequences of actually using it in a war with the Soviet Union. In any case, as Klare emphasizes, the firebreak is ultimately psychological and moral. Whatever technological developments occur, it could become stronger or weaker depending on how leaders think about nuclear war. Long before the neutron bomb was developed, American leaders were publicly calling nuclear weapons conventional to lend credibility to nuclear threats. In 1955 President Eisenhower said that he saw no reason "why [nuclear weapons] shouldn't be used just exactly as you would use a bullet or anything else." Richard Nixon, then vice president, added that "tactical atomic weapons are now conventional and will be used against the targets of any aggressive force ."^[22] The answer to such statements now, as in the 1950s, is to insist on the truth—nuclear war is inherently irrational, uncontrollable, and unwinnable—and to ensure that civilian and military leaders recognize it.

The only technical factors that can greatly alter credibility and the firebreak are those that are responsible for the doomsday machine: (1) the fragility of nuclear command and control and (2) the presence of nuclear weapons at the front lines in Europe, Korea, and the Middle

East, at sea, and elsewhere. As we have seen, not much can be done about the fragility of command and control. As Ball's painstaking study concludes, "Command-and-control systems are inherently relatively vulnerable…. The allocation of further resources to improving the survivability and endurance of the strategic command-and-control capabilities cannot substantially alter this situation…. The capability to exercise strict control and co-ordination would inevitably be lost relatively early in a nuclear exchange." In Europe, another group of experts adds that "conditions … are not congenial to major improvements in C³ I."^[23] And whatever can be done has less to do with the race to build or control weapons—the focus of the nuclear debate—than with the organization of the command system. For example, NATO presumably could make it much more difficult for field commanders to obtain the authority and technical ability needed to detonate nuclear warheads in the event of a war in Europe.

Much more can be done about the way nuclear weapons systems are scattered around the globe. Although modest changes would not make much difference, radical changes could. If NATO units guarding the central European front did not have nuclear weapons, then their involvement in conventional battles could not erupt spontaneously into nuclear war. If American aircraft carriers were not dual-capable, or if they were not sent into war zones, their commanders could not ignite a cataclysm. A thorough removal of tactical nuclear weapons from flash points for conventional war could significantly widen the nuclear firebreak, though possibly at the expense of credibility. Ironically, this one aspect of superpower arms control that bears directly on the risk of global holocaust has received almost no attention during the public debate in the United States.

The Politics of Credibility

The debates about credibility and the firebreak, like first strike, have been suffused with weaponitis. The most important factors, as usual, are not technological but political. States will consider crossing the nuclear firebreak, and their threats to do so will be taken seriously, so far as (1) they believe vital interests are at stake and (2) they do not mind risking the lives of hundreds of millions of people to pursue those interests. As Blechman and Hart observe,

Only in certain places and very special circumstances might attempts to manipulate the risk of nuclear war be credible. For the United States, these

probably include military contingencies involving Europe, Japan, and Korea, for which a willingness to make first use of nuclear weapons has long been articulated policy. Elsewhere, such U.S. threats probably would only be credible in the Middle East (including the Persian Gulf), and only when taken in response to Soviet, not local, actions. For the Soviet Union, military challenges to its position in Eastern Europe would no doubt trigger credible threats of nuclear war, as would a serious military confrontation with China in Central Asia. In all other places, Soviet nuclear threats would only be credible in the context of direct military confrontation with the United States and, even then, would depend on the circumstances which precipitated the conflict to begin with.^[24]

The historical record shows that confrontation actually can erupt in surprising places because the definition of vital interests is not straightforward. Witness the willingness of both superpowers to go to the brink over Soviet nuclear weapons in Cuba, which, as we show in Chapter 8, had little military significance for either side. Vital interests are subjectively defined by the people who run governments. In the case of Cuba, as in many others, the true vital interest judged worthy of defense at any cost was not something concrete, such as territory or natural resources, but the perception of resolve —the state's reputation for standing firm when challenged. Many accept the U.S. government's long-standing contention that any capitulation on a matter that has been declared important compromises all future commitments and invites further challenges to the government's power. On such a theory, anything can be instantaneously transformed into a "vital interest" by nothing more than a statement that it is one.

Blechman and Hart conclude from the 1973 Middle Eastern nuclear crisis that

within fairly permissive boundaries, the effectiveness of nuclear threats may not be influenced by the aggregate strategic balance. The threat is not so much to go deliberately to nuclear war as it is to participate and persevere in an escalatory process, even though it might result in nuclear war. Accordingly, the credibility of the threat would not, from a first approximation, be influenced by calculations of just how badly off each side would be if the escalation ran its course, presuming, of course, that both sides had maintained substantial forces. Its credibility would depend on the ability of the nation making the threat to demonstrate convincingly that it perceived such vital interests at stake that it was even prepared to fight a nuclear war, if that became necessary.

The numbers and performance characteristics of the weapons are again irrelevant. Indeed, beyond promoting the doomsday effect, particular

weapons are at most only symbolic contributors to images of resolve. An Aspen Institute report observes:

Deployments of U.S. nuclear weapons in and around Europe—designed for tactical and intermediate-range missions—serve a preeminently political purpose: to symbolize the American commitment to the defense of Europe, and to demonstrate in a concrete way that any large-scale Soviet attack would risk escalation. In large measure, then, their contribution to security stems simply from their presence. Their ostensible military role—to bolster NATO's conventional insufficiencies—remains open to debate, since Soviet nuclear capabilities in the region ensure that any use of nuclear weapons by the West could also risk the destruction of Europe.

Indeed, former assistant secretary of state Richard Burt, responding to a possible delay in Pershing II deployment because of technical problems, reportedly said, "We don't care if the goddamn things work or not…. What we care about is getting them in."^[25] We find it remarkable that both supporters and opponents continued to debate the range, speed, and accuracy of this missile after Burt's statement.

No matter what interests are at stake, morally responsible leaders could never credibly threaten the existence of life on a planetary scale. At the other extreme, viciousness or insanity undoubtedly promote nuclear credibility. Schelling speaks of "the rationality of irrationality," the enhancement of credibility that comes from instability and unpredictability. Nixon spoke more bluntly of the "madman theory."^[26]

The United States, the Soviet Union, and other nuclear powers have long been governed by men and women prepared to endanger the planet periodically for the goals of their states. Consequently, to use Schelling's phrase, a "competition in risk taking" has become the way to resolve the most extreme conflicts between them. Each side takes actions that could, through a chain of ensuing events, lead to the destruction of everything. As in the teenage game of chicken—where two cars speed toward each other until one driver "chickens out" and turns away—states ratchet up the chances of disaster until one saves itself, and the planet, by giving in. The outcome is determined not by the meaningless details of the arsenals that would bring about the holocaust but by the leaders' willingness to run the ultimate risk.

The Paradox of Deterrence

What if a development in the arms race someday does what now seems impossible: truly narrows the gulf that separates conventional

from nuclear war? The credibility of nuclear threats would increase; hence potential combatants would be more cautious about starting a conventional crisis or war. But the firebreak would narrow; hence if war broke out anyway, the use of nuclear weapons would be more tempting. The net effect on the risk of nuclear war is unclear.

For example, assume (falsely, in our view) that if deployed in Europe the neutron bomb would significantly lower the nuclear threshold, increasing the credibility of NATO's threat to use nuclear weapons by decreasing the firebreak between conventional and nuclear war. If the Soviet Union contemplated an invasion of Western Europe, it would presumably be deterred by the increased credibility of NATO's nuclear guarantee, as theorists from the nuclear right and mainstream often argue. But should the Soviets decide to invade anyway, the existence of the neutron bomb would accomplish what much of the nuclear left and the arms control community say it would, namely, increase the likelihood that NATO would actually use nuclear weapons. After all, the credibility of NATO's nuclear guarantee only increases because NATO appears more likely to use the weapons. Thus even if the neutron bomb did significantly narrow the firebreak and boost NATO's nuclear credibility, its effect on the risk of nuclear war would be unclear.^[27]

The same is true for almost all other new weapons that might be invented and almost all nuclear arms control treaties that might be negotiated. Since most weapons and treaties cannot greatly alter credibility and the firebreak and in any case produce contradictory effects on the risk of nuclear war, they are of little significance.

What about technical changes in the doomsday machine, which can significantly affect credibility and the firebreak? If we removed tactical nuclear weapons from Europe, Korea, Navy ships, and so on, would we, by widening the firebreak, reduce the credibility of deterrence and thereby invite conventional war? In situations where there is any real chance of war—that is, where the superpowers believe vital interests are at stake—the answer is no. Here there is no great credibility problem to begin with. Both sides know that superpower combat would pose a terrifying risk of nuclear escalation regardless of the doomsday threat, since either side might deliberately use nuclear weapons rather than concede defeat. In Europe, for example, doomsday weapons are superfluous to deterrence.

Where the superpowers do not claim vital interests, the credibility of threats to deliberately use nuclear weapons is of course inherently low. Here doomsday weapons—weapons that might be launched in combat

without an executive order—probably breed caution on both sides. But war is quite unlikely anyway in such situations, and in any case the creation of a doomsday machine is indefensible given the unthinkable consequences should something go wrong.

Unlike other nuclear arms control measures, then, those designed to reduce the "massive multiplication of a nuclear hair trigger" are of great importance. But even these measures could be only partially effective because any conventional conflict involving the superpowers poses an inescapable risk of sparking an unauthorized or unplanned holocaust, as we discuss in Part II. Preventing conventional conflicts must be the first priority of any serious effort to avoid nuclear war. But removing the nuclear weapons most likely to be used first when crises occur would be a hopeful sign of nuclear sanity.

Chapter Four
What about Misperceptions?

I don't know any American officer, or any Soviet officer, who really believes either superpower can achieve a true first-strike capability, that one side could ever so disarm the other as to leave it without the ability to retaliate…. [Both] strongly agree that neither side can win a nuclear war in any meaningful sense.
—General David Jones, former chairman of the Joint Chiefs of Staff

Are Leaders Misinformed about Nuclear War?

One of the most basic objections to our argument is that the people in charge of nuclear weapons might not understand it. Weapons might matter, if only because people think they do and act accordingly. The realities we have described may be unimportant if irrational perceptions actually govern the decisions of real national leaders.

The nuclear establishment takes this possibility seriously. As psychologist Steven Kull shows, "perception theory" has dominated establishment thinking for years.^[1] The idea that one side or the other could achieve objective, militarily useful advantages in the nuclear balance has been widely, though not totally, discredited. Yet influential theorists argue that because the Soviets and other key audiences may misperceive the meaning of new weapons, we must build them to appear strong even if they do not really make us any stronger or protect against any real threat. One version of this argument is that the Soviets mistakenly believe that nuclear war can be fought and won and that they will soon be confident of the capability to do so unless the United States takes strong actions to disabuse them of this notion. Thus if MX or Star Wars or Trident II makes the Soviets reconsider their reckless position, it must be built whether or not it has real military value. Even if the Soviets might not actually start a nuclear war, the argument goes, their confidence

about winning one, should it occur, could strengthen their hand in political conflicts with the West.

As Warner Schilling explains, "Throughout the 1970's, the United States has been concerned with the appearance of the strategic balance, as well as with its reality, and has been intent on maintaining forces that are not only equal to those of the Soviet Union but are perceived as equal." Schilling believes that "this concern with perceptions is the result of the fact that the pace of the Soviet build-up has enabled the Soviet Union, over time, to surpass the United States in such measures as the total number of missiles, the total number of delivery vehicles, and the total amount of megatonnage," which must then be judged against U.S. advantages in the total number of warheads. As we have seen, such measures have little meaning. In most cases American "disadvantages" resulted from unilateral U.S. decisions to limit force levels to avoid gross redundancy—for example, to let the total megatonnage of the U.S. nuclear stockpile decline from a historical high of around 8,000 equivalent megatons in the early 1960s to about half that today, and to let the total number of U.S. warheads shrink from a high of 32,000 in 1967 to around 26,000 in 1983. Yet "Soviet numerical advantages have led many Americans to fear that the Soviet Union might be tempted to use the threat of nuclear war to intimidate or blackmail the United States, its allies, or other states." Worse, "allies or other states, believing that the Soviet Union has the superior nuclear force, might be led to yield or accommodate to Soviet interests, and the Soviets could gain the political results of nuclear superiority without even having to threaten, much less fight, a nuclear war."^[2] A recent version of this fear concerned the Soviet deployment of more SS-20s aimed at Europe—a militarily meaningless addition to the vast Soviet nuclear threat, but one widely predicted to intimidate U.S. allies in the absence of a compensating U.S. deployment of Euromissiles.

In our view, the left exhibits its own version of perception theory in arguing that even if the Pentagon's new weapons do not actually give the United States a first-strike option, the generals, the national security adviser, and the president might not recognize this reality. With MX, Trident II, and the Stealth bomber in hand, they might foolishly press the button under the delusion that victory is finally possible. Howard Moreland, for example, acknowledges that "to launch a first strike would be risking suicide." But he argues that since "the only discernible military purpose of the nuclear arms race in the 1980's is to acquire a first-strike

capability or to prevent the enemy from acquiring one," leaders apparently do not understand the risk.^[3]

U.S. weapons programs, then, are taken as a signal of a U.S. perception that a successful first strike may be achievable. The problem therefore lies not in MX and Trident II, which do little to change meaningful U.S. nuclear capabilities, but in the leadership's misperception of them. The Soviets may also be misinformed. If they come to believe that the U.S. arsenal poses a real threat of disarming them—or that American leaders believe it does—then in the heat of a crisis they might preempt, desperately and senselessly using their weapons rather than risk losing them.

As we have seen, others fear that nuclear innovations at the low end of the destructive spectrum, such as the neutron bomb, could promote the dangerous misperception that nuclear war can be fought and contained like conventional war. Michael Klare, for example, worries that if, irrationally, the distinction between conventional and nuclear weapons faded in leaders' minds, "the inhibition against nuclear escalation would decrease, and the risk of global annihilation would skyrocket."^[4]

Nuclear Schizophrenia

There is little solid information about how leaders in various countries perceive the importance of new weapons systems—a scandalous omission since purported misperceptions have become a central part of the official case for more U.S. nuclear weapons. As Schilling writes, there is

very little knowledge about how perceptions of the strategic balance are actually influenced (if at all) by such household words as throw-weight, one-megaton equivalents, or prompt hard-target kill capability. The Soviets keep their perceptions to themselves…. As for allies and other states, there is no evidence that the United States government has engaged in any systematic research as to how relevant foreign elites reach their judgments about the state of the strategic balance or even what those judgments are.^[5]

But there has been sufficient real-world experience with nuclear weapons in international affairs to suggest the probable role of perceptions. True, leaders on both sides urgently decry the dangers of their opponent's nuclear weapons and constantly seek new systems of their own. In the peacetime budgetary and domestic political process, weapons

certainly matter to them. But leaders do not necessarily act in the international arena on the basis of these exaggerated assessments of the weapons' import.

Historical experience suggests that military and political leaders on both sides have a "schizophrenic" view of nuclear weapons. For in real foreign policy decisions and in the handling of real crises, their behavior does not appear to be affected by which weapons each side has built. They may still be willing to run the risk of nuclear war—as we document in Part II—but when they do so the details of the nuclear balance do not influence them and certainly do not delude them about the risk they are running.

At the time of the Cuban missile crisis, for example, the United States had a far larger and more advanced nuclear arsenal than the Soviets and had a highly evolved counterforce strategy for launching a disarming attack against Soviet nuclear facilities. If differences in hardware affect leaders' perceptions, then that would have been the time. Many assume that a widespread perception of American nuclear superiority was in fact decisive—motivating Chairman Khrushchev to put nuclear missiles in Cuba; leading President Kennedy to demand the missiles' removal to avoid an unfavorable change in the nuclear balance of power; and allowing the United States to prevail in the crisis. But as we will see in Chapter 8, Khrushchev and Kennedy considered the missiles important not for military reasons but mainly for symbolic ones. Kennedy certainly knew full well that the missiles caused no significant change in the nuclear balance of power.

Six of the president's senior advisers at the time—Dean Rusk, Robert McNamara, George Ball, Roswell Gilpatric, Theodore Sorensen, and McGeorge Bundy—wrote in a Time essay in 1982, "American nuclear superiority was not in our view a critical factor, for the fundamental and controlling reason that nuclear war, already in 1962, would have been an unexampled catastrophe for both sides." Existential deterrence was in place despite the great asymmetries in nuclear forces. American leaders, at least, perceived clearly that "the balance of terror so eloquently described by Winston Churchill seven years earlier was in full operation," and they had little doubt that the Soviets knew it, too. The former advisers reported: "No one of us ever reviewed the nuclear balance for comfort in those hard weeks. The Cuban missile crisis illustrates not the significance but the insignificance of nuclear superiority in the face of survivable thermonuclear retaliatory forces." Other factors determined the real balance of power in the conflict, especially the

"clearly available and applicable superiority [of the United States] in conventional weapons within the area of the crisis."^[6]

In a comprehensive study of newly released documentation of the missile crisis, historian Marc Trachtenberg provides strong additional evidence that American decision makers did not even consider the nuclear balance in their extensive deliberations:

There is no evidence that President Kennedy and his advisers counted missiles, bombers, and warheads, and decided on that basis to take a tough line…. One of the most striking things about the October 16 transcript [of key off-the-record White House meetings between the president and his senior advisers] is that no one even touched on the issue of what exactly would happen if the crisis escalated to the level of general war…. One does come away from the transcript with the sense that even rough calculations of this sort were not terribly important.^[7]

Despite the Kennedy administration's urgent peacetime concern about deficiencies in the strategic balance and its fancy new theories of nuclear strategy, in the actual crisis none of them mattered. Trachtenberg reports that "no one discussed what American counterforce capabilities were—that is, how well the United States might be able to 'limit damage' in the event of an all-out war. It was as though all the key concepts associated with the administration's formal nuclear strategy, as set out for example just a few months earlier in McNamara's famous Ann Arbor speech—in fact, the whole idea of controlled and discriminate general war—in the final analysis counted for very little." The president himself said in one of the October 16 meetings: "What difference does it make? They've got enough to blow us up now anyway." At the point of command decision, he needed no lessons in existential deterrence.^[8]

The same schizophrenia between peacetime rhetoric and crisis actions apparently operated during the next great nuclear confrontation, the Middle Eastern superpower crisis of October 1973, which we discuss in Chapter 6. The United States undertook the most urgent military alert, both conventional and nuclear, since the Cuban crisis and obliquely threatened direct American military action if the Soviets intervened in the Middle East war. Military confrontation between the superpowers loomed. Nixon later wrote that "we neared the brink of nuclear war."^[9]

Like Kennedy before him, Nixon was deeply concerned about the nuclear weapons balance and sought new American weapons to improve it. Yet during an actual crisis neither he, Kissinger, nor other key leaders appear to have paid any attention to the specific nuclear weapons on

either side that so obsessed them before and afterward. Nowhere in Nixon's or Kissinger's memoirs can one find any reference to considerations of the nuclear balance in making these fateful decisions. Trachtenberg's conclusions about the Cuban crisis seem equally relevant here. One gets the impression that the humiliating events of Watergate had more to do with the decision to demonstrate "our ability to act" and our "resolve" than any count-up of nuclear warheads.^[10]

Other accounts confirm this picture of the crisis. Barry M. Blechman and Douglas M. Hart, for example, characterize the nuclear alert as an effort to communicate to the Soviets that "if you persist in your current activity, if you actually go ahead and land forces in Egypt, you will initiate an interactive process between our armed forces whose end results are not clear, but which could be devastating…. The United States is prepared to continue escalating the confrontation up to and including a central nuclear exchange between us, even though we understand that the consequences of such an interaction potentially are 'incalculable.'" The results "could" be devastating, and the consequences are "incalculable"—an assessment consistent with an understanding of existential deterrence. As these authors confirm, Nixon did not misperceive what nuclear war would mean or what new weapons could do for the American position in such a war. Blechman and Hart emphasize that "the Soviet Union clearly did not back down because the United States had an edge in strategic weaponry and could 'win' a nuclear exchange…. Neither side possessed the capability for a disarming first strike, and each would have expected to suffer devastating retaliation if it launched nuclear war." Richard K. Betts agrees that the nuclear balance was unimportant: "Kissinger's public remarks at the time of the crisis made not the remotest suggestion of U.S. nuclear advantage and referred only to the awesome danger of mutual annihilation." If American leaders "sensed some significance in a marginal U.S. nuclear advantage, it could not have seemed more than a remote and trivial one."^[11]

Other crises and foreign policy initiatives have followed the same pattern. As Bundy points out, contrary to the right's fears, "there has been no Soviet action anywhere that can be plausibly attributed to the so-called window of vulnerability." In our interviews with Pentagon and National Security Council officials, we often heard that the Soviets "might well" derive political advantages from that window, opened by their prodigious new nuclear systems. But when challenged for evidence, not one could connect Afghanistan, Central America, the Middle East, or the conflicts of any other region with the nuclear tally sheet. As

Raymond Garthoff sensibly observes: "An American warning to the Soviet Union not to intervene in Hungary in 1956 would not have been heeded, despite the clear American strategic superiority. The Soviet stake in Hungary was great if not vital; the American was not." And the Soviets enjoy tremendous conventional military superiority in their border areas; hence any threat of provoking a major crisis would lack credibility. "For the same reason, not because of changes in the strategic balance, the Soviet leaders ignored repeated warnings from the United States not to intervene in Afghanistan in 1979, and not to build a base in Vietnam in the same year."^[12]

Similarly, American militarism and foreign policy have not borne a clear relation to weapons balances or arms negotiations. No one has plausibly demonstrated that American power concretely benefited anywhere in the world from an American strategic edge since at least 1950. As Henry Kissinger asked rhetorically in his memoirs: "What in the name of God is strategic superiority? What is the significance of it, politically, militarily, operationally, at these levels of numbers? What do you do with it?" The Center for Defense Information notes that in the late 1960s, "when the U.S. still had so-called 'strategic superiority' over the Soviets, it was unable to translate that into political or military clout in the Vietnam War." More recently, as Bundy writes,

It is not self-evident, to put it very gently, that the "victory" of December 1983, and the safe arrival in Europe of the first ground-launched cruise missiles and Pershing II's, has left the [NATO] Alliance stronger and more self confident than it would have been if it had been decided in 1977 and thereafter that there was nothing in any new Soviet deployment of any sort that required a change in the decision of the 1960's that the right place for American mid-range nuclear weapons supporting NATO was in submarines at sea.^[13]

The nuclear weapons balance does not seem to have affected the outcome of superpower confrontations or the division of real power in the world. Bundy believes that even during the four years of U.S. nuclear monopoly, 1945–1949, "aside from the debatable European case, there is very little evidence that American atomic supremacy was helpful in American diplomacy…. To whatever degree atomic diplomacy may have tempted this or that American leader at this or that moment in those years, it did not work."^[14] World opinion alone proved an enormous barrier not only to actual nuclear use but also, as Truman and Eisenhower discovered, to believable nuclear threats.

American leaders understood as soon as Soviet nuclear weapons appeared

in 1949 that the U.S. lead in the arms race might be of little value when weighed against the destructive potential of even a few Soviet atomic warheads. Betts writes: "There was never a time when leaders were confident that the United States could wage nuclear war successfully … restricting damage of the West to 'acceptable' levels." From the start, American leaders worried about both conventional and nuclear Soviet retaliation against Europe, and "U.S. leaders also had no confidence, even in the early period, that they could prevent significant Soviet nuclear retaliation against the American homeland." Even in 1949 the U.S. Joint Chiefs of Staff estimated that Soviet bombers could "reach every important industrial, urban and governmental control center in the United States on a one-way mission basis." Army Intelligence told the National Security Council that just eighteen Soviet nuclear weapons could "wipe out one-third of U.S. steel and iron production, cripple governmental operations in Washington, and hamper and delay mobilization and retaliatory efforts." Secret testimony in 1951 congressional hearings asserted that already the Soviets could obliterate American cities.^[15]

At the beginning of the Eisenhower administration, in 1953, Paul Nitze told Secretary of State Acheson of a National Security Council study revealing that "the net capability of the Soviet Union to injure the United States must already be measured in terms of many millions of casualties"—in fact, 22 million casualties for a postulated Soviet attack with 150-kiloton weapons. In 1954 Eisenhower said, "Atomic war will destroy civilization." An elaborate 1955 nuclear war simulation suggested that fifty-three American cities could be bombed, causing 8.5 million immediate fatalities, leaving an equal number injured, and depriving 25 million of food and shelter. Reportedly, "the president's one comment was: staggering." Eisenhower's diary for January 23, 1956, records his impressions of an Air Force briefing on the consequences of a hypothetical Soviet surprise attack: "The United States experienced practically total economic collapse…. A new government had to be improvised by the states. Casualties were enormous … something on the order of 65 % of the population would require some kind of medical care, and in most instances, no opportunity whatsoever to get it." Even when the scenario gave the United States a month of warning and had the Soviets concentrate on American air bases rather than cities, the president wrote, "there was no significant difference in the losses we would take." Eisenhower was so disturbed by such findings that he asked the National Security Council to study how much destruction the United States could "absorb and still survive."^[16]

By 1957 Eisenhower was talking about 25 million American dead and 60 million injured. The Joint Chiefs of Staff suggested that estimate might be optimistic: considering the effects of radioactive fallout, between 46 million and 117 million Americans could die. In 1960, when Eisenhower was told of possible plans for sharp increases in strategic nuclear weapons, he reportedly responded sarcastically: "Why don't we go completely crazy and plan a force of ten thousand?" (Ironically, that is about the number of U.S. strategic weapons today.) How many times, the president asked, "could [you] kill the same man?" The Air Force estimated in 1960 that a Soviet first strike could kill 150 million people, three-quarters of the American population; even Soviet retaliation after an American first strike could kill 110 million. John F. Kennedy reportedly absorbed the figure of 150 million dead from his first briefing by the Joint Chiefs of Staff. By that time the Soviet Union had so much nuclear megatonnage that Khrushchev announced cutbacks in production of some missiles because rockets "are not cucumbers, you know—you don't eat them—and more than a certain number are not required to repel aggression." Nuclear war, he said, "is stupid, stupid, stupid! If you reach for the button, you reach for suicide."^[17] Existential deterrence was securely in place and both sides knew it.

Garthoff, who has painstakingly studied the history of U.S.-Soviet relations, writes, "The global strategic balance is much less important in deterring or resolving crises than many have assumed, because the prospect of nuclear war deters even leaders who command an overwhelming superiority, as shown in 1962, and all the more so leaders on both sides with larger but more equal forces, as demonstrated ever since." If the Cuban missile crisis recurred, he believes that "the outcome today under strategic nuclear parity could be the same…. The mere possibility of nuclear war was the deterrent."^[18]

Referring to the great Suez, Berlin, and Cuban crises, Bundy similarly asserts: "In none of the three cases, I feel confident, would the final result have been different if the relative strategic positions of the Soviet Union and the United States had been reversed." In all three, the United States enjoyed an immense superiority in nuclear hardware. But it counted for little, since the Soviets had enough to pose an unacceptable threat to New York, Washington, Chicago, and other U.S. cities. As strange as it may seem, if the United States had been the laggard in the nuclear arms race its power would not have been less, because "a stalemate is a stalemate either way around."^[19]

In a major study of American military operations, Force Without War, Barry Blechman and Stephen S. Kaplan also conclude: "Our data

would not support a hypothesis that the strategic weapons balance influences the outcome of incidents in which both the United States and the U.S.S.R. are involved." To the contrary, in surveying dozens of incidents from 1946 to 1975, they found that, from the point of view of American objectives, "short-term outcomes were positive 43 percent of the time when the U.S. strategic advantage was 100 to 1 or greater, 82 percent when the U.S. strategic advantage ranged between 10 and 99 to one, and 92 percent when the U.S. advantage was less than 10 to 1…. If we look at [only) those incidents in which the Soviet Union used or threatened to use force, these figures were 11, 50, and 90 percent, respectively." These conclusions cover not only the period of existential deterrence but also the earlier period of enormous American nuclear superiority. The numerous detailed case studies in Force Without War likewise "provide little support for the notion that decisions during crises are strongly influenced by aggregate strategic capabilities."^[20]

We likewise found no evidence of even the most casual discussion of relative nuclear strength or the characteristics of nuclear weapons systems during the many superpower crises and confrontations discussed in Part II. This neglect stands in surreal contrast to the near obsession with the numbers and performance characteristics of nuclear arms in noncrisis times documented throughout Part I. It is as if everyone takes a powerful weaponitis pill on the days when the world is relatively safe and an equally potent reality pill when it is not. Thank goodness it is not the other way around.

Some conservative analysts, such as General Daniel Graham and Richard Pipes, continue to insist that the Soviets reject existential deterrence. According to them, the Soviets believe that in the nuclear age, as in previous ages, war is just a "continuation of politics," that it can be won, and that "socialism" will be victorious. In this view, Soviet restraint so far reflects only their analysis that the "correlation of forces" has not yet moved in their favor. This restraint could change after the Soviets complete their nuclear rearmament unless the United States undertakes several major new offensive and defensive nuclear programs. Other students of Soviet military thought, such as George Kennan, dismiss as farfetched the idea that the Soviets believe they could fight a nuclear war without risking unacceptable damage and casualties.^[21]

Apparently, Soviet leaders are as schizophrenic on nuclear issues as their American counterparts. In public pronouncements, Soviet leaders often stress the importance of preparing for nuclear war and building the weapons that would permit them to fight it. But they do not necessarily

act in international affairs on this basis or even privately harbor delusions about what nuclear war would mean. Robert L. Arnett concludes his careful study of Soviet attitudes as follows: "What Soviet spokesmen have been saying about nuclear war does not support the claims of various Western analysts who argue that the Soviets believe they can win and survive a nuclear war." Rather, as in the United States, public pronouncements reflect only half of a split view, one shaped largely by domestic ideological needs. "Publications written for internal consumption," however, "contend that nuclear war cannot serve as a practical instrument of policy, and they continually talk about the dire consequences of such a war." Similarly, David Holloway's authoritative study argues that underneath their rhetoric the Soviets understand that their relationship with the United States "is in reality one of mutual vulnerability to devastating nuclear strikes…. There is little evidence to suggest that [the Soviets] think victory in a global nuclear war would be anything other than catastrophic." Bundy agrees that despite the continuing search for "marginal benefits," both sides ultimately recognize that "effective superiority is unobtainable."^[22]

The most objective analysts observe that Soviet strategic policy advocates preemption, not premeditated first strike, and does not reflect a victory-is-possible mentality. As Leon Wieseltier (no admirer of the Kremlin) emphasizes, this policy "means only that the Soviet Union would fire its missiles first if a crisis with the United States reached the point at which a nuclear exchange seemed inevitable"—a position identical to that advanced by American officials and reflected in U.S. nuclear policy. Soviet policy is indeed "a provision for the brink of war, not the breaking of the peace. There is no evidence at all that the Soviets believe in 'first strike,' or that they have plans to start a nuclear war, or that their nuclear strategy is, as Pipes put it, 'not retaliation but offensive action.'"^[23] And as we saw in Chapter 2, preemption, whatever its merits, has little to do with weapons balances and everything to do with political conditions and the perceived risk of war.

In honest moments outside the budgetary process, senior professional soldiers on both sides sometimes acknowledge that nuclear weapons hardware does not really affect basic military realities. Marshal Ogarkov, commander in chief of the Warsaw Pact, said in a May 1984 interview in the Soviet press that "the deployment of U.S. intermediate-range missiles in Western Europe did not increase the possibility of a 'first strike' against the Soviet Union. Both sides fully recognize the inevitability of a retaliatory strike." General David Jones, former chairman

of the Joint Chiefs of Staff, agreed: "I don't know any American officer, or any Soviet officer, who really believes either superpower can achieve a true first-strike capability, that one side could ever so disarm the other as to leave it without the ability to retaliate … [both] strongly agree that neither side can win a nuclear war in any meaningful sense."^[24] In other contexts each man has undoubtedly denounced the first-strike weaponry of his enemy and urged new weapons systems for his own side. Yet clearly on another level they both understand the meaninglessness of this constant arming and counterarming.

Steven Kull provides still more evidence of nuclear schizophrenia in a series of interviews with high-level military and civilian nuclear planners and strategists. He examines the extent to which planners "conventionalize" nuclear weapons by viewing a nuclear war in the same light as a conventional war that can be limited in scope, meaningfully won, and deeply influenced by the relative armaments of the contending powers. Although the academic literature often suggests that the military wildly conventionalizes scenarios of nuclear war, "even among those who overtly conventionalized, the majority of people [Kull] interviewed, when questioned directly, did recognize the key elements of the nuclear revolution." Most "shifted between a mindset that conventionalized and another that recognized key features of the nuclear revolution," suggesting "a fragmentation or lack of integration."^[25] Respondents often began by railing against the dangers of Soviet nuclear weapons systems and the need for American systems to counter them. But under Kull's questioning, they would quickly acknowledge, as any informed person must, that both the United States and the Soviet Union are completely vulnerable to nuclear destruction, that none of the new weapons on either side has changed this vulnerability, and that it will remain in place for as long as anyone can foresee.

Even in peacetime national leaders and others involved with the nuclear issue are schizophrenic, sometimes recognizing the fallacy of weaponitis. "Virtually all respondents," Kull reports, "would at some points recognize that, given the military realities of the nuclear era, the policies they were proposing were questionable or even invalid. To elicit this second perspective, I simply had to direct their attention to key features of nuclear reality that they were ignoring or suppressing. … not a single respondent consistently presented a conventional mindset perspective; all would at some point recognize the key features of nuclear reality." Kull found, for example, that among the eighty-four former and present U.S. military officers, civilian security officials, congresspeople,

and nuclear strategists he interviewed, "the vast majority … readily recognized that … both sides have the capability, even after absorbing an all-out first strike, to retaliate in such a devastating fashion that neither side could meaningfully benefit from such a first strike." Moreover, "it was frequently stated that it would be impossible to keep a nuclear war limited and, sometimes, that anybody who thought so was crazy." Kull is "convinced that, perhaps with the exception of a few … individuals the defense establishment as a whole does fully perceive the American population as fundamentally vulnerable." "It was particularly striking," he observes, "how easily many individuals, even members of the current administration, would blithely dismiss the military value of weapons proposed or currently being deployed."^[26] Those weapons, they realize, cannot really alter the superpower nuclear stalemate for better or for worse.

Kull found a similar schizophrenia among Soviets: "In every case the clear implicit message was that they were fully aware that the military rationale for maintaining the balance [of nuclear weapons between the superpowers] did not really make logical sense. They also seemed to feel that I should know that they understood this, and they thought me a bit thick-headed for being so impolite as to force the issue." "The dominant theme of the answers," Kull writes, "was that any superpower war would almost certainly be all out and both sides would be effectively annihilated. Soviet civil defense efforts were dismissed as simply the result of 'bureaucratic inertia' or efforts to 'calm down the population.'"^[27]

George Rathjens and Laura Reed ask, "Can one really believe that an American president (or the Soviet leadership, or that of a third country) would behave very differently in a crisis if the United States had no MX missiles, or a thousand instead of the number now envisaged; or if the Soviet Union had never developed the SS-20, or alternately had thousands of them?" They are on firm ground when they answer, "It is unlikely."^[28]

A Drawing-Room Comedy

Bernard Brodie would be disappointed to see how little progress has been made since he noted a quarter century ago that "eager acceptance of the new is coupled, not only within the same organizations but often within the same persons, with stubborn insistence upon retaining also much of the old." The nuclear schizophrenia of leadership partly reflects this inner split that Brodie noticed. In the calm peacetime budgetary

process, many officials get caught up in the old thinking: the weapons paradigm. But in a crisis involving much higher stakes, when abstract analysis must give way to practical command decisions, leaders are much more influenced by common sense. They recognize that it does not much matter who has what nuclear weapons at this point in the arms race since everything will probably be destroyed anyway in the event of a nuclear war. "In even contemplating a first strike," as John Steinbruner writes, leaders "would not be very sensitive to the fine details of the technical force balance. The destructiveness of nuclear weapons delivered in retaliation, even in modest numbers, is compelling enough to dominate [their] attention and behavior. … even clearly demonstrable technical advantages would have very little effect on this basic result."^[29]

As Brodie wrote in 1978, "The defense community of the United States is inhabited by peoples of a wide range of skills and sometimes of considerable imagination. All sorts of notions and propositions are churned out, and often presented for consideration with the prefatory words: 'It is conceivable that …' Such words establish their own truth, for the fact that someone has conceived of whatever proposition follows is enough to establish that it is conceivable."^[30] When asked why he spent so much effort and money on military capabilities he knew were marginal, if not utterly unrealistic, one high-level nuclear war planner told us that it was his job to take no chances when it came to deterrence. He asked, "Do you want me to play 'you bet your country'?"

Such "marginal thinking" resembles what defense critics call "worst-case thinking." Near the end of his life, Brodie developed a profound distrust for it: "The thinking up of ingenious new possibilities is deceptively cheap and easy, and the burden of proof must be on those who urge the payment of huge additional premiums for putting their particular notions into practice."^[31] But regardless of its merits, it can logically coexist with an understanding that in the big picture the weapons balance is almost irrelevant. Marginal thinkers can work hard to get the weapons they want to achieve marginal benefits. They can then, without inconsistency, ignore those weapons when they make major foreign policy or military decisions that are logically based not on the margin but on the core of the military and political balance.

Moreover, many officials and strategists hope that over many years numerous improvements in weapons, each marginal by itself, will add up to something significant. Although they consciously exaggerate the significance of individual weapons systems, such as the MX, to get public

and congressional approval for them, the real agenda is a much longer term shift in strategic posture. But again, when they must make major foreign policy decisions, they recognize that the marginal improvements do not in fact amount to much and therefore that the real strategic balance is the same as always.^[32]

At one level, then, nuclear schizophrenia reflects flawed styles of thinking—internal confusion and obsession with marginal military capabilities—that lead officials to seek new weapons in the vain hope of increasing real military capabilities. But nuclear schizophrenia is also a deliberate strategy designed in part, as Warner Schilling explained earlier, to deal with others' supposed weaponitis by feigning one's own. Kull found that though most American officials know that new weapons cannot change the nature of nuclear war, they suspect that Soviet and other foreign leaders might not understand that. Hence they act as if the nuclear balance mattered, buying weapons to impress those leaders—in part to discourage any lingering misperception that a successful first strike against the United States is possible and even perhaps to encourage the misperception that a successful U.S. first strike against the Soviet Union is possible. If foreign perceptions of American military power (however unfounded) grow, then—as American officials have publicly argued—so will America's ability to intervene abroad without interference. The Soviets, as Kull found, follow a similar strategy—building weapons to counter supposed American misperceptions.

"The situation resembles nothing so much as a drawing-room comedy," Kull observes. "All of the key characters know a certain secret—that strategic asymmetries are militarily irrelevant in an age of overkill—but because they think that others do not know the secret they act as if they do not know the secret either. A farcical quality emerges as all the characters, more or less unconsciously, collude to establish a norm of behavior based on a failure to recognize the secret."^[33] But in crises leaders—all of whom do know the secret—do not consult the nuclear balance or indulge in the fantasy that either side can actually use its new nuclear systems without inviting worldwide destruction.

As Robert Jervis observes, "There is little evidence that European or Third World leaders pay much attention to the details of the strategic balance." In 1977 the Pentagon sponsored a rare effort to gather such evidence through a conference on "International Perceptions of the Superpower Military Balance." It provided scant support for the actors in this "drawing-room comedy": "In the overall strategic-nuclear area … the Soviets, the [French] Défense Nationale writers, and the Arabs generally

characterized [the superpowers] as equal, with many of the last two groups believing it moot to ask, 'Who is ahead?' in a situation of mutual nuclear overkill. These individuals saw the superpowers as functionally equal regardless of which had the quantitative or qualitative advantage." The conference findings also refer to the "oft-stated French belief … that overall strategic inventory totals have lost their significance due to the 'balance of terror.' It is generally assumed that both sides have more than enough." Just as important, the studies also failed to find evidence that international leaders believe "accommodations" with the Soviets should be sought because of "recent perceived shifts in some balances away from U.S. favor." Contrary to the argument that international elites mistakenly attach importance to the nuclear weapons constructed by the superpowers, at least in some cases they do not even bother to monitor these developments. One study of Arab newspapers concluded that they "do not follow the details of new weapons developments." Concluding that no one in the world seems impressed by all the weapons built for that purpose, Kull comments: "What is particularly striking [in the drawing-room comedy] … is that when the main character—in this case the Defense Department—is informed that, in fact, everybody knows the secret, it stiffens its resolve to maintain the charade."^[34]

The Pentagon maintains the charade perhaps partly to convey the impression that American leaders are dangerously irrational by building first-strike and other weapons that only lunatics would try to use. As Joseph Gerson explains, "If the Soviets believed that our leaders thought they could launch a first-strike blow—Nixon's 'madman' theory—they would use extreme caution in countering U.S. moves." Hence, as many on the right hope and many on the left fear, continual nuclear weapons modernization might be able to "shield" or "cover" U.S. intervention abroad even though everyone knows that the technical arms race is in fact permanently stalemated. Kull found that

such thinking has continued in defense circles…. Some analysts or former officials explicitly described how they would actively work to create these irrational images. One analyst described his goal of making Americans appear "wild and crazy" and imagined—with apparent delight—the effect of articles about winning nuclear war on Soviet analysts…. He also credited some of his well-known hawkish friends with intentionally appearing irrational as part of an unofficial role they had designed for themselves in the service of the United States Government.^[35]

The madman strategy, like the other sources of peacetime weaponitis mentioned earlier, produces apparent schizophrenia. To appear dangerously

irrational, government officials publicly convey the belief that their new weapons actually give the United States usable new nuclear options. But well aware that is not really the case, in crises leaders do not waste time thinking or talking about those apocryphal options.

Self-Inflicted Wounds

What should be done if—despite the evidence cited earlier—serious misperceptions about nuclear weapons either linger from the prenuclear era or emerge from the superpowers' loud claims that their new nuclear weapons actually provide new and usable military power? Potentially catastrophic misperceptions have certainly occurred before. For example, recent interviews with four of the most important U.S. Air Force generals of the early nuclear age elicited some eerie reflections on the Cuban missile crisis. The generals did not—and to this day do not—concur with the realistic assessment of President Kennedy and his top civilian advisers that even in 1962 nuclear war would have been a total catastrophe for both superpowers and much of the rest of the world. General David A. Burchinal, a former Strategic Air Command (SAC) wing commander, chief of staff of the Eighth Air Force, and senior staff of Air Force Headquarters and the Joint Chiefs of Staff, insisted that

the Kennedy administration … both the executive leadership and … McNamara … did not understand what had been created and handed to them, and what it had given them. SAC was about at its peak. We had, not supremacy, but complete nuclear superiority over the Soviets…. Our politicians did not understand what happens when you have such a degree of superiority as we had, or they simply didn't know how to use it. They were busily engaged in saving face for the Soviets and making concessions, giving up IRBMs, the Thors and Jupiters deployed overseas [see Chapter 7]—when all we had to do was write our own ticket.

General Curtis E. LeMay, father of the Strategic Air Command and Air Force chief of staff during the crisis, added: We could have gotten not only the missiles out of Cuba, we could have gotten the Communists out of Cuba at that time." "You bet we could have," agreed General Leon W. Johnson, with the National Security Council at the time. According to LeMay, when a U.S. fighter plane strayed into Soviet airspace near the height of the crisis, potentially provoking an accidental war, Defense Secretary McNamara rushed to "apologize to the Russians" [see Chapter 7], but the Joint Chiefs of Staff said, "No. Tell them, don't touch that thing, or they've had it."

These men believe that at least for that brief time the United States

could really have fought a nuclear war with the Soviet Union without risking devastating destruction. Indeed, LeMay contends that the United States had nothing at all to fear: "During that very critical time, in my mind there wasn't a chance that we would have gone to war with Russia because we had overwhelming strategic capability and the Russians knew it." Burchinal is puzzled that in "publications about the Cuban missile crisis all claim that we were so close to nuclear war; ninety-nine percent of the people who write about it don't understand the truth." He even claims that "we were never further from nuclear war than at the time of Cuba, never further." But even Burchinal recognizes that "as the Russians built up their capacity during the 1960's and into the early 1970's, that situation [of overwhelming superiority] no longer obtained."^[36]

Even today the superpowers' propaganda about their new weapons may foster dangerous misperceptions about nuclear war. The political and military institutions of both sides may, too. The Soviet Union and the United States have long planned, equipped, and trained their forces to fight a nuclear war like a conventional one. Both countries promote elaborate strategies for limiting a nuclear war and trying to win it. For example, a U.S. Army field manual published in 1980 states: "The U.S. Army must be prepared to fight and win when nuclear weapons are used."^[37] Military bureaucracies, then, may be institutionalizing dangerous misperceptions. In crises to come, as in crises past, some military officers and civilian officials may urge the U.S. president or the Soviet general secretary to enact such war plans in the insane belief that they could actually be carried out without risking everything.

Irrational beliefs about nuclear war certainly persist among professional nuclear analysts and could be dangerous if these analysts got the ears of top national leaders. Today most (but not all) reputable analysts recognize that neither side can gain a meaningful advantage by launching a first strike. Most (but not all) also recognize that a nuclear war cannot confidently be kept limited and could well lead to the total destruction of both superpowers and much more. Still, as we have seen, the so-called nuclear experts, exhibiting their own schizophrenia, often speak of "incentives" for getting in the first blow or of the possibility of using "limited nuclear options" to influence the outcome of a crisis. And in some areas the experts flatly express dangerous misperceptions. An important example, which we noted in Chapter 2, is the seventeen highly regarded specialists who write that an American leader (perhaps Bush or Quayle) must make a split-second decision about whether to

launch thousands of warheads in the event of a warning that the Soviets have already pushed the button.

Even if leaders do harbor dangerous misperceptions about nuclear war, however, building or opposing weapons systems is not the answer. Ironically, that can easily reinforce the very misperceptions that one is trying to counter. When the U.S. government argues for new weapons to create an image of strength, for example, it decries glaring weaknesses in American forces to justify these weapons to Congress and the public. This outcry actually fosters an image of weakness, not of strength, as Kull points out. The Pentagon-sponsored review of the evidence on Soviet, French, and Japanese perceptions of the nuclear balance acknowledged that "the tendency of many U.S. spokesmen (particularly government officials at budget time) to emphasize Soviet strengths and U.S. weaknesses often had a negative impact on the perceived U.S. standing." The former secretary of defense James Schlesinger calls these "self-inflicted wounds," by-products of taking false fears seriously and thereby reinforcing them.^[38]

The Soviet government plays the same crazy game. It fears that new U.S. weapons, though militarily irrelevant, will create a perception of American gain. So it denounces the weapons and vows to build its own to counter them. But this response plays into the Pentagon's hands—reinforcing the perception that the United States has shifted the balance of power. The U.S. air-launched cruise missile, for example, was supposed to counter perceptions of American strategic inferiority. "Whether or not such perceptions" of new American strength are warranted, writes Richard K. Betts, "they should be enhanced by the Soviet Union's vigorous complaints about the new threat they feel from the U.S. cruise missile."^[39]

The peace movement likewise opposes so-called destabilizing weapons, in part to prevent the misperception that the United States is acquiring a first-strike knockout capability. But to succeed, the peace movement must decry the dangers of first-strike weapons, thus unwittingly adding legitimacy to the erroneous and dangerous idea that in nuclear war it matters who strikes first. George Rathjens is one of the few critics of American nuclear policy who grasps this problem. He writes that he opposed the MX "only with great diffidence and selectiv[ity]," not because he doubted the valuelessness of the weapon but because in making the standard arguments against it he did not want to "concede implicitly that there was some legitimacy to the 'window of vulnerability' claim and/or that 'first strike' attacks were a very serious problem—when

… there were many problems more demanding of attention." With the same logic Rathjens "chose not to testify for or against" SALT II: "To do either with vigor would have suggested the agreement more significant than I believed it to be."^[40]

The only sensible way to counter dangerous misperceptions about nuclear war is to correct them. Concerned citizens and specialists alike must challenge false statements about nuclear war wherever they appear, whether in newspapers, academic publications, classrooms, government press releases, or the annual report of the U.S. secretary of defense. That is a big job, and a crucial one; as we have seen, most academic and political commentary about nuclear war, befuddled by weaponitis, continues to suggest that some technological deficiency or advancement could give a state a rational reason to fire nuclear weapons first or to fire them in haste. We must also work to unmask and replace the many military officers and civilian leaders who spout nuclear nonsense. During his 1979–1980 campaign for the Republican presidential nomination, for example, George Bush had the following interchange with journalist Robert Scheer:

SCHEER : Don't you reach a point with these strategic weapons where we can wipe each other out so many times and no one wants to use them or be willing to use them, that it really doesn't matter whether you're 10 percent or 2 percent higher or lower?

BUSH : Yes, if you believe there is no such thing as a winner in a nuclear exchange, that argument makes a little sense. I don't believe that.

SCHEER : How do you win in a nuclear exchange?

BUSH : You have a survivability of command and control, survivability of industrial potential, protection of a percentage of your citizens, and you have a capability that inflicts more damage on the opposition than it can inflict on you. That's the way you can have a winner.^[41]

We must also challenge the institutional sources of nuclear misinformation, exposing and rebutting official U.S. plans for nuclear "war fighting," "controlled escalation," "prevailing," or "damage limitation." And we should work to change those absurd plans, not indirectly, by opposing the weapons requested to implement the plans, but directly: by insisting that official nuclear strategy reflect the obvious fact that nuclear war cannot be fought without seriously risking the destruction of everything. Then the government would not be training thousands of soldiers to think about nuclear war and conventional war in the same way.

If foreign leaders should misperceive, say, the importance of Soviet advantages in throw weight or megatonnage, then "the appropriate American response," as Schilling argues, "is not to add to some component of its strategic forces in order to change the direction of a curve on a chart, but to explain to its allies and friends (and to the Soviet Union if need be) that the numerical differences in question would have no significant bearing on the outcome of a nuclear exchange."^[42] The same explanation is in order if the Soviets or anyone else should fear that American innovations or numerical advantages allow the Pentagon to launch a successful first strike or to control nuclear war.

As for the madman strategy, American and Soviet leaders may indeed try to convince each other, and others, that they are crazy enough to launch the missiles if challenged in the Third World. But even if the strategy works, the actual possession of giant new weapons is not the main reason for success. Hence working to stop those weapons or even the arms race generally may do little to stop the diabolical psychological game.

No amount of MX and Trident II missiles is likely to convince the Soviets that American leaders fail to recognize the probable result of launching thousands of such missiles in an attempted first-strike knockout: a global calamity with no winners. The same is true in reverse. The balance of terror is, by now, too obvious for either side to imagine that the other would not realize the catastrophic consequences of large-scale nuclear war.

Of course, the first-strike risk cannot be totally discounted by those who would be the victims. But more plausible by far, and the real heart of the madman strategy, is that leaders would attempt a very limited use of nuclear weapons if seriously challenged in the Third World or elsewhere—not because of a wild misperception that one side could actually win a large-scale nuclear war should it come to that, but simply because of recklessness: a mad willingness to run the risk of escalation and total mutual destruction.

As we will see in Part II, American and to a lesser degree Soviet leaders have repeatedly rolled the nuclear dice in past Third World conflicts to intimidate each other, taking actions they knew would raise the risk of a mutually unsurvivable holocaust. No matter what the state of the essentially irrelevant nuclear arms race, neither superpower is likely to underestimate the possibility that the other would go even further in the future, actually using a small number of nuclear weapons in an extreme crisis.

True, appearing mad enough to use nuclear weapons may be somewhat easier if one is constantly spending billions on building up and improving those weapons. But what really makes an impression on people around the world are the mad actions that have actually propelled the superpowers toward confrontation many times. Even after a START treaty or a nuclear freeze, the superpowers could maintain the madman strategy, since the details of the nuclear armories are not really its basis. To the extent that particular nuclear weapons do reinforce the madman strategy, it is not the so-called first-strike weapons but the small doomsday weapons (discussed in Chapter 3) that would probably be used first on some future superpower battlefield. Those doomsday weapons, as we have argued, are in fact dangerous and should be abolished, unilaterally if necessary.

To denounce as dangerous other nuclear weapons systems, which in reality add little to the nuclear danger, may seem to the Soviets and others as only more evidence that perhaps those weapons really do pose a new threat in the hands of lunatic American leaders. The most realistic answer to the madman strategy is to counter it directly by preventing madmen from becoming leaders and, failing that, by imposing large political costs on leaders who say or do crazy things in the nuclear age.

Pretending that weapons matter is thus in every way more dangerous than the weapons themselves. Such a pretense—whether by the superpower governments or by the peace movements opposing those governments—can only confuse people about the existential risks and horror of a war fought with any nuclear weapons. This pretense also distracts attention from the occasional new weapon that actually adds to the risk of nuclear war. The furor over so-called U.S. first-strike weapons, for example, has almost totally eclipsed discussion of the real hazards of the U.S. sea-launched cruise missile (discussed in Chapter 9).

If the arms race doesn't matter, we should simply say so, point out the genuine exceptions, and confront all who harbor or spread misperceptions or who speak and act like madmen. Otherwise we only add to the confusion, and to the danger.

Chapter Five
What about Star Wars?

Every new weapon will eventually bring some counter defense to it.
—Harry Truman, addressing the U.S. Congress, October 1945

After five centuries of the use of hand arms with fire-propelled missiles, the large number of men killed by comparable arms in the recent war indicates that no adequate answer has yet [been] found for the bullet.
—Bernard Brodie, The Absolute Weapon, 1946

Nuclear Science Fiction

In a fund-raising letter for Americans for Democratic Action, Isaac Asimov writes that the Star Wars plan for defending the United States by shooting down Soviet nuclear missiles is nothing but "Hollywood science fiction." He warns, nevertheless, that Star Wars is "dangerous" and "destabilizing," and even "a threat to world peace" and "to our national security." But how can a fictional weapon endanger the survival of the real world? Similarly, Harrison Brown, then editor in chief of the Bulletin of the Atomic Scientists, reveals that on learning of Star Wars he "laughed" and is "quite certain that [his] laughter blended with that of thousands of other scientists and engineers," because the design of an impenetrable nuclear shield is "virtually impossible." But if, as Brown believes, the Star Wars concept is "reminiscent of the concept of perpetual motion," how can it also carry "unprecedented dangers," and why does he say that "those of us who laughed when the Star Wars concept was first suggested should be crying"?^[1]

This strange contradiction is at the heart of the debate about the Strategic Defense Initiative (SDI), otherwise known as Star Wars. Almost everyone now accepts that SDI cannot defend America from nuclear war. The world's premier scientific journal, Nature, stated flatly that

"the scientific community knows that [Star Wars] will not work." The Pentagon nevertheless insists that SDI is essential for U.S. security. Moscow incorporated opposition to Star Wars into the new Communist party program, thereby making it "one of the most basic precepts of the party." And the American peace community, having worked diligently to prove Star Wars impractical, continues denouncing it as an unprecedented threat to peace and committing many precious resources to stopping it.^[2]

The preoccupation with Star Wars is perhaps the classic case of weaponitis. Star Wars cannot shift the nuclear equation, for better or for worse, any more than a perpetual-motion weapon could if President Reagan had gone on national television to announce that the United States was determined to build it.

Let us examine three of the central claims made about Star Wars: (1) it will protect people from nuclear war, (2) it will make nuclear war either more likely or less likely, and (3) it will radically transform the nuclear arms race.

A Nuclear Umbrella?

President Reagan's original vision, presented in his startling speech to the nation of March 23, 1983, was apparently of an impenetrable shield that could repel any offensive nuclear barrage, protect the U.S. population, and even "give us the means of rendering … nuclear weapons impotent and obsolete." A 1985 fund-raising letter of the Citizens for a High Frontier was in full agreement: "This plan … [will] make all of us safe from nuclear missile attack … [because] it will actually render harmless virtually all nuclear missiles anyone might fire at us."^[3]

No one doubted that it was possible to shoot down ballistic missiles—and had been since the original ground-based antiballistic missile (ABM) weapons of the 1960s. But soon after the first ABMs were deployed, most authorities agreed that actually protecting populations and industry was well beyond their means. Indeed, the ABM treaty of 1972 was essentially a formal recognition by both sides that a practical and economic missile defense of society was unreachable in the 1970s.

The prognosis is unlikely to change in the 1990s or even in the twenty-first century. Former secretary of defense James Schlesinger writes that "there is no realistic hope that we shall ever again be able to protect American cities" because a single hydrogen warhead can destroy one. As McGeorge Bundy, George Kennan, Robert McNamara,

and Gerard Smith point out, "even a 95-percent kill rate would be insufficient to save either society from disintegration in the event of general nuclear war." To work effectively, the shield cannot leak, but as the entire history of military technology shows, "there is no leak-proof defense."^[4]

No complex technology works perfectly, and Star Wars would be by far the most complex ever attempted. But it "must work perfectly the very first time [it is used], since it can never be tested in advance as a full system." A true operational test would require an actual Soviet attack. As prominent software engineers have pointed out, even if the gamut of exotic Star Wars hardware by some miracle performed flawlessly in its first full-scale use, the unprecedentedly large software programs that control it would not, since it is impossible to build large "bugless" programs. David Parnas, a computer scientist on the president's Strategic Defense Initiative advisory panel, resigned with the following explanation: "Because of the extreme demands on the system and our inability to test it, we will never be able to believe, with any confidence, that we have succeeded."^[5]

In fact, a leak-proof shield against current Soviet nuclear forces would be only the beginning, and possibly the easiest part, of strategic defense. For "any prospect of a significantly improved American defense is absolutely certain to stimulate the most energetic Soviet efforts to ensure the continued ability of Soviet warheads to get through." This is the answer to those who compare the prevailing skepticism about Star Wars' feasibility to that about putting a man on the moon in the 1960s: "The effort to get to the moon was not complicated by the presence of an adversary. A platoon of hostile moon-men with axes could have made it a disaster."^[6]

One way to foil Star Wars is simply to attack it. Schlesinger predicts that "an effective opponent will develop defense suppression techniques and will punch a hole through any space-based defense that is deployed." And "no one has been able to offer any hope that it will ever be easier and cheaper to deploy and defend large systems in space than for someone else to destroy them. The balance of technical judgment is that the advantage in any unconstrained contest in space will be with the side that aims to attack the other side's satellites."^[7]

The Soviets could also expand and improve their offensive nuclear strike forces until they could overwhelm, elude, and spoof the defensive screen. A study for the U.S. Office of Technology Assessment (OTA) by Ashton Carter, a Pentagon systems analyst in the first two years of the

Reagan administration, concludes that "for every defense concept proposed or imagined, including all of the so-called 'Star Wars' concepts, a countermeasure has already been identified"; that these "could be implemented with today's technology, whereas the defense could not"; that "in general, the costs of the countermeasures can be estimated and shown to be relatively low, whereas the costs of the defense are unknown but seem likely to be high"; and that "in general, the future technologies presupposed as part of the defense concept would also be potent weapons for attacking the defense." To take just one example, leaks of a classified Defense Intelligence Agency study revealed Pentagon evidence that by 1993 the Soviets could build "fast-burn" rocket boosters able to evade the critical "boost-phase" intercept by U.S. space-based weaponry. That phase is the only chance to attack Soviet missiles before they decompose into numerous warheads and decoys. The new Soviet rockets would "burn out" so fast after leaving the earth's protective atmosphere that space-based weapons would not have enough time (less than one minute) to destroy many of them before the hot exhaust used to track them was gone.^[8]

Even with a near-perfect antimissile defense—a technical absurdity—America would remain absolutely vulnerable to nuclear devastation; ballistic missiles are only one of many ways to deliver nuclear warheads. Others include air-breathing delivery vehicles (manned bombers and cruise missiles) and covert delivery by commandos, infiltrators, and saboteurs. Former secretary of defense Caspar Weinberger suggested that Star Wars would defend against these threats as well, but that hardly seems credible. Schlesinger points out that "the United States Air Force has long argued that air defense systems [against manned bombers] are penetrable and will always be penetrable…. The United States has long seacoasts. In contrast to the Soviet Union, the bulk of our population lies along the coast." (Indeed, the United States long ago dismantled the bulk of its air-defense system because of its impracticality in the nuclear age.) Furthermore, "there is no foreseeable way that we can preclude [cruise] missiles' impacting on our cities—even if we had a perfect ballistic missile defense." The Carter OTA background paper adds that "a desperate Soviet Union could introduce nuclear weapons into the United States on commercial airliners, ships, packing crates, diplomatic pouch, etc." And it notes that even if all these methods somehow failed, "other methods of mass destruction or terrorism would be feasible for the U.S.S.R., including sabotage of dams or nuclear power plants, bacteriological attack, contaminating water, producing tidal waves with near-coastal underwater detonations, and so on."^[9]

As Bundy and his colleagues observe, "the overwhelming consensus of the nation's technical community is that in fact there is no prospect whatever that science and technology can, at any time in the next several decades, make nuclear weapons 'impotent and obsolete.'" Ashton Carter's OTA background paper, based on "full access to classified information and studies performed for the executive Branch," concludes as follows: "The prospect that emerging 'Star Wars' technologies, when further developed, will provide a perfect or near-perfect defense system … is so remote that it should not serve as the basis of public expectation or national policy." Even Lieutenant General James Abrahamson, former commander of the Star Wars program, has flatly stated that "a perfect defense is not a realistic thing." Under Secretary of Defense Richard DeLauer, the department's senior technical official, admitted in congressional hearings that "there's no way an enemy can't overwhelm your defenses if he wants to badly enough." The Reagan administration suppressed one of its own high-level technical reports that came to the same conclusion about SDI: "The report doesn't boldly state that the plan is idiotic. But after reading the list of disadvantages, only a fool could come to the wrong conclusion."^[10]

Reactions such as this prompted several new twists in the president's argument to shore it up politically and lend it at least minimal scientific credibility. It was necessary to show how an imperfect defense might still protect Americans in a nuclear attack—a very small one, that is, launched by terrorists, a minor nuclear power, or even by the Soviet Union if it heavily cut back its offensive nuclear forces or made a small accidental launch. This justification is reminiscent of one rationale for the original American ABM of the 196Os: the need to defend against the nascent Chinese nuclear threat.

An imperfect defense against a small-scale attack is a far less demanding goal than foiling a large-scale attack and could possibly be achieved to an extent. But the importance of such a defense should not be exaggerated. A "mad captain" on either side might be able to launch an unauthorized attack—but an attack by even a single modern nuclear submarine would not in fact be very small and probably would not be calmly regarded as accidental by the nation at the receiving end. And as Ashton Carter argues, "Emerging nuclear powers or terrorists would be unlikely to use ICBMs [what Star Wars defends against] to deliver their small nuclear arsenals to the United States."^[11]

A committed nuclear attacker or terrorist will probably always be able to detonate one or more warheads on American soil no matter what defenses are built. And Rathjens and Ruina note that "a reasonably

thin area defense intended to protect against some unauthorized, accidental, or third-country attacks can be developed with existing technology. With time, such systems can surely be improved, but we need not await breakthroughs in technology resulting from the SDI program to address realistically the question of the net benefits of such a defense and to decide whether it should be deployed." Noted weapons scientist Richard Garwin asks, "If protecting the United States against accidental launches of Russian ICBMs is truly important, why wait for an elaborate defense? The United States and the Soviet Union could more easily and cheaply protect themselves against accidental launches by installing the command-destruct radio receivers commonly used in test firings of their operational missiles." Physicist Sherman Frankel points out that Star Wars only "detracts from getting on with deploying" such devices "on our huge arsenal of missiles."^[12]

Fred Hoffman told a Senate armed services subcommittee in March 1985 that the inability of Star Wars to defend U.S. cities is irrelevant since the Soviets, for fear of retaliation in kind, probably would not attack cities. A Soviet attack on U.S. military targets, Hoffman contends, may "leave the bulk of Western civil society undamaged." Hence even an imperfect defense could increase the number of survivors. But as Sidney Drell and Wolfgang Panofsky point out, "While neither the United States nor the Soviet Union has made the destruction of enemy populations in response to enemy attack an explicit policy objective, both recognize that should a large fraction of the superpowers' arsenals be used—under any doctrine, any choice of pattern of attack, or for any purpose—then the threat to the survival of the two societies is very grave indeed."^[13] In fact, Star Wars could easily increase the number of Soviet missiles hitting the United States if, as is likely, the Soviets acted on worst-case assumptions about the defense, building and firing more missiles than they otherwise would have, and if, as is also likely, the defense did not in fact work.

Most of the technical debate about Star Wars is simply beside the point. There is no way to defend populations against hydrogen weapons, period. It does not matter whether nuclear-pumped lasers, orbital battle stations, and the like can be made to work. The issue of the "popup" defense—Star Wars satellites that would be launched into orbit just before use to replace those destroyed by Soviet antisatellite weapons—illustrates how absurdly technical the debate has become. During 1984 Senate hearings, Harvard's Albert Carnesale conjectured that "if we could deploy a popup system, they could probably deploy a popup system

to destroy ours." Senator Paul Tsongas asked the next logical question: "But you could not assure us that we could not develop an antipopup popup system?"—that is, a way to pop up American satellites to destroy the satellites the Soviets popped up to destroy the satellites the United States popped up to destroy the missiles the Soviets popped up to pop the United States!^[14]

A Shield for Using the Sword

As Schlesinger writes, the debate about Star Wars has undergone a "remarkable transformation": "The argument is no longer that somehow we can protect American cities perfectly. Instead the argument has become that maybe, not definitely but maybe, strategic defense would permit us to improve deterrence—and that the mix of offense and defense would lead to a more stable world."^[15] The rationale is no longer that Star Wars can reduce the consequences of a nuclear war, but rather that it may reduce the risk by decreasing the incentives for a Soviet first strike.

In July 1985, for example, former national security adviser Zbigniew Brzezinski called in the New Republic for "reformulating [SDI] politically and strategically." He argued that "the U.S. should drop or at least de-emphasize President Reagan's idealistic hope for total nuclear defense for all our population." The goals, rather, should be to "reinforce deterrence and promote nuclear stability." On the anniversary of Hiroshima, the Economist wrote that "the case for Star Wars" rested not on a "wildly implausible" defensive screen for American society but on "the possibility of a defensive screen capable of stopping quite a lot of the Soviet warheads aimed at America's nuclear forces and command centres," which "would make it almost impossible for the Russians to risk a disarming first strike against America."^[16]

The new goal of SDI, according to Gerold Yona, the Pentagon's chief Star Wars scientist, is to make "leakage of such low military value as to discourage a first strike." The Defense Department general counsel, William H. Taft IV, wrote that "the purpose of the President's initiative is to strengthen our ability to deter … to reduce the likelihood of war" and not to "save lives in time of war." Physicist Robert Jastrow ("the single most influential proponent of SDI outside the government") contends that Soviet first-strike weapons have "knocked the stuffing out of deterrence" and that only Star Wars can restore stability. Jastrow actually denies that President Reagan ever claimed that Star Wars was intended

to defend American cities. The purpose of Star Wars, for now at least, is to "strengthen and preserve the American deterrent to a Soviet attack."^[17]

Star Wars advocates claim, then, that the system will stabilize deterrence by discouraging a first strike. But critics counter that Star Wars will destabilize deterrence by encouraging a first strike. Schlesinger, for example, worries that Star Wars would "create instabilities during the entire period of deployment…. The advantage of striking first, for either side, would be greater than is the case for terrestrial capabilities." The late Herbert Scoville, president of the Arms Control Association and former high-level official of both the CIA and the Arms Control and Disarmament Agency, agreed that "developing defenses encourages a first strike." Harvard's Stanley Hoffmann, writing in the New York Review of Books, called Star Wars "a major threat to stable deterrence."^[18]

The anti-SDI pledge, signed by most of the faculty of the nation's topranked physics departments as well as by many in other science and engineering departments, says: "The program is a step toward the type of weapons and strategy likely to trigger a nuclear holocaust." Hundreds of prominent scientists, including dozens of Nobel laureates and a majority of members of the National Academy of Sciences, signed a Union of Concerned Scientists appeal stating that space-based missile defenses and related antisatellite weapons "would increase the risk of nuclear war." The Mobilization for Survival's anti—Star Wars literature quotes the extraordinary claim of Robert Bowman, head of the Air Force advanced space programs between 1976 and 1978, that all Star Wars proposals "would be extremely destabilizing, probably triggering the nuclear war which both sides are trying to prevent." Writing on behalf of the same organization, Benjamin Spock warns that "our very survival depends on" stopping Star Wars."^[19]

So we are told on the one hand that Star Wars is required to prevent, and on the other that it would trigger, nuclear war. Both claims, fortunately, are wildly overblown. Star Wars probably cannot affect the risk of nuclear war one way or the other. Because Star Wars cannot change the probable outcome of a nuclear war, it should not change the incentives on either side to start one.

As we saw in Chapter 2, a successful first strike by either side is simply impossible. Star Wars cannot change that. The American deterrent force does not need to be protected more than it already is. Considering in particular the invulnerability of U.S. strategic submarines, further protection of U.S. land-based missiles is of little importance. And

Star Wars would be perhaps the least plausible and most expensive way to approach the job. The United States could protect its missiles by superhardening their silos, moving them around on railways or on roads (as with the Midgetman missile), or defending them with simple ground-based antiballistic missiles (ABMs).^[20] As for protecting command and control, as Charles Glaser points out, there are fewer than one hundred critical U.S. command and control points, and many of them are very fragile; hence "even very effective [ballistic missile defense] could not deny the Soviet Union the ability to destroy these targets with a first strike."^[21]

The fear of an American first-strike capability, backed by Star Wars, is as baseless as the fear of a Soviet one. Many argue that while a leaky space umbrella is indeed no good in a downpour (that is, a Soviet first strike), it can help a lot in a drizzle (that is, ragged Soviet retaliation following an American first strike). But against hydrogen warheads that can each level a city, a leaky umbrella is meaningless in either case. Even after absorbing an American first strike, the Soviet Union could send off enough ballistic missiles (to say nothing of bombers and cruise missiles) to utterly destroy the United States no matter what defenses we build.

As E. P. Thompson acknowledges, with or without Star Wars a successful American first strike is nothing more than an "ideological fiction."^[22] Both sides recognize this. True, the Soviets fear Star Wars. But what they really fear is losing a Star Wars technology race (which could give the United States a big advantage in spin-off technologies) and being forced to develop expensive countermeasures (a hedge against the small chance that Star Wars will actually work to a significant extent). The Soviet Union, we should recall, expressed urgent fear about many previous American nuclear systems, such as MIRVed missiles and the Pershing II and cruise Euromissiles. The United States has expressed similar fears about many Soviet systems, from sputnik to the "window of vulnerability." But these systems never tempted either side to preempt or even to take other steps—such as adopting a policy of launch on warning or putting bombers on airborne alert—that would indicate a genuine fear of first strike. Star Wars will not tempt them either.

As George Rathjens writes, the effects of Star Wars on crisis stability are "only of academic interest…. The prospects for a technically effective defense are so poor that I cannot imagine deployments in this century that would make a difference to the outcome of a nuclear exchange." Rathjens and Jack Ruina add that "there is no realistic prospect of defenses … even … being perceived as that effective."

None would be sufficient to affect "the behavior of nations in times of crisis." And even if we grant the improbable idea that Star Wars could slightly alter strategic stability or perceptions of it, we still cannot know the direction of the change. As Glaser observes, any U.S. nuclear defense program would probably be matched by the Soviets, and in that case "an attacker would face greater uncertainty about both the effectiveness of his attack and the effectiveness of the adversary's retaliation. The net effect of defenses is, therefore, indeterminate."^[23]

Star Wars could change the amount of warning time between the detection of a possible Soviet missile attack and the decision about whether to respond. The best way to shoot down Soviet missiles would be to attack them from space within minutes of launch—so-called boost-phase interception—before they decompose into numerous warhead-carrying reentry vehicles and decoys. With time so short, many fear that U.S. commanders may panic or take a fateful step based on incomplete information, or even that the United States might computerize the split-second decision and thereby invite an accidental war.

But the exotic technology required to attempt boost-phase interception does not exist and may never be perfected. And activation of a boost-phase interception system, whether by accident or following a false alarm of Soviet attack, would probably mean nothing more than laser beams or other weapons shooting harmlessly into empty space. That should in no way be confused with the firing of Minuteman missiles after a false alarm, which would of course be a catastrophe but has nothing to do with Star Wars.

But what if Star Wars is really an offense masquerading as a defense? Could it then contribute to an American first strike? One study by R & D Associates, a Marina Del Ray think tank, warns that "in a matter of hours, a laser defense system powerful enough to cope with the ballistic missile threat can also destroy the enemy's major cities by fire—each city perhaps requiring only several minutes for incineration." An Argonne National Laboratory physicist adds that the fires caused by such an attack might "generate smoke in amounts comparable to the amounts generated in some major nuclear exchange scenarios," possibly triggering "a climatic catastrophe similar to 'nuclear winter.'"^[24] These are appalling prospects, to be sure, but they have nothing to do with a first strike, and the superpowers can of course incinerate each other's cities and perhaps blacken the atmosphere in well under thirty minutes with technology already at hand.

Star Wars might also be used offensively, many critics contend, to attack

Soviet satellites—a logical first step in a U.S. first strike. But as Ashton Carter points out, the loss of photoreconnaissance, communications, and even early warning satellites may seem dramatic but still cannot make a successful first strike possible and hence would not increase the incentives to preempt.^[25]

In 1986 Robert English, a former Pentagon policy analyst, proposed the ultimate scenario for a U.S. first strike using Star Wars. It begins with a lightning "laser attack on the other side's communications networks and early-warning systems, including satellites." High-altitude nuclear explosions would "further paralyze command and control systems" through electromagnetic pulse. Then, "with the victim effectively blinded," the coup de grace: "the launch of space-based missiles against such targets as silos, command bunkers, airfields, and other military facilities…. Lasers might also be used to pin down missiles in their silos until the silos could be destroyed." Thus, "in a matter of minutes, the victim of such an attack might find the bulk of his ICBM and bomber force gone and his command systems in disarray—without having endured any significant damage to his cities and industries."^[26]

True, there would be little if any warning of a science fiction attack from space, but of course the victim could still destroy the attacker, if only from submarines that would remain invulnerable. Unsavory as offensive weapons in space may be, they seem unlikely to upset the nuclear stalemate any more than weapons on land have.

A Logical Circle

Though Star Wars cannot reduce the damage or change the likelihood of nuclear war, it may well transform the nuclear arms race. Advocates contend that the United States can ultimately bargain it away in return for dramatic Soviet concessions on offensive missiles, perhaps as part of grand strategic reductions along the lines of the START treaty. Critics predict the opposite: the demise of arms control, a runaway offensive arms race, and a new arms race in defenses and the militarization of space as the Soviets scramble to counter the American program. Four former high-level U.S. officials wrote in Foreign Affairs of "the president's choice: Star Wars or arms control."^[27]

Whoever is correct, nothing of importance to deterrence is likely to change. Many times in the past the experts on nuclear war have attributed to some new weapon a decisive importance, either positive or negative, and when that proved difficult to sustain they have back-tracked

to the argument that even inconsequential new systems will set off a cycle of responses and counterresponses that will make a difference. As Richard K. Betts points out, concerns about the arms race have mostly been "future-oriented." Thus, "during the 'missile gap' controversy, the issue was whether the Soviets might achieve superiority a few years later. In the late 1970's the raging controversy was whether the 'window of vulnerability' would open by the early 1980's." Yet "when the early 1980's arrived, and the actual operating balance of intercontinental launchers and warheads had moved further in Soviet favor, U.S. concern evaporated because [of] Reagan's plans for strategic modernization and buildup."^[28] Few notice that the feared or hoped-for developments never materialize because deterrence is in fact existential.

Perhaps someday the United States or the Soviet Union will build a weapon that actually changes the nuclear peril. But that is a remote prospect. In the meantime the superpowers are periodically endangering the planet, as they have over the past forty years, through different actions. To these we now turn.