In search of a doctrine for the new atomic age.

The nuclear order seems to be falling apart. Gone is the uneasy balance between the cold war superpowers. We now face a slew of new nuclear actors. North Korea has reprocessed enough plutonium for perhaps ten bombs, in addition to the two it has already tested. Iran’s centrifuge program seems poised to produce weapons-grade uranium. And Syria was apparently constructing a clandestine nuclear facility, before it was destroyed by Israeli air strikes in 2007.

It’s not just enemies that pose a problem. India and Pakistan, which both conducted nuclear tests in 1998, have now each built several dozen weapons, deliverable by missile or aircraft. Tensions between the two nations remain high, and a nuclear war between them would affect us all—the climatological aftermath could theoretically kill enough crops to starve one billion people, according to a recent report. What’s more, it is not inconceivable that an Islamist sympathizer in Pakistan could transfer fissile material or weapons components to Al Qaeda—a danger that would only increase if the government, already unstable, collapsed. A.Q. Khan, after all, proliferated nuclear technology to North Korea, Iran, and Libya, and, although he has been stopped, the illicit trade in nuclear technology has not. Globally, the threat of fissile material theft is so great that President Obama is convening an international summit next week to address it.

In this morass, there seem to be an infinite number of things that could go wrong. The problem is obviously not just the number of nuclear actors—it’s that some of them seem quite mad. American security in the nuclear age has relied on deterrence—the idea that one nuclear state would never attack another because it knew that it would suffer a devastating retaliatory strike—but deterrence seems to rely on a basic degree of rationality that is not readily apparent in either North Korea or Iran, to say nothing of Al Qaeda. The Hermit Kingdom is run by a self-indulgent and often absurd leader who seems to have stepped from Ian Fleming’s imagination, replete with repeated threats to turn Northeast Asia into a “sea of fire.” Ahmadinejad, too, can be buffoonish, but, even if his calls to wipe Israel off the map are nothing more than rhetorical flourish, his country’s embrace of suicide as a tactic—first during its war with Iraq, when the regime used boys to “clear” minefields, and then in its support for Hamas and Hezbollah—begs the question of whether Tehran might not someday launch a nuclear weapon at Tel Aviv, even though it would be destroyed in return. The regime’s disregard for human life is made all the more haunting by the millenarian aspects of Shiism as interpreted by the Iranian leadership, elements of which believe that confronting evil can hasten the return of the “Hidden Imam”—the twelfth direct descendant of the Prophet Muhammad, who went into occultation in 874 C.E. and whose return, it is said, will bring the establishment of a just society.

In the face of such threats, deterrence, never the most reassuring of concepts, seems downright inadequate—to both the left and the right. Conservatives have responded to these new nuclear challenges chiefly with military answers, emphasizing robust missile defenses, flexible offensive forces, and a doctrine of preemption that they hope will enable us to escape the vulnerability of the new nuclear age. Liberals, joined by an assortment of erstwhile hawks like George Shultz and Henry Kissinger, have responded in precisely the opposite way—by promoting nuclear disarmament. Concerned that the spread of nuclear technology threatens international stability, they have pushed for stricter limits on weapons, in the hope of slowing or even rolling back proliferation and reducing the opportunities for terrorists to get their hands on a bomb or its ingredients. President Obama, who has called for a “world without nuclear weapons,” is thus pursuing not only cuts to the U.S. and Russian arsenals, but also the ratification of the test-ban treaty and an accord to end the production of fissile material for weapons.

In short, it is a time of great doctrinal tumult. And yet, it shouldn’t be. Yes, the nuclear landscape has shifted considerably since the cold war. But the questions underlying the nature of deterrence and the utility of nuclear weapons are the same ones we’ve been trying to answer for years—and, in fact, the lessons from our standoff with the Soviets are of great relevance and import today. The principal danger lies not in the uniqueness of our current situation, but in the temptation to overvalue its novelty. In truth, the debate over nuclear weapons is trapped in a loop, and, as the administration submits the new start treaty for ratification, releases its nuclear posture review, and attempts to strengthen nonproliferation efforts, we’re about to go another lap. We have had the answers to nuclear mysteries for a long time. The only real question today is whether we pay attention to them.


The basics of nuclear deterrence were established early. In a book titled The Absolute Weapon, published less than a year after the destruction of Hiroshima and Nagasaki, Bernard Brodie, a professor and naval strategist, wrote that the atomic bomb had changed the very nature of warfare. A few bombs could destroy a country’s cities, and, because no defense was perfect, any nuclear war was likely to result in intolerable devastation. This meant that our military goal had to be not victory, but deterrence. “Thus far the chief purpose of our military establishment has been to win wars. From now on its chief purpose must be to avert them. It can have almost no other useful purpose.” Besides the capability to respond after an enemy attack, all that it seemed deterrence required was the rationality to understand the devastation a nuclear war would cause for all involved.

Unfortunately, it wasn’t clear that the Soviets possessed even that modest degree of reason. Nikita Khrushchev, for one, delighted in taunting the Americans with the Soviets’ supposed nuclear superiority, boasting that the Soviet Union was turning out missiles “like sausages.” During the Suez crisis in 1956, he threatened the British and the French with “rocket weapons” if they did not withdraw their forces from the canal. And, in 1958, during a visit to Moscow by Hubert Humphrey, Khrushchev asked the Minnesota senator what town he was from and then circled Minneapolis on a map, saying, “[T]hat’s so I don’t forget to order them to spare the city when the rockets fly.” Today, there is a tendency to remember the Soviet Union as a known quantity—a “generally status quo, risk averse adversary,” as George W. Bush’s administration put it in the 2002 National Security Strategy—but the contemporary reality was quite different. NSC-68, the famous study commissioned by President Truman and written by Paul Nitze in 1950, warned that, when Moscow “calculates that it has a sufficient atomic capability to make a surprise attack on us, nullifying our atomic superiority and creating a military situation in its favor, the Kremlin might be tempted to strike swiftly and with stealth.”

Deterrence, after all, was a defensive strategy which implied that one accepted the world as it was, and the Soviet Union, with its belief in world revolution, obviously did not. What’s more, the communists, as leaders of a totalitarian society, were thought to have an advantage in the stealth that Nitze feared. (Long before Ahmadinejad used knowledge gleaned from his time with the Iranian Tunneling Association to push his country’s nuclear program underground, Khrushchev, who had overseen construction of Moscow’s subway system, thought to bury Soviet missile silos so they could evade detection and survive a nuclear strike.) Finally, Soviet losses during World War II were thought to show indifference to human life: Given that the Soviets had absorbed 20 million casualties during World War II, strategists like Albert Wohlstetter argued they would be willing to tolerate such carnage in a future conflict. Indeed, their basic moral fiber was considered alien. General Thomas Power, who had headed the Strategic Air Command, wrote, “Moral principles would deter us strongly from launching a pre-emptive war … but moral considerations and the prospect of losing Russian lives and cities would not deter the Soviets.”

These critiques persisted throughout the cold war, but the Soviets were not the only ones making wild-eyed threats. The Eisenhower administration advanced a doctrine of “massive retaliation,” whereby the bomb would be used to respond to a wide range of aggression. Ike’s secretary of state, John Foster Dulles, publicly lauded the virtues of “brinksmanship,” a strategy by which it was rational to seem irrational, in order to frighten your adversary into backing down. The war plan approved in the final year of Eisenhower’s presidency allowed for the preemptive use of the entire U.S. nuclear arsenal—nearly 3,500 weapons—against targets in not only the Soviet Union but also Eastern Europe and China. In a 1961 briefing, Power, so critical of Soviet morality, joked to Robert McNamara, “Mr. Secretary, I hope you don’t have any friends or relations in Albania because we’re just going to have to wipe it out.” In both its willingness to take life and its willingness to risk retaliation, then, the U.S. nuclear posture in the late ’50s and early ’60s was itself of debatable rationality.

Ironically, attempts to introduce greater rationality to our war plans only made things worse. In the 1960s—at the recommendation of a new crop of defense intellectuals, many practicing game theory, the academic epitome of reason—massive retaliation was replaced by a strategy known as counterforce, whereby the United States would fight a nuclear war just like any other war, attacking the enemy’s forces and defending against its attacks in an attempt to limit and win the conflict. Its greater flexibility seemed far more reasonable than a “wargasm,” as Herman Kahn derided the Eisenhower approach. But it didn’t actually make nuclear war any more survivable. Targeting enemy missiles and bombers could never prevent Soviet retaliation because the Soviets would always make sure they had enough weapons to respond after a first strike. Missile defenses and bomb shelters were of little solace because study soon showed that they could be overcome easily and cheaply. Finally, if game theory suggested that counterforce was rational, it also demonstrated how ostensibly rational decision-making could lead to disaster. In a crisis, some advantage, however Pyrrhic, would accrue to the side that struck first, and, as theorist Thomas Schelling warned, the resulting “reciprocal fear of surprise attack” could lead to a nuclear war that neither wanted and that could destroy both. Rationality would not preclude catastrophe; it might well precipitate it.

As Brodie had feared, no matter what strategy either side pursued, a nuclear war would destroy both the United States and the Soviet Union. This condition became known as mutual assured destruction (MAD), but the strategy it suggested—a defensive posture designed for retaliation instead of attack—never became the basis for U.S. war plans. True, McNamara accepted the constraints of MAD—and they became the rationale for SALT I's limits on offenses and the ABM Treaty’s restrictions on missile defenses—but conservatives hated the perpetual vulnerability that MAD implied and made it a political issue. (Indeed, MAD was given its memorable acronym in a National Review article, which argued that it was insane.) Others, like Nitze and Wohlstetter, worried that the Soviets could use nuclear superiority to throw their weight around, leaving us trapped between suicide and surrender. And, given that we were using these weapons not just to deter a nuclear attack but to deter a conventional attack on Europe, how did MAD apply if the Red Army came trundling through the Fulda Gap?

So, even as it realized the inescapability of MAD, the United States rejected it. The Nixon administration pursued new multiple-warhead missiles designed for warfighting, and the Carter administration introduced ever-more-nuanced counterforce options. Reagan argued that the Soviets did not believe in MAD and so he not only called for increased spending on offensive nuclear weapons, but pursued civil defenses and, of course, the “Star Wars” missile-defense program. His Pentagon, meanwhile, prepared to fight and win a “protracted” nuclear war. And, just as Schelling and others had predicted, these attempts put the United States in greater danger: The Soviets became convinced the United States was preparing for a first strike, and, during a NATO training exercise in 1983, actually concluded that war was imminent, bringing the superpowers closer to a nuclear exchange than any time since the Cuban missile crisis.

The debate over Moscow’s strategy did not end until the cold war did. But today, we have a much better idea of what the Soviets thought. Last fall, for example, the National Security Archive released a newly declassified study commissioned by the Pentagon in the 1990s. According to interviews with former Soviet military officials, the Soviet Union had realized by the late ’60s that a nuclear war was unwinnable, and its strategy was far more defensive than ours was. Its doctrine called for a massive nuclear response if the United States used nuclear weapons (“all against any”), and, although the Soviets did contemplate more proportional responses, they “doubted that a nuclear war could remain limited for long.” The Soviets constantly feared a U.S. first strike and thought they might be called upon to act preemptively. In other words, the Soviets did understand that a nuclear war would be mutually suicidal and therefore had to be avoided, and U.S. efforts to get around deterrence via counterforce were destabilizing.

Of course, one oral history is not dispositive. But, regardless of what the Soviets thought, nuclear war did not happen. That is, in the face of the most aggressive, most highly armed, most revolutionary power the United States has ever known, deterrence worked. It worked despite serious fears about the enemy’s rationality. Indeed, it may have demonstrated that rationality is not the appropriate prerequisite for nuclear stability. Rationality can produce undesirable outcomes; it does not preclude crisis situations (Khrushchev was not insane when he ordered missiles to Cuba, he was just wrong); and in the heat of nuclear battle, rationality is unlikely to guide decisions in any country, regardless of ideology. Demanding rationality of our enemies is therefore both asking too much and asking too little. It is perhaps best that, as the scholar Kenneth Waltz has noted, “Deterrence does not depend on rationality. It depends on fear.”

Fear, after all, is an evolutionary imperative in a way that reason is not, and it induces caution in a way that can be understated by cold cost-benefit analyses. No state that values its continued existence would launch an attack that meant its own certain devastation, and there is every indication—from their oppression at home and their manipulations abroad—that the leaders of Iran and North Korea have every desire to survive. True, historically, leaders have made strategic errors that resulted in their downfall. But they did so because they miscalculated their odds of success—an error that is impossible to make in launching a nuclear strike against an adversary that clearly has the capability to retaliate. The only plausible suicide would be an assisted one in which, say, Pyongyang’s leaders feared total military defeat—deterrence does not cover “dictators in the mood of Hitler when he found himself in his final dugout,” as Churchill once put it. That means the United States shouldn’t push nuclear-armed leaders to the brink of extermination, but otherwise deterrence should hold.


For many hawks, however, the most troubling aspect of proliferation is precisely that it will markedly reduce U.S. freedom of action—that third-rate nations will suddenly become potent challengers, leaving the United States vulnerable to coercion. Thus, the real threat from an Iranian bomb would be not direct attack on the United States or Israel, but a raft of emboldened theocrats with ambitions to regional dominance. As Ashton Carter, whom Obama would appoint under secretary of defense, said in September 2008, “I think possession, even if [the Iranians] never use it, just simple possession is a disaster all by itself. I think it will change their behavior. I think it’s a shield behind which they will be emboldened to do things contrary to our interests, and to the stability of the region, that they wouldn’t dare if they didn’t have the bomb.”

There is some evidence that nuclear-armed states behave more aggressively. For example, in 1999, Pakistani forces, arguably emboldened by their nuclear arsenal, crossed the Line of Control in an attempt to cut off Indian positions on the Siachen Glacier. But, after weeks of fighting, Pakistan was ultimately forced to back down, in what became known as the Kargil crisis. Emboldenment is not the same thing as empowerment.

Nuclear weapons have certainly not given the United States carte blanche. Consider that, when Washington’s nuclear advantage over the Soviets was greatest—when it had a nuclear monopoly in the late ’40s—it was unable to prevent them from absorbing Eastern Europe and blocking access to Berlin. Similarly, the Chinese entered the Korean war in 1950 even though our atomic arsenal could have wiped out Beijing and Shanghai as easily as Hiroshima and Nagasaki. More recently, nukes have provided no assistance in Iraq and Afghanistan. If nuclear weapons yielded significant diplomatic or even military leverage, the United States would be a juggernaut, able to get its way in most, if not all, confrontations. This has not been the case.

That does not mean the United States and others have not tried to throw around their nuclear weight in an attempt to wield influence—particularly during crises—but evidence concerning the utility of such threats has been highly ambiguous. The tendency to rattle one’s nuclear saber may be less a function of the balance of power than what strategists call the “balance of resolve”—the measure of how much a state cares about a given issue. And, because states care most about their continued existence, nuclear weapons are far better at deterrence (preventing a country from attacking you) than at compellence (getting that country to do something you want).

The explanation for why Iran would feel emboldened—that suddenly it had become immune to full-scale attack from the United States—rests on the false assumption that the United States currently uses the threat of such an attack to influence Iranian policy or constrain its activities. But we don’t, and therefore removing it as an option would not change the status quo. As Negeen Pegahi of Harvard notes, “The spread of nuclear weapons carries real strategic costs for the United States: we will no longer be able to invade Iran or otherwise act to forcibly change its regime. That’s the bad news. The good news is that this constraint on U.S. action will not provide new opportunities for Iran. Since the United States does not rely on threats to the survival of the Iranian state or regime to coerce Iran, the loss of these options will not embolden Iran.” Besides, to the extent that the United States has implicitly threatened Iranian regime change—for example, through comments about keeping “all options on the table”—it has clearly been ineffective. Not only has Iran accelerated its nuclear efforts, but it remains “the most significant state sponsor of terrorism,” according to the State Department, and it has challenged U.S. ships in the Gulf. In other words, Iran is already quite bold.

Even if a nuclear-armed Iran were more aggressive, the United States could still deal with it forcefully. India and Pakistan, after all, fought directly and bloodily even when both states had nuclear weapons; and an Iranian incursion into a neighboring country could be met with force. Indeed, non-nuclear states have attacked nuclear states without apocalyptic consequences: In 1973, Israel’s nascent nuclear capability was not enough to prevent Egypt and Syria from attacking it, and, of course, in the subsequent decades it has found itself under near-constant challenge from state-sponsored terrorism and even missile attack during the Gulf war. Nor are nuclear states immune from nonmilitary regime-change efforts—there is no reason we could not support the Green Movement in a nuclear-armed Iran. We simply lose the ability to invade or militarily overthrow the regime. Which makes the rationale for attacking Iran’s nuclear program seem vaguely ridiculous: Regardless of whether we stopped or even delayed the program, we would essentially be taking military action against Iran in order to preserve our ability to take military action against Iran.

But the situation raises an important question as well: If nuclear deterrence is stable and if nuclear weapons don’t provide their owners a measurable increase in influence, why worry at all about proliferation?


On the evening of September 26, 1983, Stanislav Petrov, a lieutenant colonel in the Soviet Union’s Strategic Rocket Forces, was on duty in the early warning station he commanded. It was a fraught period in U.S.-Soviet relations. In March, Reagan had called the Soviet Union an “evil empire”; U.S. Pershing missiles were on their way to Europe; and, only three weeks earlier, the Soviets had shot down a Korean Airlines jet. The Soviet premier at the time was Yuri Andropov, a former spy and unrepentant paranoid who thought it entirely possible that the Americans would someday launch a first strike. It was amid this tension, as David Hoffman recounts in his book The Dead Hand, that Petrov’s warning panel began repeatedly flashing the word LAUNCH, indicating that American icbms were inbound. Petrov had only minutes to decide what to do; a missile fired from the continental United States would take only half an hour to reach the Soviet Union, giving the Soviets little time to authenticate an attack and launch their own missiles before they were destroyed on the ground. After reviewing other data, Petrov took a risk and declared the “attack” a false alarm, despite what his instruments said. He was right, of course, but, given the available information—or a more panicked state of mind—he could easily have gone the other way.

It is important to remember moments like Petrov’s because, despite the general stability of deterrence and the limited diplomatic and military utility of nuclear weapons, they remain an enormous threat. Yet some have taken the relative peace of the nuclear age as a sign that we can relax. A recent article in the prestigious journal International Security, for example, lamented the prevalence of “nuclear alarmism,” and, last year, scholar John Mueller published a book titled Atomic Obsession, which argued that fears of nuclear weapons are baseless. Kenneth Waltz has long and famously argued that “more may be better”—that is, he has lauded the benefits of nuclear proliferation—and his argument is experiencing a resurgence. Last August, Newsweek ran a story headlined: “WHY OBAMA SHOULD LEARN TO LOVE THE BOMB.”

Waltz and his fellow travelers present a simple and persuasive line of reasoning: Because they make the prospect of all-out war unthinkable, nuclear weapons keep the peace, or at least limit conflict. States can fight only for small gains because any attack against the vital interests of an adversary risks nuclear retaliation. Certainly, the United States and the Soviet Union tried to avoid war, fighting only via proxies despite mutual antagonism and conflicting interests, and the result was a tenuous peace.

Beneath the cold war’s veneer of “nuclear stability,” however, lay significant volatility. Stability relied on the belief that each side’s ability to retaliate against a first strike made war unthinkable, but, although the United States and the Soviet Union both had that capability, neither ever felt secure. The result was an arms race in which relations were undermined by fear and pocked by world-threatening crises—as in 1962, when Soviet missiles were discovered in Cuba, and in 1983, when Moscow mistook that NATO exercise as preparation for a U.S. first strike. Each of these incidents was a crisis precisely because one or both sides feared that war, whatever its consequences, was not “unthinkable.”

One could argue that, however frightening such episodes were, they were resolved without recourse to nuclear weapons—proving Waltz’s point. Indeed, some South Asia scholars, like Šumit Ganguly in his new book India, Pakistan, and the Bomb, argue that, even though nuclear weapons may have encouraged Pakistan to pursue its revanchist claims, they also kept a lid on the resulting conflict, via what political scientists call the stability-instability paradox. During the Kargil crisis, for example, Indian policymakers not only rejected an all-out conventional assault against Pakistan, they made sure their troops limited their attack to Pakistani forces on the Indian side of the Line of Control. And yet, the peaceful resolution of cold war crises often seemed less a function of caution than of luck, in no small part because states are not always in control of their own actions. The care that deterrence requires often flounders amid the absurdity of the human experience.

The Cuban missile crisis, for example, was plagued by accidents and freelancing. In the middle of the “13 days,” officers at Vandenberg Air Force Base in California mated nuclear warheads to nine of their ten Titan, Atlas, and Minuteman missiles—an act that the Soviets could reasonably have perceived as preparations for launch. As if that weren’t enough, they then went ahead on October 26 with a pre-scheduled flight test, actually launching the tenth ICBM over the Pacific. On the tensest day of the crisis, Soviet troops in Cuba shot down a U-2 on a reconnaissance mission, a provocation that Khrushchev later lamented. The shootdown led the Joint Chiefs to redouble their arguments for an air strike on the island—a move that could have resulted in nuclear war because Khrushchev’s field commander had deployed tactical nuclear weapons and might well have used them. Kennedy wasn’t fully in control of the U.S. nuclear arsenal, either: According to Stanford political scientist Scott Sagan, during the crisis, Air Force officers in Montana “jerry-rigged their Minuteman missiles to give themselves the independent ability to launch missiles immediately,” in gross violation of military regulations.

Whatever stability they provide, then, nuclear weapons also generate a distinctly non-trivial chance of total catastrophe, and today, that chance is accentuated by the threat from nuclear terrorism. It is unlikely, but certainly not inconceivable, that a state like Iran, long a sponsor of terrorism, could share nuclear technology with radicals—or that a radical sympathizer could divert fissile material from one of its enrichment facilities. States have long had trouble maintaining a perfect grip on their arsenals. When the Soviet Union broke up, it found itself riddled with poorly guarded weapons and fissile material that we are still struggling to lock up today. Pakistan’s government insists that its arsenal is secure, but the government itself is not secure. Even the U.S. arsenal has suffered numerous accidents and security problems (most notably the loss of eleven nuclear bombs during the cold war), and, in January, a group of peace activists managed to break into a NATO base where U.S. tactical nuclear weapons are stored.

Deterrence, then, whatever its strength, is susceptible to Murphy’s Law. Shit happens, it happens with remarkable frequency, it happens even when the consequences could be nothing less than the destruction of an entire country or worse, and it is obviously more likely to happen when more states have nuclear weapons.


If the history ofnuclear weapons contains clear lessons, those lessons don’t provide simple answers. Deterrence is strong but imperfect. Nuclear weapons are an excellent defense, but of debatable influence. We can likely live with proliferation, but we should fight it fiercely because the chance of error grows with every new nuclear state. Formulating policy that incorporates these complexities is challenging, particularly when it comes to nuclear terrorism.

Nuclear deterrence is of little utility against terrorists, who, unlike states, lack a “return address”—a city or other suitably valuable target that could be destroyed in retaliation. It might be possible to retaliate against the terrorists themselves, but that is not a mission for nuclear weapons; and, if U.S. military efforts to date had failed to deter them, it is unclear why further threats of punishment would. The more effective deterrent might be defensive rather than punitive, raising the difficulty of mounting an attack—whether by preventing access to fissile material or installing radiation monitors in ports or anything in between—to convince a would-be nuclear terrorist that the low odds of success would not be worth the effort.

The United States could threaten—indeed, the Bush administration did threaten—to retaliate against any state that helps a terrorist carry out a nuclear attack. (Terrorists don’t have the ability to produce the highly enriched uranium or plutonium needed to build a bomb, so a state has to be involved, wittingly or not.) But, as Michael Levi of the Council on Foreign Relations has pointed out, even threatening to retaliate against states that aid in a nuclear terrorist attack could undermine the cooperation on nuclear security that we currently have with Russia and Pakistan. In fact, it could, perversely, discourage them from reporting any theft of fissile material out of fear they would be held accountable for its use, thereby reducing the chances of recovering that material before it was used. True, such a threat might scare Iran and North Korea, with whom we have little cooperation anyway, into jealously guarding their nuclear material. But, even then, a threat might be less than credible: Despite advances in nuclear forensics—the study of the signature produced by each enrichment and reprocessing facility—our ability to trace fissile material to its source is imperfect.

The good news is that America’s ability to deter other nuclear states is stronger than it was during the cold war. Not only do we need not fear a nuclear war with the Soviets, we no longer need nuclear weapons to offset enemy conventional weapons. Our new START agreement with Russia only further stabilizes the countries’ nuclear relationship, adding predictability and transparency to it; and the 1,550 deployed warheads we will retain are more than enough for any conceivable mission—certainly enough to deter Iran and North Korea. Conservatives, however, fearing the fragility of deterrence and the possibility of blackmail by rogue states, have argued for a more aggressive nuclear posture. The Bush administration called for improved offensive weapons, including new low-yield and bunker-busting bombs; robust, multi-tiered missile defenses; and a nuclear doctrine that emphasized preemption over retaliation. Much as during the cold war, the right hopes not simply to strengthen deterrence by preparing to fight a nuclear war, but to escape it, eliding the vulnerability we feel when an enemy has nuclear weapons.

It’s an understandable instinct, and there is some reason to believe that counterforce could be more valuable now than during the cold war. A nuclear-armed Iran, for example, would have far fewer weapons than the Soviet Union did, and, given advances in our missile accuracy and in our missile-defense capabilities, some believe we could escape a nuclear exchange unscathed. But that is a fantasy built on a risk we would never be willing to take. Most Iranian missiles are mobile, mounted on large trucks, and therefore difficult to eliminate in a preemptive strike. Our ballistic-missile defenses, meanwhile, are far from perfect, and even a perfect system could not stop airplanes or cruise missiles or nuclear suicide bombers in a boat or a truck—all ways in which a nuclear Iran could respond to an attack. What’s more, although Washington’s relations with Moscow and Beijing are hardly tense enough for either to fear a bolt from the blue, the Taiwan Strait remains one of the world’s most dangerous flash points, and perceived attempts to maintain a nuclear “advantage” will only undercut the cooperation we need from both to deal with challenges like Iran.

By contrast, nuclear restraint could prove useful. Even though our nuclear posture can play little role in deterring terrorism, a less aggressive posture can play a role in preventing it. The Nuclear Nonproliferation Treaty (NPT) requires most states to forsake nuclear weapons, but, in exchange for their tolerance of the two-tiered system this creates, they get several benefits: They have the reassurance of knowing that rival states will remain nuclear-free, they are guaranteed access to nuclear technology for peaceful purposes—like energy production—and they were promised that the nuclear-weapons states would work toward disarmament. Admittedly, states like North Korea and Iran are not basing their nuclear decisions on U.S. compliance or lack thereof with the NPT, but other states, whose cooperation we need to pressure violators and create a more secure nuclear world, do. For example, Mohamed ElBaradei has said that the Senate’s 1999 rejection of the Comprehensive Test Ban Treaty was “a devastating blow to our efforts to gain acceptance of more intrusive inspections of nuclear facilities around the world.” Our own nuclear posture, therefore, can retard the spread of nuclear weapons. And, to the extent that we discourage proliferation, we reduce the opportunities both for catastrophe and for aspiring nuclear terrorists to acquire the tools they need.

That is one reason why the new START agreement is crucial, as are efforts like the test-ban treaty. Particularly given our overwhelming conventional might, the less salient nuclear weapons become in international affairs, the more secure the United States becomes. The president’s stated goal of abolition is the logical end of this chain of reasoning. And—ironically, since conservatives deride it as naïve—it is the only true escape from deterrence. Abolition may or may not be feasible in our lifetimes, but, for the time being, it doesn’t matter. In stark contrast with counterforce, the steps taken toward disarmament do not undermine the nature of deterrence itself. Deterrence can remain stable with fewer nuclear weapons, without new nuclear weapons, and without nuclear testing. The vulnerability of deterrence may be unpleasant, but the only way to escape it will be to move through the nuclear age. We certainly cannot go around it.

Peter Scoblic is executive editor of The New Republic. Research for this article was supported by the Ploughshares Fund.