Courtesy Reuters


Jan Lodal

Keir Lieber and Daryl Press ("The Nukes We Need," November/December 2009) argue that to deter the growing number of nuclear-armed states against which it might have to fight a conventional war, the United States should develop a new generation of accurate low-yield nuclear weapons. They contend that "the least bad option in the face of explicit nuclear threats or after a limited nuclear strike may be a counterforce attack to prevent further nuclear use."

It is true that for the United States to maintain nuclear deterrence, the president must have credible options to respond to nuclear threats or attacks. Lieber and Press rightly assert that the capability to destroy enemy cities with high-yield weapons is not enough. But their argument for new counterforce capabilities attacks a straw man. The United States already has the flexibility to carry out low-yield counterforce attacks, and there are no plans to eliminate this. The B-61 nuclear bomb has a variable yield that can be set quite low and is highly accurate, especially when carried by the stealth B-2 bomber. Cruise missiles with low-yield warheads have similar capabilities. Even long-range ballistic missiles can be targeted to minimize collateral damage.

Lieber and Press go beyond urging low-yield counterforce capabilities and propose a bizarre and dangerous nuclear strategy for the United States: to develop the capacity for attacks against a threatening enemy that would prevent the enemy from launching any subsequent nuclear attacks. These disarming strikes would be launched even if the enemy had attacked an isolated military target, such as a carrier battle group at sea. Astoundingly, the authors also propose preemptive nuclear attacks against "explicit nuclear threats." The states against which such attacks might be used include Iran, North Korea, other new nuclear powers, and even China.

Lieber and Press provide a detailed analysis of how such an attack would have a more than 95 percent chance of destroying all of China's fixed silo-based intercontinental missiles, with less than 700 fatalities. (Much of this analysis is identical to the arguments made by the George W. Bush administration for its program to develop a low-yield nuclear bunker buster -- a program ultimately blocked by Congress.) Yet there is no chance whatsoever that any U.S. president would launch such an attack -- certainly not against China, nor even against a much less capable nuclear power. The challenge in modern warfare is not hitting a target at a known and fixed location; the challenge is to know the target's location. China's capabilities are not limited to the 20 land-based silos Lieber and Press confidently predict could be destroyed in a single attack -- China also has mobile missiles, bombers, ships, and submarines. Do the Chinese believe the United States could destroy these mobile systems?

Until recently, the United States overlooked an entire nuclear material processing facility in Iran. Washington does not know where North Korea stores its crude devices, nor for that matter where India, Israel, or Pakistan keeps its weapons. A nuclear weapon could even be hidden on a pleasure boat, tens of thousands of which traverse U.S. waterways each day. Nuclear-armed pleasure boats could never defeat the United States or prevent the ultimate defeat of an enemy that has them, but the possibility of their existence does make the idea of a totally disarming attack against an adversary's nuclear forces nonsense.

Fortunately, there is no plausible scenario for a conventional war with China that might trigger a Chinese nuclear attack against the United States, given the United States' enormous conventional and nuclear superiority and the lack of any geopolitical motive for such a confrontation. Other powers would also be deterred from threatening or using nuclear weapons by the United States' highly accurate and flexible conventional and nuclear forces.

Lieber and Press' concept for nuclear war fighting, like all the many others before it, collapses under even the most superficial examination. It does not even address the most immediate threat: that a terrorist organization will acquire a nuclear weapon. By proposing yet another unworkable nuclear strategy, their article emphasizes that only by eliminating nuclear weapons can the world be protected from the existential threat they pose.

Elimination presents two immense challenges: convincing all states that there is no valid purpose for nuclear weapons other than to prevent the use of them by others (so that countries can give up nuclear weapons if all others will do so also) and putting in place a comprehensive worldwide nuclear-control regime to give all states confidence in the weapons' elimination. These goals should be the focus of U.S. nuclear policy -- not unworkable and unnecessary new missions for nuclear weapons.

JAN LODAL is immediate past President of the Atlantic Council of the United States and was a senior U.S. Defense Department and White House official in the administrations of Richard Nixon, Gerald Ford, and Bill Clinton.


James M. Acton

Keir Lieber and Daryl Press raise a salient and important question: How can the United States deter nuclear use in a conflict against a nuclear-armed adversary that has much more at stake than the United States does? They are right that the threat of a massive nuclear strike against an adversary's cities would lack credibility (such a strategy is, of course, not current U.S. policy). To bolster deterrence, they suggest that the United States further refine its counterforce capabilities so that it can confidently eliminate any enemy ballistic missiles based in hardened silos.

If the United States really did face an adversary with a small nuclear arsenal consisting purely of silo-based missiles -- or were likely to face such an enemy in the future -- there may indeed be a case for a force consisting of more low-yield, high-accuracy weapons. Even then, however, because enhanced U.S. counterforce capabilities would simply encourage an adversary to launch more of its weapons initially (leaving fewer behind as targets), there would not be a clear-cut case for such a force.

The main problem with Lieber and Press' argument is that no state actually has a small arsenal consisting solely of silo-based missiles. China, Iran, and North Korea are all focusing on the development of road-mobile missiles (in fact, the latter two do not appear to have any silo-based weapons at all). Although China has very few road-mobile missiles that could reach U.S. soil, and Iran and North Korea have none, each country has plenty that could reach the territory of key U.S. friends and allies.

The challenge with destroying road-mobile weapons is locating them. If their location is known, conventional munitions will suffice. If their location is not known, even nuclear weapons are useless (discounting the possibility of wide-area nuclear bombardment, which Lieber and Press would presumably not advocate). Locating mobile ballistic missiles is exceptionally hard. According to the Gulf War Air Power Survey (an official analysis of U.S. Air Force operations during the Persian Gulf War), the United States launched about 1,500 sorties against Scud launchers in Iraq during the 1991 war; not a single mobile launcher was confirmed destroyed.

Granted, U.S. capabilities and doctrine have improved markedly since then. Nonetheless, it is still fiendishly difficult to locate mobile missiles hidden by a well-prepared enemy. The bottom line is that because the bulk of China's, Iran's, and North Korea's missile forces are mobile, the United States could not eliminate their entire arsenals with a high probability. Increasing the United States' ability to eliminate only silo-based weapons would add very little to deterrence.

Rather than seeking to deny China's second-strike capability, the United States would be better off acknowledging that its relationship with China is one of mutual vulnerability (as it did with the Soviet Union and then Russia). Even in the face of U.S. missile defense measures and enhanced counterforce capabilities, China can preserve its capability to inflict unacceptable damage on the United States. It can fit its missiles with multiple warheads, develop more sophisticated countermeasures, or simply build more missiles.

Instead of pursuing strategic competition with China, therefore, the United States should seek strategic stability through dialogue and arms control, recognizing that mutual vulnerability is a fact, not a choice. The United States' failure to accept this is not the only -- or even the main -- reason for the failure of the strategic dialogue that President George W. Bush tried to initiate with China; China's opacity about its nuclear doctrine and its unwillingness to enter into a serious, candid discussion are also significant barriers. Nevertheless, by acknowledging mutual vulnerability, the United States would increase the chances of having a much-needed dialogue.

North Korea and a putative nuclear-armed Iran require a different approach. First, the United States should continue its policy of seeking behavioral change and not regime change. Lieber and Press rightly argue that deterrence is most likely to fail when a regime believes that it is about to be toppled and that the only chance to save itself is to use a nuclear weapon. The United States could change this calculus -- and give an embattled regime a very strong reason for nuclear restraint -- by not seeking regime change, unless the United States or its friends and allies are attacked with nuclear weapons. The problem here, as Lieber and Press note, is that the U.S. military's battlefield doctrine of decapitating the enemy undermines this strategy. The solution is to change tactics; if the president says he does not seek regime change, the military should not pursue it.

Second, the surest and best way of avoiding nuclear use is preventing nuclear proliferation (or, in the case of North Korea, rolling it back). A successful nonproliferation regime requires a broad-based international consensus. As President Barack Obama recognized in his April 2009 speech in Prague, creating such a consensus will require the United States to work toward a world without nuclear weapons. Devaluing nuclear weapons is an important step on this path. The further development of nuclear counterforce capabilities will only make this goal harder to achieve.

JAMES M. ACTON is an Associate in the Nuclear Policy Program at the Carnegie Endowment for International Peace.


Hans M. Kristensen, Matthew McKinzie, and Ivan Oelrich

Keir Lieber and Daryl Press assert that a joint study by the Federation of American Scientists and the Natural Resources Defense Council from 2006 exaggerates the radioactive fallout that would result from a U.S. nuclear attack on China's estimated 20 missile silos. Instead, they suggest, the silos could be destroyed by high-precision, low-yield weapons, causing little radioactive fallout and very few casualties.

But Lieber and Press have overestimated the effectiveness of the highly accurate, low-yield nuclear airbursts they prescribe. The probability of destroying a target depends on the accuracy of the missiles and the yield of the explosion. The hypothetical attack that Lieber and Press modeled could have appeared in their computer simulation to destroy a silo and confine the radioactive fallout, and raising the height of the airburst could reduce the fallout further. But the relationship among accuracy, yield, and probability of destruction cannot simply be extrapolated to an arbitrarily high accuracy and an arbitrarily low yield.

This is because destroying a buried missile requires a certain minimum total ground shock. Lieber and Press' model assumes that 3,000 pounds per square inch of overpressure is the level at which a silo cannot survive. In fact, a lower-yield explosion would require higher peak pressure than a higher-yield explosion to damage or destroy a silo, because as the yield of a weapon increases, so does the duration of the blast wave; short blast waves need more overpressure to demolish a silo. The low-yield explosions of the attack presented by Lieber and Press, in other words, would not be effective enough.

Another feature that makes the authors' hypothetical attack less effective is that it relies on airbursts, rather than groundbursts. This is because the main mechanism for damaging missile silos is ground shock waves. Air explosions transfer their energy to ground shock waves much less efficiently than ground explosions because of the large difference in the densities of the air and the earth. Direct ground shock waves are only produced by detonations on or very close to the ground. In addition, only groundbursts form craters and kick up walls of debris that would help destroy what are likely the most hardened strategic targets in China.

The proposed attack calls for the accuracy of cruise missiles but the promptness and penetration of ballistic missiles. A very high confidence of success would require that multiple warheads be targeted against each missile silo, and guided bombs and cruise missile warheads are the only nuclear weapons capable of achieving the required level of accuracy. Given that, it is unlikely that a strike on 20 Chinese silos could be achieved with near-certain success, as Lieber and Press claim.

The United States has only 20 B-2 bombers, which are normally based in the continental United States. Destroying the 20 Chinese silos -- taking into account defensive actions by China, U.S. operational failures, and the roughly 600 miles the aircraft would have to travel over Chinese territory -- would require several attack waves in an assault that would last for hours and could not escape notice.

Even if the U.S. attack were to get through, several of the silos would likely be empty, their missiles having been launched against the United States before the attack arrived. This dilemma is aggravated by China's deployment of dozens of mobile long-range missiles that, although highly vulnerable to conventional attacks if they can be found, nonetheless complicate U.S. planning against the silo-based missiles.

The imponderables in calculating the exact probability of destroying all the targets are no small matter, given the stakes. The range, yield, number, and accuracy of China's intercontinental ballistic missiles suggest that their only purpose is to threaten U.S. and Russian cities and serve as the core of China's deterrent. The massive destruction of U.S. cities that could be inflicted by even a few surviving Chinese missiles makes it doubtful that any U.S. war planner would be content with the extremely narrow margin of success in Lieber and Press' low-yield scenario. The high yield of the warheads on U.S. ballistic missiles is a consequence not simply of a lack of pinpoint accuracy but also of a desire to be absolutely certain that an enemy's silos are destroyed.

Lieber and Press warn that as the Obama administration reduces the U.S. nuclear arsenal, it should retain the ability to launch "precise, very low-casualty nuclear counterforce strikes." This reflects a poor understanding of the current U.S. nuclear posture. The administration is not considering transitioning to an arsenal "comprised solely of high-yield weapons." And the destruction of Tehran or Pyongyang has never been and will never be the only U.S. option for retaliation. The United States currently deploys approximately 500 nuclear cruise missiles and nuclear bombs on long-range bomber aircraft (with more in reserve) capable of explosive yields between 0.3 and 1,200 kilotons.

The current arms reduction negotiations between Moscow and Washington, aimed at creating a successor to the 1991 Strategic Arms Reduction Treaty and the 2002 Moscow Treaty, are centered on long-range ballistic missile weapons systems capable of rapid global counterforce and population targeting, artifacts of the Cold War postures. Mutual nuclear annihilation executed in under an hour is still held at bay by complex early warning and communications systems and political-military chains of command.

The posture presented by Lieber and Press risks re-creating this classic Cold War security dilemma on a smaller scale. They contend that their scenario would be a deterrence mission to be carried out during a conventional war in response to an enemy acting to "introduce nuclear weapons -- that is, threaten to use them, put them on alert, test them, or even use them." But an adversary such as China would find this indistinguishable from a first-strike posture.

"The Nukes We Need" is an unfortunate distraction on the path to understanding how a safer, more stable deterrent relationship with China and Russia can and should be achieved. The challenge today is not to make the use of nuclear weapons more credible. It is to reduce their salience.

HANS M. KRISTENSEN is Director of the Nuclear Information Project at the Federation of American Scientists. MATTHEW MCKINZIE is Senior Scientist at the Natural Resources Defense Council. IVAN OELRICH is Acting President of the Federation of American Scientists.


Nuclear weapons are a boon for vulnerable states. During the Cold War, the United States deployed them in Europe to defend NATO because Soviet conventional forces seemed overwhelming. Now, the tables are turned: the United States' potential adversaries see nuclear weapons as a vital tool to counter U.S. conventional military superiority. Facing defeat on the battlefield, adversaries would have powerful incentives to use nuclear forces coercively, just as NATO planned to do during the Cold War. The fates of Manuel Noriega, Slobodan Milosevic, Radovan Karadzic, and Saddam Hussein have taught a grim lesson: use every weapon at your disposal to prevent defeat.

When Jan Lodal and James Acton call for the elimination or devaluation of nuclear weapons, they assume that U.S. adversaries can be convinced to accept perpetual vulnerability. The Soviet Union could not talk NATO into surrendering its nuclear arsenal during the Cold War, nor can the United States dupe its adversaries into disarming today. The challenge is to grapple with the problem of deterring nuclear escalation during conventional wars, when U.S. adversaries will have every incentive to use their nuclear arsenals to compel a cease-fire. Toward this end, Washington must retain a range of counterforce capabilities, including conventional and low-casualty nuclear weapons.

Hans Kristensen, Matthew McKinzie, and Ivan Oelrich raise several technical objections concerning the United States' ability to launch a successful counterforce strike. They dispute whether 3,000 pounds per square inch (PSI) of overpressure produced by low-yield airbursts would be enough to wreck Chinese silos. The use of 3,000 PSI in our model, however, is conservative. Many analysts believe that U.S. Cold War estimates exaggerated the hardness of enemy silos, and analysts with considerable technical expertise on this matter believe that our estimated requirement of 3,000 PSI probably overstates the hardness of China's silos. Most important, our results are not sensitive to moderate changes in assumptions about silo hardness. The United States could conduct a low-casualty nuclear strike -- producing fewer than 1,000 fatalities -- against all 20 Chinese silos even if they were built to withstand 5,000 PSI.

Kristensen, McKinzie, and Oelrich also contend that airbursts alone cannot destroy missile silos. This is incorrect. Airbursts can produce sufficient overpressure to crush the caps that protect missiles in the ground. In fact, the Pentagon assigns "vulnerability numbers" to silos on the basis of their resistance to overpressure. And McKinzie co-authored a 2001 Natural Resources Defense Council report that contradicts the claims that he, Kristensen, and Oelrich make here. The report listed the overpressures required to destroy various Russian missile silos, and it argued that even Russia's silos -- which are probably much more robust than China's -- are highly vulnerable to a U.S. airburst attack.

Our critics further suggest that the existence of mobile missiles obviates our analysis. If the launchers can be located, the argument goes, conventional weapons are sufficient to destroy them; if the launchers cannot be found, even nuclear weapons are useless. But the greatest challenge of targeting mobile missiles is not locating them momentarily; it is continuously tracking them and identifying where they have stopped. Hitting mobile launchers with conventional weapons requires near-perfect real-time intelligence -- locating them within a few dozen yards. Even low-yield nuclear warheads would significantly reduce the targeting problem; locating the launchers within about half a mile would suffice if a five-kiloton warhead were used.

Kristensen, McKinzie, and Oelrich also note that the U.S. military's current delivery systems are not optimized for a counterforce mission: the most accurate systems (bombs and cruise missiles) are not prompt, and the most prompt systems (ballistic missiles) are not the most accurate. This is true. But current U.S. delivery systems are adequate given the weakness of the adversaries the United States now faces. If Washington wishes to retain effective low-casualty counterforce options, the next generation of nuclear delivery systems should further combine prompt delivery with high accuracy.

Lodal tries to link our discussion of counterforce options with the views held by senior officials in the George W. Bush administration. The fact of the matter is that nuclear counterforce options have been a core element of U.S. deterrence doctrine during every administration since Harry Truman's. U.S. strategic planners have understood that for deterrence to be credible, the president needs retaliatory options that he might actually use. Especially today, low-yield nuclear counterforce strikes are a better retaliatory option than high-yield nuclear strikes that, regardless of their target, would kill millions of civilians. The latter would be a disproportionate response to many possible enemy uses of nuclear weapons.

Critics of our policy prescriptions must confront two core issues. First, nuclear weapons have fundamentally changed since the Cold War. They once produced stalemate, and nuclear war once meant mass slaughter. For good or ill, that has changed. The revolution in accuracy means that enemy arsenals can be destroyed, and in ways that produce few civilian casualties. Theories of deterrence and beliefs about strategic stability and nuclear force requirements must be reevaluated accordingly.

Second, any proposals regarding the future of the U.S. nuclear arsenal must explicitly address four key questions: What enemy actions are to be deterred with U.S. nuclear forces? Under what circumstances might those actions be taken? What threats might a president wish to issue to deter those actions? What capabilities would give the United States the ability to carry out those threats? Working through these fundamental questions is the only basis of a responsible deterrence strategy.