Following are excerpts from the report just issued by the Union of Concerned Scientists on “Space-Based Missile Defense”—often referred to as “Star Wars.” The report is of urgent interest in view of the announced plans of the Reagan administration to develop a spacebased missile defense during the coming years. The authors of the report are listed in the box below.


I call upon the scientific community who gave us nuclear weapons to turn their great talents to the cause of mankind and world peace: to give us the means of rendering these nuclear weapons impotent and obsolete.

—President Reagan, March 23, 1983

These words unveiled the President’s Strategic Defense Initiative, a “comprehensive and intensive effort” with the “ultimate goal of eliminating the threat posed by strategic nuclear missiles.” It proposes to rely on unborn generations of sophisticated space weapons that the Secretary of Defense told Meet the Press would provide a “thoroughly reliable and total” defense. We shall adopt Mr. Weinberger’s words, and refer to the President’s goal as total ballistic missile defense, or total BMD—what in the vernacular is now called “Star Wars.”

Every sane person yearns to escape from the specter of nuclear annihilation. But that consensus still leaves a host of unanswered questions: will these BMD systems, which still are just conceptual designs, provide a total defense of our civilization against the Soviet missile force? That force now carries 9,000 nuclear warheads, each far more powerful than the Hiroshima bomb, and able to arrive on US targets within thirty minutes. (The US arsenal is, of course, equally devastating.) If these defenses of the distant future could protect us totally against today’s threat, could they cope with the Soviet strategic weapons of their own era?

What would the Soviets’ response be? Would they devote themselves to a similar effort, and agree to reduce their offensive nuclear forces? Or would they perceive this new American program as an attempt to nullify Soviet nuclear forces—as a supplement to the emerging US capacity to destroy Soviet missiles in their silos? If so, would they not respond with a missile buildup and “counter-measures” to confound our defenses, so that they could still destroy the United States (just as the US can destroy the USSR)? Or would the Soviets not have this option, because our defense would be truly total—robust enough to foil any offensive countermove?

This report addresses these questions. It examines the proposed BMD systems in the light of the scientific facts and principles that will govern their performance, and with the aid of the information released to the press by the Administration’s Defense Technologies Study Team headed by Dr. James C. Fletcher. There is general agreement that a defense of our population is impossible unless the vast majority of Soviet missiles can be intercepted in the first phase of their flight, while their booster engines emit a brilliant flame and before their multiple warheads are released. Otherwise, the subsequent layers of the BMD system will not be able to cope with the attack. We therefore devote the bulk of our attention to “boost phase” defense.

All boost phase interception must be carried out at long distance. Hence it is essential to transmit a blow to the enemy booster with a projectile that can travel quickly. The highest velocity attainable is the speed of light (186,000 miles per second). For that reason, laser beams which move at that speed, and beams of atoms or electrons which are nearly that fast, would be ideal if they could be made intense enough to cause damage at such large distances. Such devices are called directed energy weapons.

A laser is a device that emits a beam of light composed of rays that are almost perfectly parallel. We shall consider several types of lasers: chemical lasers that emit infrared light, excimer lasers that emit ultraviolet light, and a laser that is pumped by a nuclear explosion and emits X-rays.

In a weapon, the beam from an infrared or ultraviolet laser is concentrated on the target by adopting the familiar trick of lighting a fire with a magnifying glass that focuses the sun’s rays. In a space weapon, the task of focusing and aiming the laser beam is carried out by a suitably oriented and shaped mirror or system of mirrors. The laser itself could be in space or on the ground, but the mirror must be in space if it is to send the beam toward the booster.1

In assessing each BMD system, we first assume that it will perform as well as the constraints imposed by scientific law permit—that targets can be found instantly and aiming is perfect, that the battle management software is never in error, that all mirrors are optically perfect, that lasers with the required power output will become available, etc. Above all, we assume that the Soviets’ forces remain static—that they do not build more missiles or install any countermeasures. Hence, our initial optimistic appraisal ignores the critical question of whether BMD will eventually work as well as it possibly could, and does not depend on classified information.


Even in this utopian regime, our findings concerning, the proposed BMD schemes are that:

—Chemical laser “battle stations” in low orbits, or “space trucks” carrying “kill vehicles,” will have to number in the thousands to give adequate coverage of the Soviet silo fields; simply lofting these stations into orbit would cost upward of $70 billion.

—Excimer lasers on the ground, whose beams would be reflected toward boosters by over a thousand orbiting mirrors, would require power plants which alone would cost some $40 billion.

—The atmosphere and the earth’s magnetic field combine to make particle beam weapons wholly implausible into the foreseeable future.

These cost estimates do not include research and development, or construction of space platforms, lasers, kill vehicles, mirrors, and command and control facilities. Just the R&D portion of this program has been described by Dr. Richard DeLauer, Under Secretary of Defense for Research and Engineering, as having at least eight components “every single one…equivalent to or greater than the Manhattan Project.” Furthermore, all costs will climb rapidly should the mirrors be imperfect, the time for aiming exceed several seconds, redundancy be desired, etc. The full costs cannot even be estimated because the proposed technologies are still too immature, but it is clear that many hundreds of billions of dollars would be needed.

The proposal to launch X-ray lasers pumped by nuclear explosions at the time of an attack would require a new fleet of submarines, since there is no suitable base on land close enough to Soviet silos to allow interception in the time available. The laser’s soft X-rays cannot penetrate the atmosphere, and they deliver a rather light blow from which the booster can readily be protected. These facts, when combined with the feasibility of shortening the boost so that it ends before the missile leaves the atmosphere, imply that the X-ray laser is not a viable BMD weapon.

These findings assume a minimal Soviet reaction to a US missile defense. But the Soviets have made it clear that they view the quest for a total BMD as an unacceptable threat. They fear that such a BMD system would give us the option to strike first—an understandable fear since Mr. Weinberger has said that he would view a similar Soviet system as “one of the most frightening prospects” imaginable. And they have heard Administration officials speak of space-based BMD as a lever for stressing the USSR’s technologically less sophisticated economy.

In the real world we must therefore expect a determined Soviet reaction, unconstrained by all existing agreements, because the very testing of our defensive weapons would violate our obligations under the ratified Anti-Ballistic Missile (ABM) Treaty. The Soviet reaction is likely to include:

—Offensive missiles designed to circumvent BMD, such as submarine-launched cruise missiles that cannot be intercepted from space.

—Fitting ICBMss with more powerful engines so that the boosters would burn out quickly and inside the atmosphere, which would stress any BMD system, and eliminate interception by kill vehicles and X-ray lasers.

—Cheap decoy ICBMs—boosters without warheads in fake silos—to overwhelm boost phase interceptors.

—Weapons that would exploit the fact that even a battleship’s armor could not protect a space station from quite primitive types of attack.

—A Pandora’s box of largely developed countermeasures that would vastly complicate the problem of targeting boosters and warheads.

All these countermeasures would exploit off-the-shelf weapons and techniques that exist today, in contrast to the unproven and improbable technologies on which our proposed defenses would rely. Hence, the Soviet response will be cheaper and far more reliable than our defenses, and available as those defenses emerge.

While this quest for a total defense against nuclear missiles would be endless, the decision to embark would have immediate political repercussions. Indeed, the first repercussion has already been heard: the US rebuff to Soviet overtures to negotiate constraints on anti-satellite (ASAT) weapons—a stance dictated by the plans for a space-based BMD system. The ABM Treaty could not survive the start of this endless journey, and with it all constraints on offensive forces would go overboard. The impact on NATO would be profound. Our allies in Europe would not be protected by an American BMD system, and this would inflame existing suspicion that the US intends to conduct nuclear operations in Europe without risk to itself. Alliance cohesion would erode because Europeans would hold the US responsible for exacerbating East-West tensions.

The risk to our survival would mount dramatically were we ever to begin erecting the BMD system. This budding system would be exceedingly vulnerable to attack. Nevertheless, its capabilities would be overvalued by our adversaries, and its installation could well be perceived as an attempt to disarm the Soviet Union. These circumstances could in themselves provoke open conflict.


If we get through this hazardous passage, will we have reached the promised land where nuclear weapons are “impotent and obsolete”? Obviously not. We would then have a defense of stupefying complexity, under the total control of a computer program whose proportions defy description, and whose performance will remain a deep mystery until the tragic moment when it would be called into action.

The President and his entourage occasionally argue that we must pursue this quest because the benefits of success outweigh the costs and risks. However, that is only an argument for a research program in strict conformity with the ABM Treaty. Such a program has always had our support. It is needed to protect us from Soviet surprises, and it might uncover concepts that could actually provide a viable defense. But there is an enormous gulf between such a program and a call from the ramparts for a national “experiment” to mount a defense based on untried technologies and provocative doctrines. We have delineated the costs and risks of such an “experiment.” At best, the outcome would be a defense of precarious reliability, confronted by offensive nuclear forces designed to circumvent and overwhelm it, and a host of new “anti-BMD” weapons to attack our armada of space platforms which, in turn, would have to be defended by yet another fleet of anti-anti-BMD weapons.

It is difficult to imagine a more hazardous confrontation. And it is equally difficult to understand how anyone can believe that this is the path toward a less dangerous world. A direct and safe road is there for all to see—equitable and verifiable deep cuts in strategic offensive forces and immediate negotiations to ban all space weapons. If we are to take that road, we must abandon the misconception that nuclear explosives are military weapons, and the illusion that ever more sophisticated technology can, by itself, remove the perils that science and technology have created. We must, instead, recognize the overriding reality of the nuclear age—that we cannot regain safety by cleverly sawing off the thin, dry branch on which the Soviets are perched, for we cling to the same branch.

The Problem of Cruise Missiles

A cruise missile is, in essence, a small ground-hugging pilotless airplane that can carry a nuclear warhead over distances of thousands of miles. It measures the altitude of the overflown terrain with an on-board radar and matches that altitude against a map stored in its computer’s memory. An accuracy sufficient to threaten hard targets is therefore attainable, though the time from launch to impact is much greater than it is for ballistic missiles. Cruise missiles capable of penetrating into the Soviet Union are already on our B-52 strategic bombers and are being deployed on the ground by NATO. Both superpowers could develop cruise missiles for submarines that could strike both civilian and strategic targets far inside their adversary’s borders from an unpredictable launch point.

None of the space-based defense systems that are under discussion can touch cruise missiles. That is not their purpose. But until a virtually perfect shield against cruise missiles is developed, there is no such thing as a total missile defense. It would take us too far afield to examine the prospects for defense against cruise missiles. But a few remarks will suffice. No one doubts that a significant number of US strategic bombers (not just their cruise missiles) could penetrate the highly touted Soviet air defense system to deliver their high-yield bombs on target; the experiences of two Korean airliners have shown that this is the case. Cruise missiles are far harder to detect with radar than much larger and higher-flying airplanes, and as the so-called STEALTH techniques develop, cruise missiles will become even more elusive. Since they are unmanned, a high attrition rate is quite acceptable. Given these facts, it is very difficult to envisage a shield against air-and sea-launched cruise missiles that would protect our population.

Systematic Problems and Vulnerabilities

Defense Suppression

One of the most effective tactics that can be employed against a Ballistic Missile Defense is to attack the ground and space assets on which it depends. Some of these assets are of surprising vulnerability while at the same time being crucial to the defense. Nuclear explosions in space…can blind infrared sensors and blackout radars. Detonated ahead of the flight of reentry vehicles, at altitudes of 60–80 km, a single precursor burst will ionize a region of space some tens of kilometers across, hiding the vehicles for several minutes. Anti-radiation homing vehicles may be used to destroy radars in space and on the ground. Space mines or inert objects, including sand, may be used against fragile space-based lasers and mirrors. A variety of measures can be taken to jam, spoof, and confuse the data transfer links of the battle management system.

Submarine-launched ballistic missile and cruise missile attacks can be a key factor in defense suppression in consequence of the immense damage that they would do to their targets. Submarine-launched missiles have short flight times, allowing as little as 3–5 minutes’ warning for near-coastal targets. They have unpredictable launch points, which can make viewing angles poor for the defense. Moreover they can be launched on depressed trajectories, low enough to allow the missiles to evade most defenses. Cruise missiles are air-breathing, lowflying, and nearly invisible to optical and radar trackers. The BMD system cannot defend itself against such attacks. In the future, if an extensive missile defense moves toward deployment, increasing attention would have to be paid to submarine missile attacks.

Some improvements in sea-launched ballistic missile defense can be expected, but it is difficult to see what can be done to mitigate substantially the cruise missile threat. Targets would include ground facilities for battle management, rockets and basing facilities associated with pop-up sensors and weaponry, and communications and control stations. Successful pop-up launches from silos inside the United States which had survived the disarming strikes could be forestalled by nearby exo-atmospheric explosions, or by bursts in the fringes of the atmosphere to destroy missiles in flight. Well-executed strikes of this sort, in advance of the main offensive missile launch, would in all likelihood disable the entire defense structure. There appears to be no way that this vulnerability can be adequately reduced.

Software and Algorithms

The battle management systems of a total BMD must deal with thousands to hundreds of thousands of objects. This requires computers with the capacity to carry out many hundreds of millions, if not billions, of arithmetic operations per second. Advances in computer technology suggest that the hardware to accomplish this monumental task may become available in the future. There are, however, several challenges whose solutions are doubtful. One is the problem of designing and writing the programs (software) required to guide the computers. Experience with earlier defense system software, as well as examples from non-defense experience, suggests that it will be exceedingly difficult, if not impossible, to construct software that would operate properly in the environment of a nuclear attack, for which it could never have been fully tested.

A related problem is that of developing algorithms for tracking and, especially, for discrimination and weapons assignment. Algorithms are the rules or criteria by which the sensor data on range, velocity, maneuvering, and other target properties are assessed, weapons commitments made, damage evaluated, and weapons reassigned and committed. There would be one or more algorithms, for example, to decide which objects were decoys and which were lethal objects. These algorithms must be prepared long in advance of conflict and embedded in software and, to some extent, hardware. They are critical to the performance of the defense. Not only must these algorithms cope with the opponent’s counter-measures, whose nature and effectiveness could only be guessed at in advance, but they must be free of internal flaws.

As with the software, it would be extremely difficult to be confident that a fatal flaw was not embedded somewhere in these criteria which might, for example, guide far too many weapons toward unusual or unexpectedly outfitted decoys or away from warheads. They must not…direct interceptors through the empty centers of composite targets. The inability to fully test a ballistic missile defense system before it is employed makes the prospect of errors an unsettling one. It is with respect to the battle management system—the hardware, software, and the algorithms—that this deficiency can have some of its gravest consequences.


Suppose for a moment that a total missile defense system without obvious flaws had been developed and deployed. This defense, in a time of confrontation and nuclear attack, would represent the only prospect of avoiding overwhelming ruin. Could there be enough confidence in this defense that the US could safely reduce its nuclear forces unilaterally or that it could, during a crisis, ignore Soviet nuclear threats with impunity? The confidence that is needed is not just that each component of the myriad array of parts—the sensors, the weapons, the computers—would, individually, perform as expected. More important is the need for confidence that the entire assembly would operate as a harmonious machine and capably blunt the attack. Consider the nature of the defensive system: an enormous, intricate, and complex assemblage, novel in design, pushing the limits of technology, intended to provide a defense against a threat and, under circumstances which will be fully defined only when the attack comes, forced to meet an uncommonly high standard of performance. And, remarkably enough, it cannot ever have been adequately tested.

No amount of testing under simulated battle conditions could confidently explore the response of the defensive system to an actual nuclear attack. This is only in part because the nature of the attack and of the attacker’s countermeasures cannot be known in advance. One simply cannot simulate the stress and the demands on the system of the circumstances of war.

The matter of testing is crucial. The performance of complex devices can rarely be confidently predicted before they are set in realistic operation even when their tasks are well defined. Complex designs breed complex problems. And this is also the case with computer software. All large programs contain “bugs,” hidden flaws, and while their number dwindles over time as the programs are used, no one can ever be certain that the bugs are gone. The testing establishes a widening range of confident operation, no more. New circumstances can bring new flaws to light. There is no known way to get around this. We conclude that even if an apparently effective total defense could be prepared, it is highly unlikely much trust could be placed in its working properly when needed.

Net Assessment of Total Missile Defense

In this section, we pull together our technical assessments of total ballistic missile defense, taking into account the interdependence of the various elements and the fact that the vulnerabilities of one portion of the system critically affect the performance of others.


The requirement that a total BMD operate in large part in space creates an insoluble basing dilemma. On the one hand, orbiting stations, whether for weapons, sensors, or mirrors, are inherently fragile and vulnerable to attack by space mines and other anti-satellite (ASAT) techniques. This vulnerability would make them provocative targets during a developing crisis, and thus might actually help precipitate the outbreak of conflict. On the other hand, the alternatives to orbital basing—airborne, ground-based, or pop-up systems—are inferior in important respects in their ability to perform BMD functions in boost phase and midcourse. There has been no satisfactory scheme put forward for using the X-ray laser, the only directed-energy weapon that could in theory be popped-up, as a component of a BMD system. X-rays cannot penetrate the atmosphere and it is perfectly practical [for the Soviet Union] to reduce boost phase so burnout occurs inside the atmosphere. Furthermore, the impact of the X-ray beam is too weak to damage warheads following boost phase. The ground-based excimer laser, another prominent boost phase candidate, relies on orbiting mirrors that would be extremely vulnerable.

Sensors and Battle Management

Both active and passive tracking and discrimination systems and the associated battle management facilities can be seriously compromised by countermeasures and by the very large number of largely indistinguishable objects comprising the threat cloud. This is likely to be an insuperable problem in midcourse. It will be necessary to forego any attempt at tracking each individual object, making individual weapons assignments impossible. Having given up the unique assignment of interceptors to targets, it is necessary to allow the interceptors to home at random on all the objects in the threat cloud. But this is highly inefficient and would require on the order of one million homing vehicles for the expected Soviet threat.


Particle beam weapons are the least promising of the potentially available types of weapons. The prospects of developing them to a satisfactory level of performance seem dim at best. Laser prospects, although better than those of the particle beams weapons, are very far from bright. The improvements that are required are of such magnitude that important technical breakthroughs appear to be needed. While we cannot categorically rule out the possibility of ultimate success, we do not expect it. All of the particle beam weapons, and most of the lasers, must be based in space with all the vulnerability that implies. The one exception, the excimer laser, is still not much more than a laboratory curiosity and no one can be sure what its future will be. Any ground-based laser system must use a large number of mirrors in space which are extremely vulnerable. In any case, the threat of countermeasures, such as atmospheric nuclear bursts in the beam path or attacks on the mirrors, will disable such weapons. At the present time, self-propelled homing weapons are the most fully developed potential BMD weapon. The anti-satellite Miniature Homing Vehicle is approaching deployable status. Technical improvements in speed, guidance accuracy, and maneuverability appear feasible. Whether a satisfactory defensive system could be built around them is highly doubtful, however.

Ground Assets

In the above assessments we did not factor in the consequences of defense suppression attacks against the ground assets, such as satellite communication stations, radars, and pop-up and ground-based weapons. The operability of the entire defense is totally dependent on their survival. The minute leakage [of reentry vehicles, i.e., warheads] that can be tolerated necessitates that the ground assets survive intact. Because the ground installations cannot be confidently defended against sea-launched ballistic missiles, and probably not against cruise missiles, we conclude that this vulnerability represents a major Achilles’ heel in a prospective defense.


Our analysis makes clear that total ballistic missile defense—the protection of American society against the full weight of a Soviet nuclear attack—is unattainable if the Soviet Union exploits the many vulnerabilities intrinsic to all the schemes that have been proposed thus far. In none of the three phases of attack can one reasonably expect the success rates that would allow a layered BMD system to reduce the number of warheads arriving on US territory sufficiently to prevent unprecedented death and destruction. Instead, each phase presents intractable problems, and the resulting failure of the system compounds from one phase to the next.

A highly efficient boost phase intercept is a prerequisite of total BMD, but is doomed by the inherent limitations of the weapons, insoluble basing dilemmas, and an array of offensive countermeasures. As a result, the failure of midcourse systems is preordained. Midcourse BMD is plagued not so much by the laws of physics and geometry as by the sheer unmanageability of its task in the absence of a ruthless thinning out of the attack in boost phase.

Terminal phase BMD, finally, remains fundamentally unsuitable for area defense of population centers, as opposed to hard-point targets. There seems no way of defending soft targets on a continent-wide basis against the broad variety of attacks that could be tailored to circumvent and overwhelm terminal defenses.

Political and Strategic Implications

The political and strategic dangers raised by the “Star Wars” initiative are at least as important as its technical flaws. Indeed, these dangers would weigh heavily against development of ballistic missile defenses even if the technical prospects for such systems were much brighter than they are. A US commitment to BMD would precipitate Soviet responses and a chain of actions and reactions that would radically change the strategic environment to the detriment of both countries’ security. The offensive arms race would be greatly accelerated, arms control treaties undermined, and the nuclear peace made more precarious.

Of course, the technical and political issues are not completely unconnected. If it were possible to put in place overnight a fully effective, invulnerable defense against nuclear weapons, there could hardly be serious objections to doing so. But, as the preceding analysis has shown, such a system cannot be built now or, in all likelihood, ever. In the real world, BMD systems will be imperfect. Even under very optimistic assumptions about their ultimate performance, the process of improvement would be incremental and prolonged. During this extended and highly unstable transition period the strategic and political implications of BMD become critical.

While the alleged benefits of BMD are distant and hypothetical, the dangers are near-term and predictable. The adverse consequences of a commitment to BMD would be felt long before the actual deployment of mature technological systems, and quite likely even while the ABM Treaty was still technically being observed. These consequences would follow the familiar anticipated reactions syndrome, driven by the highly threatening nature of BMD and the worst-case assumptions that would dominate nuclear planning amid large uncertainties about the effectiveness of BMD systems and ambiguities about the intentions behind them. Accordingly, the dangers posed by a US policy of ballistic missile defense would be virtually independent of the level of performance that BMD systems might, decades in the future, finally achieve.

Consequences for the Arms Race and Arms Control

A collapse of the ABM Treaty and the initiation of a BMD competition between the superpowers would have a devastating impact on the prospects for offensive arms control. Following an inevitable action-reaction pattern, the Soviets are certain to respond to an American BMD with new offensive measures. Both a quantitative and a qualitative escalation of the arms race would ensue. Adherence to the terms of the SALT II treaty—not ratified by the US but until now informally observed by both countries—would end, and hopes for new agreements would be undermined.

The enormously threatening character of the policy goal announced by President Reagan in his March 1983 speech guarantees a strong Soviet reaction. Despite presidential rhetoric that defenses might be in the mutual interest of the superpowers, the Soviet Union will certainly view a serious US commitment to BMD as an attempt to achieve military superiority by negating the Soviet deterrent. After all, it is Soviet weapons that would be rendered “impotent and obsolete” by an American BMD breakthrough.

The Soviets will interpret a US BMD program not in relation to some utopian future, but in the context of the ongoing US nuclear buildup, particularly the conversion of virtually all US strategic forces to a counterforce role, and the war-fighting doctrine that this buildup is meant to implement. From this standpoint, one can readily appreciate that the Soviet Union might view an American BMD as part of a larger US effort to acquire a first-strike capability—the ability to carry out an attack against Soviet nuclear forces and to defend effectively against a heavily degraded Soviet retaliatory strike. The Soviet Union is no more likely than the US to accept such a development.

In a world of BMD deployments, each superpower’s first priority would be the maintenance of forces able to penetrate or circumvent the other’s defenses. The resulting stimulus to the arms race would be aggravated by uncertainties about the effectiveness of defenses. Operating as usual on conservative, worst-case planning assumptions, each side would tend to exaggerate the effectiveness of the other’s defense while discounting its own. As a result, offensive responses would tend to surpass the level actually needed to maintain second-strike or retaliatory capabilities. For example, US defenses assessed by American defense planners as 50 percent effective might elicit a Soviet build-up based on the assumption of 90 percent effectiveness, and vice versa. Each side would then perceive the other’s reaction as excessive and threatening, and would respond in kind, creating a vicious cycle of escalation.

In these circumstances, the hope that BMD might improve the prospects for negotiated force reductions, as suggested by the Reagan Administration, is totally unrealistic. Even less plausible is the idea that a US BMD could be used as a lever, in the words of Dr. George Keyworth II, the President’s Science Advisor, to “pressure the Soviets to take our arms reductions proposals more seriously than they do now.”2

Instead, BMD would be doubly fatal to the prospects for controlling offensive forces—raising fear and suspicion that would poison the political atmosphere for negotiations, and creating technical problems of comparison and verification far more complex than those associated with SALT or the Strategic Arms Reduction Talks (START). In effect, defenses function as a wild card, making the nuclear balance much less calculable than if only offensive forces need to be taken into account. Given the large uncertainties and controversy that already surround efforts to compare US and Soviet nuclear capabilities, the addition of BMD to the mix would place an unbearable strain on arms control. An arms control process healthy enough to support reductions in an environment of strategic defenses would be more than strong enough to produce such reductions in the absence of BMD systems. Conversely, an arms control process that is already faltering badly would be unlikely to survive at all in the face of BMD deployments by the two sides.

Ironically, the administration’s own “build-down” proposal in the START negotiations would be directly undermined by a US BMD initiative. This approach emphasizes cuts in the heavy, multiple-warhead Soviet ICBMs that threaten American land-based missiles, and seeks an overall restructuring of strategic forces away from MIRV missiles toward smaller, single-warhead ones. However, missile defenses would place a premium on large ICBMs, which offer the most efficient means of delivering large numbers of warheads, decoys, and other penetration aids to overwhelm a BMD system. As noted earlier, the US first developed MIRV, largely in response to the Soviet ABM program in the 1960s. More recently, the Scowcroft Commission on US Strategic Forces cited BMD penetration as an important rationale for US deployment of the 10-warhead MX missile:

The possibility of either a sudden breakthrough in ABM technology, a rapid Soviet breakout from the ABM treaty by a quick further deployment of their current ABM systems, or the deloyment of air defense systems having some capability against strategic ballistic missiles all point to the need for us to be able to penetrate some level of ABM defense. This dictates continued attention to having sufficient throwweight for adequate numbers of warheads and of decoys and other penetration aids.3

Administration officials have some-times suggested that a US-Soviet BMD competition would play to US technological strengths. Whether or not this is true, it is clear that a BMD-driven offensive arms race would give the Soviet Union important advantages. Due to its large advantage in missile throwweight, as well as the absence of political constraints comparable to those in the US, the Soviet Union is much better positioned than the US for a rapid offensive buildup. In particular, if SALT II becomes a dead letter, the Soviet Union has an option for a sudden breakout from the SALT limits on MITVs that the US does not enjoy. The Soviets could, for example, double their ICBMs warheads simply by additional MIRVing of their 308 S-18 missiles, which are limited under SALT II to ten warheads each but could accommodate up to thirty.

In addition to offensive efforts to overwhelm US missile defenses, the Soviets can also be expected to invest in delivery systems that would circumvent those defenses. Cruise missiles and depressedtrajectory ballistic missiles have already been alluded to in this report. Unconventional approaches, including “predelivery” of atomic weapons by emplacing them clandestinely on US territory, can also be envisaged. In many cases these responses would not only foil US missile defenses but would result in reduced warning times and verifiability, thereby further undercutting nuclear stability and arms control.

Superpower BMD deployments would also have an adverse impact on the prospects for theater nuclear arms control involving British and French forces. Both countries, as well as China, would feel threatened by a Soviet missile defense that undercut their ability to penetrate Soviet territory, and as far as possible would modernize and increase their forces to maintain the standard of minimum deterrence on which their nuclear strategies rest. For example, among the conditions set by France for participation in future “Euromissile” negotiations is that “defensive system…must remain limited.”4

More generally, a US BMD would cause major political and strategic strains in the Atlantic Alliance, inspiring European fears of a US retreat into “Fortress America.” Existing perceptions that the US is complacent about nuclear risks, and willing to contemplate a “limited” nuclear war confined to European territory, would be likely to grow.

Consequences for Deterrence and Crisis Stability

In addition to the arms race consequences discussed above, the administration’s BMD proposal would have a profoundly destabilizing effect on the nuclear balance, increasing the risk of nuclear war at times of US-Soviet confrontation and reducing the chances of bringing hostilities under control if war did occur.

These consequences run directly counter to the arguments often made by BMD advocates that US defenses would strengthen deterrence and, in the event deterrence fails, play an important damage-limitation role. Such arguments, it should be emphasized, are attempts to construct strategic rationales for only modestly capable BMD systems. As such, they represent a very large retreat from President Reagan’s vision of transcending (not reinforcing) the system of nuclear deterrence by making nuclear weapons “impotent and obsolete.” Nevertheless, these justifications for imperfect BMD systems are important to address. As the President’s original vision is increasingly understood to be illusory, a US BMD program is likely to be promoted primarily on grounds of deterrence and damage-limitation. Administration officials and supporters have already begun to argue in these terms during the year since President Reagan’s speech. In this regard, a close link has often been noted between these more modest roles for BMD systems and the administration’s nuclear war-fighting strategy—a link that will not be overlooked by the Soviet Union.

The argument that BMD would strengthen nuclear deterrence rests mainly on the claim that it would reduce the vulnerability of US land-based missiles to preemptive attack. By protecting the US ability to retaliate, BMD would make a Soviet first strike less certain of success and therefore less likely. However, this is an argument for terminal, hard-point defense of US missile silos, not for the layered, area defenses being proposed by the administration. The administration’s initiative is not only vastly more expensive and complex than is necessary for the protection of retaliatory forces, but it is provocative to the Soviet Union in a way that would reduce, not enhance, deterrent stability.

Area defenses undermine deterrence because they magnify the advantage of striking first. Indeed, the modest BMD systems likely to be attainable in the foreseeable future would be useful only to the attacker. They would be easily defeated by a well-executed first strike, but might perform with some effectiveness against a poorly coordinated and weakened retaliatory strike.

As a result, these systems are likely to be perceived as components of a first strike strategy rather than as deterrent weapons, and to create strong incentives for preeemptive attacks during periods of high tension. At such times, the fact that a first strike would be complicated by the adversary’s BMD would be judged less relevant than the fear that, if one fails to attack first, effective retaliation may be impossible.

BMD, then, would aggravate the dangerous “use them or lose them” pressures that are already increasing due to the trend toward offensive weapons designed for counterforce. The result would be a serious weakening of mutual deterrence precisely at those times when it is most needed.

The damage-limitation rationale for BMD is as dubious as the deterrence argument. This rationale has two variants. First, in the event deterrence fails and nuclear war occurs, it is argued that defenses could save lives, reducing the threat of “assured destruction” that exists in the present offense-dominated system. Second, we are told, this damage-limitation effect would strengthen deterrence by making the threat of nuclear retaliation more credible. BMD advocates who emphasize these points generally subscribe to the theory that credible deterrence requires forces designed for actual warfighting, and capable of being used in a selective, flexible manner. In this context, the damage-limitation role of a BMD system is seen as useful not only to limit population fatalities but also to protect nuclear command and control systems.

These arguments are implausible in light of the size and destructive power of superpower nuclear arsenals and the adjustments in targeting and nuclear strategy the BMD deployments would bring about. The overkill capacity of both superpowers is such that only a near-perfect defense could hope to reduce fatalities appreciably in the event of major nuclear exchanges. For example, if the Soviet Union were to target its missiles to maximize damage to the US population—a likely response to a serious American attempt to protect cities—it would need only 5 percent of its current ballistic missile warheads to kill up to half of the US urban population immediately (see illustration on this page). In other words, even a 95 percent effective BMD would leave the US with the prospect of tens of millions of prompt fatalities in a nuclear war, leaving aside all the subsequent deaths from fire, disease, and social disruption. Moreover, enough nuclear explosions would occur even in this very optimistic case to pose a serious danger of triggering a climatic catastrophe (the “nuclear winter” phenomenon).

The vulnerability of the US to destruction by Soviet nuclear forces, in short, cannot be mitigated by any foreseeable defensive shield as long as nuclear weapons exist in their current numbers. Only if offensive forces were radically reduced, to perhaps a tenth of their present size, could a moderately effective defense begin to make a dramatic difference in the vulnerability of populations to nuclear destruction. As we have seen, the prospect of negotiating such reductions would become virtually non-existent amid a major US-Soviet BMD competition.

In the absence of radical cuts in offensive arsenals, damage-limitation could be achieved, in theory, only through deliberate strategies of controlled, limited nuclear strikes, with the bulk of each superpower’s nuclear forces being held in reserve and cities being spared. The nearly unanimous conclusion of those who have studied this issue is that a nuclear war could not in practice be controlled in this manner.5 Furthermore, there is good reason to believe that strategic defenses would make this possibility even more remote.

Contrary to the assumptions of those who view BMD as a useful adjunct to a limited nuclear war strategy, it is likely that BMD deployments would reduce both the incentives and the capabilities of the two superpowers to contain nuclear war below the threshold of all-out exchanges. First, as noted above, to the extent that defenses pose a serious threat to the “assured destruction” capability of either side, they invite retargeting to retain such destructive capacity. The fewer warheads the Soviets can expect to arrive on US territory, the more likely these warheads are to be assigned to the softest and most valuable targets—major urban areas. Warhead accuracy would become less important, and sheer destruction, with maximum collateral damage, more important, thus reversing the priorities associated with limited war strategies. Targetting for strikes against cities would also be encouraged since missile defenses might be able to break up precisely timed attacks against hardened and defended targets.

Second, the space-based command and control systems necessary to limited war strategies would be put at risk if “Star Wars” defenses were deployed. Plans for controlled, protracted nuclear conflict depend critically on survivable satellites for communications, navigation, early warning, and reconnaissance. However, a growing vulnerability of these systems to antisatellite attacks would be an unavoidable side-effect of the development of space-based BMD. In a strategic environment characterized by space-based missile defenses and unrestricted ASAT competition, attacks on military space assets would occur in the early stages of a superpower conflict. This situation would not only exacerbate mutual fears of preemptive attack, but would create incentives to use nuclear forces in massive strikes at the outset of hostilities to take advantage of the capabilities of command and control systems before they are destroyed.

Far from contributing to a strategy of limited nuclear war, BMD points in the opposite direction—toward massive, indiscriminate exchanges and the erosion of control over strategic forces. Instead of damage limitation, a nuclear war fought under these circumstances could well produce higher numbers of fatalities than one fought in the absence of defenses.


The superficial attractions of a strategy of nuclear defense disappear when the overall consequences of BMD deployments are considered. More than any foreseeable offensive arms breakthrough, defenses would radically transform the context of US-Soviet nuclear relations, setting in motion a chain of events and reactions that would leave both superpowers much less secure. Deterrence would be weakened and crisis instability increased. Damage-limitation would be undermined by a greater emphasis on the targetting of cities and the increased vulnerability of command and control systems. And virtually the entire arms control process would be swept away by the abrogation of the ABM Treaty, the launching of a new offensive round of the arms race, and the extension of the arms race into space.

The members of the study panel that issued the report are:

KURT GOTTFRIED, Panel Co-chair; member, Board of Directors, Union of Concerned Scientists; Professor of Physics and Nuclear Studies, Cornell University; former Chairman, Division of Particles and Fields, American Physical Society.

HENRY W. KENDALL, Panel Co-chair; Chairman, Board of Directors, Union of Concerned Scientists; Professor of Physics, Massachusetts Institute of Technology.

HANS A. BETHE, Professor of Physics Emeritus, Cornell University; Nobel Laureate in Physics; Chief, Theoretical Division, Manhattan Project; member, President’s Science Advisory Committee (1957–1960); Strategic Military Panel (1957–1969).

PETER A. CLAUSEN, Senior Arms Analyst, Union of Concerned Scientists; former Fellow, Woodrow Wilson International Center for Scholars; former policy analyst, Department of Energy and Central Intelligence Agency.

RICHARD L. GARWIN, IBM Fellow at the Thomas J. Watson Research Center; member, President’s Science Advisory Committee (1962–1965 and 1969–1972); member, Defense Science Board (1966–1969).

NOEL GAYLER (Admiral, USN, Ret.), former Commander-in-Chief, US Forces in the Pacific; former Director, National Security Agency; former Assistant Chief of Naval Operations (Research and Development).

RICHARD NED LEBOW, Professor of Government and Director, Peace Studies Program, Cornell University; former Professor of Strategy, National War College; Scholar-in-Residence, Central Intelligence Agency.

CARL SAGAN, Professor of Astronomy and Space Sciences and Director, Laboratory for Planetary Studies, Cornell University.

VICTOR WEISSKOP, Institute Professor Emeritus, Massachusetts Institute of Technology; former Director-General, European Organization for Nuclear Research (CERN), Geneva; former President, American Physical Society.

This Issue

April 26, 1984