The past eight years have been a time of enormous flux in those strategic factors that affect stability among the great powers in Europe, Asia and North America. Some of the changes offer the possibility of reduced tensions in East-West relations; others, involving differences among allies or within the U.S. government, are deeply worrisome. This period of change is not over. Within the next year, decisions must be made regarding virtually every important area of American defense policy, from nuclear doctrine to the conventional balance in Europe.

Moreover, the Soviet Union is in the throes of generational change in its leadership. Self-interest alone would have led General Secretary Mikhail Gorbachev to attempt reforms that are essential to any hope of the Soviet Union’s achieving sustainable increases in productivity and output. We cannot know whether he will succeed in carrying out fundamental reform of the party’s control over the economy. But whether or not he succeeds, for the West the question is: What changes, if any, are in store in Soviet foreign policy?

In the short term (five to ten years), we can probably look forward to a period of gradual improvement in U.S.-Soviet relations. Gorbachev needs to focus on the resolution of internal problems, and he wants access to Western help. He can best achieve this through "good behavior"; hence his self-interested desire for a period of calm.

I am skeptical about the probability of any fundamental reorientation in Soviet global purposes before the turn of the century, if then. But because we cannot divine the future, U.S. strategic policy ought to be guided by a realistic evaluation of Soviet military capabilities—which remain formidable in every area: strategic offensive, defensive and conventional.

Some analysts believe that we can expect a cutback in the level of resources the Soviets will direct to their defense establishment. This may be small cause for comfort, however, because the motive for any cutback in military investment—if there is a cutback—seems likely to be to assure future military competitiveness, not a retrenchment in expansionist goals or a diminution of the power of the military in the state. How, then, should we proceed?


The goal of our nuclear strategy has been to convince the Soviet leaders that it would never make sense for them to initiate a strategic nuclear conflict, because of the survivability and retaliatory capacity of U.S. forces. In designing this capacity for our forces we have had to keep in mind that the Soviet Union might well discover an effective means to neutralize a particular system or technology, e.g., our land-based intercontinental ballistic missiles (ICBMs). Consequently, the United States has been careful not to put all its eggs into a single basket; instead it has followed a policy of maintaining a triad of forces comprised of ballistic missiles deployed both on land and at sea, and bomber aircraft, some armed with cruise missiles.

Our reliance through the years on the ability to threaten a credible, devastating offensive counterattack—the concept of deterrence, as opposed to defending against an incoming missile attack—has been the subject of periodic debate. The 1972 agreements of the Strategic Arms Limitation Talks (SALT) capped this debate for a time. In those agreements, the two sides decided to limit strategic offensive missile launchers and, in the Anti-Ballistic Missile (ABM) Treaty, to forgo the significant defense of those forces or of populations. At the time, America’s readiness to accept this decision reflected our hope that the Soviet Union would adopt as doctrine the principle that an approximate balance of forces and mutual vulnerability could provide for stable mutual deterrence. It also reflected the practical reality that the state of the art in defensive technologies made effective defense infeasible.

There were important but unstated American qualifications concerning what constituted "balance." Specifically, our postwar military strategy rested not only on nuclear deterrence, but on collective security as well, through the maintenance of strong alliances. In Europe our strategy—called flexible response—sought to compensate for a conventional military imbalance favoring the Soviet Union by committing, in advance, American nuclear forces to be used to check and then reverse any potentially decisive Soviet conventional breakthrough in Western Europe.

For as long as the United States held an indisputable superiority in nuclear arms vis-à-vis the Soviet Union, the credibility of our stated willingness to use it was largely moot: we were unlikely to be challenged. But starting in the mid-1970s, with the advent of rough nuclear parity between the two sides, inevitable doubts began to emerge on both sides of the Atlantic as to the credibility of America’s willingness to risk its own nuclear devastation over a conflict in Europe.

These doubts led to U.S. efforts to lend credibility to our pledges of nuclear intervention. One such measure was to move away from reliance, in our targeting doctrine, on the threat to destroy Soviet cities, and instead to make clear our capability and intention to maintain survivable, highly accurate systems that could attack Soviet military targets in Europe and elsewhere. We also modified our strategy to include plans for the possible execution of "limited nuclear options"—for attacking Warsaw Pact and other military targets discretely—while maintaining better control over the pace and character of escalation.

In addition, faced in the late 1970s with the Soviet deployment in the western U.S.S.R. of SS-20 intermediate-range missiles armed with multiple independently targeted reentry vehicles (MIRVs) that could target all of Western Europe, NATO decided in December 1979 to deploy its own corresponding capability, although in much smaller measure. It was to consist of 108 U.S.-built Pershing 2 ballistic missiles and 464 ground-launched cruise missiles (GLCMs) stationed in Western Europe.

These missiles, which were to be targeted on the Soviet Union, were ostensibly meant to deter the Soviet SS-20 force. But the more plausible NATO motive was to link any European conflict more concretely to U.S. strategic nuclear weapons. It was assumed that if we had already crossed the nuclear threshold by using Pershings and/or GLCMs in Europe—an act that Europeans assumed would be easier for us than launching ICBMs from North Dakota were the Soviets to cross the East-West frontier—then the decision to launch central strategic systems against targets deep inside the Soviet Union would be less anguishing. If the Soviets appreciated such calculation, deterrence would be enhanced.

These several actions—the revision of our nuclear doctrine, the development of limited nuclear options and the deployment of intermediate-range nuclear forces in Europe—were designed to strengthen American credibility at both the nuclear and conventional levels.

The doctrine was fine. But it was dogged by our inability to maintain survivable, responsive and effective nuclear forces to back it up. In a nutshell, our 550 Minuteman III ICBMs, each armed with three warheads, were our only responsive missiles with sufficient accuracy and high enough nuclear yield to threaten hardened Soviet missile silos and most other military targets. But the Minuteman III force was too small to hold at risk a significant percentage of the Soviet target base. And it was deployed in fixed silos that were becoming ever more vulnerable to a Soviet attack.

In the mid-1970s, the Soviets took two actions that made matters worse. First, they built a very large force of new ICBMs (the SS-18 and SS-19 missiles, carrying ten and six warheads respectively) that were capable of destroying our silos and hardened military targets. Second, the Soviets hardened their own missile silos so as to reduce their vulnerability to attack. By the end of the decade, the Soviet Union had raised an alarm in the United States: Would the U.S.S.R.’s advantage tempt it, during a crisis, to destroy most of our counterforce systems in a preemptive strike, leaving us with no options other than surrender or a suicidal attack on Soviet cities? There was also concern that the Soviet Union was carrying out active research and development in defensive technologies.

In sum, within ten years after the SALT I agreements were signed, questions arose regarding the two axioms on which they were based—our hope that the Soviets would come to accept our concept of balance and mutual vulnerability, and our assumption that cost-effective strategic defenses were infeasible.

The United States responded in the mid-1970s by proposing to modernize its ICBM force. We sought to deal with the problems of imbalance and vulnerability by designing a new, highly accurate missile with up to ten warheads, the MX, which we planned to deploy in a basing arrangement that would ensure its survivability. The latter effort focused upon several variations of deceptive or mobile basing. (It should be noted that the Russians have since moved to mobile missiles, the multiple-warhead SS-24 and the single-warhead SS-25, which are currently being deployed.) But in the United States, a considerable popular backlash emerged, as antinuclear sentiment combined with partisan political opportunism and legitimate military criticism to roll back President Carter’s plan for a multiple protective shelter deployment mode. This same combination of sentiments subsequently killed a series of basing schemes proposed by President Reagan.

At that juncture in late 1982, with the charged political climate promising continued paralysis, President Reagan appointed the Bipartisan Commission on Strategic Forces (known as the Scowcroft Commission) in an effort to establish a political consensus on the modernization of our strategic nuclear arsenal, especially our ICBM forces. In its report, submitted in April 1983, the commission reconfirmed the conceptual soundness, military effectiveness and technical feasibility of deterrence. It stated clearly that, considering all military, political and economic factors, our best course from now until some time in the early 21st century was to seek to deter a Soviet attack by maintaining a balanced triad of modern, survivable, offensive forces. Its program recommendations emphasized the importance of assuring the survivability of all our forces, and especially of moving away from high-value (MIRVed), vulnerable (silo-based) systems in the land-based leg of the triad. The commission proposed instead the development of a small single-warhead missile to be deployed in a mobile configuration (the Midgetman).


Two weeks before this report appeared, a major new factor was introduced by President Reagan’s Strategic Defense Initiative (SDI). It is now clear that, even before he was elected, the president held the conviction that it was irresponsible to leave Americans exposed to the risk of nuclear devastation, and that scientific discovery could provide the means for effective protection against such a threat. As a corollary, he believed that such a discovery, by devaluing the threat of nuclear attack, could render nuclear weapons "impotent and obsolete." He believed that, ultimately, nuclear weapons could be eliminated from the face of the earth. And at a time when he saw nuclear weapons threatening to proliferate into dozens of countries, he thought an attempt should be made to stem the tide.

The president’s SDI proposal of March 23, 1983, raised two threshold questions and a host of subsidiary ones. First, had the state of the art in defensive technologies advanced to the point where it would be feasible in the foreseeable future to defend the entire population, and at tolerable cost? If so, how should that prospect affect our strategy and force planning? Second, could effective defenses make the elimination of offensive nuclear weapons feasible and likely?

The president’s speech provoked a firestorm of debate, and the five years since then have been marked by bitter, passionate advocacy and criticism. Concurrently, however, the scientific community has been evaluating the potential of several remarkable discoveries that might be applied to the problem of defeating tens of thousands of incoming objects (warheads and decoys). Initially, these efforts were polarized by, on the one hand, the romantic and manipulative hyperbole of the Administration and, on the other, by the flatly dismissive rhetoric of scientists who should have known better. Since this initial conflict, however, a more sober climate has emerged that makes possible a more objective evaluation of the proper role that defensive technologies might play in the years ahead.

It is useful to enumerate the basic criticisms of defensive concepts: 1) they would only provoke offsetting countermeasures, including expanded and more modern offenses; 2) they would present the appearance to Soviet leaders that we might be aspiring to a first-strike capability—a destabilizing impression that could provoke Soviet preemption; and 3) to the extent that development and testing—not to mention deployment—could involve violation of the ABM treaty, the utility of arms control would be undermined.

Truly impressive gains have been made in the technologies associated with ballistic missile defense since the late 1960s. The most dramatic has been in the technology of battle management—specifically in the speed and problem-solving capacity of computers. With regard to weapons applications, substantial improvements have been made in efficient missile propellants and in guidance systems that make possible high-speed, non-nuclear, ground-based interceptors of greater reliability. Separately, the technology of kinetic-kill systems has advanced considerably and holds a high probability of being adaptable to ground- or space-based platforms within ten years. As for the more exotic technologies—directed-energy and particle-beam systems—much of the information about U.S. programs is classified. But it is fair to say that truly remarkable strides are being made, particularly in the free-electron laser.

Yet there is no basis for saying that any of these systems will be adaptable as effective weapons systems before the turn of the century, except for two traditional ground-based interceptors (called the exoatmospheric reentry vehicle interceptor system, or ERIS, and the high endoatmospheric defense interceptor, or HEDI) and a relatively primitive space-based kinetic-kill system. In short, the answer to the first threshold question—can defensive technologies provide the means to defend us against ballistic missile attack?—is that we do not know and cannot know for the next ten to 15 years.

A critical military weakness of space-based systems is vulnerability to Soviet countermeasures. Ground-based systems such as ERIS and HEDI are susceptible to being easily overwhelmed by Soviet missile warheads, which can be produced relatively cheaply. Systems operating from low earth orbit in space, moreover, are vulnerable to a family of countermeasures, both ground- and space-based, at costs that also favor the Soviet Union. Pentagon planners believe that the vulnerability of space-based systems can be overcome by building into each platform a maneuvering capability and other protective measures, but it is not clear that these measures could be adopted at a cost approaching that of the Soviet countermeasures. As a separate matter, there remains the difficult problem of how to enable a U.S. interceptor to distinguish between a real nuclear warhead and thousands of decoys.

The central reason for opposing exclusive reliance upon defensive systems for protection against nuclear attack is that to do so is to misrepresent the character of the threat, and, more cynically, to mislead our own people concerning feasible and sensible options for dealing with it.

It is irresponsible to assert either that, once developed, an instrument of mass destruction can be eradicated or that it portends inevitable catastrophe. To suggest that the Soviet Union, or any other nuclear power facing a nuclear threat, would voluntarily do away with all its own nuclear weapons is to assert that the differences that divide great powers are unimportant matters, of no consequence; that Soviet leaders are not serious about their international ambitions or the defense of their homeland and therefore will not see it as imperative to maintain the most modern weaponry for promoting and defending their national interests. Nuclear weapons cannot be disinvented. The technology essential to the construction of these weapons has spread throughout the world.

The Soviet Union has been deterred from the use of nuclear weapons, and a strong basis exists for believing it will remain deterred if faced with the enduring impossibility of any rational concept of gain. Because there is no current basis for confidence that a survivable defensive shield is within reach, and in view of the very substantial problems of political, economic and military stability attendant to making a commitment to such a shield, the adoption of a defensive doctrine, even as a goal, would be imprudent.

At the same time, a vigorous strategic defense program is very sound policy. A properly structured program is essential to our national security. In the event that an arms control agreement providing for deep reductions in strategic weapons is reached with the Soviet Union, the diminution of the threat measured in terms of Soviet ballistic missile warheads will affect the scale of any strategic defense program designed to deal with it. If we are no longer dealing with many tens of thousands of incoming warheads, but, say, only 5,000 to 10,000, the program requirements for strategic defense should be much more modest.

In this context, it is clear that a review of the SDI program is appropriate. As such a review is undertaken, the concept of strategic defense should be broadened to encompass how to preserve the survivability of each leg of the triad—not just ICBMs—against the putative threat before us.

While there is much to criticize in the misleading simplicity of the Administration’s announcement of the SDI program and the way in which it exploited popular antinuclear aspirations, one cannot ignore the positive practical consequences. In truth, the analysis behind the original concept was oriented toward dealing with an intractable military problem—a worsening counterforce imbalance which we appeared unable to check.

While we can argue whether SDI could conceivably deal with this problem and whether, even if it could, other solutions would not be preferable, it must be said that, in retrospect, SDI—as evaluated by the Soviets—seems to have played an important role in leading the U.S.S.R. to engage seriously in an effort to meet our concerns over the counterforce imbalance. In addition, the prospect of SDI may have been an important influence in encouraging Kremlin leaders to reassess their economic system’s fundamental ability to compete, in leading them to the ultimate conclusion that it could not, and in bringing about their acceptance of the need for perestroika.

For those of us involved in shaping the original SDI proposal, two conclusions stood out regarding the ability of the United States to restore and preserve a stable balance of offensive systems.

First, the United States had been reduced to competing on terms according to which the Soviet Union enjoyed a comparative advantage—that is, in building and deploying ICBMs. The problem was not that we could not build as good a missile, but that we could not fashion a political consensus behind a deployment plan, so divisive were the effects of partisanship, environmentalist opposition, legitimate military misgivings and antinuclear sentiment.

Second, those of us in the White House who proposed the Strategic Defense Initiative believed that a reorientation of our investment strategy to emphasize an area of our comparative advantage—excellence in high technology—could persuade the Soviets that we could outstrip an entire generation of Soviet military investment. If we could do so, we would remove their only claim to superpower status and perhaps lead them to deal more constructively with our concerns about their forces. The Soviets’ commitment to a 50-percent reduction in their heavy missile launchers and warheads—a possibility they would not discuss seriously until after SDI was proposed—represents a vindication of that strategy.


If a reorientation of our strategic doctrine is infeasible and imprudent at this time, then we must return to the basic task—how otherwise to address the offensive imbalance in a way that enhances overall stability. The goal for U.S. strategic offensive forces is to deter the Soviet Union from committing aggression. This is best assured by our having strategic forces that are survivable, have a short time of flight, are highly accurate and within easy, reliable reach of communications.

Submarine-launched ballistic missiles (SLBMs) fulfill the first two qualifications very well; they are survivable and can reach their targets quickly. The design criteria for the new generation of such missiles, the D-5, include sufficient accuracy and yield to enable SLBMs to attack hardened Soviet military targets successfully. Unfortunately, however, the difficulties in discrete, dispersed targeting inherent in any multiple-warhead missile, when combined with the problematic communications characteristic of submarines, make this a poor weapon of choice. If the goal is high confidence of assuring rapid initial response as well as measured responsiveness through the course of the battle, SLBMs do not fill the bill.

That leaves land-based forces, ICBMs, as the only alternative. They possess the requisite accuracy and short flight time, and communications with them are reliable. Survivability is the problem with ICBMs.

Our goal should be to move both sides in the direction of more survivable forces. This objective is promoted by reducing the ratio of each side’s warheads to the number of aim points on the other side that would have to be attacked in a preemptive strike. By definition, the objective of reducing the warhead/aim point ratio can be approached by reducing the numerator—the size of the Soviet counterforce arsenal—through arms control negotiations, by increasing the denominator by adding American aim points, or by a combination of the two.

It is to President Reagan’s credit that he has succeeded in getting Soviet commitment to reductions in its ballistic missile forces, and especially in its heavy, silo-based MIRVed ICBMs (SS-18s). If the current Soviet agreement to reduce this force by 50 percent can be translated into a final agreement while the current warhead/aim point ratio is maintained or lowered, this would represent a very substantial gain for the stability of the East-West nuclear balance. If the Soviets do in fact cut their SS-18 force by 50 percent to 1,540 warheads, stability will improve or remain the same, but only as long as we keep 500 or more aim points on our side. The important issue to keep in mind, however, is the ratio, with or without arms control.

But reducing the number of nuclear weapons—decreasing the size of the numerator—is not without its own set of problems. After reductions, we may have no more than 20 ballistic missile submarines in our force and perhaps no more than 12 at sea, instead of as many as 41 boats, with 22 at sea, as in years past. This smaller submarine force would be inherently more vulnerable, simply because the Soviets could concentrate their entire submarine attack force—the size of which is unaffected by the Strategic Arms Reduction Talks (START)—on finding fewer U.S. boats.

Regardless of the outcome of arms control negotiations, the toughest issue remains whether we will make the hard decision today that we have avoided in the past—to make our new ICBMs survivable by adopting a deceptive-basing or mobile-deployment scheme (whereby the number of aim points becomes substantially larger).

The choice is complex. We seek to maximize the number of our aim points, and thus survivability, with the least disruption of the civil/environmental setting and the least financial cost. We should proceed from the assumption that within the lifetime of the system we are designing, any START agreement reached may expire. Since, therefore, our missile force could face attack from huge numbers of Soviet ICBMs and/or SLBMs, ideally it should be able to survive an attack without warning. It should be a system that, first and foremost, meets our military requirements; if possible, it should also be compatible with the verification desiderata of sensible arms control agreements.

There are several approaches that, to varying degrees, would fulfill these criteria. Those considered have included horizontal multiple protective shelters (MPS) and closely spaced basing (CSB), both of which were turned aside earlier because of different disadvantages.

Horizontal MPS, proposed in the late 1970s, provided for 200 MX missiles to be moved randomly within a matrix of 2,300 horizontal shelters, each separated from the others by a mile and hardened against 200 pounds per square inch of overpressure. It would have taken at least 4,600 Soviet warheads to barrage such a deployment successfully, and the shelters would have cost between $40 billion and $50 billion to construct. Horizontal MPS would have occupied an enclosed land area of approximately 4,000 square miles.

CSB was based on a different principle; it called for deploying a given number of MX missiles in a corresponding number of silos spaced near to each other. This system relied upon the probability that, if an attack were launched against it, the first incoming warheads would, upon detonation, create conditions in which successive waves of the attacking force would be destroyed or would be rendered inoperable for a period of several minutes (a phenomenon called fratricide). It would have required a substantial hardening of the silos to avoid damage to more than one from each single blast and was undermined by uncertainty over largely undemonstrated theories of fratricide effects.

The newer approaches are rail garrison, the so-called carry-hard approach and road-mobile basing.

Under rail garrison, up to 100 MX missiles would be deployed on trains based on military reservations. Current Pentagon thinking calls for the trains to operate outside the bases periodically, without nuclear warheads. In time of tension the trains would operate constantly, moving to prearranged assembly sites for warhead mating once an alert condition was directed. The chief drawback to this system is that it would require a matter of two to three hours to move to full alert. Such warning time may well be available, but to rely on it would be to subvert our historical resistance to a policy of "launch on warning." Throughout the nuclear age the United States has operated its strategic forces in the expectation that we will not have, and should not plan for, more than a few minutes warning. This is why we have kept ballistic missile submarines on patrol and bombers on alert. There is no reason to alter this planning assumption now, and to do so would be imprudent.

The "carry-hard" MPS plan was designed to deal with a number of criticisms of the horizontal MPS system, notably: the high costs of each shelter due to the need to harden each one; the large land area required; the relative softness of the shelters, even with hardening, due to modifications required to meet verification needs; and persistent doubts over the system’s penetrability by Soviet intelligence. The carry-hard version’s improvements are achieved by putting each missile in a hardened canister, thus requiring only 100, not 2,300, hardened housings. These hardened canisters (resistant to 10,000 pounds per square inch), with missiles inside, would be moved randomly among over 2,200 shelters, as with the original MPS system, although the shelters would be vertical. As a consequence of the additional hardening, much greater Soviet accuracies would be required, and each shelter could be separated from its nearest neighbor by as little as 1,500 feet (as opposed to a mile between the shelters in the original MPS plan), thus requiring much less land area—250 square miles as opposed to 4,000 square miles. And unlike the original MPS system, civilian observation would be prohibited to a distance of three miles. This system demonstrates an impressive ability to meet the criterion of assuring high survivability, particularly in a strategic climate where warheads are constrained by an arms control agreement. It also withstands higher overpressures and is relatively cheaper.

The road-mobile concept calls for single-warhead Midgetman missiles to be mounted on specially designed trucks and operated from military bases, preferably in the southwestern United States. It provides better survivability than any other system, since its operating area is larger (and therefore the number of Soviet warheads needed to barrage the area would be far greater). The 15-year cost would be about $29 billion for a 500-missile force. Its greatest advantage over the alternative systems is that it would assure survivability against even the most extravagant Soviet threat. If we consider the worst-case threat imaginable—no arms control limitations and a Soviet decision to build to the limit of their capability—only the road-mobile deployment concept would fulfill the survivability criteria.

If, however, a START agreement were to be concluded along the lines of the current U.S. position (but with mobile missiles allowed on both sides), we could improve system survivability under any of these alternative concepts, with the exception of the rail-garrison proposal. The lengthy alert time required by rail garrison would still make it less desirable than the others.

Turning to the modernization of missiles themselves: the basic wisdom of the Scowcroft Commission’s proposal, that a new small single-warhead missile be developed, remains sound. It is also clear, however, that new techniques for updating current systems, and the fiscal realities before us, warrant consideration of alternatives. These include modification of the Minuteman II and/or Minuteman III systems. The Air Force has analyzed the military effectiveness and cost of refurbishing each of these systems. The Minuteman II, which is currently a single-warhead missile, could be modernized by refitting it with new propellant grains and a new third stage, as well as the single warhead and guidance system from the Minuteman III, at a cost of approximately $8.1 billion for 500 missiles. The Minuteman III missile could be modified by downloading two of its warheads at an even smaller cost.

If either alternative were selected, however, we would have a new problem: we would have to declare a system with the demonstrated capacity to deliver three warheads—the Minuteman III system or the upgraded former Minuteman II—to be considered henceforth a single-warhead system; this would lead to inevitable Soviet insistence on reciprocal treatment for their MIRVed systems. Unfortunately, even if agreed, the larger throw-weight of Soviet missiles would give them an enormous breakout advantage, were they to cheat by reverting to multiple warheads. For this reason there is a strong argument for choosing the Midgetman, which is designed at the outset to carry only one warhead. In addition, considerable research and development work has already been done on the Midgetman toward meeting the demanding conditions of road mobility; a converted Minuteman system would likely prove less reliable in this area.

The foregoing analysis might best be summarized by presenting the most desirable approach to meeting the most demanding requirements, and then proceeding by regression to less demanding levels.

—If there is no arms control regime and we must deal with an unconstrained Soviet threat, leaving aside fiscal limitations, our best course would be 1) to proceed with the MX, to be deployed in a carry-hard configuration at a force level of at least 100 missiles, and 2) to plan for the deployment of a large Midgetman force—at least 600 missiles—in a road-mobile configuration. Such a deterrent would present an extremely demanding challenge to any Soviet attack, because of the enormous proliferation of aim points to be attacked.

—If, however, an arms agreement is concluded at or about the currently proposed warhead levels, and we continue to assume no fiscal constraints, then the 50-missile MX program in silos, augmented by at least a 500-missile Midgetman program in a road-mobile configuration would fulfill our needs.

—Proceeding further, however, if fiscal realities and our political process foreclose the expense of proceeding to a $29-billion Midgetman deployment at this time, then the same 50-missile MX program (in silos), paired with a program calling for 500 Midgetman missiles in a carry-hard configuration, is the next best choice.

In my judgment, the overall economics argue for going directly to a mobile deployment mode for the Midgetman system. I believe that such a course is affordable without undue damage to other major procurement programs. At the same time, there is certainly room for argument. Putting a Midgetman or upgraded Minuteman force of 500-600 missiles in existing silos on an interim basis (assuming a START agreement along currently agreed numerical lines) would represent an improvement upon the current situation in terms of stability. Consideration should be given to this course.


To sum up: deterrence, through maintenance within a triad of forces of survivable retaliatory capability to inflict prompt, intolerable damage on Soviet military and related targets, remains the most cost-effective, stable and militarily sensible way to avoid nuclear war and lower-order aggression in Europe.

In the future it may be necessary for the United States to adopt some form of strategic defense to maintain the credibility of its nuclear strategy. There are foreseeable circumstances in which an accidental or unauthorized launch might endanger military and/or civilian targets. And there are other reasons for a serious program of research and technological development applicable to defense against ballistic missiles and air-breathing systems (aircraft and cruise missiles). These include:

—assessment of the cost-exchange ratios for defending various targets to various levels of survival in the face of responsive threats, including countermeasures, to determine the vulnerability of the defenses to direct attack and to analyze the stability of configurations of defensive and offensive deployments.

—deterrence of, and if necessary response to, a Soviet ABM breakout.

—preservation of possible options for the active defense of selected retaliatory forces and strategic command, control, communication and intelligence against certain small attacks, such as accidental, unauthorized or third-country launches.

Among the political decisions the United States should make are: to abide by the restrictive interpretation of the ABM treaty; to protect fully, in any negotiations, the ability of the United States to do research under the treaty; and to continue to evaluate SDI by the criteria of cost-exchange ratios at the margin, degree of vulnerability to direct attack, effect on stability, and impact on the strategic balance. The cost and value of SDI must also be weighed against other critical military programs for both strategic and general-purpose forces.

The SDI research program should avoid "spectaculars" carried out for their own sake. It should explore ground-based terminal defense of mobile or fixed hard points, ground-based and space-based sensors and advanced directed-energy weapons (such as ground-based free-electron lasers and space-based chemical lasers). Insofar as space-based kinetic-energy weapons appear to be vulnerable to straightforward countermeasures, that approach should be de-emphasized.

This general program should also proceed at a measured technological pace, with annual funding increasing from the fiscal year 1988 level ($3.7 billion over the next two or three years) at a rate consistent with efficient progress—on the order of ten percent per year.

Some of the funds thus made available should be diverted to high-technology conventional force initiatives, in order to reduce the risk that the rapid increase in SDI research and development funding may starve U.S. conventional forces of the technological edge they need to offset the Soviet numerical advantage in conventional forces. Any decision on full-scale engineering development of a multi-tier SDI system should be delayed at least until the early 1990s.

With regard to arms control, our purpose should be to seek agreements which contribute to the establishment of balance and stability, as described in this article. In that context, I believe that the current U.S. proposal in the START negotiations is generally sound, but only if our strategic modernization program is guided by the need to reduce or at least preserve warhead/target ratios. It is particularly noteworthy that the Soviets have agreed to reduce their SS-18 force by 50 percent. Presumably, the United States will drop its demand that mobile missiles be banned, although as a stratagem to underscore the importance of assuring adequate verification, this demand has not yet served its purpose.

The past five years have been a time of enormous change. We entered those years with public support for a more assertive U.S. international role. Much of that enthusiasm has been usefully channeled into sensible national security programs, but some of it has been wasted, in part as a consequence of our ambivalence in regard to strategic concepts.

The next five years promise to be equally tumultuous as a result of fiscal pressures, technological progress and the normal, disarming swing of the political pendulum in the West back toward the romantic side of the arc. This latter trend is most unsettling. For unlike times past when, as these swings occurred, our superior strength afforded us a margin of error to experiment with various concepts for securing our interests, we no longer have that luxury. It has become much more important that we be guided by a concrete understanding of the threat before us and the path that, historically, has been the best way to deal with it.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Robert C. McFarlane was National Security Adviser to President Reagan (1983-85); he is currently a Counselor at the Center for Strategic and International Studies in Washington, D.C. Copyright © 1988 by Robert C. McFarlane.
  • More By Robert C. McFarlane