THE subject of defense against ballistic missiles probably occupies a unique position among strategic issues of the nuclear era. It has been more intensely debated in the United States than any other weapon system selected for deployment, such as the air-defense system or Polaris or Minuteman, or than any arms-control measure adopted to date, including the ban on nuclear tests. In fact, the decision to deploy such defenses may well be more important than any other single decision so far made concerning our strategic nuclear forces. Most of the published articles relating to defense against ballistic missiles (BMD for short; sometimes, but not here, denoted ABM) have opposed deployment of missile defense by the United States.[i] This is odd, in view of the fact that, at least through 1968, the Administration and Congress both clearly supported American deployment of BMD, and that most senior American academic strategists and many prominent advocates of arms control favor deployment at least under some conditions. The doubts increasingly expressed in the Congress in early 1969 may well result from the highly one-sided literature on the subject.

The present article is an attempt to clarify some of the reasons why many analysts think it entirely reasonable to favor American deployment. It will concentrate on considerations relating to an American BMD system intended to have substantial effectiveness against major Soviet attacks, rather than on a so-called "light" or "thin" system intended to be effective only against Chinese or other marginal attacks. The system now being deployed- the Sentinel-has been identified as an anti-Chinese system, but much of the support (both inside and outside the Government) for the Sentinel decision came from those who believed that the system would eventually have significant capability against large Soviet attacks. It seems very likely (and desirable) that whatever system finally emerges will eventually have such a capability, and therefore it seems appropriate to concentrate here on the actual policy issues this prospect presents.


In an important sense, the key issues in the debate about BMD are not technical. Within the community of people who have carried out the BMD development, certain estimates have been made as to the plausible range of technical and economic effectiveness of systems that can be achieved in the near future. These estimates are subject to some controversy, yet even if the uncertainty and controversy concerning these estimates were wholly removed, most of the articulate critics of BMD would remain critical: their objections are rooted in other concerns. Nevertheless, technical issues are vital to the debate, since most of the support for BMD deployment depends on the fact that a defense of substantial effectiveness is judged to be feasible.

The most common way of characterizing the effect of a BMD system is to estimate the number of lives it might save in various specified circumstances. In a table of such estimates given in McNamara's 1968 posture statement,[ii] it was indicated that there could be 120 million American fatalities in certain possible wars of the mid-1970s if no significant BMD were deployed. Assuming opposing forces and attacks of the same strength, it was indicated that BMD systems costing from $10 to $20 billion could reduce American fatalities to between 10 and 40 million, depending on the level of the defense and the details of the war. Damage to production and transportation resources would, of course, be similarly reduced, a result that could not be achieved with economically feasible civil-defense shelter programs. Thus, such a defense might change the postwar situation from one in which over half the U.S. population was gone, and recovery in any time period would be problematical, to one in which perhaps 90 percent survived and economic recovery might be achieved within five to ten years. This difference would be enormous.

It is this possible difference that constitutes the major reason for deploying heavy defenses. In effect, procuring such defenses is like buying "insurance" that would limit the consequences of a war; the outcome would still be a disaster, but probably one of a very different order than would result from having the same offensive forces expended in a war with no missile defense.

It is possible that a BMD system might perform in an actual war much less well than expected, because of some unforeseen technical failure; it is, however, about equally likely that the opposing offensive forces will perform much less well than expected, which is to say that the defenses may perform much better than expected. (Critics of BMD are prone to emphasize the first of these points much more than the second.)

Of course, the U.S. defense might have to face a larger offensive force than the Soviets would have created if the U.S. defense did not exist. In other words, the Soviets could increase their offensive force so as to nullify, partly or wholly, the U.S. defense. Whether or not they seem likely to try to do so, it is important to consider how difficult it would be to achieve. This is one of the important technical characteristics of offense-defense interactions.

A useful way of characterizing the degree of difficulty is with a parameter called the "cost exchange ratio," defined as follows. The United States might deploy a particular defense system at a particular cost. It would generally be feasible for the Soviets to add an increment to their strategic offensive forces that would offset or nullify the opposing defense, and this increment of offense would also have a particular cost. The relation between the cost of the offsetting increment of offense and the cost of the defense is called the "cost exchange ratio." Thus, relatively ineffective defenses would have a relatively low cost exchange ratio, while a defense system that was relatively difficult to penetrate would have a relatively high ratio.

Several years ago, it was widely believed that missile defenses were easy to penetrate-so easy that offensive increments costing only one or a few percent of the cost of the opposing defense would serve to nullify it. In recent years, however, it has become apparent that cheap forms of decoys and other penetration aids cannot be relied upon to nullify modern defense techniques. A good defense can be overcome, but it is difficult. This is reflected in the fact that cost exchange ratios for a good defense are now believed to be in the region of one to one-perhaps one-third or two, but not one-tenth or ten. Thus, it is about as expensive to nullify a good defense as to build it.

Some specific examples were provided in the 1967 posture statement. Two postulated defense systems were considered, one costing $10 billion and the other $20 billion. Under the conditions of a hypothetical war in which U.S. fatalities were estimated to be 100 million without any missile defenses, these BMD systems reduced the fatalities to 30 million and 20 million, respectively, if the Soviets did not change their offensive forces in reaction to the defenses. If the Soviets added an increment to their offensive forces costing one-quarter as much as the American BMD, they could raise American fatalities to 40 million; if they spent one-half of the cost of the defenses, they could raise the level to 60 million; and if they spent as much as the full cost of the defense, the fatalities would rise to 90 million. To raise the level of fatalities back up to the undefended level (100 million)-the usual criterion of offsetting a defense- would require an increment in the Soviet expenditure that would exceed the cost of the defense. These calculations assumed that the Soviets had advanced technology for their offensive forces and made effective use of it. Since these estimates were published by Secretary McNamara, who was a strong critic of missile defenses, it can be assumed that they are not biased in favor of missile defenses.

Although McNamara published information showing that exchange ratios were about unity as far back as January 1967, and although information to this effect was known in the Government for perhaps two or three years before that, many critics of missile defenses continue to assert that it is cheaper to offset defenses than to build them. For example, Carl Kaysen, in his recent article in this journal, said: "Were we to deploy much stronger and more costly defenses [than the initial Sentinel program] it would remain within the Soviet capability to counter them by corresponding increases in the size of their offensive deployments-which could probably be made at significantly less cost than that of our increased defenses" (emphasis added). If "counter" means "wholly or largely offset," as is usual and as Kaysen's context suggests, there is no objective basis for the italicized statement, and there has not been for some few years.[iii]

Many critics of BMD have questioned the estimate of the cost. McNamara, in speaking about the $20 billion program, several times referred to the fact that many weapon systems proved to cost twice as much as originally estimated, as indeed they have, and suggested that this program would more likely cost $40 billion when completed. Some journalists and others have in turn doubled McNamara's estimate once again, and one can find articles and newspaper editorials referring to BMD systems alleged to cost $80 to $100 billion. It is difficult to escape the feeling that these latter estimates are made purely for their political effect and have no significant basis in reality. At the time official estimates were worked out, missile defense systems had been much more extensively studied than many of the weapon systems whose early cost estimates proved much too low.

Obviously the relationship between cost and performance of systems as large and complex as those considered here cannot be specified in advance with anything like precision. This should not obscure the fact that the range receiving major consideration or advocacy is, say, $8 to $20 billion. For some perspective on this range of costs, it is instructive to consider that the United States has spent perhaps $50 billion on air defense since World War II.[iv] We are currently still spending about $2 billion per year on air defense. And, while estimates of its effectiveness are not publicly available, it is a good bet that the current air-defense system is and has been less important as insurance for the country than the major proposed BMD systems would be.

Will our estimates of cost effectiveness stand up when we get to the mid- 1970s? A question in this form does not admit an unequivocal answer; one cannot say with confidence whether inventions not yet made or developments not yet realized will favor the offense or the defense, although there are some reasons for thinking the trend will continue to favor the defense. But a parallel may be useful. We have fairly high confidence that the Polaris submarine force is a reasonably secure component of our strategic offensive forces. The main reason for this confidence resides in the fact that we have conducted a major research and development program-on the order of $500 million per year-in anti-submarine warfare for many years, and no cheap and reliable way of attacking such submarines has been found. A similar statement is beginning to be true in relation to BMD: for several years, we have conducted a major research and development program-on the order of $350 million per year-in means of penetrating missile defenses, and no cheap and reliable method has been found. This does not guarantee that one cannot be found, and it does not say that expensive ways of penetrating are not known-as are expensive ways of attacking Polaris submarines. But it does suggest that the technical prospects for missile offense-defense interactions are likely to prove acceptably stable, and in particular should be a great deal better than the early estimates of air defense proved.

It is worth noting that a BMD system may possibly have important effects even if it later failed to perform as expected in a war. If the Soviets were to react to a U.S. defense by fitting their existing missile force with decoys and other penetration aids, without a major increase in the number of rockets, they would reduce the total payload available for warheads, and thereby reduce the potential damage the U.S. might incur in a war even if the defense failed utterly. This effect, which is known in the trade as "virtual attrition" and is often encountered in defensive systems of other kinds, is likely to be quite modest. A more important possibility is that a U.S. defense would induce the Soviets to re-target their offensive force; they might concentrate most or all of their offensive missiles (apart from those used for missile bases or other military targets) on the largest cities to be sure of destroying them, and leave unattacked many medium and small cities. Or they might attack only undefended areas, and leave aside most or all the largest (defended) cities. In either case, the mere presence of the defense could result in saving many millions of lives whether it "worked" or not.

There are three other positive reasons for favoring BMD deployment. First, the time may soon arrive, if it is not already here, when there will be some possibility of attacks of anonymous or disguised origin. Since these would not be subject to standard threats of deterrence, active defenses may be the primary protection against them. Second, the possibility of a purely accidental launch of some part of the Soviet force is probably very remote, but the possibility of an unauthorized launch, especially during an intense crisis, may not be so remote and should be protected against. Even a modest defense might be very effective against such an attack; this has been one of the arguments used-rightly, I believe-in support of the current Sentinel program. Third, missile defenses (even light defenses) would considerably complicate the planning of an attacker who hoped to penetrate them; this phenomenon seems likely to serve as an additional "firebreak" to the initiation of a strategic nuclear war.


One of the main areas of concern to critics of missile defense has to do with its impact on fundamentals of deterrence. The problem has often been related to what McNamara termed "Assured Destruction," defined by him in his 1968 posture statement as "an ability to inflict at all times and under all foreseeable conditions an unacceptable degree of damage upon any single aggressor, or combination of aggressors-even after absorbing a surprise attack." McNamara recognized that what constituted "an unacceptable degree of damage" was not subject to precise specification, and in spelling out this requirement, he said: "In the case of the Soviet Union, I would judge that a capability on our part to destroy, say, one-fifth to one-fourth of her population [i.e. about 50 million Russians] and one-half of her industrial capacity would serve as an effective deterrent."

Because McNamara came to regard the ability to destroy 50 million Russians as the keystone of Western security, he viewed Soviet deployment of BMD as a potential threat to that security, and intended to nullify any Soviet defenses with added U.S. offensive forces. He also appeared to believe, and frequently asserted, that the Soviets had a similar requirement for an Assured Destruction capacity, and that any U.S. interference with this requirement would in all likelihood only cause the Soviets to increase their offensive forces. It appears that this perception was at the core of McNamara's opposition to BMD deployment, as it was for many other opponents of missile defenses. The U.S. strategic posture that has evolved from this perception in recent years has been aptly dubbed "Assured Vulnerability" by Steuart Pittman.[v]

Although there were many criticisms made of the "massive retaliation" policy when it was adopted in the early 1950s, there were few if any criticisms to the effect that the U.S. threat was inadequately deterring. Yet it is doubtful if the Soviet fatalities that our force could then have produced would have exceeded a few million. And this was the Soviet Union of Stalin, believed to have a six-million-man army, a Soviet Union which had only recently subjugated Eastern Europe, and which was believed to have instigated or at least approved the Korean War. Not even the recent Czech invasion makes the Soviet Union of today seem nearly as threatening.

Of course, when major direct Soviet attacks on the United States became possible, the structure of the situation changed considerably. It might then have become possible for the Soviets to have eliminated the United States as a fighting society, and thereby to secure Europe and eliminate their principal opponent at one blow. Thus, many strategists judged that there were two reinforcing reasons why it was appropriate to react to this possibility with a large strategic threat to deter it. This was the fundamental driving force behind the evolution in the late 1950s and early 1960s of a U.S. strategic threat measured in thousands of megatons-a deterrent threat for which Assured Destruction was indeed the correct phrase.

It is clear from the origin of this perfectly reasonable logic that, to the extent the maximum possible Soviet motivation for such an attack might have been reduced, or to the extent the U.S. motivation to deter it might have been reduced, the scale of the retaliatory threat might have been reduced in some corresponding degree. It is this linkage that appears to have been wholly overlooked. Instead, we preferred to fix once and for all some intended level of Assured Destruction believed capable of deterring the Soviets from any actions whatever. There are a number of problems resulting from such a posture, of which the most immediate for our purposes is that it tends to make it difficult for us to limit damage to ourselves-i.e. it tends to force us into a posture of Assured Vulnerability.

In particular, a determination to maintain a strong capability for Assured Destruction led McNamara in 1967-68 to respond to incipient Soviet missile defenses by scheduling an increase in our offensive forces, and the theory that the Soviets would do likewise led him to oppose American deployment of an anti-Soviet BMD system. In view of the effectiveness of modern defense, we might better use the U.S. resources committed to increasing our offensive forces to increase our defenses instead. By thus reducing the Soviet threat, rather than increasing our own, we could reduce both the extent to which the Soviets might gain by attacking us, and the extent to which we are intensely motivated to deter the attack.

Consider a situation in which Soviet and American offensive forces are fixed at similar unchanging levels on each side, and in which both the United States and the Soviet Union are building comparable levels of defenses. If the American and Soviet defenses are both "light," then our BMD would not much diminish Soviet ability to destroy us, but then Soviet BMD would not much reduce the effectiveness of our assured-destruction forces either; if both American and Soviet defenses are "heavy," then our assured-destruction forces would be significantly degraded, but so too would the capability of the Soviets to eliminate the United States as a fighting society.

This points to the following formulation of a reasonable requirement for a conservative American strategic posture: Following any plausibly feasible strategic attack by the Soviet Union, the United States should have the capacity to inflict as much or more total damage (of similar kind) on the Soviet Union as the Soviets had inflicted and could still inflict on the United States. (This principle implies that the strategic offensive forces must be quite well protected, which appears to be feasible.) In short, we should have a reliable capability to do at least as badly unto the Soviets as they had done or could do unto us. The Soviets could not achieve a significant military advantage by a strategic attack, and an irrational, coercive or punitive attack-whether large or small-would risk bringing as much or more destruction on the Soviets as they could or did bring on us. This would make the initiation of nuclear blackmail unattractive to any reasonable decision-maker at any effective level of strategic forces.

Let us note explicitly that this posture does not imply that a U.S. capability to destroy 50 or 74 or 100 million Russians is a fundamental requirement of nature, without regard to the circumstances. It indicates that both the United States and the Soviet Union might reasonably engage in some measures to limit the possible damage of a war without necessarily impairing American security in the process. Such measures might include, for instance, direct reduction of strategic offensive forces by agreement- or deployment of defenses with or without explicit agreement. It is interesting that the United States has, in fact, proposed direct reductions in strategic offensive forces. Now, if certain necessary requirements are met, as they can be, a symmetric increase in active defenses deployed by each of the superpowers would have approximately the same kind of potential impact on possible war outcomes that percentage cuts in offensive forces would have.

One may ask: If BMD deployed by both superpowers would have roughly the same effect on possible war outcomes as direct reductions in offensive forces would have, why not simply reduce the latter, and save the money and trouble of BMD? The answer is that in some circumstances one might, but the circumstances are not those now prevailing. To reduce U.S. and Soviet offensive forces to a level where the possible casualties on each side (without defenses) would not exceed, say, 20 million, is likely to be acceptable to the United States only with a degree of inspection that is most unlikely to be acceptable to the Soviets. In other words, offensive- force cuts on such a scale do not now seem politically feasible, whereas BMD deployment is entirely feasible and, rather than increasing U.S. needs for inspection of Soviet offensive forces, might actually reduce our dependence on such information. This effect, indeed, would facilitate later direct reductions in offensive forces.

From the mid-1950s to the mid-1960s, the strategic postures of the superpowers were dominated by the logic that, since we could not defend, we had to deter. This position, for which there was originally ample justification, now seems to be interpreted in some minds-chiefly certain American ones-to mean that, since we must deter, we cannot defend. This should count as the non sequitur of the decade.


The final major concern of critics of missile defense is that it will lead to an arms race or complicate arms-control problems. The kind of arms race usually envisaged is an offense-defense race in which, say, the United States procures some missile defenses and the Soviet Union then feels obliged to respond by increasing its offensive forces sufficiently to nullify the defense. This increase in Soviet offensive capability might in turn motivate the United States to increase its defense capability still further, and so on. For example, McNamara said in his 1967 posture statement: "It is the virtual certainty that the Soviets will act to maintain their deterrent which casts such grave doubts on the advisability of our deploying the NIKE-X system for the protection of our cities against the kind of heavy, sophisticated missile attack they could launch in the 1970s. In all probability, all we would accomplish would be to increase greatly both their defense expenditures and ours without any gain in real security to either side." (The whole of this passage is underscored in the original.) If arms races of this kind occur, it will not be in response to some fundamental law of nature, but will result from interactions between achievable technology and prevailing attitudes and budgets. Each of these factors is of importance.

It is useful to discuss this problem in terms of the concept of cost exchange ratios. If cost exchange ratios were as low as 1:100, there would be little doubt that, within the present political environment, a defense system deployed by one superpower would be nullified by the other. For example, if the Soviets built, say, a $20 billion system, and if it could be neutralized at a cost of only $200 million, there would scarcely be any debate about the matter within the U.S. bureaucracy, the Congress or the public; the Defense Department would simply go ahead and neutralize the Soviet defense. On the other hand, if cost exchange ratios were in the neighborhood of 100:1, there would again be little debate: defense would reign supreme, and deterrence would rest on very different threats and counter-threats than those that loom as dominant today. In this case, to nullify a $20 billion Soviet defense would cost the United States $2,000 billion. The expenditure of such a sum is not, of course, a realistic possibility, and the United States would instead be building its own defense-about which there would be little controversy, and which the Soviets could not offset either. In both cases, technology would dominate prevailing attitudes.

In the actual world, with cost exchange ratios near unity, the attitudes prevailing are not driven by the technology toward either deterrence or defense, but may (and do) go in either direction. It is much more a matter of preference and conscious decision whether we and the Soviets wish to spend our strategic-force budgets chiefly to increase the level of "hostages" on the other side or to decrease our own. Thus, whether we are to have an offense-defense arms race depends in the first instance on whether or not Soviet and American attitudes are lined up the same way. If both governments are willing to accept a high level of hostages on each side, with modest or no defenses, there will be no great pressures for an offense-defense race; the same will be true if both are willing to live with substantial defenses and relatively low levels of hostages, with little or no attempt on either side to offset the defense of the other. But if one side attempts by defensive measures to reduce its own hostages below the level demanded by the other, or attempts to increase the hostages of the other above a level the other considers acceptable in the circumstances, and especially if both happen together in any combination, the resulting pressures for an arms race may well be limited almost entirely by budgetary forces.

The fact that these issues are importantly influenced by attitudes is not always clearly recognized. There is, however, something of a "fashion" in some circles against interfering with the so-called requirements for Assured Destruction on either side, especially when (almost only when) the interference comes from missile defense. But, as I have argued earlier, maintaining very high levels of possible damage to the Soviets does not seem to be a requirement of U.S. security in circumstances where we can correspondingly limit damage to ourselves, and appreciation of this seems likely to prove dominant in formulating U.S. policy in the future.

Soviet attitudes seem much less mixed than ours; they are much more friendly to the deployment of defenses. In particular, contrary to many statements by American critics, it appears that the Soviets are not significantly antagonized by the prospect of American missile defense. The Soviets have never been bashful about making political and diplomatic statements designed to deter American military programs they did not like; for example, they have recently made public statements opposing the operation of Polaris submarines in the Mediterranean and the operation of nuclear-armed bomber forces on air-alert. It would be remarkable in the extreme if the Soviets had the attitudes attributed to them by McNamara and others and yet did absolutely nothing to discourage U.S. defenses. The fact is, however, that there seems to be no reliable report of a single instance of a Soviet attempt to deter our BMD by statements indicating they would increase their offensive forces in response.

On the contrary, there have been many statements by individual Russians at both official and unofficial levels, both impromptu and carefully considered, pointing in a quite different direction. In a press conference in London on February 9, 1967, Premier Kosygin was asked: "Do you believe it is possible to agree on the moratorium on the [deployment] of an anti- missile defense system [a then-current American proposal] and if possible on what condition?" He replied in part:

I believe that defensive systems, which prevent attack, are not the cause of the arms race, but constitute a factor preventing the death of people. Some argue like this: What is cheaper, to have offensive weapons which can destroy towns and whole states or to have defensive weapons which can prevent this destruction? At present the theory is current somewhere that the system which is cheaper should be developed. Such so-called theoreticians argue as to the cost of killing a man-$500,000 or $100,000. Maybe an antimissile system is more expensive than an offensive system, but it is designed not to kill people but to preserve human lives. I understand that I do not reply to the question I was asked, but you can draw yourselves the appropriate conclusions.

Indeed, one can.

A detailed rebuttal to the standard arguments against BMD was given by the late Soviet military publicist Major-General N. I. Talensky in the quasi- official Soviet journal International Affairs in October 1964. Many other contacts with Russians at official and unofficial levels, including several meetings I personally have had the chance to observe, illustrate that the attitude exemplified by the Kosygin quotation is very widely held in the Soviet Union, while the attitude held by American critics of BMD with regard to the alleged importance of maintaining high hostage levels seems not to be held at all. The only criticisms of BMD that have been expressed by any Russians, so far as I am aware, have come from a few scientists who have been skeptical of the effectiveness of defenses and therefore regarded them as wasteful of resources. It is probably significant that these scientists seem not closely associated with current Soviet weapon programs; I believe it likely that they are following the lead of some of their Western friends and counterparts, but even these Russian critics seem not to share in the theoretical opposition to BMD.

One additional bit of evidence is worth mentioning. After the announced decision in September 1967 to deploy the Sentinel BMD system, the United States came under attack from some of our allies and neutral friends, especially from several countries participating in the Eighteen-Nation Disarmament Committee then meeting in Geneva, which complained that the American deployment decision would be bad for the incipient nonproliferation treaty and would only heighten the arms race. There was one country that came to our assistance, holding that the decision would not harm the prospects for the nonproliferation treaty: that country was the Soviet Union. So far as I am aware, this is the only case on record in which the Soviet Union has defended us diplomatically against our friends and allies. This is all the more remarkable in view of the fact that the Soviets did not, then or since, believe that the Sentinel system was intended only for Chinese attacks.

Some American critics of BMD have described these Soviet attitudes as "unsophisticated," and have argued that the Soviets will "get over it," and come to respond to an American defense as McNamara predicted even if they are not yet committed to it. It is true that strenuous efforts to "educate" the Soviets have been made by McNamara and others, and one cannot say with certainty that these efforts will fail; it would not be the first time that the Soviets lost their senses over some issue just as we were coming to ours.[vi] But their present views should not be denigrated; it seems to me, and to many other Western strategists, that the defense-oriented philosophy exhibited by the Soviets makes good sense. In fundamental matters of deterrence and defense, I should say that we in the United States might better acquire some "education" from the Soviets.

In view of the fact that Soviet attitudes already seem to favor defenses, and that American attitudes, while currently mixed, ought to favor defenses, the obvious way of trying to improve prevailing strategic postures through arms-control arrangements is to limit offensive forces primarily, and to limit defenses only secondarily, if at all. This is exactly the reverse of the order of priorities frequently suggested in the United States in the past few years. It is, however, more in keeping with the traditional aims of arms control. The primary objectives of arms control have often been stated to be reduction of the likelihood of war or mitigation of its consequences if it occurs. It seems to me highly probable that deployment of missile defenses will contribute to both of these objectives, while abstaining from defenses will likely contribute to neither. If the deployments are managed with at least modest intelligence on both sides, there need not be an arms race nor appreciably higher expenditures.

A major objection to a posture with a heavy component of defense is that the Soviets might find some way of attacking our offensive forces with sufficient effectiveness so that their defenses could intercept most or all of our remaining offensive forces. By striking first, they might escape relatively unscathed-which, if true, could motivate the initiation of a first strike in an intense crisis. Obviously no one could give an absolute guarantee that such a weakness could not occur. Major studies of BMD performance do not show a dramatic dependence on who strikes first, so this concern has no current substantial basis; it is rather an apprehension about some future possible vulnerability that might occur. It would require a vulnerability in our offensive forces of the kind that people have been trying to prevent for years. If the United States, through BMD, provides added protection to its offensive forces, there would seem to be no reason whatever for believing that heavy defenses on both sides would leave us more vulnerable to this type of problem than we are at present.

In fact, such vulnerabilities might be reduced. There are risks in any strategy and our offensive forces of the present and of the recent past may well have had important vulnerabilities, including undetected ones. For example, Senator Henry Jackson, in a Senate speech of last September, discussed electromagnetic-pulse phenomena from nuclear bursts in terms that suggest these phenomena are a continuing source of concern for the security of our offensive forces. A potential vulnerability of a different type might have threatened our land-based missiles if the Soviets had developed multiple, independently targeted reëntry vehicles (MIRV) before we had understood the concept ourselves. In view of these and other vulnerabilities of major or potentially major concern, it is not at all obvious, to say the least, that a posture of pure deterrence, based wholly on offensive forces, would be more secure against unpleasant surprises (such as a suddenly discovered, extreme vulnerability to a first strike) than postures with heavy emphasis on defense. Even if there were some added risk associated with the defensive posture, this would have to be weighed against the possibility that the defense might save 50 or 80 million Americans in the event of a major war, and make possible much more rapid economic recovery.

There is also a line of argument which is sometimes summarized by saying that, in matters of defense, one cannot be "a little bit pregnant"-that is, that public and other pressures for "more" will lead to a defensive arms race. Our experience in both air defense and civil defense is so much against this idea that it should be dismissed without further discussion. Similarly, it is sometimes said that the American public would not tolerate any inequalities in local defense effectiveness. We have had a major air- defense system with substantial inequalities for perhaps 15 years, and scarcely anyone other than experts associated with the system is even sure whether his own home city has a local defense, much less how it compares to those of other cities. The degree of awareness of these matters that has prevailed in the past in the general public is reflected in a public- opinion poll conducted in 1964 which found that two-thirds of the U.S. public believed we had a BMD system all deployed and ready for action.

Space has not permitted here a discussion of several other problem areas sometimes associated with missile defense, such as the implications of an American BMD for the Western Alliance, or the frequently repeated assertion (which I believe is false) that BMD would require as a prerequisite a major expansion of civil-defense programs. Several scholars have, however, studied these problems, and I believe it is fair to say that, on examination, they seem even less likely to prove troublesome than the problems already considered.

It appears to many that achievable defenses could reduce potential American fatalities by many tens of millions, and greatly improve the prospects for postwar economic recovery-and do this without substantially increasing traditional strategic budgets, and without provoking a major new arms race. If the American public and their leaders come to this view, they are most unlikely to judge that pursuit of Assured Vulnerability is a proper objective of the Department of Defense.

The defense rests.

[i] Including three recent articles in Foreign Affairs: J. I. Coffey, "The Anti-Ballistic Missile Debate," April 1967; Robert L. Rothstein, "The ABM, Proliferation and International Stability," April 1968; and Carl Kaysen, "Keeping the Strategic Balance," July 1968. An important exception is Charles M. Herzfeld, "Ballistic Missile Defense and National Security," Annals of the N.Y. Academy of Sciences, November 22, 1965, p. 119-125.

[ii] The main published sources of information for use in considering the prevailing technical estimates are the unclassified "posture statements" issued annually in recent years by the Secretary of Defense, especially those issued by McNamara in 1967 and 1968, although some supplementary information has been contained in Government hearings and speeches.

[iii] Other examples of mistaken reference to exchange ratios are found in the much-cited article in Scientific American, March 1968, by Richard L. Garwin and Hans A. Bethe, and in the Look article by Jerome B. Wiesner, November 28, 1967.

[iv] Herbert Roback, Staff Administrator of the Military Operations Subcommittee, U.S. House of Representatives, in E. P. Wigner, Editor, "Who Speaks for Civil Defense?" New York: Scribner, 1968, p. 95.

[v] Steuart Pittman, former Assistant Secretary of Defense for Civil Defense, "Government and Civil Defense," in Wigner, op. cit.

[vi] Since it will be clear that I am critical of McNamara's handling of BMD, I should mention that I still believe he was the greatest Secretary of Defense the United States has had.

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