A LearnWorld Text

Race to Oblivion

A Participant's View of the Arms Race

Herbert F. York




Table of Contents57Missile-Gap Mania125
INTRODUCTION78The McNamara Era147
Prologue: Eisenhower's Other Warning9PART TWO: UNBALANCING THE BALANCE OF TERROR171
1The Arms Race and I159MIRV: The Multiple Menace173
2The Race Begins: Nuclear Weapons and Overkill2711Other Lessons from the ABM Debate213
3The Bomber Bonanza4912The Ultimate Absurdity228
4The Elusive Nuclear Airplane60A Glossary of Acroyms241
5Rockets and Missiles75Index245




One of the most misleading aspects of the arms race is the notion of defense. The technical dynamics of nuclear strategy have blurred many of the distinctions between offensive and defensive weapons. A great irony of the arms race is that measures taken ostensibly for defensive purposes can seem to the enemy as great a threat as or perhaps even a greater threat than avowedly offensive measures. In this situation, defensive programs such as the air defense system and the ABM have contributed as much as anything else to intensifying arms competition.

By 1960, the United States had spent more than $20 billion on a continental air defense system. The over-all system consisted of many separate components for detecting, tracking, and intercepting incoming bombers.

The northernmost part of the detection and tracking system was known as the DEW line (distant early-warning). It consisted of a line of radars located in Alaska and Canada, with naval vessels providing seaward extensions on both ends. Other radars were located farther south in Canada and within the United States itself. The interceptors included a variety of aircraft, a cruise missile (the Air Force's Bomarc) and two short range ballistic missiles (the Army's Nike-Ajax and Nike-Hercules). Each of these interceptors required its own special


command and control system for getting it to the appropriate point in space at the right time to intercept or "kill" the attacking bomber. These individual command and control systems consisted of additional radars, computers, and special communications networks.

Clearly, such a complex system, consisting as it did of so many different components, would work better if the whole thing were tied together and operated as a unit. A system for doing so was designed and developed during the fifties; it was called SAGE, for Semi-Automatic Ground Environment. SAGE was to consist of about a dozen command centers located around the periphery of the United States. Each of them was to be equipped with a computer suitable for directing and coordinating the entire air defense battle in its zone. Data from the various lines of radar detectors were to be fed into the appropriate computers. The computers were to decide which of the various interceptors would be assigned to a particular attacker, and the necessary orders and detailed instructions for making the intercept were to be sent out to the weapons in the field. These peripheral command centers were in turn to be tied in to one central command post located in Colorado Springs, headquarters of NORAD, the North American Air Defense Command. The name SAGE eventually came to be commonly used to describe the over-all continental air defense system.

As time went on, it turned out to be harder to design such a system than had been anticipated. Serious doubts arose about the ability of the system to cope with the threat as originally perceived. The usual programmatic delays resulted, and before SAGE could be placed in operation ICBMs passed from theory to reality.

The whole matter of continental defense had to be restudied. SAGE was found to be totally vulnerable to a combined bomber and missile attack. It became painfully evident that SAGE would be obsolete before it could be operating. At the


same time, technical difficulties were experienced in the development of other components of the over-all continental air defense system, and unanticipated political problems arose over the deployment of nuclear armed interceptors in Canada. Together these factors in 1959 resulted in a substantial cutback in our plans for the deployment of both SAGE and the associated weapons systems. Further cutbacks were made in the first years of the Kennedy administration. As a result, the air defense system we now have is only a shadow of what we once had in mind. Even so, we spent over $1 billion a year on it during the sixties.

Each of the individual components of this over-all system works quite well in the test-range sense. That is, on a missile test range test crews can detect a suitable drone target with their radars and can then intercept and destroy it with a Bomarc or Nike-Hercules or other weapon. Even so, experts generally agree that the present over-all system would not produce any important level of attrition in the event of a real attack on it. Nor would the original system be capable of doing so had it been built.

The same conclusions apply to the attempts of the U.S.S.R. to build an air defense. The Russians have always been more defense- minded than we have. As a result, not only have they spent a greater fraction of their strategic-arms budget on defense systems, they have invested much more money in defensive systems even on an absolute basis. The best estimates say that the Soviets have spent to date (1970) about $75 billion, in contrast to our $30 billion. Even so, our Strategic Air Command believes that it can penetrate Soviet air defenses with a sufficiently high probability to carry out its mission. Almost all independent authorities agree with this estimate.

The problem of air defense is entirely different in the nuclear age from what it was before. In World War II the damage from the amount of chemical high explosives that a bomber could carry on one trip was so small that very many


repetitive trips had to be made by each bomber and each crew in order to attain a significant total level of damage. An attrition rate of only ten percent or so by the defenses was sufficient to prevent the necessary number of repetitive strikes or sorties. Now the damage that a single bomber armed with a hydrogen bomb can do is so enormous that an attrition of ninety percent, allowing a penetration of only ten percent, is wholly inadequate as a defense. For example, a "mere" ten percent of our SAC bomber forces alone could deliver more than a hundred thousand times as much explosive power as killed those one hundred thousand human beings in Hiroshima.

The Soviet air defense equipment deployed in North Vietnam during the late sixties was not able to achieve anything like the ten percent necessary for effective defense against bombers carrying chemical explosives. The home defenses of the U.S.S.R. are, of course, much more extensive and consist of more modern and advanced types of equipment than the surface-to-air missiles (SAM) and antiaircraft artillery deployed in North Vietnam. However, I know of no estimate that says they could achieve the ninety-five-percent-plus level of attrition necessary against a massive sophisticated nuclear attack. On the contrary, estimates by persons in whom I have confidence range from pessimistic (for us) estimates of forty percent attrition of our SAC bombers by Soviet air defenses down to more optimistic estimates of only ten percent attrition. My personal view is that the optimists are probably more nearly right in this case.

If we have not bought an effective air defense for our $30 billion and they have not bought an effective air defense for their $75 billion, what have we each done? One answer is that each of us has prevented the other from having a free ride. That is, if we had no air defense, then presumably they could attack us with simple and cheap bombers, and vice versa. This is not strictly true, however, since the need for speed and great range would still be there even in the absence


of any defense' and, furthermore, the possibility of at least some modest defense capability would have to be anticipated. But, even so, their air defense has forced us to spend much more money on our bomber fleet than we otherwise would have. We have bought many more aircraft and each one is more expensive because of the demand for high performance and the need to provide it with various kinds of penetration aids such as radar jammers and chaff. We must suppose the same has been true for the Soviets. Thus, one result of air defenses has been to induce each side to spend more money on bombers.

But that isn't all. The possibility of successful defense against bombers was one of the principal driving forces behind the development of missiles. Anyone who took seriously the articles and especially the advertisements in the missile press during the fifties could only conclude that not even a fly could penetrate the borders of the United States without proper clearance. Ballistic-missile warheads, with their greater speed and smaller size, then as now appeared to be much less vulnerable than bombers to defensive measures and hence more certain of being able to deliver their payload to their assigned targets. There were, of course, other rationales and arguments for the development of long-range ballistic missiles, but leapfrogging defensive technology was perhaps the most important on both sides.

Thus, the net result of the development and deployment of bomber defenses by each of the superpowers has been to force the other to increase greatly the number of individual weapons in its offensive forces and to multiply the number of different kinds of forces making up its strategic offense. We have here before us another simple, clear example of how it is that purely defensive weapons accelerate the arms race and how it is that each turn around the arms-race spiral ends up with more human lives at risk than before. In the mad logic of the arms race, bombers led inexorably to bomber defenses, and these


in turn led to fancier, more complex, and more deadly bombers as well as to entirely new kinds of offense, the intercontinental ballistic missile and the submarine-launched ballistic missile.

It became evident, in the early fifties, that the state of the art of nuclear weaponry, rocket propulsion, and missile guidance and control had reached the point in the U. S. where a strategically useful ICBM could be built. At about the same time, the fact that a major long-range-missile development program was in progress in the U.S.S.R. was confirmed. As a result of the confluence of these two events the tremendous U. S. long- range-missile program, which dominated the technological scene for more than a decade, was undertaken.

Also at about the same time, the Army, which had had the responsibility for ground-based air defense (including the Nike-Ajax and Nike-Hercules surface-to-air missiles, or SAMs) , began to study the problem of how to intercept ICBMs, and soon afterward (in 1956) it initiated the Nike-Zeus program. This program was a straightforward attempt to use existing technology in the design of a nuclear-armed rocket for the purpose of intercepting a single uncomplicated incoming warhead. The Air Force, as is its style, proposed more exotic solutions to the missile- defense problem, but these were subsequently either dropped or absorbed into the Defender program of the Department of Defense's Advanced Research Projects Agency (ARPA). The Defender program included the study of designs more advanced than Nike-Zeus, and it also incorporated a program of down-range measurements designed to find out what really went on during the terminal phases of missile flight.

By 1960, indications that the Russians were taking the ABM prospect seriously, in addition to progress in our own Nike-Zeus program, stimulated the designers of our offensive missiles into seriously studying the problem of how to penetrate missile defenses. Very quickly a host of "penetration-aid" concepts came to light: light and heavy decoys, including bal-


loons, tank fragments and objects resembling children's jacks; electronic countermeasures, including radar-reflecting clouds of the small wires called chaff; radar blackout by means of high-altitude nuclear explosions; tactics such as barrage, local exhaustion and "rollback" of the defense; and MRV (Multiple Reentry Vehicles). These last were good only against large-area targets (cities), but MRV soon developed into MIRV (Multiple Independent Reentry Vehicles), which eventually will be useful against smaller, harder targets such as missile silos, radars and command centers.

This avalanche of concepts forced the ABM designers to go back to the drawing board, and as a result the Nike-X concept was born in 1962. The Nike- X designers attempted to make use of the more sophisticated and up-to- date technology developed under the Defender program in the design of a system that they hoped might be able to cope with a large, sophisticated attack. All through the mid-1960s a vigorous battle of defensive concepts and designs versus offensive concepts and designs took place. This battle was waged partly on the Atlantic and Pacific missile ranges but mostly on paper and in committee meetings. It took place generally in secret, although parts of it were discussed in the open literature and before Congressional committees.

This intellectual battle culminated in a meeting that took place in the White House in January, 1967. In addition to President Johnson, Secretary of Defense Robert S. McNamara and the Joint Chiefs of Staff, there were present all past and current Special Assistants to the President for Science and Technology (James R. Killian, Jr., George B. Kistiakowsky, Jerome B. Wiesner and Donald F. Hornig) and all past and current Directors of Defense Research and Engineering (Harold Brown, John S. Foster, Jr., and myself). We were asked that simple kind of question which must be answered after all the complicated ifs, ends and buts have been discussed: "Will it work and should it be deployed?" The answer in relation to defending our people against a Soviet missile attack was


No, and there was no dissent from that answer. The context, of course, was the Russian threat as it was then interpreted and forecast, and the current and projected state of our ABM technology. There was also some discussion of this same question in relation to a hypothetical Red Chinese missile threat. In this latter case, there was some divergence of views, although the majority view (and my own) was still No.

Later that year, Secretary McNamara gave his famous San Francisco speech in which he reiterated his belief that we could not build an ABM system capable of protecting us from destruction in the event of a Russian attack. He did state, however, that the decision had been made (I presume by the President) to build an ABM system able to cope with a hypothetical Chinese missile attack, which by definition would be "light" and uncomplicated. In announcing that we would go ahead with a program to build what came to be known as the Sentinel system, he said, "There are marginal grounds for concluding that a light deployment of U. S. ABMs against this probability is prudent.,' A few sentences later, however, he warned, "The danger in deploying this relatively light and reliable Chinese-oriented ABM system is going to be that pressures will develop to expand it into a heavy Soviet-oriented ABM system." The record makes it clear that he was quite right in this prediction.

The Congress, in the course of authorizing and appropriating funds for fiscal year 1969 (beginning July 1, 1968), approved deployment of the Sentinel system without much debate, and soon afterward the Army started to acquire property for use as missile sites near Boston, Chicago, and Seattle. It sent teams out to those areas to win local support, but instead the citizenry became alarmed and began to object strenuously to having their neighborhoods turned into what they thought would be prime targets. Many people wrote their Representatives and Senators about their views, and some of these in turn began to give more serious attention to the whole matter.

Meanwhile, Richard Nixon was elected and inaugurated


President of the United States. Some of his technical advisers had for some time strongly favored deploying an ABM system, but an ABM system of a different sort and for reasons different from those used to justify Sentinel. They held that defending our deterrent against the Soviets should have a higher priority than defending our cities against a hypothetical Chinese threat.

Thus, in the early months of 1969, both the Congress and the Administration were reconsidering the Sentinel decision, and many informed (and uninformed) private citizens, including myself, seized the opportunity to make their views known. The Subcommittee on International Organizations and Disarmament (chaired by Senator Albert Gore) of the Senate Foreign Relations Committee and a similar subcommittee of the house (chaired by Representative Clement J. Zablocki) held well-publicized open hearings on the subject.

The new Defense Secretary, former Representative Melvin R. Laird, and the new Deputy Secretary, David Packard, also reviewed the question of whether to deploy Sentinel. They were assisted in this review by John Foster, the Director of Defense Research and Engineering, and Stanley Resor, the Secretary of the Army. Both of these men had served in the same positions under Robert McNamara and Clark Clifford. Similarly involved in this review were two of the abler general officers in the United States Army: Alfred D. Starbird, in charge of the Sentinel program, and Austin R. Betts, the Chief of Army Research and Development I knew both of these officers well; I had worked closely with Starbird when he was Director of Military Applications for the AEC and I was director of the Livermore Laboratory; I had worked closely with Betts when he was Director of ARPA and I was Director of Defense Research and Engineering.

Also, as usual, new intelligence information was continuing to come in. It was studied and collated by those responsible for doings so, and was fed into the review process.


On March 14, while the Senate hearings were still in progress, President Nixon announced his decision in the matter He said he had considered three possible courses: going ahead with the Sentinel system as proposed at the end of the Johnson administration, not going ahead with any ABM at all, or going ahead with a modified system. He chose the last of these, and he said that he wanted a system that would protect a part of our Minuteman force in order to increase the credibility of our deterrent, and that he had overruled moving in the direction of a massive city defense because "even starting with a thin system and then going to a heavy system tends to be more provocative in terms of making credible a first-strike capability against the Soviet Union. I want no provocation which might deter arms talks." (At that time, governments on both sides of the Atlantic were looking forward to strategic-arms-limitation talks in the summer of 1969.)

The top civilian defense officials gave this same rationale, although they put a little more emphasis on the "prototype" and "growth-potential" aspects of the system. For simplicity and clarity I shall focus on the Administration's proposal as presented in open session by responsible officials.

This new ABM soon became widely known as the "Safeguard system" for its primary purpose of defending or safeguarding a part of our deterrent force.

The Senate hearings, which began a little over a week before the President announced his decision, continued into the summer. The majority of those who testified against deployment of an ABM system were persons who had previously held government positions of major responsibility and authority. This group included all of the former Special Assistants to the President for Science and Technology: Killian and Kistiakowsky, who served under Eisenhower; Wiesner, who served under Kennedy; and Hornig, who served under Johnson. It also included Ruina, a former Director of ARPA; Kaysen, an assistant in the White House under Kennedy; George W. Rath-


jens, a former Chief Scientist of ARPA; and myself. Two other scientists who testified against deployment, Hans Bethe and W. K. H. Panofsky, had been members of the President's Science Advisory Committee Bethe during the Eisenhower administration and Panofsky during the Kennedy and Johnson administrations. A law professor, Abram Chayes, and two political scientists, Marshall Shulman and Allen Whiting, also testified negatively.

The majority of those who testified for deployment (excluding then current government witnesses) were, curiously, theoretical scientists and mathematicians. This group included William McMillan, Richard Latter, Donald Brennan, Frederick Seitz, Albert Wohlstetter, Freeman Dyson, and Edward Teller. In addition, four former high government officials testified in favor of deployment. Two of these had primarily political backgrounds: Paul Nitze, a former Deputy Secretary of Defense, and Dean Acheson, a former Secretary of State. The other two had primarily technical backgrounds: Charles Herzfeld, another former Director of ARPA, and Daniel Fink, a former Deputy DDRE. One then current member of PSAC, Gordon MacDonald, testified in favor of deployment.

The Senate vote on the ABM came late in July. In order to vote against it, a Senator had to vote for one of two amendments to the Defense Authorization Bill. An amendment offered by Margaret Chase Smith, Republican of Maine, lost on a tie, 50-50 and an amendment by John Sherman Cooper, Republican of Kentucky, lost 49-51.

The Senate opposition to the ABM had been bipartisan. Among Democrats, Fulbright, Gore, Hart, Kennedy, and Symington led the opposition. On the Republican side the leaders were Cooper, Case, and Percy.

The ABM had been widely discussed on other occasions also. On March 4, meetings were held in a number of major academic and research institutions all over the country for the purpose of discussing the proper relationship between such in-


stitutions and "war research.". Much of these discussions focused on the ABM. At M.I.T. where the idea for the meetings originated, Hans Bethe spoke before a multitude of students and faculty. He started by saying, "You're here because you're against the ABM. I'm here to tell you why you are." The audience was, happily, able to see the joke and laugh at itself. (A few days later, I spoke about the same matter to a large interested group at the Salk Institute in La Jolla.)

Later, as the Senate hearings warmed up, city clubs, local foreign- relations councils, and the like started to sponsor speakers and debates. National as well as local television gave prominent coverage to the subject in their news programs and ran news specials in order to go into greater detail. America had finally become concerned about what the generals and the technologists were doing in her name.

The Administration's spokesmen based their case mainly on the continuing deployment of the very large Soviet ICBM we call the SS-9. They described this ICBM as being able to deliver a twenty-megaton warhead, and they claimed that a Soviet program for developing a MIRV payload capable of delivering three five-megaton warheads was in progress. They further asserted that by the mid-seventies the delivery accuracy of these three individual warheads could be such as to result in a kill probability of ninety-six percent against a Minuteman silo with its current hardness of three hundred pounds per square inch. In effect, this means they were predicting a CEP of about a quarter of a mile. It also means that the Soviets would have to have a capability for replacing each of their missiles that failed during launch, thus completely making up for any residual unreliability in the system. They estimated that 150 SS-9s were then deployed and they projected a growth rate in this figure of about forty per year.

Putting together the projections of numbers and the assumptions about future accuracy, reliability, and ability to replace failed missiles, the Defense Department spokesmen


calculated that a preemptive attack in 1975 or 1976 by the Soviet SS-9 force on the U. S. Minuteman force could result in the destruction of all but forty Minutemen. And this number, they asserted, was so small that the ABM system the Soviets could have installed by that time might be able to cope with it. All this long train of estimates, assumptions, and possibilities led Secretary Laird to conclude that the only logical explanation for the SS-9 buildup was that the Soviets were "going for a first-strike capability.', The logical response on our part, he argued, was to install an ABM system to defend at least a part of our deterrent so as to deny them such a first-strike capability. One must admit that almost anything is conceivable as far as intentions are concerned, but there certainly are simpler and, it seems to me, more likely explanations.

Probably the simplest of all was apparent in a chart presented before Senator Gore's subcommittee by Deputy Secretary Packard. It showed clearly that until about the beginning of 1967 the Russians had been quite content to be a very poor second in the missile race. Up to that time, we enjoyed a three- or fourfold numerical advantage over them in ICBMs. This is made more puzzling by the fact that all during this period U.S. defense officials found it necessary to boast about how far ahead we were, in order to be able to resist internal pressures for still greater expansion of our offensive forces. It is not unlikely, therefore, that the Russians simply decided to do something dramatic about this large imbalance.

Another possible reason for the deployment of the SS-9, and one that I believe added to the other in the minds of the Russian planners, was that their strategists concluded in the mid-1960s, that, whatever the top officials here might be saying at the time, certain elements would eventually succeed in getting a large-scale ABM system built, and that penetration- aid devices, including multiple warheads, would be needed to meet the challenge. If they did make such a hypothetical prediction, they were, unfortunately, correct. Let us, however.


pass on from this question of someone else's intentions and consider whether or not the proposed Safeguard ABM system is a valid, rational, and necessary response to the Russian deployments and developments outlined above.

In order to do so, it will be well to have in mind a picture of the over-all balance in strategic forces. The main components of these forces are presented in a table given below. This table is adapted from Defense Department briefings given during the 1970 ABM debates. The small land- based missiles are Minutemen in the U. S. case and mostly the roughly equivalent Sells in the Soviet case. The large missiles are Titan IIs in the U. S. case, and the still larger SS-9s in the Soviet case. Not covered are a number of relatively less important "extras" such as the Soviet FOBS (fractional orbital bombardment system), the IRBMs, and the very many U. S. aircraft deployed on carriers, in Europe, and elsewhere.

In the area of defensive systems designed to cope with the offensive systems just outlined, both the U. S. and the U.S.S.R. have defenses against bombers that would be adequate against a prolonged attack using chemical explosives (where ten percent attrition is enough) but certainly inadequate against a nuclear attack (where ten percent penetration is enough). In addition, the U.S.S.R. has an ABM system, much like our old Nike-Zeus system, deployed around Moscow. It appears to

U.S. U.S.S.R.
Intercontinental ballistic missiles in hard silos: large 54 about 250
small 1,000 about 850
Submarine-launched ballistic missiles 656 about 110
Heavy bombers 581 about 140


have virtually no capability against our offensive-missile systems, and it is usually estimated as consisting of less than one hundred antimissile missiles. As we have seen, it was the stimulus, though not the provocation, behind our MIRV development, and thus it has been extremely counterproductive in its goal of defending Moscow.

From a technical point of view and as far as components are concerned, President Nixon's Safeguard system of today is very little different from President Johnson's Sentinel system. There are only minor changes in location of certain components (away from cities), and elements have been added to some of the radars so that they can now observe submarine- launched missiles coming from directions other than directly from the U.S.S.R. and China. As before, the system consists of a long-range interceptor carrying a large nuclear weapon (Spartan), a fast short-range interceptor carrying a nuclear weapon (Sprint), two types of radar (perimeter acquisition radar, or PAR, and missile-site radar, or MSR), a computer for directing the battle, and a command and control system for integrating Safeguard with the national command. I shall not describe the equipment in detail at this point, but shall pass on directly to what I believe can be concluded from the hearings and other public sources about four major questions.

1. Assuming that Safeguard could protect Minuteman, is it needed to protect our deterrent?

Perhaps the clearest explanation of why the answer to this first question is No was given by Wolfgang K. H. Panofsky before the Senate Armed Services Committee in April, 1969. He described how the deterrent consists of three main components: Polaris submarines, bombers, and land- based ICBMs. Each of these components alone is capable of delivering far more warheads than is actually needed for deterrence, and each is currently defended against surprise destruction in a quite different way. ICBMs are in hard silos and are numerous. Polarises are hidden in the seas. Bombers can be placed on various levels of alert and can be dispersed.


Since the warning time in the case of an ICBM attack in the mid- seventies is generally taken as being about thirty minutes, the people who believe that the deterrent may be in serious danger usually imagine that the bombers are attacked by missile submarines and therefore have a warning of only fifteen minutes or less. This is important, because a thirty- minute warning gives the bombers ample time to get off the ground and even fifteen minutes allows about half to get away, according to a report by Clark Clifford when he was Secretary of Defense In that case, however, an attack on all three components cannot be made simultaneously; that is, if the attacking weapons are launched simultaneously.

Thus, it is incredible that all three of our deterrent systems could become vulnerable in the same time period, and it is doubly incredible that this could happen without our having sufficient notice to do something about it. There was therefore no basis for a frantic reaction to the hypothetical Russian threat to Minuteman. Still, it was sensible and prudent to begin thinking about the problem, and so we turn to the other questions. We must consider these questions in the technological framework of the mid-1970s, and we shall do this now in the way which defense officials currently seem to favor: by assuming that this is the best of all possible technological worlds, that everything works as intended, and that direct extrapolations of current capabilities are valid.

2. Assuming that Safeguard works, can it in fact safeguard Minuteman?

One good approach to this problem is the one used by George W. Rathjens in his testimony before the Senate Armed Services Committee on April 23, 1969. His analysis took as a basis of calculation the implication in Secretary Laird's early testimony that the Minuteman force may become seriously imperiled in the mid-1970s. Rathjens then estimated how many SS-9s would have to be deployed at that time in order to achieve this result. From this number and the estimate of


the current number of SS-9s deployed he got a rate of deployment. He also had to make an assumption about how many Sprints and Spartans would be deployed at that time, and his estimates were based on the first phase of Safeguard deployment. These last numbers had not been released, but a range of reasonable values could be guessed from the cost estimate given. Over a fairly wide range of such guesses, Rathjens found that by prolonging the SS-9 production program by a few months the Russians would be able to cope with Safeguard by simply exhausting it and would still have enough warheads left to imperil Minuteman, if that is indeed their intention.

The length of this short safe period does depend on the numbers used in such calculations, and they, of course, can be disputed to a degree. Thus, if one assumes that it takes fewer Russian warheads to imperil Minuteman (it later became evident that defense officials were assuming one for one), then the assumed deployment rate is lower and the safe period is lengthened; on the other hand, if one notes that the missile-site radars in our system are one-tenth as hard as even today's silos, then the first attacking warheads, fired directly at the radars, can be smaller and less accurate, so that a higher degree of MIRVing can be used for attacking these radars and a shorter safe period results. To go further, it was suggested that the accuracy-yield combination of the more numerous SS­11s would be sufficient for attacking the missile-site radars, and therefore, if the Russians were to elect such an option, there would be no safe period at all. In short, the most that Safeguard can do is either delay somewhat the date when Minuteman would be imperiled or cause the attacker to build up his forces at a somewhat higher rate if, indeed, imperiling Minuteman by a fixed date is his purpose.

3. Will it work?

By this question I mean: Will operational units be able to intercept enemy warheads accompanied by enemy penetration


aids, in an atmosphere of total astonishment and uncertainty? I do not mean: Will test equipment and test crews intercept U. S. warheads accompanied by U. S. penetration aids in a contrived atmosphere? A positive answer to the latter question is a necessary condition for obtaining a positive answer to the former, but it is by no stretch of the imagination a sufficient condition.

This basic question has been attacked from two quite different angles: by examining historical analogies and by examining the technical elements of the problem in detail. Design-oriented people who consider this a purely technical question emphasize the second approach. I believe the question is by no means a purely technical question, and I suggest that the historical-analogy approach is more promising, albeit much more difficult to use correctly.

False analogies are common in this argument. We find that some people say, "You can't tell me that if we can put a man on the moon we can't build an ABM." Others say, "That's what Oppenheimer told us about the hydrogen bomb." These two statements contain the same basic error. They are examples of successes in a contest between technology and nature, whereas the ABM issue involves a contest between two technologies; offensive weapons and penetration aids versus defensive weapons and discrimination techniques. These analogies would be more pertinent if, in the first case, someone were to jerk the moon away just before the astronauts landed, or if, in the second case, nature were to keep changing the nuclear-reaction probabilities all during the development of the hydrogen bomb and once again after it was deployed.

Proper historical analogies should involve modern high-technology defense systems that have actually been installed and used in combat. If one examines the record of such systems, one finds that they do often produce some attrition of the offense, but not nearly enough to be of use against a nuclear attack. The most up-to-date example is provided by the


Russian SAMs and other air defense equipment deployed in North Vietnam. This system worked after a fashion because both the equipment designers and the operating crews had plenty of opportunities to practice against real U.S. targets equipped with real U.S. tactics.

The best example of a U.S. system is somewhat older, but I believe it is still relevant. It is the SAGE system, which I discussed at the beginning of this chapter. To quote from the recent book on the ABM decision prepared by Wiesner, Abram Chayes, and others: "Still, after fifteen years, and the expenditure of more than $20 billion, it is generally conceded that we do not have a significant capability to defend ourselves against a well-planned air attack. The Soviet Union, after even greater effort, has probably not done much better." And in 1970, Secretary Laird said, "The current bomber defense system, as the Congress is well aware, has a limited combat effectiveness and is expensive to operate.... We plan to continue the phase-down of the ... system...."

So much for analogies; let us turn to the Safeguard system itself. Doubts about its being able to work were raised during the public hearings on a variety of grounds, some of which are as follows:

First, and perhaps foremost, there is the remarkable fact that the new Safeguard system and the old Sentinel system use virtually the same hardware, deployed in a very similar manner, and yet they have entirely different primary purposes. Sentinel had as its purpose defending large soft targets against the so-called Chinese threat. The Chinese threat by definition involved virtually no sophisticated penetration aids and no possibilities of exhausting the defense; thus were "solved" two of the most difficult problems that had eliminated Nike-Zeus and Nike-X.

Safeguard has as its primary purpose defending a part of the Minuteman force against a Russian attack. It is not credible that a Russian attack against the part of the Minuteman force so defended would be other than massive and sophisti-


cated, so we are virtually right back to trying to do what in 1967 we said we could not do, and we are trying to do it with no real change in the missiles or the radars. It is true that defending hard points is to a degree easier than defending cities, because interception can be accomplished later and at lower altitudes, thus giving discrimination techniques more time to work. Moreover, only those objects headed for specific small areas must be intercepted. These factors do make the problem easier, but they do not ensure its solution, and plenty of room for doubt remains.

Second, there is the contest between penetration aids and discrimination techniques. This was discussed by Richard Garwin and Hans Bethe in their March, 1968, article in Scientific American and was mentioned also in varying degrees of detail by many of those who testified in the spring of 1969 concerning the ABM issue. The Russian physicist Andrei D. Sakharov, in his essay Thoughts on Progress, Coexistence and Intellectual Freedom, put the issue this way:

Improvements in the resistance of warheads to shock waves and the radiation effects of neutron and X-ray exposure, the possibility of mass use of relatively light and inexpensive decoys that are virtually indistinguishable from warheads and exhaust the capabilities of an antimissile defense system, a perfection of tactics of massed and concentrated attacks, in time and space, that overstrain the defense detection centers, the use of orbital and fractional-orbital attacks, the use of active and passive jamming and other methods not disclosed in the press--all of this has created technical and economic obstacles to an effective missile defense that, at the present time, are virtually insurmountable.

I would add only MIRV to Sakharov's list. Pitted against this plethora of penetration aids are various observational


methods designed to discriminate the real warheads. Some of the penetration devices obviously work only at high altitudes, but even these make it necessary for the final "sorting" to be delayed, and thus they still contribute to making the defense problem harder. Other devices can continue to confuse the defense even down to low altitudes. Some of the problems the offense presents to the defense can no doubt be solved (and have been solved) when considered separately and in isolation. That is, they can be solved for a time, until the offense designers react. One must have serious reservations, however, whether these problems can ever be solved for any long period in the complex combinations that even a modestly sophisticated attacker can present. Further, such a contest could result in a catastrophic failure of the system in which all or nearly all interceptions fail.

Third, there is the unquantifiable difference between the test range and the real world. The extraordinary efforts of the Air Force to test operationally deployed Minutemen in the late sixties show that it too regards this as an important problem. Moreover, the results of those tests did reveal important weaknesses in the deployed forces. The problem has many aspects: the possible differences between test equipment and deployed equipment; the certain differences between the offensive warheads and penetration aids supplied by us as test targets and the corresponding equipment and tactics the defense must ultimately be prepared to face; the differences between the installation crews at a test site and at a deployment site; the differences in attitudes and motivation between a test crew and an operational crew (even if it is composed of the same men); the differences between men and equipment that have recently been made ready and that everyone is watching and men and equipment that have been standing ready for years during which nothing happened; the differences between the emotional atmosphere where everyone knows it's not "for real" and the emotional atmosphere where no one can believe what


he has just been told. It may be that all that enormously complex equipment will be ready to work the very first time it must "for real," and it may be that all those thousands of human beings have performed all their interlocking assignments correctly, but I have very substantial doubts about it.

Fourth, there is the closely related "hair-trigger-stiff-trigger" contradiction. Any active defense system such as Safeguard must sit in readiness for two or four or eight years and then fire at precisely the correct second following a warning time of only minutes. Furthermore, the precision needed for the firing time is so fine that machines must be used to choose the exact instant of firing no matter how the decision to fire is made. In the case of offensive missiles the situation is different in an essential way: Although the maintaining of readiness throughout a long, indefinite period is necessary, the moment of firing is not so precisely controlled in general, and hence human decision-makers, including even those at high levels, may readily be permitted to play a part in the decision- making process. Thus if we wish to be certain that the defense will respond under conditions of surprise, the trigger of the ABM system, unlike the triggers of the ICBMs and Polarises, must be continuously sensitive and ready in short, a hair trigger for indefinitely long periods of time. (This distinction applies only to the past and present. In the future, some offensive systems will also require "hair triggers.")

On the other hand, it is obvious that we cannot afford to have an ABM missile fire by mistake or in response to a false alarm. Indeed, in 1968 the Army went to some pains to assure residents of areas near proposed Sentinel sites that it was imposing requirements to ensure against the accidental launching of the missile and the subsequent detonation of the nuclear warhead it carries. Moreover, Army officials assured the public that no ABM missiles would ever be launched without the specific approval of "the highest authorities."

These two requirements a hair trigger so that the system


can cope with a surprise attack, and a stiff trigger so that it will never go off accidentally or without proper authorization

are, I believe, contradictory ones. If the system cannot be fired without approval of "the highest authorities," then the probability of its being fired under conditions of surprise is less than it would be otherwise. This probability depends to a degree on the highly classified technical details of the command and control system, but in the last analysis it depends more on the fact that "the highest authority" is a human being and therefore subject to all the failures and foibles pertaining thereto.

This brings us to the principal question: Anyway, what harm can it do?

We have just found that the total deterrent is very probably not in peril, that the Safeguard system as described by the Nixon Administration cannot safeguard Minuteman even if it "works," that there is, to say the least, considerable uncertainty whether or not it will "work." Nonetheless, if there were no harm in it, we might be prudent and follow the basic motto of the arms race: "Let us err on the side of military safety." There seem to be many answers to the question of what harm it would do to build an ABM system. First of all, such a system would cost large sums of money needed for nondefense purposes. Second, it would divert money and attention from what may be better military solutions to the strategic problems posed by the Administration. Third, it would intensify the arms race. All these considerations were discussed at the 1969 hearings: I shall comment here only on the third, the arms-race implications of the ABM decision.

It is often said that an ABM system is not an accelerating element in the arms race, because it is intrinsically defensive. For example, during the 1969 hearings Senator Henry M. Jackson of Washington, surely one of the best-informed Senators in this field, said essentially that, and he quoted Premier Kosygin as having said the same thing. I believe such a notion


is in error and is based on what we may call "the fallacy of the last move."

I believe that in the real world of constant change in both the technology and the deployed numbers of all kinds of strategic-weapons systems, ABM systems are accelerating elements in the arms race. At the beginning of this decade we began to hear about a possible Russian ABM system, and we became concerned about its potential effects on our ICBM and Polaris systems. In response the MIRV concept was invented. Today there are additional justifications for MIRV besides penetration, but that is how it started. Now the possibility of a Russian MIRV is used as one of the main arguments in support of the Safeguard system. Thus, we have come one full turn around the arms-race spiral. No one in 1960 and 1961 thought through the potential destabilizing effects of multiple warheads, and certainly no one predicted or even could have predicted that the inexorable logic of the arms race would carry us directly from Russian talk in 1960 about defending Moscow against missiles to a requirement for hard-point defense of offensive-missile sites in the U. S. in 1969.

By the same token I am sure the Russians did not foresee the large increase in deployed U. S. warheads that will ultimately result from their ABM-missile deployment and that made it so counterproductive. Similarly, no one today can describe in detail the chain reaction the Safeguard deployment would lead to, but it is easy to see the seeds of a future acceleration of the arms race in the Safeguard deployment. Russian offense planners are going to look at it and say something such as: "It may not work, but we must be prudent and assume it will." They may then plan further deployments or more complex penetration systems, or maybe they will go to more dangerous systems such as bombs in orbit. A little later, when some of our optimistic statements about how it will do the job it is supposed to do have become part of history, our strategic planners are going to look at Safeguard and say


something such as: "Maybe it will work as they said, but we must be prudent and assume it will not, and, besides, now look at what the Russians are doing."

This approach to strategic thinking, known in the trade as "worst- case analysis," leads to a completely hopeless situation in which there is no possibility of achieving a state of affairs that both sides would consider as constituting parity. Moreover, our defense establishment, and presumably theirs too, will not even consider arms-control measures except, as an absolute minimum, under conditions that it considers to constitute parity. Thus, there is no possibility of stopping the arms race except by political action outside the two defense establishments.

I also mentioned in my own testimony at the ABM hearings that "we may further expect deployment of these ABM systems to lead to the persistent query 'But how do you know it really works?' and thus to increase the pressures against the current limited nuclear-test ban as well as to work against amplifying it." I mentioned this then, and I mention it again now, in the hope that it will become a self-defeating prediction.

Nuclear Designs: Great Britain, France, and China in the Global Governance of Nuclear Arms
[Transaction Publishers, 1996]

British SSBNs

French SNLEs
Other Sites
"Comprehensive Test Ban" [28 February 1996] and a 21 June 1996 addendum on China's CTB policy. The Acheson-Lilienthal Report [16 March 1946]: Report on the International Control of Atomic Energy. Re CTB

LearnWorld Explained FAQ Bosnia War Theory Templates

Our Home Page Our Home

Please direct comments and corrections by email to Bruce D. Larkin, who has prepared this page.