Foreign Relations of the United States, 1958–1960, Volume III, National Security Policy; Arms Control and Disarmament, Microfiche Supplement

145. Presentation by McMillan to Eisenhower¹

Washington, March 4, 1959

An Analysis of Technical Factors in the Strategic Posture of the United States—1956–64

What I have to present is an analysis—illustrated with charts—relating to our strategic posture over the next few years.

The particular material we will see today has developed out of a study by the Panel of the Science Advisory Committee concerned with anti-ballistic missiles, a study of the technical possibilities for defending our striking force. I think you will see, however, that what we have to say falls into the pattern of earlier conclusions of the Science Advisory Committee about the influence of continuing technological change upon our strategic position, and about the steps we must take to keep technically abreast of this change.

In looking at ways of defending the striking force, the anti-missile Panel was forced to consider two related, but nevertheless distinct, questions. We have organized our presentation around these two questions.

The first question is: what array of targets does the continental United States present to the Soviet attacker, and how much force would the attacker need to destroy or neutralize these targets? This is a natural question to ask in examining the defense of the striking force. An answer to the question would show the extent to which defensive measures might be useful in increasing the enemy’s force requirements against us. Indeed, one would like to find, or to bring about by defensive measures, that the attack the enemy would need to destroy our striking force is much larger than any attack he could possibly mount. In such a case, one could be fairly sure, without further study, that an enemy aware of the circumstances would not attempt an attack except under extreme provocation.

Should it appear, in looking at force requirements, that the enemy might possibly have a force large enough for a completely destructive attack against our strategic bases, then one must probe more deeply into the situation. This possibility raises the second question: what retaliatory force can we muster even if our bases are attacked with the enemy’s best capabilities? A later part of our discussion is devoted to this second question.

For the purpose of our discussion, it has been necessary to prepare charts illustrating both some facts and some estimates we have made. It is in the nature of such graphic presentations that they tend to make all data look like fact—even mere estimates—by emphasizing exact numerical values. We ask you to bear with us in this, because we do not wish to present our charts as exactly descriptive of the strategic position of the United States. Rather, we wish to use them as a basis for discussing some uncertainties in that strategic position. Perhaps our main message is that, because of changing technical factors, the strategic position of the United States is now subject to a range of uncertainty and that this uncertainty can lead to a condition of instability internationally. We hope we can also show that there are some relatively simple things that this country can do to reduce uncertainty and increase stability.

We can turn now to the first of our two questions—how much force would the Soviets need to destroy as completely as possible the ability of the continental United States to mount retaliation? The first chart sets a scale by picturing something of the range covered by recent and current intelligence estimates. Specifically, the numbers on the vertical scale represent operational intercontinental ballistic missiles—missiles produced, deployed, and supported by trained troops with the necessary logistics. The dots show specific capabilities discussed in recent intelligence estimates. The August, 1958, national estimate, for example, grants the Soviets the capability to begin deployment during 1959, to reach 100 missiles one year later, and to reach 500 a year after that. The estimates are not committal as to when, during 1959, one should reckon the starting date. A supplement to this August estimate was issued in November, emphasizing that dates about a year later than the original ones seemed more probable for the Soviet capability to reach 100 and 500 respectively. This supplement also emphasized the difficulties the Soviets might encounter in reaching even these objectives. What has been drawn on this chart are then two curves which include between them something of the range of possibilities allowed by the August and November estimates. They picture, of course, a range of estimates; not the whole range allowed by the published NIE. They illustrate one of the variable factors that enter into an examination of the strategic situation. These same curves have been repeated on some of the subsequent charts.

The second chart, plotted to the same scale, shows for comparison the present United States intercontinental missile programs. As you well know, these have proceeded rapidly in the recent past, and operational intercontinental missiles will begin to appear in 1959. Present programs terminate in early 1964 with 180 Atlas and Titan missiles on site, and nine Polaris submarines capable of carrying a total of 144 missiles. All 144 Polaris missiles have been shown, even though, typically not more than half of them will be at sea at one time.

We turn specifically now to the question of the force needed by the Soviets to attack us. The third chart shows, against the background of the intelligence estimates, an estimate of the number of aiming points we present in the continental United States, and an estimate of the number of intercontinental ballistic missiles the Soviets would need against them. The lower curve counts the targets the Soviets would have to attack in order to destroy or neutralize:

1. all SAC air bases in the United States,

2. key air defense installations in the United States,

3. all missile bases in the United States.

This curve begins with about 40 targets at the present time and grows to about 140 as we add SAC bases, missile bases, and air defense control centers.

The blue curve above estimates the number of Soviet missiles needed to attack these targets. In addition to these missiles, the Soviets would probably wish to bring over some aircraft in a follow-up blow, but at no time would he need more than [text not declassified] over the United States to complete the destruction of all strategic targets.

The number of intercontinental ballistic missiles here estimated as his needs makes generous allowance for unreliability and aiming error in his missiles. It is assumed that these missiles are only 60% reliable with an aiming error—CEP—of 3-1/3 miles. Thus, five missiles are directed at each SAC base, and up to 30 missiles at each hardened missile site. Fifty to 100 attacking missiles are allotted to installations other than bases of retaliatory force, e.g., air defense and communication centers. It is felt that an attack by even half this many missiles could deal us a destructive blow.

Using this generous basis of estimate, the chart shows the Soviet needing 200 missiles to attack us in 1959. The figure then increases with time as we activate more SAC bases, add SAGE centers, and later add missile sites to our target system.

One cannot, from this chart, draw with certainty the conclusion he would like to draw—namely that the Soviet force will always be inadequate for a decisive attack against our continental forces.

This is a rather weak conclusion—the conclusion that one can’t conclude—and is not the primary reason for setting up this chart. Rather, we would like to pursue the argument on force requirements beyond this point, in order to illustrate some of the technical possibilities. We would like to discuss here how to get out of the red.

In looking at defense against ballistic missiles, our Panel was impressed by the relative simplicity and effectiveness of certain passive means available to our striking forces. For example, using the kind of missile postulated for the construction of this chart, the Soviet attacker needs five missiles to be certain of destroying an array of SAC aircraft on the ground or a cluster of three Atlas missiles of the early type. These are targets which can be destroyed by a one megaton bomb burst, 3-1/2 miles away. If these targets are sheltered to withstand 100 pounds per square inch of blast pressure, a one megaton bomb would need to burst within ¾ of a mile. Correspondingly, the Soviet would need almost 200 of the missiles postulated to be certain of destruction—in other words, he wouldn’t even attempt direct destruction with such a missile. He would have to await a better missile, or resort to artillery tactics and pin his target down until he can kill it with his aircraft. Even pinning his target down would cost him 20 to 40 missiles, in place of his original five.

It has seemed to the Panel that there are no technical reasons why this tactic of hardening cannot be used as a means for greatly increasing the enemy’s force requirements against us. If we could harden effectively overnight, for instance, the Soviets would now, in 1959, need something more than 1000 missiles to neutralize our SAC.

In terms of an example more practical than hardening overnight, the overlay shows an estimate of the effect on the Soviets force requirements of what appears to be a reasonable series of steps for sheltering of SAC aircraft and further dispersal and hardening of missile sites. To develop this rapidly rising curve of Soviet needs, it was assumed that some construction could begin at SAC bases late in fiscal 1960; this then first bears fruit in calendar 1961. Also in 1961, our present Atlas and Titan programs begin to add to the target system bases hardened in differing degrees. Here it has been assumed that about 1/3 of the Titan missiles—all of which are expected to be hardened to 100 pounds of blast—would be deployed to individual aiming points, rather than sited in groups of three. It has also been assumed that the Atlas missiles—now expected to appear in groups of three hardened to 25 pounds—would either be deployed to individual sites or else hardened to 100 pounds. We have just heard that Defense has been examining the possibility of further dispersal within present programs, and we are encouraged.

There isn’t room on the chart to show the full effect of a program of this kind; and certainly not room to show the effect of a more ambitious program of sheltering and dispersal. What has been shown is in fact a conservative estimate of the effect of the program I have described. It shows the steps toward stability which even a relatively simple program of passive protection can bring about. Until such time as the Soviets have a highly advanced missile, one hardened aiming point added to our target system costs the Soviets from 10 to 20, or even many more, missiles to negate. This factor applies to every missile of ours we can add to the system.

In sum, with respect to the question of force requirements, we find the situation now unclear. There seems to be no technical obstacle to creating a picture both more clear and more favorable; indeed, this appears possible by steps relatively much simpler than the creation of new weapons systems.

We can turn now to the second question—what kind of retaliatory force can we get off against the USSR if he does elect to attack us? We have prepared a chart which attempts to answer this question under different a assumptions about the size of the Soviet attack.

You will observe that this chart, except in its reference to Polaris missiles, is directed specifically to the state of affairs in the continental United States. It is recognized that the United States does have elements of force elsewhere in the world that could, potentially, contribute further retaliation. However, virtually all of these forces are within the reach of Soviet tactical air and missiles. The intelligence has been clear for sometime that the Soviets have a tactical air force and tactical missiles adequate to cope with any targets we may present to them. There is no basis for assuming that this circumstance will change under our present programs.

In the presence of this Soviet superiority in the tactical theater, if we are to make our overseas forces effective when the Soviets initiate intercontinental war, we must achieve, between our continental and our overseas forces, a very high degree of coordination. In fact, the technical conditions imposed upon us in this situation seem much more stringent than those imposed upon the Soviets if they choose to negate our efforts by careful coordination of their own attack.

We can see no technical obstacles to prevent the Soviets from attaining the kind of coordination they would need. Therefore, in fixing our attention in the next chart upon forces in the continental United States, plus Polaris, we feel that we are not presenting an inaccurate description of our whole strategic position.

Two different estimates are here shown of the retaliatory force which we might be able to mount after a Soviet attack upon us. In presenting these estimates I must emphasize how many assumptions one must make to arrive at them, and how difficult it is to verify these assumptions. For example—how does one know how fast SAC can take off aircraft in 1961? We have access to, and have used, the figures they have developed for planning purposes, but there is no possibility of testing these against fact. Similarly, we have had to postulate a rate of arrival for an initial salvo of Soviet missiles. So as not to be guilty of painting too black a picture, we have assumed for this chart a rather ragged initial salvo against SAC—a full hour’s duration for the attack, with three quarters of the attack arriving in the first half of the hour.

The estimates given on the chart then correspond to attacks based on the two curves projecting Soviet capabilities as displayed earlier. That is, the upper curve in this chart assumes that the Soviet missile force builds up with time in the manner of the smaller intelligence estimate and the lower curve assumes an attack based on the larger intelligence estimate. Thus, for example, in mid 1961, an attack by 100 Soviet missiles allows us the upper estimate of retaliation, an attack by 500 missiles, the lower estimate.

It will be helpful first to examine the effect upon our strategic aircraft of a Soviet attack; the important relations are best displayed by looking first at the lower blue curve.

Prior to mid 1959, the intelligence estimate allows the Soviets no significant force of missiles. Therefore, during this period the Soviets first blow upon the continental United States would be an attack by aircraft and something like two hours warning would be available to all SAC bases. During 1959 then, the curve of aircraft available follows SAC’s present estimates of the number of aircraft they will have available in two hours. The rise in the curve reflects SAC’s expectations for improving readiness. By the end of 1959 the intelligence estimate allows the Soviets a significant force of missiles and an attack under these conditions would take its toll of our aircraft. This curve reaches its minimum during 1961, at a time when the Soviets are estimated to have about 500 missiles and before our Ballistic Missile Early Warning System is in full operation. Thereafter, the curve rises again under the assumptions that the BMEWS system will be installed as scheduled, and will be operational.

The smaller intelligence estimates of November lead to the upper estimate of aircraft available. Here, because until 1961 there is no significant force of Soviet missiles, the upper curve continues to follow SAC’s present estimates of the number of aircraft to be available in two hours. The continued rise in the curve reflects SAC’s expectations for improving readiness.

After January 1961, the Soviet missiles begin to take their toll, while, on our side, BMEWS is going into operation.

So far we have been discussing the force of aircraft which we could mount in retaliation. As our missile forces build up, they contribute to the picture, as shown on the overlay. Following the lower curve, we find a final total of about 70 missiles surviving and available for retaliation. These correspond to that half of the Polaris force which, because it is at sea, escapes attack.

Under the lighter Soviet attack of the upper curve, a fraction of the hardened Titan missiles also are likely to survive attack. This fraction diminishes as the forces available to the Soviets increase.

We find, in the end of the period, a retaliatory force estimated at about 700 bomb carriers. Such a force is not to be dismissed lightly, but we must recognize that there are many factors which can influence the actual size of this force and cause it to differ from these estimates. We will want to discuss some of these factors, not with the aim of degrading these curves to a more gloomy picture of our condition, but in order to illustrate our main thesis.

As we said at the start, our main thesis is this: that our true strategic position is unclear, and that our estimates of it are set about by doubts. Our actual position is greatly sensitive to factors over which we have no control, necessarily then our estimates of this position are sensitive to the assumptions we make about these factors.

We feel that there are several things further that this country can do to remedy the situation, but before discussing solutions, I would like further to explore the anatomy of the problem.

What are the main sources of uncertainty in our situation as illustrated by these charts?

First, we are greatly dependent upon warning and upon decisive and impressively rapid response thereto.

Every blue aircraft on this chart is one that took off after a decision that attack was imminent or that war was unavoidable. When the real attack comes, every minute of indecision costs us about 40 bombing aircraft; five minutes of indecision cost us 200 bombing aircraft; every SAC base covered with snow costs us about one aircraft per minute. A complete failure of BMEWS—perhaps because of jamming or spoofing—would cost us about 450 bombers.

It is worth noting also that, especially in the early period, as we respond to the immediate warning that an attack is in progress, those aircraft which escape destruction do so by winning a race against the attacker. The force that escapes, therefore, is not exactly of our choosing, but is selected by the fortunes of war. It may not therefore conform in its structure to a pre-determined war plan; it may have to be regrouped and re-directed to targets for which it is adequate, according to information and a plan developed during the attack.

It may be felt that the emphasis in this chart upon warning is unrealistic because the chart assumes an attack which, apart from the last minute warning we get from BMEWS or a bomb burst, takes us completely by surprise. But the situation postulated here is really quite otherwise. In fact, this chart assumes a great deal of strategic warning; indeed a background of international tensions so ominous that the United States has undertaken to bring the whole of SAC to a peak of readiness. There remains only enough equivocation in the picture of Soviet intent to keep us hopeful that war can be avoided. In such a situation, we cannot respond freely to possible false alarms, lest we on the one hand fall victim to a feint and get caught on the ground after our forces are recalled, or on the other hand provoke an otherwise avoidable attack upon us. We must then await a clearly overt act by the Soviets to remove our doubts. This act could be the penetration of BMEWS, or the burst of a bomb on one of our bases. In sum, the assumptions here about our readiness seem to be favorable to us rather than otherwise.

A second important feature of our position is that we are greatly sensitive to the mass, coordination, and technical excellence of a Soviet missile attack. Sensitivity to mass hardly needs more emphasis than the difference between the upper and lower curves already gives it. Let us also recognize that the lower curve does not call for a Soviet attack as large as the national estimates leave room for. Sensitivity to coordination prevails throughout the period.

To see the sensitivity of these estimates to the coordination of the Soviet attack, we have tried to estimate what would happen to our aircraft if the Soviets mounted a perfectly coordinated missile attack—by perfectly coordinated, I mean an attack in which 100 to 300 missiles arrived in the first five minutes. In mid 1960, the lower blue curve would be depressed to about 130 aircraft. In mid 1961 it would go virtually to zero. In late 1963, even with the full operation of BMEWS, the curve would go to less than 200. At the other extreme, of course, in a ragged and very poorly timed attack we could expect to get off almost all of our available aircraft—from 300 to 800 bombers.

The disparity between these two extremes is impressive. This disparity is the measure of our sensitivity. Whatever one’s private judgment about where, between these extremes, reality lies, it is clearly evident that a tremendous premium is being offered to the Soviets for skill in coordinating their forces. Again, we can find no basic technical reasons for assuming that the Soviets cannot reach a high degree of coordination, particularly in the initial salvo of an attack for which they had long prepared.

Sensitivity to technical excellence is similar to sensitivity to coordination. If the Soviet missile had a one mile aiming error, rather than the two to three miles assumed in constructing the chart, we would get off less than 400 aircraft in 1963.

A third important feature of our position as shown by this chart is that the bulk, if not the total, of our potential retaliation is borne by aircraft. We know that the USSR has an extensive and improving air defense system. We must assume that this system will in time include nuclear weapons. It is difficult to know exactly how effective this system will be against our aircraft, and even more difficult to guess how effective the Soviets think it will be. Therefore, it is hard to know exactly how impressed the Soviet planner will be with our ability to do him retaliatory damage.

It is fairly certain, however, that if we are to do him much damage with our aircraft it will be because we have been able to get off a fairly large and well coordinated attack. One need not count the trees in order to see the outline of the forest: we are not only dependent upon aircraft for the bulk of our retaliation, but also we are dependent upon their ability to survive as a coordinated military unit.

In mentioning these sources of uncertainty, we are not calling into question the value to the United States of current programs to provide warning of attack and to accelerate reaction to that warning. Indeed, the discussion of the uncertainties should serve to emphasize the presently supreme value of these measures. But, as long as the success of these measures is sensitive to factors over which the United States has little control, it is in the nature of the measures themselves to create a condition of instability. For in relying upon warning and rapid response, we offer to pay a great premium to the Soviets—in fact, either to ourselves or to the Soviets—for haste in reacting to a threat.

To restore stability one needs certainty. To create certainty, one needs a retaliation which is inevitable, is inevitably decisive, and is clearly recognizable as such by a potential attacker. In the long run, to do this requires a force which is large enough, is so secure against attack that it neither provokes attack nor need react in haste, and is insensitive to defensive action.

The United States is, of course, building the foundation for such a retaliatory force through its missile programs. It seems to us, in fact, on technical grounds, that a force of the right kind of missiles, adequately dispersed and protected by sheltering or mobility, can always be relied on as a prime deterrent: it is relatively insensitive to attack and to defensive action, and is therefore able to react deliberately and—if in sufficient force—decisively.

In the meantime, major reliance must remain upon aircraft. Here we feel that there are some further simple steps that the United States can take to increase the certainty of its retaliation.

First, we have already mentioned the possibility of sheltering aircraft to increase the Soviets’ force requirements. Sheltering also increases the survivability of a useful fraction of the force, and decreases dependence upon haste.

Second, we can rather easily do more toward getting, or toward better insuring, warning. It may be possible to hasten the Ballistic Missile Early Warning System; it appears possible, and indeed attractive, also to supplement this radar system with an airborne infra-red system capable of detecting missile launchings. We refer here to a simple system that might be implemented soon—not to more remote satellite-borne system. The United States can also simply, and therefore soon, provide SAC bases and other key centers with automatic bomb alarms tied together by communications to warn all bases immediately and certainly when any one base is attacked. This could be a very simple system indeed. It appears that it could be available by the end of fiscal 1960; there is no funded program for such a system at the present time.

The Panel also recognizes the value of more certain strategic warning and would encourage any steps in this direction.

Third, we can increase the ability of SAC to take off rapidly by the simple step of providing more runways.

Such further measures as these just mentioned pay off well. In talking earlier of hardening, for example, we noted that each hardened aiming point, when added to our target system, costs the Soviet from 10 to 20 or more missiles to negate. This rate of exchange would apply, for example, to every new missile that we could add to the force in a hardened configuration.

We can also estimate the effect of better warning in the same terms used to develop the present chart. For example, and to set the scale, in 1963 the difference between having a fully functioning BMEWS system—as postulated for the chart—and no such system, is about 450 aircraft. A similar difference prevails on the lower curve as early as 1961. Presence of an infra-red system alternative and supplementary to BMEWS is therefore insurance for a precious possession. It also insures against under-shooting BMEWS on low angle trajectories. Furthermore, since an infra-red system is alerted by launchings, rather than by penetration during flight, it can be expected to speed up our responses; if for example it added ten minutes to our warning time in mid 1961, this would be worth 120 aircraft. A similar value attaches to a bomb alarm system in this period. Either type of warning would have a much greater effect were we to assume a more sharply timed Soviet attack.

Doubling SAC’s take-off rate would add 100 or more aircraft in 1961, perhaps 50 in 1963, under the attack postulated for this chart. It could have more value in a more highly coordinated attack, and it is always insurance against damage to, or accidental clogging of, runways.

To summarize our views then:

We feel that it is very clear, and for good technical reasons, that the most effective way for the United States now to invest its resources in military measures is to proceed in the direction of hardening, dispersal, and increasing dependence upon protected forces of intercontinental ballistic missiles. For the near future we must also look to further measures for the protection of our aircraft. There are programs now in the Department of Defense aimed at increasing our security, which simply do not attack the heart of the problem as we see it. We have raised questions about some of these programs in an earlier discussion of technical factors relating to the Defense budget. We feel that if some of these programs were given their proper emphasis, most of the steps that we have here emphasized as important could be implemented effectively and without adding to the Defense budget.

1. Source: “An Analysis of Technical Factors in the Strategic Posture of the United States—1955–64.” Top Secret. 18 pp. Eisenhower Library, White House Office Files, Additional Records of the Office of the Special Assistant for Science and Technology.