Foreign Relations of the United States, 1961–1963, Volume VII, Arms Control and Disarmament

22. Memorandum From Secretary of Defense McNamara to the National Security Council⁰

Washington, May 15, 1961.

The Defense Department is scheduled to review with the National Security Council on Friday, May 19, certain data relating to the resumption of atomic weapons testing.

Attached for your information is a copy of the material to be presented to the Council. The document reflects the views of the Joint Chiefs of Staff, the Director of Defense Research & Engineering,¹ Mr. Gilpatric, and myself.

Robert S. McNamara

Enclosure

Paper Prepared by the Director of Defense Research and Engineering (Brown)

QUESTIONS BEARING UPON THE RESUMPTION OF ATOMIC WEAPONS TESTING

Answers to the following questions are pertinent to the position which should be taken by the United States with respect to the resumption of atomic weapons testing.

(1) What evidence do we have that the Soviets are continuing to test?

Answer: We have no proof that the Soviets are testing, nor do we have any proof to the contrary. Since the beginning of the moratorium in 1958, there has been a continuing and vigorous association between development laboratories and individuals known to have been associated with testing programs prior to the moratorium and activities at sites which are particularly suitable for underground testing. It is questionable whether this evidence can support a firm belief that the Soviets are doing nuclear testing. The proper conclusion to reach from this situation, however, is that under present circumstances intelligence indications cannot in fact produce any conclusion about whether or not the Soviets are testing which is firm enough so that national policy can be based upon it.

The Soviets evidenced great interest in pure fusion devices as early as 1952, and in 1958 produced scientific papers on experimental work dating back to the earlier days, which had been done in a related area. The work done at that time involved very low levels of fusion reaction and was far from constituting an explosion, although the Soviets mentioned possible explosive uses as a conceivable future development. [12-1/2 lines of source text not declassified]

(2) What does the Department of Defense consider to be the military objectives (or information) to be attained through the resumption of testing?

Answer: The objectives of resuming testing are important, numerous, and varied. The following are indicated as the ones of most importance and interest to the Department of Defense:

(a) To pursue the development of [less than 1 line of source text not declassified] pure fusion weapons. These are particularly applicable to limited warfare, but if successfully developed, also for general application to the whole spectrum of defensive and offensive weapons systems. Such weapons provide anti-personnel effects over a well defined radius without wholesale physical destruction or radioactive fallout. [5 lines of source text not declassified]

[12 lines of source text not declassified] Its effects on limited war could be very great.

On the other hand, it should be pointed out that the U.S. has some tactical nuclear weapon capability [less than 1 line of source text not declassified]. If no cutoff of fissile material production takes place, tens of thousands of such weapons using [4 lines of source text not declassified] could be made available without unacceptable diversion from other uses. By exploding them at a height above the ground of several fireball radii, the military problem of fallout is avoided although there may still be a substantial psychological problem. These fission weapons would, however, be considerably less effective (covering perhaps 1/20 of the area) for a given yield, and the yield would tend to be the limiting factor with regard to acceptable blast or thermal effects.

(b) To develop a spectrum of reduced size and weight weapons, without undue sacrifice of yield, for particular application to strategic missile systems for the purpose of improving survivability and penetrability. This development affords the prospect of strategic missile warheads in [less than 1 line of source text not declassified] (i) for use singly in lightweight missile systems capable of high mobility and a high degree of invulnerability at relatively low cost or (ii) for cluster warheads in present or future systems to improve penetrability against an anti-missile defense. The development of these weapons would proceed through straightforward extension of the present state of the art.

The effect of the availability of such weapons of lower weight on our deterrent capability depends on how marginal that capability is otherwise considered to be. It is clear that going from [2-1/2 lines of source text not declassified]. Advances in the [less than 1 line of source text not declassified] pound range are likely to have a smaller influence, since the U.S. already has a considerable degree of mobility in its retaliatory force, and could get greater and more diversified mobility even without nuclear tests.

Developments of small strategic weapons certainly will enable us to create new systems which plug soft spots in our arsenal of second-strike weapons, by allowing us to use smaller missiles which are thereby made easier to deliver from mobile (land, sea, or air) launchers, or by increasing the multiplicity of warheads and thus saturating possible defenses, or by any combination thereof. It is also possible to reduce such weaknesses by developments in other areas of weaponry. But cutting off nuclear tests clearly inhibits the U.S. from developments toward reduced vulnerability in what has been one of the most rapidly changing areas (reduction of warhead weight).

(c) To determine the lethal ranges for various kill mechanisms against ballistic missile re-entry vehicles for warheads. The range of some of these mechanisms is fairly well known against specific warhead designs. There are others whose existence is known but whose range is not. In view of the frequency with which such mechanisms have been found, it is likely that there are still others, whose nature is still unknown, with perhaps still larger lethal radii. Knowledge of this kind is important both in examining U.S. possibilities for defense against ballistic missile attack, and in deciding how our ballistic missile attacks (including the nature of the re-entry vehicle, degree of saturation, and the necessary separation between warheads) must be planned if the Soviets should begin to deploy an AICBM system of their own.

On the other hand, considerable work can be done without nuclear explosions to reduce the uncertainties in the lethal radius for those effects whose nature is known. The uncomfortably large factor between the large value we must assume for possible lethal radii against our own warheads and the small lethal radius which we must assume our own AICBM’s would have against Soviet warheads is a result of four factors. One of them is the lack of nuclear tests which could provide more precise lethal radii produced by a given warhead against a re-entry vehicle. Another is our uncertainty about the nature of the Soviet re-entry vehicle which is probably not matched by a corresponding uncertainty on their part about ours. A third is that we do not know the yield of possible AICBM’s which the Soviets might deploy, and the lethal radius goes up with the square root of the yield. The fourth is that we must deal with a Soviet first strike against us, and consider a second strike by us against the Soviets; under these circumstances we are talking about our small warheads, which tend to be more vulnerable, and Soviet larger ones which can be made very much less so.

Only one of these factors of uncertainty can be reduced by nuclear testing. However, nuclear testing might conceivably disclose other effects which could create very large lethal radii, which are now unknown and therefore not subject to examination in terms of past nuclear experiments. Furthermore, it is possible that very cheap and very light warheads might allow a barrage approach to AICBM at acceptable cost. [1-1/2 lines of source text not declassified]

Much of the nuclear testing of use in this connection would be underground. Some of the more complex phenomena might have to be examined in outer space.

It should be added that some nuclear tests could disclose smaller lethal radii than had been expected. Such nuclear tests by either side would tend to increase stability, since it is hard to think of anything more dangerous than an AICBM system which was believed to work but actually did not.

(d) To provide adequate assurance that new weapons entering stockpile to meet our most critical needs will function properly. A number of important weapons under development or in production have not been tested as finally designed (e.g., warheads for Polaris, Minuteman, and Skybolt). Functional tests can be performed underground at yields which are small fractions of the design yields.

Though these weapons have not been tested in their operational configuration, it should be understood that they have been certified by the AEC as having the same assurance of functioning as those which have been tested. By testing them in this way, however, more precise predictions of their full yield may be made.

(3) What would be the relative advantages to the United States and the USSR of testing by both sides?

Answer: Assuming both sides test and disregarding the effect, if any, of a treaty or a continued uncontrolled moratorium on nth power acquisition of strategic weapons, it appears that the strategic situation would be stabilized by the development by either or both sides of smaller thermonuclear warheads. Our development of such weapons would enhance our deterrent capability much more than Soviet development of large or small thermonuclear weapons would improve their first-strike capability. Correspondingly, their development of small thermonuclear weapons would degrade our counterforce capability.

This last point is relevant only if counterforce is otherwise a feasible U.S. strategy. What has made counterforce unattractive to the U.S. even from the purely military point of view has been Soviet secrecy which has been able to keep the location of even quite large missiles hidden from the U.S. So long as this continues the Soviets have no particular need for smaller warheads.

How important this increase in stability is likely to be will depend on whether our deterrent strategy now or at a later time has severe soft spots, in terms of such things as vulnerability of our missiles either before launch or to Soviet AICBM, etc.

It is generally assumed that the U.S. knows more about the effects of nuclear explosions in space and the upper atmosphere than the Soviet Union. If one believes that the Nike-Zeus warhead is the right AICBM warhead, then it is probably true that the corresponding Soviet warhead would either weigh more or give less yield. This argues that resumed testing by both sides would enable the Soviets to “catch up”. On the other hand, despite our “lead” in these areas, we have not been able to develop a satisfactory AICBM system. This argues strongly that to “catch up” in nuclear warheads will not produce such a system for the Soviets either. If in fact they do deploy an AICBM system of the terminal interception Nike-Zeus type, it indicates either they know more about suitable nuclear (or non-nuclear) AICBM warheads than we do or that they are making a mistake. Both of these results would be cause for great concern, which a continued moratorium would in no way remove.

It may be that some of the new warhead developments [less than 1 line of source text not declassified] or effects which are unknown or about which we know little could have a large influence on AICBM. Any such possibility depends on knowledge about nuclear weapons and their effects which we do not possess, and therefore resumed testing by both sides is unlikely to make our relative position any worse.

One may also want to consider the situation in which residual military forces have an important effect on the nature and condition of the nation, and on the termination of the war. Then quantitative improvement by a factor of three to ten in weight of strategic offensive weapons could be an important influence on both sides. In very light strategic weapons (below 2000 pounds) the Soviets appear to be somewhat behind the U.S., and might therefore gain more than we if both sides resume testing and eventually become equal. Of course, their relative gain would be still greater if they test clandestinely and the U.S. does not test.

When it comes to the matter of nuclear weapons for limited war, we know much less about the Soviet program than about their strategic weapons. In particular, in the case of the pure fusion bombs, there is no way of telling whether the Soviets are ahead of us or behind us, and if both sides test in this area there is no particular reason to expect alteration in relative capability. It may be argued that the U.S. has a greater need of these than the Soviet Union in order to overcome a disparity in conventional armament. This asymmetry might well instead be overcome by a free-world build up in conventional armament (and anti-guerrilla capability), but it is a fact that the asymmetry now exists and will persist for some time. Even if the pure fusion bombs are developed, it may not be possible to use them in many areas of the world without fear of escalation. Whether threats of such escalation would deter a side which had sole possession of such weapons cannot be foreseen.

All the above has been on the basis that both sides resume testing. In view of the answer to question (1) above, which indicated that U.S. intelligence can in fact not provide convincing evidence either that the Soviets are or are not testing in the low-yield range, the situation in which only one side tests must be offered for comparison. Recognizing these limitations of intelligence under a continuing moratorium or under a treaty of the kind the Soviet Union now insists on, the Soviet Union could if it so desired carry out low yield underground tests without fear of U.S. acquisition of information on which we could act. There would thus be no deterrence to Soviet clandestine testing.

[3-1/2 lines of source text not declassified] The precise military effects of such a situation are difficult to evaluate, but the Soviets would clearly be able to gain in the strategic areas, and might produce a very dangerous unbalance in AICBM. [2 lines of source text not declassified] In any event, under such circumstances they could do far more than “catch up”.

(4) What specific devices would be tested in the near future?

Answer: This question is best answered by the Atomic Energy Commission since the Commission would design the test program to meet the objectives outlined above and other important goals. The attached list, however, indicates a number of the devices which we understand would be readied for testing within the next year.

(5) How do the above conclusions compare with those of the Fisk Panel?

Answer: Although this document weighs the various factors differently than does the Fisk Panel Report, the technical evaluations are consistent with those in that report. Several new factors have appeared which change the situation somewhat, although not drastically.

The first is the further development of the pure fusion work in the AEC laboratories. This makes the achievement of such weapons appear quite likely, although the Fisk Panel Report already assumed that they were possible.

The second is that further study indicates the likelihood of larger lethal radii against U.S. ICBM’s than previously has been believed to be the case. Although nuclear tests would resolve some of the most important questions about these lethal radii, much can be done by non-nuclear experiments and calculations. The relative importance of nuclear tests and non-nuclear tests and calculations is subject to the usual debate. Whatever the merits of this argument, the existence of larger lethal radii against existing warheads narrows the margin of strategic deterrence. It thus tends to make possible advances in strategic nuclear warheads a more critical item in the deterrent balance.

A third and perhaps more important factor is that the Fisk Report assumed that even small-scale violations would be subject to some risk of detection in case of acceptance of the U.S. draft Treaty. Therefore, although it considered a variety of cases, it predicated its approach on the basis that the Soviets would have to weigh their possible gains against the consequences of being caught and a real risk of detection. The Fisk Panel Report therefore evaluated the military situation by comparing the case in which both sides test and that in which we do not and the Soviets do some cheating, with the situation in which neither side tests. As the political situation is now evaluated, there will be either no treaty, or if there is a treaty the control system will be virtually powerless. The deterrence of violation would appear to be either very much weakened or non-existent. The case in which both sides refrain from testing probably can no longer be the standard to which one refers the other cases for comparison.

Harold Brown

Attachment

Tentative and Partial List of Weapons Tests To Be Undertaken Within the Next Year

1. [1 line of source text not declassified]

2. [less than 1 line of source text not declassified]

3. [2 lines of source text not declassified]

4. [1 line of source text not declassified]

5. [2-1/2 lines of source text not declassified]

6. [less than 1 line of source text not declassified]

7. [1 line of source text not declassified]

0. Source: Washington National Records Center, RG 330, OSD Files: FRC 65 A 3464, Atomic 400.112 8 Apr 61. Secret; Restricted Data. A copy was sent to McCloy.
1. Harold Brown.