Draft - 12/2/66

THE SECRETARY OF DEFENSE
WASHINGTON

MEMORANDUM FOR THE PRESIDENT

SUBJECT: Production and Deployment of the NIKE-X.

A number of events have occurred during the last year which, taken together, tend to bring to a head the long-standing issue of whether to produce and deploy a U.S. anti-ballistic missile defense:

1. The Soviet Union has accelerated the deployment of hard ICBMs beyond the rates forecast in last year's NIE (but not beyond the "higher than expected" case on which the U.S. Defense Program was based).

2. The Soviet Union has started the deployment of an anti-ballistic missile system around Moscow and perhaps a second type of ABM system in other parts of the country.

3. The Chinese Communists have launched and demonstrated a nuclear-armed, 400-mile range ballistic missile, and there is some evidence that they may be preparing to test a booster in the ICBM range.

4. Our own anti-ballistic missile system, the NIKE-X, has now reached a stage of development where it may be feasible to start concurrent production and deployment.

5. The Joint Chiefs of Staff have reaffirmed their recommendation that a decision be made now to deploy, with an initial operational capability in FY 1972, a NIKE-X system which would provide for area defense of the continental U.S. and local defense of 25 cities against a "low" Soviet threat.

6. The Congress for the first time since 1959 has appropriated funds to prepare for the production and deployment of an ABM defense system.

There are three somewhat overlapping but distinct purposes for which we might want to deploy an ABM system:

1. To protect our cities (and their population and industry) against a Soviet missile attack;

2. To protect our cities against a Chinese Communist missile attack.

3. To help protect our land-based strategic offensive forces against a Soviet missile attack.

After studying the subject exhaustively, Mr. Vance and I have concluded we should not initiate ABM deployment at this time for any of these purposes. We believe that:

1. The Soviet Union would be forced to react to a U.S. ABM deployment by increasing its offensive nuclear force with the result that:

a. The risk of a Soviet nuclear attack on the U.S. would not be farther decreased.

b. The damage to the U.S. from a Soviet nuclear attack, in the event deterrence failed, would not be reduced in any meaningful sense.

The foundation of our security is the deterrence of a Soviet nuclear attack. We believe such an attack can be prevented if it is understood by the Soviets that we possess strategic nuclear forces so powerful as to be capable of absorbing a Soviet first strike and surviving with sufficient strength to impose unacceptable damage on them (e.g., destruction by blast and radiation alone of approximately 20%-30% of their people and 50% of their industry). We have such power today. We must maintain it in the future, adjusting our forces to offset actual or potential changes in theirs. (a)

There is nothing I have seen in either our own or the Soviet Union's technology which would lead me to believe we cannot do this. From the beginning of the NIKE-ZEUS project in 1955 through the end of this current fiscal year, we will have invested a total of about $4 billion on ballistic missile defense research — including NIKE-ZEUS, NIKE-X and Project DEFENDER. And, during the last five or six years, we have spent about $1.2 billion on the development of penetration aids to help ensure that our missiles could penetrate the enemy's defenses. As a result of these efforts, we have the technology already in hand to counter any offensive or defensive force changes the Soviet Union is likely to undertake in the foreseeable future.

We believe the Soviet Union has essentially the same requirement for a deterrent or "Assured Destruction" force as the U.S. Therefore, deployment by the U.S. of an ABM defense which would degrade the destruction capability of the Soviet's offensive force to an unacceptable level would lead to expansion of that force. This would leave us no better off than we were before.

2. With respect to protection of the U.S. against a possible Chinese Communist nuclear attack, the leadtime required for China to develop a significant ICBM force is greater than that required for deployment of our defense — therefore the Chinese threat in itself would not dictate the production of an ABM system at this time.

3. Similarly, although the protection of our land-based strategic offensive forces against the kind of heavy, sophisticated missile attack the Soviets may be able to mount in the late 1970s might later prove to be worthwhile, it is still premature to produce and deploy the NIKE-X for that purpose.

In the pages which follow I will explore in detail the foundation for these conclusions:

1. The Soviet Strategic Threat

The latest National Intelligence Estimate, dated Oct. 20, 1966, indicated that the Soviets have accelerated the deployment of two hard ICBMs, the SS-11 and SS-9. (The SS-9 is a large, storable liquid-fueled missile, roughly the size of our TITAN II, with a warhead yield of 12-40 MT and a CEP of 0.5-1.0 n.mi. The SS-11 is a small, storable liquid-fueled missile, about the size of our MINUTEMAN, with a warhead yield of 0.2 – 1.2 MT and a CEP of 1.0 – 1.5 n.mi.) The November 1965 NIE estimated that by mid-1968 the Soviets would have operational about 100-110 SS-9s and 200-250 SS-11s; we now estimate that they will have 130-140 SS-9s and 320-400 SS-11s by that date. (a)

[ (a) In addition to the SS-9s and SS-11s the NIE forecasts that the Soviets will have in mid-1968 273 other missiles, including missiles on the test ranges.]

By mid-1971, we believe they could have a total of 800-1100 operational ICBMs on launchers, compared with last year’s estimate of 500-800 by mid-1970. We believe the higher end of the range of estimates will prevail if the Soviets decide to emphasize quantity in an effort to match the size of our ICBM force, and the lower end if they choose to emphasize quality. In the first case, they would concentrate on the SS-11 which is a relatively simple and cheap missile. In the second case, they would place added emphasis on the SS-9 which is a more expensive and also, for certain purposes, a much more effective missile. The SS-11 because of its relatively poor CEP and small payload would have little value against hard targets such as our MINUTEMAN silos, and it is therefore essentially a retaliatory weapon for use against cities. The Soviets also have some older ICBMs but these are already being phased out and few are expected to be left in the operational force by 1971.

Although we still have no direct evidence of such an effort, the Soviets might also develop and install multiple independently-aimed reentry vehicles (MIRVs) on their SS-9s. However, an effective capability with such reentry vehicles would require much greater accuracies (lower CEPs) than have thus far been achieved by Soviet ICBMs. If they were to start now, they could probably achieve an operational capability by about 1971-72; and we would probably be able to detect the testing of such a system perhaps two years earlier. Improvements in both accuracy and penetration capability could also be made in the SS-11s, and in addition the Soviets might deploy a new solid fuel, highly accurate small ICBM.

We have known for some time that the Soviet Union was working on anti-ballistic missile defense. After several false starts the Soviets now appear to be deploying one type of system (which is definitely designed for ballistic missile defense) around Moscow and another type of system, designated “Tallinn”, (which my be designed for defense against manned bombers, or ballistic missiles, or both) across the northwest approaches to the Soviet Union and at a few other locations.

The Moscow system appears to consist of a series of complexes deployed at some of the outer ring SA-1 sites, about 45 n.mi. from the center of the city. Each complex has two “Triads” (one large and two small radars operating together) and 16 launchers apparently designed for the "GALOSH" missile which the Soviets displayed in 1964. (six complexes are under active construction and a seventh is now dormant.) In addition, there is a large phased-array radar southwest of Moscow (called Dog House) oriented towards our ICBM threat corridor and additional large phased-array radars (called Hen House) sited at two locations to the northwest. These three radars my be intended as forward acquisition radars for the Moscow system, while the Triad radars handle the target and interceptor missile tracking functions. The Moscow system could have an initial operational capability in 1967 or early 1968, and a full operational capability with six complexes (96 launchers) by 1970-71. (By that time the Soviets could also construct two more complexes to fill out the southern part of the ring, for a total of 128 launchers.)

The GALOSH itself is a large, relatively slow acceleration missile probably designed for exoatmospheric interception much like our new, extended range interceptor missile, which we now call "OLYMPIA". We have no evidence thus far of a Soviet terminal defense missile such as our SPRINT. If used for both area and terminal defense, the GALOSH system would be very expensive, at least $15 million per missile on launcher (dividing the total investment cost by the number of missiles on launchers) where only 16 missiles are provided per complex. Even if two reload missiles were provided for each launcher, the cost per missile would still amount to about $6 million. But there is a real question whether the reloading speed of the GALOSH (now estimated at 10-30 minutes after arrival of the missile at the launcher) would be fast enough to be of any use in a single engagement. Similarly, there is a question whether a single Triad, the radars of which are mechanically steered, could handle more than eight launchers. (We ourselves have abandoned this type of radar for ABM defense because of its grave limitations.)

With regard to the second defensive system, there is still disagreement within our Intelligence Community as to its primary purpose. One view is that it is primarily an advanced surface-to-air missile system designed against high altitude, high speed manned bomber attacks, and the pattern in which it is being deployed, the configuration of the sites and their equipment, and the characteristics of the radars, all lend credence to this view. Several "farms" of missile launchers are located in a barrier line across the northwestern part of European Russia and around Leningrad and Moscow, and seme parts of the southern approaches. At least 22 complexes have been definitely identified, most of which consist of three launch sites, each with six launch positions and one radar. These could be operational by 1967-68, and more may be under construction. The local radars associated with the launchers are of limited capability and would appear to need the support of the much larger but vulnerable Hen House radars if the system is expected to perform with a reasonable degree of effectiveness in the ABM role.

If it is indeed designed as an advanced surface-to-air, antiaircraft missile, it would be most effective in defending against high-altitude penetrating bombers of the B-70 or SR-71 type; it would be ineffective against low-altitude penetrating bombers such as the B-52 or FB-111. It is this incongruity, together with the fact that this type of ABM system would be much cheaper than the GALOSH, which leads the proponents of the other point of view to believe that it is an ABM system, or at least has some minimal ABM-capability. And while we know something about the geographical deployments of this system, and about its launchers and radars, we still know very little about the interceptor missile itself.

The latest intelligence estimates (HIE 11-3-66) concludes "...that the Moscow ABM system will have a good capability against a numerically limited attack on the Moscow area by currently operational missiles, but that its capabilities could be degraded by advanced penetration systems and it could not cope with a very heavy attack. Moreover, the present deployment will not cover all of the multi-directional POLARIS threat to Moscow." With regard to the Tallinn system in the ABM role, the NIE concludes:

“Many of the Tallinn system complexes are so located that presently known Hen House or Dog House radars could not furnish useful target tracking data on them. Where this is the case, or if the Hen Houses or Dog House were destroyed or blacked out, the capabilities of the system would be seriously reduced and limited to local and self-defense. Thus, under these assumptions including the alternate characteristics which would have to be assumed for the missile to give it an ABM capability if Hen House or Dog House data were available, the Tallinn complexes could defend areas large enough to provide a strategic ABM defense; without such data, they could not.”

In summary, we have firm evidence of Soviet ICBM deployment through mid-1968 and fairly good estimates through mid-1971. Our knowledge of Soviet ABM deployments is much more sketchy. We are reasonably certain that the deployment of the GALOSH around Moscow will be completed but until we know more about the Tallinn type of system or see evidence of GALOSH deployments around other cities, we can only conjecture about the ultimate scale, effectiveness and cost of the Soviet ABM effort. However, knowing what we do about past Soviet predilections for defensive systems [*], we must plan our forces on the assumption that they will have deployed some sort of an ABM system around their major cities by the early 1970s. Whether made up of GALOSH only, or a combination of GALOSH and a [DELETED] type system, or even some combination of GALOSH and a terminal missile of the SPRINT type, a full scale deployment would cost the Soviet Union something on the order of $20 to $25 billion.

[* – The Soviets for more than a decade have spent substantially more on air defense against strategic bombers than has the U.S. The bulk of the Soviet expenditure has been wasted — throughout the period the U.S. Strategic Air Command stated, and it was generally agreed within the United States Government, that approximately 85 percent of the U.S. incoming bombers could penetrate the Soviet defenses and reach their targets.]

2. History of the U.S. ABM Effort

In considering the issue of whether to deploy the NIKE-X, it might be useful to review briefly the history of the U.S. ABM effort, the kind of system originally envisioned, the evolution of technology in that field and the attitudes of past Presidents, Secretaries of Defense, Chiefs of Staff, the Congress, etc.

The predecessor of the current ABM development program, the NIKE-ZEUS, was begun in FY 1955. Up until the launching of the SPUTNIK in October 1957, the project proceeded at a leisurely pace. Congressional attitudes towards the program ranged from incredulousness regarding its operational feasibility (especially in view of the problems then being encountered in anti-bomber defense) to concern over a new "roles and missions" fight between the Army (ZEUS) and the Air Force (WIZARD).

In the aftermath of SPUTNIK a new sense of urgency developed with regard to all aspects of advanced military technology. From FY 1955 through FY 1957, a total of only $12.2 million was applied to NIKE-ZEUS R&D but in FY 1958 alone the total rose to $66 million and in FY 1959, to $237 million. By the spring of 1958 when the FY 1959 Budget was before the Congress, the Army had already proposed the production of initial sets of equipment. Secretary of Defense McElroy, however, argued that "we should not spend hundreds of millions on production of this weapon pending general confirmatory indications that we know what we are doing.” His view prevailed for the moment.

It was not until the FY 1960 Budget that NIKE-ZEUS deployment became a real issue. The Army’s initial request included $875 million for ZEUS — $35 million for R&D, $720 million for procurement and $115 million for construction. President Eisenhower, however, sent to the Congress a request of $300 million for R&D and test facilities only. The House Appropriations Committee recommended the addition of $200 million "for the acceleration of the NIKE-ZEUS and/or the modernization of Army firepower." Secretary McElroy agreed to accept $137 million for the acceleration of NIKE-ZEUS and $63 million for Army modernization. The Senate approved these amounts and added $200 million more for Army modernization. The final enactment provided $375 million for NIKE-ZEUS and/or Army modernization.

In the fall of 1959, in connection with the development of the FY 1961 Budget, the Army proposed a new NIKE-ZEUS deployment plan consisting of 35 local defense centers (one for each defended area), 9 forward acquisition radars and 120 batteries. The typical battery was to consist of 50 missiles on launchers and 16 radars, a missile-to-radar ratio very close to that of the current Soviet GALOSH system. An initial operational capability was to be achieved by FY 1964 and the entire program completed by FY 1969, with a total investment cost estimated at $13 to $14 billion, of which $1.5 billion would be required in FY 1961.

The system was designed around a relatively slow speed and limited range interceptor missile and mechanically steered radars. Because of the missile’s slow speed, it had to be fired long before the incoming target reentered the atmosphere, thereby precluding the use of the atmosphere as a means of distinguishing real warheads from other objects such as decoys or tankage fragments; and the limited range of the missile reduced the potential kill radius. (Indeed, the plan called for the firing of three ZEUS against each attacking ICBM.) Because the radars were mechanically steered (like the local GALOSH radars), the traffic-handling capabilities of the system were low, leaving it vulnerable to saturation attacks.

This plan was rejected by President Eisenhower, who pointed out in his FY 1961 Budget message that:

"The NIKE-ZEUS system is one of the most difficult undertakings ever attempted by this country. The technical problems involved in detecting, tracking, and computing the course of the incoming ballistic missile and in guiding the intercepting ZEUS missile to its target—all within a few minutes—are indeed enormous.

"Much thought and study have been given to all of these factors and it is the consensus of my technical and military advisors that the system should be carefully tested before production is begun and facilities are constructed for its deployment. Accordingly, I am recommending sufficient funds in this budget to provide for the essential phases of such testing. Pending the results of such testing, the $137 million appropriated last year by the Congress for initial production steps for the NIKE-ZEUS system will not be used."

The Joint Chiefs of Staff, with one dissenting vote, supported the President's position and the Congress agreed to limit the program to research and development.

The weaknesses in the NIKE-ZEUS system led in 1961 and 1962 to the development of a new and different system known as NIKE-X. To help solve the problem of discriminating actual warheads from decoys and other objects, a new, high acceleration terminal defense missile, the SPRINT, was designed. Because of its fast reaction time, this missile would permit the defense to wait until the enemy attack penetrated well into the atmosphere where the lighter objects, such as unsophisticated decoys, would be separated from the warheads, thus permitting the defense to concentrate more of its fire on the latter. To solve the problem of limited handling capacity, a new family of phased-array radars was developed. These radars employ a relatively new principle; instead of scanning the skies with an electronic beam by mechanically rotating the entire radar structure, the structure is covered with thousands of sensors and is kept stationary while the electronic beam does the rotating. Because an electronic beam can be rotated a million times faster than a mechanical structure, the phased-array radar has a far greater search and tracking capacity. In other words, it can simultaneously handle many more incoming objects, thus eliminating one of the major limitations of the old NIKE-ZEUS system.

With the phased-array radar and SPRINT missiles, the defense battery could bring firepower to bear on all targets entering an area 20 miles high and 25 miles in radius. However, even if these batteries were deployed around all our major cities, a large part of the nation would still be left undefended and the attacker would have the option of groundbursting his warheads outside the defended areas, thus producing vast amounts of lethal fall-out which could be carried by the winds over the defended areas. Moreover, a terminal (or local) defense compels the defender to allocate his resources in advance, leaving the attacker free to concentrate his resources against whatever targets he may choose at the moment of the attack.

To fill in this gap, we initiated in the spring of 1965 the development of a new long-range interceptor with a high yield, high temperature X-Ray warhead. This missile, the OLYMPIA, is designed to reach out over 400 nautical miles from its launcher and attack incoming objects at altitudes of up to 280 nautical miles. Its warhead is to be capable of destroying ballistic missile reentry vehicles at ranges of five to ten miles if they are hardened, and 10 to 100 miles if they are not. About a dozen properly located batteries of such a missile could provide some coverage over the entire United States. Together with the SPRINT, it could provide a defense in depth, permitting all incoming objects to be attacked first well above the atmosphere and then the surviving objects a second time as they enter the atmosphere. Moreover, by overlapping the coverage of the OLYMPIA batteries, some of the attacker’s inherent advantage against terminal defenses alone could be overcome, since the defender at the moment of the attack would also have the choice of concentrating his resources over those targets he chooses to protect.

The deployment of an ABM system did not become a serious issue again until earlier this year. It was clear to us from the beginning, i.e., 1961, that the NIKE-ZEUS as then conceived would not be an effective ABM system against the type of ballistic missile attack the Soviets would be able to launch by the end of the decade. Accordingly, both in President Kennedy’s and your administrations, we have steadfastly maintained that the development of a more effective ABM system should be pursued on an urgent basis but that no production or deployment, should be undertaken until much more was known about the system’s technical capabilities and its likely effect on the strategic situation generally. This view found substantial support within the Executive Branch and in the Congress up until recently, although an abortive attempt was made by some members of the Senate in 1963 to authorize an appropriation for the deployment of the NIKE-ZEUS. However, in acting on the FY 1967 Defense Budget, the Armed Services Committees and the Defense Appropriations Subcommittees of both Houses recommended, and the Congress appropriated, about $168 million to prepare for the production of the NIKE-X system. It is, therefore, clear that the deployment of this system will be a major issue in the next session of the Congress.

3. Technical Feasibility of the NIKE-X System as Presently Visualized

Attachment 1 provides a description of each of the major elements of the NIKE-X system and its current development status. Briefly, the system would consist of a number of different types of phased-array radars and two types of interceptor missiles, which could be deployed in a variety of configurations:

a. Multi-function Array Radar (MAR) — a very powerful phased-array radar which can perform all the defense functions involved in engaging a large simultaneous attack: central control and battle management, long-range search, acquisition of the target, discrimination of warheads from decoys or "spoofing" devices, precision tracking of the target, and control of the defense interceptor missiles.

b. TACMAR Radar — a scaled down, slightly less complex and less powerful version of the MAR, which would be used in a "light" deployment of the NIKE-X. It can perform all the basic defense functions on a smaller scale.

c. VHF Radar — a relatively low frequency, phased-array radar required for the very long-range search and acquisition functions involved in area defense. To achieve the full potential of the extended-range OLYMPIA, the target must be picked up at much greater distances in order to compute its trajectory before the OLYMPIA is fired.

d. Missile Site Radar (MSR) — a much smaller, phased-array radar needed to control the SPRINT and OLYMPIA interceptor missiles during an engagement. It can also perform the functions of the TACMAR but on a considerably reduced scale. Actually, a number of different sizes are being studied. This "modular" approach will permit us to tailor the capacity of the radar to the particular needs of each defended area.

e. The extended-range OLYMPIA — a three stage missile with a hot X-ray. [DELETED] warhead capable of intercepting incoming objects at a range of over 400 nautical miles and at altitudes of up to 280 nautical miles. This missile makes use of some of the components of the old NIKE-ZEUS.

f. SPRINT — a high-acceleration interceptor missile which can climb to 80,000 feet in 10 seconds. It is designed to make intercepts between 5,000 and 100,000 feet at a range of 25 miles.

In addition to these major elements of the system, an entire new infrastructure, including base facilities, communications, logistics support, etc., will be required. The exact cost of this infrastructure cannot be determined until a specific deployment plan is decided upon, but it would surely be substantial for any deployment.

The technical principles involved in the radars are now fairly well established. One R&D MAR-type radar has been constructed at the White Sands Missile Range. A contract has been let for the power plant of a second MAR-type radar, which is to be constructed on Kwajalein Atoll. The Missile Site Radar is well along in development and the construction of one of these radars on Kwajalein Atoll has also begun.

Testing of the SPRINT missile was started at White Sands in November 1965 with one complete success, two partial successes and three failures. The failures are attributed mostly to insufficient quality control but some of the missile’s components may have to be redesigned. The tempo of testing will steadily increase during the current fiscal year and we are advised by our technical people that the missile will eventually reach its design goals. The nuclear warhead is also well along in development and does not appear to present any particular problem.

The OLYMPIA is still on the drawing boards. It represents a very substantial redesign of the original NIKE-ZEUS and we will not know until it is flight tested a year and half hence how well it will perform. However, we are less concerned with the missile itself than we are with its warhead. A significant number of development tests will have to be performed, all underground, before the design parameters can be established; and then we will have to proof test the resulting warhead, again underground. (The feasibility of a full yield test underground has still to be established.) Accordingly, there is still considerable technical uncertainty concerning the warhead. Although alternative warheads could be used on the OLYMPIA, their effectiveness would be much lower.

Facilities for testing both the SPRINT and the OLYMPIA will be constructed on Kwajalein Atoll. These, together with the TACMAR and Missile Site Radar (MSR) and the programs for the computers will give us all of the major elements of the NIKE-X system which are essential to test its overall performance against reentry vehicles fired from Vandenberg Air Force Base in California. (We feel we know enough about the VHF radar technology to be able to use the mechanically steered radars already on Kwajalein as simulators.) The system will be tested in stages, starting With the MSR and SPRINT tests in January 1969, then the OLIMPIA missile in July 1969 and the TACMAR radar between July and December 1970. Upwards of 100 test shots will be launched from Vandenberg to Kwajalein during the period 1969-72 to test the system thoroughly as a whole. [*] The most important objective of this effort is to determine proper system integration and computer programming, since the individual components of the system will have already been tested ahead of time.

[* – This schedule alone raises a question whether the R&D program has advanced far enough to warrant a commitment to production in FY 1968.]

But even after this elaborate test program is completed, a number of technical uncertainties will still remain unresolved. Chief among these are the following:

1. Large Sophisticated Attacks. Notwithstanding the number of test-shots planned, the ability of the system to cope with a large sophisticated attack will still remain to be demonstrated, except to the extent that such attacks can be simulated in the computers.

2. Discrimination of Decoys and Other "Spoofing" Devices. Although the MAR-type radars are specifically designed to deal with this problem, discrimination will always remain an unresolved issue. We have been studying and developing such devices for many years and we are now installing some of them in our offensive missiles. No doubt new devices and the counters to them will be invented in the future, and the contest between the offense and the defense will continue as it has in the area of manned bombers.

3. Blackout. Detonation of nuclear devices high in the atmosphere can seriously degrade the effectiveness of the defense's radars. These detonations can be either the defensive warheads (self blackout) or deliberate explosions of the incoming warheads (precursor blackout). They have the effect of producing an area in the atmosphere similar to an opaque cloud which the radars cannot see into or through. The size of the area is a function of how high the burst occurs and of the frequency of the radars. The blacked-out region is larger at higher altitudes and appears larger to lower frequency radars. At the lower altitudes, the blackout region is essentially the visible fireball. For the terminal defenses employing SPRINT missiles in the lower atmosphere and radars in the microwave region (about 1200 megacycles), the blackout effects can be minimized and are well understood from previous testing.

For the area defense the problem is more severe. For one thing, the number of tests conducted by both the U.S. and the Soviet Union at the altitudes of interest for area defense (above 200,000 feet) is relatively small. In the U.S. tests, the data collected are not complete enough to answer all the technical issues, although out continuing study of the available data is increasing our knowledge of the blackout effects. However, we can never resolve all the uncertainties with the existing data. We know there will be blackout effects and we know that we can choose a radar frequency and proliferate radars to minimize them. But we do not know how many precursor nuclear blasts the Soviets would have to place over the United States to black out our radars. And, we do not know how much they learned from their nuclear tests. Consequently, we do not know precisely what their uncertainties would be in using this as an offensive tactic. We do know that the blackout effects can be offset by raising the frequencies of the radars, and we are doing this in the case of the VHP radars. However, because the area defense radars must detect small targets at long ranges and because the price of a radar set operating in this manner increases with the frequency, there is a limit on how far we can go in this direction to counter blackout.

4. Programming the Computers. The management of a sophisticated ballistic missile attack engagement presents an extremely complex problem. To control the phased-array radars and guide the missile, powerful computers and sophisticated "programs" are needed. The size of the computer varies with the type of radar. For the area defense (with VHF/UHF radars), computer speeds and capacities equivalent to the best of today's commercial computers are adequate. The MSR and the MAR will need much more powerful computers, development of which has been underway since 1962. However, it is not the computer itself which is our major concern, but rather the production of the "programs" which must be designed in advance to reflect every conceivable eventuality the system may confront. Our experience in programming the SAGE computers against manned bomber attacks has revealed some of the complexities, and the costs, of such an undertaking. Whether we can provide for all of the variables involved in such a vastly more complex problem as anti-missile defense has yet to be demonstrated. Here, again, we will have a much better idea of what is actually involved in programming the computers when the prototype system on Kwajalein is demonstrated in the 1970-72 period.

5. Production and Operational Problems. We have learned from bitter experience that even when the development problems have been solved, a system can run into trouble in production or when it is put into operation. All too often the development prototype cannot be produced in quantity without extensive re-engineering. Production delays are encountered and costs begin to spiral. Sometimes these problems are not discovered until the new system actually enters the inventory and has to function in an operational environment. The TERRIER, TALOS, and TARTAR ship-to-air missiles are a good example; after spending about $2 billion on development and production of these missiles, we had to spend another $350 million correcting the faults of those already installed and we still plan to spend another $550 million modernizing these systems.

In this connection, it is worth noting that had we produced and deployed the NIKE-ZEUS system proposed by the Army in 1959 at an estimated cost of $13 to $14 billion, most of it would have had to be torn out and replaced, almost before it became operational, by the new missiles and radars of the NIKE-X system. By the same token, other technological developments in offensive forces over the next seven years may make obsolete or drastically degrade the NIKE-X system as presently envisioned. We can predict with certainty that there will be substantial additional costs for updating any system we might consider installing at this time.

4. Assuming the NIKE-X System is Technically Feasible, Should It Be Deployed Now?

This question can be answered only within the context of the general nuclear war problem as a whole and our overall national security objectives. For many years the overriding objective of our national policy with regard to general nuclear war has been to deter the Soviet Union (or any other nation) from launching a surprise nuclear attack against us or our Allies. As long as that remains our overriding objective, the capability for "Assured Destruction" must receive first call on all of our resources and must be provided regardless of the cost and the difficulties involved. Programs designed to limit damage to our population and industrial capacity in the event the deterrent fails can never substitute for an "Assured Destruction" capability in this context, no matter how much we spend on them. It is our ability to destroy the attacker as a viable 20th century nation that provides the deterrent, not the ability to limit damage to ourselves.

What kind and amount of destruction we would have to be able to inflict on an attacker to provide this deterrent cannot be answered precisely. However, it seems reasonable to assume that in the case of the Soviet Union, the destruction of, say, one-fifth to one-third of its population and one-half to two-thirds of its industrial capacity would mean its elimination as a major power for many years. Such a level of destruction would certainly represent intolerable punishment to any industrialized nation and thus should serve as an effective deterrent to the deliberate initiation of a nuclear attack on the United States or its Allies.

Once sufficient forces have been procured to give us high confidence of achieving our "Assured Destruction" objective, we can then consider the kinds and amounts of forces which might be added to reduce damage to our population and industry in the event deterrence fails. But here we must note another important point, namely, the possible interaction of our strategic forces programs with those of the Soviet Union. If the general nuclear war policy of the Soviet Union also has as its objective the deterrence of a U.S. first strike (which I believe to be the case), then we must assume that any attempt on our part to reduce damage to ourselves (below what they would estimate we would consider "unacceptable levels") would put pressure on them to strive for an offsetting improvement in their deterrent forces. Conversely, an increase in their "Damage Limiting" capability would require us to make greater investments in "Assured Destruction", which, as noted earlier in this memorandum, is precisely what we are now doing. It is in this context that we should examine the desirability of increasing our "Damage Limiting" capabilities.

As I noted earlier, the major elements of the NIKE-X system, as they are now being developed, would permit a variety of deployments; two have been selected for the purposes of this analysis. The first, which I will call "Posture A", represents a light U.S. defense against a Soviet missile attack on our cities. It consists of an area defense of the entire continental United States, providing redundant (overlapping) coverage of key target areas; and, in addition, a relatively low-density SPRINT defense of the 25 largest cities to provide some protection against those warheads which get through the area defense. [*] The second deployment, which I call "Posture B", is a heavier defense against a Soviet attack. With the same area coverage, it provides a higher-density SPRINT defense for the 50 largest cities.

[* – This is essentially the deployment now recommended by the Joint Chiefs of Staff.]

Shown on the following table are the components and the costs (which, if past experience is any guide, are understated by 50 to 100 percent for the systems as a whole) of Posture A and Posture B, together with the time frames in which the deployments can be completed:

In addition, if technically feasible, we would have to provide some improvement in our defense against manned bomber attack in order to preclude the Soviets from undercutting the NIKE-X defense; we would also want to accelerate the fallout shelter program. The investment cost (including R&D) of the former is estimated at about $1.5 to $2.4 billion and would provide for a small force of F-111 or F-12 type interceptors (e.g., 48 F-111s or 32 F-12s) and about 42 aircraft warning and control aircraft (AWACS). With the introduction of these new types of aircraft, we might be able to phase out most of the present interceptor aircraft and a large part of the ground-based aircraft warning and control network, thus producing an actual saving in operating costs over the longer term. The expanded fallout shelter program would cost about $800 million more than the one we are now pursuing. We would also need some of our anti-submarine warfare forces for use against Soviet missile submarines, but we are not yet clear whether these ASW forces would actually have to be increased over the currently planned levels. In any event, the "current" estimates of the investment cost of the total "Damage Limiting" package would amount to at least $10.5 billion for Posture A and at least $20 billion for Posture B ("final" costs for each of these Postures would probably be 50 to 100 percent higher). [*]

[* – Even before the systems became operational, pressures would mount for their expansion at a cost of still additional billions. The unprotected, or relatively unprotected, areas of the U.S. (e.g., Alaska, Tampa, Birmingham, Sacramento) would claim that their tax dollars were being diverted to protect New York and Washington while they were left naked. And, critics would point out that our strategic offensive force is premised on a much larger Soviet threat (the "possible", not the "probable" threat); they would conclude that the same principles should be applied to our strategic defensive forces. For these and other reasons, I believe that, once started, an ABM system deployed with the objective of protecting the United States against the Soviet Union would require an expenditure on the order of $40 billion over a ten year period.]

To test the contribution that each of these NIKE-X deployments might make to our "Damage Limiting" objectives, we have projected both the U.S. and Soviet strategic nuclear forces (assuming no reaction by the Soviets to the U.S. ABM deployment) to FY 1976, by which time Posture B, the heavier defense, could be fully in place. These forces are shown on the table which follows:

The fatalities which these Soviet forces could inflict upon the U.S. (with and without a U.S. ABM defense) and the fatalities which the U.S. forces could inflict on the Soviet Union (with a Soviet ABM defense) are shown on the following table:

The first case, "Soviets Strike First, U.S. Retaliates”, is the threat against which our strategic forces must be designed. The second case, "U.S. Strikes First, Soviets Retaliate", is the case that would determine the size and character of the Soviet reaction to changes in our strategic forces, if they wish, as clearly they do, to maintain an "Assured Destruction" capability against us.

These calculations indicate that without NIKE-X and the other "Damage Limiting” programs discussed earlier, U.S. fatalities from a Soviet first strike could total about 100 million; even after absorbing that attack, we could inflict on the Soviet Union about 65 million fatalities. Assuming the Soviets do not react to our deployment of an ABM defense, which is a most unrealistic assumption, Posture A might reduce our fatalities to 15 million and. Posture B, to about 10 million.

Although the fatality estimates shown for both the Soviet Union and the U.S. reflect some variations in the performance of their respective ABM systems, they are still based on the assumption that these systems will work at relatively high levels of efficiency. (In fact, for the purpose of these calculations we have assumed that the Soviet ABM system will be just as good as the NIKE-X, even though we believe the system, or systems, which they are now deploying are, in fact, far inferior.) If these ABM systems do not perform as well as our technical people postulate, fatalities on both sides could be considerably higher than shown in the table above, or the costs would be considerably higher if major improvements or additions had to be made in the systems to bring them up to the required level of performance.

If the Soviets are determined to maintain an "Assured Destruction" capability against us and they believe that our deployment of an ABM defense would reduce our fatalities in the "U.S. Strikes First, Soviets Retaliate" case to the levels shewn in the table above, they would have no alternative but to increase the second strike damage potential of their offensive forces. They could do so in several different ways: by deploying a new large, land-based ICBM (either mobile, or hardened and defended), or a new submarine-launched missile like our POSEIDON, or by adding large numbers of hardened but undefended SS-9s or SS-11s. They have the technical capability to deploy any of these systems with highly accurate MIRVs (or single warheads) by the mid-1970s. Shown in the table below are the relative costs to the Soviet Union of responding to a U.S. ABM deployment with a land-mobile ICBM system:

If the Soviets choose to respond to our ABM deployment with such a system (200 missiles against Posture A and 650 against Posture B), the results would be as shown below:

But, if the Soviets choose to respond to our ABM deployment by installing MIRVs and other penetration aids in their SS-9s, they could do so at a fraction of the cost of the U.S. ABM defense, thereby substantially neutralizing that defense in a Soviet first strike. In such circumstances, even with Posture B, our fatalities would again total about 60 million, and with Posture A, about 90 million.

In short, the Soviets have it within their technical and economic capacity to offset any further "Damage Limiting" measures we may undertake, provided they are determined to maintain their deterrent against us. It is the virtual certainty that the Soviets will act to maintain their deterrent which casts such grave doubts on the advisability of our deploying the NIKE-X system. In all probability, all we would accomplish would be to increase greatly both their defense expenditures and ours without any gain in real security to either side.

5. Effect of U.S. ABM Deployment on Relations with Other Nations

With regard to our NATO Allies, two questions arise: (1) What would be their reaction to our deployment of an ABM system?; and (2) Would they want to deploy such a system?

Some European governments and many European specialists in defense and arms control matters have exhibited a growing interest in ABM defense. At the insistence of several European countries, ABM defense was discussed at the recent NATO arms control experts conference. The European and Canadian attitude as expressed at the NATO meeting was generally hostile to a U.S. ABM deployment. The same attitude was expressed by the U.K. delegation at the recent U.S.-U.K. bilateral talks on ABMs. This reaction appears to be based on a desire to avoid an accelerated arms race which Europeans believe would upset the detente. There is also some fear on the part of the British that an ABM race would price them out of the nuclear business.

Even if the U.S. offered the Europeans a similar system, it is unlikely that they would accept; only Germany has expressed a mild interest thus far. This is so for several reasons. First, the cost (at least $6 billion, and probably more, for a meaningful system) would involve a substantial increase in their defense budgets. Second, the European preoccupation with deterrence rather than defense makes it unlikely that they would pay for such an ABM system. Third, the Europeans are unlikely to achieve the degree of political and decision-making unity which would be necessary to deploy an effective ABM system.

6. Attitude of U.S. Public Toward ABM Defense

Perhaps the most difficult problem we will have to face in a decision not to deploy at this time an ABM system for defense of our cities against a Soviet ballistic missile attack is the attitude of our Congress and our people. (Deployment of the NIKE-X as a defense against a Chinese Communist nuclear attack is discussed in Attachment 3 and for the protection of our land-based offensive missile forces in Attachment 4.) The first reaction of most Americans to the events I have described at the beginning of this memorandum will inevitably be in favor of our immediate start on production and deployment, if for no other reason than the Soviets are deploying such a system. More mature reflection on all of the factors involved in this vastly complex problem should swing at least the majority of the informed public behind the decision. But we will have to undertake a massive program to present all of the relevant information, and in an understandable form, to both the Congress and the general public. Without giving them a thorough understanding of all the factors involved, we cannot hope to gain their support for a decision which on the surface appears to place dollars before lives.

7. Conclusion

Mr. Vance and I recommend that we not deploy the NIKE-X system at this time for the defense of our cities against a Soviet missile attack, but that we hold open the option to deploy this system at a later time for defense against a Chinese Communist (or accidental) attack and/or for the defense of our strategic missile forces. There is much work remaining to be done in the development, test and evaluation of the system and we recommend that this effort be pursued with undiminished vigor.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Attachment 1.

TECHNICAL STATUS OF NIKE-X SYSTEM COMPONENTS

SPRINT

SPRINT, the two-stage, high acceleration, short range interceptor missile, has been undergoing tests at the White Sands Missile Range (WSMR) since March, 1965 with the following results:

These test flight failures are a source of concern and we are working intensively to correct their causes. However, there does not seem to be a single basic fault at the root of these failures; rather each failure appears to stem from a different cause susceptible of correction.

The SPRINT warhead is in Phase 3 development at the AEC and is proceeding normally on schedule. Present test ban limitations are not a limitation on this program because of its relatively small yield. About a half-dozen successful firings of the ARROW-type warhead, on which the SPRINT warhead design is based, have been conducted, and there is very high confidence that the SPRINT warhead development will be successful.

Despite the initial flight test problems, it is believed that adequate time remains to correct these deficiencies prior to the presently scheduled start of system tests in January, 1969.

EXTENDED RANGE OLYMPIA MISSILE

The predecessor of the extended-range OLYMPIA missile, the NIKE-ZEUS DM-15-C, was extensively tested from August 1959 to May 1966. The following table summarizes the results for all classes of tests:

Ten out of 14 attempted intercepts at Kwajalein of ICBM targets launched from Vandenberg were successful. In addition, there was one successful intercept of an AGENA satellite at 98 n.mi. altitude (there were no failures).

The primary change between the new OLYMPIA (DM-15-X2) now under development and the previous ZEUS (DM-15-C), is the substitution of the old first stage as the second stage of the OLYMPIA. This increased the missile’s length from 50 to 56 feet, its weight from 23,700 to 35,890 pounds, its range from about 55 to 300 n.mi., and its payload capability from 460 to 2900 pounds (allowing it to carry a 5-6 megaton warhead). In addition, the auto-pilot has been redesigned and minor structural changes to reduce weight are now being made.

A static firing of the second stage has been successfully conducted. The first flight test will not take place until February, 1968. Because of the use of NIKE-ZEUS technology, there is a high confidence that the development of the missile itself will meet its design goals. However, there are more risks in the warhead development.

The design goals for the warhead include a yield of [------------ONE SENTENCE DELETED-------------]. This is the high temperature feature required for optimum area defense effectiveness (high lethal range). The feasibility study showed this to be feasible in principle but the actual demonstration of the Exploding Case Principle (ECP) has not yet taken place although a test is imminent (Greely Event). Assuming this is successful, other problems remain. The AEC will not guarantee the warhead yield without full yield tests. Under the present test ban, this cannot be accomplished unless we can obtain a high-yield test site. This matter is being actively investigated but its feasibility is still in question. If we are constrained to test at a one megaton level or less, then we may have to sacrifice yield. If the ECP is shown not to be feasible, we may have to have a dirtier bomb than we wish. The disadvantage here is that the dirtier the bomb, the greater chance there is that the detonation of the OLYMPIA warhead will black out its own radars.

In summary, we may have to sacrifice a certain desirable margin of performance in the warhead unless we are successful in tests not now known to be possible.

VHF RADAR

The VHF radar is the key component of the area defense and operates at a relatively low radar frequency. Because of the work on Project DEFENDER’S Electronic Scanned Array Radar (ESAR) program and on the Air Force’s SPADATS program (AN/FPS-85), the VHF radar is considered a low risk item. In fact, it will not be built at Kwajalein and the development model will be located at an operational site. Phased-array technology in this frequency range is considered well in hand.

MSR RADAR

The Missile Site Radar (MSR) is the primary terminal defense radar, providing acquisition, tracking, some discrimination, and control of the SPRINT missile. The MSR is an array radar operating at S-Band (2700 mcs). Although the basic technology has been demonstrated by Project DEFENDER’S HAPDAR Radar, the MSR represents more of an advance than the VHF radar and considerable effort is being devoted to it. The antenna elements, the key performance item, will be tested within the next year at the Raytheon plant. A model will be installed at Kwajalein for testing in CY 1968.

MAR/TACMAR

The MAR and TACMAR are the highest performance and most complicated radars in the NIKE system. The TACMAR is essentially half of a MAR radar in terms of power and performance. A prototype called the MAR I has been constructed at White Sands and tested. However, the next generation MAR/TACMAR family of radars are considerably more powerful and employ a complex beam steering system. A portion of the beam steering system (called MOSAR) is now under test at the General Electric plant in Syracuse. The other critical component of this system, the transmitting chain, is under test at the Raytheon plant in Boston. A TACMAR will be installed on Kwajalein and tests will start sometime in 1970. This is probably the highest risk radar in the system but it is not as important in early deployments since its principal role is for discrimination against a very sophisticated attack employing advanced decoys and penetration aids.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Attachment 2.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Attachment 3

DEFENSE AGAINST THE CHINESE COMMUNIST NUCLEAR THREAT

The Chinese Communists' nuclear weapons and ballistic missile development programs are apparently being pursued with high priority. Recent evidence suggests that it is possible they will conduct either a space launching or a long range ballistic missile launching before the end of 1967. Such an event would suggest that the Chinese might be aiming at an initial operating capability (IOC) for an ICBM as early as 1969, though an IOC in the 1970s still seems most likely. A launching in 1967 would also probably be taken as signifying an even more immediate threat than would actually exist. However, it still appears unlikely that the Chinese could have a significant number of ICBMs before the mid-70s or that those ICBMs would have great reliability, speed of response or substantial protection against attack.

An austere ABM defense against the Chinese threat for the decade 1975-85 might consist of 5 VHF and 12 Missile Site Radars, together with some 700 OLYMPIA and 200 SPRINT missiles (the latter to protect the principal radars). The total investment cost of such a program might amount to $3.3 billion, including the cost of the nuclear warheads.

The effectiveness of this deployment in reducing U.S. fatalities is shown in the table below:

The austere defense can preclude damage in the 1970s almost entirely. As the Chinese force grows to the level it might achieve by 1980-85, additions and improvements might be required, but relatively modest outlays for improvements in the system could limit the Chinese damage potential to low levels well beyond 1985.

It is not entirely clear that we need an ABM-defense against China. In any event, the lead-time for deployment of a significant Chinese offensive force is longer than that required for U.S. ABM deployment; therefore, the decision for the latter need not be made now.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Attachment 4.

ABM DEFENSE OF U.S. OFFENSIVE MISSILE FORCES

The most severe threat we must consider in planning our "Assured Destruction" forces is an extensive, effective Soviet ABM deployment, combined with a Soviet deployment of a substantial hard-target kill capability (i.e., highly accurate MIRVed SS-9s or SS-11s). By reducing to 0.3 n.mi. or less the CEP of their SS-9 missile, and equipping each booster with 6 MIRVs of 3 MT each, the Soviets could destroy large numbers of our MINUTEMAN missiles. We could not count on more than two years of warning between the first intelligence indications of such a development effort and the start of deployment of the system.

Assuming that the Soviets start such a development immediately, they might achieve the build-up in their threat shown below:

The effect of such a deployment on the U.S. MINUTEMAN force appears below:

To offset a possible decline in our missile force’s damage potential, we have considered the construction of additional POSEIDON submarines, and an Advanced ICBM (a new large payload missile with an as yet undetermined basing system designed to reduce vulnerability to a Soviet MIRV threat). We have also considered the defense of MINUTEMAN with the NIKE-X.

Shown below is the possible effect of a NIKE-X area defense of the approved U.S. offensive force (with [DELETED] option) against a greater-than-expected Soviet threat, compared with the "No Defense" case:

As shown, an austere area defense (with an investment cost of about $3 billion) might be able to maintain the MINUTEMAN force’s retaliatory capability. Without the ABM defense, our offensive missile forces alone could inflict at least 40 million fatalities on the Soviet Union in 1974, and with the bombers 65 million (27%). Therefore, the already approved and proposed additions to our strategic offensive forces make such action at this time unnecessary even in relation to a greater-than-expected Soviet threat.