STATEMENT OF DR. GEORGE W. RATHJENS TO THE SENATE COMMITTEE ON ARMED SERVICES, APRIL 23, 1969*

(Excerpt)

I welcome the opportunity, and am honored, to appear before you to comment on the question of our deployment of an anti-ballistic missile defense system. .

...

In considering defense of our strategic retaliatory forces two questions must be asked:

1. Are they likely to be so vulnerable to preemptive attack in the near future that a decision must be made now to remedy that vulnerability?

2. If so, is the deployment of the proposed SAFEGUARD ABM defense for MINUTEMAN the preferred way of dealing with the problem?

The Administration's decision implies an affirmative answer to both questions. In my judgment the answer is almost certainly negative.

Our strategic forces now have the capability to deliver over four thousand nuclear warheads against an adversary. Less than one tenth of that force could, according to former Secretary of Defense McNamara, destroy over 75% of the industry and 30% of the population of the Soviet Union, the latter figure being almost certainly a low estimate since it is based on immediate and easily calculable fatalities with those that are delayed and difficult to quantify being neglected. During the time when SAFEGUARD is being deployed the number of warheads our strategic force can deliver will be increased to ten thousand or so assuming implementation of present plans to replace large fractions of our MINUTEMAN and POLARIS force with new missiles carrying several warheads each.

While these new warheads will each be of lower yield than those they replace, they will nevertheless be very much more powerful than those that destroyed Hiroshima and Nagasaki. The net effect of these changes is that our capability for inflicting damage on an adversary will go up sharply during the next few years. Even without adding to this the fact that we also have several thousand nuclear warheads in Europe, some of which could be delivered by tactical forces against the USSR, a preemptive strike against us in the mid-70's would seem like madness on the part of the Soviet leadership unless they could have extremely high confidence of being able to destroy at least 95%, and more likely 98%, of our retaliatory force.

This implies an ability to destroy nearly simultaneously our ICBM and our POLARIS forces, a requirement that could be met only if the Soviet Union were to develop very large numbers of highly accurate missiles and an extremely effective anti-submarine warfare capability. In this connection it is to be noted that we are probably ahead

*From hearings, "Authorization for Military Procurement, Research and Development, Fiscal Year 1970, and Reserve Strength," Senate Committee on Armed Services, 91st Congress, 1st session, part 2.

of the USSR in all of the relevant technologies: the development of MIRV's, the attainment of very high accuracies with our missiles, and in ASW developments. Yet even if the Soviet SS-9 missile force were to grow as rapidly as the Defense Department's most worrisome projections, even if the Soviet Union were to develop and employ MIRV's with those missiles and even if they achieved accuracies as good as we apparently expect with our MIRV forces (according to figures released in late 1967 by former Deputy Secretary of Defense Nitze), a quarter of our MINUTEMAN force could be expected to survive a Soviet preemptive SS-9 attack. That quarter would alone be more than enough to inflict unacceptable damage on the USSR. The possibility of the Soviet Union attaining an ASW capability that would imperil our POLARIS force is even more remote. Despite our superiority in ASW there is no reason to believe that even we will have really effective ASW capabilities during the next few years. (If there were any real basis for hope we would presumably be expanding our forces when in fact the Nixon Administration budget involves a cut of $105 million for attack submarines.)

In addition to means of destroying our ICBM's and SLBM's, the Soviet Union would also have to have a highly effective air defense including capability to cope with improved air-to-surface missiles if it were to rationally consider a preemptive attack against us. This follows because of the impossibility of coordinating a preemptive strike against all our retaliatory forces, and because of the possibility that some of our aircraft may be airborne at the time of attack. If an adversary were to time an attack so that our ICBM's and bomber bases would be struck at the same time, early warning of the attack against our ICBM's would permit us to launch at least 40 percent of the bomber fleet. If an attempt were made to strike the bomber bases by surprise using FOB's or SLBM's, then we would have at least 20 minutes after destruction of the bombers during which to launch our ICBM's. Thus, a coordinated attack that would knock out ICBM's, SLBM's, and bombers would appear to be impossible.

From the Soviet point of view an effective first strike is an even more difficult undertaking than the above implies. While we have designed our ICBM force so that it can "ride out" an attack against it, Soviet decision makers could never be sure that the force would not be launched before their weapons actually detonated.

Thus, the determination that action is now needed to cope with a potential mid-70 threat to our retaliatory capability seems decidedly premature. I know of no basis for believing that the concatenation of events required to place our retaliatory forces in peril by the mid-70's is even a remote possibility.

But, however tenuous the basis for the argument, the Administration has clearly decided that our retaliatory capability may be threatened in the mid-70's, and that action is now required to insure that it will not be. If indeed such an essential element of American strength as our retaliatory posture may be in jeopardy, I would suggest that defense of MINUTEMAN using SAFEGUARD is hardly a high confidence solution.

There is, as I am sure you are aware, substantial doubt in the scientific community about whether such a complex system as SAFEGUARD can be counted on to function reliably, particularly consider

ing that it will never be possible to test it in an environment that even remotely simulates that in which it will have to operate. Experience with far less complex systems reinforces those doubts. Recollection of our experience with our ICBM programs should certainly give us pause.

Despite the fact that the requirements for reliable performance in the case of a defense of ICBM's may not be as high as for defense of cities, it seems to me that the Administration is asking great forbearance on the part of the American public when it attempts to persuade them that our retaliatory capability is in jeopardy, and then offers as a solution to the problem a system in which they can have no more confidence than in SAFEGUARD.

However, even if SAFEGUARD could be counted on with high confidence to perform exactly in conformance with specifications, there would still be reason to doubt that it would be a good solution to the question of possible MINUTEMAN vulnerability.

Appendix: Defending MINUTEMAN Missiles With the SAFEGUARD System

I

The Administration has stated that the total cost of Phase I of its SAFEGUARD plan will be 2.1 billion dollars. Two ICBM sites at Great Falls, Montana, and Grand Forks, North Dakota, will be defended.

It is unlikely that either will have more than 75 SPRINT missiles considering that the Administration has announced that the number of SPRINTS would not be increased significantly over the earlier plan, and that all twelve SAFEGUARD sites are now to have SPRINT defenses.

If one assumes accuracy for Soviet SS-9 missiles comparable to that expected for U.S. MIRV programs, an optimum Soviet attack would employ MIRV warheads in the yield range near one megaton and each Soviet SS-9 missile could carry several of these. Several such warheads would have to be targeted on each MINUTEMAN to have a reasonable chance of destroying it, but one such warhead could easily destroy the missile site radar.

The defense would have to assume that the radar would be targeted in any attack, and it would have to be defended. If the defense assumes its interceptors will work perfectly it might allocate only one to each Soviet reentry vehicle that might destroy the radar. More reasonably the defense might allocate at least two SPRINTS to attempt intercept of each incoming reentry vehicle. To exhaust the defenses in these circumstances the Russians would have to allocate at most 200 reentry vehicles to attack the two MSR's. If they did so they would be able to destroy some 40 fewer MINUTEMAN than if there had been no defense.

Thus, the defense will have saved that number of MINUTEMAN for a cost of $2.1 billion, or at a cost of 50 million per MINUTEMAN saved.

There are many reasons why this estimate is likely to be much too low.

1) If some of the SPRINTS are deployed too far away from the MSR to defend it, or if there are fewer than 75 SPRINTS per base, the

costs per MINUTEMAN saved could go up sharply-perhaps severalfold.

2) If the offense could count on at least two SPRINTS being allocated to attack each reentry vehicle targeted against the radar, and it probably could, it would have to allocate only half as many reentry vehicles to radar attack to exhaust the defenses, the effect being to double the cost per MINUTEMAN saved.

3) If the adversary warheads did not have a kill probability of one against the MINUTEMAN (or the radar), the number of MINUTEMAN saved will be decreased assuming, as is likely, that at least some warheads that would have gone wide of their marks would nevertheless have been intercepted.

4) If, after review, the Administration should decide to terminate the program after completing only the Montana and North Dakota bases, then all of the development and production tooling costs would have to be written off against just the two sites.

5) While we might hope that the Administration's $2.1 billion cost estimate is correct, costs of similar programs have generally escalated upon implementation. (Note also that the $2.1 billion figure does not include AEC costs for SAFEGUARD.)

Considering these reasons, the 50 million dollar figure above should be multiplied-perhaps by a factor of three to ten. Obviously then, an estimate of 50 to 100 million dollars per MINUTEMAN saved is very conservative.

Two arguments can be made against the foregoing line of reasoning: 1) It can be said that the discussion neglects the effects of SPARTAN missiles in degrading an adversary attack. On technical grounds this seems like a reasonable approximation. While the offense would of course have to allocate some effort to insuring penetration of SPARTAN defenses, it would hardly be enough to change significantly the above estimates. In any case, the Administration can scarcely argue otherwise since, in claiming as it has that the Soviet Union need not react to a nationwide SPARTAN defense by improving its offensive forces, it has implicitly acknowledged SPARTAN'S inutility as a defense against a sophisticated attack.

2) It can be argued that the discussion fails to consider the bonus effect implicit in the fact that the sites at Grand Forks and Malstrom would offer protection of population against Chinese attack. It is probably reasonable to neglect this considering that the relatively small cities in this area, particularly in the area protected by the Malstrom site, would be unlikely to be hit by the Chinese. This would be particularly so if only the Phase I program were implemented in view of the fact that they could as easily attack other far more populous targets which would not be defended at all.

II

For the Soviet Union to have a capability to destroy 90% of the MINUTEMAN force in a preemptive attack using a MIRVed SS-9 force, about 850 missiles would be required (assuming accuracies similar to those we expect and reliabilities of about 75%). For the capability to be achieved by January 1975 requires production at the rate of about one missile every three days. Assuming 4 or 5 MIRV's per missile, from about 20 to 50 missiles would be needed to exhaust 150

SPRINT interceptors, the number depending on the reliability of the offensive missile and the number of SPRINTS to be used against each attacking reentry vehicle. With a production rate of one ICBM every three days, from two to five months' production would suffice to neutralize the SAFEGUARD Phase I deployments. If the offense chose to use somewhat smaller yield reentry vehicles for radar attack than would be optimal for ICBM attack the production time would be further reduced.

STATEMENT OF DR. ALBERT WOHLSTETTER TO THE SENATE COMMITTEE ON ARMED SERVICES, APRIL 23, 1969*

(Excerpt)

I appreciate the honor of testifying before this Committee on the role of an anti-ballistic missile system in the 1970's. . .

Understanding of the complex problems of designing a protected and responsible nuclear strategic force has grown slowly among scientists as well as laymen, civilians as well as soldiers, Democrats as well as Republicans. But it has grown, and I think decisively. The United States has designed and deployed a second strike force capable of riding out an attack; and there have been large improvements in protecting responsible command.

This was accomplished not by merely expanding nuclear bombardment forces, but in essence by shifting to forces with protection against the changing threat. The stereotype repeated throughout the 1960's that our security has declined while our strategic force grew at an accelerating rate is grossly wrong on both counts.

In the past some key programs increased the protected second strike capacity of the force, while cutting at the same time billions of dollars from the spending projected, and our security is much greater in the 1960's, since we have protected and made more responsive our strategic force.

In the 1970's, unless we continue to make appropriate decisions to meet technological change, once again the viability of a large part of our second strike force will be put in question. Several related innovations, but in particular the development of a rocket booster carrying many reentry vehicles each aimed precisely at a different target (MIRV's), raise once again the possibility of attack ratios favoring the attacker. One reentry vehicle may kill a booster carrying several. One booster can carry the means of destroying many boosters.

Raising a question about the future second strike capacity of any part of our strategic force implies nothing about the present intentions of an adversary to strike first or even to be able in the future effectively to strike first. The recent debate on whether the SS-9 is a "first strike weapon" or whether the Russians intend it to be seems beside the point. If by maintaining our second strike capability we can make the risks of striking very great, this can affect an adversary's intentions favorably to ourselves.

From hearings, "Authorization for Military Procurement, Research and Development, Fiscal Year 1970, and Reserve Strength," Senate Committee on Armed Services, 91st Congress, 1st session, part 2.