Foreign Relations of the United States, 1977–1980, Volume IV, National Security Policy

150. Memorandum From Secretary of Defense Brown to President Carter¹

Washington, August 24, 1979

(C) I have made my final review of the MX program preparatory to your decision. Upon your concurrence, I will initiate full scale engineering development of the basing system. Development of the large missile (190,000 pounds and carrying 10 RVs) has already begun, as we agreed in June.² The baseline MX program envisions deploying 200 of these missiles, each mounted on a transporter-erector-launcher (TEL). Each TEL would have a deployment area consisting of 23 horizontal shelters interconnected with a loop-road. We would conceal which of the 23 shelters contained the TEL, but would periodically (or on demand) open the shelters for inspection by satellites. As an additional security feature we would be able to redeploy the missiles from one shelter to any other in a time less than that of ICBM flight.

(C) The basing concept is described in more detail in the attached paper, which also covers costs, survivability (including resilience to escalated threats), verification and environmental impact.³ The material presented is essentially that discussed at the PRC meeting of 7 August 1979, with additional clarification on matters pointed out by the participants at the meeting and in subsequent conversations with them.⁴ The paper has been reviewed in draft by other agencies at the staff level. I’m sending Zbig copies to distribute to Cy Vance, Stan Turner, Frank Press, Jim McIntyre, and George Seignious.

(C) The system is estimated to cost about $33 billion in FY 80 dollars, including development, production, and installation; operating costs will be $440 million per year. That is a substantial expense, but comparable with other major strategic programs. Corresponding costs for the Minuteman program (in FY80$) are $40 billion, with operating costs running $340 million per year. By comparison, the TRIDENT program will cost $39 billion (FY80$) if we build enough TRIDENT submarines (25) to maintain a force of 600 SLBMs (slightly fewer than now deployed).⁵ Operating costs for the SLBM program run more than $1 billion per year and for the B–52 program about $1.6 billion per year (including tanker costs). Alternative MX concepts with adequate survivability would be even more costly (Air Mobile MX and Trench MX for example), except for the vertical shelter MPS system. That would cost about $27 billion (FY80$) but would not be as satisfactory a solution either from a verification or a survivability point of view.

(C) The costs have been carefully prepared, but with any complex system projected 10 years into the future, there is bound to be uncertainty. Historically, major DoD projects have averaged 15% higher than their originally estimated costs. I have taken that as a challenge to prepare careful and conservative estimates in the first place and to maintain tight management control during the execution phase. The most critical phase in terms of cost control is the first year or two, when requirements tend to balloon (“gold-plating”) and when the contract structure is established. There is already pressure to “improve” the system in various ways (installing radar intrusion sensors at each shelter instead of each complex, e.g.) which I intend to resist. There are some features which probably will ultimately be added to the system (mass simulators, e.g.), but which I intend to defer until their need is clear. Finally, there is the possibility that the Soviets will increase their ICBM threat more than we have estimated, especially in the absence of SALT or termination of SALT. In that case we would expand the MX system in response and that would, of course, increase the cost. Expanding to 300 missiles and 6900 shelters, for example, would cost an additional $8 billion.

(C) It may also be possible to reduce costs by actions we take in the engineering development phase. The shelter spacing is set conservatively at 7,000 feet. We will explore the possibility of reducing this to 6,000 feet with concomitant reduction in road costs. We will also investigate the development of automated mass production techniques to reduce the cost of the shelters. Finally, there is another side to the threat uncertainty. We may be able to negotiate significant reductions in the Soviet ICBM threat in SALT III, which would allow us to deploy fewer MX missiles than planned, with a corresponding cost reduction.

(S) A key technical-military issue with this system is whether it poses a sufficiently difficult task to the formidable Soviet counterforce capability projected for the 1990’s and beyond so that our ICBM forces can then be considered survivable. The system will not achieve its full operational capability for 10 years. During that period we must depend primarily on our SLBM and bomber/cruise missile forces for high confidence deterrence because of the possibility of a crippling surprise attack on our Minuteman force. While the bombers and cruise missile carriers can also be attacked at their bases by Soviet SLBMs, I believe that our at-sea SLBM force will be invulnerable to attack during this 10-year period. However, we cannot count on such SLBM invulnerability indefinitely, especially if we give the Soviets a chance to concentrate on it. From that, in major part, flows the importance of restoring the ICBM survivability by the end of the 1980’s. Other factors include the C3 and accuracy associated with the ICBM force.⁶

(S) The MX achieves survivability by (a) forcing the Soviets to target shelters instead of missiles, and (b) building more shelters than they can realistically attack effectively. We achieve the first objective by concealing which one of the 23 shelters contains the missile. I believe the elaborate measures we have devised will allow the Soviets to know that only one shelter has a missile but will not allow them to know which one. If this belief proves unfounded, that is even if the Soviets were to locate the missiles and launch an attack, we could move the missile to a different shelter, before the ICBM reached its target. This rapid dispersal capability is intended primarily to complicate the task of the Soviet war planner (thereby deterring him from attacking) and to give the President an insurance policy against loss of security, or assertions of such loss. As a result of these measures, we can have high confidence that any attacker wanting to destroy the MX missile force would have to target all of the shelters. [10 lines not declassified] In short, the Soviets would be in a worse position after the attack than before, even without considering a US retaliatory strike against residual Soviet forces.

(S) Larger Soviet ICBM forces are possible to envisage. These would require a correspondingly larger MX “force” (more shelters, and perhaps more deployment areas, depending on the nature of the threat) to achieve comparable levels of survivability. In the attached paper we describe, in addition to [21 lines not declassified]

[1 paragraph (15 lines) not declassified]

(S) Another crucial issue is verifiability of the system. I believe we have designed a system which deals effectively with that issue, both in fact and in perception. This system should be easier to verify than other mobile systems—SLBMs, bombers, and road mobile systems like the SS–20 that are not confined to a designated deployment area. CIA and [less than 1 line not declassified] analysts have done an outstanding job of assessing verifiability (and as a result have influenced some of our design features). They conclude that the system can be verified either as a US deployment or as a Soviet deployment. The verification issue is treated in detail in the attached paper. Also, Stan has prepared a separate memo on this subject which you will want to review.⁷

(U) Environmental impact and public reaction have been handled very well by the Air Force. There is good public support in the probable development areas—Nevada and Utah. The Governors of both states have written to you expressing their interest. The attached report contains a section on this matter. We plan a vigorous effort to make MX energy-independent via solar techniques. Though very much a side-issue in military terms, such an effort could be of significant national value in maturing that technology so that costs reduce to acceptable levels for other applications.

(C) I believe that we should move out promptly with a decision to start full scale development.⁸ We now have the funding (the FY 79 Supplemental Appropriation) and we can demonstrate to Congress and the public a forceful step to strengthen our strategic forces. By announcing a decision before 5 September, we can take the initiative rather than appear to be responding reluctantly to the criticisms of delay that will be made when Congress reconvenes.⁹ Before you make a public statement, you or I should personally call Governor List of Nevada and Governor Matheson of Utah, as well as Senator Robert Byrd and Speaker O’Neill, to give them advance notice.

Harold Brown

1. Source: Carter Library, National Security Affairs, Staff Material, Defense/Security, Ermarth, Box 7, MX Missiles: 8/79. Top Secret; Codeword. Sent via Alpha Channel.
2. See Document 142.
3. Not found attached.
4. See Document 147.
5. Carter wrote in the right margin next to this and the previous sentence: “looks like SLBM’s are cheaper.”
6. Carter wrote in the margin to the right of this and the previous sentence: “ex schizophrenia.”
7. See Document 151.
8. Carter underlined “move out promptly” in this sentence.
9. Carter underlined “before 5 September” in this sentence.