June 5, 1996

Dear Congressman Spratt:

I am writing in response to your May 24, 1996, letter which expresses concerns about National Missile Defense (NMD).

Overall Costs. Let me begin with a summary of our estimates of the cost of various proposed deployments of NMD systems, subject to the qualifications and explanations set out in this letter and its enclosure.

• The Administration's current program to develop and deploy the "3+3" system would cost about $10 billion. Costs for development alone would be about $2.5 billion. This amount, but not the cost of deployment, is included in the Administration's five-year budget projections.

• Adding to the Administration's 3+3 program the Space and Missile Tracking System (SMTS) for improved sensing capability would add about $5 billion to the cost of a deployed system.

• The Army and Air Force have proposed emergency-deployment anti-ballistic missile systems, able in theory to be fielded more rapidly than the 3+3 system, at a cost they estimate to be $8 billion and $3 billion, respectively. These cost estimates have important omissions and may be substantially understated.

• Larger and nominally more capable systems, such as those envisaged in the Defend America bill, would include more complex and difficult-to-develop concepts such as space-based lasers for a layered defense. The Congressional Budget Office (CBO) has estimated their cost in the $31-60 billion range. The CBO estimate is roughly in line with prior DoD estimates of the cost of comparable systems. More recent cost estimates of some space-based systems have been higher, suggesting that, if anything, the CBO range is low.

The balance of this letter explains these concepts in more detail.

The "3+3" System. The NMD program described in the President's budget request is designed to field as early as 2003 a system able to deal with threats of ballistic missile attack by rogue states (e.g., Iran, Iraq, Libya, or North Korea). This program, based on a three-year development phase followed by a three-year deployment phase, has been commonly referred to as "3+3". When fielded at a single site and depending on its architecture, the resulting system is projected to be able to defend all 50 states against a limited attack by a rogue nation. It would also have some capability against a small accidental launch from more nuclear-capable states.

This sensible approach uses the best technology available because, in contrast to the Defend America Act, it does not lock us into deployment now, but permits us to defer a deployment decision until we know it is necessary. Consequently, it enables us to make the best assessment of the threat before making a costly commitment of defense resources and to adapt the system to be deployed to the actual threat it would have to meet.

Recognizing the importance of NMD, we have recently designated it a Major Defense Acquisition Program and directed the establishment of a Joint Program Office. This will add important discipline into program management and execution to ensure that we fully understand and deal with program risk areas, and have an executable program with a validated cost estimate.

At this early stage of development we have not yet fully defined the 3+3 program nor completed a rigorous cost estimate. But we expect it would cost about $10 billion--$ 2.5 billion to develop, and an additional $7.5 billion to produce and deploy an initial system which would include 100 interceptors. I want to note that the Administration budget and program now include the approximately $2-5 billion needed to develop the 3+3 system. Because we do not recommend that a deployment decision be made until the system has been adequately demonstrated and a rogue nation threat appears more likely to emerge our program does not include funding for fielding the system. of course, if a decision were made to proceed with the system, the costs of deployment would have to be allocated. In addition, we would have to program the operating and support costs for the system.

The enclosure describes the 3+3 system in more detail.

SMTS. The Air Force's SMTS system would add significant sensing capability to the 3+3 system, thereby maximizing the possible defended area of the ground-based interceptors. SMTS would cost about $5 billion to develop and deploy as an adjunct of a ground-based NMD system. Such deployment is one of the options being examined in the 3+3 program. Accordingly, SMTS would, if added to our base 3+3 system, make the total 3+3 system cost about $15 billion. This estimate is in line with the CBO estimate of $14-17 billion for an essentially similar system.

Army and Air Force Emergency Systems. The Army and Air Force have described NMD options that could provide a limited capability against simple, rapidly-emergent threats somewhat earlier than the 3+3 system. Although the two services believe their respective systems could be deployed at a somewhat lower cost than the 3+3 system, the proposals have some obvious omissions. Accordingly, as LTG O'Neill noted in his May 23, 1996 letter, their costs could be substantially higher. More importantly, although these options would need to be considered in an emergency, we do not believe they provide a long-term approach to the NMD challenge. But in developing the 3+3 program we have sought to capture the strongest features of each.

Space-Based Systems. The Defend America Act calls for a layered NMD system that would ultimately include space-based weapons--either space-based interceptors, space-based lasers, or both. The CBO assessed that such a system would cost $31-60 billion. This range is consistent with earlier DoD estimates for these types of systems. And it is somewhat below more recent estimates of some components, particularly space-based lasers. Thus, the CBO cost estimate for a layered system appears to be in the right range, although it may be optimistic. Further, it does not appear feasible to deploy such a system in six years.

None of the approaches to NMD is low risk, but the Defend America Act approach contains the greatest risk. And no architecture that could be deployed within six years could defend against an unauthorized launch which might contain a large number of warheads (e.g., a full load of warheads from a Russian SSBN). We believe the 3+3 program maintains the best balance of matching our most capable NMD technologies against the most plausible threats.

Congress needs consistent information on this important subject. To help ensure this, I have asked Dr. Kaminski and BMDOto review all NMD data as we send it to Congress.