Foreign Relations of the United States, 1981–1988, Volume XLIV, Part 1, National Security Policy, 1985–1988

136. Memorandum From Sven Kraemer and Robert Linhard of the National Security Council Staff to the President’s Assistant for National Security Affairs (Poindexter)¹

Washington, October 2, 1986

SUBJECT

• Interim Restraint Policy Options Paper

In response to tasking under NSDD–236,² a specially compartmented (MUSE) paper has been prepared under Arms Control Support Group auspices on options for US interim restraint policy. The basic study, and a verification annex (Annex B) prepared by the DCI, were completed over the last weekend, but we were not able to achieve closure until last night on a technical annex (Annex A) tasked to JCS and OSD on military sufficiency. A memo from Secretary Weinberger (included with Annex A at OSD’s request) takes strong issue with the JCS analysis.

In the memorandum to agencies attached for your signature at Tab I, it is requested that agencies provide their comments no later than Wednesday, October 8, 1986. However, keeping in mind the NSPG meeting now scheduled for Monday, October 6,³ your memo encourages the addressees to familiarize themselves with the paper at their earliest opportunity and seeks their comments as early as possible. Since agencies have had copies of the basic paper and, at least the verification annex, in their possession for several days, they should be able to provide their views to you very rapidly.

Prior to Monday’s⁴ NSPG, we will provide a brief assessment of the attached options paper, as well as any preliminary views we have received from agencies by that time.

RECOMMENDATION

That you sign the memorandum to agencies at Tab I seeking their early comments on the specially compartmented paper on Interim Restraint options attached at Tab A.⁵

Tab I

Memorandum From the President’s Assistant for National Security Affairs (Poindexter) to Secretary of State Shultz, Secretary of Defense Weinberger, Director of Central Intelligence Casey, the Chairman of the Joint Chiefs of Staff (Crowe), and the Director of the Arms Control and Disarmament Agency (Adelman)⁶

Washington, October 4, 1986

SUBJECT

• Interim Restraint Policy (U)

In response to tasking under NSDD–236, a study has been prepared on options to be considered with regard to future U.S. Interim Restraint Policy. To complete the planned review of this subject, it is requested that you provide comments on the options, as described in the attached study, no later than Wednesday, October 8, 1986. However, given the rapidly approaching meeting in Iceland and the likely Soviet interest in this subject, I would encourage you to familiarize yourself with this paper at your earliest opportunity and to provide your comments as rapidly as possible so that they can be most useful to the President in his preparations. (S)

In view of the sensitive nature of this specially compartmented study, it is requested that you restrict access to those in your agency who are currently cleared for the MUSE compartment. (C)

FOR THE PRESIDENT:

John M. Poindexter

Attachment

Study Prepared by the Arms Control Support Group⁷

Washington, September 26, 1986

[Omitted here are the title page and the table of contents.]

MUTUAL INTERIM RESTRAINT EXECUTIVE SUMMARY

In his May 27 decision, President Reagan stated our current restraint policy, inter alia, that (1) in the future, the United States must base decisions regarding its strategic force structure on the nature and magnitude of the threat posed by Soviet strategic forces; (2) the United States will exercise utmost restraint, seeking to meet its strategic needs, given the Soviet buildup, by means that minimize incentives for continuing Soviet offensive force growth. In the longer term, this is one of the major motives in our pursuit of the Strategic Defense Initiative; (3) as we modernize, we will continue to retire older forces as our national security requirements permit. We do not anticipate any appreciable numerical growth in US strategic offensive forces and assuming no significant change in the threat we face, as we implement the strategic modernization program; and, (4) the United States will not deploy more strategic nuclear delivery vehicles (SNDVs) or more strategic ballistic missile warheads than does the Soviet Union.

This study is prepared in response to NSDD–236. In that Directive, the President noted that the US and the USSR need to build a sound new foundation of truly mutual restraint and real arms reductions. The Directive requested that a study examine:

— Numerical limitations, in addition to, or as substitutes for, the May 27 statement;

— Supporting definitions and rules;

— Other qualitative limits, pending completion of START; and,

— What legal form such elements could take.

This study examines a variety of possible elements for a new US restraint policy, including those of the President’s May 27 statement, the SALT limits, and other possible limits. All elements were analyzed in terms of a number of criteria for determining their desirability and utility for possible inclusion in a new interim restraint regime: impact on US forces; impact on Soviet forces; consistency with US policy and our START position; monitorability and verifiability; whether or not the Soviets have violated limits; complexity of getting Soviet agreement to mutual restraint; and, finally an overall assessment of desirability or utility for possible inclusion in a new interim restraint regime.

Form of Undertaking

The study examines whether or not a regime should be based on unilateral statements or bilateral agreements, and concludes that a unilateral statement approach would be strongly preferred. It would reduce the likelihood of diverting the focus from the START negotiations. A unilateral statement approach would also allow the US more flexibility in changing policy and reduce the likelihood that the resulting restraint regime would be argued to require Congressional consent.

Duration

The NSDD does not specify a duration for possible mutual restraint regimes. The analysis of military sufficiency of the options presented looks at a number of timeframes (i.e., through 1988, 1991 and 1996).

The President’s Restraint Limits

The study concludes that the President’s restraint limits from his May 27 statement—that the United States will not exceed the Soviet levels of SNDVs or strategic ballistic missile RVs—are appropriate for our restraint policy. It is particularly appropriate in the case of SNDV restraint to use the actual Soviet level rather than a fixed limit, as the Soviets have violated the SALT fixed limit on SNDVs.

Central SALT Limits

The study concludes that the central SALT limits (aggregate SNDV limits (other than the President’s proposal), MIRV subceilings, limits on new types of ICBMs, and the restrictions on telemetry encryption), are not appropriate for a regime of interim restraint. This is concluded in light of their obsolescence, their permissiveness in allowing a continued and highly threatening Soviet strategic force buildup, their inconsistency with current START policy and the fact that Soviet violations have made the limits largely irrelevant.

Non-Central SALT Limits

The SALT II Treaty’s variety of lesser provisions were examined. It is concluded that several of these non-central limits could be appropriate for inclusion in a restraint policy. Incorporation of these limits into a restraint regime would be straightforward in all cases if we were to use the SALT II provisions without change. If negotiated, discussions could become complex as changes were made—changes that might be better left to START. In many cases, our START position rejects a direct carryover of SALT II limits. Hence, this study has narrowed those provisions to a small number, as summarized in the decision matrix.

Definitions and Counting Rules

The study also examined the SALT II definitions and counting rules and concluded that these are inadequate and, therefore, inappropriate for a new restraint regime pending a START agreement. While we have not as yet addressed definitions and counting rules in the START negotiations, any eventual START agreement will have new and different definitions and counting rules as well as improved formulations of past definitions and counting rules; and we would not want to give credence to past formulations that we have found to be inadequate. There is consensus, therefore, that we should say nothing publicly about definitions or counting rules or seek to negotiate these with the Soviets outside of START. Some believe, however, that we will need appropriate definitions and counting rules for internal use, both in assessing Soviet compliance and in assessing implications for US forces with regard to a new interim restraint regime. Others believe that counting rules and definitions are needed only for bilateral agreements. In the case of unilateral statements, intelligence assessment can and should be made on the basis of evidence, analysis and inference to judge Soviet compliance and implications for US force structure, not predetermined rules and definitions.

In any event, depending on the complexity of the regime, it may require some discussion with the Soviet side to establish the ground rules the US intends to employ with regard to assessing Soviet behavior under a new interim restraint regime.

Other Suggested Constraints Not Based on SALT

The study examined supplementing the President’s proposed restraint policy with new restraints. These fall into four general categories: limitations on US forces (generally in areas where we are currently ahead); limitations on Soviet forces (generally in areas where they are ahead); limitations on US forces that are explicitly conditional on reciprocal Soviet restraint (generally in areas where the sides have similar capabilities); and, confidence-building measures (CBMs), derived from those proposed by the US in 1982 and 1983 in the context of the START negotiations.

These restraints could be complemented by the SALT limits (or modifications of those limits), which were assessed as being positive in the matrix.

Options

A number of options have been constructed.

A. Current approach with its two limits on SNDV and strategic ballistic missile RVs at the Soviet level. This option would be the simplest approach and would leave all possible future options open, including the adoption of additional constraints pursuant to agreement on a START framework.

B. Additional constraints. This option, based on previous analysis, suggests possible candidates to supplement our existing regime. These include both SALT and non-SALT restraints which are either conditional on Soviet restraint, or unilateral on the part of the Soviet Union, in light of US restraint. These include:

— A US unilateral undertaking to limit its heavy bombers to 450 and not appreciably increase its planned ALCM deployments, given no significant change in the threat we face.

— A US offer:

— Not to test or deploy more than 10 RVs per ICBM or more than 14 RVs per SLBM.

— Not to test or deploy more than 20 ALCMs per heavy bomber, if the Soviets do the same.

— Not to encrypt telemetry on strategic ballistic missile flight tests or interfere with national technical means of verification if the Soviets do the same.

— A statement of US expectation that the Soviet Union would not:

— Significantly increase strategic ballistic missile warheads, SNDVs and throw weight.

— Expand their heavy ICBM force and not test new or modernize heavy ICBMs.

— Test or deploy more than 10 RVs per ICBM or more than 14 RVs per SLBM.

— Test or deploy more than 20 ALCMs per heavy bomber.

— Encrypt strategic ballistic missile telemetry or interfere with national technical means of verification.

— TWO CBMs:

— A data exchange derived from that proposed by the US in 1982 and 1983 on deployed strategic systems by type.

— Notification of all strategic ballistic missile launches.

C. Reductions approaches. Two reduction approaches have been suggested as possible candidates to supplement our existing regime:

(a) One variant which expands on the two constraints in the May 27 policy, would be for the US to announce that not only is the US committed not to deploy more SNDVs or strategic ballistic missile RVs than the Soviet Union, but it is prepared to restrict the number of its SNDVs and ballistic missile RVs to limits five percent below the current Soviet levels [less than 1 line not declassified], provided that the Soviet Union will reduce its SNDVs and ballistic missile RVs to the same limits. In this context, the United States is also prepared to restrict its deployed ALCMs to a limit five percent below the US planned deployment level [less than 1 line not declassified], provided that the Soviet Union will restrict their deployed ALCMs to the same limit.

(b) Another alternative would be for the US to announce that, as a clear demonstration of restraint (particularly in the context of a Summit), the US would be willing to dismantle strategic ballistic missile launchers containing 500 warheads on strategic ballistic missiles, on a one-time basis, if the Soviets do the same. The strategic ballistic missiles and their warheads would be removed and the launchers dismantled over a one-year period beginning on or about December 1, 1986. This would not necessarily involve any net reductions in strategic ballistic missile warheads or SNDVs. Nor would it involve any US commitment to a new strategic ballistic missile warhead or SNDV ceiling beyond the existing US commitment not to deploy more strategic ballistic missile warheads or SNDVs than the Soviet Union.

Pros and Cons for these options are presented in Section VIII.

Monitoring Considerations

Overall, the DCI representative judges that, given current Soviet practices, the three basic policy options can be monitored now by US intelligence [less than 1 line not declassified], depending upon the policy element in question. [3 lines not declassified] Whatever the situation, these confidences could be improved by what appear to be relatively easy and modest steps in the areas of definitions, counting rules, procedures, et al. Particularly for Option A, these steps could be entirely within the US Government. If these steps were undertaken, in combination with others to maintain a robust potential to react to Soviet actions, they collectively should lead fairly quickly to as strong a verification regime as it is possible to create in practical terms.

With regard to the specific options, it is clear that as these options progress in complexity, the monitoring difficulty increases as well. And the more complex Options B and C appear to require some sort of an understanding between the sides as to how weapons are to be counted and in the case of Option C, their launchers (or carriers) are to be removed from the inventory. Further, Options B and C increasingly imply the need for some discussion with the Soviets so that they understand the ground rules and standards of behavior the US intends to employ.

While not quarreling with many of the basic monitoring assessments presented by the DCI’s representative, others believe definitions, counting rules, and procedures to enhance verification would be difficult to develop and negotiate in the context of interim restraint—and could set negative precedents for START.

[Omitted here is the body of the report.]

Annex A (1)

Memorandum From Secretary of Defense Weinberger to the President’s Assistant for National Security Affairs (Poindexter)⁸

Washington, October 1, 1986

SUBJECT

• Military Sufficiency Assessment of Interim Restraint Options (U)

(S) NSDD 236 directed that there be “An assessment, led by the Secretary of Defense and the Joint Chiefs of Staff, of the answers [of the interim restraint study]⁹ in terms of their impact on U.S. and Allied security and military sufficiency.”

(S/MUSE) The attached paper by the Joint Chiefs of Staff concludes that all four interim restraint options under consideration are militarily sufficient under a number of assumptions which, among others, include:

— The United States strategic modernization program will be carried out to the degree permitted by each option.

— Adequate C3I will be available.

— No strategic defenses are deployed beyond the limits of the ABM Treaty.

The Joint Chiefs note that, “These assumptions are not made to prejudice policy, but are necessary if analysis is to be accomplished in a timely fashion . . .”

(S/MUSE) The JCS analysis is reasonable as far as it goes and in light of the assumptions made. However, there are a number of considerations which affect the desirability of the options under consideration. First, as the President stated on 23 September, the House of Representatives is not funding defense at an adequate level. While it is clearly impossible to predict precisely what will be funded over the next several years, it would be highly imprudent to assume complete funding of the strategic modernization program.

(S/MUSE) Second, the assumption that the Soviets will not deploy widespread missile defense capability may also be optimistic. The Soviets are in clear violation of the ABM Treaty and, as the President concluded in his December 1985 compliance report, the pattern of Soviet activity, “suggests that the U.S.S.R. may be preparing an ABM defense of its national territory.” Recent discovery of [1 line not declassified] by the U.S.S.R. and evidence concerning the [less than 1 line not declassified] reinforce this concern.

(S/MUSE) Third, when we consider the larger issue of U.S. and Allied security, we face the issue of proposing “cosmetic” or token reductions. The United States Government has consistently refused to do so since 1981. The Joint Chiefs of Staff note that, “The enclosed assessment does not prejudice the advisability of any of those options as appropriate arms control constraints for the United States.” As the President noted in his 27 May statement, “. . . no policy of interim restraint is a substitute for an agreement on offensive nuclear arms, provided that we can be confident of Soviet compliance with it. Achieving such reductions has received my highest priority.”

(S/MUSE) The dangers inherent in the Option C approach of a mutual 500 RV reduction proposal should be obvious. In short, it does not conform to our stated arms control principles: deep reductions to equal levels under verifiable agreements. If one factors in the underlying premise, i.e., that both we and the Soviets would be free to build back up to levels at or above those we had before the retirements, the cosmetic nature of this proposal would make even the “fatally-flawed” SALT II Treaty appear responsible by contrast. Furthermore, if one anticipates—as we must—that the Soviets will attempt to turn this proposal to their advantage (e.g., a freeze by both sides at the new, slightly lower but unequal levels), it should be self-evident that we must eschew this approach for military as well as arms control policy reasons.

(S/MUSE) Thus, we believe that Option A is the most militarily advantageous. Our existing policy is most consistent with our modernization program and meeting our military requirements. It gives us the most flexibility in responding to changes in the threat. It maximizes our leverage in the START negotiations. (The other options would tend to reduce Soviet incentives to agree to significant reductions.) It is consistent with the policy this Administration has pursued for almost six years—a policy which for the first time has forced the Soviets to agree at least to the principle of substantial reductions. We believe this policy must be continued.

Cap

Annex A (2)

Memorandum From the Acting Chairman of the Joint Chiefs of Staff (Welch) to the President’s Assistant for National Security Affairs (Poindexter)¹⁰

Washington, September 24, 1986

SUBJECT

• Military Sufficiency Assessment of Interim Restraint Options (C)

1. (S/MUSE) The Joint Chiefs of Staff have reviewed the four options developed in the NSDD–236 study. Attached is an annex to that study, certifying the military sufficiency of those options.

2. (S/MUSE) The Joint Chiefs of Staff have reached their conclusion based upon the specific assumptions and constraints described in the annex. Insofar as they are providing their military judgments of sufficiency prior to completion of the study, they are prepared to review and comment upon the study in its finished form in the near future.

For the Joint Chiefs of Staff:

LARRY D. WELCH
Acting Chairman
Joint Chiefs of Staff

Annex A (3)

Paper Prepared in the Joint Chiefs of Staff¹¹

Washington, undated

Annex: Military Sufficiency

NSDD 236 directs a study of mutual interim restraint, to include an assessment of the military sufficiency of candidate mutual interim restraint regimes. This annex addresses the military sufficiency of four options developed and described in detail within the body of the companion study: Option A (Current Approach), Option B (Additional Constraints), Option C1 (Percentage Reductions Approach), and Option C2 (Numerical Reductions Approach).

In anticipation of the Nuclear and Space Talks (NST) which commenced in 1985, the President reaffirmed earlier guidance that any approach or alternative approaches to those talks recommended for his approval should, as a minimum, permit the United States to develop and possess sufficient military capability relative to that allowed the Soviet Union, to permit execution of U.S. national military strategy with reasonable assurance of success. The President specifically directed the Joint Chiefs of Staff to certify the military sufficiency of any approach submitted to him in this regard.

For purposes of this study, the same general procedures used in arriving at military sufficiency judgments for NST options have been used in assessing candidate mutual interim restraint regimes. It is important to note that the military sufficiency judgments apply to regimes as a whole, not to elements of a regime or to regimes with elements other than those explicitly described in this annex. In addition, the following assumptions were made in arriving at the military sufficiency judgments:

○ The United States strategic modernization program will be carried out to the degree permitted by each option.

○ There will be no change to national security policy and strategy.

○ Adequate C3I will be available.

○ No strategic defenses are deployed beyond the limits of the ABM Treaty.

○ Soviet compliance, consistent with the nature of each option.

These assumptions are not made to prejudge policy, but are necessary if analysis is to be accomplished in a timely fashion and with sufficient detail to support the requirements of NSDD 236. The assumptions apply to the three time periods under review: through 1988, through 1991, and through 1996.

Option A (Current Approach) is the existing interim restraint regime which was established by the President’s 27 May 1986 announcement. The Joint Chiefs of Staff examined this option in terms of illustrative Soviet forces provided by CIA for purposes of this study. U.S. forces under this option would be constrained by the factors announced by the President:

○ The United States will continue to modernize its strategic nuclear forces, continuing to retire older forces as national security requirements permit.

○ Assuming no significant change in the threat, the United States does not anticipate any appreciable numerical growth in its strategic offensive forces.

○ The United States will not deploy more strategic nuclear delivery vehicles or strategic ballistic missile warheads than does the Soviet Union.

Option B (Current Approach with Additional Constraints) involves an expanded restraint regime. The Joint Chiefs of Staff also examined this option in terms of illustrative Soviet forces provided by CIA for purposes of this study. Soviet and U.S. forces in this option were constrained by a number of proposed limits. The major numerical limits were:

○ No increase in the current level of Soviet strategic nuclear delivery vehicles; the United States may deploy up to that level.

○ No expansion of the Soviet heavy ICBM force or testing of new heavy or modernized ICBMs, and no increase in Soviet ballistic missile throw-weight.

○ No increase in the current level of Soviet ballistic missile warheads; the United States may deploy up to that level.

○ No testing or deployment of more than 10 RVs per ICBM and 14 RVs per SLBM (a conditional restraint—the analysis assumed compliance by both the United States and the Soviet Union).

○ No deployment of U.S. heavy bombers beyond an operational limit of 450 (the analysis assumed staying within current plans).

○ No appreciable increase in planned U.S. ALCM deployment (the analysis assumed staying within current plans).

○ No testing or deployment of more than 20 ALCMs on a heavy bomber (a conditional restraint—the analysis assumed compliance by the United States and the Soviet Union).

The same general assumptions that applied to Option A also apply to Option B.

Option C1 (Percentage Reductions Approach) is a variation on expanding the two numerical limits expressed in the President’s 27 May statement. As with Options A and B, the Joint Chiefs of Staff examined this option in terms of illustrative forces provided by CIA. These forces were constrained by a five percent reduction below current Soviet levels of strategic nuclear delivery vehicles and strategic ballistic missile warheads. Thus, the limits for U.S. forces become:

○ The United States will not deploy strategic nuclear delivery vehicles or ballistic missile warheads to a level beyond ninety-five percent of current Soviet levels in these two categories, assuming the Soviets reduce and remain at that level.

○ The United States also will not deploy ALCMs to a level beyond ninety-five percent of its planned deployment level, in the context of the above limit and assuming the Soviets are prepared to show equal restraint.

Although this option allows for incorporating additional constraints of Option B, bounds had to be placed on the option for the purpose of supporting analyses. Thus the only numerical limits examined in reaching a military sufficiency judgment are those explicitly described above.

Option C2 (Numerical Reductions Approach) also is a variation on expanding the two numerical limits in the Presidents 27 May statement. The Joint Chiefs examined this option in terms of illustrative forces provided by CIA and constrained by a one-time reduction of 500 warheads on strategic ballistic missiles, with the reduction taking place from 1 December 1986 through 1 December 1987. It was assumed that the reduction would take place only if the Soviets agreed to do the same. Thus the limit on United States forces essentially is the same as in Option A, with an additional one-time reduction in strategic ballistic missile warheads. This option also allows for incorporating additional constraints of Option B; however, for the same reasons discussed earlier, such additional constraints were not considered in this assessment.

Conclusion: Given the assumptions and the numerical limits in the options laid out in this annex, the Joint Chiefs of Staff find that all four options permit the United States to develop and possess sufficient military capability relative to that allowed the Soviet Union, to permit execution of U.S. national military strategy with reasonable assurance of success. Therefore the Joint Chiefs certify with those understandings that the proposed mutual interim restraint options are militarily sufficient. This certification does not prejudge the Joint Chiefs of Staff assessment of the advisability of any of these options as appropriate arms control initiatives for the United States.

Annex B

Paper Prepared in the Central Intelligence Agency¹²

Washington, September 26, 1986

Annex B: Assessment of Monitorability and Verifiability

Overview

Overall, the DCI Representatives judges that, given current Soviet practices, the three basic policy options can be monitored [less than 1 line not declassified] depending upon the policy element in question. [2½ lines not declassified]. Whatever the situation, these confidences could be improved by what appear to be relatively easy and modest steps in the areas of definitions, counting rules, procedures, et al. Particularly for Option A, these steps could be entirely within the US Government. The associated verification judgments ought to track directly with such steps. If these steps were undertaken, in combination with others to maintain a robust potential to react to Soviet actions, they collectively should lead fairly quickly to as strong a verification regime as it is possible to create in practical terms. [portion marking not declassified]

As the three basic policy Options progress in complexity, the monitoring difficulties increase as well. At least one element of each Option implies creation of a verification regime. And the more complex Options B and C seem to require some sort of understanding as to how weapons are to be counted and/or their launchers (or carriers) are to be removed from the accountable inventory. [portion marking not declassified]

Further, Options B and C seem to increasingly imply the need for discussion with the Soviets so that they understand US standards of behavior and so that the US does not employ a standard of behavior which the Soviets have rejected already or would reject if it were exposed to them. And finally, some elements of Options B and C—without an explicit verification package of some type, directly connected to these constraints—seem to imply political and military risk beyond what the US has judged itself willing to accept in the recent past, although comparable risks from pre-existing accords have been judged tolerable for short periods of a few years. [portion marking not declassified]

Supporting definitions and other features, such as counting rules, are judged necessary to ensure as strong a verification regime as possible for these options. The regime needed varies somewhat from option to option and from element to element in each option. A few, such as the limit on SNDVs, are relatively easy to express, based on our experience in monitoring Soviet forces and tacit “rules of the road” that have developed over the years and seem to be understood by both the US and USSR. For these, the US could unilaterally express the standard of behavior for the Soviets which—should they choose to participate—they could readily accept and understand. [portion marking not declassified]

However, most of these specific elements—for example, RV counting, throw weight, and maximum ALCM loadings—involve difficult technical, and controversial political, issues resulting in some monitoring uncertainty and associated military risk. [8½ lines not declassified]

Introduction

1. Policy Options: This annex examines the monitorability and verifiability of the variety of possible elements for a new US interim restraint policy, including those of the President’s May 27 statement, selected SALT limits, and other possible limits. These elements are currently embodied in three policy options. [portion marking not declassified]

To support whichever option is chosen, US intelligence will have to perform the monitoring tasks outlined below:

— Option A: Assess whether there has been any increase (or decrease) in deployed Soviet strategic nuclear delivery vehicles (SNDVs) and strategic ballistic missile warheads;

— Option B: Determine whether the Soviets were:

— Significantly increasing (or decreasing) strategic ballistic missile warheads, SNDVs, and throw weight.

— Expanding their heavy ICBM force and not testing new or modernized heavy ICBMs.

— Testing [or deploying] more than 10 RVs per ICBM or more than 14 RVs per SLBM.

— Testing [or deploying] more than 20 ALCMs per heavy bomber.

— [Encrypting strategic ballistic missile telemetry or interfering with national technical means of verification.]

— There are two confidence building measures included in this option:

— [A data exchange derived from that proposed by the US in 1982 and 1983 on deployed strategic systems by type.]

— [Notification of all strategic ballistic missile launches.]

— Option C: In addition to portions of Options 1 and 2, or by itself, determine whether the Soviet Union was:

— Reducing its deployed SNDVs and ballistic missile RVs by 5 percent, and restricting its deployed ALCMs to a limit 5 percent below the US planned deployment level (about [number not classified]; or

— Dismantling strategic ballistic missile launchers containing 500 warheads on missiles over a one-year period. [portion marking not declassified]

2. [1 line not declassified]

— [4 lines not declassified]

— [2 lines not declassified]

— [2 lines not declassified]

— [2 lines not declassified]

— [1 line not declassified]

These stages make it possible to offer verification assessments. [portion marking not declassified]

3. Value of Monitoring/Verification Assessments: The above factors are the necessary inputs for a considered verification assessment of the interim restraint options. It is clear that as the options progress to include more complex restraints involving greater monitoring uncertainties, it becomes more necessary to provide detailed verification measures, tailored to those restraints, to maximize the capability of the US to know whether or not the Soviets may have engaged in activities (including non-compliance) in enough time to allow the US to exercise the safeguards put in place to cope with Soviet behavior. Depending on the monitoring uncertainties, cooperative measures to reduce uncertainties, appropriate safeguards, and a JCS assessment of military sufficiency of each policy option, it ought to be possible to make an assessment of the acceptability of any particular verification regime. [portion marking not declassified]

Assessment of Option A

[7 lines not declassified]

5. [1 paragraph (11 lines) not declassified]

6. [1 paragraph (12 lines) not declassified]

7. [1 paragraph (9 lines) not declassified]

8. Cooperative Measures: Although there are no specific cooperative measures attached to this option, it is clear that there needs to be some type of an understanding concerning terms, definitions, counting rules, and procedures. This could be entirely within the US Government or between the US and USSR. Without this understanding, not only would the Soviet Union not be able to know how it should act to satisfy the restraints, but US intelligence would have unnecessary difficulty in accurately assessing Soviet activities. Further, without such elements of a verification regime, a disparity could arise between the operational force and the force defined under the policy option. SNDVs taken “off line” for whatever reasons might still be judged accountable. [portion marking not declassified]

9. In the case of SNDVs, it is clear that both sides have a good understanding of what constitutes a SNDV in today’s world. A unilateral statement by the US to this effect could suffice. For RVs, however, it would not be possible to improve the prospects for verification without some sort of communications with the Soviets, either through diplomatic channels or in the SCC. Because of the different ways a side could determine the maximum number of deployed RVs on each type of fractionated missile, it is judged essential for assessments that it be made clear to (or with) the Soviets what criteria it is using to determine the maximum RV loadings. For example, it is possible to use the maximum number of RVs flight tested with (or without) the maximum number of simulated releases on a missile type during flight tests. These particular illustrative criteria, however, would not necessarily capture the maximum number. Monitoring confidences would not be likely to improve a lot—and for all types of missiles—simply because of this one counting rule, however, as long as Soviet telemetry encryption continues at its current levels. [portion marking not declassified]

10. [1 paragraph (20 lines) not declassified]

11. [1 paragraph (14 lines) not declassified]

12. Possible Safeguards: In order for the US to be in the strongest possible verification situation for this Option, the US needs to decide on (and create and declare as necessary), as well as put in place, the following steps:

— A vigorous effort on options for US strategic forces; this would include RDT&E, as well as deployment, of all elements of the Strategic Modernization Program.

— A strong national intelligence program, which incorporates the capability to monitor SNDVs and to estimate the RV loadup on deployed operational ballistic missiles. Some additional funding, now undetermined in cost and duration, in addition to restoring cuts directed so far, might be prudent.

— A policy statement that the US will react politically and/or militarily to significant changes in either level.

To the degree that these steps already exist, sufficient (and perhaps strong) safeguards are in place for Option A. These safeguards could be strengthened by extending these steps into an Executive-Legislative understanding. One conceivable way to do this is by means of a binding joint resolution of Congress. [portion marking not declassified]

13. Verification Assessment: [2 lines not declassified]. Moreover, a number of steps could be created which, if undertaken, in combination with actions to maintain at least a plausible potential to react to Soviet actions, ought to lead to as strong a verification regime for this option as it is possible to create in practical terms. Depending on the JCS assessment of military sufficiency, which is not available at present, this verification regime might be at least acceptable. [portion marking not declassified]

Assessment of Option B

14. Relation to Option A: Option B contains all the restraints in Option A but adds other restraints, as well as confidence building measures. [portion marking not declassified]

15. Monitorability: With regard to SNDVs, the assessment for Option A above is unchanged, except to note that a data exchange on deployed strategic systems by type would marginally assist US intelligence in its accounting of SNDVs, particularly land-based mobile ICBMs. To be most practical and useful, a data base should be established at the beginning of the interim restraint regime with data exchanges of updates repeated every 6 months. Also helpful would be notification of SNDV launchers that enter and exit the inventory and new types of systems coming on line. Notification of a few days or weeks would suffice for the first; several months or a year’s notice would be helpful for the second. Any such data exchanges and notifications could take place via diplomatic channels or through the SCC. [portion marking not declassified]

16. Ballistic Missile RVs: The assessments concerning the monitoring of ballistic missile RVs would differ in this option from Option A. This option includes an implicit end to telemetry encryption, [7 lines not declassified]

17. [1 paragraph (14 lines) not declassified]

18. [1 paragraph (20 lines) not declassified]

19. [1 paragraph (12 lines) not declassified]

20. [1 paragraph (13 lines) not declassified]

21. [1 paragraph (4 lines) not declassified]

22. Cooperative Measures: Among the measures that would be needed for this wide-ranging option to be monitored [less than 1 line not declassified] would be: an agreement or understanding as to what constitutes throw weight (an issue of past contention), as well as what constitutes a new heavy or modernized ICBM (also an issue of contention). Moreover, an end to encryption, a data exchange periodically updated on all the types of deployed strategic systems, and notification of ballistic missile launches would assist US monitoring of this overall option. An end to encryption would have a particular impact in [less than 1 line not declassified] the estimates of the numbers of RVs and the amount of throw weight flight tested. Further, on-site inspection/monitoring [less than 1 line not declassified] could confirm that an ICBM (or SLBM) had been deployed with no more than 10 (or 14) RVs. And the data from the on-site inspection/monitoring would be valid only during the time of the inspection/monitoring. It is clear, however, that the list of restraints noted above and these cooperative measures would require at least a discussion with the Soviets. The measures would also be most beneficial to monitoring efforts if they were kept together as a package. [portion marking not declassified]

23. Significance of Violations: The same risks and projections that apply in Option A also apply to Option B. In addition, although there would be a significant risk to US security were there a major Soviet expansion of its heavy ballistic missile force, and new heavy or modernized ICBMs (assuming no US response), [less than 1 line not declassified]. Any resulting increases ought to lead to only slightly more military risk for the US in this option than they would in any other option. [6 lines not declassified]

24. Possible Safeguards: Approximately the same as in Option A; the details ought to follow from the policy elements selected. [portion marking not declassified]

25. Verification Assessment: The DCI Representative believes that the verifiability of this wide-ranging option (as stated) ought to be [less than 1 line not declassified] depending on which element(s) are selected and how they are combined. Moreover, a number of steps could be created which, if undertaken in combination with actions to maintain at least a plausible potential to react to Soviet actions, ought to lead to as strong a verification regime as it is possible to create in practical terms. Depending on the JCS assessment of military sufficiency, which is not available at this time, this verification regime might be at least acceptable. [portion marking not declassified]

Assessment of Option C

26. Monitorability: [3 lines not declassified]. However, procedures can be devised that would allow [less than 1 line not declassified] confidence in monitoring the number of launchers and missiles removed from the USSR’s operational inventory. But this would require, among other things, specific procedures governing removal from deployment (e.g., dismantling and destruction) of the weapon systems and launchers in question. [2 lines not declassified] it is assumed that the dismantling and destruction of missiles and launchers would also involve at least the removal of the warheads from their operational launcher. [portion marking not declassified]

27. Monitoring the removal from deployment depends on the steps the Soviets use. For example, monitoring the destruction of fixed silo launchers, SSBNs, and large aircraft, is relatively easy and can be done [less than 1 line not declassified] if the USSR cooperates. Such monitoring for mobile missile launchers is somewhat harder [less than 1 line not declassified] unless the USSR cooperates. [2½ lines not declassified]. In contrast, even with past types of cooperation, the destruction of missile airframes requires determination of the types of missiles destroyed. This is particularly important for “newer” missiles—such as the SS–17, SS–18, SS–19, SS–25, and SS–X–24, [5 lines not declassified].

28. [1 paragraph (9 lines) not declassified]

29. Programmed changes or additions to US intelligence capabilities will help the monitoring of this option somewhat. Monitoring here is much more dependent upon appropriate cooperative measures, properly designed and implemented. [portion marking not declassified]

30. Cooperative Measures: In the absence of active cooperative measures, agreement on detailed passive cooperative measures is necessary for successful monitoring of this option. Specifically, agreement ought to be reached on at least:

— Prenotification of reductions, dismantlements, and destruction (including locations, if possible).

— Procedures for destroying launchers or cutting up airframes.

— Length of time for such activities and for display to NTM.

— If necessary, positive identification of the airframe.

— Disposition of remaining (including destroyed) parts.

These types of procedures have been a source of contention on occasion between the Soviets and the US and ought to require more detailed discussion. [4 lines not declassified]

31. Significance of Violations: With a SNDV total of about [number not declassified] (depending on the definition of SNDV), and a total deployed ballistic missile warhead count in the range of about [less than 1 line not declassified] (depending on counting rules), there ought to be minimal (if any) military significance to a Soviet failure to reduce either 5 percent in SNDVs (i.e., [number not declassified]; or 500 warheads. [portion marking not declassified]

32. Possible Safeguards: Approximately the same as in Option A; the details ought to follow from the policy elements selected. [portion marking not declassified]

33. Verification Assessment: The DCI Representative believes, [2½ lines not declassified]. These steps and others could be created which, if undertaken in combination with actions to maintain at least a plausible potential to react to Soviet actions, ought to lead to as strong a verification regime for this option as it is possible to create in practical terms. Depending on the JCS assessment of military significance, which is not available at present, this verification regime might be at least acceptable. [portion marking not declassified]

1. Source: National Security Council, National Security Council Institutional Files, Box SR–093, NSDD 236. Secret; Muse. Sent for action. Copies were sent to Cockell, Mahley, and Brooks. A stamped notation at the top of memorandum reads: “Signed.”
2. See Document 129.
3. An unknown hand crossed out “Tuesday, October 7,” drew a line to the right-hand margin, and wrote “Monday Oct 6.”
4. An unknown hand crossed out “Tuesday’s” and wrote “Monday’s” above it.
5. Poindexter approved the recommendation.
6. Secret; Muse.
7. Secret; Noforn; Wnintel; Muse.
8. Secret; Muse. Weinberger wrote “John” above the recipient line.
9. Brackets are in the original.
10. Secret; Noforn; Muse.
11. Secret; Noforn; Muse.
12. Secret; [handling restrictions not declassified]. Brackets, except those indicating material not declassified, are in the original.