Foreign Relations of the United States, 1969–1976, Volume XXXII, SALT I, 1969–1972

14. Paper Prepared by the Interagency SALT Steering Committee¹

Washington, undated.

SUMMARY OF NSSM–28 REPORT

I. Introduction

In response to the NSSM–28 directive,² we have approached the problem of Strategic Arms Limitation Talks (SALT) with the USSR from three separate but related perspectives. We have considered (1) general considerations relevant to any negotiations; (2) specific issues arising from peculiarities of different weapons systems and their interrelationships; and (3) several illustrative packages of arms restrictions which might serve as the basis for a U.S. position.

Many of the issues to be decided in connection with this report are closely linked to the issues raised in NSSM–3, Review of U.S. Strategic Posture.³ Indeed, decisions on certain weapons systems for our posture will affect which specific weapons systems could be controlled under an agreement with the USSR.

II. General Considerations

A. U.S. Objectives

There is general agreement that the primary military purposes of our strategic forces are:

• —to reduce the likelihood that nuclear war will occur;
• —to protect ourselves and our Allies from the destructive consequences of nuclear wars, in so far as we can; and
• —to be capable of controlling strategic nuclear conflict so that the possible outcomes leave the United States and its Allies in a relatively advantageous position.

These objectives are also, in part, criteria for judging a strategic arms limitation agreement with the Soviet Union. This report recognizes that these basic objectives can be pursued unilaterally, without any agreement with the USSR. Certain arms control measures in the realm of strategic systems, however, could help achieve some of our objectives at lesser costs. They might contribute to a reduction in political tensions and some of the uncertainties surrounding strategic competition. Depending on the nature of the agreement, however, some new uncertainties might be created.

B. The Scope of an Agreement

A basic issue is whether it is in our interest to aim for a measure limited in scope, perhaps only symbolic in nature, or whether we should seek a more comprehensive agreement designed to stop as much of the strategic arms competition as possible. Stated another way, what restrictions would we be willing to accept on our own ongoing programs and future options in order to inhibit Soviet programs and future options. The principal implications of each approach are:

1. Limited Measures

Should we decide to seek a minimum agreement affecting only one or two systems, leaving aside more complex problems and excluding some weapons systems or future programs altogether; as examples we might propose:

a. a freeze on the number of fixed land-based ICBM’s only, or

b. a freeze on the total number of ICBM’s, plus a freeze on the initiation of further construction of submarine-launched ballistic missiles; or

c. a limit on land-and sea-based ICBM’s to total numbers now under construction, but with freedom to vary the mix.

These and other variants are discussed and analyzed in Section V of this summary (Alternatives I and II).

2. Intermediate Measures

Should we decide to go beyond the kind of agreements outlined above, we could consider the following:

A proposal (Alternative III), which not only freezes land and sea-based systems at existing levels (including those under construction), but also bans enlarging, modifying, or relocating missile silos or launchers, and prohibits mobile missiles.

3. Comprehensive Agreements

Should we decide to seek more comprehensive coverage of qualitative as well as quantitative aspects, we could add a prohibition on MIRVs to the intermediate type program outlined above (Alternative IV).

Or, we could go still further and base our position on a quantitative and qualitative freeze on all aspects of strategic offensive and defensive systems that are subject to adequate verification by national means. In effect this would mean that we would stop where we are and the Soviets would do likewise.

In order to choose among these alternative approaches, several specific issues will have to be resolved. The main ones are:

(1) what level of ABM defenses we would agree to, under whatever offensive limitations we propose;

(2) whether to retain or prohibit MIRVs;

(3) what level of confidence we consider prudent in our ability to verify agreements.

Also, depending on our ultimate negotiating objectives, we should decide what moratoriums, if any, we will seek during negotiations.

C. Moratorium

With regard to the issue of what will be permitted while negotiations proceed, the approach we adopt will be affected by decisions on which systems we hope to control in a final agreement. In addition, we would have to decide how long to maintain a moratorium. We have several choices:

1. No moratorium;

2. A moratorium on offensive deployments only (i.e., construction of offensive systems—land-based ICBMs and/or ballistic submarines);

3. In addition, a moratorium on ABM construction;

4. In addition, a moratorium on flight testing of new offensive systems, including MRVs, MIRVs etc., or MIRVs could be treated separately.

The principal arguments are:

On the one hand,

• —negotiating a moratorium might prove as complicated and contentious as negotiating a final agreement;
• —without any moratorium, the U.S. would be free to continue programs which may be necessary should talks fail;
• —a moratorium could prove difficult to terminate.

On the other hand,

• —a moratorium could restrict some of the Soviet buildup in offensive systems;
• —if we want to ban some systems, especially MIRVs, a moratorium on testing might contribute to the achievement of a final agreement;
• —if negotiations are protracted, continuing deployments could make an agreement unlikely.

D. Verification

The basic issue is whether we should rely solely on national means to verify an agreement or should insist on some means of on-site inspection.

An intelligence estimate⁴ concluded that there are aspects of potential agreements that can be verified by national means with high confidence. While some Soviet violations could get underway without our knowledge, detection in most cases would be assured prior to the creation of a substantial Soviet buildup.

Our confidence in Soviet compliance in certain areas could be increased if we could obtain some form of on-site inspection or other additional means of assurance. For example, inspection could be very helpful in dealing with problems such as illegal upgrading of a defense system (Tallinn), retrofitting of ICBMs and IR/MRBM launchers, and deployment of MIRVs on operational missiles. Such inspection could not only provide hard evidence of violations but more important would act as an additional deterrent to attempts at evasion.

It should be noted that national means of verification involve highly sensitive intelligence sources and methods and for a U.S. administration to be able to demonstrate publicly that the Soviets were complying with, or evading an agreement could involve compromising these sources or methods.

Insistence on on-site inspection, however, could pose a major obstacle to an agreement. At a minimum, efforts to obtain on-site inspection would probably lead to prolonged negotiations. Nevertheless, the Soviets might accept token inspection if this were the only obstacle in the way of an agreement they considered highly favorable. They might accept a mechanism for examining or adjudicating problems, which, in time, could lead to an informal, but not institutionalized, inspection system.

In addressing the verification issue, we should assess the extent to which the range of uncertainties in our national verification capabilities, and their associated risks, would affect the net military posture of the U.S. and Soviet Union.

The areas in which there may be verification problems are:

1. Destruction of Silos and Relocation of Launchers. There is no way national means can determine that a silo has been deactivated unless there is an agreed procedure for doing so, such as cratering.

2. Mobile Missiles. If mobile missiles were permitted up to a prescribed total, they could not be counted accurately, although we could estimate the general magnitude of the force. The extent of possible violation before detection would be dependent upon the agreed limit. Since detection of a single mobile missile would indicate a violation, we could verify a ban with considerably greater confidence, although the Soviets might be able to deploy as many as 200–300 before we would detect the violation. Thus, a ban would constitute a greater deterrent to cheating than would limitation to an agreed number.

3. ABM’s. Stationary ABM launchers can be detected with high confidence; mobile ABM’s would present the same problems described above. Depending on how defined, verification of ABM-associated radars could pose problems. However, if there is no limit on ABM-associated radars, warning of violations of agreed limits on launchers would be greatly reduced, particularly if mobile launchers are not prohibited. We would have high confidence of detection and probable identification of the conversion of a surface-to-air (SAM) system to an ABM prior to IOC. Lesser modifications to give a SAM system a limited ABM capability (e.g., for a point-in-space intercept) would be more difficult to detect, and some might escape detection completely.

4. MIRV’s. If flight-testing continues to the point where either side feels that the other might be able to deploy MIRV’s without further flight-testing, verification of a MIRV ban by unilateral means would be impossible. Verification of a ban on MIRV deployment could then be accomplished only by highly intrusive inspection of the missile itself. Notwithstanding the net judgment of the Intelligence Community that MIRV tests would be detected, some argue that our assurance in verifying a MIRV prohibition through a ban on MIRV flight-testing could be uncertain. They argue that such uncertainties would result because of the question of defining exactly what is an MIRV, since the Soviets may have a system different from ours; they believe that to reduce such uncertainty it would also be necessary to ban flight-testing of all systems related to MIRVs, including some penetration aids which the U.S. might want to develop as a hedge against cheating or abrogation of an agreement, and including flight-testing of systems which the U.S. has already deployed.

5. IR/MRBM Upgrading. We might not be able to detect the deployment of ICBM’s in the existing 135 hard IR/MRBM silos.

E. Soviet Objectives

Soviet objectives in arms limitation talks are probably to determine if it is possible to obtain an agreement which preserves and perhaps improves their present relative strategic position and which enables them to avoid being drawn into a costly round of strategic procurements in response to reprogrammed and potential U.S. offensive and defensive deployments. There is reason to believe that the present balance of Soviet leadership favors such talks and sees value in a possible agreement. However, they may also hope that talks by themselves would serve to delay the introduction of new U.S. weapons systems and perhaps weaken Allied unity and concern for strengthening NATO. Thus, the Soviets may well have several objectives which are not mutually exclusive and which could be realized simultaneously.

F. Negotiability

Little is known of Soviet views as to the details of a strategic arms limitation agreement. They have agreed that it should apply to both offensive strategic delivery vehicles and ABM’s, and that the first step should be a limitation and not a reduction of armaments. From our past experience with them in arms control negotiations, we can expect them to advocate an agreement of broad coverage. This does not mean that they would never accept a limited measure, but at least they are likely to insist that any limitation placed on their ongoing or planned strategic weapons programs be matched by comparable restrictions of our own.

In this connection, the alternative packages we have considered contain a number of features that may be challenged by the Soviets. Principal among these are: that bombers should not be included in an agreement; that Soviet mobile missiles should be banned or limited; that IRBM’s and MRBM’s should be limited, but British and French weapons and U.S. tactical nuclear weapons in Europe should not be included.

G. U.S.-Soviet Political Relations

The conclusion of an actual agreement to limit strategic arms could have far-reaching effects on U.S.-Soviet relations. Most of these changes would be of a long-term and intangible nature, so precise predictions are not possible.

H. NATO Reactions

Our NATO allies favor the initiation of SALT. Doubtless, the majority of Europeans would welcome an actual agreement as heralding a more meaningful détente in East-West relations. However, our allies will have certain specific concerns, which they would expect to be resolved in the course of consultations and negotiations. They will be primarily concerned with those elements of a possible agreement directly affecting their security interests; e.g., its effect on the U.S. deterrent, the restrictions it places on Soviet IR/MRBM’s (frozen under all packages), and its effect on third-country nuclear forces and U.S.-controlled tactical nuclear weapons in Europe (none under any package). Our allies would also be concerned that an agreement not adversely affect our capabilities against Soviet weapons systems targeted upon them. Given a satisfactory answer to these questions, our allies would probably prefer a comprehensive agreement, though they would probably welcome an agreement which was at least a significant first step toward curbing the strategic arms race.

The SALT talks will focus attention of the NATO Allies on serious strategic questions, such as the effect of a SALT agreement on the deterrence of conventional war in Europe and the implication of a U.S.-Soviet condominium. Such concerns may grow as the negotiations proceed.

I. Cost Savings

It is difficult to estimate with any precision what cost savings, if any, would result from the various types of agreements we have considered (in Section V, Alternatives I–IV). Rough calculations, comparing planned U.S. strategic budgets with the various illustrative agreements indicate that savings would probably not exceed $0.5–1.0 billion a year for the first five years after an agreement. Budgetary savings beyond that time could be greater, depending on whether the U.S. could forego deployments of certain systems that we might otherwise have to make and on what programs were adopted as hedges against abrogation.

Estimates for the USSR also do not show major savings in the short-run.

III. Weapons Systems Issues

The following major specific issues related to weapons systems or elements strongly influence the nature and scope of positions which the U.S. would find preferable or acceptable. These issues are discussed more fully in Section IV of the basic study report.

A. ICBM’s and SLBM’s

1. We have no plans to develop land-mobile ICBM’s, while the Soviets may be close to a capability to deploy them. Even though the Soviets may oppose a ban, we should attempt to ban, rather than limit, these systems since their deployment could create large uncertainties in the strategic balance, and an agreed limit would be considerably more difficult to verify than a total ban.

2. The potential vulnerability of fixed ICBM’s to counterforce attacks in the 1970’s will need to be addressed. Alternative considerations which could alleviate this problem include the following:

a. a MIRV ban would substantially delay and mitigate, but not necessarily eliminate ICBM vulnerability (because of the possibility of improvements in missile accuracy and payload).

b. ABM defenses of ICBM sites could help maintain ICBM survivability, but ABM’s could also adversely affect our retaliatory capability by providing protection for Soviet cities.

c. Superhardening of silos could be permitted, but we would have to work out replacement procedures relating to verification of destruction of old silos.

d. Limiting throw-weight of offensive missile forces, in which the Soviets have a current lead, could help reduce future threats to fixed ICBM’s, although improved missile accuracies could still lead to problems. This measure would pose problems of definition and verification.

e. If practicable, preventing significant improvements in missile accuracies could minimize future threats, but this is a limit on technology, and verification would be complicated, if not impossible. Development of high accuracy probably can be verified, but the degree of accuracy attained may not be known.

3. We may want to allow freedom to mix land- and sea-based missile forces within an agreed number of total launchers. If greater fractions of each side’s forces are moved to sea, this could have the effect of reducing potential instabilities, principally of fixed ICBM vulnerability, and decreasing incentives for both sides to pursue counterforce options but, unless there were agreed procedures for substitution, could create verification problems and lead to uncertainties as to ultimate force levels.

4. Since we are well ahead in numbers of SLBM’s, we may want to impose limits to check the current Soviet buildup. As a means of simplifying an agreement or on the grounds that these are stabilizing retaliatory weapons, we could choose to impose no limits on SLBM’s, but this would leave the arms race wide open in one important area.

B. MRBM’s and IRBM’s

The Soviets have about 700 MRBM’s and IRBM’s, and we have none. We want to freeze further construction and prevent them from being further hardened, made mobile, or converted to ICBM’s. This is clearly in NATO’s interest as well as our own, since it would limit a targetable threat.

C. ABM’s

1. Although there could be some tactical negotiating benefits in not offering an ABM level initially, it is necessary for the U.S. to establish a minimum and maximum ABM level which it would prefer or be willing to accept in the context of an overall agreement. The Soviets already have a small ABM system, which complicates prospects for a possible total ABM ban.

2. Under an agreement, the higher the ABM level the more tendency there is for each side’s retaliatory capabilities to be eroded for a given set of offensive limitations.

In general, however, this erosion of capabilities is more drastic for the USSR than the U.S., with or without MIRVs.

A decision on acceptable levels of ABM is closely related to a corresponding decision on whether both sides are free to deploy MIRV’s and develop penetration aids. In general, our analysis indicates high ABM levels rule out banning MIRV’s, and lower levels of around 200–500 or so ABM interceptors would be compatible with a ban on MIRV’s.

3. The possibility that both the U.S. and the Soviet Union may wish to deploy an ABM for defense of their population against Chinese ICBM’s or accidental launches could establish a lower limit on the acceptable levels of ABM’s.

4. To the degree that the potential vulnerability of ICBM’s may require ABM’s also for defense of ICBM sites, the overall ABM level may have to be raised or else separate limits placed on ICBM defenses to the extent practicable.

5. If ABM associated radars were limited, it would increase our confidence in our ability to verify limits on ABM systems. If not, they could contribute significantly to a capability to execute rapidly a planned abrogation of ABM limits or to provide clandestinely some ABM capability for Soviet surface-to-air missile systems.

6. The implications of alternative ABM levels can be summarized as follows:

a. Low Level (0–500). Limiting to low levels will be necessary under a MIRV ban which also bans sophisticated penetration aids, with a level of around 200–500 being preferable in order to reduce the sensitivity of our retaliatory capability to possible Soviet cheating. Might not be sufficient for damage denial against potential Chinese ICBM threat. Could protect a significant fraction of ICBM sites if MIRV’s are banned.

b. Medium Level (500–1500). With MIRV’s, the U.S. could maintain its retaliatory capability over this range, but the Soviets might find the upper levels unacceptable because their retaliatory capability with missiles would be sharply reduced. Would allow damage denial against Chinese ICBM’s. At upper range, could provide reasonable defense of ICBM sites.

c. High Level (1500–2000). With MIRV’s, the U.S. could maintain its retaliatory capability, but such high ABM levels could create uncertainties in hedging against Soviet evasions and make the U.S. retaliatory capability more sensitive to Soviet cheating. Probably less acceptable to Soviets, at least on military grounds, since their retaliatory capability would be severely degraded. Would permit damage denial against Chinese ICBM’s and extensive U.S. ICBM site protection.

D. MIRV Limitations

1. The conditions, if any, under which the U.S. might prefer or accept a ban on MIRV’s involve complex judgments. A MIRV ban has implications for related technology (such as penetration aids and maneuvering reentry vehicles), the consequences of which must be taken into consideration. A MIRV ban could delay or diminish the potential threat to ICBM survivability. With a low ABM level, this could enhance stability. However, there could be greater sensitivity of our retaliatory capability to possible levels of Soviet cheating or abrogation compared to cases in which MIRV’s are allowed. This sensitivity would be minimized if Soviet ABM capabilities were tightly constrained. The Soviets may prove unwilling to stop their offensive missile buildup without a MIRV ban.

2. The urgency associated with a U.S. position on MIRV’s stems from the fact that continued U.S. MIRV flight-testing and possible Soviet MIRV testing before or during negotiations could make it difficult, if not impossible to reach an enforceable MIRV deployment ban. Actual MIRV deployment cannot be verified without intrusive on-site inspection of the launchers and examination of missile components. We may, therefore, want to consider proposing a moratorium on MIRV testing either prior to or during negotiations in order to keep open the option of an effective ban. This step could greatly increase our chances of obtaining a moratorium on ongoing Soviet programs, particularly ICBM and SLBM construction.

3. A number of considerations suggest that a MIRV ban may not be desirable or negotiable even if desirable.

a. The Soviets may not be interested in a MIRV ban, either because they want MIRV’s themselves or are willing to let us have MIRV’s in exchange for something they want, such as mobile ICBM’s or more SLBM’s.

b. Notwithstanding the net judgment of the Intelligence Community, some argue that our assurance in verifying a MIRV prohibition through a ban on MIRV flight-testing could be uncertain.

c. There may be complications in determining what is to be banned since the Soviets may have a system somewhat similar to one which we have already deployed (the Poseidon A–3).

E. Bombers and Air Defense

1. The Soviets may well raise this issue and as a minimum we need a contingency position. The issue is whether the U.S. should be prepared to accept limits on bombers or air defense if the Soviets insist, and if we could reach agreement on the definition of the systems to be included.

2. Bombers can provide a stabilizing hedge against cheating. However, limiting bombers would close a loophole in the arms race. In the absence of undetected Soviet cheating, a freeze on bombers could preserve the U.S. position in an area in which we have a major advantage. It would, however, also open the complex air-defense problem, which may be non-negotiable with the Soviets. At the same time, a ban on new or upgraded air-defense systems could increase our confidence in verification of limitations on ABM deployments.

IV. Results of Strategic Analysis

Analysis using war-gaming techniques and a range of scenarios produces the following game results concerning the effects upon programmed U.S. and projected Soviet forces of the various arms control proposals through 1978. (For a fuller discussion, see Section V–C and Annex A⁵) The results of this type of war-gamed strategic analysis provides one of the tools for arriving at strategic force judgments but not an exclusive means.

A. Under all the cases considered, U.S. retaliatory capability remains over 30% except in the cases where a MIRV ban is combined with area ABM levels over 500. In these cases, U.S. retaliatory capability remains over 25%. At these higher ABM levels, U.S. retaliatory capability is increasingly dependent on bombers, and is entirely so above ABM levels of 2000.

B. Under all the cases considered, Soviet retaliatory capability remains high except for area ABM levels above 500. At levels above 500, there are points at which Soviet retaliatory capability declines to levels presumably unacceptable to the Soviets. For Package I, this point would not be reached until somewhere between 1000 and 2000 ABM’s, because of the high level of SLBM’s it is assumed the Soviets would deploy under this option. For the other packages, the point would probably be reached between 500 and 1000 ABM’s. Based on these calculations, we believe the Soviets would prefer a lower limit on ABM’s, and should favor the more comprehensive packages, including a prohibition of MIRV deployment. Other, non-military factors would also be involved in determining the Soviet position.

C. At the higher ABM levels examined in the strategic analysis, the U.S. could save a significant number of fatalities by striking first, but would still suffer unacceptable fatalities. In all cases examined, the Soviets could not appreciably reduce fatalities to themselves by striking first. Based upon these considerations, neither side would have an incentive to strike first.

D. Predicated on U.S. estimates of our intelligence capabilities and hedges available, Soviet cheating could be overcome in every case. However, under a MIRV ban the U.S. might be more sensitive to substantial Soviet cheating than it would in the absence of such a ban. In case of Soviet substantial cheating on ABM’s under a MIRV ban, the U.S. might have to deploy MIRV’s in order to retain a confident retaliatory capability that does not rely on bombers alone.

V. Illustrative Arms Control Packages

We have studied several illustrative arms control restriction packages consisting of four options with three variants. These packages range from relatively limited to relatively comprehensive agreements, and the analyses in Sections IV and V of the basic report also permit the selection of other combinations which may be preferable to any of these packages. (Not yet studied is the Alternative of both sides stopping where they are.)

While all of the packages contemplate some limit on numbers of ABM launchers, none of them includes a specific designation of ABM level; a range of different ABM levels was analyzed for each package. None of the illustrative packages restricts bombers, air defense, or missile throw-weight and accuracy improvements, and none of them calls for force reductions. All of the packages limit numbers of ICBM’s, prohibit mobile IR/MRBM’s, and limit the numbers and hardness of fixed IR/MRBM launchers.

Advantages and disadvantages of each package are summarized briefly below, and more fully in Section V of the basic study report.

Package I: Freezes ICBM launchers to number now operational and under construction, and bans land-mobile missiles. No restrictions on SLBM’s nor on missile or launcher characteristics or location.

Pro’s

a. A first step, without disrupting present U.S. programs;

b. Each could build more survivable systems (SLBM’s), which would reduce incentives for first-strike capability and thus increase stability;

c. Could lead to quick agreement;

d. Easy to verify; little incentive to cheat.

Con’s

a. Could channel arms competition to other areas, such as MIRV’s, SLBM’s, more SS–9’s, and new generation of ICBM’s;

b. Could permit Soviet Union to catch up with the U.S. in area in which U.S. presently has a substantial lead (SLBM’s);

c. Might not be negotiable because Soviets would have to limit ABM’s and two of their key offensive programs, ICBM buildup and land-mobile missile development, without limiting current U.S. offensive programs;

d. Not convincing politically as an arms control measure;

e. Verification procedure for launcher replacement necessary, thus increasing negotiating difficulties.

Package II: Similar to I, but SLBM’s limited and land-mobiles permitted within total number of ICBM’s.

Pro’s

a. Limit on SLBM’s would preserve U.S. lead;

b. Might appeal to Soviets since it permits land-mobile ICBM’s;

c. Possible quick agreement.

Con’s

a. Probably less negotiable than other packages;

b. Only marginal effect on arms race;

c. Hard to verify agreed number of land-mobile missiles;

d. Verification procedure for launcher replacement necessary, thus increasing negotiating difficulties.

Package II–A: Permits varying the ICBM–SLBM mix within agreed total. Except as noted below, pro’s and con’s similar to Package II.

Pro’s

Could enhance survivability of retaliatory forces and reduce incentives to strive for first strike capability, thus promoting stability.

Con’s

a. Would permit Soviets to reduce U.S. lead in SLBM’s;

b. Land-sea mix would add verification problems;

c. Would introduce additional strategic uncertainties, since future force structures would be less predictable.

Package III: Freezes numbers of all offensive missile launchers; bans land-mobile, offensive missiles; prohibits enlarging silos or changing launcher configuration or location; bans land-mobile and sea-based ABM’s.

Pro’s

a. More comprehensive than I and II, since it affects all categories of strategic launchers;

b. Would allow U.S. to proceed with all ongoing offensive programs, but stops Soviet buildup in offensive launchers;

c. Could be acceptable to Soviets as permitting them to maintain an adequate strategic posture;

d. Could be step to broader agreement;

e. Would allow application of new technology to each side’s offensive missile force to enhance its effectiveness in target coverage and penetration and provide hedges against cheating and abrogation.

Con’s

a. Would permit MIRV’s and other missile improvements leading to possible future counterforce threat to Minutemen, depending on the agreed levels of ABM’s;

b. The responses to MIRV’s would probably lead to a continuing qualitative arms competition within the constraints of the agreement;

c. Soviets might reject because would limit their program, and because Soviet retaliatory capability is degraded at higher ABM levels;

d. Difficult to keep ABM low because of MIRV’s.

Package III–A: Permits superhardening, enlargement, and relocation of silos. Otherwise, pro’s and con’s similar to Package III.

Pro’s

a. Make counterforce attack more difficult;

b. U.S. would have greater opportunity to increase throw-weight and reduce present Soviet advantage than in Package III.

Con’s

a. Soviets could replace smaller silos with large ones for SS–9’s or new missiles;

b. Verification procedures for silo replacement necessary, thus increasing difficulties of negotiation.

Package III–B: Similar to III but permits varying the ICBM–SLBM mix. Since Package III places more restrictions upon offensive systems than does Package II, the advantages of the freedom to mix are greater under Package III–B than under Package II–A. Otherwise, pro’s and con’s similar to Package II–A.

Package IV: Same as III, except for addition of ban on MIRV’s.

Pro’s: In addition to arguments given earlier for MIRV ban,

a. Would do most to curtail the strategic arms race, particularly the MIRV–ABM escalatory action-reaction cycle;

b. Would be possible for both sides to accept low ABM levels and maintain confident retaliatory capabilities;

c. Assuming the Soviets want a MIRV ban, would make them more willing to accept the other limitations proposed;

d. U.S. budgetary savings could be greater.

Con’s

a. MIRV ban would reduce freedom to hedge against Soviet cheating by restricting U.S. capabilities to penetrate ABM’s;

b. Would give up what might be a significant U.S. technological advantage;

c. Would prevent attainment of U.S. capabilities to cover all Soviet time-urgent nuclear-threat targets which threaten the U.S. and our NATO allies;

d. Soviets may want to develop their own MIRV capability;

e. Initial U.S. proposal to ban MIRV’s might forego significant bargaining card.

f. A MIRV ban could be an initial step in placing controls on further development of technology for strategic systems, which some consider neither desirable nor verifiable.

1. Source: National Archives, Nixon Presidential Materials, NSC Files, NSC Institutional Files (H-Files), Box H–139, National Security Study Memoranda, NSSM 28. Top Secret. Sent under a May 30 covering memorandum from Smith to Kissinger. The full report is ibid. The SALT Steering Committee was under the chairmanship of ACDA and included representatives of the Secretaries of State and Defense, the Chairman of the JCS, the Director of Central Intelligence, the Chairman of the AEC, the President’s Science Advisor, and the Assistant to the President for National Security Affairs. The first meeting of the SALT Steering Committee was held on March 19, and meetings were held biweekly until the report was issued. Memoranda about those meetings are in Washington National Records Center, RG 383, ACDA Files: FRC 383–98–098, Director’s Files, Smith Files, NSSM 28, Preparation of US Position for Possible SALT Talks.
2. See footnote 2, Document 8.
3. See Document 4 and footnote 2, Document 12.
4. See also Annex B to basic study report. [Footnote in the original. Annex B, attached to the full NSSM 28 report, is entitled “Verification of Possible Alternative Options for Strategic Arms Limitation Talks.” It summarizes a report by the Verification Panel Interagency Working Group. (Washington National Records Center, RG 383, ACDA Files: FRC 383–98–0162, Verification of Possible Alternative Options for SALT—NSSM 28, May 1969)]
5. Annex A, attached to the full report, provides “Strategic Exchange Results.”