100. Memorandum From the President’s Assistant for National Security Affairs (Kissinger) to President Ford1


  • Analysis of Soviet SALT Proposals

I have provided below my preliminary analysis of the current Soviet positions on the outstanding SALT issues.2 This analysis will pro [Page 427] vide background for your meeting with General Secretary Brezhnev in Helsinki.

By way of summary, I believe we are now in the following situation on the major issues:

—On MIRV verification, the Soviets have made a major move in agreeing to count missiles tested with both MIRVs and single RVs as MIRVed when deployed. However, a major problem still remains with counting MIRVs on SLBMs. If MIRVs are deployed only on part of a submarine class, we probably cannot verify that the remaining missiles on that class are not also MIRVed.

—Regarding cruise missiles, they have not made a major movement. However, our only strategically important interest is in air-launched cruise missiles, and we could probably trade off cruise missiles launched from other platforms for what we want in this area.

—Their position favoring a ban on the deployment of land-mobile and most air-mobile ICBMs could create some bureaucratic problems, but on balance, a combined ban would be in the US interest.

—The Backfire bomber issue could be a problem, unless Gromyko’s failure to explicitly reject a tanker prohibition indicates some flexibility on this point.

—Regarding the definition of a heavy missile, there seems to have been a breakthrough. It appears that what they have in mind for the definition of a heavy missile may not be much worse, and in some ways may be even better, than what we had proposed.

—Their proposal on limitations on increases in the size of ICBM silos is unacceptable. However, they were confused on this even among themselves, and I am confident we will be able to reach agreement on this issue.

—Finally, the Soviet position on the resumption of follow-on negotiations is essentially what we had proposed.

A more detailed discussion and analysis of each of these issues follows.

MIRV Verification

On verification of the 1320 MIRV limit, the Soviets accepted our rule counting all missiles tested with both MIRVs and single RVs as MIRVed when deployed. This is significant for resolving potentially serious ambiguities associated with such systems as the SS–18. However, there are two other potentially serious verification ambiguities which remain unresolved:

—The deployment of MIRVed SLBMs on only part of a submarine class if the MIRVed launchers on submarines of the class are indistinguishable from unMIRVed launchers.

[Page 428]

—The deployment of single RV missiles in silos modified in a manner which will permit the deployment of MIRVed missiles.

Concerning SLBMs, there could be a significant verification uncertainty regarding the total number of MIRVed launchers if the Soviets develop MIRVed and unMIRVed SLBM launchers which are indistinguishable. Because of this uncertainty, we have proposed in Geneva to count all SLBMs in a submarine class when the first submarine in the class is equipped with a MIRVed missile.

Although the Soviets have rejected this approach to SLBM MIRV verification, we have some latitude in our position which we have not yet exercised. In particular, we could lessen the immediate impact of the rule on the Soviets by tying the SLBM MIRV count to an overhaul or conversion cycle. For example, counting submarines of a class as MIRVed when they complete overhaul or when they go to sea after construction or conversion could be more negotiable than counting all the launchers of the class at the time the first MIRV is deployed. If at the same time we specify that, after an agreed period, all submarines of a MIRVed class would be counted as MIRVed, we would not have lost our ultimate verification goal of counting all SLBMs by submarine class. Nevertheless, this concession on timing is not likely to be seen as a major one by the Soviets. This issue may end up being the most difficult one remaining.

We can afford to be more flexible on our remaining ICBM counting rules which address the deployment of ostensibly unMIRVed missiles in silos designed to accommodate, or which could accommodate, missiles equipped with MIRVs. The Soviets may well decide to modify single RV silos to MIRV-type silos for hardening purposes and then deploy single RV missiles in the upgraded silos. They have already done this for 60 silos, although they will probably soon replace the unMIRVed missiles with MIRVs.

The problem, of course, is that if the Soviets choose to upgrade all of their ICBM silos to MIRV-type silos, we may have no way of determining which, if any, of these silos contain unMIRVed missiles. The resulting ambiguities could cause us to conservatively estimate that Soviet MIRV deployment is significantly in excess of the actual number. To resolve the ambiguity, our current position specifies that any ICBM launcher which will permit the deployment of MIRVed missiles must count within the 1320 MIRV limit.

The problem with our present counting rule is that it is not sufficient to cover all potentially difficult situations. It only resolves the most obvious ambiguities. For example, the Soviets could still deploy a new MIRVed missile which fits in existing silos or harden existing silos under a new design which they claim is not MIRV-related but which [Page 429] we could not distinguish from MIRV-related activity. Our present counting rule would not resolve either of these ambiguities.

Consequently, we can probably consider modifying our present rule on silo modifications to permit greater Soviet latitude in modernizing unMIRVed silos, a point they strongly emphasized at my last meeting. At the same time, however, we can still eliminate glaring ambiguities of the type which currently exist in which unMIRVed missiles are deployed in unique, MIRV-type silos. For example, we could ask the Soviets for assurances that they will not modernize unMIRVed silos in the same manner that they have modernized silos for MIRVed missiles.

In sum, the Soviets have made a significant movement toward our verification position by accepting our “once tested with MIRVs, all versions count as MIRVed” counting rule. However, significant verification ambiguities remain unresolved. The most important of the remaining ambiguities, MIRVed SLBMs which are compatible with all launchers in a submarine class, may be resolved by modifying our present SLBM position to lessen the immediate impact of the counting rules on the Soviets. In addition, we can probably live without our present rule regarding modified ICBM silos; however, before dropping such a rule altogether, we should seek Soviet assurances on silo modifications which will prevent them from modifying unMIRVed silos in exactly the same manner that they have modernized silos for MIRVed missiles.

Mobile ICBMs

Before my last meeting with Gromyko, the Soviet position regarding mobile ICBMs had been extremely one sided: their proposal would have effectively banned air-mobile ICBMs, a system in which we have considerable interest, while permitting land-mobile ICBMs, which have been under active development in the Soviet Union for several years.

We rejected the Soviet proposal on mobile ICBMs, but indicated that we were prepared to consider equitable proposals under which the limits on mobile systems would impact equally on the two sides. Responding to this, the Soviets indicated at my last meeting that they were prepared to ban the deployment of land-mobile ICBMs for the duration of the new agreement. Coupled with their present formal position effectively banning air-mobile ICBMs, the new Soviet proposal amounts to a combined ban on both air- and land-mobile ICBM systems (although a limited air-mobile system might still be permitted).

The bureaucracy is very divided on the mobile issue. OSD (and to a lesser degreee JCS) is likely to object strongly to a combined mobile [Page 430] ban. They will probably argue on behalf of a US mobile program in light of new Soviet developments in anti-submarine warfare (ASW) which could allegedly threaten our sea-based deterrent. However, there remains considerable interest in the other agencies in a combined mobile ban. CIA continues to emphasize the serious verification problems associated with mobiles, while ACDA and State argue that public and Congressional acceptability of SALT II, as an arms control agreement, would be enhanced substantially by including a ban on mobiles.

In spite of OSD’s concerns, there still remains no strong rationale for mobile ICBM deployment. The principal strategic rationale for developing a mobile missile—a hedge against future vulnerability of not only silo-based ICBMs, but SLBMs and bombers as well—is an extremely unlikely contingency during the life of this agreement. OSD’s concern about recent Soviet developments in ASW are greatly exaggerated, and Soviet developments in ASW which may occur over the next decade should be more than offset by our own Trident program. In short, the retaliatory capability of both our SLBMs and bombers should remain intact.

Beyond this, there continues to be two practical problems associated with mobile systems:

—We do not have a feasible land-mobile option owing to land availability and political factors.

—It is unlikely that we could develop and deploy an operational air mobile ICBM prior to the end of the 1985 agreement period in any event.

On this last point, OSD currently projects an operational capability for a US mobile ICBM system (air- or land-based) by 1985. Although this could probably be moved up by two or three years under an accelerated development schedule, it would be virtually impossible to get the required funds from Congress over the next few years.

In sum, it would be in the US interest for the 10 years of the agreement to agree to a combined ban on air- and land-mobile ICBM systems. However, in view of the strong DOD interest in doing development work on mobile ICBM systems, we should confirm the apparent Soviet position that the ban applies only to deployment, and not to development and testing as well.

Cruise Missiles

The Soviets are continuing to claim that the air-to-surface missile (ASM) limitation agreed to at Vladivostok (counting ASMs greater than 600 km range in the aggregate) should apply to all ASMs, both cruise and ballistic, while we are holding to the position that this agreed 600 [Page 431] km limitation should only apply to ballistic ASMs. Their overall cruise missile position can be summarized as follows:

Air-Launched Cruise Missiles (ALCMs). Count ALCMs of range greater than 600 km deployed on bombers in the 2400 aggregate; ban ALCMs of greater than 600 km range on all aircraft other than bombers.

Sea-Launched Cruise Missiles (SLCMs). Ban the development, testing, and deployment of sea-launched cruise missiles of range greater than 600 km (includes SLCMs launched from both submarines and surface ships).

Land-Based Cruise Missiles. Ban the development, testing and deployment of land-based cruise missiles of range greater than 5500 km.

In your channels we have told the Soviets that we could agree to a limitation under which all cruise missiles of greater than 3000 km range would be counted in the aggregate, independent of launch platform; nevertheless, they are continuing to insist on the rigid limits described above.

The only strategic issue with respect to cruise missiles concerns the need to retain the option to deploy long-range air-launched cruise missiles; we could readily accept the proposed 600 km limitation on sea-launched cruise missiles (SLCMs) and the 5500 km limitation on land-based cruise missiles.

The issue with respect to ALCMs is the need to retain long-range ALCMs as a hedge against future threats to bomber penetration, and in particular as a hedge against the uncertainties regarding eventual B–1 deployment.

The air defense threat of principal concern is Soviet deployment of an extraterritorial barrier defense consisting of airborne radars such as US AWACS (Airborne Warning and Control System) and long-range fighter interceptors. Such a defense, capable of engaging US bombers at a distance of about 1000 km from the Soviet border, is projected to be deployed in the early 1980s. There are two possible approaches to penetration of such a defense:

—Launch long-range stand-off air-launched cruise missiles outside the defenses (i.e., 1000 km from the Soviet borders) which penetrate by saturating the barrier defense with a very large number of attacking objects.

—“Fight your way in” using anti-AWACS and anti-interceptor air-to-air missiles to punch holes in the barrier defense.

At the present time, there is much greater confidence in the feasibility of the stand-off cruise missile approach; the required hardware is in an advanced state of development and the penetration technique, saturation, is presently considered much more effective. In contrast to the ALCM, work is just beginning on an anti-AWACS and anti- [Page 432] interceptor missile. The required technology is much more sophisticated and the penetration approach is much more vulnerable to countermeasures such as decoys, air-to-air missiles, etc. For this reason we have favored retaining the option to deploy a long-range ALCM as the preferred method for insuring bomber penetration in the 1980s. In addition, there is concern that the B–1 might never be deployed in significant numbers because of cost considerations, independent of SALT; in such a situation, B–52 penetration, even with an anti-AWACS missile, could be questioned. In contrast, the B–52 would make a very good stand-off cruise missile carrier.

The range required for a stand-off cruise missile option depends on the target coverage required. A 3000 km ALCM will provide coverage of about 70 percent of the Soviet urban/industrial target base, assuming an AWACS barrier at 1000 km from the border; 2500 km will provide coverage of about 50 percent of these targets including Moscow. However at 2000 km only 25 percent of the Soviet urban/industrial target base could be covered, and coverage of Moscow would be lost.

On the other side of the argument is the question of how long-range ALCMs would affect Soviet bomber capability. With a 2500–3000 km ALCM, the Backfire can cover most of the more populated regions of the US and recover in the Soviet Union. On the other hand on one-way missions complete coverage of the US is possible, with or without long-range cruise missiles. However, even if Backfire is not classed as a heavy bomber, there would still presumably be a restriction on equipping medium bombers such as Backfire with long-range ASMs (e.g., more than 600 km). Thus a 2500–3000 km ALCM limit would not give the Soviets a unilateral advantage because of extension of Backfire capability.

With respect to sea-launched cruise missiles (SLCMs), the US strategic SLCM program has no strategic utility; even with a 3000 km SLCM limit, SLCM submarines would be less survivable than our SSBN forces because they must approach much closer to Soviet shores to launch their missiles. If our SSBN force survives, the increased target coverage is certainly not needed. Similarly the SSBN force can satisfy any requirements for a more survivable theater tactical nuclear capability to replace the current theater nuclear-capable tactical aircraft. In addition, equipping attack submarines with missiles for use against land targets could detract from their principal mission of anti-submarine warfare and the protection of shipping lanes.

The Navy has recently expressed some interest in a tactical ship-to-ship SLCM with range in excess of 600 km. However, almost all tactical SLCM missions can be satisfied by a cruise missile of less than 600 km range, as is evidenced by the choice of 550 km for the range of the [Page 433] tactical variant of the strategic SLCM currently under development. It is unlikely that the Navy will ever rely on attacking Soviet surface ships from very long ranges with tactical SLCMs since US attack submarines are projected to be able to easily approach within a few hundred kilometers of Soviet surface vessels without fear of detection into the foreseeable future.

With respect to intercontinental range land-based cruise missiles, which would be banned by the Soviet proposal, there has been some minor interest in DOD in preserving the option to develop recallable land-based intercontinental cruise missiles (ICCMs). However, there is no anticipation of deploying such systems before 1985. There is also some interest in deploying long-range land-based cruise missiles as tactical weapons in Europe. The Soviets are probably proposing a 5500 km limit on land-based cruise missiles to avoid prejudicing their position on air-to-surface (ASMs) where they emphasize the difficulty of distinguishing between cruise and ballistic ASMs. They may also wish to retain the option to deploy long-range land-based cruise missiles for use against China and Western Europe. In any event, there are no problems with accepting their proposal for a 5500 km limit on land-based cruise missiles.

There has been some concern expressed about the verification problem of different range limits on different types of cruise missiles since cruise missiles are readily adapted to different launch platforms. However, the verification problem is probably tractable in a situation where the range limits on ALCMs and land-based cruise missiles are substantially greater than that on SLCMs; it is highly unlikely that the Soviets could conduct a covert test program for long-range SLCMs of the magnitude required to operationally deploy such SLCMs without US detection of the test program. In addition, it is unlikely that by 1985 the Soviets could develop a long-range cruise missile which could be launched from torpedo tubes. The Soviets do have large launchers on current Soviet SLCM submarines, but these submarines are old and currently must surface to launch their missiles. Thus, a 2500–3000 limit on ALCMs, a 5500 km limit on land-based cruise missiles, and a 600 km limit on SLCMs would be acceptable from a verification standpoint if the limit on SLCMs bans testing, development, and deployment as the Soviets have proposed. Since ALCMs would be permitted and counted above 2500–3000 km, the longer range limit on land-based cruise missiles does not create a verification problem.

It will be important to obtain some progress on this issue at your meeting with Brezhnev in Helsinki. I suggest you impress on him that long-range cruise missiles on heavy bombers are bomber armament and not strategic delivery vehicles like ICBMs and SLBMs. In addition, [Page 434] we should tie our concessions on the cruise missile issue to our position on Backfire.

ASMs on Non-Bomber Aircraft

From the US viewpoint, there are two issues which bear on the Soviet proposal to ban all types of ASMs of range greater than 600 km on aircraft other than bombers:

—Whether air-mobile ICBMs are going to be banned or permitted.

—The need to retain the option to deploy long-range cruise missiles on transport aircraft (e.g., C–5 and 747) as well as on bombers such as the B–1 and the B–52.

With respect to deploying air-mobile ICBMs on non-bomber aircraft, as noted above, it is extremely unlikely that we could retain this option and also obtain a ban on land-mobile ICBMs. On the other hand, if we decided to forego the ban on land-mobiles, we would have to retain the right to deploy long-range ASMs on non-bomber aircraft (in particular on transport aircraft) if we are to have a viable air-mobile ICBM option.

Regarding the second issue, the B–1 and the B–52 would both be adequate cruise missile launch platforms in the 1985 timeframe. Although deployment of long-range ALCMs on modified transport aircraft is an option which has been given serious consideration, it is not imperative that this option be retained in a 1985 agreement.

If we do agree in principle to accept a ban on long-range ASMs on non-bomber aircraft, this ban should apply to non-heavy bombers as well as to non-bomber aircraft in order to avoid the problems of long-range ASMs on Backfire and the other Soviet medium bombers. The Soviet proposal in Geneva was somewhat ambiguous on this issue since it was couched in terms of a ban on long-range ASMs on “strategic” bombers.


The Soviets have shown no sign of movement on this issue. This is hardly surprising. We have known from the beginning that we would have to compromise on Backfire, since the Soviets clearly had no intention of including the Backfire in the 2400 aggregate when this limit was negotiated at Vladivostok.

Disturbingly, however, the Backfire has become something of a public issue, and the longer the issue drags on the more people will dig in their heels and claim it is a heavy bomber. As you know, our intelligence seems to confirm Soviet statements that the Backfire will not be used for intercontinental roles, and consequently we could probably fall off designating the Backfire as a heavy bomber if deployment indicators continue to confirm Soviet statements.

[Page 435]

To nail down specific indicators which would improve our confidence that Backfire was not being used for intercontinental roles, but which still permit the Soviets to exempt the Backfire from the 2400 aggregate, we have tried through your channels to obtain Soviet agreement to limit Backfire deployment to certain bases in the Southern USSR and to prohibit a companion tanker force from supporting Backfire operations. To date the Soviets have rejected this compromise approach, although there are some indications that they may consider a tanker prohibition.

Consequently, we should continue to explore the possibility of a tanker prohibition for Backfire operations. The ban on long-range ASMs on all non-heavy bombers discussed above will also assist in inhibiting development of an intercontinental capability for the Backfire. In addition, we should continue the linkage, which we established in our note to Dobrynin,3 between Soviet acceptance of constraints on Backfire operations and US movement toward the Soviet position on cruise missiles.

Silo Dimensions

We agreed at Vladivostok to retain the Interim Agreement provision which limits increases in silo dimensions to 10–15 percent. This limit is ambiguous in that it is not clear whether one or both dimensions of a silo (depth and diameter) may be increased by 10–15 percent. To remove this ambiguity we have proposed that the sum of the individual increases in depth and diameter should be limited to 15 percent. In Geneva, Gromyko proposed that a 32 percent limit on volume increases be substituted for the 10–15 percent limit on increases in dimensions. Since we want to retain the 15 percent limit on depth increases, a reasonable compromise would be to supplement a 15 percent limit on increases in dimensions with the 32 percent volume limit proposed by the Soviets.

Heavy ICBM Definition

The Soviets have finally conceded that the new agreement should include an explicit heavy ICBM definition. However, they have not agreed to our proposal for a definition based on throw weight, but instead have proposed that a heavy ICBM be defined on the basis of missile gross weight, with the SS–19 as the dividing line between heavy and non-heavy ICBMs. Our analyses indicate that missile gross weight is less desirable than throw weight as the basis for a definition, but significantly more desirable than basing a definition only on missile volume. For example, if the heavy ICBM limit is pegged to the SS–19 [Page 436] gross weight (about 185,000 lb), the best advanced technology missile which the Soviets could deploy would have about 9,000 lb throw weight, as compared to the 7000 lb throw weight of the current SS–19. A 185,000 lb limit would not affect the US M–X advanced ICBM design which is estimated to have a gross weight of 150,000 lb.

The only problem with missile gross weight is that it does not take into account the possibility of “cold” launch, i.e., a launch assist device, such as the Soviets use on the SS–17 and SS–18. This technique can substantially increase the throw weight for a given gross weight missile. In principle, two approaches to this problem are possible:

—Include the weight of the launch assist device for cold-launched missiles in calculating missile gross weight.

—Have a separate definition for cold-launched and hot-launched missiles, e.g., the definition of cold-launched missiles could be tied to the SS–17.

For the sake of simplicity in the negotiations, the first approach of adding the weight of the launch assist device is preferable. This can be worked out by the Delegation in Geneva.

The problems of verifying missile gross weight are comparable to those for verifying throw weight. Telemetry is an important factor, although knowledge of missile volume and fuel type can produce acceptable estimates of gross weight.

If this issue comes up in your discussions with Secretary Brezhnev, you could indicate to him that you believe that agreement on a heavy ICBM definition can be reached using missile gross weight as the defining parameter. However, you should indicate that there is also a need for placing an upper limit on heavy ICBMs as well and argue for limiting heavy ICBMs to the gross weight of the SS–18. Without such a limit the only restraint on heavy missile size would be the limit on silo dimension increases.

How to Proceed

I believe we should prepare a response to the Soviet proposals that we could transmit to Brezhnev through Dobrynin.4 This will set the agenda for your meeting with Brezhnev and insure that the important issues are covered.

The principal issue which you will have to face in making your decisions is the necessity for keeping open options which the military believes might be of interest in the 1980s. Clearly all options can’t be kept [Page 437] open or we would never have substantive arms control agreements. The Soviets are apparently willing to close a number of possible areas of possible strategic arms competition. By accepting some of their proposals, we could clearly enhance the arms control impact and public acceptability of the new agreement.

  1. Source: Ford Library, NSC Institutional Files, Box 5, Verification Panel Meeting, 7/24/75–SALT. Top Secret; Sensitive. Lodal drafted the memorandum for Kissinger on July 16 and sent it under a covering memorandum that explained it would help prepare the President for his Helsinki meeting. Kissinger did not initial the memorandum, but it was included as Tab E to Kissinger’s briefing materials for the Verification Panel meeting, July 24. (Ibid.) The unsigned memorandum was also included as Tab F to Ford’s briefing materials for the NSC meeting scheduled for July 25. (Ibid.)
  2. See Document 98.
  3. See Document 97.
  4. Lodal and Sonnenfeldt prepared a note to the Soviet Union on SALT and sent a draft to Kissinger under a covering memorandum, July 25. There is no indication that the note was given to Dobrynin. (Ford Library, National Security Adviser, KissingerScowcroft West Wing Office Files, Box 21, SALT, Chronological)