Foreign Relations of the United States, 1989–1992, Volume XXXI, START I, 1989–1991

157. Memorandum From Arnold Kanter of the National Security Council Staff to the President’s Assistant for National Security Affairs (Scowcroft)¹

Washington, September 18, 1990

SUBJECT

• Your September 19 Meeting on START

Objectives

As the President wrote to Gorbachev,² Shevardnadze’s presence in the U.S. next week probably is the make-or-break opportunity to complete START and CFE this year. We do not know whether the Soviets will come ready to do business. This meeting on START and the September 20 meeting on CFE are intended to put us in the position to drive for agreements on all of the remaining issues if the Soviets are prepared to do so.

Assuming you and your counterparts agree that we are in the end of the end-game, the goal of your two meetings should be to identify the bottom-line U.S. positions that distinguish between treaties we will and won’t sign. A let’s-see-what-the-Soviets-have perspective or other moves to defer decisions (e.g., we need more analysis) will virtually guarantee that START and CFE will drift into next year and, quite likely, an MBFR-like existence.

Later this week, the President will need to approve whatever consensus emerges from your group as well as resolve any differences. Armed with the President’s decisions and authority to close the deal, Secretary Baker will then need to engage Shevardnadze intensively next week on the remaining issues, supported by the Bartholomew-Karpov working group. (There is the separate, but related, matter of persuading Baker to be willing to use all of the flexibility the President gives him so that the President does not find himself in a detailed negotiating session with Shevardnadze.)

Attached at Tab A are brief papers on the START issues listed below. Copies have been distributed to the other principals to help them prepare for the meeting. All of these issues are essentially political and require a judgment about what we need to sign START and what we can settle for. Given what has already been agreed (on terms largely favorable to the U.S.), none would be likely to have a significant impact on the strategic balance no matter how they are decided. Yielding on too many of them, however, could foster the impression of a United States that was more eager, if not desperate, for START than a weakened Soviet Union.

Backfire

Everything except the number of Backfires allowed under START is agreed. In Irkutsk, Shevardnadze offered a limit of 600. In Moscow, we countered with a limit of 450. The obvious compromise is 525, but to close this issue Baker should have (and be prepared to use) the flexibility to agree to any number up to the 600 proposed by Shevardnadze.

RV elimination

Both sides have agreed to destroy missile launchers as part of the START reductions. The issue is what else, if anything, needs to be eliminated to minimize the political embarrassment over a reductions treaty that does not result in any weapons being destroyed. INF ratification suggests that this could be an issue on the Hill. Soviet proposals increase our potential vulnerability.

The U.S. has rejected missile elimination on the grounds that expensive assets which have value for other purposes (e.g., as space boosters) should not be destroyed just to make a political point, especially when we could—and would—simply build new boosters in their place. The bureaucracy has come up with a scheme for the “elimination” of RV shells (described in the attached paper) in an effort to answer the political mail, but no one is enthusiastic about it.

The approach is carefully designed to have a negligible effect on the availability of warheads for U.S. systems. We probably cannot stand anything more far-reaching. But precisely because it would have so little effect on the forces of the two sides, its political impact also is likely to be negligible. Indeed, it could be worse than useless, coming in for criticism as one more START sham masquerading as arms control and opening the door to the Senate, if not the Soviets, to insist on more far-reaching destruction measures. We should continue to resist RV “elimination” and take our lumps.

Non-Circumvention

This is an issue on which there is a strong difference about a principle (our unlimited right to transfer strategic nuclear systems to the British) that is likely to have no practical effect (because the British will not have any strategic modernization beyond Trident for the life of the START treaty). Nevertheless, it could be a treaty-stopper.

This issue matters to the Soviets: both Shevardnadze and Gorbachev have pressed the matter with Baker. It also matters to the British. Mrs. Thatcher will strongly resist anything in START that looks like a limitation on the “existing pattern of cooperation.” At the same time, she also has a strong interest in not letting British nuclear modernization become a final stumbling block to a START treaty.

The Soviets said they are prepared to accept continuing U.S. cooperation with the British provided it is limited to SLBMs (i.e., Trident and beyond but no cruise missiles). The British would surely reject such a limitation and we would do damage to our relationship by even sounding them out on their reaction.

We have little alternative but to hang tough with the Soviets on this issue, perhaps moving in their direction on some other issue (e.g., heavy ICBMs throw-weight) in a tacit political trade. We might, however, consider working with the British on a fig-leaf for the Soviets that could take the form of a unilateral statement by HMG that it has no plans for strategic modernization beyond its SLBM program for the duration of the START treaty and/or a unilateral U.S. statement that we have no plans for strategic cooperation with the British beyond SLBMs for the duration of the START treaty. Mrs. Thatcher will not like even these non-binding statements, but she also does not want to become the last obstacle to a START agreement.

Heavy ICBMs

The Soviets insist that they already have agreed to all the limits they are prepared to accept in START, arguing that any further limitations on heavy ICBMs should be the subject of the follow-on negotiations. In Moscow, we again proposed a limit of seven flight tests every three years as a way of constraining further improvements in SS-18 capabilities. We also invited the Soviets to give us their ideas for achieving this objective. They were unresponsive on both counts.

The Soviets are very unlikely to accept anything like our flight test limit. NSD–40 says we can settle for the already agreed limitations on heavies. If we do, we will need at a minimum to insist on a Soviet commitment not to increase the SS–18’s launch weight and throw-weight. (Otherwise, our “new types” definition would give them a free ride on further SS–18 upgrades.)

My Ungroup was distinctly unenthusiastic about your idea to get a more explicit Soviet commitment to deMIRVing in the follow-on talks in exchange for accepting the currently agreed SS–18 limits in START. They argued that it would reopen the joint statement agreed at the Summit at the same time that we are criticizing the Soviets for trying to reopen agreements on issues like ALCMs. That is a silly argument against a proposal that is worth trying and would give us some domestic political cover if it were accepted. You should try the idea of a Soviet deMIRVing commitment with your counterparts to see whether it gets a more positive response.

Throw-weight

The prospect of a “50 percent” reduction in aggregate Soviet throw-weight remains one of the most oft-cited payoffs of a START treaty. The Soviets, however, have always been careful to commit themselves only to a reduction of “approximately” 50 percent. In Geneva, they are interpreting “approximately 50 percent reduction” to mean a 45 precent reduction. Rick has given them reason to believe that, in exchange for Soviet acceptance of other elements of the U.S. position on throw-weight (of which OSD was the strong champion), the U.S. could accept a mid-point compromise (e.g., 47–48 percent) or even 45 percent.

The difference between a 45 and 50 percent reduction is roughly equivalent to the throw-weight of the SS–25 force. But the other central START limits (6000/4900/1500) would substantially constrain the Soviet ability to take advantage of any additional throw-weight. Agreeing to any number other than 50 percent will probably generate (perhaps substantial) political flak. On the other hand, Soviet agreement to 50 percent probably is not in the cards, at least not without a high level political effort on our part. If we decide to move off of 50 percent, it does not matter much—militarily or politically—where in the 45–50 percent range we compromise.

Baker should try to get Soviet agreement to a 50 percent throw-weight reduction. If that fails, we should compromise on any number between 45 and 50 percent in exchange for Soviet acceptance of a U.S. position of importance to us (e.g., noncircumvention). Whatever deal is struck on throw-weight should be made contingent on the acceptability of the Soviet declared throw-weight of the missiles they will reduce. Otherwise, they could inflate the magnitude of their claimed aggregate throw-weight reduction.

START/ABM Issues

The immediate issue is whether to accept a reference to “Article XI of the ABM Treaty” in the START preamble as the price for getting the Soviets to fall off several other efforts to relink START to the ABM Treaty. The Soviet motive almost certainly is to try to retain some vestigial linkage. The objections for accepting the deal stem from concerns that any reference to the ABM Treaty, no matter how innocuous, will cause us problems on the Hill with SDI supporters. SDI supporters also argue that any reference to the ABM Treaty amounts to a reaffirmation of a bargain—no defenses in exchange for continuing limitations on offenses—that the U.S. no longer supports.

Assuming that SDI remains too much of a political hot potato, we can try to outlast the Soviets on the preamble issue, recognizing that doing so will take time when we don’t have much. Alternatively, we can decide that the innocuous reference to Article XI is a small price to pay to settle this issue and get on to less symbolic disputes. Cheney probably will propose the former and Baker likely will support the latter.

There are two other SDI-related issues. One concerns our efforts to balance limits on Soviet encryption with our need to encrypt SDI tests (the “data denial” issue). OSD and the intelligence community have come up with a compromise that both can live with. It was presented to the Soviets in Moscow and the ball is in their court.

The other issue relates to Soviet efforts to ban the conversion of missiles to boosters for the deployment of “space strike weapons.” There are no present plans to use converted ballistic missiles to deploy SDI. Some tests also could be captured by the Soviet proposal, but work-arounds would be available at some increase in cost. The main argument against the Soviet proposal is that it would “relink” START and ABM/SDI.

This issue appears to be very important to the Soviets and we might be able to get leverage on some other issues if we were forthcoming. The question, as above, turns largely on the domestic politics of SDI.

Other START Issues

The Soviets are still trying to tie us in knots on ALCMs, particularly in their continuing efforts to constrain conventional ALCMs and press us for a very intrusive inspection regime. Colin Powell may raise the issue. If he does, Baker will reassure him that we will continue to be stalwart in the defense of U.S. interests. No additional decisions need be taken at your meeting.

There also are a collection of issues related to non-deployed missiles (NDMs). These include tagging, perimeter-portal monitoring, and the magnitude of the numerical limits. The Ungroup needs to do a better job of focusing the issues before they are considered at your level. One way or another, we will need to be in a position to say something to the Soviets on them next week.

Concurrences: John Gordon, Rich Davis

Attachment

Paper Prepared in the National Security Council³

Washington, September 17, 1990

Backfire Limits in START

The U.S. and the Soviet Union reached agreement in principle on the treatment of Backfire in START at the 1990 Washington Summit except for the number of Backfire aircraft that would be permitted. This agreement meets the requirements for Backfire limitations stated in NSD–40 (“Decisions on START Issues”).⁴ We need to decide what is the largest acceptable number of Backfires under a START agreement.

Internal guidance developed during the Summit was to seek an overall limit on Backfires as close to 500 as possible. In a July 29 letter to Secretary Baker, Foreign Minister Shevardnadze stated that the Soviets could accept a limit of 600 Backfires. In the recent Ministerial discussions in Moscow, we offered 450 Backfires.

The Soviets currently have about 370 Backfires (210 LRA and 160 SNA). They are projected to have about 600 Backfires (375 LRA and 220 SNA) by the year 2000. This number of Backfires implies a somewhat smaller Soviet medium bomber force.

At present the Soviets owe us an answer to the move we made in Moscow. In the end, however, the issue will come down to haggling over the number. Our approach to the issue in upcoming meetings would be substantially shaped by our bottom line: What is the maximum number of Backfires between 450 and 600 can we accept?

Attachment

Paper Prepared in the National Security Council⁵

Washington, September 17, 1990

RV Elimination

The U.S. has rejected missile elimination in START. The U.S. position in START would also not require the elimination of ballistic missile RVs. (In contrast, the Soviet position calls for the elimination of missiles, RVs, ALCMs and other bomber weapons.) We need to decide what kind of RV elimination proposal, if any, to adopt in START.

The USG has developed a specific limited proposal for eliminating ballistic missile RVs in a way that is consistent with U.S. programs. This proposal envisions the destruction of most of the RV shell, after the warhead and other internal components have been removed. The proposal would require the destruction of a number of RVs equal to the net number of warheads on missiles retired under START. The actual RVs destroyed could be older RVs that are currently in storage. This flexibility would permit us to reuse warheads and RVs from retired Trident C–4s on the D–5s now entering service. An elimination regime that required the destruction of the actual RVs removed from service would be much more costly to the SLBM modernization program. An elimination proposal that required the destruction of warheads in addition to RV shells would be even more expensive and could deny us the ability to deploy future Trident submarines so long as Rocky Flats is not operational.

RV elimination is largely a political and perceptual issue. START will be compared to the standard set by the INF Treaty and improvements over SALT II. The U.S. has claimed from the beginning of START negotiations that it must address a major flaw of SALT II by limiting missiles as well as their launchers. During INF ratification, the Administration was criticized because INF did not require the elimination of warheads.

In START we already have agreed to eliminate submarines, heavy bombers, and ICBM launchers. The question is whether we need additional destruction requirements and, if so, whether the RV elimination proposal outlined above is better or worse than no proposal.

On one hand, an RV elimination requirement could mitigate pressure during ratification for more extensive—and expensive—destruction. On the other hand, it is not clear that the draft U.S. proposal would succeed in this regard. First, RV elimination would have much less visual impact than the pictures of missiles being destroyed in INF. Second, the draft U.S. proposal would be criticized by some as an expensive sham because it would permit everything except the RV shell to be salvaged before destruction, would permit the elimination of older RVs in storage, would not require the elimination of RVs on missiles actually retired, and would not prohibit the construction of new RV shells.

Soviet statements in Geneva suggest that their motivation in proposing RV and weapon elimination is primarily a matter of their domestic politics. If so, the Soviets may be flexible on the details. On the other hand, a limited U.S initiative might make it more difficult to resist Soviet (or Senate) pressure to adopt additional elimination requirements that would be much more disruptive and expensive to implement.

Options

1. Maintain current U.S. position; no RV destruction.

2. Propose to require the destruction of a number of RV shells equivalent to the net number of RVs deployed on missiles that are reduced. There would be no commitment to destroy the RVs actually on the missiles removed from the force.

3. Propose to require the destruction of RV shells or RVs from the missiles that are actually removed from the deployed force under START.

Attachment

Paper Prepared in the National Security Council⁶

Washington, September 17, 1990

Non-circumvention

We and the Soviets have agreed to texts of a treaty article and an agreed statement which would ban international obligations or undertakings which would conflict with the treaty, and ban transfer of ICBMs, SLBMs, heavy bombers and nuclear ALCMs, except for existing patterns of cooperation. The Soviets made a unilateral statement that they have no programs of cooperation involving the transfer of strategic offensive arms to third countries. The Soviets object to our unilateral statement that makes clear our pattern of cooperation with the UK is essentially open-ended, going beyond Trident II and beyond SLBMs. We need to decide what, if any, additional steps to take to resolve this issue with the Soviets, and begin high-level consultations with the U.K. on any such steps.

The Soviets accept the Trident II program, but claim an open-ended provision for transfers to the U.K. would cause them a ratification problem. Both Gorbachev and Shevardnadze have highlighted this “non-circumvention” issue in their meetings with Secretary Baker.

The U.S. at the summit took the position that U.S./U.K. cooperation could not be constrained by START, and there is not a problem here because U.S./U.K. cooperation will not make the U.K. a nuclear superpower. U.K. Foreign Minister Hurd subsequently wrote Shevardnadze making clear the importance of U.S./U.K. cooperation to the British, but assuring the Soviets about the relative small scale of future U.K. strategic forces.

In the recent Bartholomew-Karpov discussions in Moscow, the Soviets offered to accept future U.S. cooperation with the U.K. if it were specifically limited to SLBMs, apparently permitting a follow-on to Trident, but prohibiting any transfers of other systems, such as nuclear ALCMs. We have repeatedly assured the British since SALT II in the late 1970s that we would accept no limits on the existing pattern of U.S./U.K. cooperation.

Options

1. Maintain the present U.S. position.

2. Offer a package that would resolve both non-circumvention and heavy ICBMs: We settle for the already agreed limitations on heavy ICBMs (but permit continued flight-testing) in return for Soviet acceptance of the U.S. position on non-circumvention.

3. Offer a U.S. (and possibly a U.K.) unilateral statement of plans along the following lines: The U.S. has no plans to transfer strategic offensive arms to a third country, over the duration of the Treaty, beyond the existing obligations (to the U.K.) related to SLBMs. (A U.K. unilateral statement could indicate there were no plans for cooperation with the U.S. on strategic modernization beyond SLBMs.)

4. Agree to a binding limit on future cooperation with the U.K. to SLBMs.

Attachment

Paper Prepared in the National Security Council⁷

Washington, September 17, 1990

Heavy ICBMs

The Soviets have agreed to cut their heavy ICBM force by 50 percent to 154 missiles with 1540 warheads, ban new types of heavy ICBMs, and ban mobile heavy ICBMs. We need to decide what additional constraints on heavy ICBMs, if any, are necessary to complete the START agreement.

NSD–40 states that an acceptable package of constraints for the U.S. would be the agreed constraints plus “such limitations on the modernization of existing types of heavy ICBMs as may be agreed between the sides.” The Soviets resist all further limits in START, saying the 50 percent cuts are already drawing domestic criticism and any further limits must be deferred to follow-on talks. The June 1990 Summit joint statement on future negotiations calls for measures related to the question of heavy missiles and MIRVed ICBMs.

At the Washington summit, the U.S. pressed unsuccessfully for a limit of 7 flight tests over a three-year period. Since then the Soviets have taken the position that the heavy ICBM issue is resolved for START and that there is nothing left to discuss. In the recent Moscow meetings, the Soviets did not reiterate their position that the issue is closed, nor object to the U.S. presentation on heavy ICBMs which reiterated the proposal for a limit of seven flight tests every three years, although they also did not respond to it nor to an invitation to present their ideas for additional limits.

Options

1. Settle for the already agreed constraints on heavy ICBMs.

2. Accept the already agreed constraints on heavy ICBMs in return for constraints which would permit little, if any, additional modernization of SS–18s.

3. Offer a package that would resolve both heavy ICBMs and non-circumvention: We accept the already agreed limitations on heavy ICBMs in return for Soviet acceptance of the U.S. position on non-circumvention.

4. Accept the already agreed constraints on heavy ICBMs in return for a Soviet commitment to a deMIRVing commitment in START II that goes beyond the summit statement. (The summit statement commits the two sides to “seek measures that reduce the concentration of warheads on strategic delivery vehicles as a whole, including measures related to the question of heavy missiles and MIRVed ICBMs.)

5. Continue to press for a flight-test limit on heavy ICBMs.

Attachment

Paper Prepared in the National Security Council⁸

Washington, September 17, 1990

Ballistic Missile Throw-weight

After years of resisting the U.S. proposal for throw-weight limits, the Soviets appear ready to accept the U.S. concept. The primary outstanding issue is whether to accept a Soviet-proposed 45 percent reduction in Soviet throw-weight, or insist on a 50 percent reduction that the U.S. has sought.

Throw-weight has been one of the most politically visible concepts in strategic arms control. SALT II’s failure to address the 3:1 advantage in Soviet throw-weight was one of the principal objections to that treaty. As a consequence, the U.S. proposed in START to reduce Soviet throw-weight by 50 percent. This objective of a “50 percent throw-weight reduction” has been one of the most popular and easily understood features of START.

The Soviets agreed in 1987 to reduce their throw-weight by approximately 50 percent, but insisted that it be measured by the weight actually tested on missiles, not the full capability of missiles. Now the Soviets accept the U.S. proposal to measure the full capability of missiles, but want a 45 percent reduction rather than a 50 percent reduction.

This is largely a political issue. Accepting the Soviet proposal would be criticized as abandoning a longstanding position, to which the Soviets have agreed in principle, for 50 percent reductions (although the Soviets have been careful to accept only a commitment to reduce throw-weight by “approximately” 50 percent). From a political perspective, the fact that the U.S. has settled for anything less than a 50 percent reduction may be as much or more important than the particular percentage reduction that is agreed (e.g., 45 vs 47 percent). Militarily, under the Soviet proposal, their post-START force could be 10 percent larger in terms of weight than a 50 percent throw-weight reduction, the equivalent throw-weight of the SS–25 force. In principle, this additional throw-weight could be exploited to deploy larger warheads or penetration aids. In practice, however, the military significance of this additional throw-weight would be mitigated by the numerical limit of 4900 warheads, which would remain unchanged.

Finally, in order to make clear what limit will actually apply to Soviet throw-weight, we need to establish an agreed numerical baseline from which the reductions will be taken. We need to make clear that any proposal to resolve the throw-weight reduction is contingent upon the Soviets tabling (and our acceptance of) throw-weight data for their existing missiles.

Attachment

Paper Prepared in the National Security Council⁹

Washington, September 17, 1990

START/ABM Issues

START Preamble

The Soviets have pushed to relink START to the ABM Treaty by proposing that the START Treaty Preamble state that the sides are “Mindful of their obligations under the ABM Treaty.” They have also proposed an agreed statement linking START and the ABM Treaty and unhelpful language on follow-on negotiations. Recently the Soviets agreed to drop all three positions (although they would issue a unilateral statement on START-ABM Treaty linkage) if we could accept an explicit reference to Article XI of the ABM Treaty. Article XI states: The parties undertake to continue active negotiations for limitations on strategic offensive arms.

In Moscow last week, we stated that the U.S. is unwilling to accept any explicit reference to the ABM Treaty, but could agree to incorporate the substance of Article XI, as follows, “Taking into account the undertakings of the Parties to continue active negotiations for limitations on strategic offensive arms.” The Soviets referred the issue to Geneva.

Options. In weighing options, the balance to be struck is between removing several irritating Soviet positions by accepting a reference to an innocuous article in the ABM Treaty, versus seeming to acquiesce to START-ABM Treaty linkage. The appearance of such acquiescence would be used by some to argue against START ratification. Moreover, citation of Article XI could be regarded as reaffirmation of the essence of the 1972 deal: forgo defenses in order to make offensive limitations possible. Today; our position in NST is that offensive force reductions and measured deployment of defenses can be compatible. Thus, the preamble language could be seen to contradict one of our central positions.

Data Denial

In Moscow last week we presented a position on data denial in START that strikes a balance between permitting limited telemetry encryption for SDI tests and prohibiting encryption that is most important to monitor START limits. The Soviets have not yet responded to our proposal. We may have to return to the issue if the Soviets press, either for a more comprehensive ban on data denial, or for greater freedom to encrypt data on flight-tests.

Space Launch Vehicles

The U.S. has proposed to permit the conversion of ICBM and SLBM stages into space launch vehicles to launch payloads into the upper atmosphere or orbit. The Soviets have revived their old proposed ban on “space strike arms” by proposing to ban the use of these converted ballistic missiles for, “the purpose of placing into orbit weapons intended to strike targets on earth, in the atmosphere or space.”

In Geneva the Soviets have said this provision is intended to prevent the use of converted ballistic missiles to deploy a SDI system, such as Brilliant Pebbles, into orbit. Their proposal would, however, appear to apply to tests in orbit as well.

We intend to use converted ballistic missile stages to launch satellites and sounding rockets, including payloads for SDI tests. Up to three of these launches are currently planned for Brilliant Pebbles, which could possibly be captured by the Soviet proposal. Although converted ballistic missiles could be used to deploy Brilliant Pebbles, there are no present plans to do so.

Options. In terms of the practical effect on U.S. plans, the Soviet proposal could increase the cost of certain SDI tests, but would not prevent any planned tests. In weighing options, the balance to be struck is between removing this issue, which is very important to the Soviets, versus seeming to acquiesce to the Soviet position on banning space strike arms, which has been directed at prohibiting SDI. The appearance of such acquiescence would be used by some to argue against START ratification.

1. Source: George H.W. Bush Library, Bush Presidential Records, National Security Council, John A. Gordon Files, Subject Files, OA/ID CF00953–021, START—September 1990. Secret. A stamped notation indicates Scowcroft saw the memorandum.
2. See Document 156.
3. Secret.
4. See Document 117.
5. Secret.
6. Secret.
7. Secret.
8. Secret.
9. Secret.