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

220. Paper Prepared in the National Security Council¹

Washington, undated

POINTS TO BE MADE FOR NSC MEETING ON START

— As the President indicated, we want to follow the same basic strategy used for CFE. A letter from the President to Gorbachev to engage him directly and encourage him to break down internal Soviet roadblocks.

— The letter will lay out the issues and our proposal for dealing with each of them. We will invite Gorbachev to send a high-level team to Geneva for intensive negotiations with a group led by Reg Bartholomew and offer an additional meeting between Jim Baker and Foreign Minister Bessmertnykh, if necessary, to break any remaining logjams.

— I want to turn now to each of the issues we need to address in the letter.

REAFFIRM HOUSTON DEAL

— We begin by telling Gorbachev that we are proceeding on the basis of the agreements reached by Jim with Shevardnadze at the Houston ministerial, making clear that this is the only basis on which we can quickly reach final agreement on START.

— In Houston, agreements were reached on:

– monitoring of mobile missile production, i.e., the “PPCM” provisions,

– accountability and inspection of heavy bombers,

– limitations on throw-weight,

– construction of new silos for heavy ICBMs, and

– attribution of warheads to new types of ballistic missiles, including the application of the “40 percent” rule.

DOWNLOADING/NEW TYPES TRADE-OFFS

— We would then address two major, unresolved issues - downloading and the new types definition.

— These issues are related in the sense they both address concerns about breakout potential.

— We can assess the breakout potential inherent in downloading existing missiles with some confidence because we know how many warheads each missile has proven capable of carrying.

— For new types, the breakout potential is more theoretical since it is based on judgements concerning capabilities that have not been directly demonstrated.

— In terms of the larger U.S. objectives, it may be better for us to hold firm on limiting the known breakout risks associated with downloading, and allow some latitude on the new types definition.

— Such a trade-off accepts that new types of missiles may, or may not, have a breakout potential, but it also builds on the realization that to take advantage of the new types definition, the Soviets will need to make significant investment in development and deployment of new missiles.

— That’s the perspective that leads to the following approach.

DOWNLOADING

— The first issue concerns the downloading of warheads from existing types of ballistic missiles.

— We want to be able to download Minuteman III from three to one or two RVs. The Soviets already have de facto downloaded their SS–N–18 from seven to three RVs.

— We ought to draw the line on downloading at these two systems. Otherwise, we will just allow the Soviets to keep adding to their breakout potential up to the more than 2000 warheads they proposed.

— That would not make strategic sense and would not make political sense. The START treaty would be roundly attacked on the Hill and the critics would be right.

NEW TYPES DEFINITION

— On the question of how to tell when a missile is a “new type,” we have maintained a simple principle: a new type missile must be genuinely new and verifiably different from any existing ballistic missile.

— The two sides have already agreed on criteria that reflect this principle.

— The Soviets have argued recently that an increase in throwweight should be added as a criterion for defining a new type of missile.

— We don’t like using throw-weight this way. [3 lines not declassified]

— But the Soviets are insistent that increases in throw-weight be an additional criterion for identifying new types.

— Given the relationship between downloading and new types I described, this may be an area in which we can show some flexibility, especially if we continue to insist on the U.S. position on downloading.

— The problem is finding the right throw-weight threshold.

— If we set the throw-weight threshold for new types of missiles too low, the Soviets might deploy variants of existing missiles masquerading as new types that would be counted with a fewer number of RVs. That, in turn, could provide the potential for breakout.

— Picking the right number is a matter of judgment, balancing how much risk we are prepared to run that the Soviets might exploit this approach to preserve some future breakout capability, against making a move that could lead to Soviet agreement on other issues, such as downloading, of particular importance to us.

— Given this background, it is not surprising that our staffs have not been able to come to agreement on what throw-weight increase we should propose as a new types criterion. Some believe we must insist on not less than a 30-40 percent increase. Others believe we could live with a 20 percent increase if we have to. (In both cases, these throw-weight increases would be in combination with a 5 percent change in the length of the first missile stage, a modification that by itself goes part of the way toward giving us confidence that we are dealing with a “new type” of missile.)

— I don’t propose that we try to resolve the issue here, Mr. President.

DATA DENIAL

— The final issue concerns the ban on denying performance data about ballistic missiles.

— Arnie Kanter’s “Ungroup” is wrapping up its review of a new proposal to be forwarded to Geneva this week.

— A fundamental principle of this proposal would be agreement that each side would use telemetry practices which would allow the other side to acquire data on ballistic missile performance.

— In addition, the sides would agree to ensure that the other side could interpret and understand the data acquired.

— However, extensive expert-level talks will probably be required to resolve the details of this issue even if the Soviets react favorably. The letter proposes these experts meetings.

— The Soviets have said that because of the complexity and the time required to negotiate acceptable provisions, final agreement on this issue could be put off until after treaty signature.

— That would be a serious mistake. We must have these data denial provisions nailed down before we can sign a START treaty.

CONCLUSION

— We don’t know how much flexibility the Soviets have on these matters. They may accept, but more likely the President’s letter will result in a counteroffer.

— This will at least give us an early indication of whether or not we can sign a solid treaty within the next several weeks.

1. Source: George H.W. Bush Library, Bush Presidential Records, National Security Council, H-Files, NSC Meetings Files, NSC0058—June 06, 1991—Strategic Arms Reduction Talks. Secret. Gordon, Davis, and Kuehne sent the paper to Scowcroft through Kanter under a July 5 covering memorandum: “At Tab I is a meeting memorandum to the President for the June 6 NSC meeting on START. At Tab II are talking points for your use at the meeting to lead the participants through the issues in the proposed letter to Gorbachev.”