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

172. Memorandum From Richard Davis of the National Security Council Staff to the President’s Assistant for National Security Affairs (Scowcroft)¹

Washington, December 5, 1990

SUBJECT

• Principals’ Meeting on Data Denial in START, December 6, 1990, 1:15 P.M.²

Data Denial is the most important outstanding issue in START, because of its potential implications both for treaty ratification and for U.S. national security interests. DOD and IC positions remain far apart and agreement in your meeting is unlikely. Accordingly, the primary objective is to gain a common understanding of the issue and reach a consensus that there is a suitable basis for decision by the President. The Intelligence Community proposal described below has been flatly rejected by the Pentagon at the working level, but nevertheless may be the best vehicle for framing the issues and leading the discussion.

Background

The U.S. first proposed a ban on data denial in the political context of SS–25 flight-testing, which we charged was a violation of SALT II throw-weight provisions. We also charged that Soviet telemetry encryption prevented us from monitoring SS–25 flight tests accurately, in violation of the limited ban on telemetry encryption that impeded verification of the treaty limits. The two sides have now agreed in START to the U.S.-proposed ban on data denial by “encryption, encapsulation, jamming, or any other means.”

The current U.S. position on data denial would have the effect of permitting us to continue denying Soviet access to U.S. telemetry data (through complex coding) while gaining U.S. access to Soviet telemetry (by banning encryption). The Soviets reject this asymmetry and have proposed either to exchange data in sufficient detail to permit a full understanding of telemetry by the other side, or not to exchange any data. They have indicated that they must be free to adopt U.S. practices—[less than 2 lines not declassified]

There are two interrelated issues:

— how to define treaty commitments to ensure that each side can receive the other’s telemetry transmissions (the “power levels” issue),

— what obligations should apply to explaining or interpreting the telemetry that is received (the “data exchange” issue).

Power Levels

Both sides currently can read each other’s telemetry. The Soviets, however, insist on the right to adopt “U.S. practices.” [2½ lines not declassified] Raising the U.S. power level could cost $700 million–$1.7 billion and take five years to achieve, but the result would not significantly improve the already good Soviet ability to receive U.S. telemetry. [less than 2 lines not declassified]

The IC would like to freeze both sides to their current practices, a proposal the Soviets have refused. A fallback is to freeze current practices for existing types and to define transmission practices that would apply equally to both sides for future types of missiles. This approach has merit and should be tried out on the Soviets.

[4½ lines not declassified] The Soviets have linked this issue to data exchange (see issue below), arguing that there is no need for a tape exchange if they cannot read the tapes.

As we discussed, “power levels” is a shorthand term for a set of technical parameters. Your meeting should not get diverted into a discussion of these technical issues.

Data Exchange

The critical data denial issue is to weigh the benefits for treaty monitoring and threat assessment of gaining access to Soviet telemetry against the costs of permitting the Soviets to read and understand U.S. telemetry. The various protagonists in this debate can more or less describe what additional information we or the Soviets might gain, but no one is very good at explaining how exactly the additional information could be exploited to our security benefit or detriment. The IC can provide a somewhat clearer picture of the implications of the ban on data denial for the narrow issue of START monitoring.

[3 paragraphs (19 lines) not declassified]

On the other side of the issue, U.S. ICBMs and SLBMs have had reliability and performance problems that appeared only in telemetry. DOD argues that these flight-test problems would give the Soviets valuable information on the status of U.S. strategic forces and their capabilities. Some have suggested that such information could alter perceptions of the balance in ways that would decrease deterrence and crisis stability. There have also been suggestions that the Soviets could gain information useful in designing ballistic missile defenses.

There is little doubt that the Soviets will be able to learn more about the details of U.S. missiles with access to our telemetry. The question is what they would be able to do with that knowledge. The Services believe strongly that the details, which may not be available to the Soviets through other sources, are vitally important. So far they have not explained their case well, and Don Atwood seemed skeptical when they tried to do so with him.

Politically, the U.S. has had a clear position since 1982 to ban data denial and has offered this ban as a primary benefit of START. Although the Administration will face criticism during ratification regardless of the outcome, the SSCI has already made clear that it regards this issue as of primary importance both for the treaty and for future Intelligence collection programs.

Proposals

The IC proposes a full tape exchange and data exchange (i.e., enough information to allow confident interpretation of the tapes provided by the other side). They would base the requirements for information exchange on parameters included in a presently unclassified Air Force manual. (The counterpart Navy manual is presently classified Confidential.) [less than 6 lines not declassified]

DOD seems prepared to offer a small number of telemetry channels related to booster acceleration, but strenuously resists offering in anything approaching a full data exchange. That position is far short of what the IC says it needs for monitoring START.

This is a case in which a clean decision is necessary. That will require referring the issue up to the President for resolution, not down to the working level for further analysis. Any attempt to split the difference will suffer the disadvantages of both approaches and the advantages of neither. Absent such a clean decision, we would be likely to find ourselves in protracted bureaucratic negotiations when we are trying to wrap up START, and in the end come up with a proposal that both the IC and the Pentagon might grudgingly accept, but the Soviets almost surely would reject.

Art Kuehne concurs.

Tab 1

Paper Prepared in the Central Intelligence Agency

Washington, undated

[Source: George H.W. Bush Library, Bush Presidential Records, National Security Council, John A. Gordon Files, Subject Files, OA/ID CF01033–006, START—December 1990. Secret; [handling restriction not declassified]. 6 pages not declassified.]

Tab 2

Paper Prepared in the Office of the Joint Chiefs of Staff³

Washington, undated

What Do We Need to Know About the Soviet Ballistic Missile Threat in a START Environment?

The national security strategy of the United States clearly states that deterrence of nuclear attack remains the cornerstone of U.S. national security. Regardless of improved U.S.-Soviet relations and potential arms control agreements, the Soviet’s physical ability to initiate strategic nuclear warfare against the United States will persist and a crisis or political change in the Soviet Union could occur faster than we could build neglected strategic forces. A START agreement will allow us to adjust how we respond to the requirements of deterrence, but tending to those requirements remains the first priority of U.S. defense strategy.

Intelligence assessments of Soviet strategic ballistic missile capabilities and force deployments under START remain one essential element in the calculus of evaluating the military sufficiency of a START treaty, and how well the national security is protected under that treaty. But our information concerning the Soviet ballistic missile threat will never be perfect. The critical question is how much information we need, and whether that information can be obtained even in the absence of U.S. access to fully interpretable ballistic missile telemetry.

Military assessments of the post-START Soviet threat have assumed that Soviet ballistic missiles would become increasingly accurate and lethal through the 1990s. A refined knowledge of potential range, accuracy, footprint capability, time-on-target control, navigation updates and maneuvers, and so forth adds little to the fundamental conclusion that the Soviets are negotiating a START treaty which allows them to achieve their most important targeting goals. [less than 2 lines not declassified]

On the other hand, if the Soviets had full access to U.S. ballistic missile telemetry data for developmental or operational tests, assessments of the sufficiency of START—the ability of the U.S. to deploy and operate nuclear forces sufficient to carry out Presidential guidance—could be dramatically affected. Since the first U.S. strategic ballistic missile went on operational alert in 1959, there have been a number of instances in which tests have revealed system or engineering problems which degraded the operational status or potential performance of U.S. ballistic missiles. The problem often is not one of visibly catastrophic failure, but of subtle malfunctions or problems which only are detected through exploitation of telemetry data. Denying the Soviets full access to ballistic missile telemetry has given the U.S. time to analyze and fix such problems in the past, without endangering national security.

It also is instructive to keep in mind that knowledge of problems affecting an entire class of ballistic missile systems favors the side which has the larger number of systems and which has an employment strategy which includes real options for first use (preemptive or otherwise) of strategic nuclear weapons.

Another aspect of the military equation affected by U.S. knowledge of Soviet ballistic missile capabilities is the prospect for strategic defense. [less than 6 lines not declassified]

And even if the U.S. does not have full access to Soviet ballistic missile telemetry, it will have considerable knowledge about the size and scope of the Soviet ballistic missile threat. Telemetry is one source (albeit a very important one) of information; it is not the only source.

In weighing the above considerations, there is a strong case for concluding that the U.S. has more to lose than gain by allowing full access to ballistic missile telemetry.

[1 paragraph (2 lines) not declassified]

In advising ratification of the INF Treaty, the Senate Select Committee on Intelligence and the Senate Foreign Relations Committee singled out militarily significant cheating as a central verification standard. The Final Report of the Senate Foreign Relations Committee concluded: “The START verification regime should be designed to detect militarily significant cheating in time for the United States to respond effectively.”

Full access to Soviet telemetry will provide a comprehensive understanding of the internal functioning of the missile system. The question is whether that full understanding is necessary to detect militarily significant cheating in time for the United States to respond effectively. In assessing the military sufficiency of the emerging START treaty, the JCS have concluded that post-START Soviet forces would be able to meet Soviet targeting goals within the constraints of the treaty. [3½ lines not declassified] There is a web of verification provisions in addition to the ban on telemetry encryption which decreases the risk that the Soviets could exploit ballistic missile testing in a way which would fundamentally weaken deterrence.

Does Full Access to Ballistic Missile Telemetry Set a Good Precedent for START II?

The debate over full access to ballistic missile telemetry largely involves the question of what kinds of cooperative practices will be mandated by the Treaty. Until early this year, the Soviets demanded that telemetry encryption be banned for cruise as well as ballistic missile tests. The U.S. determined that unencrypted cruise missile telemetry was not in the U.S. interest, and successfully resisted this ban.

[1 paragraph (5 lines) not declassified]

1. Source: George H.W. Bush Library, Bush Presidential Records, National Security Council, John A. Gordon Files, Subject Files, OA/ID CF01033–006, START—December 1990. Secret. Sent for information. Sent through Kanter.
2. No minutes were found.
3. Secret; Noforn.