184. Memorandum From the Director of the Arms Control and Disarmament Agency (Warnke) to President Carter1


  • On-Site Inspections (OSIs) Under a Comprehensive Test Ban

I believe that our exploration of the OSI issue with the Soviets has already gone as far as it productively can and that, if we are to avoid an impasse in the CTB negotiations, we must make a formal proposal on OSIs when the talks resume later this month. Specifically, I recommend that we now propose a compromise approach that is responsive to the Soviet principle of “voluntariness” and, at the same time, meets basic U.S. requirements for an effective inspection provision. The principal reasons for this recommendation are that:

—a U.S. proposal for mandatory OSIs would almost certainly deadlock the negotiations, raise doubts in the minds of the Soviets and others about the sincerity of our commitment to conclude a CTB in the near future, and jeopardize our efforts to obtain further Soviet concessions on other CTB issues of greater importance to us;

—the actual technical contribution to our CTB verification capability of any form of on-site inspection, whether mandatory or voluntary, is quite limited;

—difficult practical and legal problems would arise for the U.S. under a mandatory OSI system because of the need to deny access to sensitive U.S. national security facilities and to respect constitutional rights regarding private property;

—we would encounter serious domestic problems if we formally proposed mandatory OSIs and then, as must be expected, had to fall back in order to reach an agreement; and

—the overall verification package that now seems achievable should put us in a good position to explain to the Congress and to the American people that effective and reasonable measures have been worked out for verifying the treaty. This package is likely to include improved national technical means, authenticated data from seismic stations in the USSR, and a new challenge OSI approach (described [Page 435] below) that eliminates the key defects of the Soviet proposal while preserving the basic deterrent quality of the mandatory approach.

These reasons, as well as my suggestions for a specific solution, are discussed below.

Discussion of OSI during the December Round. Although we tabled a U.S. Working Paper in December outlining the key elements of a CTB treaty and protocol, the paper did not contain a proposal on OSIs.2 Instead, as instructed, the U.S. Delegation expressed a desire to narrow the differences between traditional U.S. and Soviet positions and explored at considerable length Soviet receptivity to an annual quota of mandatory inspections.

Soviet reactions to our probes on mandatory OSIs, which were made on a number of occasions and at several levels, were uniformly and categorically negative. Morokhov, head of the Soviet Delegation, said that we should have no illusions: his government would reject any U.S. proposal for mandatory OSIs and would regard it as a backward step intended to complicate the task of negotiating a treaty at a time when the USSR had made several important concessions in order to make agreement possible. He maintained that the 1962 Khrushchev offer of 2–3 annual inspections had been discredited and formally withdrawn by the Soviet Government and that all segments of their bureaucracy were opposed to it.3

In response to our calls for narrowing the differences between our traditional positions, the Soviets asserted that they had already demonstrated their willingness to compromise. Morokhov pointed out that, despite the USSR’s long-standing opposition to any form of OSI, they had accepted the Swedish-developed concept of voluntary, or “challenge”, inspections. He also cited Soviet readiness, announced during the recent October round, to work out in advance the detailed rights and functions of inspection teams.

Limited Technical Value of OSIs. The Interagency Working Group’s study of September 19774 demonstrated that the actual verification value of on-site inspection must be considered quite limited, given the small probability of discovering evidence of a nuclear explosion with current OSI techniques if the evader has taken certain precautions (e.g., sufficient depth of burial, careful stemming and emplacement hole construction). Because of uncertainties in fixing the precise location of suspicious events, OSI visits would, in most cases, be like looking for a needle in a haystack. As the study points out, success would not be as[Page 436]sured even if the location of a covert test were known within ten meters.

For these reasons, the interagency study concluded that the value of an OSI provision in a CTB was largely political. Any state rejecting a demonstrably reasonable OSI request would create the impression internationally that it had something to hide and would give another state grounds for withdrawal. This would be the case whether the rejection was prohibited under the terms of the treaty (i.e., mandatory approach) or legally permitted (i.e., voluntary approach).

Practical/Legal Problems of Mandatory OSIs. Even under a nominally “mandatory” system, neither we nor the Soviets could tolerate OSIs at sensitive locations, such as missile sites. The attempt to work out criteria for exempted locations would be very difficult, and any criteria that the Soviets could support—given their broad definition of sensitive facilities—might well be so general as to raise domestic concerns about OSI “sanctuaries” in the USSR. In addition to the question of national security exemptions, serious legal problems would arise in seeking to grant the USSR mandatory inspection rights on privately-held territory in the U.S. Given constitutional and statutory protections of U.S. citizens and their property, the U.S. Government’s ability to accede to a Soviet OSI request could well depend, in a number of circumstances, on judicial determinations.

Both of these legal dilemmas would be minimized under a voluntary OSI system. In the case of a sensitive area, an OSI request could either be turned down for valid national security reasons or the area to be inspected could be altered through negotiations to exclude the sensitive facility. In the case of a privately-held area, the U.S. Government either could turn down the request on the basis of the landowner’s and judge’s objections or could negotiate with the Soviets and the landowner to restrict the scope of the OSI visit to a mutually acceptable level.

Domestic Considerations. If we were to propose mandatory OSIs, it is only realistic to assume we would have to fall back at a later stage in order to reach an agreement. When our “concession” became public, as it certainly would, it would be characterized by critics as a major U.S. retreat and could even affect prospects for ratification. It seems far preferable, in my view, to make a determination now, based on the limited technical utility of OSIs, that a mandatory OSI provision would contribute little, if anything, to adequate verification of a CTB and that therefore there is no good reason to seek such a provision.

Critics of a CTB will no doubt try to make an issue of verifiability. However, I believe that the verification measures that are likely to be achievable—including authenticated data from seismic stations in the Soviet Union, challenge OSIs, as well as enhanced national technical [Page 437] means—will strike the American people and Congress as an impressive package. Moreover, such a package would represent a considerable advance beyond where we were just a few years ago, when the Soviets were insisting on national technical means alone.

Negotiating Considerations. There is little doubt that a U.S. proposal for mandatory OSIs would be unacceptable to the Soviets and would lead to a stalemate in the CTB negotiations. But the cost of making such a proposal is not just that resolution of the OSI issue would be delayed until we decided to abandon the idea of mandatory inspections. U.S. insistence on mandatory OSI has long been regarded worldwide as a reflection of U.S. unwillingness to stop nuclear testing, and a formal U.S. proposal would be seen by the Soviet leadership—and by the rest of the world when it became generally known—as a throwback to earlier U.S. attitudes toward a CTB.

In addition to stalemating the OSI issue and casting doubt on our desire to conclude a CTB in the near future, a proposal for mandatory OSIs would work against our efforts to obtain further Soviet concessions in areas of much greater actual importance to us, especially internal seismic stations and PNE-related issues, and would thus be getting our priorities wrong. The Soviet representatives have stressed that they expected reciprocal movement by us for their November 2 initiative on PNEs and entry into force, which they regard as having made agreement possible, and that they were disturbed by our failure in December to take any tangible steps toward them on the OSI issue. Indeed, by the end of the December round, we could discern some stiffening of Soviet positions, particularly on internal stations. I believe we have to face up to the prospect that, until we meet the Soviets half way on OSIs, they will not be inclined to meet our concerns on other issues.

The British have recently informed us via diplomatic note (attached)5 that they do not believe further progress can be made toward a CTB without Western willingness to abandon mandatory OSIs and that they hope we will be prepared to do so when the negotiations resume later this month.

Seeking an Acceptable Compromise. The present Soviet OSI proposal contains a number of serious deficiencies. Specifically, OSI requests are to be based on seismic evidence that meets certain technical criteria, thus making it impossible even to request an inspection if the available evidence is either non-seismic (e.g., overhead photography) or is seismic but does not meet the criteria. Then, once a request is made, the party on whose territory the questionable event has taken place can simply reject it out of hand, without even providing an explanation.

[Page 438]

I recommend that we make a counterproposal designed to shift the burden of proof from the requesting party (where it is placed under the Soviet proposal) to the suspected party as well as to raise the political costs that the suspected party would incur if it refused a reasonable OSI request without giving justifiable grounds for doing so. The formulation I recommend, which would be incorporated in the separate verification agreement that will be concluded with the Soviets to supplement the multilateral CTB treaty, would contain the following elements:

—Each party would have the right to request an OSI at any time (i.e., there would not be a ceiling on the number of requests that could be made).

—Any type of evidence, seismic or otherwise, could justify an OSI request. There would be no technical criteria for screening out events as not eligible to invoke an OSI request.

—As in the Soviet proposal, the party on whose territory the event took place would have the right to turn down the request. However, if it decided to do so, it would be obligated to provide a detailed justification for its decision, including any seismic, photographic, or other evidence that it believed demonstrated that no violation had been committed.

—If the requesting side were not satisfied by the explanation, it could call for further information. If it remained unsatisfied by such additional information, it would be entitled to bring the matter to the U.N. Security Council.

—The parties would issue an agreed interpretative statement to the effect that arbitrary refusals to grant OSI requests could undermine confidence in the treaty and could create a situation in which a party whose requests had been arbitrarily denied might consider that its supreme national interests were being jeopardized. The clear implication would be that a party could legitimately withdraw from the treaty if it felt that its requests for OSI were being turned down on insufficient grounds.

The principal distinction between a mandatory OSI system and a purely voluntary one is that a refusal to grant an inspection under the former type would constitute a treaty violation and would thus justify withdrawal from the treaty by a party whose interests are threatened. The agreed interpretative statement described above would provide the functional equivalent of that advantage, and would serve as a significant disincentive against arbitrary rejections of OSI requests. The recommended approach, by requiring suitable justification for any OSI refusal and by eliminating Soviet-proposed barriers to making an OSI request, would tend to create the presumption that the state which is asked to accept an inspection visit will either grant that request or provide a convincing explanation why the visit is unnecessary.

[Page 439]

By providing explicitly for bringing the matter to the Security Council, the Soviet Union would be put on notice that it can turn down a reasonable U.S. OSI request only at considerable political cost. Given the Soviet Union’s veto, we would not, of course, ask the Security Council to take corrective action. However, the threat that its arbitrary behavior could be exposed publicly in a forum that commands wide international attention would strengthen the overall value of the provision as a deterrent against arbitrary rejection of an OSI request.

Recommendation. Although I believe that OSIs, in general, have very limited utility as a CTB verification measure, I am confident that the approach outlined above for initiating OSIs, together with prearranged rights and functions of inspection teams that would be triggered when an OSI request is accepted, would give us as much value as could be derived from any OSI provision, including the idea of a mandatory quota. I therefore recommend that you authorize us to propose that approach when the CTB negotiations resume.

I have already sent copies of this memorandum to State, Defense, JCS, Energy and CIA, and they all agree with the basic approach regarding challenge OSIs that I have outlined. OSD, JCS, and Energy have commented to me that we should use our move on OSI in a manner that enables us to obtain Soviet concessions on matters of importance to us, especially on duration of the PNE ban and internal seismic stations. JCS suggests the approach be explored, but not tabled at this time. If you approve my recommendation, I will present our new position in the way I believe will best maximize our pressure on the Soviets to move our way on the critical remaining issues. I would make clear that our willingness to accept the challenge OSI approach is contingent on reaching a satisfactory solution on the entire treaty.

Paul C. Warnke6
  1. Source: National Archives, RG 383, Records of the Arms Control and Disarmament Agency, Office of the Director, Paul C. Warnke Files, December 1974–July 1979, Accession #383–98–0154, Box 1, Memoranda to the President Regarding SALT, Cruise Missiles, CTB, NPT, and Indian Ocean Arms Control, January–October 1978. Secret. A copy was sent to Keeny. Drafted by Robert Einhorn (ACDA/MA).
  2. See footnote 2, Document 180.
  3. See footnote 6, Document 180.
  4. The Working Group study of OSIs is available in Washington National Records Center, OSD Files: FRC 330–80–0017, Box 63, A–400.112 TEST BAN (27 SEP 77) 1977.
  5. Attached but not printed is a December 29, 1977, paper from the British Embassy entitled “Comprehensive Test Ban.”
  6. Printed from a copy that indicates Warnke signed the memorandum.