159. Letter From the Director of Central Intelligence (Colby) to President Ford1

Dear Mr. President:

In early September, I received a memorandum2 from your Assistant for National Security Affairs summarizing certain recommendations submitted to you by your Foreign Intelligence Advisory Board regarding the production of National Intelligence Estimates. Attached to that memorandum was a draft Presidential directive for undertaking and evaluating an experimental modification of the present process for developing estimates in two specific strategic areas: anti-submarine warfare and the accuracy of ICBMs. I was asked to give you my comments on the PFIAB recommendations and on the proposed experiment. This letter constitutes my response. In addition to my own views, it also reflects the views of my colleagues in CIA and in other components of the Intelligence Community responsible for contributing to our strategic assessments of Soviet capabilities. The draft text of this letter was reviewed, discussed and unanimously endorsed by the United States Intelligence Board.

As summarized in the memorandum and the accompanying draft directive, the new procedure would involve:

a. The development of an estimate of Soviet capabilities in these two key areas by “an independent analysis group composed of Intelligence Community and non-government representatives.” This experimental estimate would be a “purely intelligence document which avoids net assessments.” It would be something independent of, and prepared separately from, the National Intelligence Estimate in which Soviet capabilities in these areas are already considered: NIE 11–3/8–75.

b. A subsequent detailed net assessment of Soviet and U.S. capabilities. In the two experimental areas, the draft directive suggested that the net assessments be prepared by an ad hoc working group established under the auspices of the Interdepartmental Political-Military Group.

c. A thorough critique of the net assessment by an independent entity. In the experiment, as suggested in the draft directive, the NSC Under Secretaries Committee would make a comparison and critique of the independently prepared estimates and the net assessments de[Page 715]scribed above, and compare both with the treatment of the same subjects in NIE 11–3/8–75.

Through subsequent discussions with the NSC Staff and the PFIAB Secretariat, we learned that:

a. The NSC Staff’s summary recommendations were intended to implement those contained in the PFIAB’s memorandum to you of 8 August 1975.3

b. The recommendations for change were not intended to apply to all National Intelligence Estimates but only to estimates in the NIE 11–3/8 series (Soviet Forces for Intercontinental Conflict).

c. In suggesting the above-described experimental procedure, neither the NSC Staff nor the PFIAB intended to disrupt or delay the preparation of this year’s NIE 11–3/8–75, which was then in its final stages of preparation. It has now been approved by the United States Intelligence Board and is being published.

I would like to comment on some of the points raised by the PFIAB in its 8 August memorandum to you, which served as the stimulus for these recommendations. That memorandum expressed the PFIAB’s view that last year’s National Intelligence Estimate on Soviet strategic capabilities—NIE 11–3/8–74: Soviet Forces for Intercontinental Conflict Through 1985—“is seriously misleading in the presentation of a number of key judgments and in projecting a sense of complacency unsupported by the facts; as a consequence it is deficient for the purposes it should serve.” This view appears to be based on a belief that:

a. NIE 11–3/8–74 was not sufficiently explicit on important uncertainties underlying our intelligence judgments, particularly on a few vital technical issues such as the accuracy of Soviet ICBMs and the progress of Soviet research on anti-submarine warfare.

I certainly share the PFIAB’s view that “National Intelligence Estimates should be among the most important documents issued by the Intelligence Community.” NIE 11–3/8–74 was the product of a still-continuing evolutionary process through which the Intelligence Community is endeavoring to make each of these major annual assessments of Soviet strategic capabilities better than those of preceding years. While I would not contend that NIE 11–3/8–74 was a perfect document, I cannot agree with the PFIAB’s contention that it errs by “projecting a sense of complacency” or, for that matter, in offering any judgments “unsupported by the facts.”

b. NIE 11–3/8–74 contained what appeared to be “net assessments” of U.S. and Soviet strategic capabilities, when detailed operational analysis of strategic conflict required for genuine net assessment was lacking.

[Page 716]

There are clearly specific issues on which individual members of the PFIAB differ with the Intelligence Community’s conclusions. But the estimate as a whole depicted Soviet strategic capabilities that are steadily improving in many areas and will continue to improve, even in a climate of détente and even if a SALT Two agreement is successfully negotiated. I hardly consider this judgment any valid basis for complacency, even though NIE 11–3/8–74 also concluded—on the basis of rigorous analysis of all available evidence—that the Soviets are unlikely within the next decade to have deployed operational weapons systems enabling them to launch an attack that would prevent devastating U.S. retaliation.

With respect to the specific proposals of the NSC Staff, my comments are as follows:

a. I welcome the evaluation, by consumers, of the utility of our intelligence products and any suggestions on how those products can be made more informative and enlightening to the policy officials for whom they are written.

b. I also welcome any improvements in the U.S. Government’s procedures for developing net assessments of U.S. capabilities with respect to those of potential or putative adversaries. This task goes well beyond the scope of intelligence estimates—which, by definition, are focused on the capabilities and intentions of foreign powers. It is a task, however, to which a sound intelligence input is essential. As you know, at various times over the past several years, the net assessment function has oscillated between the NSC Staff and the Department of Defense. The responsibility for net assessments needs to be more clearly assigned and a better mechanism needs to be developed for producing them on a regular, systematic basis—drawing on intelligence inputs plus the details of U.S. capabilities and operational plans. The Intelligence Community will, of course, provide any support or assistance it can to new procedures, or experiments with new procedures, designed to improve the quality of U.S. net assessments.

c. The intelligence estimating experiment proposed by the NSC Staff, however, gives me some trouble. Our annual estimates on Soviet strategic capabilities—the NIE 11–3/8 series—utilize all the information known by and the best analysis available to the U.S. Government. Undergirding the production of the actual estimate itself—e.g., NIE 11–3/8–75—is an extensive research program examining specific aspects of Soviet capabilities in considerable detail, a research program involving not only all concerned elements of the Intelligence Community but also drawing on the views and talents of knowledgeable experts in specific fields outside the government. It is hard for me to envisage how an ad hoc “independent” group of government and non-government analysts could prepare a more thorough, comprehensive assessment of Soviet strategic capabilities even in two specific areas—than the Intelligence Community can prepare.

An “independent” group could, of course, produce a sharply drawn set of scenarios, outlining various capabilities the Soviets might be able to develop. Such alternative scenarios or hypotheses were in[Page 717]deed discussed, and carefully weighed, in and during the process through which NIE 11–3/8–75 was prepared. The actual estimate, however, reflects my strong belief that intelligence has a dual set of responsibilities to those for whom it is produced. It clearly has the responsibility of warning its consumers of risks and potential problems, of various things the Soviets might do. What some miss or ignore is that intelligence also has a responsibility for making an assessment of the relative likelihood of such unpleasant contingencies, of saying what capabilities—in its best judgment—the Soviets are not likely to develop in given time frames. Our present process for producing national estimates is designed to discharge both sets of responsibilities, not just the first.

All of us in the Intelligence Community are constantly seeking ways in which we might improve the quality and utility of our estimates. This year’s NIE 11–3/8–75, in fact, has incorporated several innovations, including the discussion and assessment of developments of low probability but of great potential significance, should they occur. Two separate sessions of the United States Intelligence Board were devoted to this estimate before it was issued. On 14 November,4 the Board spent the entire day on a thorough presentation, which included adversary debate, of the evidence and alternative judgmental conclusions in seven critical areas, including both ASW and ICBM accuracy. On 17 November, the Board addressed the actual text of the estimate, and its Key Judgments, page by page.5

The published version of 11–3/8–75 will be in the hands of concerned consumers, including the PFIAB, within the next few days. I would suggest that the best, most efficient way to proceed would be for those consumers—especially the PFIAB—to scrutinize NIE 11–3/8–75 and ascertain the extent to which it overcomes or rectifies what they may have perceived as deficiencies in NIE 11–3/8–74. After this process of review has been completed, my representatives or those of my successor—can then sit down with members of the PFIAB and the NSC Staff to discuss specific courses of action most likely to be of value in our joint, continuing quest for a better national intelligence product.

W. E. Colby
  1. Source: Central Intelligence Agency, National Intelligence Council, Job 91M00696R: Subject Policy Files, Box 7, Competitive Analysis Background, 1975. Secret. Colby forwarded the letter to Scowcroft under a November 24 covering memorandum. (National Archives, Nixon Presidential Materials, NSC Files, NSC Institutional Files (H-Files), Box H–282, Intelligence Files, PFIAB (1) [1971–1975] [3 of 3])
  2. Document 156.
  3. Document 155.
  4. Colby presumably means 13 November. See Document 157.
  5. See footnote 1, Document 157.