I. The Institutional Framework




A. The Director of Central Intelligence and the NIE

The National Intelligence Estimate--spelled thus with capital initial letters--was one of the major innovations of General Walter Bedell Smith, the fourth Director of Central Intelligence [DCI], whose incumbency bridged the period 7 October 1950-24 January 1953.(2)

The title itself proclaims at least two important messages. First, the use of the word "estimate"--as distinct from "report" or "study"--shows the Director's concern to emphasize this particular form of intelligence utterance and its importance in his thinking. In this General Smith reflected a similar bent of his deputy, William Harding Jackson, who as an intelligence officer during World War II had had a first-hand experience with estimates, had made a deep study of the institution as practiced at high levels of British intelligence, and had himself written a section on national estimating in the Dulles-Jackson-Correa report.(3)

The second, the use of the word "national" was employed with equal purpose. It not only designated a type of subject matter suitable for purposes of national security policy formulation, and a hoped-for quality appropriate for use at highest levels of government, but more especially an intelligence production effort which would engage the knowledge and talent of the national intelligence community over which the DCI was the presiding officer. Indeed that thing often referred to as "national intelligence" had been declared to be one of the three principal charges on the DCI.(4) He and he alone was under obligation to produce it. Terming the estimates-to-be national would put them clearly within the larger canopy of "national intelligence" and as such within the personal jurisdiction of the DCI.(5)

Thus the first, and by all odds most important, legal and constitutional aspect of the National Intelligence Estimate is that it was and is the Director's estimate, and its findings are his. Although many experts from perhaps all intelligence components of the community participated in the production of the papers in the NIE series, and although the intelligence chiefs themselves formally passed on the final text, they could not bend its findings to suit their own judgments contrary to the will of the DCI. They could try to win him to their sides by full and free discussions, but they could not outvote him and force him to join them, nor could they make him dissent from them, even though they constituted a clear majority of the Intelligence Advisory Board, Intelligence Advisory Committee, or the Untied States Intelligence Board as it was successively known. By the same token, the DCI could not oblige them to join him in a matter at dispute. They could of their own accord concur with his findings, or, not being able to, they could dissent and make their alternative views known in footnotes to his text.

In his very first full dress meeting with his IAC on 20 October 1950 General Smith tactfully but forcefully made the matter clear.

The minutes for that historic meeting are gratifyingly full; they contain a verbatim rendering of a memorandum which General Smith read to his colleagues.(6) He began with the title: The Responsibility of the Central Intelligence Agency for National Intelligence Estimates and went on to read: "One of the principal duties assigned to the CIA--is to "correlate and evaluate intelligence relating to the national security and provide for its proper dissemination.'" The memo elaborates the intended significance of this phrase from the National Security Act of 1947, and continues: "The CIA is thus given the responsibility of seeing to it that the United States has adequate central machinery for the examination and interpretation of intelligence, so that the national security will not be jeopardized by failure to coordinate the best intelligence opinion in the country, based on all available information."

The logical construction goes on abuilding: Although the National Security Act provided that the departments and agencies of the government shall continue to collect, evaluate, correlate, and disseminate departmental intelligence, it does not limit the duties of the CIA vis--vis its intelligence mission except by the standard of national security. In fact, "the Act apparently gives the CIA the independent right of producing national intelligence. As a practical matter [such national intelligence emanating in the form of] estimates can be written only with the collaboration of experts in many fields of intelligence and with the cooperation of several departments and agencies of the Government. A national intelligence . . . estimate as assembled and produced by the CIA should reflect the coordination of the best intelligence opinion based on all available information."

The memo went on: The concept of national intelligence estimates underlying the statute is that of an authoritative interpretation and appraisal that will serve as a firm guide to policy makers and planners. A national intelligence estimate . . . should be compiled and assembled centrally by an agency whose objectivity and disinterestedness are not open to question. "Its ultimate approval should rest upon the collective judgment(7) of the highest officials in the various intelligence agencies." Finally, it should command recognition and respect throughout the Government as the best available and presumably the most authoritative estimate. Although the task is made more difficult by a lack of general acceptance of the concept of national intelligence estimates in the Government, it is, nevertheless, the clear duty and responsibility (8) of the Central Intelligence Agency under the statute to assemble and produce such coordinated and authoritative estimates.

The "statute" to which General Smith had referred was, of course, The National Security Act of 1947,(9) notably its section 102, subsection (d)3, which reads:

(d) For the purpose of coordinating the intelligence activities of the several Government departments and agencies in the interest of national security, it shall be the duty of the Agency,(10) under the direction of the National Security Council--

(3) to correlate and evaluate intelligence relating to the national security, and provide for the appropriate dissemination of such intelligence within the Government using where appropriate agencies and facilities. . . .

Had General Smith desired, he could have given the background to those cryptic and not wholly satisfactory words of section 102 (d), (3). The fact is that President Truman used almost these exact phrases in his letter of 22 January 1946 addressed to the Secretaries of State, War, and Navy and in which he designated them the so-called National Intelligence Authority and directed them (and a fourth officer to be named by him) to plan, develop, and coordinate "all Federal foreign intelligence activities so as to assure the most effective accomplishment of the intelligence mission related to the national security." His letter went on to say that the addressees would assign persons and facilities from their departments, "which persons shall collectively form a Central Intelligence Group" under a Director of Central Intelligence, "who shall be designated by me."

The text immediately following says that the new DCI shall:

Accomplish the correlation and evaluation(11) of intelligence relating to the national security, and the appropriate dissemination within the Government of the resulting strategic and national policy intelligence. In so doing, full use shall be made of the staff and facilities of the intelligence agencies of your [i.e. State, War, and Navy] Departments.

A few paragraphs later on, the President ordained an Intelligence Advisory Board(12)--the first name given to the body which in General Smith's time was known as the Intelligence Advisory Committee. The letter did not describe the right of Board members to register dissents to decisions of the DCI, but that came soon in the very first directive which the National Intelligence Authority issued.(13)

The President's letter and the NIA directive were given additional strength (perhaps) and precision (certainly) in the first intelligence directive issued by the National Security Council a few months after the passage of the Act that called it into being. Paragraphs 4 and 5 of NSCID #1 (12 December 1947) read thus:

4. The Director of Central Intelligence shall produce(14) intelligence relating to the national security, hereafter referred to as national intelligence. In so far as practicable, he shall not duplicate the intelligence activities and research of the various departments and agencies but shall make use of existing intelligence facilities and shall utilize departmental intelligence for such production purposes. For definitions see NSCID #3.

5. The Director of Central Intelligence shall disseminate National Intelligence to the President, to members of the National Security Council, to the intelligence chiefs of the IAC agencies,(15) and to such governmental departments and agencies as the National Security Council from time to time may designate. Intelligence so disseminated shall be officially concurred in by the intelligence agencies or shall carry an agreed statement of substantial dissent. (Emphasis added)

Fast on the heels of this document came NSCD #3 (13 January 1948) which reiterated the DCI's duty to produce and disseminate national intelligence,(16) and two of the early DCIDs [Director of Central Intelligence Directives] which set forth the Standard Operating Procedures for Departmental Participation in the Production and Coordination of National Intelligence(17) and Policy Governing Departmental Concurrences [and Dissents] in National Intelligence Reports and Estimates. (18)

In other words, when General Smith told his colleagues of the IAC how he construed his powers under the National Security Act, he could have invoked a number of other forceful and explicit texts (which antedated the Act and followed it) to bolster his position. Of course, he did not need them, nor did he need them to support three other decisions which were essential parts of his new deal for estimates.

First was his announcement of his formation of a new office, the Office of National Estimates [ONE], whose only concern would be the production of national estimates and closely related matters. General Smith set great store by this office and indicated that "in his opinion it would be the heart of the CIA and of the national intelligence machinery."(19)

Not revealed in the official minutes, but in a memo for the record drafted by Col. Hamilton Howze, USA, an aide to the G-2 who was present at the meeting, was General Smith's mention of the future Board of National Estimates [BNE].(20) Colonel Howze's memo reads:

9. Within the new Estimates Division of ORE [sic] there will be a panel of five or six individuals constituting the top brains. General Smith is looking hard for a retired General or Admiral to head. He tried to get Admiral [Leslie] Stevens (recent Naval Attach, Moscow) and asked Admiral Johnson [Felix Johnson, the DNI (Director of Naval Intelligence)] to talk once more to Stevens in an effort to persuade him. General Smith also said he was anxious to get General [Clarence Ralph] Huebner to be a member of the panel, and possibly to head the Division.

(Be it said that General Smith did not get the services of Leslie Stevens, nor did he put General Huebner in charge of the new office. Huebner did accept a place on the Board, and the distinguished Harvard historian, William L. Langer, became its first chairman.) With this sort of official announcement, the ONE with its own administrative machinery was off to an auspicious start.

Second, the National Intelligence Estimate would be known as just that, not an "ORE" with a number, nor yet an "ONE," nor a "CIA" for that matter. It "would be published under a cover showing plainly that the estimate was a collective effort, the result of which would be labeled as a national intelligence estimate."(21)

Third, General Smith indicated his intention of holding IAC meetings "more often and for longer periods, although as chairman he would make every effort to keep the meetings as brief as possible. He stated that the IAC must be geared for rapid cooperative work."(22) In this he was true to his promise; the IAC began meeting regularly (and once a week) with the DCI seldom absent from the chair. As the NIEs moved into production, NIE business--whether the laying on, the clearing of scope notes, or pronouncing upon a finished product--became a staple of IAC fare. This was of course in marked contrast to the Hillenkoetter regime, where IAC meetings were rarely called and when called, never to participate in any phase of the pre-Smith brand of national estimates.

At this first IAC meeting, there was another piece of NIE business which was not exactly an innovation. It was in large measure a reminder of the production procedures which had first appeared two years earlier in DCID 3/1 and DCID 3/2. General Smith's restatement of these procedures was official notice of his desire to have things done according to the book.(23)

Perhaps to maintain the momentum which he had already given to the NIE, General Smith ended by calling for another meeting in five days to discuss "national estimates priorities and the frame of references and assumptions to form the basis of an intelligence estimate of the situation in Indo-China."(24)

In his rendering of the established procedures for doing NIEs General Smith added something new and important to the law as it was then understood. It was the content of his first sentence (paragraph "8a" in the Minutes): "The Intelligence Advisory Committee will adopt an intelligence plan, or more specifically, a list of required national estimates in an order of priority."

With this came into being two significant developments. The first had to do with the initiation of the NIEs.

Henceforward NIEs would be formally initiated by IAC action. Requests could come in from many quarters and did: a few times the President himself, often from the members of the NSC (especially from the Secretary of Defense in Mr. McNamara's time) or from the NSC Staff's chairman,(25) often from the second echelon in the Departments of State and Defense, from the DCI, IAC members, from the Board of National Estimates [BNE], and others. Such requests were usually referred to the BNE in the first instance, which would put the item on the agenda of the next IAC meeting or get an IAC authorization by telephone if time pressed. Upon occasion, when a request came in which was clearly not a suitable topic for the NIE treatment (something more akin to a National Intelligence Survey or a research study) the chairman of the BNE would try to deflect it to another component of the CIA. Failing this, the chairman of the BNE was bound to take the request to the IAC and try to make his case there for declining the honor. The point is, of course, the actual initiation of an NIE which would engage the talents of scores of people throughout the community was the decision of the community's highest body.

The second institution General Smith set in motion was that of planning the program of NIEs to come.(26) At the next meeting of the IAC, that of 26 October 1950, a program of 11 estimates was adopted in the following order of priority: the Philippines, Indochina, Soviet Capabilities and Intentions, Germany, Chinese Communist Capabilities and Intentions, Yugoslavia, Iran, Greece, Turkey, India, and Austria. At this moment there was not yet an ONE nor a BNE. General Smith turned to Ludwell Montague, who had handled the burden of the estimating in ORE, and announced that pending the establishment of an ONE, Montague would be in charge.(27)

In the next four weeks, while the ONE was in its formative stage, Montague placed six coordinated estimates before the IAC for final clearance. Three of them were from the original program, and three others were crash estimates related to the Chinese Communist intervention into Korea. By the end of November, the ONE was well established and Montague handed over the charge to Mr. Langer, who had the twin titles of Assistant Director for National Estimates (and as such was in charge of the new ONE) and Chairman of the Board of National Estimates.

My own appearance dated from about this time, and I well remember Montague turning to his new colleagues on the Board and suggesting pointedly that they begin to share the burden.


B. The Office of National Estimates (28)

From this time forward until 1 November 1973, the Office of National Estimates acted as the Director's executive agent for the acquittal of his responsibility for the production and dissemination of national intelligence estimates. One may date the Office's formal legal beginnings from the appointment of its chief, William L. Langer (13 November 1950). In these days before the creation of the Office of the Deputy Director/Intelligence [DDI], the AD/NE [Assistant Director/National Estimates] (along with five AD's of the so-called overt offices) reported to the Director through his deputy (the DDCI). Mr. Langer's mission and functions were spelled out in "CIA Regulation No. 70" of 1 December 1950. With the exception of one of its paragraphs, this document described the duties which he, his successors, and the office they presided over followed in guiding the procreation of more than 1,500 National Intelligence Estimates over 23 years. The paragraph which became inapplicable was #6, which had assigned to the AD/NE the current intelligence task and the issuance of the Daily Summary. In a matter of a few weeks, Mr. Langer had disengaged from this responsibility to concentrate his resources on the main task of the estimates.

The Office of National Estimates(29) took shape speedily. It should be viewed as consisting of three components: The Board of National Estimates, the professional staff, and the support staff.

The Board was the principal departure from what had gone before. In the thinking of General Smith and Mr. Jackson, the Board was to consist of an indeterminate number of senior officers (say, more than five and less than twelve), who came from a variety of professional backgrounds, and who, paid handsomely in the supergrade categories, had, (contrary to normal civil service practice) no administrative duties whatever. Their task was wholly substantive. Their days were spent in individual and more often collective efforts on every aspect of the estimates. They met first thing in the morning to hear the day's news and perhaps discuss it in terms of NIEs in the works or to come; they met again often with the ONE staff, often with representatives of the IAC agencies to talk about the schedule, to produce terms of reference, to review drafts, and to arrive at duly coordinated texts suitable to present to the Director and the IAC. They invited and listened to ambassadors, officers of the foreign aid program, attachès, members of the numerous military assistance groups (MAG, later MAAG), CIA officers in from the field, and many others. Above all they studied the new intelligence. Each day their reading room received a wide spectrum of the daily take which ranged from routine items like the FBIS [Foreign Broadcast Information Service] reports, CIA, attach, and State Department cables to the most sensitive materials that lay in the arcane codeword areas on the far side of Top Secret. This was the daily grist for thought and discussion. Indeed, almost as much as the labor on the draft estimates, the reading of the highly privileged news made its contribution to the collegial nature of the Board. And it was this very group effort that so often resulted in the posing of the right questions and the struggling for the best answers. As one Board member has pointed out, the collegial spirit also made its contribution to a finished product of high quality. There were always, he remarks, one or two colleagues who had not been so immersed in a paper as to be bored with it and willing to let it go forward irrespective of flaws. Seemingly there was almost always one of these fresh brethren who stepped in as a potent "no" man.

At the start, the Board consisted of Mr. Langer, myself, who was named his deputy early in 1951, General Clarence Ralph Huebner and Admiral Bernard Bieri (General Smith here deferred to his own background and the important role of the military in the intelligence community), Maxwell Foster (a Boston lawyer nominated by Mr. Jackson), Raymond J. Sontag and Calvin B. Hoover (Mr. Langer's choices: two outstanding professors of modern history and economics respectively), and DeForest Van Slyck and Ludwell Montague (senior officers of CIA's Office of Reports and Estimates [ORE]). The latter two, who had had many years of intelligence experience including three or four as estimators in ORE, brought with them a high competence for the task, and a rich first hand knowledge of the grandeurs and miseries of coordinating speculative intelligence at the national level.(30)

Along with Van Slyck and Montague, ONE inherited a much broader legacy from ORE. Most obviously, we recruited our full staff, both the professional and support components, from ORE. Let me speak of the professionals first.

In the beginning there were about 25 of them, two decades later, a few more than 30. Most if not all of them had had graduate school work in history or the social sciences, and most if not all had served in wartime intelligence work (with one of the military intelligence organizations or OSS). They had improved their regional or functional competence in their duties with CIA. They also, like Van Slyck and Montague, knew a lot about the post-war intelligence community, its strengths and weaknesses, and how to do business with it. They set a pace for a quality of workmanship that we were able to maintain during the lifetime of the Office. For 20 years they were the best staff in town and so proclaimed by a good number of very knowledgeable outsiders.(31)

The support staff, also recruited from ORE, was made up of about the same number of skillful women (growing eventually to about 35) who controlled the distributing in ONE of the daily flood of incoming intelligence materials, ran the ONE library, did the general secretarial work for the Board and the professional staffs, and attended to the reproduction in multiple copies of the endless stream of NIEs in every stage of their creation, first, second, third, and nth draft right up to the final manuscript for dispatch to the printer. The capabilities of our little reproduction staff were a nine-days' wonder throughout the community's band of estimators.

Thus the ONE at the beginning owed much to what had gone before. If all of us in the office had been newcomers like the members of the Board, and if all of us had had to learn the complicated trade from scratch, our fast start and speedy accomplishment would not have been.

With time there were great changes in the manning of both Board and Staff. We were careful about replacements and maintained the standards of excellence. One thing greatly in our favor was a refusal to try to build an empire and stretch our table of organization [T/O] to imperial dimensions. In the beginning our T/O was set at 85, a figure we never reached. For 1951 we had fewer than 60 people aboard. Ten years later, with a considerably larger work load, we reached a total of something under 70, perhaps a dozen of whom were on the Board. Some of the latter were new recruits from outside and some were former members of our staff or other CIA staffs whom the Director raised to Board status.

The original concept was that Board members should be "generalists" without specialized expertise in, or estimative responsibility for, particular geographic or functional areas. Over the years, certain specialization began to emerge informally. A Board member by virtue of being assigned to chair a succession of papers on a particular area, or by reason of his own growing interest and study, would become more knowledgeable than his colleagues about a particular problem or part of the world.

Furthermore, as members of the staff, which was organized on a regional basis, began to become members of the Board, they of course brought with them the more profound knowledge of the areas to which they had been assigned. Papers on "their" areas were more often than not given to them to shepherd through the trials of examination by the Board and coordination with the Reps. Thus, without any very conscious plan, a sort of specialization developed within the Board. This had the notable advantage of enabling the Board member so qualified to be more useful in the various stages of drafting and coordination.

Some anomalies developed, for example, Middle East specialists from the staff were appointed to the Board in numbers out of proportion to the other area experts so that, to the extent Board members were admitted to have specialties, we were overendowed with Middle Easterners. But the unsystematic system worked pretty well. The chairman of a paper would see to it that a couple of his colleagues would follow its development closely enough to be able to lend a hand if trouble developed in a Reps' meeting, and most of the other Board members would have had their say before then.

Later, when Abbot Smith took over as head of the ONE with John Huizenga as his deputy, a more formal effort at specialization was launched. Board panels were established, each responsible for a particular area, and each with a Board member in charge, with two of his colleagues also assigned. This was well enough, but there was a corollary: Board members were at least tacitly discouraged from concerning themselves with the doings of a panel to which they were not assigned. Doubtless this saved time in the Board consideration of an estimate, but it also narrowed the range of inspection to which an estimate was subjected. In this situation, the views of a panel chairman sometimes came to have inordinate weight.


C. The Representatives of the Other Intelligence Agencies

With the beginning of the ONE came a marked change in the manner of coordinating estimates with the other members of the intelligence community. In the days of ONE's predecessor (CIA's Office of Reports and Estimates) man-to-man contact between ORE analysts and their opposite numbers in the community had been irregular. A good deal of the coordination of estimates had been achieved via a challenge and response ballet conducted in writing. ORE would initiate an estimate and request contributions. Not receiving adequate help, ORE would draft the paper on whatever resources available and send it out for comment. When the comment came in, it was often given in written form. ORE would attempt to conform its text to well-founded exceptions and forget the rest. It would circulate the paper once more--this time for concurrence or dissent. Throughout, the bulk of the transaction was conducted by memo.

When General Smith asked Ludwell Montague to serve as the CIA officer to coordinate a number of NIEs, and in a great hurry, he insisted upon a man-to-man contact with his opposite numbers in the IAC agencies. Thus Montague was able to get a far higher degree of helpful compliance than heretofore. The six papers which he shepherded were thrashed out around a table with living representatives of the four principal intelligence services (State and the three military services).

By the time I had entered on duty in late November, the meeting of representatives (the Reps) to coordinate a text was a going institution. Throughout the history of the NIE, between 1950 and 1973, the Reps were one of the elements which made the whole enterprise a success.

A word about the Reps: IAC members, perceiving that the NIE was a deadly serious undertaking by General Smith, and cheerful at the way the account was being handled, gave ready support. Of their officers, they continued to designate one who would be their principal staff operative for the NIE account. We, as ORE before us, recognized these officers as the IAC Senior Representatives. They were the ONE's first point of contact within the IAC agencies for all business affecting the NIE.

Below each of these Senior Reps was a pool of intelligence officers most of whose duties included the area of the NIE. They were usually experienced men and women with a regional or functional specialty and an ability to discuss the substance and the rhetoric of draft estimates. They attended the meetings where text was coordinated and where agreement was achieved when possible. They were the people who when agreement was not possible, were the articulators of tentative dissent.

The institution of the Reps, which had had its informal beginnings in the ORE days, flourished with the coming of the ONE and its heavy schedule of NIEs. Its existence rested solidly upon the stuff of the customary law. I can so assert because there is no reference to "Representative" in DCID 3/2 (8 July 1946), devoted to the standard procedures of national intelligence production nor, of course, in DCID 3/2 of 13 September 1948 devoted to concurrences in national intelligence. In General Smith's rough outline of procedures, there are references to "discussion" between "ORE, or . . . the ONE when it is established . . . and the several intelligence agencies," but no word of "Representatives." However all NIEs produced from that point on involved the Reps in one way or another. It was not until the issuance of DCID 3/5 of 1 Sept. 1953 (which superseded DCID 3/1 cited above) that the word "Representatives" (and the institution) passed from the customary to the statute law. Paragraph 3 (c) reads:

Consideration by Representatives of the IAC Agencies--Representatives of the IAC Agencies will meet with the Board to review, comment on and revise the draft as necessary.(32)

Of the scores or even hundreds of Reps that we encountered, two things may be said: (1) They were indispensable to the production of NIEs, and (2) there was no other uniformity. Some were skilled intelligence professionals; others were unhappy time-servers; most fell between these poles. I will have more to say about them in a later section.



(1) The following general histories contain the essential background and a wealth of elaborating detail of the subject of this essay:

Arthur B. Darling, The Central Intelligence Agency: An Instrument of Government, to 1950 (12 vols.) 1953. (HS-1) [A declassified version of this history has now been published under the same name by Pennsylvania State University Press (1990).]

George S. Jackson and Martin P. Claussen, Organizational History of the Central Intelligence Agency, 1950-53 (10 vols.) 1957. (HS-2) [A declassified version of this history is held by the National Archives.]
Ludwell L. Montague, General Walter Bedell Smith as Director of Central Intelligence, October 1950-February 1953 (5 vols.) 1971. (DCI-1) [A declassified version of this history has now been published under the same name by Pennsylvania State University Press (1992).]

Wayne G. Jackson, Allen Welsh Dulles as Director of Central Intelligence, 26 February 1953Ð29 November 1961 (5 vols.) 1973. (DCI-2) [A declassified version of this history is held by the National Archives.]

George S. Jackson, Office of Reports and Estimates, 1946-51 (5 vols.) 1954. (MS-3) [Classified.] Hereafter I will cite the first four of these works as Darling, The CIA; Jackson and Claussen, History; Montague, Smith; Wayne Jackson, Dulles.

(2) There can be no question that the NIE, spelled with capital initials, was a Smith innovation. This is not to say that the CIA, and the CIG before it, had not produced finished intelligence utterances which contained estimates and which met most or all of the criteria of the word national as used in the context. The unit of the Agency which produced such papers was the Office of Reports and Estimates [ORE]. It was a large office which engaged in a number of intelligence research and analysis tasks. It published, inter alia, a current intelligence daily and current intelligence briefs, straightaway intelligence research studies on a wide range of subjects--world wide--situation reports, and an otherwise undesignated series known as "OREs." As a general rule, "OREs" were designed for consumption by policy makers at the national level and hence narrowly focused on problems of prime import to the national security. Further, they represented not only the best effort of the originating office, but also were coordinated within the community. They constituted the nearest thing to the pre-Smith national intelligence estimate.

They did, however, differ considerably with the successor institution (the NIE): 1) they contained much more narrative and descriptive data and probably less estimative material; 2) the coordinating process which attended their completion was quite different from and almost certainly less effective than the one which became possible under General Smith's leadership. That the DCI did not personally "submit" them to the NSC and that the IAC members did not personally, and in solemn conclave, approve them (with or without dissent) robbed them of a certain cachet enjoyed by the NIEs. Furthermore and perhaps more importantly, the absence of this high level review permitted a certain amount of captious (analyst's) dissent and an undue (analyst's) discursiveness.

(3) The full title of this report, usually cited as The Dulles Report is: Allen W. Dulles, William H. Jackson, and Mathias Correa, Report to the National Security Council on the Central Intelligence Agency and National Organization for Intelligence (1 January 1949).

Upon receipt of The Dulles Report, two principal officers of the NSC (the Secretaries of State and Defense) solicited comment from all parts of the intelligence community and in the light of the Report and comment wrote and submitted to the NSC A Report to the National Security Council by the Secretaries of State and Defense on the Central Intelligence Agency and National Organization of Intelligence, 1 July 1949. The President accepted this report and issued it as NSC #50. One of its principal recipients was General Smith who always referred to its group of recommendations as his marching orders from the President.

The importance of what I am calling the NIE in this essay received its due (though not in these exact words) in both The Dulles Report and NSC #50. The latter clearly ascribed to the DCI the personal responsibility for the issuance of national intelligence.

(4) The other two (in shorthand) were the coordination of the intelligence community and the undertaking of certain services of common concern.

(5) Readers of this essay will not miss the distinction between national intelligence on the one hand and departmental intelligence on the other. The early texts are signally emphatic in identifying departmental intelligence as something gathered, evaluated, and issued in support of departmental missions and functions and not to be trifled with by a supra departmental intelligence authority such as the DCI and his Agency.

(6) IAC-M-1, 20 October 1950. The memo in question had been composed a few weeks earlier by the DDCI, William H. Jackson, who had had Walter Lippman in mind as a chief recipient. At some time before 20 October, Mr. Jackson had shown a copy to Lawrence Houston, General Counsel of the CIA. Mr. Houston pointed out to Mr. Jackson that the memo erred in its attribution to the community of the "responsibility" for the NIEs. Mr. Houston emphasized the all-important point that this was a "responsibility" of the DCI alone. One document shows where exactly this correction was made in Mr. Jackson's typescript. Note: passages in single quotation marks are from the National Security Act of 1947.

(7) In Mr. Jackson's text, this word "judgment" had been "responsibility."

(8) General Smith (or Mr. Houston) added this "and responsibility" to the Jackson text.

(9) The effective date of the Act was 18 September 1947. Though the Act was signed into law on 26 July 1947, section 310 states that it would not be fully in effect until the day after the day upon which the Secretary of Defense, first appointed, takes office or the sixtieth day after the date of the enactment, whichever is the earlier. Mr. Forrestal was sworn in on 17 September 1947.

(10) Back in 1965 when I began putting down my thoughts on this subject, I sent a memo to the General Counsel asking him inter alia how was it the Congress had used the word "Agency" in this context rather than the "DCI" as had appeared in all prior texts. Mr. Houston answered me at length:

The most important thing about the Act itself is the congressional intent behind it, and no matter how ambiguous the wording of the Act, it is crystal clear that what the Congress wanted to do was place the responsibility at one single point for the coordination of intelligence and intelligence support to the policymakers. Also, it became clear that by one point the Congress meant one person. They were strongly influenced by the lessons brought out by the congressional investigation of Pearl Harbor, and while they were not too interested in organization or techniques, they had seen that the information by and large which would have warned of the Japanese attack was available and in the hands of various components of the executive branch and no one brought the pieces together and made an adequate evaluation to warn the President. They had received some testimony that such evaluation should be arrived at through board or committee action, but it is quite clear that they discounted any such dispersing of responsibility and were thinking of responsibility placed in one man. This led, among other things, to their designation of this man as Director of the [Central Intelligence] Agency, to connote his over-all responsibility. Thus, when you look at the Act you have behind it a pretty clear expression of the intent of the Congress, which has for the most part been consistent with the organizational concepts of the various Presidents.

I have studiously avoided getting into a legal hassle on the question you raise that in the Act the duties are given to the Agency, yet responsibilities in the NSCIDs are put on the Director. Since the Director is the head of the Agency and the Agency responds to his direction and control, I could see nothing inconsistent with the Act giving the responsibility to the Agency, particularly when you knew the legislative history.

(11) These unfortunate words, "correlation and evaluation," themselves have an interesting history. The word "synthesize" would have done the trick and indeed was used in an early draft which Admiral Sidney Souers (the principal draftsman of the President's letter) had submitted to Mr. Truman. Souers had relied heavily upon the thought and language of a document relating to a future central intelligence service which the Joint Chiefs of Staff (JCS 1181/5, 19 September 1945) had forwarded to the President. From their text, Souers had borrowed the phrase that the director of the service "shall accomplish the synthesis of departmental intelligence relating to the national security. . . ." Mr. Truman didn't like "synthesis" or "synthesize." Souers told Ludwell Montague that he thought Mr. Truman did not know the intended meaning of the word. Souers guessed that he thought it sounded derogatory (cf. synthetic). (Memo to SK from Ludwell Montague, 26 November 1965.)

(12) "The Director of Central Intelligence shall be advised by an Intelligence Advisory Board consisting of the heads (or their representatives) of the principal military and civilian intelligence agencies of the Government having functions related to national security, as determined by the National Intelligence Authority."

(13) National Intelligence Authority, Directive #1 (8 February 1946), paragraph 6.

The Central Intelligence Group will utilize all available intelligence in producing strategic and national policy intelligence. All intelligence reports prepared by the Central Intelligence Group will note any substantial dissent by a participating intelligence agency. (emphasis added)

(14) Some hero finally bit the bullet and substituted the word "produce" for "correlate and evaluate." By this time the CIA was very much of a going concern with a significant capability to collect a good deal of raw information through its own efforts. Hence it did not need to confine itself to simply synthesizing what it learned from other intelligence organizations of the community.

(15) The Act failed to mention an Intelligence Advisory Board or Committee, although it had had an important place in the President's letter and in the history of national intelligence from January 1946 on. The first paragraph of NSCID #1 rectifies matters with a note on the composition and advisory functions of the (now) IAC:

1. To maintain the relationship essential to coordination between the Central Intelligence Agency and the intelligence organizations, an Intelligence Advisory Committee consisting of the respective intelligence chiefs from the Departments of State, Army, Navy, and Air Force, and from the Joint Staff (JCS), and the Atomic Energy Commission, or their representatives, shall be established to advise the Director of Central Intelligence. The Director of Central Intelligence will invite the chief, or his representative, of any other intelligence agency having functions related to the national security to sit with the Intelligence Advisory Committee whenever matters within the purview of his agency are to be discussed.

A revised edition of this NSCID (7 July 1949) directs that the DCI shall be the IAC chairman and that the Director of the FBI will be on the Committee. (He was always represented by one of his officers, a matter officially recognized some nine years later--NSCID # 1 of 25 April 1958).

(16) See esp. paragraph 1 (e) National Intelligence.

(17) DCID 3/1, 8 July 1948.

(18) DCID 3/2, 13 September 1948.

(19) IAC-M-1, 20 October 1950, para. 7. In the context of the chairman's remarks, Mr. Jackson indicated that the fact that the [former] Office of Reports and Estimates has in the past produced both national estimates and miscellaneous reports in various fields, which could not possibly be construed as national estimates, had blurred and confused both the product and function of the Office of Reports and Estimates. There has been insufficient differentiation between the form and the coordination procedure in connection with the two products and in their methods of production.

(20) A copy of the Howze memo is on file in HS/HC 266.

(21) Quoted from para. 8 of IAC-M-1 above, note 9. In actual fact the cover of NIE 1 (3 November 1950) did not plainly show that it was the result of a collective effort. The lay-out of the cover was National Intelligence Estimate/The title/The CIA Seal/NIE-1/Published 3 November 1950/Central Intelligence Agency. The first page immediately after the cover contained the dissemination and distribution notices. The next page was the proper title page: NIE-1/National Intelligence Estimate/The title/followed by "The intelligence organizations of the Departments of State, Army, Navy, and the Air Force participated in the preparation of this estimate and concur in it."

Perhaps a year passed before this latter bit of text appeared on the front cover.

(22) Quoted from para. 5 of IAC-M-1.

(23) As reported in IAC-M-1:

9. After discussion the following procedural steps were agreed upon in the production of national estimates:

    a. The Intelligence Advisory Committee will adopt an intelligence plan, or more specifically, a list of required national estimates in an order of priority.

    b. In the case of a particular estimate, a frame of reference and the assumptions on which the estimate is based will be discussed and approved by the Intelligence

    Advisory Committee.

    c. Work on the estimate will be referred in the first instance to the Office of Reports and Estimates, or to the Office of National Estimates when it is established in the Central Intelligence Agency, and the several intelligence agencies will be consulted and a timetable fixed for contributions to the national estimate within the fields of their respective interests.

    d. On the basis of these contributions, the Central Intelligence Agency will produce a first draft of the proposed national estimate.

    e. This draft will be sent back to the agencies for comment and modification and for further discussion if required. On the basis of such comments and discussion, the Central Intelligence Agency will produce a second draft of the estimate.

    f. This second, or later drafts if required, will be submitted to the Intelligence Advisory Committee for final discussion, resolution of differences and approval.

    g. If differences cannot be resolved and approval obtained, the estimate will be published with notation of substantial dissent and reasons therefore. It was made clear by General Smith that this procedure would not and could not be followed in the case of so-called "crisis estimates." In the event of need arising for a quick or crisis estimate, a procedure similar to that used in the recent instance when the President called for a series of estimates prior to his departure for the meeting with General MacArthur would be followed. That is, a special meeting of the Intelligence Advisory Committee will be called and representatives of the various intelligence agencies assigned at once to the production of a draft of the required estimate for immediate submission to the Intelligence Advisory Committee for discussion, revision and approval.

      (24) IAC-M-1, para. 10.

      (25) In the Eisenhower years, the staff work for the NSC was conducted along military lines and with military precision. Mr. Cutler, who was the President's man in charge of NSC business, took the chairmanship of what was called the NSC Senior Staff. One of his activities was a continuing tour of the horizon of US foreign relations and security policy and the identification of situations that called for policy adjustment. Another was seeing to the preparation of coordinated policy papers (with recommendations) relating to all of the likely trouble spots. Mr. Cutler planned his papers for months in advance and relied upon the intelligence community to produce an NIE on each upcoming subject. Deadlines for the NIE were set so that it would be ready when the Senior Staff began its policy deliberations. The Staff's finished paper often quoted liberally from the NIE. During Mr. Cutler's time and that of Mr. Dillon Anderson who succeeded him, upward of perhaps 80 percent of NIEs were produced for this particular account.

      This is not to say, however, that Mr. Cutler and the NSC, or the NIEs for that matter, had an important role in all major foreign policy decisions of the Eisenhower administration. There were those situations of particular concern to Secretary of State Dulles. These he watched over personally and made his recommendations to the President without reference to Mr. Cutler's complicated staff machinery and its equally complicated intelligence support.

      The Kennedy administration changed matters very considerably. Nevertheless with McGeorge Bundy as the President's Special Assistant for National Security Affairs, many NIEs were produced at his request for the consideration of the President and members of the Council, and as well for Mr. Bundy's own NSC staff.
      The sort of relationship between Mr. Bundy and the ONE continued with Mr. Rostow who served in the Johnson years. Mr. Kissinger, President Nixon's man in the same job, seemed to have had considerably less interest in the NIE.

      (26) See pp. [62-64] about the scheduling of the estimates.

      (27) See Montague, Smith History [p. 69 ff. in the Penn State Press edition].

      (28) For a discussion of the formative period of the ONE, see Jackson and Claussen, History, IX, 32-51. I succeeded Mr. Langer on 3 January 1952 as the AD/NE and held the position until 31 December 1967. Abbot E. Smith was my successor (1 January 1968Ð17 April 1971). John Huizenga followed him (17 AprilÐJune 1973). For the last months of ONE's existence (June-November 1973) Ramsey Forbush was the acting chairman of the Board of National Estimates.

      (29) See Appendix B for the official description of the organization and mission and functions of the ONE.

      (30) See Appendix C for two charts relating to the Board of National Estimates. The first shows the changing membership of the Board between 1950 and 1963 with a graphic indication of each member's professional background. [Editor's Note: The second chart, Board of National Estimates Members' History, is missing from the original copy of this study.]

      There were a number of members of the Board who do not show up on either of these documents. Among them were Admiral Jerauld Wright whose last active service in the Navy had been as CINCLANT, Livingston Merchant, who had held many important positions in the Department of State including Undersecretary for Political Affairs and Ambassador to Canada, and Llewellyn Thompson, one of the nation's leading Sovietologists and twice our Ambassador to the USSR.

      (31) Almost from the beginning, the organization of the staff followed regional lines: Western Europe, Middle East, East Europe (which included the USSR), and Far East. As the demand grew for NIEs concerning Latin America and Africa small staffs were formed to handle these accounts. Later still, when the number of NIEs devoted to Soviet military and technical matters (e.g., atomic energy, space exploration) grew, we formed a special Soviet Military/Technical Staff.

      (32) This identical language is repeated in para. 3c of DCID 1/1 of 21 April 1958, which superseded DCID 3/5, and in para. 3c of DCID 1/1 (5 August 1959) which superseded the version of 21 April 1958.


      Historical Document
      Posted: Mar 19, 2007 11:00 AM
      Last Updated: Jul 07, 2008 02:10 PM