CIA HISTORICAL REVIEW PROGRAM
RELEASE IN FULL 22 SEPT 93
No Foreign Dissem
Uphill struggle to achieve coordination of departmental intelligence activities, June 1946 through March 1947.
WITH VANDENBERG AS DCI
Arthur B. Darling
Part II: Coordination in Practice1
ORE 1, published on July 23, 1946, as the first product of the new Office of Research and Evaluation, was a masterly demonstration of what could be done by a single person in correlating, evaluating, and producing strategic intelligence. It had involved coordination too, of a sort, but not the kind that its author, Central Reports Staff chief L. L. Montague, wished to have. From his wartime experience on the Joint Intelligence Staff, he had proposed that full-time assistants in the new Central Reports Staff should both represent their respective departments and at the same time work with him to synthesize departmental intelligence and produce national estimates. As Staff Chief he would decide, subject to DCI ratification, what the CIG estimate would be; the departmental representatives would record, subject to their chiefs' approval, any substantial dissents from that estimate.2
Coordination of Estimates
An effort was made in October to effect this arrangement which had been set up on paper for the Central Reports Staff. Admiral Inglis, Director of Naval Intelligence, objected to the fact that ORE 1 bore no indication that the intelligence agencies of the departments had concurred prior to its dissemination. He argued that the Intelligence Advisory Board, that is the departmental intelligence chiefs, should approve the CIG estimates through a voting system such as that used by the joint intelligence Committee of the Joint Chiefs of Staff. He was not willing to leave these estimates, as opposed to the "factual" Daily and Weekly Summaries, to the judgment of the Navy men in ORE; they should be reported severally to the DNI, and he should have at least two or three days to consider each paper. In case of delay, the estimate could go forward with a statement that a dissent or concurrence would follow. Inglis was willing to have a part-time. Navy representative assigned to the estimating staff, but that officer, he said, should be only a "messenger" to ONI.
At its best, such a system meant that the proposed estimate would receive painstaking and diligent review by the chief intelligence officer in each department. At its worst, it would be obstructive and time-consuming. In any case, it did not provide what Montague gee and others with estimating experience sought—representation and responsibility for the departments at the working level. With the Inglis system there was likely to be no real fusion of departmental intelligence into a national estimate, whereas if the departmental representatives worked day after day with the evidence, giving their full time to the business, they would come to make more effective syntheses of the materials. They did not have to lose their sense of responsibility to their departments because they became expert in the common concern of all.
Montague's plan was taken up by General Vandenberg before the JAB on October 31. It was debated at length, or rather the debate ranged for some time over many phases of the relation between the DCI and the IAB. At first glance, it seemed to have been adopted and put into operation by administrative order on November 1. Closer examination reveals that it was not. It was agreed that each member of the JAB should designate a personal representative as liaison in the estimating division of ORE, who would concur or present dissenting opinions as directed by his chief. But there was no stipulation that he spend his full time participating in the development of estimates.
A review of the record on April 15, 1947, at the end of Vandenberg's administration, showed that the departments had in fact been more than wary in their cooperation. The JAB members appointed their personal representatives as agreed, but none of them gave his full time to the work of estimating. Only one even had an office in CIG. They were, as Admiral Inglis wished no more than messengers to their chiefs.
The average lapse of time between submission of estimates and receipt of concurrence or dissent from the departments was seventeen days. A later survey, covering the twenty reports and estimates which had been fully coordinated by August 1947, when CIG became an Agency, showed differences in promptness among the several departments. The median, average, and extreme delays for the Air Force were seven, eight, and fourteen days respectively, for the Navy eight, nine, and seventeen, for Army eight, eleven, and twenty-seven, and for State eleven, fourteen, and fifty-five.
The story of the central estimating function has been carried into the summer of 1947 to show that ORE had not effectively produced coordinated national estimates up to that time. The failure was due in largest part to the fact that the intelligence officers of the departments were not ready to make the work of the central agency swift and definitive. But there were other handicaps as well.
[Top of page]
For months ORE could not undertake the research and evaluation it was intended to do. For example, Assistant Director for Special Operations Galloway inquired on August l, 1946, about getting evaluations of OSO reports, and Montague had to reply that ORE had neither the personnel nor the working files for appraising them. The Reports Staff was at half strength and equipped only for current intelligence and attempting to synthesize departmental estimates. CIG would have to ask G-2 to continue grading these secret reports. Even at the end of the year Assistant Director Huddle reported that ORE (now called the Office of Reports and Estimates, rather than Research and Evaluation, at the request of the Department of State) was still operating at only 20 percent of its proposed strength. It was not until June 1947 that OSO began to get a file of evaluations from ORE, which even then had to strain its facilities to produce them.
Another problem was friction that developed between the Intelligence (formerly Reports) Staff—Montague and his deputy, Deforest Van Slyck—and the regional branches of ORE. Beyond personal antagonisms it was the ancient dispute between area experts and those who temper their expert findings with over-all judgments, coupled with the chronic irritation of writers at editorial meddling. The Intelligence Staff maintained that it was to have the final review of the finished ORE product, subject of course to the approval of the Assistant Director and the DCI. But as the number of the regional branches increased, some of them came to insist that the only sound practice was for intelligence to pass directly from the experts in the branches to those who used it, with any necessary review performed by the branch chiefs, who were the authorities in daily contact with the problems of their areas.
In the end the experts won a temporary victory. At the beginning of May 1947, as Admiral Hillenkoetter became Director of Central Intelligence, the Assistant Director for ORE assumed the duties of the Chief of the Intelligence Staff. Montague and Van Slyck were placed off to the side in a new Global Survey Group.
[Top of page]
Admiral Souers' Central Planning Staff, whose chief job had been to prepare studies and recommendations for the departmental chiefs of intelligence to consider, individually or in the IAB, took upon itself, we saw, the additional function of planning for the other CIG staffs and offices.3 As a result of this activity, which proved irksome to others, Vandenberg came to the directorship inclined to let each office make its own plans and policies for his approval. On July 20 he broke up the Central Planning Staff and distributed its members among the various offices. But his struggle with the IAB over the fifth NIA directive4 convinced him that be needed a staff of departmental representatives to prepare the way in future dealings with the IAB.
He established the Interdepartmental Coordinating and Planning Staff for that purpose. Its membership represented the departments, its chief the Department of State. They were to work with him on behalf of the IAB as he formulated his opinions and reached his decisions; the chief intelligence officers of the departments would then know in advance what entered into his thinking. In short, ICAPS was to have been a working staff within CIG for the IAB. Like the Central Planning Staff preceding it, however, it gained more of a reputation for action inside CIG than for coordinating the activities of the departments.
The Central Planning Staff bad conceived of a whole series of interdepartmental coordinating committees which should handle matters of foreign scientific, military, political, economic, and geographic intelligence. This scheme was now abandoned with the explanation that such committees and boards were not necessary; all CIG personnel were authorized and encouraged to establish relations and consult with persons of similar positions in other intelligence agencies. (An exception, as we have seen, was the Office of Special Operations.) One interdepartmental coordinating committee was sufficient. ICAPS instructed itself to act for the DCI in coordinating the intelligence activities of the departments, assuring that the facilities of each department were ample, that each was covering its proper fields of intelligence, and that its methods, procedures, and controls were adequate for the collection, integrated research and evaluation, and dissemination of strategic and national policy intelligence.
The most optimistic advocate of central intelligence could not have imagined in August 1946 that the intelligence services of the departments would tolerate such supervision and control. It would have meant inspection—a right which neither Souers nor Vandenberg dreamed of exercising at this stage—of the most vigorous and persistent nature. But ICAPS could perform much more effective inspection and direction within CIG, where it proposed to maintain continuous supervision over planning and coordination.
Its Chief at once sent a memorandum to the heads of the Offices of Collection, Research and Evaluation, and Dissemination requesting information on the Peace Conference that opened in Paris on July 28. What steps had been taken by the State, War, and Navy Departments to provide reports and disseminate the information when it arrived in Washington? The offices should find out from the departments and report. It is revealing that a staff established to work with the IAB members should have taken this indirect approach to getting information from them.
On August 5 ICAPS asked that ORE undertake a general study of the periodic and special reports of the intelligence agencies of the government. Montague replied on August 9 that the task fell within the jurisdiction of the Office of Dissemination, which had already issued one directive on the subject.
In January 1947 ICAPS proposed a program of production for ORE. It should have a staff giving its whole time to current intelligence; it should issue "situation reports" on the world's several geographic and strategic areas monthly; it should draw from the personnel of its branches a group to prepare for "National Intelligence Digests" the eventual NIS program.
ORE favored the idea of area "situation reports," which had been requested earlier by the Navy, but thought they should not be reissued rigidly month by month but governed by events, With respect to forming a group for the production of National Intelligence Digests, ORE simply did not have the staff, nor any immediate chance of obtaining it. It had fifteen key persons in its six regional branches and only two more in prospect where a total of seventy-seven had been authorized. The Chief of CAPS, however, insisted that ORE undertake the enlarged program of production against preposterously close deadlines. He ordered one of his men to make a tour of inspection in ORE by July 10 to see hove it was progressing.
In dictating a program unrealistic in terms of possible achievement, the planning staff was also distorting the central purpose of ORE, the production of "definitive estimates of the capabilities and intentions of foreign countries" as they affected the security of the nation. General Vandenberg himself had made some contribution to this distortion when he insisted upon taking over the function of research and evaluation and greatly enlarging CIG as an independent producing agency. This meant duplication and intensified efforts on the part of established services to defend their prerogatives. It impeded establishment of the close interdepartmental cooperation indispensable to the production of strategic intelligence.
ICAPS members submitted weekly reports of their activities, conferences, and accomplishments. The candor of one reporter deserves whatever immortality this study can give him. He wrote on October 14, 1946, that he had made "no progress worthy of reporting." A week later he had finished the organization charts, but there was "no other progress of note," he said, "except in frustration."
At the end of 1946 the Chief compiled an annual report listing projects completed and pending and surveying the difficulties CAPS had encountered in endeavoring to plan and coordinate with the departments for the DCI. They were the difficulties which General Donovan, General Magruder, Colonel McCormack, Admiral Souers, and others had experienced whenever they sought to bring the intelligence officers of the departments together. These men seemed not to understand each others' problems. They did not like to turn "operational" information over to civilians. They shied away from the centralization of common functions. They deplored others' duplication of their efforts but were unwilling to give up their own activities. They came to interdepartmental meetings poorly prepared to discuss matters which bad been for some time on the agenda. And there were frequent changes in announced policy which threw all negotiation back to the beginning.
The Intelligence Advisory Board did not accept ICAPS in the role of working staff that Vandenberg intended. Instead, it sent ad hoc committees to confer with ICAPS, and these specially appointed representatives carried back to the IAB members word about what was being pondered for the DCI. The procedure did not make for speed or decisiveness in coordinating the intelligence activities of the departments. It certainly did not expedite the central formulation of policies and procedures in matters of common concern.
[Top of page]
The Director of Naval Intelligence offered on September 6, 1946, a plan for merging the "static intelligence function" of the State, War, and Navy Departments in CIG. By "static intelligence" he meant stable political, economic, sociological, topographic, and technical information such as composed the joint Army-Navy Studies—papers prepared for the joint War Planners of the two departments—and the Defense Project5 then under way in the Pentagon. This subject of common concern had been discussed before. The chief obstacle in CIG was the lack of personnel and equipment. It made sense to have. the work done centrally if for no other reason than to eliminate the waste of public funds by duplication.
The matter came before the IAB on October 1. The representative of the State Department, Mr. Eddy, caused some distraction by declaring that his department was not willing to turn over to CIG its responsibility for obtaining economic and political intelligence. The representative of the War Department, General Chamberlin, caused more by wondering if it were not possible to "parcel out functions on the basis of primary responsibility." Each department would furnish finished intelligence, he said, to the others and they could then "rework it to meet their particular needs." In other words, there would be no central operation.
Admiral Inglis answered the objection of the State Department by stating that he was talking about processing, not collection. But there was further rambling discussion as to whether the departments wished to have CIG as a "middleman" and whether one department used the same kind of political and economic intelligence as another. Admiral Inglis reduced it to simple terms. Each department should retain its own operational intelligence but should supply the rave materials of static intelligence to CIG, which would do the processing very much as a publishing house and turn the product over to the departments. These could put it in a different final form, if so desired by their customers, and assume the responsibility for its dissemination.
The IAB came to no conclusion at this meeting; a special committee was to make further study of the problem. This committee met on October S. It too ranged the fields of intelligence near and far from the question at issue. The Office of Naval Intelligence offered to transfer to CIG its personnel engaged in strategic intelligence if other agencies did so. But the Military Intelligence Division of the War Department declined on the ground that it would lose direct control over intelligence functions related to its problems of staff and command. The representative of the State Department emphasized again its "inescapable responsibilities," particularly for political and economic intelligence.
The special committee suggested recourse to bilateral agreements in particular fields between CIG and the departmental agencies, and asked to be relieved. CIG on its own responsibility made further study of the problem of basic intelligence over the following winter and spring. A member of ORE became chairman of the Working Committee of the Defense Project in March, 1947. Eventually the Joint Army and Navy Studies were taken over by CIG on October 1, 1947, as it began its program of National Intelligence Surveys.
[Top of page]
Coordination of Collection
Another problem of importance before the IAB in the fall of 1948 was the collection of intelligence, though it became entangled with production and dissemination as the debate progressed. The State Department offered on October 18 to eliminate duplication by assuming complete responsibility for the overt collection of intelligence in foreign areas on all political, economic, social, and cultural matters. There would be some overlapping and mutual assistance, but State would be responsible. It would share with the armed services the collection of scientific intelligence.
The Military Intelligence Division of the War Department matched this proposal with a plan to coordinate the intelligence activities of the three departments. This was not so much an answer to the State offer, however, as notice to General Vandenberg that the departmental secretaries and not the DCI would manage the intelligence activities of the departments. The plan would have the departments not only retain determination of their primary interests but also do their own research, evaluation, and analysis in those areas. Collection in the field would follow the same lines of primary interest. Matters of principle would be coordinated between the departments in Washington. Coordination of collection in the field would be the function of the Chief of Mission.
Regardless of source, reports would be sent directly to the agency primarily concerned. Estimates too would be the responsibility of each agency; they would provide information on their respective subjects to other agencies as the needs required. Communication would be direct from one agency to another, not through CIG. Each would collect and maintain files of biographic information within its primary responsibility, though a central file should be maintained for common reference. Each agency should contribute chapters of a Strategic Intelligence Digest like the Defense Project.
The central agency, thus carefully segregated from the departmental activities of "primary responsibility," was nevertheless to maintain supervision over interdepartmental cooperation and production. No explicit indication was given as to what authority should determine which department had primary responsibility in case two or more claimed it; presumably the departmental secretaries in the National Intelligence Authority would do that, with the view of the President's representative decisive.
This War Department plan came before the Intelligence Advisory Board on November 26, together with the State proposal, the draft of an NIA directive combining these two prepared by ICAPS, and specific reservations by General Vandenberg. The directive, if adopted by the NIA, was to be accompanied by a DCI directive providing definitions and detailed arrangements for coordinating collection. Vandenberg's reservations concerned the secret activities of the Office of Special Operations, overt collection by FBIS, and the information obtained by the Contact Branch from American businesses and travelers. What had been drawn into the central organization Vandenberg was not disposed to put back into the field of departmental activity and control.
The ICAPS draft added to the State and War proposals provisions for coordinators in foreign areas, along with the DCI in Washington. The stress was on his position as the chief coordinator of the whole system of collection and that of CIG as the central intelligence organization. The duties of the field coordinator were precisely indicated according to the situation in the area. He might be the chief of the diplomatic mission or the senior military commander; if neither had sole responsibility the DCI, with IAB approval, would designate someone to supervise them. (We shall find this provision, with the DCI in central position, missing from the final directive.)
By agreement among the departments the fields of primary responsibility would be assigned in careful detail: to the State Department political, cultural, sociological, economic, and international matters; to the War and Navy Departments their respective military and naval concerns; scientific intelligence to each agency according to its particular interests. A specific list was constructed to fix the assignments and reduce misunderstandings and conflicts of interest to a minimum.
Admiral Inglis commended ICAPS for its fine paper but wished to have another week to study it. General Vandenberg asked if it might not be approved in principle as a guide, with controversial parts to be considered later; he felt that there should be no further delay. General Chamberlin too praised the paper, but he still wished to assign fields of primary responsibility for activities other than collection; and because ambassadors as political officers went abroad without training in intelligence, certain principles should be defined for their guidance. And so the IAB agreed to have another ad hoc committee study the matter with ICAPS and bring back another report. A companion piece should also be prepared on the coordination of production and dissemination of intelligence. It was clear by now that ICAPS was not a representative working staff of IAB.
The ad hoc committee met with ICAPS OD December 3. They defined "area" to mean a country which had a station of the Foreign Service. They recommended that the word "coordinator" be dropped and "senior U.S. representative" be used. Reference to theater commanders should be omitted because coordination in occupied areas was a temporary thing. They listened to an Army division of intelligence into categories in which "factual" took the place of basic and "staff" replaced departmental. Strategic and national policy intelligence did not enter the discussion. One cannot escape the conclusion that the representatives of the Departments were not present to advance the case for the Director of Central Intelligence and the Central Intelligence Group.
The ICAPS draft as modified by the ad hoc committee went without the supplementary DCI directive to the IAB on December 17 in what proved to be its last meeting with General Vandenberg as DCI. The DCI directive, he explained, which did not require IAB approval, had been gone over by ICAPS and their ad hoc committee and the latter's recommendations accepted. There should be no further need for extensive discussion. The requirement was urgent. The NIA directive regarding collection should be completed and put into effect.
But there was further discussion. Should collectors in the field make their reports on duplicating mats or not? Should political intelligence be listed separately from economic intelligence, after having been bracketed with it in the State Department for years? Mr. Eddy asked for State that the section in the DCI directive concerning the allocation of primary responsibilities be incorporated in the NIA directive. General Vandenberg consented. With it was included a provision that collectors in the field might send copies to their own agencies when they transmitted materials directly to the field representatives of the agencies primarily concerned.
The IAB adjourned without taking a vote, but the final draft of the directive went to the members individually. By the end of the month all had approved without further change. It was issued on January 2, 1947, as Directive No. 7 of the National Intelligence Authority.
There was no mention in the directive of the Central Intelligence Group nor of the Director of Central Intelligence. The several members of the IAB bad obtained control over the collection of intelligence for their departments, except for the secret intelligence of OSO and the collections of the Foreign Broadcast Intelligence and Contact branches in the Office of Operations. Vandenberg let the directive pass. He might have withheld his approval formally, but there was no point in arguing further with the chiefs of intelligence. He was about to take his case up to the secretaries in the NIA.
[Top of page]
Now that the general matter of collection had been settled for the time being, the specific question of collection in China perhaps could be answered. This had been proposed as an area in which to try out the coordination of collection. As a member of the IAB under Admiral Souers, General Vandenberg had urged that the Strategic Services Unit which was keeping alive the secret operations of OSS be kept at work in China until it could be replaced, as it now was by OSO. Admiral Goggins had gone to the Far East during the summer to make arrangements with General MacArthur and Admiral Cooke. Vandenberg was in correspondence with General Willoughby in regard to CIG stations in Tokyo and Seoul.
ORE had prepared on October l, at the request of ICAPS, a draft set of intelligence requirements for China. From this beginning ICAPS had developed, in conference with another ad hoe committee for the IAB, the draft of an NIA directive, "National Intelligence Requirements on China." It did not include requirements for basic intelligence, and it omitted those for scientific information on the assumption that there would be little there. It focused upon current intelligence. Little exception was taken to the draft in the IAB meeting on November 7, except that the phrase "essential elements" was changed to "current essentials" and "requirements" to "objectives." But General Chamberlin blocked its acceptance and issue as a directive until he could study a specific directive for collection which should be based upon these requirements.
It was not until January 8, 1947, after the general NIA directive on collection had been issued, that General Chamberlin came again to the question of intelligence on China and gave his views to General Vandenberg. The general directive on collection, he said, and a plan for coordinating the production of intelligence which seemed about to be approved by the IAB appeared to make "the China experiment" unnecessary. He thought it should be withdrawn or at least postponed "to insure that it be brought into full consonance with the broader directives."
Vandenberg apparently found this a last straw. He replied on January 23 that the Chamberlin proposal indicated a "misunderstanding of the national intelligence program" as it was being developed by CIG with the "advice and the assistance" of the IAB. The DCI was endeavoring to develop the program laid down in the President's Directive of January 22 the year before—the foundations of a central intelligence system which should furnish the President and the departments with strategic and national policy intelligence. The program was designed to facilitate departmental intelligence as well as national; but it had been assigned by the President to the DCI alone. He had therefore to determine its requirements and procedures for collection, research, and dissemination.
General Vandenberg became peremptory. He requested that General Chamberlin concur in the immediate release of NIR China and that he furnish the G-2 personnel he had already named to assist CIG in preparing similar papers on the Soviet Union and the Near East. Moreover, said Vandenberg, he was withdrawing the collection directive for NIR China from IAB consideration and referring the matter to his Assistant Director for Collection and Dissemination, who would issue such requests as might be necessary. So on February 12 NIR China appeared as the eighth NIA directive, with the concurrence of the IAB and without further official comment of any kind.
[Top of page]
Agent for the Secretaries
This was the day of the historic ninth session of the NIA. It met to discuss the correlation of intelligence in the field of atomic energy (a problem whose history will be traced later) and to hear a report from the DCI. Secretary Patterson made a brief statement on the transfer of intelligence personnel and files from the Atomic Energy Commission to the CIG, and then the NIA members heard General Vandenberg pass quickly over CIG accomplishments since he last addressed them—on October 16, 1946, concerning the budget for 1948—in order to concentrate on his present difficulties.
They grew from uncertainty with regard to the directive authority of the DCI. He found this adequately stated in the President's Directive of January 22, 1946, and the fifth NIA directive on July 8 of that year: he was to "act for" the NIA in coordinating foreign intelligence activities. The interpretation of the agencies, however, was coordination "by mutual agreement"; and in some instances this had taken from six to eight months. He requested authority to act as agent for the secretaries of the departments. The alternative was that CIG should forward its directives to the NIA members for issuance from their own offices. This would be cumbersome and it would involve great loss of time for all concerned.
The production of strategic and national policy intelligence by CIG, its primary purpose as the central intelligence organization of the government, was further hindered by uncertainty among the agencies over its definition. Vandenberg asked the NIA to approve the definition established in CIG thinking ever since it had picked up the torch from Donovan and Magruder. It was intelligence collected from every available source, both covert and overt, and then verified, appraised, and synthesized in estimates for the benefit of the policy-makers of the government.
After listening to General Vandenberg's statements, Secretary Patterson saw no alternative to approving his request, provided that any aggrieved agency might appeal to the NIA itself through the secretary of the department concerned. Vandenberg acknowledged such a right as inherent. Admiral Leahy agreed with Patterson. Secretary Forrestal gave his consent. Mr. Eddy of the IAB, who was present with Secretary Marshall for the State Department, assumed that normally any directive would have prior discussion by the IAB. Vandenberg assented.
The NIA now approved the statement that the Director of Central Intelligence should "operate within his jurisdiction as an agent of the Secretaries of State, War, and the Navy," and delegated the necessary authority to him so that "his decisions, orders and directives" should have full force and effect as though emanating from the secretaries. Any aggrieved departmental agency might have access to its own secretary and through him to the NIA. And then to make General Vandenberg's satisfaction complete, the NIA authorized the definition: "Strategic and national policy intelligence is that composite intelligence, interdepartmental in character, which is required by the President and other high officers and staffs to assist them in determining policies with respect to national planning and security in peace and in war and for the advancement of broad national policy. It is in that political-economic-military area of concern to more than one agency, must be objective, and must transcend the exclusive competence of any one department."
It would seem as though these decisions should have been enough. They were not; the departmental intelligence chiefs did not give up so easily. Admiral Hillenkoetter was to inherit a still bitter controversy.
[Top of page]
The Joint Chiefs
Vandenberg also brought the relationship between CIG and the joint Chiefs of Staff up before this NIA meeting. The matter had had a considerable recent history. In August 1946 Kingman Douglass and William H. Jackson had made a report showing that the British had brought about an effective articulation of their intelligence system with their Chiefs of Staff, in spite of a permeating influence of the Foreign Office that gave this country's military and naval authorities much to consider. There had been representation from the State Department, the Foreign Economic Administration, and the office of Strategic Services on the joint Intelligence Committee of the JCS during the war, but the presence of civilians in military councils was not generally acceptable to the Army and the Navy.
On August 12 Admiral Inglis had proposed to General Vandenberg that a channel be established between the CIG and the JCS to avoid useless duplication when they were working on subjects of common interest. Specifically, the Joint Intelligence Staff of the JCS's Joint Intelligence Committee might serve as the staff also of the Intelligence Advisory Board; the JIC and JAB were composed of the same persons and could logically have the same staff. But General Vandenberg had other ideas. He replied on September 4 that with a full-time staff the IAB would have to act unanimously, whereas at present its recommendations could be submitted to the NIA even though a member did not concur. He urged Inglis to join in sponsoring a joint study of the problem by ICAPS and the Joint Intelligence Staff. Then, before Inglis could make another move, ICAPS, working with Secretary Lav of the NIA, had prepared a counterplan to his.
The ICAPS plan would establish the DCI as the chief adviser on intelligence to the JCS, thus ranking him above the joint Intelligence Committee. He would meet with the Chiefs of Staff in the same way as be sat, without a vote, in the meetings of the National Intelligence Authority. He would submit appropriate matters to the joint Intelligence Committee, of which he should be chairman, as he was presiding officer of the IAB. The subcommittees of the joint Intelligence Committee and its joint Intelligence Staff would be integrated into CIG to create a compact and efficient intelligence organization serving both the joint Chiefs of Staff and the secretaries of the departments constituting the National Intelligence Authority.
This plan had features which the military and naval authorities were reluctant to accept. Colonel Carter W. Clarke, deputy to General Chamberlin, summarized for him its weakness as he saw it. The DCI would be chief intelligence adviser to the joint Chiefs of Staff, but they would have no authority over him; he was responsible to the NIA. An external agency which the Chiefs did not control would thus come between them and their subordinates. This violated the principle of command.
In further conversation Colonel Clarke remarked for himself that the joint Intelligence Committee should be abolished; the one good reason for keeping it was its relationship with the British committee. Clarke believed that the DC1 should be a member of the joint Chiefs of Staff and he felt sure that General Eisenhower would agree. He was doubtful whether Navy would. If the DCI were included with the JCS, the Central Intelligence Group would be in an echelon above the joint Intelligence Committee, and there would be little use for this Committee whose membership sat elsewhere as the IAB.
At this juncture, on September 25, 1946, Vandenberg had first taken the matter before the NIA. He set the scene by declaring that CIG could not produce national intelligence unless it had all of the information available in the government. He was getting complete coverage, he thought, from the State Department and the Navy; but he had not obtained access to the President's messages, General Marshall's, or those of the War Department marked "OPD Eyes Only." He would like to have a C1G officer in the message center of each department to review and transmit, under any necessary restrictions, the items of intelligence value. Secretary Patterson thought such an arrangement could be made. Vandenberg could have added that he was getting from his Office of Special Operations secret intelligence which he should have received from G-2.
Against this sharp background Vandenberg pointed out to the NIA that the JCS Joint Intelligence Committee was duplicating work of the CIG, and the coordinating activities of the two often conflicted with each other. The studies of the JIC got priority in the intelligence agencies of the War and Navy Departments because the Chiefs of Staff were the immediate superiors of the heads of those agencies. It had been suggested, he said, that CIG should be combined with the joint Intelligence Staff. Secretary Patterson responded that he saw no reason why the Staff should not be disbanded. Admiral Leahy agreed with Patterson and remarked that he had so stated to the joint Chiefs.
But when Vandenberg offered his plan to the IAB on October 1, there was random demur and objection in small detail. The real cause for hesitance on the part of the military men became obvious with the question from General Chamberlin whether it would be acceptable to the Planners for the JCS, whether they would be willing to let an outside agency know their secrets. He knew well, in fact, that the Planners would not. General Vandenberg knew that too. He suggested further study. General Chamberlin, agreeing, said he would like to discuss the problem with General Eisenhower.
When the IAB met on November 7 Admiral Inglis had a new proposal. It modified Vandenberg's plan by confining the joint Intelligence Committee to representatives of the armed services, eliminating the State Department. The joint Intelligence Staff would also lose its representation from the State Department. State should have contact henceforth only through the IAB and NIA. The DCI should be responsible to the NIA in peacetime but to the joint Chiefs of Staff in time of war. Admiral Inglis had evidently been in touch with higher authority in the Navy. The Chief of Naval Operations proposed a similar plan to the joint Chiefs on the following December 9.
General Chamberlin told this IAB meeting of five principles which emerged from his discussion with General Eisenhower. Eisenhower too had misgivings about civilian participation in the committees of the joint Chiefs of Staff; it must remain essentially a military agency. But he would hesitate to change the existing arrangements for coordination with the Department of State. The second fundamental in Eisenhower's thinking was that the DCI had duties beyond the scope of the joint Chiefs of Staff; this fact should be allowed for. Third, there should be no obstacle between the President and the Chiefs of Staff in wartime. Fourth, no civilian agency should be interpolated between the Chiefs of Staff and their agencies engaged in making war plans. Similarly, no agencies not strictly military should have access to military plans.
General Vandenberg accepted these principles but replied that CIG proposed to have a "watertight compartment" for military secrets. The war planners needed the best intelligence available; CIG should therefore work closely with them. The talk went on and on but reached no conclusion. General Chamberlin doubted the wisdom of mixing with the joint Chiefs of Staff the head of an agency which reported to civilian authority, the NIA. Admiral Inglis said that since the DCI reported to the NIA, he should have additional duties for the joint Chiefs of Staff. It did not seem to matter to either that the NIA "civilian" members were the superiors of the joint Chiefs of Staff. Vandenberg brought the discussion finally to an end by proposing further work on the problem by staff members. Another ad hoc committee therefore was named to deliberate with CAPS and bring in majority and minority reports.
This committee agreed on December 3 that each agency should submit its own plan, but the proceeding came to naught. In the meantime the members of the joint Intelligence Committee had endorsed Admiral Inglis' original plan to let the joint Intelligence Staff serve them also when they sat as the Intelligence Advisory Board. Vandenberg saw no point in discussing the question further with these men and took it to the NIA in this meeting of February 12.
General Vandenberg stated his position to Secretaries Marshall, Patterson, and Forrestal and Admiral Leahy—men who understood the operations of the joint Chiefs and the concept of central intelligence. Those who bad created CIG, Vandenberg understood, had in mind that it would replace the joint Intelligence Committee. This, however, had not occurred; nor had any working relationship been achieved. The two organizations continued with parallel responsibilities and no effective coordination. There was constant friction with the intelligence agencies of the War and Navy Departments over priorities. The duplication was unnecessary. lie recommended that the joint Intelligence Committee be abolished. CIG should provide the necessary intelligence for the joint Chiefs of Staff.
Secretary Forrestal inquired if this plan had been taken up with the joint Chiefs. Vandenberg was sure that it had, through the members of the IAB, that is the JIC. Mr. Eddy, present for the State Department, thought it important to abolish the JIC and have all interdepartmental intelligence under CIG. Without further comment the NIA agreed that the Joint Intelligence Committee should be abolished and its functions assumed by CIG, subject to the views of the joint Chiefs, to be obtained by Admiral Leahy.
What the Joint Chiefs were likely to decide had already been indicated. On the preceding December 9 the Chief of Naval Operations, Admiral Nimitz, declared that the time had come to reorganize the JIC on the principle that it should consist only of representatives from the militan, services. Coordination with other agencies should be done through CIG. Of the wartime civilian representation only that of State remained; it should now be removed. The Joint Chiefs approved this concept on February 21, 1947; and there the matter rested for months, as all were far more concerned with actions in Congress over the merger of the armed forces and the creation of a National Security Council. When the question rose again, General Vandenberg was no longer Director of Central Intelligence, was soon to be a member of the joint Chiefs of Staff.
[Top of page]
Soon after CIG was established, Admiral Soucrs took the initiative to coordinate scientific intelligence with the war-created Office of Scientific Research and Development, directing his Central Planning Staff to look into the problem. He obtained Dr. H. P. Robertson as his scientific consultant. The Secretaries of State, War, and the Navy through their Coordinating Committee, predecessor of the National Security Council, investigated Japanese research in nuclear energy and examined the results of the atomic tests at Bikini in the summer of 1946. On August l, Congress passed the act creating the Atomic Energy Commission.
Anticipating the AEC's takeover of the Manhattan Engineering District from the War Department, General Vandenberg had a directive prepared to place intelligence about foreign development of atomic energy within the jurisdiction of the CIG. Vandenberg knew from his experience with the fifth directive in July that his plan would never get by the IAB with phrases like "control and supervision" in it. He accepted the looser concept of "coordination." The papers were circulated to the NIA members on August 13. Secretaries Patterson and Forrestal approved. Admiral Leahy wished only a few editorial changes. But Dean Acheson, Acting Secretary of State while Byrnes attended the peace conference in Paris, would not permit the directive to issue and called a meeting of the NIA for August 21.
Secretary Patterson opened the discussion. It seemed senseless to him that the small division in General Groves' office engaged in collecting information about foreign activities in the field of atomic energy should be kept apart from the CIG. This division was not concerned with the production of atomic energy in this country; there was no reason to leave it under the AEC. Secretary Forrestal agreed, if the AEC were not denied the information. Admiral Leahy favored the plan. Vandenberg assured them that it was designed to provide and not to deny information; he would certainly furnish the intelligence to the AEC as directed by the NIA. But Acheson demurred. He did not want to interfere with the organization which was searching for uranium ore. In any case, they should wait until President Truman had appointed the AEC and it could establish its policies. He had reason to believe that the President would so prefer.
Secretary Patterson persisted in saying that the matter was already within the province of the NIA and was urgent. Admiral Leahv gave Patterson further support. And so Acheson suggested that Leaky clear the directive with the President. This Leahy did by telegraph, noting that the NIA could make any change in the future which the AEC might desire. The President replied on August 23 that he wished to postpone action until he had returned to Washington. This delay was prolonged into December when the appointment of David Lilienthal to the AEC chairmanship aroused opposition.
In the meantime ICAPS became much interested in coordinating the intelligence of several agencies on guided missiles and the requirements of scientific intelligence concerning Russia. General Vandenberg was endeavoring to obtain a close relationship with the joint Research and Development Board which Secretaries Patterson and Forrestal had established on July 3, 1946. President Truman had been inclined to continue the Office of Scientific Research and Development which had done so well during the war under the chairmanship of Vannevar Bush. But it was Bush's own opinion that the OSRD had come to the logical end of its career and should go out of existence, much as the OSS and other wartime agencies had done. He and other members of the old organization were willing to stay at the request of the secretaries and carry on those functions and programs which should be continued.
The new joint Board was to be more concerned with planning than with the operations which had engaged so much of the OSRD time. It was not to be an independent executive agency associated with the joint Chiefs of Staff, but was directly responsible to the Secretaries of War and the Navy. It would not collect intelligence on foreign scientific activities as the OSRD had, but would like to rely on the new Central Intelligence Group for this while it concentrated on planning for the Army and Navy.
It was in this spirit that the Technical Advisors of the JRDB held a preliminary meeting on October 23, 1946. At their next meeting, November 20, they had an ORE estimate to read and consider on Soviet capabilities for developing an atomic bomb, guided missiles, heavy bombers, fighters, radar, and submarines during the next ten years. It was, as it said, at best "educated guesswork." But it was impressive as an interdepartmental product, and in the end it proved surprisingly accurate. With this substantial evidence before them of CIG capabilities, the Technical Advisors listened to Dr. H. P. Robertson, scientific consultant of the DCI, explain the organization of the CIG, discuss problems in the field of scientific intelligence, and suggest ways of mutual assistance for CIG and the JRDB.
The third meeting of the Technical Advisors on December 6, 7, and 8 brought together General Vandenberg, Allen W. Dulles, and General Donovan for a thorough discussion of foreign intelligence. There was no stenotypist present to make a record of their remarks, but a member of the secretariat recalls that Mr. Dulles related his wartime experiences as OSS station chief at Berne, Switzerland. General Donovan repeated with his usual fervor the principles he advocated and the corresponding criticisms of CIG which he had made in Life for September 30, 1946. He did not like having the NIA as a board of control; the DCI should be responsible directly to the President, and the secretaries of the departments should serve as his advisers, not superiors.
General Vandenberg reviewed his difficulties with the intelligence Advisory Board. He was at that time, as we have seen, at odds with the chief intelligence officers of the Army and Navy over his authority in relation to them over requirements and the coordination of collection, and over the place the DCI and CIG should have with the joint Chiefs of Staff.
The result of this three-day conference was agreement that the joint Research and Development Board should find a bead for a CIG section to have charge of evaluating scientific and technical intelligence and should help him obtain the experts necessary for his work. It was further agreed that there should be a statement of the scope of the term "scientific and technical intelligence," a general plan for securing coverage of foreign developments in science, and a definite relationship between the JRDB and CIG.
Vannevar Bush and General Vandenberg issued their program for cooperation in the field of scientific intelligence on January 10, 1947. It provided that the Scientific Branch of ORE should assume the initiative and responsibility for developing a national program of scientific intelligence. The head of the Branch, serving as adviser on scientific intelligence to the DCI, should have direct access to JRDB activities pertaining to his work.
The Scientific Branch would formulate requirements for scientific intelligence in collaboration with JRDB and the departments and agencies concerned. It would be responsible for planning and coordinating collection. It would prepare estimates on the scientific capabilities and intentions of foreign countries. It would correlate these scientific estimates with those in other fields of intelligence for the production of strategic intelligence.
CIG undertook to provide the JRDB with the intelligence to meet its needs, particularly in regard to foreign items of specific interest. On its part, the JRDB undertook to cooperate in supplying CIG with qualified personnel, special facilities, and close day-to-day liaison on scientific matters.
General Vandenberg had endeavored to arrange a meeting of the NIA for January 6 to approve this agreement with Bush before, issuing it, but was unable to do so. Nor had the person to head the Scientific Branch been obtained when it was formally established on January 23. Vandenberg wrote to Bush on March 13 to say that he was appointing an acting chief from within CIG who would report to the chairman of the JRDB and make himself and the Branch as a whole fully available. It was a long while, however, before: the Branch was equipped to perform the functions stipulated in the agreed program of cooperation.
In accordance with Secretary Patterson's report to the NIA on February 12 about the arrangement for transferring the files and personnel in the intelligence division of the Manhattan Engineering District—not including information about uranium deposits—to the CIG, the transfer was completed on February 18. On March 28 they became the Nuclear Energy Group in the Scientific Branch of ORE. A directive authorizing the DCI to coordinate all intelligence related to foreign development of atomic energy was issued on April 18.
[Top of page]
1 For Part I, "Some Functions Centralized," see Studies XII 3, p. 79 ff.
2 This plan is described in greater detail in Studies XII 1, p. 66 f.
3 Studies X11 1, pp. 62-3 and 67-8.
4 See Part I in Studies XII 3, pp. 79-83.
5 Forerunner of the NIS on the USSR.
[Top of page]
No Foreign Dissem