Foreign Relations of the United States, 1945–1950, Emergence of the Intelligence Establishment

83. Memorandum From the Secretary of State’s Special Assistant for Research and Intelligence (McCormack) to Secretary of State Byrnes⁰

Washington, December 31, 1945.

Mr. Russell’s proposal for the intelligence organization¹ modifies the one recommended by the Intelligence Advisory Board in four particulars, of which two are not important enough for the attention of the Secretary. The two important points are:

a) It provides for dismemberment of the Office of Research and Intelligence after three months (March 31, 1946); and

b) Of the proposed geographical divisions of that Office, it leaves only the American Republics unit in the status of a Division and lumps all the rest in a single Division of European, Near Eastern, African, and Far Eastern Intelligence.

With respect to the latter point, Mr. Russell’s memorandum says:

“Whatever alternative may be approved, the divisions of the Office of Research and Intelligence should be changed to conform to the geographic pattern established for the other Offices of the Department. No justification can possibly exist for different geographic breakdowns. It would place the Department in a ridiculously inconsistent position to approve a geographic division for a new Office of the Department wholly different from that established and approved for the traditional Offices of the Department.”

It is difficult to see anything “ridiculously inconsistent” in an intelligence agency geographically subdivided into (1) American Republics, (2) British Commonwealth, (3) Europe, Near East, and Africa, (4) Far East, and (5) USSR. Those are the main politico-geographic subdivisions of the world. It is believed, on the other hand, that the alternative proposal is quite illogical in setting up the Latin American unit as a Division, with 35 people, and then lumping all the rest of the world, with 326 people, into another Division. The political considerations that require an Assistant Secretary for Latin American Affairs have no bearing on an intelligence organization. Therefore that unit should have the same status as all the others. All should be Divisions of the proposed Office, or all should be reduced to the next lower status (at the expense of adequate Civil Service grades and the certain loss of key personnel). To do otherwise will lift the Latin American unit to a status that it is not entitled to have.

As proposed by me, the geographical Divisions conform generally to the Political Offices, but not exactly, because the present organization does not lend itself to precise conformity. Mr. Clayton’s and Mr. Benton’s geographical breakdowns likewise do not conform precisely to those of the Political Offices. In each case the problem is different. The Political Offices can hardly claim that their geographical breakdowns are either immutable or completely logical, when, for example, they put New Caledonia under Europe and Greece under Middle East and Africa.

It is strongly recommended that the organization as proposed by the Intelligence Advisory Board be approved.

As to whether the research operation should be decentralized to the geographical and functional offices, the argument for decentralization is made as follows:

“I believe that research at the geographical level must be under the immediate direction of those who are to use it. In my judgment, the divorce of research from the policy action taken after the evaluation of information will lead inevitably to wasteful duplication and to competing evaluations of information that will breed confusion and disorganize the operations of the Department.”

There is also a quotation (at page two) of the views of the Geographic Offices. My views to the contrary are represented by a quote from a draft prepared by a member of the Working Committee of the Intelligence Advisory Board, which was not in the Committee’s report and is not an adequate statement of either my views or theirs. A full statement of the Committee’s views appears in its report.

My own view is, briefly, that the chances of keeping the R&A Branch together, strengthening it with new personnel, depend on its maintenance as a unit. It is an integrated organization with common management, procedures and files. It is flexible, in that personnel from one geographic unit can be shifted to meet peak demands in others. It does a whole job of processing incoming information and collating it for everybody’s needs. It is independent of policy makers and adheres to the standard of intelligence offices, that they must keep out of policy and maintain objectivity, since their mission is fact-finding.

That an intelligence organization must be free of operations or policy involvements is fundamental. That such freedom could exist in the 20-odd Divisions of the Geographic Offices is unthinkable.

In my opinion, decentralization would destroy the R&A Branch. It would give the Geographic Offices some additional personnel, but it would end all possibility of organized State Department intelligence, and the President’s idea of State Department leadership in government-wide intelligence could not be attained.

A centralized intelligence operation within the Department can serve the needs, not only of the geographical desks, but of the economic organization, the Office of Public Affairs, the Office of Special Political Affairs, the various Committees which play such an important role in the Department, and such quasi-Departmental committees as the Far East Advisory Commission and the Interdepartmental Committees. A Research and Intelligence unit decentralized to the geographical divisions can serve only those divisions.

The centralized operation can look at the national intelligence problem as a whole, coordinating the work of its component parts and gearing itself into the other intelligence organizations of the government. Within each of its component units, the problems of any area can be looked at in all aspects, since the political scientists, the economists, the geographers and other specialists belong to a single unit, working closely together and having common files and a common flow of information.

The interests of an intelligence office, in its regional divisions, go far beyond the interests of the geographical desks, which are concerned with current problems arising in the conduct of our foreign affairs. The geographical desks have neither the time nor the training to engage in a systematic compilation of basic information for future intelligence purposes. Even if they had, they still could not do it, because different habits of thought and a different frame of mind are required for research work than for the daily operating job.

Decentralization of the research and intelligence operation is not going to eliminate the need for centralized activity. It is only going to make more difficult the attainment of the President’s objective of a coordinated government-wide intelligence program.

The proposal would destroy the effectiveness of the R&A Branch, assuming that all the personnel would continue to perform the functions that they now perform. That, however, would not happen. To state the problem in terms of numbers, in the regional divisions of IRIS there are 272 professional employees, half of whom are strictly economists or geographers, leaving about 136 who could be decentralized to the 19 geographic divisions of the Department, as follows:

Far East	41
Europe-Africa	59
Latin America	16
British Empire	3
USSR	17
	136

This is an average of less than 7 persons per Geographic division. Included in the 136 there are perhaps 15 key people who have kept the organization together. These 15 people are scholars who perform or supervise research and who well know that the decentralization of the R&A Branch would destroy its usefulness. I doubt if any one of those people would stay in the operation if it were decentralized. Practically all those who have discussed the problem with me have stated that they would resign at the first opportunity if such a step were taken.

If the operation is decentralized, the personnel who go to the geographical desks will have find their futures in those divisions. The problems of promotion and advancement in the geographical units will be quite different from those in a unified research organization. The good opportunities will come, not in research but in operating and policy jobs; and there will be an absorption into such jobs of such good personnel as will remain. That is what happened when the Territorial Studies unit was decentralized. No vestige of it, I am informed, now remains in the Department.

It is strongly recommended that the Office of Research and Intelligence be set up, as proposed, as a definitive organization. If, during the fiscal year 1947, changes in the direction of decentralization appear desirable in the light of experience, they can then be made. To set up the organization in a particular way does not mean that it can never be changed. It does, however, give assurance that those responsible for creating a State Department intelligence organization will have control of their personnel and the opportunity to work out their relations with other departmental units within a definite framework of responsibility.

0. Source: National Archives and Records Administration, RG 59, Records of the Department of State, Records of the Bureau of Intelligence and Research: Lot 58 D 776, Birth of the Intelligence Organization in the Department of State. No classification marking. Also sent to Under Secretary Acheson.
1. Document 82.