55. Memorandum From the President’s Assistant for National Security Affairs (Brzezinski) to President Carter 1


  • Reorganization of the Intelligence Community

There are essentially two broad alternative approaches to organization of the Intelligence Community:

—Consolidation of all predominantly national intelligence activities into one bureaucratic structure under the full control of the DCI; or

—A “community” approach which differentiates to some degree responsibilities and authorities.

To evaluate these approaches, it is necessary to analyze their impact on the key operational functions of setting requirements, tasking authority, analytic production, resource management and line authority. On the basis of the PRM/NSC–11 studies and subsequent SCC discussions, it seems clear that any approach to organizing the Intelligence Community should:

—recognize that the major consumers of intelligence should play a dominant role in establishing requirements and prioritizing them through a high-level committee system;

—give greater power to the DCI during peacetime to translate consumer requirements into detailed intelligence collection objectives and task these to appropriate intelligence collection organizations;

—leave primary responsibility for national analytical intelligence production with the DCI, who would remain your principal substantive intelligence adviser, but provide for the continuation of departmental analytic centers;

—give the DCI a strong and leading role in the resource management of all predominantly national intelligence programs.

Both Stan Turner and Harold Brown are in general agreement with these principles, although they have some differences over the details of organizing to implement. However, as advocates of specific centralization and community options, their two approaches are most clearly distinguished on the issue of line authority, particularly over the military and technically-oriented major collection programs now within the Department of Defense.

[Page 300]

—Any full centralization approach would involve the transfer of complete line authority over the major DOD collection programs to the DCI. The basic argument for doing so is the controversial assertion that this is necessary to assure proper performance of an intelligence system that is responsive to both national and unique DOD intelligence requirements. This point has not been supported with examples of failures or other case studies, and as such is subject to challenge.

—Any “community” approach by definition assumes a certain differentiation and dispersion of responsibilities and authorities. It is based on the premise that, while certain critical operational functions, like resource management and national tasking, can be performed effectively by a centralized authority, full line control over functions that have both national and departmental significance should be decentralized to assure responsiveness to both. While there has been a gradual trend toward greater centralization in recent years, this principle has in the eyes of many observers remained valid.

Agency Options

During the course of the SCC deliberations, Stan Turner, Harold Brown, and OMB each developed detailed options for your consideration. As you will recall, the major features of these options are as follows:

Brown’s “community” option would modify the status quo by (a) strengthening the DCI’s and PRC(I) collegial role in managing national intelligence resources, (b) providing for a high-level consumers committee within the NSC system to establish intelligence requirements, and (c) explicitly delegating to the DCI all responsibility for tasking collection facilities during peacetime (subject to appeal to a consumers committee) and to the Secretary of Defense during crisis or war. No changes would be made in the basic organizational structure of the Intelligence Community or in its normal daily mode operation.

Stan Turner’s “consolidation” option would involve full centralization of national intelligence activities. It would place the present CIA, NSA, NRO [less than 1 line not declassified] under the full line management control of the Director of Central Intelligence and functionally integrate some major collection systems. Departmental analysis units would remain basically independent of DCI control. A high-level interagency consumers committee would be established to identify priority national intelligence needs, subject to your approval, and a DCI controlled joint civilian-military center would actually task collection systems.

OMB would diversify the line management authorities inherent in the secretaries of most governmental departments by centralizing the most critical national intelligence management functions (tasking, resources and analytical production) under the DCI while leaving other [Page 301] operational and administrative functions normally associated with line authority decentralized. In addition, OMB would functionally integrate all technical and human source collection activities and national analytical production into new separate agency structures.

Compromise Option

I believe that the Turner, Brown and OMB options each has some constructive new elements that together could provide the basis for a reorganization decision along the following lines:

—Requirements would be established and prioritized by the Policy Review Committee, chaired by the DCI. All agree that the major consumers should set requirements, and putting the official charged with implementation and who has the greatest vested interest in success in the chair should assure that it gets accomplished effectively.

—The DCI would be decision-maker on tasking the various elements of the Intelligence Community to fulfill requirements and priorities. The Secretary of Defense would have a right of reclama to you through the NSC system and could, at your discretion, be given full intelligence tasking power during times of extreme crisis or war. This should be acceptable to both Stan Turner and Harold Brown, especially if some civilian-military tasking mechanism is created, such as Turner’s National Intelligence Tasking Center.

—The DCI would have full authority to provide guidance on the development of the national intelligence budget, approve its content prior to submission to the President, present it to Congress, reprogram funds as necessary (though it may be difficult to get the Congress to loosen their reins on reprogramming). The National Foreign Intelligence Board would replace the PRC(I) but in an advisory role to the DCI on his budget decisions (in the same manner as it now advises him on national estimates and other activities of common community concern). This goes beyond Harold Brown’s proposal which retains a collegial system but department heads still would have the right to reclama DCI budget decisions to the President. The DCI would have adequate staff and access to information to ensure that he could carry out the program audit and evaluation necessary to his budget and tasking responsibilities.

—The DCI, as the President’s principal substantive intelligence adviser, would continue to have full responsibility for the production of national intelligence in appropriate consultation with departmental analytical centers.

—The essentially implemental and administrative elements of line management authority such as personnel actions, support activities, operational control of systems and military entities and audit/inspector general functions would remain as presently assigned under departmental arrangements.

[Page 302]

The basic rationale for this approach is that it centralizes the most critical national intelligence management functions under the DCI—tasking, resources, and production—while leaving the administrative and support functions with the operational elements where they are performed adequately today. This is the same assumption on which the OMB option is based. It also builds on the concepts behind the Turner and Brown options as follows:

—It recognizes that, while there are certain major intelligence programs and tasks which should be directed at the national level, the distinction between national and tactical intelligence is increasingly artificial, and in the future intelligence systems must be responsive to the concerns of all users. This is a fundamental point made by the military services and the basic reason they resisted centralization.

—The link between consumers requirements, tasking and resource allocation is centered for the first time in the office of the DCI and should in theory result in more productive and cost effective collection and production activities. This was the most critical deficiency identified in the PRM/NSC–11 study and the basic argument for consolidation of authority.

—Finally, while some reorganization within CIA and the Defense intelligence agencies may be necessary, any approach which divided them up and reallocated their activities into new units could completely break the already low morale of their professional cadre, and would minimize the element of constructive competition that has stimulated creativity in the past. This is a basic point of reality overlooked in the OMB option.


That you sign the Presidential Directive at Tab A2 based on the above indicated compromise principles for reorganization of the Intelligence Community. Based on this we will develop a public statement for release by Jody Powell that dampens speculation about who won or lost.3

  1. Source: Carter Library, National Security Council, Institutional Files, 1977–1981, Box 12, PD 17 [4]. Confidential. Sent for action.
  2. Not found attached. The Presidential Directive is printed as Document 59.
  3. For the text of the August 4 public statement, see Public Papers: Carter, 1977, Book II, pp. 1421–1423.