42. Memorandum From Director of Central Intelligence Turner to the President’s Assistant for National Security Affairs (Brzezinski)1


  • PRM/NSC–11, Task 2 Report

1. Submitted herewith is the subject report as directed by the President. I believe it provides an instructive overview of the functions, powers, and problems of the Director of Central Intelligence (DCI), particularly in his role as leader of the Intelligence Community.

2. On the basis of my past experience and all I have learned since becoming DCI, I have formed some strong views on what is needed in the way of improvements to Community structure and to DCI authority to make the Community more effective and efficient, and to assure that its activities are demonstrably proper. I have expressed such views in this report.

3. Not surprisingly, there are those who differ sharply with some of my views. Representatives of the Department of Defense, in particular, take exception to some of them in the attached report. Secretary Brown and I have had an extensive and constructive exchange on these matters. I believe the time has come to submit them to the test of review and debate in the Special Coordinating Committee.

Stansfield Turner2


Report Prepared by the Ad Hoc Interagency Subcommittee on the Role of the Director of Central Intelligence3

The Roles of the DCI and U.S. Intelligence: An Organizational Analysis

[Omitted here is a table of contents.]

[Page 226]


In PRM/NSC–11, the President directed a comprehensive review of the missions and structure of United States intelligence entities with a view to identifying needed changes. As part of this review, the Director of Central Intelligence (DCI) was directed to chair an interagency subcommittee of the Special Coordination Committee (SCC) of the National Security Council (NSC) to analyze his own role, responsibilities, and authorities.

This subcommittee was comprised of representatives of the DCI (Central Intelligence Agency, National Intelligence Officers, and Intelligence Community Staff), the Defense Department (Office of the Secretary of Defense and the Joint Chiefs of Staff), the Department of State (Bureau of Intelligence and Research), and the NSC Staff.

Specifically, the PRM/NSC–11, Task 2, called for a report that reviews “the responsibilities and powers of the DCI in his role as Foreign Intelligence Advisor to the President, central authority for the production of national intelligence and manager of the national foreign intelligence program and budget. This examination should include an analysis of the mechanisms for:

  • —planning, evaluating, and improving the Intelligence Community performance;
  • —identifying intelligence requirements and tasking all sources;
  • —processing, analyzing, producing and distributing intelligence for anticipated activities, warning, crisis support, current and estimative intelligence and net assessments;
  • —evaluating intelligence production performance.”

Because this report is devoted, as tasked, to the roles of the DCI, who is but one of several senior authorities responsible for the activities of the Intelligence Community, it cannot completely treat the roles of other such authorities. Representatives of the Department of Defense (DOD) believe this is particularly the case regarding the roles of the Secretary of Defense, who manages nearly 80 percent of the financial resources of the National Foreign Intelligence Program, who is executive agent for several major intelligence programs of great importance to national as well as to DOD’s intelligence concerns, and whose principal functions require intimate involvement in national intelligence affairs.

DOD wishes, further, to state the following: It should finally be noted that the text was changed in many respects at the direction of the DCI after the last Subcommittee meeting.4 In DOD’s view, these changes serve to make this report principally a vehicle for the expres[Page 227]sion of the DCI’s views on the changes he believes are appropriate in the Intelligence Community structure. DOD also believes that the Executive Summary does not represent a balanced presentation of the main text.


Intelligence is a diversity of collection and production organizations serving a variety of customers with varying needs from the President down to military commanders and diplomats in the field.

—The Central Intelligence Agency (CIA) and the position of the Director of Central Intelligence (DCI) were created to afford a degree of unity amid this organizational diversity.

—The roles of the DCI and of the other officials with whom he interacts in this federated community of organizations evolved, and the size and diversity of US intelligence have grown over thirty years.

—The Department of Defense (DOD) retains a very large role in US national intelligence affairs, with management custody of some 80 percent of the National Foreign Intelligence Program (NFIP) budget, including major national technical collection programs; and DOD has major specialized intelligence needs in the areas of force and weapons development and tactical operations.

In recent years, largely as a result of the Community’s size and diversity, questions have arisen about the adequacy of the organization and management of the Intelligence Community and of the role which the DCI plays within it. The key structural questions are:

—Whether the responsibilities of the DCI are clear and sound, particularly as they relate to intelligence entities within DOD.

—Whether the authorities and powers of the DCI are commensurate with his responsibilities.

Of the DCI’s many roles, the most important are:

—Principal advisor to the President and the National Security Council on foreign intelligence matters;

—Producer of national intelligence;

—Leader of the Intelligence Community;

—Head of the CIA.

[Page 228]

The first of these roles has important implications for Community structure.

—To the extent that there is a perceived need for someone to organize and manage the intelligence affairs of the US Government as a whole, there is a tendency to look to the DCI.

—In one view, held by the DOD, this tendency can lead to an unwise deepening of the DCI’s involvement in the management of other agencies’ intelligence affairs, and an unhealthy dilution of the DCI’s primary substantive role.

—The DCI believes, however, that this tendency is both natural and legitimate. The resulting expansion of DCI responsibility can be appropriately handled through delegation of duties to subordinates.

The DCI’s substantive role as producer of national intelligence originates with the duty given the CIA in the National Security Act of 1947 to “correlate and evaluate intelligence relating to the national security.”

—Although there are weaknesses in this area, the DCI has significant power to remedy or alleviate problems; improvements are frequently more a matter of judgment and management attention than of authority.

—However, the DCI has little power over the departmental contributors on which the analysis and production of national intelligence so heavily relies.

The DCI’s resource management responsibilities in the Intelligence Community have two time dimensions: the use of existing collection and processing resources to meet current and near-term intelligence needs; and the development of new resources to meet future intelligence needs.

—Centralized mechanisms for the guidance of major current collection activities exist at the national level, under the DCI, in the case of technical collection assets. DCI powers are strong and prescriptive in the area of imagery; somewhat less strong in the case of SIGINT. Many argue that difficulties here arise not so much from lack of DCI authority or from failings of Community structure, although the fragmented structure of the Community has helped to instill in each collection discipline a disposition to want to manage its own affairs with only general guidance. Frequently, difficulties are in defining problems and designing workable improvement mechanisms—for example, managing collection tasking during the transition from peace to war and assuring reliable cooperation between the Community and overt human source collectors outside of intelligence (e.g., in the Foreign Service).

—A greater challenge for US intelligence management is to develop the best overall mix of future capabilities needed to perform effectively [Page 229] at reasonable cost. A fundamental problem is one that is common to other functional programs in government: the absence of a set of measures for assessing the value of outputs and the relative contribution of inputs in terms that find general acceptance and lead to confident decisions.

In his role as head of the CIA, the DCI has strong management powers, but the augmentation of the DCI’s role as Community leader has been perceived, in recent years, to cause increasing tension between the two roles.

—Some in the Community see the DCI as bound to favor CIA in any Community deliberation on production, requirements, or resources in which CIA has an interest, and therefore argue for some degree of DCI separation from CIA.

—Others contend that part of the problem stems from the imbalance between the DCI’s broad responsibilities and his more limited decisionmaking powers in the Community arena; this forces him into a position where he must appear to neglect the CIA to be effective as a negotiator in the Community. Those of this view tend to favor enhancing DCI authority over other Community elements.

Most of the DCI’s other roles are subsidiary to these four primary ones and have fewer implications for Community structure.

—To help protect the security of intelligence sources and methods, past DCIs have sought new legislation to punish damaging disclosures of sources and methods information; other initiatives—such as reinvigoration of the classification system within the Community—are also needed.

—The DCI is a participant in US foreign counter-intelligence policies and activities; there is a clear need for a national level policymaking and coordinating structure in this area.

—As an officer responsible for the propriety of US foreign intelligence activities, the DCI has an Inspector General and the normal mechanisms for discovery and investigation of impropriety within CIA. Although charged under Executive Order 11905 to ensure effective Inspectors General in other agencies, he has little power to act on this charge and is generally not equipped to assure propriety in the behavior of agencies other than CIA.

—Occasional confusion about the DCI’s responsibilities as coordinator of liaison with foreign intelligence services would appear to require some clarification in pertinent regulations.

—With respect to his role as principal spokesman to the Congress on national foreign intelligence, one of the foremost problems for the future may be to find a way in which the DCI can respond to the proper demands of Congress without jeopardizing Presidential prerogatives and DCI relations with the Executive.

[Page 230]

—Regardless of the organizational configuration of the Intelligence Community, the DCI almost certainly will be expected to continue the trend toward greater openness and to accept a continuing role as public spokesman on national foreign intelligence.

Three basic criteria, especially pertinent to the roles of the DCI, can be used in assessing the adequacy of management and authority structures within the Community: propriety, effectiveness, and efficiency.

—In the view of DOD, these criteria, as discussed in this paper, do not fully address other criteria important to the roles of the Secretary of Defense, especially the need for adequate integration and interoperability of intelligence with military command and control.

Assuring the propriety of intelligence activities is not solely—or, in the view of some, primarily—a matter of Community structure or authority. It is a matter of political or constitutional standards, law and regulations, oversight, and professional ethics. The DCI cannot, at present, be held directly responsible for the actions of agencies which he does not directly command.

—Although legal responsibility for the propriety of intelligence operations runs from the President down through the line managers of the several intelligence agencies, the DCI believes that the President, the Congress, and the public expect him to act as virtual guarantor of the propriety of all United States national foreign intelligence activities below the President. In the DCI’s view, his authorities to satisfy these expectations are now less than adequate, except in the case of CIA.

Improving the overall effectiveness of national intelligence production does not rest mainly on structural change or redistribution of management authority. Improvement requires problem recognition and steady management effort at all levels and in all producing agencies. But efforts to improve intelligence production do have implications for Community structure, and changes in structure sought for other reasons could affect the quality of intelligence production. Effective service to consumers requires a diversified set of producing organizations, some of which are directly subordinate to consumer entities, all of which are able to act in concert when required. The Intelligence Community today affords such a structure.

—The DCI believes that the diversified structure of the national intelligence production Community existing today is generally sound. In his view, however, more effective national intelligence production requires enhancing the DCI’s authority to:

a. Task Community production elements outside CIA for national intelligence production;

b. Task national collection assets that lie outside CIA but support national intelligence production;

c. Control the program management of the major NFIP elements.

[Page 231]

DOD disagrees with this view. It believes, moreover, that such enhancements of DCI authority could materially degrade the responsiveness of DOD collection and production elements to DOD needs.

Achieving the most efficient allocation of resources is mainly a matter of managing collection and processing resources, because that is where most of the money and manpower are. The challenge is to provide the necessary coverage of target problems and adequate service to consumers, while avoiding unnecessary effort and undesirable duplication.

—With regard to the management of current collection requirements, priorities, and tasking, the DCI believes that, notwithstanding his central role respecting technical systems today, enhanced DCI direct tasking or line authority over major national collection entities is essential to improve their responsiveness to all consumers and to eliminate the high degree of competitive overlap that presently exists.

DOD disagrees with this view. It maintains that such enhanced DCI authority would probably work to reduce the responsiveness to DOD needs of those major collection entities within DOD.

Historically, programming and budgeting aspects of US intelligence resource management, as well as line control, have been largely decentralized, both in the Community as a whole and in DOD, where most of the resources reside. But pressures to centralize the process of managing those resources labeled “national” have been increasing for several years, culminating last year in Executive Order 11905.

—The programming and budgeting decision system initiated by Executive Order 11905 is essentially collegial (in the PRC[I])6 and rests on the cooperative interaction of the DCI, departmental authorities, their staffs, and intelligence program managers. To a large extent, it places the initiative in the hands of program managers and outside critics. As a by-product, it places some strain on the dual roles of the DCI as a Community leader and as head of CIA. It also, as a practical matter, requires that departmental authority over departmental intelligence elements in the NFIP be compromised; the Executive Order does not eliminate the statutory responsibilities of the department Secretaries over their intelligence activities.

—Refinement of the programming and budget process created by that order is one way of enhancing the integrity of national intelligence resource management in the future; it has the significant virtue of an evolutionary approach that builds on existing organizations and accumulated experience. Better definition of goals and rules is desirable [Page 232] to make the process of persuasion inherent in the collegial approach more constructive.

In deciding whether significantly to change this regime, several issues are relevant, such as:

—How much emphasis should be placed on efficiency as compared with other goals;

—What intelligence activities should be involved;

—How much and what kind of centralized authority is desirable?

The last question involves at least four conceptually distinguishable management activities: definition of requirements and priorities, and issuance of guidance; reviewing and vetoing Community programs; controlling programming and budget decisions; and exercise of line management. Each activity could, in theory, be centralized or decentralized, could be unilateral or collegial, could be mandatory or advisory. The relevant options and responses are addressed in other parts of the PRM/NSC–11 response.

The DCI believes, however, that present arrangements give him responsibilities in intelligence resource management that are beyond his management authority to fulfill. Although formal responsibility for the contents of the NFIP rests with a collegial body, the PRC (I), as Chairman and as DCI he is expected by the President and the Congress to develop and take responsibility for an NFIP that is rigorously efficient and displays a close relationship between resource inputs and intelligence product outputs. In the DCI’s view, achieving the goals of efficient national intelligence resource management requires his having stronger central authority over national intelligence programming and budgeting decisions, and, in the case of key national programs, line authority as well.

DOD disagrees with this view. It maintains that the degree of centralization under the DCI implied above would be unwise and would severely prejudice the ability of major collection programs in DOD to meet important Defense needs in peace, crisis, or wartime.

I. Introduction

Intelligence can be thought of as a service industry in government, a diversity of organizations serving a variety of customers with varying needs. At the origins of post-war US intelligence, Congress and the President responded to a strongly perceived need to create some degree of overall unity amid this departmental diversity. The Central Intelligence Agency (CIA) and the position of the DCI were created to afford a degree of unity—as well as some independence from the policy process—with respect to information and judgment on intelligence questions of national importance. In the intervening years, the size and diversity of US intelligence have grown. (See Figure 1 and other [Page 233] graphics at Annex for an indication of the size and diversity of today’s Intelligence Community and its activities.)7 But so also have the pressures for unity amid diversity. As the nation’s senior, full-time functionary for national foreign intelligence, the DCI has been the focus of these pressures. He is the President’s principal advisor on foreign intelligence, and national intelligence of preeminently Presidential concern is produced under his authority. He has come to preside over Community mechanisms that decide how to use major technical collection capabilities on a day-to-day basis. Since the November 1971 directive of President Nixon, he has been increasingly expected by the President and the Congress to be the guiding authority with regard to programs and fiscal resources of US intelligence entities specified as national.

A direct line of authority runs from the President and his advisory body, the NSC, to the DCI and the CIA. Surrounding this line of authority, however, are a host of vital relationships with other entities of the Executive Branch which generate and receive intelligence. These other relationships do as much to shape the role of today’s DCI as does his line command of CIA. For many years, CIA has itself been highly dependent on them. In recent years, they have been seen within CIA to strain the DCI’s relationship with the Agency.

Of these other relationships, that with the Department of Defense (DOD) is the most significant and involved, strongly influenced by the fact that the Secretary of Defense, by virtue of his statutory responsibilities as head of the Department of Defense and member of the NSC, has his own direct line of authority from the President. Characterizing this relationship with the DOD goes a long way toward defining the role of today’s DCI. It shall be treated further in following sections. Suffice it to say here that:

a. The DOD consumes the greatest volume of foreign intelligence. In scope and variety, DOD needs for intelligence approach those of the rest of the government combined. Many of its needs arising from force planning and operational action responsibilities are large and unique.

b. Much of the raw intelligence on which the performance of the DCI as an intelligence producer depends is collected and processed by intelligence elements within the DOD. The Secretary of Defense, for example, as executive agent of the Government for signals intelligence (SIGINT), manages the National Security Agency (NSA) as a service of common concern for all agencies and departments, within the basic [Page 234] requirements framework established by the DCI with the advice of the National Foreign Intelligence Board (NFIB).8

c. Defense intelligence production entities, in addition to supporting DOD consumers, play a major role in the development of national intelligence judgments through the NFIB and the medium of national intelligence estimates. In some areas of analysis, their contributions are unique.

d. Because nearly 80 percent of the National Foreign Intelligence Program (NFIP) is located in the DOD, it is with the intelligence authorities of this department that the DCI and his Community Staff must interact most intensely to develop the consolidated NFIP and budget.

e. It is in the relationship with DOD that the interwoven complex of national, departmental, and tactical intelligence needs and capabilities arises most sharply to complicate the definition of the DCI’s role.

f. In the event of war, and even in some peacetime situations, the DCI’s role could conflict with that of the Secretary of Defense.

[2 lines not declassified]

Although not as complex, the DCI’s relationship with the Department of State is also vital. Foreign Service reporting—a form of collection not formally identified as intelligence—makes the major contribution to political and economic intelligence and also provides information on defense policies in many parts of the world. [1½ lines not declassified] The Department of State in Washington and Ambassadors overseas deal with the foreign affairs aspects of all foreign intelligence programs and projects, and play key roles in coordinating overt collection in the field [1 line not declassified] The Department is also a heavy consumer of foreign intelligence, and its Bureau of Intelligence and Research (INR) both contributes to national intelligence judgments and produces unique political analyses.

Small in size and specialized in interest, the intelligence elements of the Treasury Department, Energy Research and Development Agency (ERDA), and Federal Bureau of Investigation (FBI) flesh out the formal intelligence relationships of the federation of agencies which has come to be called the Intelligence Community. These latter agencies and the departments they serve have increased in importance as intelligence has had to diversify into new areas of international economics, nuclear proliferation, terrorism, and international narcotics traffic.

Finally, other departments and agencies outside the Intelligence Community—the Department of Agriculture, the Department of Commerce, the Arms Control and Disarmament Agency (ACDA), the United [Page 235] States Information Agency (USIA), and others—are collectors as well as important consumers of foreign intelligence (See Figure 2 for an overview of the Governmental components which have foreign reporting capabilities.)9

The purpose of this report is essentially to describe and assess the unifying roles of the DCI, along with other, in some respects conflicting, roles which he has in this Community.

II. Basic Criteria for Organizational Judgment

In understanding or structuring any management system, a first task is to establish the functioning spheres of responsibility and authority, and their limits—essentially how the cloth is divided. The second task is to establish how and to what extent that cloth is sewn back together in order to overcome the negative aspects of necessary divisions of responsibility and to make the parts function as a whole. This is a large challenge for US intelligence because of institutional and functional diversity and the countervailing necessity that the parts interact as a whole.

One approach that can be used to rationalize Community structure is to argue distinctions between national, departmental, and tactical intelligence. This tripartite formula arises largely from the relationship of the DCI and the DOD, and is reflected as well in the intelligence-related functions of other departments, e.g., in the reporting of Foreign Service Officers or Commercial attaches. This formula has serious weaknesses and frequently confuses more than it clarifies. Defining the terms usually obliges use of other terms left undefined. For example, it is said that national intelligence is that intelligence needed by the President, the NSC, and senior US officials to make national policy decisions. But what are national policy decisions? They are decisions those officials want and are able to make; they frequently reach deep into the affairs of departments and can dictate the tactics of military and diplomatic actions. (Further complications arise, for example, within the SIGINT Community, where it is asserted that collection assets are best distinguished along global and local—rather than national, departmental, or tactical—lines.)

The essence of the organizational problem in intelligence is that these concepts overlap extensively in meaning, at least some of the time. The needs of consumers overlap. The President is always interested in broad assessments of Soviet foreign and military policy. But in a crisis at sea, he is likely to be interested in the exact location of specific naval combatants, a seemingly tactical issue. By the same token, a field [Page 236] commander or foreign mission chief needs broad strategic assessments as well as tactical information. The uses to which a given intelligence fact or judgment can be put also overlap in the tripartite formula. An assessment of the hardness of Soviet missile silos, for example, can be of direct value to the operational planner of strategic strikes, to the force planner, to strategy and national policy planners, and to the arms controller; the President is likely to be interested in all these applications. The organizations and systems that collect intelligence data also overlap the categories of national, departmental, and tactical. This is particularly true with emergent space-based reconnaissance systems that may monitor arms control agreements, collect order of battle data, supply warning, and support tactical military operations.

Thus, the key organizations and systems of US intelligence can or do play extensively overlapping roles at different times. Although only imprecisely, one can distinguish among primary and secondary missions of major organizations in terms of the national, departmental, and tactical formula. But this does not resolve all cases; it leaves a middle ground for argument and a poor basis for organizational judgment.

Organization is about management, and management is about basic purposes and standards of performance. Organizational judgment must be based on a clear understanding of basic performance criteria that do or should govern US intelligence. Among such criteria, three especially pertinent to the roles of the DCI are propriety, effectiveness, and efficiency. (In the view of DOD, these criteria, as discussed in the succeeding pages, do not fully address other criteria important to the roles of the Secretary of Defense, especially the need for adequate integration and interoperability of intelligence with military command and control.)

Propriety demands that US intelligence be conducted in conformity with the legal and political standards of our country as interpreted by proper authority. In today’s conditions, propriety may tend to conflict with effectiveness and efficiency by restricting certain means of collecting or using intelligence or forbidding the collection or use of certain kinds of intelligence. It tends to conflict with intelligence requirements for secrecy on which effectiveness and efficiency depend. Assuring the propriety of US intelligence in appropriate balance with conflicting considerations is not primarily a matter of organization, although clear lines of command and management responsibility are required for this task. Assuring propriety also requires:

a. establishing a sound environment of law and regulations;

b. establishing sound oversight or policing mechanisms within and outside intelligence organizations; and

c. cultivating appropriate professional and managerial ethics within intelligence entities.

[Page 237]

The concept of effectiveness in intelligence management is output or product oriented. It is, therefore, preoccupied with consumers and with how well they are being served—with who the consumers are, what they need, when they need it, and why they need it. As already indicated, US intelligence serves a great variety of consumers with a great diversity of needs. Within the Executive Branch, they can be arrayed in the following rough hierarchy:

a. the President, the NSC, and Cabinet-level decisionmakers; those who decide the policies of the Administration on foreign, military, arms control, and foreign economic matters, and on crisis management.

b. policy and strategy planners; option developers; force posture, major program, and budget developers; planners of negotiations; those who present the Presidential and NSC level with structured choices on broad policy issues and crisis options.

c. central implementers of policy and operational planners in foreign, military, and foreign economic areas;

d. field and tactical decisionmakers; policy or plan implementers, e.g., diplomats, negotiators, and military commanders.

These kinds of intelligence consumers are found, of course, in the main departments of the US national security establishment: the Executive Office of the President and the NSC Staff, State, Defense, ACDA, and, to a lesser extent, in most other departments and several regulatory agencies. One must also count Congress as a substantial consumer of intelligence, and, to a degree, the public, which receives much of its information about events overseas, at least about the Communist world, indirectly from US intelligence. Because it must store up information and analysis to meet future or unexpected needs, intelligence is itself a major consumer of intelligence end products.

Service to the policymaking entities of the Executive Branch is the measure of effectiveness in intelligence. Their needs for intelligence are without limit in principle and constantly growing in practice. They touch upon all areas of the globe and embrace most fields of human knowledge.

Effective service to intelligence consumers dictates a number of organizational principles:

a. The service or output end of intelligence must be highly diversified and relatively specialized to meet the diverse special needs of consumers. This demands specialized intelligence production support to departments, agencies, subcomponents, commands, etc.—size, scope, and level depending on the case. The Defense Intelligence Agency (DIA), INR, the Foreign Technology Division of the Air Force, and ERDA’s intelligence element are examples of the varying levels of support necessary to meet the specialized needs of departments.

b. In addition to expert and objective analysis from departmental intelligence agencies, the President and the NSC, along with other [Page 238] major consumers, need a source of intelligence that is independent of policy institutions, broadly competent, and available to support them directly, as a first priority. This principle justifies CIA’s role as a producer of finished intelligence.

c. To the extent practicable and consistent with security, the system must fully share information within itself. All production entities in a given subject area should share the same data and analysis.

d. The Community must have the means to come together to render a collective judgment or disciplined disagreement on vital intelligence issues. This is essentially what national estimates and other interagency products have been intended to do.

e. The Community should be structured so that collection is as responsive as possible to producer and consumer needs.

These principles lead naturally to some redundancy among intelligence production agencies. It is the belief of intelligence professionals and critics alike, however, that some overlap of substantive activity and competition in analytic judgment among intelligence production agencies is almost always healthy, necessary, and affordable. Of course, effective intelligence support to consumers depends on a great many considerations other than organizational structure. But the structure for producing intelligence within the US Government must reflect the above principles to be effective at all.

The criterion of efficiency in US intelligence is concerned with resources, the processes whereby they are employed, and their impact on production. After two decades of growth during the Cold War, concern for efficiency in Community-wide resource management is a comparatively recent phenomenon, accompanying a general skepticism about national security spending and a downturn over the last half-dozen years in real outlays for intelligence. Critical scrutiny of intelligence behavior by Government and the public has intensified the concern with efficiency in the last few years. In 1971 and 1976, two Presidential initiatives relating to Community authority structure were wholly or partly directed at improving the efficiency of Community resource management.10

Efficient management of intelligence resources proceeds in two connected dimensions. Existing resources must be optimally deployed and operated to meet existing intelligence needs according to a priority scheme that managers can base predictions on but that is still flexible. At the same time and largely by the same set of managers, decisions must be made as to what magnitude and mix of resources should be [Page 239] mobilized for the future. How these two kinds of decisions are reached in the Intelligence Community will be discussed in the next section (see p. 26).11 Some attempt to state first principles can help one to understand and judge present arrangements.

Intelligence resource management is largely a matter of managing collection and processing resources, because that is where most of the money and manpower are. Many collection assets are developed to gain broad access (e.g., a broad area imaging system) or potential access (e.g., an agent with a promising future or a regional clandestine capability). Broad access systems require extensive selection and processing for useful data, not all of which can be successfully processed. Potential access capabilities may or may not yield as anticipated. Moreover, intelligence is a form of conflict. Those managing intelligence resources are in reality doing battle with others in the world whose main aim in life is to frustrate the formers’ efforts. These conditions challenge the quest for efficiency and should induce a certain modesty in one’s goals.

In terms of structure, efficient management of current resources against current needs means giving control to the party with the incentive to seek and the capability to approximate the best allocation. To the extent intelligence collection and processing resources are expensive and scarce, relative to perceived needs, there is a tendency to centralize control. But other factors limit such centralization. Control may need to be contingent on changing conditions in the case of capabilities with varied application. The question thus arises of shifting control of certain national collection assets from the DCI in peace to military authorities in war. Some collection capabilities, such as tactical reconnaissance organic to combat forces, are justified solely for the contingency of war support to those forces and must be controlled and subordinated accordingly. Some degree of decentralization is reasonable in intelligence processing (e.g., photo interpretation, signals analysis, document translation) to achieve focus and promptness in the service of analytic users.

Assigning responsibility for programming future intelligence resources for efficient satisfaction of future needs is essentially a matter of deciding what should be traded off against what, to maximize what value. What should a given program element compete against in order to justify itself? And for what goals? Desirable multipurpose capabilities may have to compete simultaneously in several trade-off and value markets.

[Page 240]

For example, a major overhead reconnaissance system that supplies data to support national intelligence production and can also provide tactical intelligence support to military commanders ought to be weighed against other national intelligence assets, against other means of tactical intelligence support, and even against additional military forces. Whatever single or collegial authority manages national intelligence resources must be capable of making or assimilating sound judgments on such trade-offs.

This logic would insist that the DCI and the main departmental custodian of intelligence assets, DOD, should be running different, if somewhat overlapping, resource trade-off markets. The DCI should be expected, in the main, to trade off intelligence resources against other intelligence resources; the DOD, on the other hand, should generally be expected to trade off intelligence resources against military forces and support programs. Others hold, however, that the DOD is, in fact, a diversified market place in which multipurpose intelligence assets can be realistically assessed both in terms of comparative intelligence value and value to operating forces.

It should also be noted that the care and incentives applied to the trade-off of interests may vary with the size of the intelligence package relative to the money market in which it competes. The DCI market place is 100 percent intelligence; the DOD market place is less than 5 percent intelligence (see Figure 3). This, of course, does not preclude someone at an appropriate level in DOD from paying 100 percent attention to intelligence resources.

Any system for allocating intelligence resources must balance contending claims from many users of intermediate and final intelligence products with a central authority capable of resolving disputes in a rational manner. It must also balance rigorous assessment of costly initiatives with enough flexibility or permissiveness to permit initiatives to be pursued in the face of uncertainty.

[Omitted here is Section III: The Roles of the DCI.]

IV. Assessment

Section II of this report advanced three basic criteria for assessing the adequacy of intelligence management and authority structures:

a. Propriety of intelligence activities with respect to legal and political standards.

b. Effectiveness in the provision of needed intelligence to all Government users.

c. Efficiency in the use and mobilization of intelligence resources, particularly the expensive collection and processing resources.

This section attempts to summarize and assess the problems of the Community in meeting these criteria, to determine how DCI responsi[Page 241]bilities respecting them compare to his powers and Community structure, and to identify causes of problems that may not involve Community structure and authority. Specific options for changing Community structure and other innovations are treated in other portions of the response to PRM/NSC–11.

A. Propriety

The intelligence agencies of the US Government operate in conformity with the law of the land, the stipulations of Executive Order 11905, special restrictions laid down by the Attorney General in 1976, and other internal regulations and restrictions pertaining to propriety. Mechanisms for assuring proper behavior on the part of intelligence agencies are in place and operative. (Further discussion of this issue will be found in the report of the Attorney General’s Subcommittee dealing with PRM/NSC–11, Task 1.)12

But the situation is far from satisfactory. Many segments of US society external to the Intelligence Community entertain doubts as to the propriety of intelligence activities and the general trustworthiness of intelligence agencies. Internal to intelligence, many professionals suffer in some degree from an atmosphere deficient in confidence, trust, and respect for their chosen vocation. Managers and operators must, moreover, contend with uncertainties and conflicts that the new “ground rules” relating to propriety have presented to intelligence.

The ability of the DCI and other intelligence authorities to protect the security of intelligence sources and methods is severely limited by the lack of appropriate laws defining and protecting official secrecy in general. But such laws will certainly not be forthcoming unless the laws and regulations [Page 242] assuring the propriety of intelligence activities generate widespread confidence.

Alone, the DCI has little power to shape this larger environment. Much depends on the leadership of the President and other key officials of the Executive Branch, and on the reactions of the Congress, the press, and the public at large. The DCI has it within his power, however, to take constructive initiatives that could contribute to an environment in which the propriety of intelligence activities is assured, believed, and consistent with effective intelligence operations. He can take measures to rationalize and make more defensible the security and classification policies applied within intelligence. He can lead in the development and promulgation of professional standards relating to propriety applicable to the Community as a whole. With line command of CIA, he can be held accountable for its activities.

Assuring the propriety of intelligence activities is not primarily a matter of Community structure. It is mainly a matter of law and regulations, oversight, and professional ethics. But the DCI cannot fairly be held directly responsible for actions of agencies other than those he directly commands.

Although legal responsibility for the propriety of intelligence operations runs from the President down through the line managers of the several intelligence agencies, the DCI believes that the President, the Congress, and the public expect him to act as virtual guarantor of the propriety of all United States’ national foreign intelligence activities below the President. In the DCI’s view, his authorities to satisfy these expectations are now less than adequate, except in the case of CIA.

B. Effectiveness

Assessment of effectiveness in meeting the intelligence needs of all Government users applies basically to production of intelligence in the broad sense, that is, the production of intelligence reports and analyses, briefings, contributions to policy studies, and other forms of information support. This criterion also embraces warning and crisis support. (Assessment of wartime support to military decisionmakers is treated in the next subsection.)

Unfortunately, however, there are no absolute or simple means to measure such effectiveness. Policymakers dealing with an uncertain world cannot offer any comprehensive or fixed standard of intelligence “sufficiency.” Their needs for information and judgment are limited only by their capacity to absorb. Consumer surveys indicate that US intelligence organizations do fairly well at supplying current news and quick information support. In other areas, customers complain of deficiencies. Those who manage and evaluate US intelligence performance are obliged, therefore, to hear complaints, assess problem areas, and seek to improve where improvement seems feasible and important.

This brief treatment cannot explore all the problem areas identified by recent assessments of Community effectiveness in intelligence production, e.g., the recent NSC Semiannual Review.13 A summary list of major criticisms and self-criticisms of intelligence production activity is instructive, however:

a. Intelligence organizations at all levels do not understand consumer needs well and have poor tools for improving their understanding. Consumers, by the same token, only poorly appreciate the capabilities and limitations of intelligence. Producers and consumers are more isolated from each other than they should or need be.

[Page 243]

b. Mid- and long-range analysis and estimating is weak, unsophisticated, and generally under-emphasized. Major national estimates are frequently too unfocused, not directly pertinent to policy, and insufficiently sharp in judgment. Producers are not adept at integrating political, military, economic, and technical perspectives on problems that demand such integration.

c. Intelligence conduct and support of net assessment efforts are inadequate, although the main deficiencies in net assessment are not primarily due to deficiencies in intelligence.

d. Users who want fairly voluminous and detailed treatment of problems find many intelligence products dominated by summary judgments without supporting evidence, explicit reasoning, and uncertainty estimates. Users who want summary judgments find many products too voluminous with little judgment in them.

e. The Community is short of expert analytical personnel in some new areas of intelligence interest, e.g., political and economic aspects of nuclear proliferation. It also suffers from shortages of trained specialists in traditional areas, e.g., expert Russian linguists and area specialists.

f. ADP and other information support services are falling behind the explosion of information. To some degree, compartmentation impedes production. Analysts do not operate in an environment that assures they have all data available to the US Government pertinent to their problem.

g. Warning and crisis support responsibilities and arrangements for responses to them are insufficiently netted together to constitute a reliable and efficient system. Warning and crisis analysis is sometimes inadequate.

h. All production organizations are beset by fire-fighting demands that inhibit quality analysis on new problems. Much time is spent repackaging old material for new users and changed situations.

i. Too little attention is paid to seemingly mundane, but vital and difficult “bread and butter” analysis, e.g., maintaining and scrutinizing order-of-battle files, studying detailed aspects of the Soviet economy.

j. All analytic organizations are spread too thin. The situation is clearly critical in DIA, where vital national and departmental needs are inadequately met because DIA has too many masters, too broad and unstructured a mission, and too little management flexibility to assemble the quantity and quality of people needed for its job.

k. As a producing organization, CIA is insufficiently attentive to the needs of DOD in general.

l. Even the best analysts in any agency suffer from parochial views and failures in judgment.

There is no “right” judgment as to which complaints ought to be on this list or as to the degree of their validity. The important points are that:

a. these complaints are sincerely voiced and valid to some degree, and

b. they impinge on the entire environment of intelligence analysis and production.

Tackling these problems and improving the overall effectiveness of intelligence production, including the kind for which the DCI is [Page 244] uniquely responsible, does not rest mainly upon structural change or redistribution of management authority. Improvement requires problem recognition and steady management effort at all levels, in all producing agencies. As noted in the previous section, the basic structure of the intelligence production community is appropriate to the provision of effective support to policymakers. It permits departmental and non-departmental production; it permits the sharing of data and judgments; it permits interagency agreement or disagreement as required.

Efforts to improve intelligence production do, however, have some implications for Community structure, and changes in Community structure sought for other reasons could affect the quality of intelligence production. The following points bear on this issue:

a. The basic structure of the Intelligence Community must afford a close interaction between analytical activity and collection activity. The efficiency of both activities depends on it; present Community structure permits it; and the DCI can encourage it. Alternative structures might or might not be as conducive.

b. The Intelligence Community should have better means for executing its warning and crisis support responsibilities matched to the needs of those who must act on warning and deal with crises.

c. Some institutional framework or process outside intelligence is required to permit effective intelligence support of national net assessment activities.

d. Unless mooted by restructuring decisions, it would be desirable to resolve the apparent tension between the national intelligence responsibilities of the DCI’s NIO mechanism and those of his DDI within CIA.

e. To the extent that the DCI’s performance as a national intelligence producer depends upon the performance of departmental production entities, the DCI has a direct interest in the resource and management factors that shape their performance, as do their departmental superiors.

f. Over the years it has been frequently asserted that a significant increase in total Community resources given to analysis and production, at modest cost to collection and processing, could yield visible benefits in the quality of analytic products. While possibly valid, such assertions are probably unprovable. Such shifts probably would require stronger central authority over Community resources to achieve. In general, however, the keys to improving product quality are more in management attention, methodological innovation, and better producer-consumer dialogue than in gross resource or organizational shifts.

The DCI believes that the diversified structure of the national intelligence production Community as it exists today is generally sound. In his view, however, more effective national intelligence production requires enhancing the DCI’s authority to:

a. Task Community production elements outside CIA for purposes of national intelligence production;

[Page 245]

b. Task national collection assets that lie outside CIA and supply vital data for national production;

c. Control the program management of the major NFIP elements.

DOD disagrees with this view. Departmentally based collection and production elements are already fully responsive to DCI needs; no significant example of unresponsiveness to DCI needs has been adduced to support the need for change to his tasking authority. Moreover, DCI control of Defense intelligence programs could materially degrade their responsiveness to DOD needs, especially in wartime.14

C. Efficiency

Achieving the most cost-effective allocation of intelligence resources is mainly a matter of managing the most costly resources—those for collection and processing. Management proceeds in two time dimensions: the use of existing assets to meet current and near-term needs; and the development of capabilities for the future. In both dimensions the challenge is to provide necessary coverage of target problems and adequate service to consumers, while avoiding unnecessary effort and undesirable duplication.

1. Current Collection, Requirements, Priorities, and Tasking

Formal, centralized mechanisms for the guidance of major technical collection operations exist at the national level, under the DCI. These mechanisms—at the center of which are the DCI’s committees, COMIREX for imagery satellites and the SIGINT Committee for satellite and conventional SIGINT operations—are structured largely to fit the systems they guide. Their basic task is to assure that the needs of information users are optimally met by the capabilities of existing collection entities. Problems and frictions arise in the course of their business and concern about the responsiveness of these systems persists. These are manageable in the current structure of the Community; they could be eased by some and exacerbated by other structural changes. These collection guidance mechanisms are the middlemen of the intelligence process. Their function is not always understood by analysts or users, collectors, or outside critics. One needed improvement is to make the process better understood.

Human source collection lacks a formal centralized system of requirement and priority definition. The large and varied array of largely overt human source collectors who reside outside intelligence entities and provide a major portion of US foreign reporting properly resist inclusion in such a system. But some reliable means, even if voluntary, of tying them into the intelligence process must be achieved [Page 246] if clandestine resources are to be used no more than necessary. The DCI and his subordinates can argue for improvements on this front, but must depend on cooperation outside intelligence for real progress.

The Community lacks a centralized standing mechanism for orchestrating current collection requirements on an all-source basis. Such competence does exist in the collection management, analytical, and operational elements of the Community. Moreover, once one moves beyond the general guidance contained in such instruments as Key Intelligence Questions and DCID 1/2,15 current requirements management must be done largely in terms of the specific collection disciplines against specific problems. This does not necessarily lead to undesirable duplication because, while many assets may be targeted against the same problem, they yield different kinds of data on it and thereby produce the all-source picture needed by national intelligence.

It would still be desirable, however, to develop a somewhat more explicit communications network among the major entities of current collection management to give assurance that effective all-source allocation is taking place. Such a network could also provide the basis for developing current or near-term collection strategies against new collection problems. This entity should not be an additional layer of requirements management between analysts and collectors, but rather a horizontal connective tissue that would allow the DCI, NFIB, NIOs, and, where appropriate, consumers to know and influence easily what the total collection community is doing on a given problem.

It can be argued that difficulties here arise not so much from lack of DCI authority or from failings of Community structure, although the fragmented structure of the Community has helped to instill in each collection discipline a disposition to want to manage its own affairs with only general guidance. The main difficulties are defining problems and designing workable improvement mechanisms.

In the DCI’s view, however, enhanced DCI direct tasking or line authority over major national collection entities is essential to improve their responsiveness to all consumers and to eliminate the high degree of competitive overlap that generally exists. DOD, on the other hand, notes that the DCI already has direct tasking authority over the major national collection entities. Moreover, DOD believes there is no “high degree of competitive overlap” in this area. Finally, DOD does not understand how the DCI’s proposals will improve the responsiveness to DOD needs.16

[Page 247]

The Peace-War Problem: One problem of current collection management that has not been adequately addressed is that of transition from a peacetime, to a military crisis, and to a wartime posture in which major national collection systems, particularly overhead imagery and the total national SIGINT capability, must support military decision-making from the President down to the field commander. This problem has become more prominent as reconnaissance satellites have become more able to supply the intelligence needs of military commanders. It strongly influences debate over Community structure. One school of thought argues that the DCI should exercise control in peace, crisis, and war for reasons of central efficiency and rational resource allocation. Another argues that the Department of Defense must exercise control to provide reliable support to the command hierarchy.

As long as intelligence collection systems not organic to combat forces can provide such support, satisfactory definition and resolution of this problem will not arise from a priori principles. Careful and detailed study, planning, and exercising are required. A major difficulty is that we have not had much practical experience with the newly available array of collection assets in a major military crisis or large-scale conflict involving US military forces. The Vietnam War and the October 1973 Mid-East war offer some practical experience for study, however. For example, during the Vietnam War, all collection assets that were deemed useful to the theater commander were either directly tasked by him or were responsive to him, including SIGINT, SR–71s, U–2s, etc. Some general observations could help structure the problem and perhaps avoid errors:

a. Whoever runs or controls the national intelligence collection posture of the US in time of deep military crisis or war will have to use it not only to serve the needs of military decisionmaking, but also those of top-level political decisionmaking and the conduct of diplomacy. Military needs will likely dominate, but not to the exclusion of other needs.

b. For support of both military and non-military users of intelligence, the problem of collection management in war will be the same as in peace: marshalling many different collection systems to serve many different users. The major difference will be the volume and time-urgency of demands placed on these systems. Moreover, whoever (person or organization) has to run these systems in war needs to have had experience in tasking, line-operating, and resource-managing them in peace.

c. The primacy of military demands for intelligence support is not likely to be challenged in wartime by any collection management system. The basic problem is to assure that collection management systems geared for non-military needs in peacetime can shift rapidly [Page 248] to the needs of military support. Difficulties for any managing authority will arise from conflicts among different levels and kinds of military needs, and also from competing civilian demands. Establishing reasoned priorities will require system-specialized methods since specific systems can play very different roles in different military scenarios.

It may be possible to select among several distinguishable philosophies for managing this problem centrally:

a. In military crisis or wartime, the Secretary of Defense should manage, as a service of common concern, an integrated collection tasking system for all assets that can support military operations.

b. The DCI should manage the integrated collection tasking system as a service of common concern taking requirements as necessary from the military command hierarchy.

c. Management of some critical collection platforms or capabilities should be transferred entirely to Defense, depending on the system and the conflict scenario.

Any of these approaches could work, but it is unlikely that any of them would work well until we know in greater detail what intelligence collection management really means in a wartime context and build working mechanisms appropriate to that understanding.

2. Assembling Resources for the Future: Programming, Budgeting, and Other Management Powers

A foremost challenge of US intelligence management is to develop the best overall mix of capabilities needed to perform effectively at reasonable cost. This challenge is met in the year-to-year process of funding the major intelligence programs of the Community. How and how well this is done is central to the issues of Community structure, the powers of the DCI, and the powers of other senior intelligence resource managers, especially the Secretary of Defense.

It should be understood, however, that efficient resource management is more than a matter of structure and authority. The most fundamental problem of intelligence resource management is one that is common to other functional programs in government: there is no management science or comprehensive and orderly set of measures which may be applied to allocation of intelligence resources. We do not have a rigorous method for assessing the value of intelligence outputs and the relative contribution of inputs in terms that find general agreement and lead to confident decisions. This problem emerges from the very nature of the intelligence business:

a. There are no agreed objective measures of output value, since the limits of the needs of intelligence consumers cannot be readily defined, and there are no ways to quantify marginal satisfaction.

b. Except in discrete technical areas, the relative contribution of the many elements of the intelligence process cannot be quantified. [Page 249] These contributions are made through highly disaggregated and usually subjective processes within the heads of analysts and evaluators.

c. There is no explicit and comprehensive way to measure the value of, or loss implicit in, unsuccessful effort, i.e., experiments that fail, collection efforts that yield less than desired, analytic labors that do not produce. By its nature, intelligence necessitates much effort that proves unsuccessful.

These shortcomings of value measurement do not preclude reasoned judgment on what intelligence resources to assemble and how to use them. Such judgments are made all the time. In some aspects of intelligence management, they rest on quantifiable or explicit analysis, albeit with incomplete information. But more often they require successive aggregations of choices based on subjective judgment, experience, intuition, institutional preference, and a large measure of arbitrary decision.

Given the prominence of subjective judgment in this decisionmaking process, it naturally leads to concern about organization and authority structure. For, lacking a science of intelligence resource management that all parties practice in harmony, organizational structure is the most frequently used approach to establish the incentives and interests that more or less integrate all the disaggregated decisions that constitute resource management from top to bottom. Those responsible for such decisions at the top or center want great authority to structure incentives, give guidance and instruction, and review or correct lower echelon decisions. Those lower in the system typically want maximum independence. Those on the periphery or outside, but dependent on the system, want influence over the parts that interest them. This produces the familiar tension between centralizing and decentralizing forces.

Historically, US intelligence resource management has been largely decentralized, both in the Community as a whole and in the Department of Defense where most resources have resided. But pressures to centralize the process of managing those resources labeled national have been increasing for several years. Going beyond mere instruction, in 1976, Executive Order 11905 initiated a relatively centralized process, but one still based on a federated institutional structure and collegial decisionmaking below the President.

The record established in one year of operation under Executive Order 11905 is mixed. A consolidated NFIP and budget were produced. Through unprecedentedly extensive interactions among the members of the CFI, their staffs, and the NFIP program elements, issues were defined, studied, and in some cases resolved, in others deferred. Such issues were initially identified by the program managers, the Intelligence Community Staff (functioning as the CFI staff), OMB, and Con[Page 250]gress. Valuable experience was gained working with this process. A major step forward was taken in forcing programmatic decisions into a process wherein it is possible to justify program inputs in terms of intelligence value across the Community.

But this record was achieved only through a difficult struggle over procedure and substance. Key players, notably in the ICS and the Department of Defense, were at odds over the basic goals, as well as the rules, of this process. Executive Order 11905 strenghtened the incentives of the DCI’s ICS to give critical scrutiny to, and to influence the specific contents of, intelligence programs. At the same time, it enhanced DOD’s incentive, growing for some years, to place one central authority, DDI/ASD(I), astride all DOD intelligence equities. These authorities inevitably came into conflict as the former attempted to deal directly with program managers on program details and the latter resisted such attempts.

Although issues examined and decisions made were dealt with in terms of cross-program implications where they could be identified, the 1976 experience did not include a major new effort to accomplish cross-program trade-offs of the most basic sort. The process did not and probably could not come to grips with major shifts of funds among programs and across the elements of the intelligence process, i.e., collection, processing, and production. The CFI did not attempt to redefine the proper contents and scope of the NFIP—notably, which Defense intelligence program elements should be included and which excluded—according to a systematic examination of each element. It elected merely to accept the NFIP as it found it and to begin making resource decisions from there.

Although opinions differ as to how this record should be read, it is clear that the system worked to a considerable degree and has potential for improvement as more able and experienced staffing of the process is achieved. It is also clear, however, that this system will occasion continued tension and struggle among the participants, especially the ICS and the DOD, unless the goals and rules of the process are better defined.

Certainly, refinement of the programming and budgeting process created in Executive Order 11905 is one option for enhancing the integrity of national intelligence resource management in the future. It has the significant virtue of an evolutionary approach that builds on existing organizations and accumulated experience.

As it presently stands, however, the system obliges the DCI, as Chairman of the PRC(I), to proceed on most matters by persuasion and negotiation. This means that, to a great extent, initiative in the [Page 251] process lies with program elements and with outside critics.17 As a by-product, this structure places significant strain on the DCI in discharging his dual roles as head of CIA and as Community leader. At the same time, this system presents those department Secretaries having intelligence responsibilities, particularly the Secretary of Defense, with an awkward compromise of their statutory duty to manage and fund the programs under them.

Although formal responsibility for development of a rational and integrated NFIP rested in the PRC(I) as a collegial body, expectations have been generated that present to the DCI resource management tasks that extend beyond his pure management authority to fulfill.

Deciding on options for Community structure that will satisfy the criterion of efficient resource management requires that certain key issues be addressed:

a. How much emphasis should be placed on resource management efficiency in structuring US intelligence?

Many would assert management efficiency to be an obviously essential goal. But it is not obvious that satisfactory intelligence performance can be achieved at lower than present costs through better allocation of resources. One could argue that declining resources have already put intelligence overall in an inefficiently austere condition, where needed initiatives and improvements are too hard to justify and, hence, are not taken. But the fact that we cannot reasonably show whether particular intelligence efforts are essentially “efficient” should not deter pursuit of a resource allocation regime that emphasizes efficiency. Failure to display a workable system that strives for efficiency and shows results is likely to produce unwise, arbitrary decrements. Moreover, there are numerous specific areas where a rigorous regime can be expected to identify needless duplication and possible savings.

b. What is the promise of better analytical methods, or management science, for improving the efficiency of intelligence resource management?

It is doubtful that [Page 252] better analysis on resource issues can fully substitute for management authority in achieving more efficient intelligence allocations. Improvements can be reasonably expected from better, more consistent data on intelligence activities at all levels, from staffing the resource allocation processes of intelligence more expertly, and from applying more rigorous methods. But in the end, the results will depend considerably on the incentives of the players to cooperate; this depends in turn partly on the authority structure in which they operate. The Department of Defense, on the other hand, believes that only better analysis on resource issues can significantly assist management authority in achieving more efficient intelligence allocations, regardless of how that authority is structured.

c. What is the appropriate scope of the intelligence activities of the US Government that ought to be brought under a centralized intelligence management system?

In part, this question is: What activities should be included in the NFIP? But because intelligence is a shaded continuum of activities, some of which probably cannot be managed as intelligence per se, it is probably necessary to distinguish several kinds of intelligence for resource management purposes, and to accept some arbitrary dividing lines. Different management regimes should probably apply to each. For example, in one view, CIA, NSA, [less than 1 line not declassified] programs clearly represent a set of assets that are primarily national in nature. Consequently, in this view, they ought to be justified in relationship to each other and managed as national assets by a senior national intelligence authority. But their value in tactical support roles argues that they also be judged against other means of supplying such support and against additional military forces, a clear responsibility of the Defense Department. Other elements, such as departmental analytical organizations and many entities within the GDIP, could be justified primarily in departmental terms, but subject to review, criticism, and stimulation from the national, or DCI, arena because of their value or the extent of their contribution to the national effort. Given the rather coarse measures available, distinguishing departmental from national needs does not offer a confident means for delineating authority and responsibility. Whatever is of departmental interest is also of national interest. Yet a third set of intelligence resources would seem essentially tactical in character, e.g., assets organic to military combat units. Here the main interest of the national intelligence manager has always been and should be to gain the benefit of their existence in ways consistent with their mission but to assume no responsibility for their management.

d. How much centralizing authority is required for efficient resource management in the national intelligence structure?

Four kinds or levels of authority can readily be distinguished, each level capturing the previous one, except where explicitly compromised by the rules of the chosen management process:

1) defining future intelligence requirements and priorities; issuing broad guidance for planning and programming;

2) reviewing and vetoing Community programs and budgets;

3) controlling program and budget decisions;

4) exercising line management, including operational control and personnel authority.

[Page 253]

Given future uncertainties and long lead times, the DCI’s power to define requirements and priorities that apply to future intelligence capabilities is only a partial means of controlling resource allocations. Vetoes can stop but not initiate actions. Direct influence over programs and budgets is required to effect such control, either by unitary or by collegial decisionmaking methods. But even then some would argue that the uncertainties and inevitable disputes that must attend intelligence resource allocations for the future demand, in some cases, the authority necessary to direct subordinate organizations and to make their members willing supporters of the goals of the center. The DCI believes this to be the case. The DOD would argue that such strong central management authority over intelligence resources is not necessary and would be undesirable in that it would excessively concentrate authority and result in programs inadequately responsive to crucial consumers.

e. Should responsibility for intelligence resource management be combined with or separated from responsibility for national intelligence production?

Separation of resource management and intelligence production responsibilities might make it easier for the production manager to justify his resources and to concentrate on improving analytic performance. On the other hand, close interaction of these responsibilities is required if large expenditures on intelligence collection and processing are to be rationalized in terms of their ultimate contribution to intelligence output. If efficient allocation of intelligence resources means anything, it must mean an orderly relationship between inputs and outputs. The greater the separation of analysis and collection management responsibility, the more difficult it would be to assure such a relationship.

f. If there is to be a national intelligence manager, with special emphasis on and responsibility for resource management, who should he be and whom should he report to? Over what elements should he have line authority, collegial influence, or some advisory responsibility?

This, of course, is the bottom-line issue. It ranges beyond the instructed scope of this report. The relevant options and arguments will be addressed in other responses to PRM/NSC–11.

The DCI believes, however, that present arrangements give him responsibilities in intelligence resource management that are beyond his management authority to fulfill. Although formal responsibility for the contents of the NFIP rests with a collegial body, the PRC(I), as Chairman and as DCI he is expected by the President and the Congress to develop and take responsibility for an NFIP that is rigorously efficient and displays a close relationship between resource inputs and intelligence product outputs. In the DCI’s view, achieving the goals of efficient national intelligence resource management requires his having [Page 254] stronger central authority over national intelligence programming and budgeting decisions, and, in the case of key national programs, line authority as well.

DOD disagrees fundamentally with this DCI view and the comparable DCI views expressed on pages 58 and 60.18 In each case these views were added at DCI direction after the last Subcommittee meeting. In DOD’s opinion, these views confuse issues of tasking, resource control, and line authority; they attempt to justify added DCI responsibility on the purported ground that the DCI is expected to take such responsibility, an approach which begs the questions; and they fail to address other basic issues such as the compatibility of DCI’s envisoned role with effective wartime operations, adequate attention to the DCI’s primary responsibility to ensure reliable intelligence judgments, etc.

  1. Source: Central Intelligence Agency, Office of the Director of Central Intelligence, Job 97M00248R: Policy Files, Office Level and Above, Box 1, Folder 14: PRM 11—Intelligence Structure and Mission (Folder 3). Secret; Handle Via Talent-Keyhole Control System Only.
  2. Printed from a copy bearing a stamp that indicates that Turner signed the original.
  3. Secret; Handle Via Talent-Keyhole Control System Only.
  4. Not further identified.
  5. DOD does not concur in this Executive Summary. Note especially Page iv (FOREWORD) and Pages 58, 60, and 69. [Footnote is in the original. See footnotes 14 and 16 below and the last paragraph of the report.]
  6. Brackets are in the original.
  7. The Annex with Figures 1–10 is attached but not printed. Figure 1 is identical to Figure 1 in Document 41.
  8. The Secretary of Defense is also executive agent for US communications security, advised by the US Communications Security Board. [Footnote is in the original.]
  9. Figure 2 is identical to Figure 3 in Document 41.
  10. Reference to President Nixon’s November 1971 memorandum (see footnote 4, Document 35) and President Ford’s Executive Order 11905.
  11. In Section III.C, not printed.
  12. See footnote 3, Document 41.
  13. See footnote 6, Document 41.
  14. See last paragraph on Page 69. [Footnote is in the original. Reference to the last paragraph of the report.]
  15. DCI Directive 1/2, “Current U.S. Foreign Intelligence Requirements, Categories, and Priorities,” January 1, 1977.
  16. See last paragraph on Page 69. [Footnote is in the original.]
  17. DOD believes this statement to be untrue, since the system established by E.O. 11905 was designed to and has given the initiative to the DCI. [Footnote is in the original.]
  18. See footnotes 14 and 16 above.