40. Memorandum From the Comptroller of the Central Intelligence Agency (Taylor) to Director of Central Intelligence Turner1


  • Further Thoughts on PRM–11 Issues

1. During your session with us the other day2 on our paper on the options available under PRM 11,3 you asked several fundamental questions about the nature of the authorities we thought you needed to do your job. Following the meeting we spent some additional time talking with Mr. [name not declassified] about his related efforts and got from him some further insight into your questions. As I understand it, you have divided the question of authorities into three basic areas: those dealing with the ability to task the Community to do your bidding, those which involve enhanced budgetary authority, and those which deal with line authority. Mr. [name not declassified] suggested that a paper dealing with some of the issues inherent in these concepts might be helpful to you, and we offer the following.

2. We see the problem similarly but would argue that line authority and tasking are in fact one and the same thing. Tasking in our view is a subset of line authority and not an independent, stand-alone variable. But let us take you through our reasoning. To do that we will talk about the tasking question first, then line authority, and then budgetary authority.

3. There is a good deal of confusion surrounding the concept of tasking. Let us elaborate on two different views as to what tasking means. You are today under the 1947 Act4 charged with pulling together intelligence from all the various producers and collectors in the Intelligence Community and integrating it for the consideration of policy makers. You thus have the legal authority to ask for the product of all Community components and to ask collectors to collect certain kinds of information. In the case of CIA you cannot only ask that the information be collected but direct that that task be accomplished; and if it is [Page 164] not done to your satisfaction, you are in a position to change that. With respect to the other collection entities in the Community, however, all you really can do at the present time is ask. The mechanisms available to you to ask the Community to contribute on problems basically consists of the DCI committee structure, which is a vehicle for the articulation to others of your requirements and needs. You have at the present time all the authority you need to ask through these mechanisms that work be done. What you lack is the ability to enforce those requests, i.e., to ensure that requests are met in whatever timeframe is appropriate. Because the DCI’s role in the Government is important and cannot simply be ignored, the collegial committee process resting essentially on the consent of the participants often works, although rarely as crisply and efficiently as is ideally possible. In short, tasking should mean not only the ability to ask for information but the ability to ensure that you get it. The former you have; the latter you lack. It is line authority over the Community components involved which would give you the latter. It is for this reason that we would argue that the concept of tasking is in fact integral to the concept of line authority.

4. What would it mean if you had the ability to task the Intelligence Community to answer to your needs in the way we have suggested above? To answer this question, we picked the management problem you mentioned at our recent meeting—how far does your present staff authority have to be augmented to gain effective control over NSA? Or, as you put it, how much of the existing dotted line between the DCI and NSA would have to be inked in to give the DCI the necessary authority to manage NSA? As the solid line representing the authority of the DCI over NSA increasingly replaced the dotted line of staff guidance, the solid line that now extends from the Secretary of Defense to NSA must be correspondingly broken to reflect the DCI’s increased authority. Thus, we have a twofold problem. Any increase in the DCI’s ability to direct or manage NSA must be accompanied by a proportionate dimunition of the power Defense now holds over NSA. The force of logic influences us to state that you cannot both have line control and not have it; or to answer that there is no such thing as a little line control. It seems to be indivisible. The owner of the heaviest solid line calls the shots and establishes the ground rules for the other players. But let’s look at what powers the DCI now has to make NSA responsive to his direction and, then, enumerate what we think he must have to carry out his responsibilities. Some place between the powers the DCI now has over NSA and those we believe he should have, the border between the dotted staff line and the solid command line will be crossed.

5. The DCI is faced with two distinct management situations as he strives to carry out his responsibilities to the President. He must manage [Page 165] the diverse resources of the Intelligence Community toward the fulfillment of long-term national intelligence objectives and, on an ad hoc basis, he must be able to utilize these same resources to support the President in crisis situations. Crisis management puts a different stress upon management capabilities than do the work-a-day problems he faces that are not time urgent. Therefore, we should examine the need for increasing the DCI’s authority over NSA in both situations.

6. The DCI’s present ability to “direct” NSA is made up of three separate but obviously interrelated approaches. First, is his unquestioned authority to promulgate broad collection guidelines in the form of Key Intelligence Questions5 and other more specific national intelligence requirements. Secondly, he can, through the budgetary process, veto some NSA activities, change the pace of ongoing activities where progress is closely related to dollar limits, and he can encourage new initiatives by providing funds to encourage NSA-originated initiatives. Lastly, he can selectively use the force of his personality and his access to the President to bring a recalcitrant Agency into line. The promulgation of broad guidelines and the selective use of special access to higher authority are textbook mechanisms that are traditionally used by staff personnel to get the job done. Strong budgetary power is one of the keystones of line authority. Thus, the DCI today has the usual staff powers plus one of the essential elements of line authority. The other essential element of line authority is the capability to reward directly those who effectively carry out their assigned responsibilities and to punish just as directly, those who do not. The rewards and punishment element of line authority encompasses the ability to hire and fire personnel, to have unrestricted access to all parts of your subordinate organizations and to evaluate the performance of subordinates against the tasking they have been given by their chief.

7. How can the DCI use the tools he now has to direct NSA? If the DCI decides that the needs of national intelligence require more economic reporting and less military reporting from NSA, he can issue collection guidance requirements that “task” collection systems to increase their economic reporting. No one will question the DCI’s right to issue collection guidance and if the Director, NSA, and the Secretary of Defense agree with the DCI, the necessary adjustments will be made. If they do not agree, the collection ratio between military and economic coverage will remain more or less the same. The DCI, in the course of time, will find out that NSA is not responding to his tasking. At this point, he can wait for the next budget cycle, or he can appeal to the [Page 166] President to tell the Secretary of Defense to honor the DCI’s request to collect more economic intelligence. The DCI may decide this is really not a proper problem to bring to the President’s attention, and the DCI will then have to pick up his budget stick. He will soon discover, however, that he cannot find an effective place within NSA to use the budget stick to cause a shift from military to economic reporting. The same collection systems serve both reporting categories. This is also true of the processing mechanism. There is nothing to veto; no unit to deprive of funds and no slots he can refuse to fund. The choice may be to cripple the ability of NSA to collect intelligence at all or to let them continue their practice of selectively responding to DCI collection guidance. Thus, all the tools in the DCI’s inventory can prove to be ineffective in the most elemental test of his powers—the bringing of collection systems into line with national intelligence needs. He can, of course, given a world of “limitless” resources, give NSA the extra funds they would need to expand their overall collection capability in general and thereby increase economic coverage, but that is rarely a real option.

8. As would be expected, a crisis situation which calls for a rapid shifting of collection emphasis to support the President’s need for the rapid formulation of foreign policy initiatives shows even more clearly the handicaps the DCI must overcome to orchestrate collection and production resources. With his present powers, the DCI can order his human source collection mechanism to respond, and the DDO will move immediately to redirect its collection assets. The DCI’s Human Resources Committee is not even relevant to this process. In fact, most DDO collectors have only the vaguest notion of this Committee. One leg of the DCI’s collection triad has responded immediately to his direction. The other two legs of the triad, represented by SIGINT and reconnaissance systems, are not as easy to redirect in crisis situations. The assets of the CCP and the NRP are owned by the Secretary of Defense. If the Secretary perceives the crisis with the same level of priority as does the DCI and if he agrees with the “trade off” involved with any redirection of collection assets, then all will go relatively well. The DCI’s SIGINT and COMIREX Committees will work their collegial magic, and the technical collection systems will slowly swing around to focus on the crisis. The DCI has effectively matched Community resources with national intelligence needs. Or has he? Maybe the Secretary of Defense played the key role. For what would have happened if the Secretary had not agreed with the importance of the crisis and refused to go along with the collection trade off that would occur if his CCP and NRP assets were moved from their standing collection responsibilities? In that case the collegial committee process would not work as harmoniously. The inevitable compromise process would [Page 167] set in with its attendant delays, and the DCI’s effectiveness in focusing Community resources on a crisis area would not be as impressive. In essence, the DCI can do anything with the resources of the CCP and the NRP that the Secretary of Defense lets him do. In short, you are not in a position to make trade off collection decisions because it is the Director of NSA who must do the balancing between your needs and those of the components or organizations which he serves most directly in a command sense. Giving you line authority over the two other parts of the Community as suggested in our earlier paper, the NRO and the CCP, would put you and not the Director of NSA in the position of weighing the competing intelligence and military needs. And it is, of course, for this reason that Defense will most strenuously argue with proposals to remove these components from the Department.

9. If Defense controls the resources of the CCP and the NRP, and if the DCI has essentially the same staff guidance relationship to both, why is it that the reconnaissance assets seem more responsive to DCI guidance than do the COMINT collectors? Of the two technical DCI resource tasking committees, COMIREX works more effectively through the collegial process than does the SIGINT Committee. In fact the COMIREX Committee has often been held up as a model for the other collegial committees to emulate. The answer to this is rather simple. COMIREX assets are limited by technology to collecting data within a narrow spectrum of national intelligence needs. Moreover there is a great degree of Community acceptance of COMIREX targets. Photographs seldom help us to understand the political process of a target nation. They are of limited use against economic targets. Pictures do not tell us much about basic research or the pre-prototype stages of weapon systems developments. Overhead photography, however, is a remarkably effective collector against targets of military significance. The importance of the military targets covered by COMIREX assets is understood and accepted. The limitations of this technology to collect against other targets is also understood. Therefore, the COMIREX Committee meets in an atmosphere of relative harmony with limited possibilities for significant “trade off” arguments. Discounting telemetry and ELINT collectors which enjoy the same relative target commonality as photographic satellites, SIGINT Committee COMINT assets have the technological potential for collecting against all national intelligence requirements. The probability of disagreement is correspondingly broad and the likelihood of agreement without extensive compromise and long delays is improbable. There are, of course, other differences between the collection programs represented by the COMIREX and SIGINT Committees but they are not as fundamental. CIA’s historical role as the technological leader in satellite photography [Page 168] and the physical location of important program managers within CIA and under the line control of the DCI also improve the DCI’s ability to match COMIREX resources against intelligence needs. Since the DCI and the Secretary of Defense have fewer disagreements over photographic, telemetry or ELINT targets, DCI requirement guidance is more effective and the need for DCI line control to match resources against requirements is not as critical. The opposite is true with COMINT collectors. Without real line authority there is no way of making sure COMINT collection will be guided by your perception of national intelligence needs.

10. In our meeting on Wednesday,6 there was a good deal of discussion about what it would mean to you if you were in fact responsible for not only the CIA but also the CCP and NRO in a line management sense. Questions were raised as to whether the management job was so large that your ability to carry out substantive responsibilities would be seriously compromised by the time required to be spent on managerial duties. Basically, we think this is somewhat of a red herring. There are many Government officers who have responsibility and authority over programs larger than that which would emerge if CIA, NRO, and CCP were combined. Further, we think there is a plausible argument that line control over those other two organizations would in fact make your Community resource and other responsibilities easier to handle than they now are. You would then have the more manageable task of making your organization responsive. The collegial Community management process developed over the years and further enshrined in E.O. 11905 is, because it is built on a Presidential order which cannot modify statutory responsibilities, necessarily a cumbersome and time-consuming apparatus. If your real authorities were clearer, it can be argued that the managerial task you would have would in fact be simpler. In the last analysis, the question is really one of delegation. In combining the three organizations, it would be important to build an effective staff organization which enabled you to focus the organization on the questions you wished addressed, and it would be necessary to build procedures to ensure that the large questions in which you wanted to be involved were brought to your attention but the others were handled by subordinate elements. In other words, the way in which you delegated your authority and indeed your management style would probably be as critical to the question of whether or not you had time for substance as would the size of the organization you would be managing.

[Page 169]

11. We have talked about tasking and about line authority and argued that one is but a subset of the other. What of the various proposals to give you expanded budgetary authority in the Intelligence Community without line authority? To answer this question let us lay out the two different models which as far as we are aware have been attempted in the Government and give you a sense of what each would mean and how it would work.

12. The first of these is essentially reflected in the existing IC staff arrangement. You were given under the Executive Order last year what is essentially a staff responsibility to the President, not unlike that of OMB, to advise him on the appropriate mix and disposition of resources within the Intelligence Community. The authority you have been given under the Executive Order is limited to making a recommendation on the proper allocation of resources. If a decision is made, it must be the President’s or the Secretary of Defense’s, and you have no legal responsibility for the defense of the program before the Congress or the execution of it once it is approved except in the case of CIA. The ability to recommend actions on the budget is a powerful tool although it has, as we pointed out in our previous paper, limitations.

13. Another model which has been suggested would involve appropriation of funds to you for that portion of the Intelligence Community for which you wish to have a budget responsibility. These funds would be directly apportioned by you among the various programs which make up the Community. In such an arrangement, you would theoretically be given the power to run an effective budget process, to raise issues and decisions with the President, and to defend the program before the Congress, and to execute the budget as you saw fit within any limitations imposed by outsiders. There is precedence for such an arrangement. The so-called poverty program set up by President Johnson in the Office of Economic Opportunity (OEO) in the early 1960s in fact was designed to function in this manner. The basic concept was that funds would be appropriated to the Director of OEO but that the responsibility for actually conducting programs would generally be delegated to other existing departments of the Government. The Director OEO would shape the budget in accordance with his priorities, defend it before Congress, but leave the day-to-day management of, for example, manpower training programs, to someone else, in this case the Secretary of Labor. By the late 1960s when OEO’s appropriation was about $2 billion, about $1 billion was appropriated to the Director of OEO but transferred thereafter by him to the Secretary of Labor for the conduct of manpower programs. The idea had a good deal of appeal but in fact was largely judged a failure. (The whole program was thought by many to be a failure; here we are discussing only this peculiar budgetary arrangement.) The fact was that the Secretary of [Page 170] Labor had vastly more influence over the budget which legally was to be prepared by the Director OEO than one would have thought, given the original concept established in law. This happened for very human reasons, and we doubt that were you, for example, to have a similar responsibility with respect to NSA today the situation would be much different. Because the Secretary of Labor operated the manpower programs, because he had good Congressional contacts, because OMB turned to him for advice on these programs rather than to Director OEO, because even the White House turned to the Secretary of Labor instead of the Director OEO for advice, OEO found itself essentially rubber stamping what the Secretary of Labor had already agreed to do with others. In fact OEO was never able to get the Labor Department to concentrate on the areas it thought were important in the manpower program area. Doubtless there have been other analogous approaches to this problem in previous times although we personally are not aware of any of significant size. In this particular case, after a fair amount of backbiting between OEO and the Department of Labor and a growing recognition by everyone that little was gained by appropriating the money to OEO, a decision was eventually made to appropriate the funds for these programs directly to the Department of Labor. No one knew the difference.

14. A net assessment of that experience is that it was not worth the trouble. In addition, our previous paper suggests to you what we believe are some of the other important limitations of the budgetary tool alone.7 Also, we explained our view that your assumption of a more far-reaching budgetary role within the Community would lead to demands from others in the Community, particularly the Department of Defense, that you separate yourself from CIA. This in turn would require that you take at least the production apparatus out of CIA so that you would be able to fulfill your most fundamental intelligence responsibility, thereby raising the question of whether CIA without the production apparatus could continue to exist. Perhaps more fundamental from your point of view, however, you would be left with line command over essentially only the production apparatus and faced with a “residual” CIA (i.e., the CIA today minus the DDI and the NIOs) which reported around you in a line command sense to either the NSC or the President. We doubt that the budgetary authorities you would gain would compensate for the losses sustained through your separation from the CIA and the end runs which would, we think, occur with some regularity.

[Page 171]

15. Thus, we return to the argument posed in the earlier paper, that it is line command over the essential elements of the Community which you need to do the job which others expect you to do. In this connection, we might explore one further option. If it is clear that it is line command over the Intelligence Community which should be established, is it necessarily clear that it is the DCI who should exercise this authority? Why not, for example, make the CIA responsible to the Secretary of Defense and establish a position of Intelligence Community czar within the Department of Defense? This solution is conceptually the same as giving line authority over the Intelligence Community to the DCI, and it would solve the Community management problem analyzed in our earlier paper. This arrangement would have the great strength of not provoking an enormous battle with the Department of Defense. In avoiding that battle, however, we believe that you would create several others which would be equally, if not more, difficult. Perhaps the only issue on which almost any Congressman (from conservative to liberal) will agree regarding CIA is that it must be independent of the policy making apparatus of the Government. A proposal to include CIA within the Department of Defense would we think provoke a very strong and negative reaction. In a large study of this question last year,8 we pursued this option at some length and considered whether there might not be some arrangement which would accommodate to those concerns. We considered, for example, the idea that the DCI might be established as a statutory official within the Department of Defense responsible for the management of all intelligence including CIA and that in an arrangement similar to the Joint Chiefs of Staff, he would be able to see the President independently on substantive or other matters of concern. The concept has a certain appeal and it would in fact solve a number of managerial concerns. In the last analysis, however, we believe that the approach is flawed. Customers in departments and agencies other than Department of Defense would see such a move as a threat to the support which they now receive. This would be particularly true in the case of the Department of State. We doubt that a CIA lodged in the Department of Defense could attract the quality of personnel it needs to do its job, primarily because the intelligence profession must always be viewed within Defense as support to the Department’s primary responsibility to guarantee the nation’s military security. Despite legal provisions guaranteeing the independence of the Director in a substantive sense from the Secretary of Defense, we doubt such independence could in fact be guaranteed or that others would believe that it could.

[Page 172]

16. We hope that this paper is helpful to you. We would be happy to either pursue some of these ideas further on paper or explore them with you in another meeting. There may also be practical problems on which you may like short papers. One of these might be concerned with the management structure you might need to exercise line control over CIA, NSA and the NRO.

James H. Taylor9
  1. Source: Central Intelligence Agency, Office of the Director of Central Intelligence, Job 97M00248R: Policy Files, Office Level and Above, Box 1, Folder 14: Intelligence Structure and Mission (Folder 3). Secret. Sent through the DDCI. Neither Taylor nor the DDCI initialed the memorandum.
  2. Not further identified.
  3. Presumably a reference to Document 38.
  4. Reference to the National Security Act of 1947, P.L. 80–253.
  5. William Colby, DCI from September 1973 until January 1976, established the Key Intelligence Questions in 1973, which were to provide the entire intelligence community with intelligence targets critical to policymaking.
  6. May 4. Presumably the meeting with DCI Turner.
  7. Document 38.
  8. Not further identified.
  9. Printed from a copy that bears this typed signature.