This is the United States, Africa, Europe, South America, et cetera. Here is the Army network worldwide, Navy network worldwide, and Air Force network worldwide. It is quite apparent that by providing integrating lines at particular points one system can serve as a backup, an emergency backup for another system, thereby greatly increasing our combat strength. And this is exactly one of the pri mary purposes of the Defense Communications Agency, and a purpose it is already meeting.

Mr. HARDY. Then, in other words, it is tying in these different communications systems together?

Secretary MCNAMARA. Systems which were incompatible and nonintegrated.

Mr. HARDY. But what is to keep it from restricting or directing the use of the entire system in a way that it might jeopardize the particular service requirements?

Secretary MCNAMARA. The Defense Communications Agency reports directly to the Joint Chiefs of Staff, it is responsive to that body and quite clearly would not, in any way, act contrary to the direction it is given from the Joint Chiefs.

Mr. HARDY. But if a Defense Communications Agency decided that it wanted to utilize Army's system for some particular service which might even embrace an Air Force or a Navy requirement, couldn't it blanket the Army's system so that the Army couldn't carry out its mission?

Secretary MCNAMARA. No, it cannot.

Mr. BLANDFORD. That is a decision of the unified commanderSecretary MCNAMARA. Perhaps you would like me to go back. I will leave these with you.

Mr. HARDY. Thank you.

The question of a unified commander would, of course, apply in a particular area, but unless it also applied at the departmental level, you would be in trouble or could be in trouble with it; couldn't you?

Secretary MCNAMARA. No, sir. The unified commander would have the authority to dictate the use of a communications system and what Defense Communications Agency has done is provide that unified commander with a capability that did not exist previously by developing the integrating, unifying mechanical links necessary for one system to supply a reserve capacity to another.

Mr. BLANDFORD. Isn't DCA, however, Mr. Secretary, an overall management control system, more in the field of coordination? I mean, to compare DCA with DSA, they are two different things entirely.

Secretary MCNAMARA. I was simply responding to the question how have these agencies increased the combat capability and there is one very good illustration of it. I will be happy to give others.

Mr. HARDY. I think that might be a good illustration. I think most of us can see the need for at least having interchange ability in use of different systems to meet combat requirements.

However, I do have some question in my own mind as to whether the same type of reasoning would apply with respect to some of the other agencies that are established. I don't know that we need to pursue that.

Secretary MCNAMARA. Mr. Chairman, I would be happy to respond to that. I would specifically like to discuss NSA in that respect, but because the matter is highly classified, we could defer that.

Mr. HARDY. I think NSA we probably better leave out, and I think all of us would place it in a slightly different category from some of the others.

Secretary MCNAMARA. All right.

The next one I would discuss is DASA. This is perhaps also in a separate category, because quite clearly, it evolved from the Manhattan project

Mr. HARDY, But I don't think that relates to the kind of a problem we are principally concerned with.

Secretary MCNAMARA. That leaves only two more, DIA and DSA. Mr. HARDY. That leaves those two insofar as the present is concerned.

Secretary MCNAMARA. Yes, insofar as this particular question is concerned, of combat efficiency and the extent to which the agencies have contributed to it.

With respect to DIA, while I was preparing for this committee hearing, I came across a series of replies from the unified commanders that had been sent in during the month of April in response to a request made to those unified commanders by General Carroll, Director of the Defense Intelligence Agency, early in April.

At that time he had cabled the unified commanders asking for their criticisms upon the activities of his agency to date, and their recommendations for changes in its operations in the future.

Since it was established on October 1, 1961, this was a normal and expected action, I think.

In any case, the replies were interesting and I would like to quote from one or two of them, particularly as it relates to your question, has the Defense Intelligence Agency added to the combat capabilities of the Defense Department.

Here is what CINCNORAD said, the commander in chief of the North American Defense Command:

Since the forming of the Defense Intelligence Agency and its assumption of the functions of the U.S. Air Force indications system, the value of messages exchanged has been greatly enhanced by the addition of the other agencies and services.

He is referring there to the addition of the intelligence that he had not previously received, but now receives because of the input from collection sources that were and are part of the Navy and Army and other agencies.

"This is extremely helpful in prompt access"-this is the point I want to make "of the enemy's posture and intentions."

CINCAL, the commander in Alaska of the Unified Command had essentially the same point to make, and the commander in chief of our European forces, CINCEUR, stated:

The cable summaries in particular are considered excellent and timely.
This again is the point CINCNORAD made:

The indications system which now includes this headquarters provides a steadily improving service for the rapid evaluation of critical global activities.

These commanders are particularly concerned about information any place in the world that bears on the potential for combat and therefore the requirement for readiness of their forces.

And finally, General Powers, of the Strategic Air Command, stated:

I would like to emphasize the outstanding relationship that exists between DIA and SAC's warning center. The manner in which the Defense Intelligence Agency indications responds to our queries has been most commendable.

I am talking here only of the extent to which DIA has contributed currency and an expansion of information relating to the indications or warning centers of the unified commanders. Their cables contain many other bits of information of the same type, but I extracted only these, and I think it bears specifically on the question you raised. Mr. HARDY. Those are helpful. I am glad to have them on this question of the Defense Intelligence Agency. Undoubtedly-I had some discussion of it, Mr. Secretary, over at CINCEUR just before Easter and we explored that to some extent, but I am still a little bit uncertain as to exactly what that Agency does and going through your statement you said, among things, to integrate and validate all DOD intelligence requirements and produce DOD estimates.

Now, how do you distinguish your DOD intelligence requirements from a service requirement, for instance, or are they distinguishable? Secretary MCNAMARA. Yes, I believe they are.

The service requirement, because of its support for training, as you pointed out, and its very, very broad responsibility at least in my opinion, for logistical planning and support starting with research and development and proceeding into procurement and production management has specific requirements for technical intelligence, the characterization of weapons systems being developed elsewhere in the world. This is clearly a service requirement as you used the term.

On the other hand, the Joint Chiefs, being responsible under my direction, and the direction of the President, for the command of the unified commanders, have clear responsibilities and clear requirements for operational type intelligence, and this is a Chief's requirement, and in that sense, a Defense Department requirement that is different from a service requirement.

The services, as you well know, are not responsible for the command of the unified commanders.

Mr. HARDY. No, but I am thinking, pursuing this just a little bit further, in connection with a unified commander's requirement for intelligence which might be of a somewhat different-well, it would be specialized in a particular theater of operations-he might have some particular needs.

Has he got to get that strictly through DOD?

Secretary MCNAMARA. No, sir. I would distinguish between the requirements for the Chiefs as commanders of the unified command on the one hand, versus the requirements of the military departments for intelligence, the kind I mentioned a moment ago, from the requirements of the unified commanders.

The requirements of the unified commanders to the extent that they can be satisfied by the collection efforts of the forces under the unified commander are directed by the unified commander to that end.

The unified commander will, however, send a copy of his requirement into the Defense Intelligence Agency so that that may be maintained in the Central Registry of Intelligence Data so that if another unified commander has a requirement for similar information, it will be known that it is already available.

Mr. HARDY. Well, that, of course, should be helpful, but then you get into a question of estimates as well as requirements, and the next one that interests me was your statement: "* * * and produce DOD estimates."

I don't want to get into any detail on this.

Secretary MCNAMARA. I think I can respond in an unclassified fashion to start with on this.

Mr. HARDY. Back on the subject of DOD estimates, would that then become, if we are talking now of estimates of enemy capabilities, for instance, in a particular area, would that then become the official estimate which all of the services would have to utilize?

Secretary MCNAMARA. I want to talk about the situation as it will be following completion of the implementation of the present plan for the Defense Intelligence Agency, rather than as it is today, because the Agency is in the process of transformation today, and I think you would be more interested in what we are heading toward, rather than exactly where we are. At the moment DIA is not equipped to prepare such estimates.

It has roughly 500 people. It is moving toward approximately 3,000 people. I will discuss the way it will operate when it has 3,000 people and if that leaves questions in your mind, we can come back to the present.

Mr. HARDY. One thing that we are talking about here now is looking toward the future under your DIA setup, if you are going to have official DOD estimates produced by DIA, then I presume that they will be the final determiner as to an estimate of capability, enemy capability, in any particular field?

Secretary MCNAMARA. They will be, but I want to emphasize that. they will be based upon a thorough analysis of the data available from all sources which data were not always available in the past to the estimators.

This has been one of the problems of the past, because the data were compartmentalized, it was exceeding difficult for estimators working on problems that had defensewide implications to have access to all of the data.

Mr. HARDY. You are familiar, I am pretty sure, with what prompted this question of mine, because it hasn't been so long ago that we had some variations in estimates of enemy capabilities by the different services, and by the CIA, and they all acknowledged that they based their estimates on-they all had available to them the same information, but they were miles apart in their final estimates.

But my question is simply this: Are we going to wind up with one estimate, are the Air Force and Army going to be told: "This is the official estimate, you can gear your thinking on it to this"?

Secretary MCNAMARA. No, I think not, in the way in which you have stated it.

The estimate will be an estimate based upon all available data, but only rarely will the available data be such as to permit only one

conclusion and therefore, the estimate must show more than a single conclusion.

It must show a range of potential, in other words, because our data are not sufficiently clear or comprehensive to permit a single value as opposed to a range of values.

Mr. HARDY. You remember a very significant one we had not too long ago, where the Air Force had one estimate, way up high, and the Army had one way down low, and the official CIA estimate just hit a happy medium?

Now, if we are going to have that kind of a system, I am a little bit worried about it, because if we come up with-if this is going to be an official estimate and maybe it would be in complete contrast to the Air Force evaluation of the information that it has

Secretary MCNAMARA. No. I think in that particular case, if I am thinking of the same case you are

Mr. HARDY. I expect we are thinking about the same one.

Secretary MCNAMARA. Each of the parties did not have access to all of the other information.

Mr. HARDY. Then the hearing record is wrong and it ought to be corrected, because they testified to it.

Secretary MCNAMARA. Certain of the individuals, because of the normal complexities of administration, operating in the estimating progress, were not even known by name to other individuals, also participating in it.

Also, I think it is entirely understandable that because of the close relationship to requirements that, consciously or unconsciously, estimates are very likely to be affected by that relationship.

Mr. HARDY. I can understand that.

Mr. BATES. For instance?

Secretary MCNAMARA. And that moving the responsibility for the estimating, if you will, from the military departments to an agency directly responsible to the Joint Chiefs of Staff, will minimize that difficulty. And therefore, the estimates

Mr. HARDY. It might minimize the difficulty and at the same time increase the possibility of error.

Secretary MCNAMARA. Quite the contrary.

The possibility of error is introduced by the possibility of conscious or unconscious bias and by reducing the potential for bias, the possibility for error is reduced.

Mr. BATES. But in the particular case to which I think Mr. Hardy also has reference, we were very careful on many occasions to use the term "The Intelligence Community," indicating to this committee that these various individuals testifying before us had all common. information and that the only difference was the question of what they would draw in conclusions from that information furnished.

Now, are you indicating they didn't all have that information? Secretary MCNAMARA. Well, the basic information they had was common-let me turn it around the other way-much of the information they had was common, but there were particular bits of information of use in evaluating and interpreting it that were not common. Mr. BATES. That certainly was not the impression that I drew from those hearings, and I think Mr. Hardy has also indicated that we all thought that they all had the same information.