to students to study hard languages if they were willing to commit their services to the government for a period afterwards. That lapsed with the passage of time————

Senator DEWINE. That's an excellent idea.

Mr. HITZ [continuing]. But it may be one way to deal with it. Senator DEWINE. Let me ask any of the panel members this question. We've heard testimony recently from Secretary Armitage and Secretary Wolfowitz that both of them believe that one of the biggest problems with our intelligence and analysis is that the agencies strive for consensus and don't always encourage dissemination of dissonant views.

I wonder if you agree or disagree with that.

Mr. HAMILTON. I agree with it. I think there's tremendous pressure on the analysts to reach a consensus. The broader the consensus, the more general and sweeping the conclusion is, the less valuable it is. I think you want to encourage competition among your analysts and have majority and minority reports. If I were the President of the United States, I'd want some sense of the bureaucratic view, both for and against a given course of action.

I want competition among analysts. I think the minority needs to be protected. I think they need to be encouraged to speak up. Senator DEWINE. Anybody else want to comment?

General ODOM. I would briefly say I would strongly encourage Members to go try to find a situation, look at it directly, where intelligence is used to make decisions. You are exactly right about reaching a consensus on the national intelligence estimates, the national documents that are produced by the National Intelligence Council. Even within some agencies like DIA, et cetera, there are consensus kinds of problems.

It's been my experience that almost no decision is made on that kind of intelligence. I'm looking for somebody to show me how an NIE caused a policy to change. The major advantage of an NIE is it makes the Intelligence Community share its information base. If we're talking about a problem of sharing for counterterrorism, that's with the FBI and people outside. We have made great progress in getting a common intelligence information base in the community.

If you go where decisions are made, you will find the user deeply involved in the intelligence process himself. To give you an example. When we were building the M-1 tank in the Army, the thickness of the frontal glaces of a T-80 tank was very important. Ten millimeters of difference would have changed it from a 120 gun to a 105 gun. It was very easy. I had the responsibility of deciding which way to go on that, and I was a big 105 gun fan, but I finally said okay, I'm going to do this way. That pushed about $18 billion around, to up the gun to 120.

Now that's where intelligence really plays. I never understood the big brouhaha of not predicting the end of the Soviet Union. It doesn't make any difference. Tell me what we would have done differently had we known it was going to end. Or, the issue of how much the Soviets are spending on defense. Nobody cared. It was a political game the CIA and the rest of the Intelligence Community played for the press. We didn't spend our money to buy based on

how many rubles or dollars they spent. I wanted to know-my army wanted to know how many tanks we had to kill.

So what we should be tested against is whether we counted the number of tanks and knew what it took to kill them. And that's where intelligence plays. And I'd go to the State Department and find where intelligence really helps them in their negotiations. And that's the test. It's not whether there's competitive analysis or these other kinds of thing.

So I have a kind of practical, hands-on view. If you want an answer to that that's clear, you're going to have to go and absolutely look and make people show you the causal linkage.

Senator DEWINE. Good. Well, let me thank all three of you. I appreciate your testimony. I think it's been very helpful for our Committee. Thank you, Mr. Chairman.

Chairman GRAHAM. Thank you, Senator. I want to thank the panel, but I also want to thank our four lead questioners today. I think they have raised very fundamental issues in a way that will be very helpful to us as we move towards our final recommendations.

We are now going to move into the five-minute questioning period, and I have the good fortune of being the first of those. I'd like to turn in my first five minutes to issues of personnel. My own experience has been that you can have the best organizational chart, but if you don't have the people who can make the system work you're not likely to be very successful. Parenthetically, I hope Steve Spurrier is not in that position.

Mr. Hitz, you talked about a number of things that might be done to enhance the personnel within the intelligence agencies, starting with the idea of more internships to introduce young people. There have been a series of other suggestions made which I'd just like to mention for purposes of stimulating your brief comments on those and then adding to your list of ideas.

There was a proposal in an issue of Foreign Affairs earlier this year of setting up an intelligence reserve corps somewhat analogous to the military reserves, which could be activated at a time of particular need, a proposal for an intelligence reserve officer training corps. Maybe that comes close to your suggestion that you just made about restoring the scholarships, the defense scholarships that used to be made.

Also, the issue in terms of getting more people in intelligence who can speak the languages and understand the cultures of the diverse areas of the world in which we're now trying to operate, that we ought to have a more aggressive recruitment policy towards non-traditional CIA and other intelligence agencies, specifically among the Arab-American population.

Could you comment on those and any other suggestions to improve the personnel quality? Let me just mention one other that I heard on a recent visit to our station in Cairo. That was that, almost like in the public schools, the agent's compensation is heavily affected by their years in service, and once they reach a certain level it's hard to show much more advancement, and at that point agents are tempted to leave their case responsibilities and move into an administrative position where they have more economic up

side, just as teachers tend to leave the classroom and go into administration for the same reason.

Is there some way that we ought to be re-looking at the compensation strategy for intelligence agencies to keep our best people doing what they're best at and being rewarded for that superior service?

Mr. HITZ. There are a lot of good points that you've made, Mr. Chairman, and also ways in which I think aspects of them have been tried in the past but not pushed enough. I think the intelligence reserve corps has been attempted after a fashion, and I think you would agree with me, sir, that presently, with deficiencies in a number of language skills and experience levels, a lot of old boys have been brought back. They are in effect an intelligence reserve corps right now. They've gotten recent retirees to agree to come back and put their shoulder to the wheel after 9/11. That's an informal way of getting at that, but it certainly makes

some sense.

On the question of trying to recruit operatives from America's Arab-American community to go back to the Middle East, I thought there was a very thoughtful unsigned editorial in the Washington Times a day or so ago making the point that that's not necessarily the answer to a problem, because natives of that region want to be dealt with by a real American, so to speak. I'm putting that in quotation marks. They want somebody that may not speak the language with absolute proficiency but is good enough.

And a perfect example of that is the Popov case in the cold war era, when the Agency sent a recent emigre from the Soviet Union to first deal with Popov when he was coming over to us from Vienna, and Popov didn't trust him. He knew what Stalin was able to do with emigres and wouldn't fiddle with it. So I think there has to be a sort of a balance on that part.

On the management side, as opposed to why promote a person who is a first-rate case officer into management and lose those talents, again that's a debate that's gone on long in the Intelligence Community and lots of people frankly, as case officers, don't want to be and aren't very good as managers. I think George Kiesevalter-I don't think I'm taking his name in vain-always considered himself to be a case officer till the day he died, even though he was a very senior intelligence officer.

But I want to be clear on the point. I don't think all of the talent for doing this work is going to come out necessarily of our finest universities. There are all kinds of skills that work in this area and, if I can be permitted just a personal anecdote to finish up with, my first boss in the Agency was a Nisei. He was a person who lived long enough ago in Los Angeles to be able to dive for abalone in Los Angeles Harbor. It's been a long time since that happened.

Well, he was sent to camp. His family were interned, and in 1942 the U.S. Army came along and recruited him to the counterintelligence corps. He served in Japan until the end of the war, joined CIA, for 30 years practicing his trade all over the world-an extraordinary person. Why there was no lingering bitterness, I have no idea. But he was absolutely first-rate. So this country has got an awfully broad range of talents out there to draw from.

Chairman GRAHAM. Thank you, Mr. Hitz. I'm sorry my time has expired, but when I rotate back I'd like to ask Mr. Hamilton and General Odom for their comments on the issue of personnel.

Senator Shelby.

Vice Chairman SHELBY. Thank you. There's been a lot of debate recently in the Senate about the creation of the Department of Homeland Security that you are all familiar with. Perhaps the most important part, at least I think so, of the Homeland Security legislation that's being debated is its provisions dealing with information analysis. Senator Graham and I have been involved, with Senator Lieberman and others, dealing with that particular piece. If homeland security analysts are to occupy, if they are to occupy a unique position and have a unique perspective in that they would have access both, General, to domestic and foreign intelligence, and to information on homeland vulnerabilities, it seems that it would be important to give them the capability for this sort of deep information access to intelligence agencies, in other words, if they needed it or thought they needed it.

What are your thoughts here regarding the role of homeland security in the Intelligence Community and should it have an analytical component? And, without an analytical component, what would it be?

General.

General ODOM. It needs an analytic component. It doesn't just need one; it needs many. It needs a central one and then it will need distributed ones. Let me offer the model of mainframe central processing versus distributed processing in computers. When you had one big mainframe computer, with slow computers, dumb terminals, people got backed up in a queue. When we came to microprocessors, we could have a lot of people processing simultaneously. Intelligence analysis is done the same way, and many users need analysis and you'd like to have the analysis with them. The collection can be more centralized. When you have the organization, homeland security, deciding where it wants those, I think it should have them near decisionmaking points, then each of those analytic capabilities need to be able to tap into NSA, the national imagery agency, to CIA's clandestine service, and to what I would see as the national counterintelligence service. The FBI will never give information out. A national counterintelligence organization would be an intelligence organization, not an arrest organization. Therefore they would have an incentive. They want their stuff used. They're not doing it for themsevles.

There's no reason that that analytic center can't draw on the whole community. The model that I think it's easy to look at right now in that regard is how the State Department works. You have INR at State, which is the general central point, but within negotiation you can have an analytic center supporting anything that's going on.

That seems to me to be sort of a straightforward, easy organizational issue to deal with.

Vice Chairman SHELBY. I agree with you, General. Congressman Hamilton?

Mr. HAMILTON. Senator Shelby, I don't know that I understand all that well the President's proposal on the Department of Home

land Security with regard to intelligence. As I understand it, the Department of Homeland Security would not be a collecting agency at all. They would not get the raw data. They would get the conclusions from the CIA and the FBI.

Vice Chairman SHELBY. The whole community.

Mr. HAMILTON. The whole community, and then they would use that to assess threats and for the primary purpose of protecting the infrastructure. And the President's talked about it being a clearinghouse, and I think George Tenet has said the Department would be a consumer of intelligence.

I'm a little skeptical of all of that. I don't think the conclusions of the Intelligence Community, if handed to the Department of Homeland Security, will satisfy the Department of Homeland Security folks. They're going to want to know, well, where did this come from and how sure are you of this information.

Vice Chairman SHELBY. They're going to vet it, in other words. Mr. HAMILTON. I think that's correct.

Vice Chairman SHELBY. And I think they should.

Mr. HAMILTON. You were mentioning they have to have some capability to examine information analysis. I think I agree with that. Now, if you do that, then what is the relationship between these three organizations-DHS and CIA and FBI? I'm not sure I know the answer to all of that, but I am a little skeptical of this idea that the raw data would not be available.

I also understand that the President would have the authority to provide the raw data, under certain circumstances, and that might work satisfactorily, but I'm reasonably sure if I were running the Department of Homeland Security and I got the conclusions from the intelligence agencies-those conclusions tend to be fairly broad and sweeping and vague at times-that I wouldn't be satisfied with it.

So I sense the Senate is quite correct in looking at this pretty carefully.

Vice Chairman SHELBY. I agree with you, Senator Graham and I. That was not the President's first proposal but we've worked out a proposal now.

My time is up. Can Mr. Hitz say anything, Mr. Chairman?

Mr. HITZ. Well, I find myself in agreement with Congressman Hamilton on that, and the only question I was going to put to you, Senator Shelby, was what is the recourse of the homeland security analyst if he finds that the intelligence he's provided is not up to snuff? This is the point. Does he have to go back to the President and knock on his door to get it right.

Vice Chairman SHELBY. He should be able to go right back and task someone what is this, what does this mean and so forth. That's what Senator Graham and I have been proposing for six months, I guess it is.

Thank you, Mr. Chairman.

Chairman GRAHAM. Thank you, Senator. The next questioners, in order, will be Congressman Roemer, Senator Roberts, Senator Feinstein. That will complete the first round. Congressman Roe

mer.

Mr. ROEMER. Thank you, Mr. Chairman. I want to welcome all three of you here today. You have really been extremely helpful to