ing, and negotiation. One way to think of the controlled war is that it is a limited general war. While this strikes most people, when they first hear about it, as a sort of academic absurdity, it does not take very long, just an hour or two, before one can make almost everybody take it very seriously. I predict that it will be taken seriously in both our country and the Soviet Union.

The President in his message to the Congress made the point that he wanted a command and control system adequate enough so that he could control the forces after the war has started. He specifically added that he wished to be able to use the force with discretion, even after the hostilities have broken out.

INADVERTENT WARS

Inadvertent wars: These are wars which are started as the result of accident, miscalculation, unauthorized behavior, a catalytic process, and so on.

There are many ways an inadvertent war could start. Some of these inadvertent wars might be much more destructive than the calculated wars simply because they have not been calculated and are therefore uncontrolled. Others are much less destructive because the planning, preparation, or tactics is poor or inappropriate. According to which kind occurs, the suggested civil defense program could perform with a wide range of effectiveness.

FAVORABLE MILITARY CIRCUMSTANCES

The next possibility on the chart is a situation with an unexpectedly favorable military circumstance. This possibility is almost always ignored.

Let me give you two examples: Consider a situation in which violence has broken out in Europe. A limited war might be raging, and, worse, it might look like it was getting out of control. I would assume we would then be more than willing to evacuate our cities. I could make the point more strongly. We could not stop the evacuation. People would leave the cities, and the only questions that would occur

are:

Is the evacuation reasonable or unreasonable, effective or ineffective? Is the population being evacuated to places of protection and safety or are they being evacuated to overcrowded facilities which are vulnerable or otherwise dangerous?

Such an evacuation could take place over a period of days or weeks or months. It is not an evacuation in which one is trying to outrun the ballistic missile. This is a strategic evacuation; we have warning. The warning was not supplied by an intelligence agency, it was supplied by the New York Times, the Post, the Evening Star, the Tribune-have I named all the proper newspapers? In other words, the events themselves gave us adequate warning.

There is another favorable military circumstance which is often ignored. This is the possibility that we may be much more competent than the enemy; we could also be militarily less competent; it could go either way.

Let me give an example: During the Korean war our fighters had a number of aerial combat duels with the fighters of the other side. For

every 1 of our planes that they shot down in these duels, we shot down 16 of theirs. If I had been discussing this problem of fighter duels before this committee in, say, 1949, I would not have had the nerve to even conjecture that we might be 16 times more potent than the other side; I would not have had the nerve to suggest a program that would work well if we happened to be 16 times better but wouldn't work very well if we were not.

This committee would rightly have said:

Mr. Kahn, you are indulging in wishful thinking. It cannot possibly be true that we are 16 times better. We might be interested in the 16 times worse case as insurance but not 16 times better.

Things like this do and can occur. When they occur, it can change the effect of the war by orders of magnitude. Trying to be in position to be able to exploit being militarily more efficient, if you happen to be, is different from the wishful thinking that would rely on being more efficient militarily.

I cannot discuss the possibilities in detail without getting into classified information, but I did want to mention the general principle, because it could be important.

CHINESE AND SMALL COUNTRY ATTACKS

Let me discuss item No. 4 on the chart. Fourth priority is still very important.

Every man in this room probably believes that by the late 1960's or the early 1970's the Chinese will have a strategic nuclear capability. It may be a rather effective one. However, unless we are careless, it is unlikely that they will be able to launch a surprise attack on the United States and prevent a reasonably massive retaliation, though they may be able to blackmail us under many circumstances. To the extent that we wish to be resistant to either accident, blackmail or irresponsibility by the Chinese, we must start putting in today the systems that we will need to handle this and other late 1960-early 1970 situations.

In spite of the consensus on this problem, it is very difficult to take seriously these potential military problems of the late 1960's or the early 1970's. I grappled with almost exactly the same problem in the mid-fifties. Everybody knew the ballistic missile was going to be an important element of military power in the early sixties. While almost nobody in the mid-fifties argued about this, it was still difficult at that time to motivate decision makers to spend large sums of money to defend against these hypothetical missile threats of the early 1960's. In the same way that the missile was a hypothetical and academic threat in 1956, the Chinese are a threat today. To the extent that we worry about China and even smaller countries, we must make preparations today. Many of the necessary measures have long leadtimes. It may be 5 or 10 years before they are really effective. To the extent that we are looking at 10-year systems, we must today prepare for 1971; 1971 certainly includes a Chinese military capability. Adequate civil defense is likely to take 10 years or so, so it would be well to get into the business.

I should say that the problem of protecting against Chinese attack is quite different from the problem of protecting against Soviet attack. you think of the Chinese as being in the same relation to the Rus

If

sians as NATO is to the United States, they are about 5 years behind, technologically, and much smaller in numbers; a system specifically designed to take advantage of Chinese weaknesses might be totally ineffective against the Soviets, but still effective against the Chinese. I believe it would be worthwhile to spend money on such hypothetical threats.

ARMS CONTROL

The arms control possibility is also important. As I mentioned earlier, we have some arms control today. Neither we nor, I believe, the Soviets are building the most destructive systems possible. There are militarily potent systems that could be built which are not being built simply because nobody wants to own them. If we had time, could discuss the most spectacular and ominous possibility, the doomsday machine, in detail. If one were simple-minded, one might believe both sides were building such devices today. Such devices would indeed render obsolete many suggested civil defense programs. But so far as I know, neither we nor they are building them.

Other types of arms control measures may be implemented, but even if we have elaborate signed agreements with the Soviets and others, that does not mean war cannot happen. Barring a world government, we will still be in the business of defending our country. The agreements may be deliberately broken, they may be abolished, or they may be accidentally violated.

With many arms limitation programs, some of the problems which look like they might become almost intolerable if we got into a fullfledged arms race are sharply alleviated. Many of the defense programs which are being considered today are completely compatible with many arms control measures, and indeed work better with the arms control measures than without them. They may be both militarily effective and tend to reinforce the arms control measures. Therefore, this possibility must be considered.

TECHNOLOGICAL BREAKTHROUGH

The next possibility is the technological breakthrough. One must admit that most of the technical people feel it is more likely that technological breakthroughs will hurt rather than help civil defense, but this is not inevitable. In particular, there are real possibilities in active defense. Of course, active defense today has serious problems; though they are perhaps not as serious as some think who discount it completely. I happen to think one can procure valuable levels of active defense, but with current technology and vigorous use of countermeasures by the attacker, active defense can have some serious deficiencies. However, these measures of active defense do tend to reinforce civil defense measures.

In addition, we are working hard at various systems, and one can have technological improvement in defense as well as offense. Some of these improvements cannot be predicted. When they occur we would like to be in a position to take advantage of them. If we have not started the necessary programs we cannot take advantage of these improvements even if they occur.

Not only does active defense complement fallout shelters, it may also complement blast shelters. Some of the earlier testimony stated

73266-61- -14

that blast shelters were not compatible with active defense. This may or may not be true, depending on what the system is.

RISE OF A HITLER

The last, and I hope, the least likely contingency on the chart is the rise of a Hitler. It is commonplace today to say that Khrushchev is not like Hitler. I think that is true. He is not as reckless, he is not as determined, not as malevolent.

Some, most notably A. J. P. Taylor, have said that Hitler was not like Hitler. In particular in the period 1933-43, Hitler and his government were relatively cautious as compared to the image we tend to remember. He acted much more rationally and prudently than most of us recall, and, in fact, came close to achieving his objectives. But even then he was an incredible threat to his warweary, peace-loving opposition. Today a Hitler type armed with thermonuclear bombs, particularly one who is crazy or realistically simulates being crazy, has the edge.

If somebody comes up to you and says, "One of us has to be responsible and it is not going to be me so it has to be you," he has a very effective bargaining advantage. If he can convince you that he is stark, staring mad and if he has enough destructive power, then deterrence alone doesn't work; you have to cave in or be annihilated.

It is difficult for Khrushchev to convince us that he is stark, staring mad because we can see he is thinking. I should add that I do not think we can convince him that we are stark staring mad.

It may happen that a leader will take over somewhere, sometime who either is or tries to act that role. Our only ability to handle him, the only way we can challenge him, is to have a method of putting our people in a place of safety so we can say, "Look, if you really are mad, we will fight it out."

Let me summarize this last point by the following: If somebody says, "I would rather be Red than dead," he is a coward, and I think very properly an object of contempt and scorn. If somebody says, "I would rather have everybody Red than everybody dead," he is making a sort of reasonable remark. It is a perfectly reasonable position to take. You may not agree with it, but I would agree with it.

I would rather have everybody Red than everybody dead, but we must not allow a situation to occur in which this last is the choice that is presented to us. We must always have an ability to say: "That is not the question. Our Nation and system will survive the worst you can do and we are willing to accept large casualties rather than surrender."

We cannot handle a Hitler on just resolve. It takes programs because he may well have more resolve than we have, particularly if he is not listening.

This finishes my rather cursory survey of the strategic background of civil defense. In spite of the superficial nature of my remarks, I have taken up more of the committee's time than I intended. I do want to finish by saying that the current program recommended by the administration seems to me the right way to get into the civil defense business. I would have some criticism of some aspects, but providing the program is pursued expeditiously, competently, and efficiently-provided they do what they say they want to do then I have no hesitation in endorsing the recommendation.

CIVIL DEFENSE RESEARCH

Mr. ROBACK. Mr. Kahn, would you comment specifically about the request for research in that program?

Mr. KAHN. Well, we suggested a program in 1958 as a result of the 1957 RAND study. The "we" is the people who did the study not the corporation, and the suggestions can be found in RAND Report RM 2206-RC. There is a revised summary of these suggestions in appendix IV of my book, "On Thermonuclear War."

The report recommended that about $300 million be spent for a capability in being. The recommended measures were much like the current program. In other words, identification, counting, and labeling of existing fallout protection, and then making plans to use this protection. In addition, we suggested about $200 million for what we called research, development, and planning for the preliminary phases of a spectrum of programs. Here the administration's recommendations are somewhat less than ours.

We called some of this research, development, and planning program a part of the so-called preattack mobilization base. The idea was that in this country we theoretically have the capability to procure a civil defense program in just a year. We have the physical capability in the construction and the other industries that are necessary to actually carry through a massive program running between. say, $50 and $100 billion in 1 year. It can be done not only because the necessary industries have the capacity, but because these industries are very deployable.

It

One can give a construction company a blueprint on Friday for a new shelter, and on Monday this company would start to build it. does not need a long leadtime like, for example, the aircraft or the traditional munitions industries.

One of the inadequacies of the current proposal is that it does not look enough at the various alternative programs and possibilities. The people involved may do all these things after they get started, but many of the possibilities are not explicitly recognized in the proposals set forth, or do not receive sufficient emphasis. I think the actual bill has a research budget for $7 to $15 million, something like that.

Mr. ROBACK. That is for research; there are other items in the budget, perhaps, that could be classified in that broad area.

Mr. KAHN. Yes. That is certainly much larger than we have had in the past by quite a bit. But it is not as large as we need.

Let me give some examples. It used to cost between $100 and $200 million to build a prototype fighter aircraft, $100 to $200 million just to do the research and development so that we could see what we might be able to procure. We used to develop five airplanes for every one we bought. In other words, we actually put out $500 million to $1 billion in order to see what five of these objects looked like; we then looked at them, picked one, and threw four away.

So far as I know, nobody criticized that process. Rather, some said, "Why don't you build more prototypes. Why are you so skimpy on research and development?"

Today the Department of Defense spends between $8 and $10 billion on research and development, if you allocate all the costs to research and development that really belong there. (This last number is different from official costs.) In other words, they spend one out of every