scratch a magnificent team of scientists and engineers and an array of first-class laboratories. In 1945 we saw the scientific team being disbanded and the research facilities transferred to other auspices. These circumstances were the cause of much thought and debate, which produced such appraisals of the needs and deficiencies of American science as Vannevar Bush's Science: The Endless Frontier (1), which still makes good reading, and the well-known Steelman Report for President Truman.

In 1951, under the stress of the Korean War and the possibility of another mobilization of the scientific community, President Truman created a Science Advisory Committee in the Office of Defense Mobilization, to provide the President with independent advice on scientific matters, particularly those of defense significance. This was the first significant step toward moving scientific advisers into the White House.

Again in 1957, in the traumatic aftermath of Sputnik, there was a call for a general reappraisal of where we stood in our national science policies and goals and the adequacy of government science organization. It resulted in the appointment of the first full-time Special Assistant to the President for Science and Technology, James Killian. Simultaneously, the President's Science Advisory Committee was established in the White House.

Five years later, in 1962, after another study and review of the White House science organization, it was decided to establish the Office of Science and Technology (OST) to provide permanent staff resources to the President for dealing with matters involving scientific and technological considerations.

Now, 6 years after the OST was created, we are again at the crossroads of introspection and examination of our national science policy and the organization needed to formulate it and carry it out. It is my feeling that, as before, changes will be made and, I hope, for the better.

Having spent the past 5 years at the bench of U.S. science policy development, I would like to review with you some of the issues and problems as I see them, with some thoughts as to the future.

The main problem areas have been perceptively indentified by the OECD (Organization for Economic Cooperation and Development) examination of U.S. science policy: academic science and the universities, the role of the government in industrial research, some of the social impacts of U.S. science policy, and the adequacy of the mechanisms in the U.S. government for dealing with these problem areas that is, to make science do for the intellectual and material welfare of the American people all the things we think it can do and that we claim for it.

FEDERAL SUPPORT OF ACADEMIC SCIENCE

With regard to academic science and the universities, the central questions are: first, how to provide training of high quality for enough scientists and engineers of the right kinds; second, how to maintain vigor and creativity in the basic reserach establishment; and third, how to set priorities and determine the relative emphasis given to different research areas.

Concern over maintaining the vigor and quality of academic science is not a new phenomenon in 1968. At each of the 5- to 6-year steps in the evolution of the government science structure to which I referred there was a peaking of public concern about the state of American science. I venture to say that this recurrent, if not continuing, concern will remain with us for the foreseeable future.

You will recall the pronouncements after World War II about the sad state of fundamental research in the United States and our unhealthy dependence on European scientific discoveries for the development of the U.S. arsenal of new weapons, most notably the atomic bomb. The case for substantially strengthening the ties of government to university science was eloquently stated in Bush's Science: The Endless Frontier, in July 1945.

The year 1950 finally saw the creation of the National Science Foundation, after long debate (and a Presidential veto) over how independent this so-called independent government agency should be.

Again in 1957, with the advent of Sputnik, there was a resurgence of concern and interest in academic research, particularly in terms of the production of new scientists and engineers with advanced training, partly out of fear that the rapidly increasing output of scientists and engineers in the U.S.S.R. would pose a long-term threat to U.S. security.

When I entered the White House scene I was confronted with the issue of academic science in a somewhat different form. The explosive growth of

government support of science in the 1950's and early 1960's had left in its wake a new array of problems of science administration, both in the universities and in the government. There was evidence of congressional dissatisfaction with what they believed to be lack of tightness and tidiness of federal controls over these large expenditures. This was, in part, based on misunderstanding of the nature and form of federal support. The question of overhead rates charged by the universities was raised, apparently from a confusion of overhead and profits-a question, I must admit, that has not been swept away (witness the recent Mansfield amendment to limit indirect costs paid under research grants).

Members of Congress had become acutely conscious that university science had entered the big league of congressional interests. The House established a Select Committee to Investigate Expenditures for Research Programs. The House Science and Astronautics Committee moved to establish a permanent Subcommittee on Science, Research, and Development. The Congress debated ways of strengthening congressional mechanisms for obtaining information and advice on scientific and technological fields.

Today there are again mutterings about a "crisis of confidence" in federal support of academic research. As in the past, this appears to be another moment of introspection, calling for self-renewal and readjustment of our sights to see clearly the goals ahead.

The current problems of academic science appear to have their origins in the budget stringencies growing out of the Vietnam war. But, in my view, the budget squeeze is only one symptom of a more general difficulty. It has brought to the surface the latent, unresolved problems which must inevitably be dealt with directly. I refer to such issues as the support of research through project grants versus broad institutional grants, and how to wed the cultivation of the best science to the training of enough scientists, broadly distributed throughout the country. Even more fundamental and serious is the failure of the university and the scientific community to effectively communicate its values, its purposes, and its contributions to the public and to the lawmakers.

Although these and other problems connected with federal support of aca demic research could be alleviated by increased funds, it is unlikely that there will ever be enough funds to satisfy all legitimate requests. In short, we had better face up to the underlying problems. With the increasing size of the academic science establishment and the proliferation in the number of promising avenues of research, failure to develop a coherent approach could bring even greater pain at a later date should the enterprise suffer a loss of public confidence and support.

As we move to unite the knots in the existing policies and arrangements for federal support of academic research, we must, I believe, find a healthy accommodation between a laissez-faire system and centralized control. Forces in the direction of detailed planning of basic research and graduate education have been resisted because of the inherent unpredictability of the results of scientific research and the needs of our society, and because of difficulties in estimating the long-term national requirements for scientists and engineers. While I agree that central direction of federal support of academic science is not conducive to the maintenance of vigorous, high-quality academic research, neither is chaos. Nor can we entirely capitulate to the vested self-interests in subgroups of the scientific community that will resist any change or trade-off that they believe would threaten their interests. What I am suggesting is a better articulated framework for federal support of science and an indicative plan, looking a few years into the future, that will provide a general guide for the allocation of funds, at least, and provide a necessary degree of stability and predictability for future planning by the universities and the government agencies involved.

There are many ways in which the federal support of academic science can be carried out-different mixes of government agencies and universities, as well as different mechanisms for the support of research and for the support of graduate training. What may make sense at one level of consideration may not make sense at another level. I believe that we do not know enough about the interrelationships of the various parts of the scientific enterprise, the various types of support, and the various objects of support to construct a comprehensive blueprint or plan for proceeding. However, I am convinced that we need to sharpen our analytical tools and capabilities, identify and acquire the necessary data, devise working hypotheses, and be willing to experiment with subaggre

gates of the system so that we will be in a steadily improving position to deal effectively with the entire set of problems. And we will have to move further toward the generation of broad-scale, long-range plans.

This problem can be likened to the continued, healthy growth of a delicate and complex organism. It is not analogous to the stages of human growth from childhood to adolescence, adulthood, and old age-and I hope the latter is not in sight. Rather, it is more like the problems of medicine and physiology, where we understand some of the pieces, but where our understandings are isolated and do not explain the functioning of the organism as a whole. The pieces I refer to are basic research, education, applications, and their coupling to technology. Our job is to make the organism healthier-not just its component organs.

IMPACT OF THE NATIONAL SCIENTIFIC EFFORT ON SOCIAL AND ECONOMIC PROGRESS

A question just as fundamental as that posed by academic science concerns the coupling between the national scientific effort and our country's social and economic progress.

During the past 2 years I have been deeply involved in two studies of the so-called "technological gap" issue. One was carried out at my direction within the U.S. government. The other was undertaken by the Organization for Economic Cooperation and Development, in Paris, in preparation for the OECD Ministers of Science meeting last March. The analysis of technological disparities among industrially advanced countries and their basic causes makes it clear that the United States does better than most countries in harnessing science and technology to economic and social progress.

Europeans tend to regard the technological gap as a new phenomenon, and in doing so overlook the long history of U.S. preoccupation with industrial growth. There was considerable debate on this issue among the "founding fathers" after the American Revolution. According to George Soule, in his Economic Forces in American History (2), Thomas Jefferson favored a nation of landowners, principally engaged in farming, to avoid the poverty and exploitation of the working classes which accompanied the beginning of the industrial revolution in England. In this debate, Alexander Hamilton's differing views prevailed, and Hamilton should be credited for the strategy America used to overcome its technological dependence on Europe. The basic elements of this strategy, reflected in Hamilton's "Report on the Subject of Manufactures," submitted to Congress in 1791 when he was Secretary of the Treasury, were the protection of "infant industries." Perhaps more importantly, he urged the promotion of immigration of technologically skilled manpower and the encouragement of capital inflow from abroad. Since that early time, numerous European observers, from Alexis de Tocqueville on, have commented on the positive American attitudes toward technological change and the introduction of new technology in industry.

Federal policies and programs aimed at stimulating American industrial technology, directly or indirectly, are simply the modern version of Hamilton's infant-industry argument. What makes it more difficult now is that we are trying to strike a balance between a national view and a world view. In Hamilton's day, government policies toward satisfying the needs of 10 million people couldn't upset any international apple-carts. Today, the currency and the military power of the United States are dominant forces in the world of commerce and international order. Government policies with short-term domestic objectives can, through international repercussions, have longer-term adverse effects on both the international and the domestic scene-witness the run on the dollar due to what others regard as overexpansion of domestic programs.

Despite the acknowledged American success in most fields of science and technology, there are some industrial people in the United States who feel that the effect of our emphasis on academic science has been to draw off too many talented people from other creative functions of society, such as industrial engineering and innovation. They feel, for example, that contemporary engineering training is not appropriate to the conduct of engineering in industry-although others dispute this allegation.

Another difference of view concerns the degree of coupling of the results of government-financed research and development, particularly in the military and space areas, with the needs of civilian industry. Again, some will allege that the federally financed research and development effort has siphoned off or otherwise deprived industry of creative talents that could be put to use in commercial R&D that it has undesirably inflated the salaries of scientists and engineers employed in nongovernmental commercial business.

With regard to the "spin-off" question, almost everyone who has look at the evidence agrees that the exploratory development programs of the Defense Department and NASA have enabled us to press the technical arts to their farthest limits. Some of our best, newest, and most thriving industries have their roots in this government-financed industrial activity. Our favorable export balances largely reflect export of products born of intensive technological effort in industrial sectors such as sophisticated electronics, computers, and aircraft, which owe much to the stimulation of federal support. But we remain unclear about the diverse effects of federal support of research and development in the aerospace and electronics industries on our industrial base as a whole. The dual fads of enthusiasm and complaint about "spin-off" are not likely to be dissipated without further intensive study of the complicated cause-and-effect relationships observed over a considerable period.

One could advance the hypothesis that, in a sense, technology per se is to industrial innovation what science is to the generation of new technology-that the general search for new technology is the industrial equivalent of basic research. In my view, though, there is an important difference between science and undirected technology. The best basic research is directed at carefully conceived problems framed by the investigator. I question whether government programs aimed at the general development of new technology would be effective in advancing civilian-directed industry. On the other hand, technology which is a product of industrial R&D contracts aimed at satisfying the exacting requirements of military and space systems-requirements which go well beyond civilian needs and which set concrete performance goals for the product-is more likely to be applicable. We have just witnessed a magnificent demonstration of this point in the Apollo 8 mission, in which the huge Saturn V had to perform flawlessly on its first flight, as did computers, a far-flung communications and tracking system, and a complex human organization-not to mention the astronauts themselves. This distinction between general technological development and the achievement of measurable goals was not well brought out in the OECD studies and seems to have been blurred in some foreign debates on government programs for strengthening the technological base of industry-say, the computer industry. It should also be observed that technological development is enormously expensive as compared to most basic research, and that, although the Department of Defense, NASA, and the Atomic Energy Commission, among others, do support exploratory development efforts, development can normally be supported as a federal expenditure only where it is aimed at specific needs that the public, expressing itself through the Congress, regards as commensurate with the investment. This, of course, raises the $64 question of the appropriate role of the U.S. government in supporting or promoting research and development for the prime purpose of advancing industrial development and growth for civilian ends. Although there is general satisfaction with the health of American industry and its rate of technological innovation, there are some areas (environmental pollution is an example) where the ordinary market rewards do not stimulate industry to develop at an adequate pace the new products and processes needed by the general public. In the field of air pollution there is a lack of strong private incentives, and the urgent need for improvements in pollution-abatement technology have called for government leadership.

The leadership for pollution abatement, as present, lies in the government through its role in standard setting and in supporting science and technology to demonstrate what can be done, and how. It will be the essential job of industry to find cheaper and improved ways of applying the new technology. In the longer run, this is bound to lead to an increase in private activity and a lessening of the financial burden on the government.

Government standard-setting has been an important indirect means of stimulating industrial incentives and competition to improve the quality of products affecting other aspects of the general health and welfare. Through food and drug legislation we have been able to maintain high standards of drug safety and efficacy. Automobile safety standards are another example. Within the interval of a few years we have seen a dramatic shift in the attitude of the automobile industry from a phobia about mentioning automobile safety in advertising to today's promotion of safety features in meeting industry competition.

A great deal more work needs to be done to sharpen the tool of standardsetting as a means of introducing product improvement and change in particular sectors of industry. Standards must be based on sound scientific evidence,

which must be continuously reexamined and improved. They must be set with regard to the industry's economic, managerial, and physical ability to respond. If there is careful regard for the sensitive interaction between incentives for innovation and requirements for protection, government standard-setting can exert a strong motive force for private investment.

At the same time, when looking to industry one should be realistic about the size of the market incentives needed to stimulate private investment. The expected market demand or dollar sales volume must be large in relation to the R & D investment that can be justified to produce the improved product.

In some areas it will be necessary for the government to directly stimulate industrial innovation in important but lagging industries. In some cases it can do this as a consumer of a large number of units (such as military housing) or through partial or full support of research, development, and demonstration projects.

The question of whether there is need for an overall governmental policy for strengthening civilian technology generally has been held in abeyance. In the absence of a direct interest in a specific industry or social problem, the government has not adopted, as a general approach, direct measures for encouraging industrial invention and innovation per se. Patent and tax incentives have long provided indirect encouragement for private investment. With the exception of selected industries somehow identified with the public interest (for example, agriculture, atomic energy, the supersonic transport, water desalting, pollution abatement, and a few others), the government has not subsidized civilian-oriented industrial research. Further measures to stimulate technological innovation may be needed, but there appears to be no need for an across-theboard, direct approach by the federal government. Nonetheless, we should watch closely the experience of Canada, the United Kingdom, and France in their new programs for subsidizing the development of new civilian technology, to see whether experiments along this line are indicated for the United States.

GOVERNMENT SCIENCE ORGANIZATION

Thus far I have dealt with some of the issues that academic science, industrial research, and social needs pose for U.S. science policy. The fourth question asked by the OECD examiners is even more elusive: how adequate is the organization of the federal government for dealing with these questions-particularly, how adequate is the organization at the Presidential level?

I believe we have the right basic ingredients. The Office of Science and Technology has grown steadily; it now has a staff of over 50, more than 20 of them professionals. This high-quality staff works closely with the agencies, with the Bureau of the Budget, with the National Security Council staff, with the Council of Economic Advisers, with the White House staff, and with the committees of the Congress. Its central concern is the evaluation of existing and potential programs, the coordination of agency programs, and participation in the larger discussions of priorities and emphases. On selected major issues it benefits from the external advice of the President's Science Advisory Committee and over 200 consultants. Internally it draws on the expertise and experience in the agencies through the Federal Council for Science and technology and its panels.

But I believe OST and the Science Advisory apparatus need strengthening. Before I get more specific, I would like to caution that the easy answer to all problems in government, scientific and nonscientific, seems to be to move them closer to the President. I do not think that answer is tenable for many thingshe is already overburdened.

My first guiding principle as regards government science organization (and most other organization) is this: decision-making should be pushed to the lowest responsible level appropriate to that decision. I question the wisdom, for example, of asking a high-level group to make decisions which could be made by a laboratory director. On the other hand, there is an important class of problems that involve general questions. In my view, the more general the question is, the more it should approach the center of the decision-making apparatus. For example, one function that can best be performed at the center is overall planning. Today we are facing a set of problems involving science and technology, and their interaction with many institutions and sectors of our society whose dimensions extend well beyond the capabilities or jurisdiction of any single department or agency of the federal government. I believe that the development of a greatly improved capability to analyze these complex problems and to forsee