« PreviousContinue »
CALIFORNIA INSTITUTE OF TECHNOLOGY,
Pasadena, Calif., August 11, 1969. Hox. EMILIO Q. DADDARIO, Committee on Science and Astronautics, U.S. House of Reprsentatives, Rayburn
House Office Building, Washington, D.C. DEAR MR. DADDARIO: Thank you very much for your letter of July 15, with the invitation to submit a written statement to your Subcommittee on Science, Research and Development in view of my inability to appear in person.
I would like to divide my statement into three parts. The first has to do with some trends which I see developing in attitudes in this country toward research and development, and toward science and engineering which concern me very much. The second has to do with the problems of applied research, some of which have recently been considered by a Panel of the National Academy of Engineering. The third has to do with the specific subject of your hearings: the matter of Centralization of Federal Science Activities.
A. TRENDS IN ATTITUDES TOWARDS RESEARCH AND DEVELOPMENT I would like to cover the first by conveying to you the following answers which I transmitted to Senator Fred R. Harris, Chairman of the Subcommittee on Government Research of the Senate Committee on Government Operations, at his request and in response to his questions :
1. What, in your considered judgment, have been our major national science policy assumptions, that is, the conceptual underpinnings of the Federal Government's support of research and development, since the creation of the Office of Scientific Research and Development in World War II ?
(a) Research and Development are an unmixed good. (1) Research support should grow with the number of people qualified to do it.
(c) Development needs to be judged on the basis of its contributions to economic health or national security.
(a) Exponential growth of research at 15 to 20 percent per year, of development at 10 percent, can go on for a long time.
(e) Research and development should be done where it can be done best. Research choices are best left to researchers.
2. Are new assumptions emerging or evolving, and if so, what are they?
Yes. Mostly they are overreaction from the earlier-held views listed in (1), and are mistaken, or only part of the truth. Nevertheless, they are becoming widely held by differing and more or less numerous groups of people.
(a) Research and Development are the source of many of society's problems, and the solution of few.
(b) Research is a competitor for funds with worthy social programs.
(c) Research and Development should be distributed geographically to build new centers of excellence.
(a) Growth in Research should be slow, and in Development near zero. (e) Derelopment should concentrate on urban and environmental questions.
(f) Research and development people should not determine the nature of their work.
(g) Uuch research is irrelevant and wasted.
3. In the light of any fundamental changes. how might our science policy mechanisms be modified to formulate and implement science and technology priorities and policies that are more responsive to national goals and that are more in consonance with the national interest?
On research (and graduate education) (a) Institutional grants on the basis of excellence at least as much as on size. (b) Full graduate student support.
(c) More interaction between researchers and supporters in determining fields of concentrat n.
(d) Clear recognition that research applications are not always predictable, so that though applied research priority judgments must involve developers and users as much as researchers, much basic research must be chosen by researchers.
(e) Continue multiplicity of support.
(f) Encourage developers to work on environmental and urban problems, but give them the flexibility to invent that the military developers have been given.
B. APPLIED RESEARCH
I. THE NATURE OF THE PROBLEM
People now ask for immediate relevance; this is clearly a change in attitude. In 1961 it was considered, by most, wonderful that the government was spending 2 billion dollars for basic research, 12 billion dollars for research and development. In 1969 it is considered, by many, tragic that we spend 2.5 billion dollars for research, 4 billion dollars for applied research, 18 billion dollars for research and development. From the public's point of view, basic and applied research are tied together in a complex way.
(a) Definitions. Basic research and applied research are defined by motivation. However, motivation may be different for the researcher, the supervisor, and the source of support, even though the research itself may be identical. We can instead divide applied research by discipline, perhaps thought of in terms of applicability of the discipline to various ends—rather than specific research in the discipline. For example, if we make optics healthy, it is applicable to library management, medicine, aerial photography, etc.
Now what is the right amount of support across the R and D spectrum?
(6) I would like to present the following theorem, which I heard indirectly from Charles Hitch: in the spectrum of activity running from basic through applied research to advanced development to engineering development to development for production, and then to testing and evaluation, the costs of an optimum program go up for many kinds of enterprises (electronics, aircraft, etc.) on a logarithmic scale. The support from industry (and from missionoriented government agencies) is likely to fall well below this optimumfurthest below for basic research ; less for applied, etc. The reasons, for industry, are the difficulty of showing a direct connection with a profit-producing enterprise, and the fact that the more basic the research, the less subject it is to patent or propriety position. There are analogous reasons for the mission-oriented agencies.
(c) To a substantial extent the National Science Foundation and the National Institutes of Health have brought basic research nearer to the optimum—at least in operating funds—though it probably never has reached optimum ; in fact, in recent years it has fallen rather badly below again and desperately needs reinforcement, especially in facilities and institutional funding. In the case of applied research, there has been no central agency. Until now, such an agency has perhaps been less needed. Some work (after all, 4-5 billion dollars is a lot) has been done by central laboratories of various industries, and by missionoriented agency support of industrial and university research. But, as society's problems become more pressing, we find that in many cases (by no means all), the applied research (in physical, biological, or social-a special case-sciences) on which to build development is missing. And the somewhat scattered and in-and-out approach of the mission-oriented agencies and of industry inevitably means that the discipline-oriented approach suffers. Furthermore, areas important to a variety of agencies or industries, but not the primary responsibility of any one, also tend to fall into the cracks. Problems in the American steel, coal, and railroad industries testify that this can have serious consequences. Thus it can be argued that the span of the bridge of research, development, testing and evaluation which is labeled "applied research”, though heavily funded, may be the weakest of all and in need of special attention from government, industry, and universities. The following chart will illustrate my point:
(a) A proposal to set up an Applied Research Agency to provide institutional, teaching and research support was, as you know, considered at one time by some members of Congress.
(b) In view of the legislation which permits (although it does not mandate) NSF to support applied research at universities and nonprofit enterprises, NSF could very suitably do this job—with some provisos and caveats which I will mention later. It is used to supporting university research and it has had experience in determining quality. It has, or at least could have, a knack for seeing important areas which mission agencies can overlook. Of course, there are also some problems, e.g.
(1) Increasing involvement of NSF and its head in mission or development problems.
(2) Dilution of effort.
(3) Differences in style. (c) The applied research gaps which concern me are in areas which universities regularly cover in their engineering schools or divisions (most engineering research falls into the applied research category-which also covers applied biological and social science research). The following are examples of applied reesarch with profound social implications:
(i) Environmental pollution applied chemistry, hydrology, meteorology, etc.
(ii) Jet noise and sonic boom—engine design, transonic and supersonic filow.
(iii) Earthquakes-prediction, prevention (or at least mitigation), earthquake engineering. (d) A linkage between basic and applied research would be desirable because often they differ only in motivation, and the public still (or again) needs education in the ultimate possible practical-as well as the certain intellectual-value of basic research.
It should be remembered that the most spectacular applications tend to come either from relatively new basic research-e.g., nuclear power from nuclear fission in the 1940's, transplants from immunology now-or from old basic research which, either under the impetus of new applied needs or from interaction with new basic research, becomes applicable when it was not so before. For example, lasers would now be gaining application much less rapidly in the absence of the energy level information from spectroscopic data up to ninety years old, which had until then been applicable principally in astronomy and other basic research. All of this reemphasizes the need for a linkage.
(e) Support of institutions and teaching, as well as of research, is needed. This is parallel to the situation in basic research, where the National Science Foundation has done well. The National Institutes of Health has spanned the basic/applied spectrum—and its performance probably offers some arguments for, as well as some lessons about, connection between the two.
Still, there are NSF problems—which lead to several caveats and provisos to which I referred earlier.
III. PROVISOS AND CAVEATS
(a) Necessary as it is, support of applied research must not come at the expense of basic research. If the applied research mission is added, applied research funds must be added beyond the needed increase in basic research. Unless basic research is supported at a fully adequate level, the applications will dry up-sooner rather than later.
And one can't be positive that Congress and the public will not lump all the NSF appropriations together and say: "They were getting $450 million, now they are up to $650 million; cut back!”—forgetting that $200 million is for the new mission.
(b) The criteria for applied research, as in basic research, include quality, measurable by peer groups. Applied research quality has sometimes left something to be desired—although not in the sense that applied research is intrinsically second-rate, which some basic research people may have felt. In any event, quality needs to be carefully monitored.
(c) However, in applied research other criteria also exist for what should be supported. Priority and relevance become important considerations. Additional reviews and judgments, by different kinds of people, are needed for this purpose. The sponsoring agency needs to work closely with development-oriented groups, and the researchers also need some means of association with such groups. Yet the researchers must be able to make indepedent judgments about some criteria for applied research. The sponsoring agency can say, for example, how critical some disciplines are to developments; the researchers can say how far from theoretical limits the current tools or approaches are. There would need to be an annual review of mission agency needs—and an annual report by them on how they had used NSF applied science efforts. FCST, by its charter, could supervise this effort (though, as a former member, I have my doubts).
(d) NSF would have to be very selective, since it could support only an even smaller fraction of total federal applied research ($4.5 billion) than it does of basic research ($2–3 billion), and would have to be careful to avoid being merely a ballast tank.
(e) Different kinds of managers (research and development manager types) and organization would be required. There are successful industrial prototoypes (e.g. Bell Telephone Laboratories) which use principles such as geographical propinquity, plus organizational separation. If we can get the quality of management for a corresponding federal enterprise which exists in a very few industrial enterprises, I would feel very much more comfortable.
(f) There are "handover” and “not invented here” problems related to mission-oriented agencies and to industry. Getting others to fund applied research even after they have picked up the corresponding development can be very difficult. In recent years, we have unfortunately seen the reverse kind of handover happening, even in basic research-for example, mission agencies giving it to NSF without the money. Doing so for applied research would be fatal to it and to NSF.
(9) If possible, there should be parallel support of applied research hy NSF to universities, with the mission agencies supporting industry in related development and universities in applied research. Last but not least, there should be close contact between universities and industry working in allied fields—with some industrial support, not closely specified as to use, of the university activities in basic and applied research helping to lubricate that contact.
I do not think I have here all or even most of the solutions to the problems currently besetting applied research on the federal level. But I think I see the outlines of the problems and have some suggested lines of approach to solutions. We need those solutions because both healthy basic research and high-quality applied research, directed at the areas where society has needs, are required. At least this is so if science and technology are to do their share in helping solve our economic, social, and even political problems.
C. CENTRALIZATION OF FEDERAL SCIENCE ACTIVITIES
I have read through the report which was presented to you by your Committee's staff and find that it gives most of the arguments, pro and con, for various degrees of centralization. I myself, both before and during my eight years of
federal service, considered the same questions frequently. The idea of a single agency for federal research and development activities can be dismissed quite easily. The various using agencies which support large-scale development must bear the responsibility for those developments; they cannot deal at arm's length with some other federal agency supported by separate appropriations, which would do those developments for them. This fact substantially outweighs any arguments to the contrary which might be made on such bases as interrelation of various kinds of developments (military, atomic energy, space, transportation, etc.), or the desirability of having a single agency which could be held responsible for the general health of development in this country.
The arguments about centralization of science activities are more difficult. There clearly is a substantial similarity between the kinds of basic research, and even applied research, which are supported by such diverse agencies as Defense, NASA, AEC, and the National Science Foundation. And, based on the identity of the institutions where much of its research is done, NIH can be argued to be in a similar situation. It might therefore seem that a single agency responsible for all of these activities would prevent duplication, set priorities, etc. This would also get the mission agencies out of what some people might regard as an undesirable area for them to be in, and an unnecessary one since, at least for basic research (and, to a much lesser extent, for applied research), those agencies and developmental activities can use the results of the basic research only in an unpredictable way at a substantially later time.
However, there are three strong arguments which lead me to conclude that although closer coordination, of a kind that the Office of Science and Technology, together with the Bureau of the Budget and the Federal Council of Science and Technology (which operates under the chairmanship of the Director of OST), would be very desirable, a single agency to do all of the basic research now carried out by the various mission-oriented agencies, by NSF (and perhaps NIH) would have grave defects which would make it a worse choice than the present arrangement.
1. Research continues to be relatively unpredictable. It would be a mistake to have an arrangement whereby a single individual either at the top of, or buried somewhere in, a single agency could make the decision not to follow a certain line of research despite the opinions of a large body of researchers. The multiplicity of support which has existed in the past has prevented this from happening, and a number of vital pieces of basic research have thereby taken place which otherwise would probably have been cut off.
2. The mission-oriented agencies have a great deal to gain by supporting basic research in terms of establishing relationships with researchers, so that those researchers may become interested in the applications and development which lie at the heart of the activities of the mission-oriented agencies. They then can act as consultants for, or have ideas on, these developments, thus providing an important input from highly intelligent and motivated people outside of the government. Researchers, faculty and students simply will not have the same attitude toward some of the applied problems if their work is enirely supported by some pure-research agency. Furthermore, I believe that the agencies which ultimately use the fruits of basic and applied research have a direct, even if not the primary, responsibility for the health of those activities. This means that these agencies should directly support such activities.
3. There is no doubt in my mind that if the basic research activities of missionoriented agencies were folded into a new agency, the budget with which that new agency started would lie between the sum of the present support and that of NSF alone, and I greatly fear that it would be nearer the latter than the former. I do not believe that the cause of basic research would be thereby served. In fact, I believe that the nation's future intellectual, scientific, technological, and eventual economic health would be damaged. Furthermore, putting all basic research into a single agency risks very gravely the possibility that a single member of Congress in a key position in regard to the affairs of that agency would have a decisive voice on our nation's science policy, and neither the scientific community nor the bulk of the nation's voters would have any appeal from his decisions.
For all these reasons, I believe that centralizing research, as opposed to coordinating it better, would be a mistake.
Thank you for the opportunity to express these opinions on a variety of subjects. I look forward to seeing you again soon. Sincerely,