Government Funding of Scientific Research



Government Funding of Scientific Research: A Working Paper of the National Science Board

The National Science Board
Ad Hoc Committee On Strategic Science And Engineering Policy Issues


With the end of the Cold War and the need to reduce the size of the Federal deficit, all facets of the Federal budget have come under scrutiny, including the Nation's investment in research and development (R&D). There has been considerable discussion on proposals to reduce the Federal R&D budget and to reorder its priorities. The National Science Board (NSB) is specifically charged with assessing the health of science in the Nation and with advising the President and Congress on matters of national science policy.1 The Board therefore offers its perspective on the important issues this country confronts today concerning the funding of scientific research by the Federal government. Consistent with its charge, the Board has focused its efforts on issues affecting scientific research as distinct from development.

Peer review of proposals has long assured the funding of the best researchers with the best ideas. However, presently there is no widely accepted way for the Federal government in conjunction with the scientific community to make priority decisions about the allocation of resources in and across scientific disciplines.2 We examine this complex issue and offer our views on this challenging task with two purposes in mind. The first is to guide future actions of the Board in reaching priority decisions about the budget of the National Science Foundation (NSF). The second is to engage the attention and participation of others in meeting this challenge by supplementing present procedures with other systematic ways to reach and prioritize decisions.

The rationale for the major Federal role in funding scientific research goes back some fifty years to the time after the end of World War II, when realization of the impact of science-based technology on the course of the war was keenly felt. The mood was expressed in Vannevar Bush's July 1945 report, Science--The Endless Frontier.3 It is natural to question the validity of the philosophy for today and, particularly, to examine the question of the coordination of federally-financed research.

The Board has studied the report, Allocating Federal Funds for Science and Technology, issued in 1995 by a committee of the National Research Council chaired by Frank Press.4 (We refer to this as the "Press report.") A major theme of that report is the need for some degree of coordination of federally-financed research. This idea is not totally new but was particularly well developed in the report. This timely and critical but highly controversial proposal merits careful attention at this time.

If it is in the Nation's interest for there to be some form of "comprehensive" and "coherent" coordination of federally-financed research,5 it is necessary to have guidelines to provide clear direction on setting priorities within the Federal research budget. The Press report pointed out that guidelines were offered in the 1993 report of the National Research Council's Committee on Science, Engineering, and Public Policy (COSEPUP) Science, Technology and the Federal Government--National Goals for a New Era.6 The Board has considered the adequacy of the COSEPUP guidelines.

This working paper presents the Board's thinking on these subjects. With this paper, the Board hopes to encourage a much needed dialogue among appropriate stakeholders. The document is divided into four sections. The first section addresses the definitions of "research" and "development" and highlights the essential differences between them, particularly as they affect the possible government role in funding. Considerable confusion has been created by imprecise and sometimes improper use of the term R&D. The Board feels it is important to clarify this issue.

The second section revisits the justification articulated by Vannevar Bush for government funding of scientific research. It addresses some of the changes in the past fifty years that may have altered the justification but concludes that the need for government funding of research is just as critical today as it was at the height of the Cold War.

The third section examines the need for comprehensive coordination of federally-financed research. It concludes that such coordination could assist the President and Congress by providing a valuable addition to and improvement over the processes presently in place. However, implementation of such a policy would involve the difficult task of developing acceptable procedures.

The final section addresses the availability of guidelines to provide clear direction on setting priorities. It concludes that further study of priority setting methodologies involving appropriate stakeholders should be undertaken. The NSB recommends such a study and pledges its support for this effort.


Because this document focuses on research, it is appropriate to define "research" as distinct from "development," recognizing that there are instances where the boundaries blur.7


Research is the search for new knowledge and concepts that unify and extend that knowledge. The work, stimulated by theoretical or practical questions, is conducted in the context of existing knowledge and paradigms. A paradigm is a guiding concept or model, based on accumulated knowledge, which is generally accepted as valid and useful.

Typically, research is designed to answer specific questions to fill gaps within the existing body of knowledge or to test the paradigm itself. Work which is intended to confirm or refine an existing paradigm may, in fact, contradict it, thus opening the way for a scientific revolution.

Practical applications of knowledge may range from new products and processes to the information base needed for management or policy decisions. An investigator may or may not have specific, practical applications for the results of his/her work when designing the research. However, extensive history has documented the fact that the most important applications and policy implications are not envisioned at the time of the research. This fact is most especially true of work that leads to new or greatly modified paradigms.


Development is the process by which a new product or process is brought into being or improved based largely on existing knowledge and theory. In an industrial setting, development encompasses a wide range of activities, such as scale-up, packaging, or cost analysis. Here we will consider only the technical development by which the concept may be reduced to feasible practice. We have chosen not to address development efforts outside of the commercial sector, that is, development directed to achieving the mission of a sponsoring agency.

In general, development cannot occur based on existing knowledge and theory only, for there are inevitable gaps in the knowledge base. Experiments are typically designed in the development process to address these specific gaps. Thus development has some important features in common with research, though the questions in a technical development program tend to be of narrower scope than in research.

While there are research aspects to technical development, research does not naturally lead to development in any linear way. Rather, research and development are iterative, with development dependent on research, and often vice versa. Taken together, research and development may be defined as "technical innovation." Invention is possible at any stage in the technical innovation process and success is necessary at every stage to produce a commercially viable product or process.

Observations on the R&D Definitions

Research and development, as here defined, are related: not every activity can be clearly classified as one or the other. Additional phrases such as "applied research" or "exploratory development" have been created to provide finer definition of the gray areas between research and development. The above definitions are simpler and adequate for present purposes.

It should also be noted that success in technical innovation is necessary, but not sufficient, for commercial success. Many other factors influence the ultimate commercial success or failure of a new product or process. Some factors, such as marketing, distribution, design for manufacturability, and testing, are primarily the responsibility of industry. Other factors, such as the cost of capital, liability laws, environmental regulations, and tax policy are dependent on government actions and general economic conditions. It is, in fact, these interdependencies that necessitate close cooperation among the sectors-- academy, industry, and government--to ensure the economic well-being of the Nation.

Our definitions distinguish research from development and also indicate the relationship between them. Discussion of support by the government must deal carefully with this relationship, while recognizing that the rationales for supporting the two are quite different.

It further should be noted that research and education are inexorably linked in U. S. higher education in science and engineering. During the undergraduate and graduate years, students learn the fundamentals of their fields. However, because the knowledge base is growing explosively, students must also learn how to learn, lest their education become obsolete. It is in this realm that research becomes a powerful part of both undergraduate and graduate education, which is one of the great strengths of the American higher education system. In research, students learn how to gather current knowledge, how to pose significant questions to further that knowledge base, and how to frame and implement an approach to address their questions. This research/ education experience is invaluable training, not only for those continuing in research, but for the broader workforce and an informed public.


Prior to World War II, support for research by the government of the United States was largely focused on government missions and carried out by Federal employees in Federal establishments. The experience with weapons development during the war highlighted the enormous potential impact of the results of scientific research on national needs. It was also realized that academic research was a powerful engine for generating such results.

The government role in supporting research in the scientific community at large was greatly stimulated by the vision enunciated by Vannevar Bush. Bush wrote, "The Government should accept new responsibilities for promoting the flow of new scientific knowledge and the development of scientific talent in our youth. These responsibilities are the proper concern of the Government for they vitally affect our health, our jobs, and our national security."8 Bush used the word "jobs" to describe what elsewhere he referred to as "prosperity" or "public welfare." The concept is now commonly referred to as "economic security." The three areas identified by Bush were those of most concern at the time. Were Bush writing today, he would probably add others, including "the environment," "green manufacturing," and "clean energy sources."

Bush saw the benefits of research accruing to a wide range of national needs rather than to a single objective, such as defense. Indeed, he concluded his letter to President Truman transmitting his report with a broad vision of the impact of science on quality of life: "Scientific progress is one essential key to our security as a nation, to our better health, to more jobs, to a higher standard of living, and to our cultural progress."9

Vannevar Bush clearly recognized that applications of research results often appear many years after the work is started and that there is no certainty as to which of the many national needs will benefit from this work. He also observed that "…basic research is essentially non- commercial in nature. It will not receive the attention it requires if left to industry."10 Today this concept is recognized as a lack of "appropriability." Because of the long-term nature of research and the uncertainties in predicting its practical applications, a company cannot be certain that investment in research will result in a competitive advantage in the worldwide marketplace. Indeed, the increase in global competition has exacerbated the "appropriability" issue. It consequently has increased the need for government support of research.

The Bush vision encouraged the mission agencies to support research universities in fields that were deemed to have probable long-term relevance to their missions. It also led to the establishment of the National Science Foundation and the gradual building of its budget to the point that it has become a major source of support for science and engineering in our universities. The National Science Board was created with its dual mission of overseeing the activities of NSF and monitoring the health of science in the Nation.

As a result of implementing the Bush vision, our research universities have become the envy of the world. The application of new knowledge and talent in science has indeed created handsome benefits in the three areas Bush identified. We will cite just one example in each area. The understanding of the structure and properties of DNA opened up totally new opportunities to address health issues and provided the basis for the vibrant new biotechnology industry. Polymer and photochemical research led to the creation of photoresists that are key to the success of the microelectronics industry, which accounts for well over a quarter of a million jobs in the U. S. today. The atomic clock, which was based on research in atomic physics and was stimulated by needs in astronomy, provided a foundation for the development of the Global Positioning System to satisfy a critical defense need. More recently, it is creating a large commercial marketplace for everything from ships to backpackers.

In the fifty years since the end of World War II, major changes have occurred here and overseas that might have an impact on the rationale for government support of scientific research. Two of the most frequently cited are the end of the Cold War and the emergence of a global technological marketplace. Another is the increasing need for information and knowledge as a basis for policy and management decisions by institutions and individuals, to enable them to contend with the modification of natural and social environments that is occurring at increasing rates, over larger scales, and in fundamentally new ways.

Do these changes call for a major change in our attitude toward research? We believe that none would invalidate the justification for wise government support of research. Health, economic security, and national security remain as imperatives, and are now joined by social and environmental concerns. Only the sense of priority has changed. Defense priorities have decreased but competition from global science- based technological industry and environmental and social concerns have increased as no one would have dreamed in 1945.

Some Asian nations, most prominently Japan, have succeeded in building excellent high-tech industries in the absence of a publicly-accessible academic research base. At the same time, U. S. industry appeared to be faltering in areas such as consumer electronics and in fundamental research in manufacturing engineering. These observations have been used to suggest to some that government funding of science might not be required to enhance national prosperity. We believe that this is an incorrect conclusion stemming from a number of misunderstandings of the characteristics of research and development and their role in the total innovation process.

First, as discussed in the section on definitions, success in bringing high-tech products and services to the marketplace involves a total innovation process including functions such as research, development, manufacturing, marketing, and others. All of the functions involved must work well. The problems with the U. S. consumer electronics industry have been thoroughly studied and are well understood.11 American firms lost market share to competitors with shorter product cycles, lower costs, and superior quality. Even excellent science will not compensate for such a weakness in the industrial environment.

Second, as also discussed in the section on definitions, the innovation process is an iterative, not a linear, process. While some very important product developments are triggered by new knowledge from research, the majority are stimulated elsewhere -- by market needs, by manufacturing advances, and by ideas from the development laboratory. These product developments can proceed largely on the foundation of existing and widely understood scientific and technical knowledge. The consumer electronics industry fits this model as does the mature semiconductor industry. Thus, even nations without ready access to research capabilities can prosper and excel in these product lines.

The most obvious situation in which research can lead to a competitive edge for industry is where there is a fundamental breakthrough, a paradigm change. Here there may be opportunities to create whole new industries. The understanding of DNA was surely one such paradigm change. When this occurs, a nation with both a strong industry and a leading scientific capability can capitalize on its closer access to knowledge and talent to become first in the world market with the most innovative, profitable products and services. It is under these less frequent and highly unpredictable circumstances that research makes a critical contribution to industrial competitiveness.

There are other research benefits that can be at least as valuable. Basic expertise is needed to evaluate new technical opportunities regardless of their source. Whatever the extent of a nation's investment in research, some breakthroughs are bound to occur elsewhere. Having expertise in a field makes it possible to catch up with the originator in the implementation phase and even get to market ahead of the originator. In planning technical programs, whether in research or development, it is valuable to understand what can work and it can be even more valuable to know what cannot work. Finally, ready access to the talent in research universities, whether as employees or consultants, is an asset to industry in all facets of the innovation process. These benefits from research can be seen in the strength of our information, chemical, and pharmaceutical industries and the competitive advantage they have gained from close access to basic science.

We conclude that changed circumstances in recent years do not reduce the desirability of continued government funding of scientific research. Changes in national priorities do not negate the potential of research benefits which are long term and uncertain in detail but have proved over time to be substantial. In the presence of global competition a nation should be strong in all facets of technical innovation and should have available a continuously renewed base of knowledge to inform its decisions and those of its citizens. A nation requires a robust high-tech industry, a scientific talent base, and a vigorous research activity to prosper over the long term.


We recognize that a degree of coordination of Federal research spending exists across disciplines and that during the last decade the Executive branch has taken steps to improve coordination of research across agencies in key areas. Indeed, the Office of Management and Budget in consultation with the Office of Science and Technology Policy provides annual budget guidance to all agencies participating in support of priority research areas in preparing the Federal budget for submission to Congress. Too, agency budget submissions must be developed in the context of the Government Performance and Results Act, which requires that agency supported research activities have measurable outcomes toward achieving agency missions. We note in particular that the committees of the National Science and Technology Council (NSTC) provide coordination in areas of special national interest, such as global change, the development of less polluting transportation, energy, specific health areas, childhood development, and the future of the U.S. program in the Antarctic.12

These efforts benefit from special Administration studies, including reports of the President's Committee of Advisors on Science and Technology (PCAST) and the NSTC.13 But, beyond those special areas, coordination depends on individual agency-to-agency agreements, informal cooperation across agencies at the program level, and the memories of Congressional committees. Sometimes important decisions about the allocation of limited resources happen by default, without explicit weighing of alternatives. There remains a need to examine and coordinate the science and engineering research budget as a whole.

We are proposing that the Federal government take upon itself the high-level coordination of the diffuse sources of Federal funds for research as suggested in the Press report. Improved coordination and decision-making at the Federal level could lead to a better alignment of expenditures with respect to national priorities without in any way replacing the spontaneous generation of ideas and proposals by individual research workers and teams. Such coordination could correct deficiencies that will inevitably surface in its absence. The main deficiencies are gaps, overlaps, and failures to meet priorities.

Decentralized allocation will sometimes result in separate agencies unintentionally pursuing the same agenda.14 Duplication of research efforts is not always a bad thing, even when funds are scarce. It may encourage competition among investigators and advances in knowledge across a broad front. Whether or not any particular duplication is desirable competition or wasteful overlap has to be decided explicitly. There is no reason to expect the optimum answer to arise by happenstance.

In exactly the same way, decentralized allocation will sometimes leave important areas of research inadequately covered. Individual funding agencies and individual researchers may incorrectly presume that others are pursuing particular topics and related areas. Although such gaps may correct themselves over time as the writers and readers of proposals see what has happened, this can be a wasteful process, and even quite destructive if young researchers decide to leave important unfunded fields. Coordination would allow one to see gaps in advance and judge whether they should be eliminated.

Sometimes there will be a clear sense within the Federal government that some areas of research merit particularly high priority for social or economic reasons (examples: climate, hydrology, violence, materials, transportation, etc.). The uncoordinated generation of research proposals will not completely ignore such priorities, but cannot be expected to reflect them with great fidelity. It was already noted that important applications of research are not always foreseen when the research is planned. This observation does not deny that research aimed at a particular application is more likely to achieve it than research aimed in some other direction. Comprehensive coordination can achieve a rough conformity between accepted priorities and the allocation of resources. This fact becomes increasingly important when funds are scarce. As an extreme example, it is a common observation that completely decentralized modes of allocation run into particular trouble when budgets must be cut. At such a time it is easy for the general interest to be overridden by parochial interests.

Whenever there is some amount of comprehensive coordination and decision-making, it is supremely important that the criteria of choice be appropriate. There is no virtue in doing the wrong thing efficiently. Any scheme of oversight must begin with explicit discussion of and agreement about the goals to be achieved.


Within the Federal budget, there should be an overall strategy for research, with areas of increased and areas of decreased emphasis. The budget as a whole should be adequate both to serve national priorities and to foster a world-class scientific and technical enterprise. To this end, Congress and the Administration need to establish a process that examines the complete Federal research budget before the total Federal budget is disaggregated. Departments and agencies should make decisions based on clearly articulated criteria that are congruent with the overall strategy.

Within the Executive branch, the interagency NSTC, and before it the Federal Coordinating Council for Science, Engineering and Technology (FCCSET), have successfully organized crosscutting research areas of national interest, such as global change, energy, transportation science, environmental science and technology, and human resources for the twenty-first century. However, in order for broader coordination and priority setting to be successful, general guidelines are required to provide clear direction.

The most recent effort by the scientific community to recommend guidelines for the allocation of research resources across all fields of science and engineering appears to be the COSEPUP report. That report proposes that Federal research resources be allocated among different scientific fields and Federal agencies and departments so that the United States will be among the leaders in all major fields of science and the leader in selected major fields.15

The National Science Board supports the spirit of the COSEPUP recommendations but believes that they may not go far enough. The COSEPUP criteria would assure that the United States would be competitive with, indeed somewhat ahead of, other nations. This, we believe, is highly desirable but may not be sufficient. In addition to questions of world leadership, one must also ask what is the appropriate scale of the investment to meet the needs of the greatest economic power in the world. Given the broad range of national needs that can benefit from the results of scientific research, the Nation may choose, and may be able to afford, to invest beyond the levels that the COSEPUP criteria would suggest. Thus the Board believes that further study is needed before a particular methodology for setting priorities is adopted.

To ensure the most effective use of Federal discretionary funding it is essential that agreement be reached on which fields and which investment strategies hold the greatest promise for new knowledge that will contribute most effectively to better health, greater equity and social justice, improved living standards, a sustainable environment, a secure national defense, and to extending our understanding of nature. It is intrinsic to research that particular outcomes cannot be foretold; but it is possible, indeed necessary, to make informed choices and to invest wisely. The need for better coordination and priority-setting is not related to cycles of fiscal constraint alone. It is, rather, an integral aspect of a sound, future-oriented strategy for the investment of limited Federal dollars.

Although the need for establishment of research priorities has been discussed often, no agreed upon method exists for carrying out this task. Moreover, no consensus has been built to support such a methodology. Several subfields of science have long-established methodologies for producing ranked lists of new construction projects: for example, the Decadal Studies in Astronomy, the periodic reports of the High Energy Physics Advisory Panel (HEPAP) ranking accelerator projects, and the occasional reports ranking investments in x-ray and neutron scattering sources.

However, these priority-setting exercises have been within fields and subfields of science. We are aware of no examples of the scientific community agreeing on the relative priorities for investment across scientific fields. Although many scientists consider the task both undesirable and undoable, the National Science Board believes that this difficult task will become increasingly important and must be faced over the next few years.

The Board has concluded that an appropriate next step is to initiate a study of guidelines that go beyond those proposed in the COSEPUP report. The purpose of this task would be not to set priorities, but rather to undertake a study of how they might best be set. Specific charges would be to: [a] review, in light of changing circumstances, the goals for Federal investment in scientific research as stated in the Administration report, Science in the National Interest;16 [b] examine what methodology and criteria might best be used to set priorities across different scientific fields and disciplines toward the attainment of those goals; and [c] consider what mechanisms will be effective in building broad public and scientific support for, and involvement in, priority setting. The study should involve the opinions of a diverse group including, among others, active researchers with breadth of vision.

The National Science Board recommends further study of priority-setting methodologies involving appropriate stakeholders. The Board believes that this task is of paramount importance to the future health of U.S. science and technology. It should be undertaken to assure the continued flow of wide-ranging benefits to society from Federal investments in science and engineering research. The Board offers its assistance on this critical task in any way that the President and the Congress would find helpful.


1 National Science Foundation Act of 1950, as amended, 42 U.S.C. Sec. seq. A particular responsibility of the Board in implementing this mandate is the biennial publication of Science and Engineering Indicators.

2 Throughout this paper, "science" includes mathematics, engineering, and materials research.

3 Vannevar Bush, Science - The Endless Frontier (40th Anniversary Edition, Washington, DC: National Science Foundation, 1990).

4 National Research Council, Committee on Criteria for Federal Support of Research and Development, Allocating Federal Funds for Science and Technology (Washington, DC: National Academy Press, 1995).

5 Ibid, p. 5.

6 National Research Council, Committee on Science, Engineering, and Public Policy, Science, Technology, and the Federal Government: National Goals for a New Era (Washington, DC: National Academy Press, 1993).

7 Definitions of "research" and "development" are congruent with operational definitions for the National Science Foundation Survey of Industrial Research and Development.

8 p. 8.

9 p. 2.

10 p. 22.

11 Richard S. Rosenbloom and William J. Abernathy, "The Climate for Innovation in Industry: The Role of Management Attitudes and Practices in Consumer Electronics," Research Policy, 11, no. 6 (1982): 209-25.

12 National Science and Technology Council, Technology for a Sustainable Future/A Framework for Action. Washington, DC: US Government Printing Office, 1994, and Infectious Disease - A Global Health Threat. Washington, DC: September 1995.

13 President's Committee of Advisors on Science and Technology, Federal Energy Research and Development for the Challenges of the Twenty-First Century. Washington, DC: November 5, 1997. National Science and Technology Council, National Security Science and Technology Strategy. Washington, DC: OSTP, 1995.

14 The funding environment is an ecosystem. Changes in a particular agency's budgets and programs may have unintended consequences by creating gaps in significant areas of research and increased pressure on other agencies that may not be in a position to respond.

15 pp. 18-24.

16 William J. Clinton and Albert Gore, Jr., Science in the National Interest (Washington, DC: Office of Science and Technology Policy, 1994).


Bush, Vannevar. Science-The Endless Frontier (40th Anniversary Edition). Washington, DC: National Science Foundation, 1990.

Clinton, William J. and Albert Gore Jr. Science in the National Interest. Washington, DC: Executive Office of the President, Office of Science and Technology Policy (OSTP), August 1994.

Committee on Criteria for Federal Support of Research and Development, National Academy of Sciences, National Academy of Engineering, Institute of Medicine, National Research Council (NAS/NAE/IOM/NRC). Allocating Federal Funds for Science and Technology. Washington, DC: National Academy Press, 1995.

Committee on Science, Engineering, and Public Policy, NAS/NAE/IOM. Science, Technology and the Federal Government/National Goals for a New Era. Washington, DC: National Academy Press, 1993.

Energy Research and Development Panel, President's Committee of Advisors on Science and Technology (PCAST). Federal Energy Research and Development for the Challenges of the Twenty-First Century. Washington, DC: November 5, 1997.

Executive Office of the President, OSTP. Science and Technology Shaping the Twenty-First Century. Washington, DC: April 1997.

National Science Board. Science and Engineering Indicators-1996 (NSB-96-21). Washington, D.C.: US Government Printing Office, 1996.

National Science Foundation. NSF Survey Instruments Used in Collecting Science and Engineering Resources Data (NSF-95-317). Arlington, VA: National Science Foundation, 1995. National Science and Technology Council, OSTP. Accomplishments of the National Science and Technology Council (NSTC). Washington, DC: 1996.

_________, Technology for a Sustainable Future/A Framework for Action. Washington, DC: US Government Printing Office, 1994.

_________, United States Antarctic Program. Washington, DC: 1996.

_________, Committee on International Science, Engineering and Technology Working Group on Emerging Infectious Diseases. Infectious Disease - A Global Health Threat. Washington, DC: September 1995.

_________, Committee on National Security. National Security Science and Technology Strategy. Washington, DC: OSTP, 1995.

_________, Committee on Transportation Research and Development Intermodal Transportation Science and Technology Strategy Team. Transportation Science and Technology Strategy. Washington, DC: OSTP, September 1997.

Rosenbloom, Richard S., and William J. Abernathy. "The Climate for Innovation in Industry: The Role of Management Attitudes and Practices in Consumer Electronics." Research Policy 11, no. 4 (August 1982): 209-25.

Back to NSB Publications