America's Basic Research: Prosperity Through Discovery
Committee for Economic Development
Maintaining U.S. Leadership in Basic Research: The CED Prescription
The remarkable scientific progress that has helped to define 20th century America is due in no small part to the unique characteristics of our basic research enterprise. We have described these elements of success in American basic researchpublic funding, merit-based allocation of resources, the central role of the research university, the primacy of the individual investigator, and a robust education pipeline for future scientists and engineersand how they have contributed to the social and economic benefits that we enjoy today. But in each of these elements there are currently signs of stress that need to be addressed if the promise of tomorrow's basic research is to meet the expectations created by the successes of the past. In this chapter we will describe the challenges to our system of basic research, offering policy recommendations that address:
Improving the Allocation of Resources for Basic Research
In addition to ensuring an adequate level of resources for basic research, policymakers have a responsibility to maximize the potential of those resources through efficient allocation mechanisms. There is waste and inefficiency in the current allocation systems for basic research. Shortcomings in allocation arise on two levels: 1) in the allocation of the research grant from the funding agency to the researcher, whether it is done competitively, based on peer review, and whether it is directed toward individual investigators rather than institutions; 2) in the allocation of funds to agencies and missions from Congress.
Mechanisms to Ensure Quality in Funding
If basic research were like any other production process, efficient allocation of resources would be a relatively straightforward matter. Resources would go toward efforts that demonstrated the highest productivity, as calculated through a measure of output. But measuring research outputs and the productivity of basic research in general, let alone for individual basic research projects, is highly problematic.
Over the years, government agencies under various mandates have attempted to develop output measures and related productivity and quality indicators for federally-sponsored research. The most recent initiative has been mandated by Congress under the auspices of the Government Performance and Results Act (GPRA). GPRA requires all government agencies to submit performance plans for measuring and assessing program impacts as a means of increasing public accountability, an exercise that has proved understandably problematic for science and research-related programs. Efforts to increase accountability in government are laudable and government-funded basic research should not be exempt. Nonetheless, GPRA should not impose one-size-fits-all criteria for measuring results. Such an approach would, at best, prove unworkable for basic research programs; at worst, a rigid imposition of quantified performance standards would undermine basic research by shortening the time frame of projects and limiting their scope to areas where the payoff is predictable at the outset.
Measuring basic research output is no less a troublesome issue for business, which companies have addressed with mixed success. The IBM case highlights that company's efforts to quantify its own research output, measuring for example, number of patents filed, number of external awards received by IBM researchers, and self-assessments of key results for the year.
Absent meaningful and practical output measures, we must rely on other means of efficiently allocating resources. The best alternative, we believe, is the system of peer-reviewed competition for research grants. Peer-reviewed awards to individual researchersand increasingly to teams of researchersfor individual research projects helps to ensure quality on a project-by-project basis. Some critics have charged that peer review encourages an "old boy" network because it favors the status quo in grant awards. Yet, we know of no better mechanism for emphasizing substance over reputation or political interests.
The worst abuse in the allocation of basic research dollars, leading to the least productive use of research funds, is Congressional earmarking. The most recent survey of these activities demonstrates dramatic growth in the number of earmarks for academic research in recent years.1 Over the course of a decade, academic earmarks increased in value from the tens to the hundreds of millions of dollars. Such earmarks frequently place narrow constituent interests or even otherwise laudable goals like regional economic development, over scientific merit.2 In bypassing the competitive, peer-review process for determining merit, earmarking for basic research is a form of pork-barrel, like any other, it should be recognized as such and acknowledged to have no place in our publicly-supported basic research enterprise.
Of course, allocation of funds through peer-reviewed grants to individual investigators does not meet all the needs of our basic research enterprise. For example, there has long been, and continues to be, a need for public investments in large-scale projects, often referred to as "big science."3 Historically, projects like the fusion reactor at Princeton University, the historic Voyager satellite missions, and NSF's deep sea drilling project have proved very important to basic research and could not have proceeded in the context of relatively small individual investigator grants. In cases such as these, when the nature and size of the research project makes individual investigator grants impractical, direct funding of institutions rather than individuals is appropriate.
Unfortunately, because of the large institutional and regional economic stakes, funding for big science often falls prey to political interests. Also, it is often threatened by budget cuts or other outside attempts to influence the character of the project. It is the responsibility of policymakers to ensure that necessary investments in big science and institutional grants proceed on the basis of scientific merit (determined by expect advisory committees) and in the larger context of national needs and priorities.
Further, it is important to emphasize that allocation for big science through institutional grants simply means investing in the infrastructure necessary to make certain fields of inquiry possible by giving researchers the tools to do their work. The scientific work that proceeds from investments in these tools should conform to the same process of competitive peer review that individual investigator projects do.
In sum, the primary mechanisms for allocating federal basic research funds in all agencies and to all institutionswhether universities, federal labs, or elsewhereshould be based on scientific merit determined through peer review and the support of individuals and projects. Political earmarks for basic research are an unproductive use of scarce funds and should be halted. The needs of big science can be met without sacrificing scientific quality, but this depends critically on the motivations of policymakers and the input of scientists.
Ensuring Adequate Funds for All Basic Research Disciplines
Peer review goes a long way toward ensuring quality in the allocation of funds from federal agencies to individual research projects. But at the start of the funding streamthe allocation of basic research dollars from Congress to federal agenciesthere is no such mechanism to ensure overall quality. Setting broad priorities in basic research is the domain of policymakers in Congress and Administration, and should be the result of informed policy debate. Scientists have an important role to play in ensuring that the debate is an informed one; but ultimately, these decisions are appropriately made by our elected leadership.
Certainly, as the growth in earmarking illustrates, there is room for improvement in how priorities are established in public funding for basic research. The Committee for Economic Development (CED) does not believe an overhaul of the congressional appropriations process in order to serve the needs of basic research is practical or remotely likely. Just the same, Congress can act in a number of ways to ensure quality basic research in its appropriations. We believe it critical that Congress work to achieve a balance among basic research missions, a diversity of missions, and that increasingly important cross-disciplinary research receives better treatment by congressional appropriators and agency administrators alike.
Balance in basic research missions
Currently, the health mission dwarfs all other categories of federal support for basic research.4 Although we recognize that missions must be prioritized to ultimately reflect the needs of society at large, we also believe that neglect of less popular areas of research is foolhardy. Health research has benefited immeasurably from advances in the computer sciences, the behavioral and social sciences, mathematics, and physicsyet these disciplines receive most of their federal support from non-health missions. The defense mission, in particular, has been an important, and in some cases a primary, source of funding for diverse scientific disciplines. But, the decline in funding for this mission alone from its Cold War levels has had negative consequences for some areas of basic research.
While the health sciences have clearly benefited from applications of advances in other disciplines, various non-health fields of science have, in turn, been furthered by biomedical research. In addition, the needs of biomedical science have created demand for further research in other fields such as chemistry, physics, materials and engineering, and information technology.
These observations underlie the concept that basic research and subsequent development is a mutually reinforcing process occurring between a variety of science and engineering fields. All of these fields will require adequate support in an increasingly multi-disciplinary environment.
The agency missions themselves are dependent on many scientific disciplines. Ultimately, progress in any single field will not be sustained if other fields wither on the vine. In order to achieve an appropriate balance in funding for scientific disciplines, appropriators need to pay close attention to the impact of agency and mission funding decisions on specific scientific disciplines. As some missions become lower priorities in the federal budget (the post-Cold War defense mission), or as Congress pursues structural reforms in major programs tied to basic research,5 the impact (direct or indirect) on scientific disciplines should be assessed and those disciplines should be accommodated in a way that is consistent with mission/agency goals and the needs of basic research generally. Often, this may mean making basic research funding a higher priority than other missions and agencies (such as the NSF), as it becomes less of a priority in declining missions (as in defense).
A diversity of missions
The National Science Foundation (NSF) has long been a model of support for peer-reviewed basic research independent of agency missions at the federal level, providing funds on the basis of merit to the entire spectrum of science and engineering disciplines. Nevertheless, much of the scientific progress of the past half-century resulted from much larger allocations to basic research funded by the Department of Defense and the National Institutes of Health. Consequently, any effort to abandon agency and mission-based support for basic research in favor of the "secular" efforts of the NSF would ultimately reduce the amount of public support for basic research in general and undermine the value of a diverse approach to funding. Federal support for basic research should be diverse in its funding sources, resisting efforts for central control or concentration in one mission area. The diverse model is most viable politically and is best-suited for the unpredictable nature of basic research outcomes. Therefore, we do not support calls for a "Department of Science" or for an NSF that would envelop all other federal sources of basic research support.
Cross-disciplinary research within missions
All sources of federal basic research funding should recognize the cross-disciplinary imperative of much of today's basic research and encourage such approaches in their funding decisions. The case studies (and particularly the Pfizer, Merck, and Harvard cases) provide compelling evidence that a cross-disciplinary approach to scientific investigation is a necessity in many areas of research today and is an approach employed successfully by businesses and universities alike.
It is not enough to ensure balance among missions; it is increasingly important to recognize research that must take place across funding missions and across scientific disciplines. Grant proposals that fall outside of the traditional categories or scientific disciplines should not be punished by a traditionally discipline-oriented funding structure and peer review bias. This is an imperative that applies to universities and peer review panels as much as it does to agency administrators and congressional appropriators. Evaluating merit in these cases may require more effort (requiring, for example, a peer review panel that consists of scientists from more than one discipline). But the effort should be made. Universities can promote collaboration by recognizing its value in tenure decisions and through university-level initiatives that are independent of federal agency directives. Harvard's Mind/Brain/Behavior program illustrates such an initiative.
Keeping Research Universities at the Core of American Basic Research
The central role of the research university
CED believes the most productive use of federal basic research funds institutionally is through the nation's research universities. The universities' exceptional track record in performing high-quality basic research is not surprising given the central role of the individual investigator and the widespread use of competition in grant awards.
The research university deserves to be the predominant basic research institution in America for yet another reason: federal support of university research ensures the future health of science and engineering by supporting the training of graduate students. Tomorrow's scientists and engineers receive the best possible training in today's university research laboratories through direct participation in the leading research of the day. There simply is no substitute for this type of training. A decline in support for research universities would terribly weaken the foundation for the basic research enterprise of the future.
Problems and solutions in the administration of federal grants for university research
Although federal funding of basic research in universities has been very successful, there are several aspects of the grants process that are problematicfor the individual researcher and the universityand should be reformed.
There are concerns among many university researchers that the competitive grant structure has become overly-burdensome for the individual researcher and has deterred young scientists from pursuing careers in academic research. The administrative burden for the individual researcher lies in: 1) short-duration grants, which require repeated preparation and re-submission of grant applications to sustain research projects; 2) the low rate of success in applications; 3) the long time period between the submission of applications and the notification of awards; 4) the very high dependency of the individual investigator on external support and the lack of seed money. To alleviate these problems we urge grant-making agencies (and universities) to review their requirements and systems of support with an eye toward reducing administrative burden.
Grant-making agencies should explore ways to offer outstanding scientists longer-term grant support, to provide young researchers with sufficient resources to launch their careers, and to ensure that established researchers who temporarily lose grant support are not forced to abandon long-term, productive research endeavors. Institutions like the Howard Hughes Medical Institute and the Scripps Research Institute provide cues for grant-making institutions and universities in this effort. In particular, we urge agencies to consider extending the funding period of their basic research grants. The continual process of applying and re-applying for grants has a very discouraging effect on young scientists and engineers considering careers in university research. Longer and more stable periods of grant support would alleviate this problem.
Many funding agencies have already begun the process of limiting administrative burden in grant awards and have made some progress. For example, some agencies now permit electronic submissions of grants. However, a large and persistent cause of administrative burden results from a cumbersome system of overhead, or indirect cost, reimbursement in the federal grants process. Federal research grant awards to university researchers include funding to cover the direct costs of the research and overhead research costs incurred by the university, reimbursements for which vary considerably from grant to grant. From the university's perspective, there are large costs associated with the investments in plant, equipment, and human resources necessary to pursue individual research grants and sustain a research enterprise at the university. The universities' tuition income and other funds received for education are designated strictly for that function. Without overhead reimbursement, universities would not have the necessary level of resources or incentives to sustain a vibrant basic research enterprise.
But the current retrospective, cost-based procedures for establishing federal reimbursement for overhead have resulted in a fractious relationship between universities and government. Present procedures are very costly and time consuming and, according to some experts, distort incentives for efficient expenditures.6 To remove these distortions, and much of the unnecessary regulatory burden, they recommend a system in which overhead rates associated with all research grants are set by benchmarks, or average overhead rates for similar universities. The benchmarks would be determined on a periodic basis by examining costs at a small sample of universities within each class of institution. Such a procedure would greatly reduce accounting and auditing costs for universities. It would also minimize government leverage, which universities consider excessive because of the government's power to demand lower reimbursement levels. Indeed, universities have argued forcefully that cost-shifting has escalated in recent years, with federal agencies increasingly unwilling to adequately cover overhead costs. Finally, benchmarking would provide universities with a strong incentive to trim overhead, because they would benefit from holding costs below the benchmark.
CED believes that the benchmarking of overhead reimbursement rates has merit in principle and encourages funding agencies and universities to explore it on an experimental basis. The determination of "similar universities" is a difficult task, which should take into account geographic variations among other factors. In general, reform of indirect cost reimbursement should be guided by the principles of fairness (to both parties) and simplicity.
Too many research universities?
Some observers are concerned about the number of research universities vying for federal funds today. As this number grows, basic research resources are spread more thinly and may be put to less effective uses. We recognize this possibility, though we do not support attempts to set a fixed number of research universities eligible for funding, or worse yet, a fixed list of specific universities eligible for funding. Historical experience indicates that it is very difficult to predict who will make discoveries and where discoveries will take place. Further, such "fixes," in our minds, violate the principles of flexibility and competition that have defined the success of America's university research system.
CED believes the appropriate number of research universities will best be determined by reinforcing existing mechanisms of peer review, as well as by distributing funds at the project and investigator levels rather than at an institutional level. Competition for grants based on scientific merit will ultimately separate the wheat from the chaff among universities. Those that do not attract quality researchers will not be able to support a research program through federal grants. And the openness of the competition to researchers regardless of institutional home ensures accessibility to high talent wherever it is based.
Unfortunately, university research that is funded outside of the peer review system is not subject to this important competitive process. Again, political earmarks are not an appropriate funding mechanism to support universities or any other research institution. University research supported by earmarks would not likely meet the peer review standard, and a proliferation of research universities due to earmarks would undermine the quality of our basic research enterprise.
Finally, although we do not support a "top-down" approach to limiting the number of research universities, we also do not believe that all universities and colleges in the United States should feel an entitlement or an obligation to pursue federal research dollars. Competition for research dollars is healthy for the basic research enterprise; it is not always in the best interest of individual universities and colleges to enter this competition. In particular, schools should not neglect their education mission in the process. For many institutions, the pursuit of a research function comes at the cost of undergraduate education, which is not critical not only to the basic research enterprise but to our economy and society as a whole.
Existing alternatives to research universities are weak
Although CED supports basic research funding through a diversity of missions and agencies, we question the degree of success that some agencies have had in supporting basic research. We believe that the further removed agencies are from merit-based, competitive grant system of allocation, the poorer the science will be which they support. The quality of research supported by the Department of Agriculture has suffered for this reason.7 In fact, just 5.4 percent of the Department of Agriculture research budget goes to nationally competitive research grants. If pay-offs to our public investments in agricultural research are going to improve in the future, this percentage must increase substantially.
The Department of Energy's national labs face an uncertain future due, in part, to inadequate mechanisms for determining merit, but also due to shifting missions. It would be a mistake to view the end of the Cold War as the end of the mission justification for the national labs. However, at the prodding of the Administration and Congress, many labs are increasingly chasing the technology du jour, with over half of their research dollars now devoted to commercial product development, often through industry partnerships.8 Although industry partnerships with the national labs may serve the competitive interests of specific industries in the development and commercialization of technologies, they do not always serve the interests of the nation and its taxpayers. As we argue later in this chapter, subsidizing civilian technology for national competitiveness purposes is not a justifiable federal mission and it should not be allowed to displace federal investments in basic research. Further, the Intel case described in Chapter 3 illustrates the political morass that such partnerships can create when mission justification is weak, raising questions of favoritism among companies, as well as concerns about indirect federal subsidies of foreign companies.
At the same time, the national labs continue to represent a tremendous potential resource for the nation's basic research enterprise, particularly oriented toward large-scale scientific inquiry. We call on the Congress and the Administration to make a clear determination of the missions of the labs and assess where realignments of the missions and functions are necessary. The recommendations of the Galvin Commission should serve as a starting point for this work. The labs themselves should be free to take actions to ensure the best researchers are attracted and retained to accomplish lab missions. In particular, more of the research conducted in the national labs, including the large-scale science that they have the capacity to perform, should be brought into a system of competitive, merit-based peer review that characterized the best of our basic research enterprise.
Beyond scientific merit and mission justification, cost-efficiency in these labs must be improved; performing basic research in the national labs should not cost more than similar projects in other basic research institutions. To this end, research projects performed in the national labs should be free of the multiple layers of micromanagement emanating from the Department of Energy and Congress that have created gross inefficiencies and rigidity in the labs.
In sum, it is clear to us that if the national labs are to continue to play a productive role in basic research, that role must be justified on the basis of strong missions, outside peer-reviewed determinations of scientific merit, and efficient management and oversight structures.
Preserving Our Capacity to Do Basic Research in the Future
Future capacity to do basic research requires long-term thinking today. Unfortunately, two threats to tomorrow's basic research enterprise currently command too little attention from political leaders and policymakers.
One is funding for basic research. Today's federal budgetary environment, with rosy scenarios for "budget surpluses as far as the eye can see," has created a decidedly optimistic attitude within the science and engineering community (and among its advocates in Washington). Speaking of a threat to basic research funding in the environment may strike many as odd if not downright foolish. Yet, the likelihood of stable and adequate support, not just for the next few years but for the next few decades, is far from ensured.
The second area that requires long-term thinking today is in sustaining the human capacity to do basic research. Sustaining the education and employment pipeline for basic research also means raising the technical and scientific ability of all student and society at large. This effort is important both so our citizenry can better exploit the increasingly flow of new knowledge in an increasingly sophisticated workplace and also so the importance of basic research is not lost by a society that is less connected to progress of science.
Ensuring Adequate Resources for Basic Research
Throughout this chapter we recommend ways that basic research resources can be used more productively. But none of these policy recommendations should be taken as support for getting along with less public support. As we described in Chapter 2, the economic and social payoff to American investments in basic research has been tremendous in this century alone. Given this level of benefit, adequate and sustained funding for basic research must be a high and consistent national priority.
With this in mind, we are deeply concerned by trends that have the potential to squeeze the level of resources for basic research in the future. Both the Administration and Congress now claim a desire to increase basic research funding substantially over the next few years. It remains to be seen, however, how these increases will materialize and if they be sustained. Windfalls derived from a booming economy or a hypothetical tobacco settlement would certainly be welcome additions to basic research funding. They do not, however, eliminate political and budgetary imbalances that stand as threats to future funding.
Recognizing the Role of Private Sector Support for Basic Research
Advocates of a smaller government role in basic research point to increasing private sector R&D budgets as evidence that the federal government's efforts are no longer as important in this area. Industry is doing more, the argument goes, therefore government can do less. However, private spending on total R&D should not be confused with spending on basic research, nor should industrial basic research be confused with government-supported basic research. As indicated in Chapter 3 and as the Pfizer case study suggests, although the "Web of R&D Innovation" is a complex one, with complex interactions between basic and applied research, it remains clear that the private sector role in basic research, and in R&D in general, is and will continue to be largely distinct from the government's role.
In sum, industry will continue to support and perform important areas of basic research; but this work should not be viewed as a substitute for the much larger role of the federal government in support of basic research. Given these trends and the misunderstanding of them that is prevalent among some political leaders and the public generally, we urge business to begin a dialogue with political leaders and the American public so that they might better understand the critical importance of steadfast government support of basic research.
Choosing Basic Research Among Competing Claims
Meeting the entitlement threat
Despite robust political support for short-term funding increases, support for basic research is likely to become less sure in years to come, due to the budgetary costs of an aging population. Basic research is one of many competing claims on a shrinking discretionary portion of the federal budget, necessitated by rapid expansion of Social Security, Medicare, and Medicaid. The considerable revenue boost the federal budget has received from the current expansionary economic cycles does not erase the underlying structural deficiencies in the budget. This is a problem that will become dramatically worse the longer we wait to deal with it; indeed, we may ultimately face devastating cuts in the basic research budget if entitlement spending is not brought under control before the baby-boom generation retires. As CED has argued frequently in recent years, our political leadership simply must deal with the growing burden of the federal entitlement programs.9 Otherwise, the federal budget will have no room left for the government's most important activities, including investment in basic research.
Federal funding for civilian technology
One of the competing claims that basic research faces in the federal budget comes in the form of spending on applied and development research, much of which is necessary to achieve objectives in various missions, such as the development of weapons systems technology for national defense. A significant initiative of recent years, however, has been to seek to improve the international competitiveness of U.S. industry by increasing federal expenditures in certain areas of applied research and civilian technology development. The Advanced Technology Program in the Department of Commerce is a product of this initiative.
CED does not believe that national competitiveness programs have the same strong claim to federal support that basic research does. Although proponents of competitiveness programs often use the same language to make their case-that public subsidies are necessary to correct for insufficient levels of funding by the private sector in key areas (i.e., "market failures")we find the case far from convincing. Indeed, too often government spending on programs of this type amounts to little more than taxpayer subsidies for favored industries and firms, supporting research that the private sector would have supported on its own or research that is not worthy of public or private funding.10
With few exceptions, we do not think government should be in the business of directly funding what we view to be a function of the private sector-development and commercialization of technologies. The exceptions apply in cases where the funding can be viewed strictly as a procurement function, as in the national defense example, or to correct a clearly defined and well substantiated market failure.
Finally, CED believes that the government should be sensitive to the basic research activities occurring in industry and should avoid duplicative initiatives. Federal funding is most vulnerable to duplication when research initiatives are overly-prescribedas in research that is targeted at specific diseasesand removed from a general peer-review process.
Sustaining the Education and Employment Pipeline
In addition to the funding of basic research, a second threat to our long-term basic research capacity lies in our ability to sustain a pipeline of future scientists and engineers, a work force adequately skilled in the sciences to exploit new knowledge, and a citizenry with enough understanding of basic science to recognize its importance and support its progress.
The United States is not in danger of running out of scientists and engineers to perform basic research any time soon. After all, they comprise only a tiny fraction of the nation's work force: just 542,000 are doctoral scientists and engineers.11 But there is a growing disconnect between the need for a highly-skilled basic research labor force on the one hand, and the quality of our K-12 math and science education and the interest in science at the undergraduate level on the other hand. Further, the desirability of basic research employment at universities is eroded by the amount of time devoted to applying for and complying with federal grants, as well as the demands of university technology transfer offices. Left unchecked, we are concerned that these trends will gradually erode the base of American students willing or able to enter basic research careers.
These trends also exacerbate the on-going challenge of attracting women and minorities to training and careers in basic research. The severe underrepresentation of these groups in the science and engineering disciplines has far-reaching effects on our basic research enterprise; in part, it contributes to concerns that the peer review process is biased against some researchers and research projects. A more diverse research enterprise has been identified as a priority by universities, agencies, and science organizations; we support this goal and urge these institutions to hasten its realization.
Finally, we are also concerned about a society that is increasingly isolated from the world of science and discovery. To a large extent, this isolation is driven by science itself. Long gone are the days when a well-educated citizen could be expected to have even a cursory grounding in all areas of knowledge (scientific or otherwise). But the isolation we are particularly concerned about involves a lack of understanding and acceptance of scientific methods and principles. This deficiency is a detriment to the skill needs of today's workplace. It also threatens to undermine the public support for basic research, an enterprise that relies primarily on the public sector for support. If society becomes less enthusiastic about science and more suspicious, or simply indifferent, the case for research support will become much more difficult to make in the halls of Congress.
Based on all of these concerns, we offer the following recommendations to strengthen the science education and employment pipeline and science education in general, not only for future scientists but also for a public that must ultimately support them.
Better math and science education at the K-12 level
In a series of policy reports on education improvement, CED's trustees have done extensive research into strategies for raising academic achievement in general and science and math performance in particular.12 The complexity of this task requires comprehensive and coordinated change in several interdependent areas:
We maintain our strong support for high achievement standards at the national level in all core academic subjects, with particular emphasis on mathematics and science. However, given the inherent difficulties in getting national standards in these subjects developed and accepted, we urge teachers and administrators to actively and continuously pursue information and research on new knowledge and effective, innovative classroom practices and to redirect the mission and goals of their schools toward raising standards for student performance.
Improved learning cannot happen without improved instruction. Therefore, our schools need to both attract and continuously support better-qualified math and science teachers, particularly at the middle and high school levels. Strategies to accomplish these ends include:
Finally, it is critical that students be actively engaged in learning in the sciences and math. We believe that improved teaching methods, together with currently available interactive learning technologies, including computers and Internet connections, can effectively motivate student performance and accelerate improvements in learning. Substantial investment in infrastructure improvements are also critical in this regardmany schools have little or no lab facilities to support science instruction.
In order to further engage students in learning and to excite students about a possible career in the sciences, they should have the opportunity to interact with members of the research community. Businesses, universities, and schools work together to place more professional scientists and engineers in the classroom as volunteers to assist in class lectures, lab work, and field trips. These volunteers raise the quality of instruction, demonstrate to the students the future job benefits of a science and math education, and communicate the role of science and math in the world today. Similar initiatives give students and teachers the opportunity to venture out of the classroom and explore the nature of basic research in a professional setting.13 High-quality and engaging videos and films can support this objective, while also overcoming geographic, time, and resource constraints that might not allow for in-person interaction.
Improving graduate training
Graduate-level scientific training is perhaps the most important segment of the educational pipeline for basic research. A science education that is 16 years in the making can continue on to a career in basic research or can be diverted to some other field, depending on whether or not the college graduate decides to complete graduate training in the sciences. For those choosing a future in basic research, their impact as scientists and engineers depends a great deal on the quality of their graduate training.
Good scientific training is grounded in hands-on exposure to basic research and scientific inquiry. It is essential that graduate students have access to hands-on research projects. Thus, training of graduate students should be a paramount criterion for universities and faculties in considering their research portfolios, particularly concerning the nature of the research.
The federal government can help to make graduate student training a higher priority in the research university by increasing the amount of scholarships and training grants available to students. Grant support that goes directly to the student, versus support that comes directly from grants for research projects, clearly places the priority on the student's training rather than the needs of any particular research project.
Prolonged time to degree in graduate training is another concern, although admittedly one that is not well understood. To the extent that more time spent attaining the graduate degree reflects a great complexity of the field of study, there is little to be done and the increased time likely produces a more knowledgeable and productive graduate. But it is undesirable for prolonged time to degree to be caused by higher student-to-faculty ratios, excess time spent on assisting faculty research, or difficulty in securing stable funding.14 The direct burden and opportunity costs of a longer time to degree are born by today's student, and will also likely have a deterrent effect on those considering graduate training in the future. We urge research universities to undertake frank and self-critical examinations of their graduate training programs in an effort to expedite the time to degree and reduce the financial and time burden on their graduate students. Again, greater investments in graduate education by science and engineering-related federal agencies can also play an important role in reducing this burden.
Academic employment is at the core of basic research and will remain so. But an increasing number of Ph.D. scientists and engineers are finding roles, whether by choice or due to a lack of academic alternatives, in the private and other non-academic sectors. Private sector employment of Ph.D. scientists and engineers plays an important role in the dissemination of scientific knowledge. In order to facilitate the transition from academic training to private sector employment, graduate schools should offer more training programs and mentorships in their curricula that prepare and orient students for employment outside of academe.
Team-oriented and cross-disciplinary workan important component of these effortsneed not detract from core disciplinary training. In fact, the Pfizer, Harvard University, and BBN case studies all suggest that multi-disciplinary, cross-functional project teams are an increasingly important element of success in many areas of basic research, regardless of whether it is private sector or university research. At the same time, some university research administrators have questioned whether there is adequate grant support for cross-disciplinary and team-oriented research, given the discipline-oriented nature of current grant systems. CED urges grant-making agencies to support more team and cross-disciplinary efforts in research projects.
CED notes that several recent reports address the graduate training issues discussed here, and that reforms in graduate education in the sciences and engineering are underway.15 Agencies such as the National Science Foundation are responding to the recommendations made by the National Academies and others. Universities are also participating in the renewal of graduate education in the sciences and engineering.
Academic employment of young researchers
Although there is an important role for private sector employment in sustaining the basic research "pipeline," academic employment of scientists and engineers remains central to the health of the basic research enterprise. The trend away from full-time faculty positions at research universities and the increased reliance on post-doctoral, part-time, and temporary employment are not positive signs. They send the wrong signals to today's college and graduate students as they embark on the long and rigorous training necessary for a career in basic research.16 Hearing horror stories of post-doctoral training that extends well into a young researcher's career, some prospective scientists and engineers are likely to opt out of this career path. Getting young researchers out of temporary positions and into stable employment should be a priority for all research universities.
As we discussed earlier in this chapter, academic employment also carries with it a large administrative and grant-raising burden. Research universities and the federal agencies that sponsor universities should be exploring ways to make the academic scientist's work environment more stable and more productive, with less time spent on raising money.
Principles for University Research in the Marketplace
American research universities have always been oriented toward practical concerns, while maintaining excellence in the most theoretical and exploratory corners of science. The transfer of knowledge and new technologies from the university to industryas embodied in graduates entering the work force, through open dissemination of research results, or through private partnerships, patents, and licensingis of great economic benefit to the United States. But universities must walk a fine line as they seek to increased the economic value of their basic research, while maintaining a public mission defined by openness and wide dissemination of research findings. This balancing act is made more difficult by the problems researchers face in obtaining public grant support, as indicated earlier in this chapter. CED believes universities, industry, and the federal government should adhere to the following principles as they pursue relationships in basic research .
1. The primary channel of benefit from university research to industrial innovation, and to society at large, is through wide and open dissemination of knowledge in research journals, at conferences, and by the education of graduate students. New knowledge generated from university research should continue to be openly disseminated. The publication of research findings, upon which future research frequently depends, should be only minimally delayed by patent preparation and other requirements of sponsored research agreements.
2. Pursuit of patent protection on university inventions, initiation of university-industry partnerships, and licensing of intellectual property for commercial development can stimulate important new research and facilitate and expedite the transfer of new technologies to industry, ultimately benefiting society at large. Therefore, such partnerships and licensing activities, when structured appropriately, should be encouraged.
3. These relationships can create conflicts of interest among the universities, their faculty, and collaborating industrial concerns. Universities should be strongly encouraged to develop, and continually improve, policies and procedures for technology transfer and industry relations, such that the basic educational and research mission for the university is neither diluted nor compromised.
4. In addition to the need for universities to make research results broadly available, these institutions also have an obligation to devise licensing agreements for technology developed using public funds in a fashion that generates the greatest benefit for society. The guiding principle should be to prevent the university technology transfer system from impeding further research advances, wherever they may occur. As a general principle, research advances that can be turned into distinct proprietary products are appropriate candidates for exclusive licensing and development. On the other hand, technology that can serve as a tool for many researchers to create new knowledge should be made widely and nonexclusively available under commercially reasonable terms. Industry should recognize and respect these distinctions.
5. The success of America's basic research enterprise will continue to depend on the use of patented inventions in basic research. In general, basic research will benefit if a patent holder's rights (whether the owner is a company or university) are not enforced in a way that restricts further basic research. Access of all parties to tools for basic research is particularly important in cases where federal funding supported the initial discoveries. When a research tool is also a research product (as may be the case in areas of biomedical research), the interested parties should work out terms that, first and foremost, do not prevent broad future use of the research tool, and secondly, permit use of the product in the marketplace. Industry and universities should continue to seek avenues which allow the fewest possible restrictions on use of patented inventions for new basic research.
6. The major research universities have developed policies that honor the public responsibilities of their research mission. These policies are particularly important in maintaining the central purpose, value and morale of the faculty who are the lifeblood of the university. In particular, it is important that the interests of industry not influence critical decisions of the university regarding hiring, compensation and promotion. The continued strength of basic research in America depends on the ability of faculty members to devote their energies to, and be evaluated by, their quality and contributions as basic researchers and their success in educating new scientists. Commercial successes are secondary to these missions.
7. Finally, private sector funding of university research supported through licensing fees should not be viewed by universities or the federal government as substitutes for federal funding. The objectives and goals of private firms are different from public missions and should not be allowed to define the character of university research in general. Ultimately, the only way to maintain basic research in the university as it has been characterized over the past half-century is through predominantly public funding.
International Challenges and Opportunities for American Basic Research
Maintaining a Commitment to Basic Research in a Global Economy
Basic research activities around the world are increasing as other countries step up the pace of their research investments. At the same time, the new knowledge derived from basic research moves more easily across national borders than ever before, thanks to the dramatic advances in information technology. Transfer of information is also facilitated by the steady pace of economic globalization.
These trends have led some to questions the viability of our national basic research enterprise in an increasingly international economy. In fact, some argue that because the products of our basic research now move freely and quickly to other countries, or to foreign subsidiaries of U.S. firms, U.S. taxpayers should not pay for it. They argue that the United States should take better advantage of other nations' basic research, or should attempt to limit foreign access to new American knowledge. CED strongly believes that arguments for reducing or restricting access to the outcomes of U.S. efforts in basic research, based on fears of foreign appropriation if American discoveries, are misguided and could undermine the central role that the federal government plays in funding American basic research.
We readily acknowledge that other countries benefit from our basic research. In some cases, U.S. firms have been slow to capitalize on technological breakthroughs. This reflects poorly on U.S. technological development and is not a failing of the basic research enterprise.
The foreign benefit from American basic research need not be our loss. Performing initial basic research should position us to exploit benefits faster than can the followers and the "freeloaders." If we reduce our efforts in basic research, we will lose this "first mover" advantage. Also, by promoting an open two-way environment for new knowledge, we are able to share in other countries' basic research findings, even to the point of exploiting those findings more quickly than firms in those countries.
Finally, a robust basic research enterprise in American universities keeps our high-tech firms on-shore, while also attracting foreign firms to invest within our borders. The scientists and engineers at our research universities are an enormous international strategic advantage.
In sum, more basic research activity around the globe will present tremendous opportunities for scientific progress, with large payoffs for all nations. Nationalistic attitudes and policies are counterproductive to this progress and have no place in a productive global basic research enterprise. It is critical, then, that the United States take a leadership position in ensuring the free flow of fundamental knowledge and basic research findings globally.
Intellectual Property Worldwide
Although CED does not view the internal spillover of non-proprietary knowledge derived from basic research as a negative, we are concerned about the impact of weak intellectual property laws in other countries on innovation in the United States and globally. U.S. patent laws play a very large role in stimulating innovative activities in this country by protecting a company's rights to discoveries and innovations that are proprietary in nature; in this way, the company is able to capture the returns on its investment. In an increasingly global environment and one in which a growing number of countries are investing in research, it is in the interest of all countries to play by the same rules regarding intellectual property. The pirating of intellectual property that occurs in other countries ultimately hurt all levels of innovative activity, from basic research to commercial development.
Pursuing International Collaboration
Some areas of science have become too expensive and too risky to be supported solely by a single country. Because of their ultimate potential benefit, such projects deserve international cooperationin funding, institutional collaboration, and sharing of scientists and results. Mistrust between countries in such projects ignores the reality of global science today. A great deal of scientific information already flows freely across national borders through publications, research conferences, and through the Internet and other forms of information technology. By cooperating in large-scale projects, the United States and other countries extend globalization to projects that are not sustainable at a national level. CED believes the United States must pursue international cooperation in "big science," in order to optimize advances in science and technology.
Supporting Foreign Scientists and Engineers in the United States
In an increasingly global scientific community, American training of foreign scientists is another welcome form of international research cooperation. The participation of foreigners in American universities and in American science and engineering programs benefits the U.S. research enterprise and the economy in general.17 Foreign students contribute to our basic research through participation in university research. Given the capacity of the American research university system, it is unlikely that these students are significantly "crowding out" high-potential American students. Many foreigner graduates will choose to remain in the United States as professional scientists and engineers (about 40 percent in recent years).18
The direct participation of foreign students and professionals in American basic research will likely decline in the future, as foreign economies become more sophisticated . How the United States reacts to these trends is critical to the future of our own basic research enterprise. CED believes the United States should continue to make reasonable efforts to attract foreign scientists. Our immigration policies should be further liberalized to allow foreign scientists and engineers to live and work in the United States through permanent visas. Also, immigration policies should encourage foreign scientists and engineers to visit the United States on a temporary basis as consultants or participants in research collaborations.
We are in the midst of a remarkable period of discovery and innovation. Ask those engaged in the discovery process what they think about the future of their particular discipline and the answer is likely to be strikingly positive, with postulations about breakthroughs that are unfathomable to us today. Add to this optimism the impact of the current economic boom, which is making the most of product innovations and brining new technologies to market at a dizzying pace. Together, these trends create an atmosphere of almost boundless enthusiasm about the future.
Raising a cautionary note about the future of the basic research enterprise in this environment, then, is no easy task. Certainly, we do not foresee calamity on the horizon. The United States will continue to see pay-offs from its basic research investments, as it has throughout this century.
Yet, today's emerging problems, left unchecked, become a potent threat to tomorrow's research enterprise. In this sense, the reforms we have recommended in this reportshoring up the system of competitive peer review, maintaining adequate resources, and correcting deficiencies in the educational and employment pipeline, to name a few-are more than just fine-tuning. Problems in each of these areas have the potential to erode the quality and quantity of output we have come to expect from American basic research.
Dealing with these challenges requires resolve. Certainly, if reforming the national laboratories is politically difficult, K-12 educational reform is even more so. The key for policymakers, and for the citizens they represent, is to keep their focus on the potential for the future. The pay-offs from a basic research enterprise that is working at its full capacity are tremendous, and far too great to forgo for lack of political will.
1. National Science Board, Science & Engineering Indicators1993, (Washington, DC: U.S. Government Printing Office), 1993, p. 139.
2. Walter Powell and Jason Owen-Smith, "Universities and the Market for Intellectual Property in the Life Sciences," Journal of Policy Analysis and Management, Vol 17 No (2), P. 268.
3. As we argue later in this chapter, some projects prove to be too expensive to support at the national level and should be pursued in conjunction with other countries.
4. The President's 1999 budget proposal would increase the funding disparity. Of the proposed $1.2 billion increase in basic research funding between 1998 and 1999, $616 million (just over half) would go to the NIH basic research budget. This would raise NIH's share of the federal basic research budget slightly, from 46 percent to 47 percent. See AAAS Analysis of R&D in the FY 1999 Budget, available at www.aaas.org on the Internet.
5. As described in Appendix 1, Medicare and Medicaid have been important sources of funding for research at university medical centers and, indirectly, a significant source of funding for university research in general. Structural reforms in these programs may result in less research funding from these resources.
6. See Roger Noll and William Rogerson, "The Economics of University Indirect Cost Reimbursement in Federal Research Grants," in Challenges to Research Universities, ed. Roger Noll, (Washington, DC: Brookings Press, 1998). Noll and Rogerson provide an analysis of the cost reimbursement problem and justification for a "benchmarking" alternative to the current system.
7. National Academy of Science, Allocating Federal Funds for Science and Technology, Committee on Criteria for Federal Support of Research and Development, (Washington, DC: National Academy Press, 1995), pp.2526.
8. National Science Board, Science & Engineering Indicators1996 (Washington, DC: U.S. Government Printing Office, 1996), NSB 96-21, pp. 4-28.
9. CED has recently proposed remedies for the Social Security component of entitlements. See Fixing Social Security (March 1997).
10. See Kenneth M. Brown, Downsizing Science: Will the United States Pay the Price?, (Washington, DC: The AEI Press, 1998).
11. Data are for 1995. NSF, NSF Data Brief, 1997 no. 3 (March 13, 1997); U.S. Department of Labor, Monthly Labor Review (January 1998), Table 43.
12. See the CED policy statements Putting Learning First: Governing and Managing the Schools for High Achievement (1994), Connecting Students to a Changing World: a Technology Strategy for Improving Mathematics and Science Education (1995), and The Employer's Role in Linking School and Work (1998).
13. Some have suggested that agencies that sponsor research (such as the NSF, NIH, and the Department of Defense) could provide a significant boost to mentoring, internship, and other educational outreach initiatives by providing additional resources to research institutions for these purposes.
14. For a recent discussion of this issue, see Committee on Science, Engineering, and Public Policy, Reshaping the Graduate Education of Scientists and Engineers (Washington, DC: National Academy Press, 1995).
15. Committee on Science, Engineering, and Public Policy, Reshaping the Graduate Education of Scientists and Engineers. See also Committee on Science, Engineering, and Public Policy, Advisor, Teacher, Role Model, Friend, On Being a Mentor to Students in Science and Engineering, (Washington, DC: National Academy Press, 1997).
16. Equally discouraging for new doctorates and those considering doctorates are certain aspects of today's academic research environment. The administrative and grant-raising burden placed on university researchers has become a formidable disincentive to pursuing academic research careers. See "Problems and Solutions in the Administration of Federal Grants for University Research," page 36, for recommendations in this area.
17. Immigrant scientists have long been very important to American basic research, particularly during and following World War II. Between 1933 and 1970, 27 of the 77 Nobel Prizes awarded to the United States went to first-generation Americans (Eric Hobsbawm, The Age of Extremes, New York: Pantheon Books, 1993).
18. National Science Board, Science & Engineering Indicators1996 (Washington, DC: U.S. Government Printing Office, 1996), NSB 96-21, Appendix Table 2-35.
Reprinted with permission from America's Basic Research: Prosperity Through Discovery, pp. 3247. Copyright Committee for Economic Development, 1998, New York, NY.