|
PITAC - Report to the President
Setting Federal Research Priorities: Findings and Recommendations |
 
|
|
|
|
|||
 |
|
||||||||
|---|---|---|---|---|---|---|---|---|---|
|
|
||||||||
|
|
|||||||||
2. Setting Federal Research Priorities
|
Setting Federal Research Priorities: Findings and Recommendations Government-funded information technology research has produced enormous innovation. The results have been made readily available to industry. Testbed activities involving academia, government research facilities, and industry have served as powerful engines for technology transfer into the private sector, for the benefit of industry, government, and the Nation. Today we enjoy the economic, strategic, and societal benefits of well-placed investments in long-term, wide-ranging information technology research begun during the Eisenhower and Kennedy Administrations. These modest investments have yielded massive economic benefits to the Nation. The empirical evidence is unequivocal: today's information technology sector vastly outstrips the current growth of all other sectors of the economy. The Federal Reserve reports that during the past five years production in computers, semiconductors, and communications equipment quadrupled at a time when total industrial production grew by 28 percent. These three industries account for one-third of the total real growth in production since 1992. No other sector contributes nearly as much to the growth of our economy. The businesses spawned by these technologies employ millions of Americans in manufacturing and information processing jobs that pay wages well above the national average. History suggests that, to be successful, Federal research investment must be sustained and flexible. Federal policies must support, encourage, and help coordinate long-range technological development. Federal R&D programs must be well designed and must not subsidize activities best left to the private sector. Only in this way can the Federal investment spur those critical areas of technology that either industry neglects, or, which the Government overlooks in the normal course of business because they cut across Federal agency missions. |
||||||||
|
|
|||||||||
|
2.1 Findings
|
Findings Finding: Total Federal information technology R&D investment is inadequate. The non-inflationary growth in high-technology businesses fuels U.S. world leadership and creates whole new industries with hundreds of thousands of high-paying jobs. Furthermore, information technology's importance goes far beyond its economic benefits: it is fundamental to the solutions of many nationally important problems. But the Federal R&D investment has been compromised by a shift toward applied R&D. The Committee further finds that the amount of Federal research investment in information technology has not kept pace with information technology's growing economic, strategic, and societal importance to the Nation. The number of proposals in computer science and information technology competitions provides evidence of the shortage of research funding. For example, the recent competition in Knowledge and Distributed Intelligence sponsored by NSF drew over 1100 letters of intent and over 850 full proposals even though it had been announced that at most seventy-five projects could be funded. The response to the ASCI level 2 centers competition was similar. The reason for the extraordinary number of responses is that funding for information technology is extremely tight. Researchers are forced to participate in nearly every competition for which they might qualify. The end result is that researchers are spending increasing percentages of their time in proposal writing to the detriment of research itself and yet many good ideas are still not being funded. This trend is not just bad for information technology researchers, it is bad for the Nation. Our ability to produce reliable software, build an information network on which the Nation can run, produce the high-end computing systems needed for advanced science, engineering, and defense tasks, and understand the socioeconomic effects of these revolutionary technologies is threatened by inadequate investments in research and development. By neglecting research, we do more than deplete the stock of fundamental knowledge -- we endanger the long-term effectiveness of the entire R&D system and threaten U.S. leadership in the emerging 21st-century information economy. Finding: Federal information technology R&D is excessively focused on near-term problems. The NSF defines basic research as the study of the "fundamental aspects of phenomena and of observable facts without specific application toward processes or products." In contrast, "applied research" is aimed at determining the means to meet specific needs, and "development" is defined as the systematic use of knowledge to produce useful materials, devices, or methods. These definitions are widely used in Federal policy and budget accounting, but they accurately describe only part of the nature of research and development (R&D). R&D is a complex non-linear interaction between concepts and theories, data and experiments, and new products and processes. Basic research is a critically important part of this interwoven system. During the past decade both industry and Government have altered the balance between basic research and the later stages of technology development and commercialization. At the same time, major corporations have cut back on basic research expenditures, shifting staff from centralized laboratories to operating divisions where applied work is closely tied to commercial products and processes. In both the public and private sectors, the interacting reasons are 1) downward budget pressures, 2) increased focus on mission, and 3) the inefficiency of transitioning long-term research to near-term product. Although total Federal spending for R&D has remained steady, there has been a marked shift toward support for applied R&D. One example is that DARPA, which funded much of the innovative research in the 1980's, revised its priorities in the 90's so that all information technology funding was judged in terms of its benefit to the warfighter. In the process of making this change it decreased the time horizon for potential technology transfer. In the early 1990's, DARPA split the Information Systems Office off from its Information Technology Office to address specific military systems. Nevertheless, total funding for basic research in the DARPA Information Technology Office is less than $20 million out of a total office budget of more than $200 million, an inadequate investment in our judgement. DARPA's new leadership plans to reverse these changes, but recent history shows that Government research managers correctly favor their agencies' missions when budgetary pressures grow and they have to choose between long-term research and short-term mission needs. Most R&D investment restructuring in the early 90's was essential for industry and the Government to maintain a competitive footing in the global marketplace and to maintain readiness against our present and future adversaries. However, this restructuring came at a high price: a serious decline in basic research activities. Research in computer science is a good example. In 1995, by the Federal Government's own measure,1 more than two out of every three Federal dollars spent research in computer science was for applied work. The Committee finds that the Federal agenda for information technology R&D has moved too far in the direction of near-term applications development, at the expense of long-term, high risk, fundamental investigations in the underlying issues confronting the field. Currently, too high a percentage of research funding comes from mission-oriented agencies whose main goal is not so much to advance the knowledge-base in information technology as it is to solve the immediate problems confronting those agencies. Meanwhile, information technology has begun moving much faster than other fields. In order to intercept tomorrow's challenges we must set long-range, stretch goals rather than focusing on incremental advances. Within information technology, the scalable information infrastructure area is moving extremely fast, making it even more important to stretch for long-term research objectives. Economic growth and defense leadership, if based on evolutionary improvements to yesterday's research results, are not sustainable -- the rest of the world is not standing still in seeking economic advantage from new information technologies. By the Committee's own calculation, basic research spending may be as low as five percent of the total Federal information technology R&D budget. As a result, promising long-term research is being passed over in order to meet the goals of near-term technology development. It is time to swing the pendulum back in the other direction and to strike a proper balance. We need more basic research -- the kind of groundbreaking, high-risk/high-return research that will provide the ideas and methods for new disciplinary paradigms a decade or more in the future. We must make wise investments that will bear fruit over the next forty years. Finding: The Federal information technology R&D funding profile is incomplete. The current funding portfolio is not properly balanced. It is deficient in the support of multiple-institution, long-duration projects. Funding for projects of longer duration and larger scope is critical to the Federal research program. Projects of larger scope allow for multiple-investigator, interdisciplinary collaboration, intramural research in academia and Federal research institutes, and joint industry-Government-academia experiments or proofs of concept. Projects of longer duration allow exploration of research problems with multiple-year horizons, which may lead to unexpected and significant discoveries. It is important that Federal investments include a range of complementary funding modes, including classical single principal investigator (PI) research, multiple-PI experimental research, and multiple-institution/multiple-year efforts. Such diversity in funding approaches and tactics is important. It provides complementary modes for research, ensuring a broad perspective in addressing problems, thus increasing opportunities for discoveries. |
||||||||
|
|
|||||||||
|
2.2 Recommendations for Research |
Recommendations for Research The current boom in information technology is built on basic research in computer science carried out more than a decade ago. There is an urgent need to replenish the knowledge base. Our legacy to future generations should be based on an intelligent, well-planned, highly disciplined investment in information technology that is commensurate with its role in the Information Age. Recommendation: Create a strategic initiative in long-term information technology R&D. The Advisory Committee recommends that the President create a strategic initiative to support long-term research in fundamental issues in computing, information, and communications. The initiative should endeavor to increase the total funding base to over two billion dollars per year after five years. The Federal funding agencies should use the resulting budget increases to encourage research that is visionary and high risk. The goal should be to recapture in the universities and research labs much of the excitement that existed at top-rated departments in the past. The Administration's proposed Federal budget for FY2000 demonstrates a commitment to sustained growth in IT research through its Information Technology Initiative (IT2). This commitment is an important first step in what must be a continuing effort on the part of the Federal government to increase research dollars and to create a new management system designed to foster innovative research. But the effort cannot stop here. Further increases and continued oversight are needed to remedy the shortfall in long-term research investments that have accrued. In order to maintain U.S. leadership, 21st Century budgets should be guided by the principle that Federal support for information technology must be increased to bring it in line with its prominence in the economy and its importance to solving critical societal problems. Recommendation: Increase the investment for research in software, scalable information infrastructure, high-end computing and socioeconomic issues. Four areas of the overall research agenda particularly need attention, and must be a major part of the strategic initiative:
Recommendation: Fund projects of larger scope and duration. We stand at the dawn of a new century, a century where leadership in information technology may well be economically and militarily decisive. To meet this 21st century challenge, we believe we must diversify the modes of research supported by the Federal government. Computer science and information technology are collaborative research fields. Hence, special emphasis should be placed on involving and supporting researchers at many institutions in large-scale research projects that can explore technologies farther into the future with teams of researchers that may be interdisciplinary and multi-institutional. Chapter 5 of this report discusses ways to diversify research support to accomplish this goal:
|
||||||||
|
|
|||||||||
|
Endnotes |
Endnotes
|
||||||||
|
|
|
||||||||
|
|
|||||||||