agency sponsors decline. By the same token, fields primarily funded by agencies with rising budgets do not necessarily fare accordingly. In the FY 1993–1997 period, medical sciences (up 14.4 percent) received far more from growth in the National Institutes of Health budget than did the biological sciences (up only 1.1 percent).

  • In the constrained budget environment of the mid-1990s there was no consistent pattern of protecting support of university research relative to in-house research and research performed in the corporate sector.

  • No single agency was serving as a “balance wheel” to ensure some stability of funding in fields whose support is declining elsewhere. In the 1990s, NSF, with the broadest research portfolio, appeared to be amplifying changes in other agencies, in most cases boosting funding for fields prospering elsewhere and reducing funding for fields being cut elsewhere.

  • In the cases where direct comparisons can be made because of identical field nomenclature in different NSF surveys, changes in university research funding of a field corresponded to changes in the number of graduate students supported by federal fellowships, traineeships, and research assistantships in that field. Where research funding was down (e.g., chemical and mechanical engineering), the number of graduate students also declined. Conversely, the number of federally supported graduate students in computer science increased, as did federal research support.

Based on those findings, the Board expressed its concern about the reduction in federal investment in fields important to such industries as electronics, software, and materials processing and concluded that the trends in federal funding, if they continued, merited “a careful assessment of their long-term implications and what steps, if any, should be taken to change them.”6 At about the same time, concern began to increase about the possible “imbalance” in the federal research portfolio based on the divergence between the declining support of the physical sciences and engineering and the growth of funding of biomedical research through the National Institutes of Health.

In the fall of 2000 the STEP Board decided to revisit its analysis and conclusions for several reasons.

The Board’s first motivation was the availability of additional data that were otherwise unlikely to be presented and interpreted in a form useful to policy makers. As a result of efforts by the NSF’s Science Resource Studies (SRS) Division to accelerate the availability of results of the Federal Funds and other surveys, data on agencies’ research obligations are now available within approximately 15 months of the end of the fiscal year. This enables examination of federal agencies’ spending on nearly two dozen research fields in FY 1998 and FY 1999 and their reasonably reliable estimates for FY 2000 by major field of research. Surprisingly, however, none of these data have been reported in the NSF’s publications of science and technology statistics—Science and Engineering Indicators, National Patterns of R&D Resources, or, with few exceptions, SRS’ periodic Data Briefs.7 Nor does the American Association for the Advancement of Science examine the subject of field allocation in its annual analyses of the current fiscal year federal R&D budget.

A second motivation was curiosity about the effects on allocations among research fields of the marked turnaround in federal research funding in FY 1998. After 5 years of stagnation, FY 1998 research expenditures were up 4.5 percent in real terms from 1993 and even more (11.7 percent) in FY 1999. Even DOD’s research budget showed modest increases over FY 1997 (although it was still much smaller than in 1993), and increases in FY 2000–2001 federal R&D budgets ensure continuing incremental growth through most of 2001. In this improved funding environment, the question arises whether the disparities in how research fields fared in the mid-1990s have been eliminated or moderated.

Third, recent articulations of the importance of “balance” in the publicly supported research portfolio by a number of executive branch and congressional policy makers makes the question of the relative growth in funding among research fields when R&D budgets are increasing even more compelling. What has been the impact of officials’ greater attention to how federal research money is being spent and their declared intention to correct any “imbalances”?

Finally, the Board decided to extend the analysis of federal research spending in certain respects. First, although this report deals primarily with changes in (basic and applied) research spending through 1999, it also examines trends in basic research support and research conducted at universities by field.8 Second, the report looks in greater detail at the relationship between research funding and graduate student support by research field. Third, the report identifies which fields changed their structure of support (principal agency sponsors and their shares) in the 1990s and which did not and with what results. Finally, it considers trends in the composition of research support from nonfederal sources, principally states, philanthropies, and industry, to cast some light on the question of whether other sponsors of research are


National Research Council. 1999. Securing America’s Industrial Strength, pp. 4. Washington, D.C.: National Academy Press.


An exception was Alan I.Rapoport. Feb. 17, 1999. “How Has the Mix of Federal Research Funding Changed Over the Past Three De-cades?” Arlington, VA: National Science Foundation.


Occasionally the report refers to trends in research and development expenditures, especially with regard to budgets after 1999 for which separate figures for research are not available.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement