The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
APPENDIX A RECENT TRENDS IN FEDERAL SPENDING ON SCIENTIFIC AND ENGINEERING RESEARCH: IMPACTS ON RESEARCH FIELDS AND GRADUATE TRAINING
Michael McGeary and Stephen A. Merrill
Aggregate and Agency Trends
The federal government has funded a large share of national research and development (R&D) since World War II. It was the largest funder until 1980 when it was surpassed by private industry. In 1998 (the most recent year for which expenditure data are available) it still provided 40.9 percent of all funding for research, basic and applied, carried out in university laboratories and medical centers, industrial and federal laboratories, and other research facilities in the United States.1
Real growth in the federal R&D budget—that is, growth in excess of inflation—began to level off in the late 1980s, and after 1992 it fell as part of the effort to reduce the federal budget deficit.2 According to data collected by the National Science Foundation (NSF) on actual research obligations, federal spending on the research part of R&D peaked in 1993 and by 1997 was 2.2 percent less in real terms.3
1
That is 56.7 percent of basic research and 30.0 percent of applied research. Industry spending on research did not exceed the federal government's support until 1995. Calculated from National Science Foundation (1998a, Tables B-2B (basic research) and B-3B (applied research)). See Box A-1 for definitions of basic research, applied research, and development.
2
Budget authority for R&D fell 8.8 percent in real terms between 1992, its historical high point, and 1997 (AAAS, 1998). Budget authority is legal authority to incur financial obligations that will result in outlays.
3
Obligations are commitments to spend money, although actual payment may be made later, for example, under multiyear contracts.
Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole. Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 53
APPENDIX A
RECENT TRENDS IN FEDERAL SPENDING ON SCIENTIFIC AND ENGINEERING RESEARCH: IMPACTS ON RESEARCH FIELDS AND GRADUATE TRAINING
Michael McGeary and Stephen A. Merrill
Aggregate and Agency Trends
The federal government has funded a large share of national research and development (R&D) since World War II. It was the largest funder until 1980 when it was surpassed by private industry. In 1998 (the most recent year for which expenditure data are available) it still provided 40.9 percent of all funding for research, basic and applied, carried out in university laboratories and medical centers, industrial and federal laboratories, and other research facilities in the United States.1
Real growth in the federal R&D budget—that is, growth in excess of inflation—began to level off in the late 1980s, and after 1992 it fell as part of the effort to reduce the federal budget deficit.2 According to data collected by the National Science Foundation (NSF) on actual research obligations, federal spending on the research part of R&D peaked in 1993 and by 1997 was 2.2 percent less in real terms.3
1
That is 56.7 percent of basic research and 30.0 percent of applied research. Industry spending on research did not exceed the federal government's support until 1995. Calculated from National Science Foundation (1998a, Tables B-2B (basic research) and B-3B (applied research)). See Box A-1 for definitions of basic research, applied research, and development.
2
Budget authority for R&D fell 8.8 percent in real terms between 1992, its historical high point, and 1997 (AAAS, 1998). Budget authority is legal authority to incur financial obligations that will result in outlays.
3
Obligations are commitments to spend money, although actual payment may be made later, for example, under multiyear contracts.
OCR for page 54
As Table A-1 indicates, the trend in research funding has not been uniform across agencies.
Much of the decline has been in defense research, because of the end of the Cold War and changed national security requirements. The Department of Defense's (DOD) support of research was down substantially in both relative and absolute terms between 1993 and 1997. Other agencies that spent less in real terms in 1997 than in 1993 included the Department of Energy (DOE), U.S. Department of Agriculture (USDA), and the Department of the Interior (DOI). But what would have been a $2 billion decline was largely offset by increases in other agencies, especially the National Institutes of Health (NIH), NSF, and the National Aeronautics and Space Administration (NASA), so that the net decrease was just $173.6 million in 1998 dollars.
Federal support of basic research did not decrease between 1993 and 1997. Although the aggregate level of basic research funding dipped for several years after 1993, it was up again slightly—by 1.5 percent—in 1997 over 1993 in real terms. This occurred in part because of real growth in research spending by agencies that favor basic research, NIH and NSF, which offset decreases at DOD, DOE, and other agencies with shrinking research budgets. In 1993, 66 percent of NIH's research budget and 93 percent of NSF's were classified as basic research;
TABLE A-1 Trends in Federal Research Funding Obligations, FY 1993-1997 (millions of 1998 dollars)
Change, 1993-1997
FY 1993
FY 1997
Amount
Percent
DOD
5,353.7
3,882.0
-1,471.6
-27.5
NASA
3,971.7
4,264.2
292.5
7.4
DOE
3,850.3
3,635.3
-214.9
-5.6
DHHSa
10,288.3
11,440.9
1,152.5
11.2
NIH
9,668.2
10,719.0
1,050.9
10.9
NSF
2,106.2
2,291.1
184.9
8.8
USDA
1,400.9
1,314.6
-86.3
-6.2
DOI
649.2
563.1
-86.1
-13.3
EPAb
404.6
416.9
12.3
3.0
DOCc
651.5
823.4
171.9
26.4
Others
1,419.4
1,290.7
-128.7
-9.1
Total research
30,095.7
29,922.1
-173.6
-0.6
aDepartment of Health and Human Services.
bEnvironmental Protection Agency.
cDepartment of Commerce.
Note: Constant-dollar conversions were made using the Gross Domestic Product (GDP) deflators in OMB (1998, Table 10.1).
Source: National Science Foundation (1998b).
OCR for page 55
BOX A-1 Definitions of R&D
Common definitions of basic research, applied research, and development are used by the Office of Management and Budget (OMB), the American Association for the Advancement of Science (AAAS), and NSF, the sources of data in this paper. NSF uses the same definitions in its survey of industry. They are also generally consistent with international definitions. The objective of basic research is to gain more comprehensive knowledge or understanding of the subject under study, without specific applications in mind. The objective of applied research is to gain knowledge or understanding to meet a specific recognized need. Development is the systematic use of the knowledge or understanding gained from research directed toward the production of useful materials, devices, systems, or methods.4
the percentages in 1997 were largely unchanged. As a group, agencies other than NIH and NSF spent 7.7 percent less on basic research in 1997 than in 1993 in real terms.
In short, although federally funded research does not appear to have suffered greatly from the decline in R&D that took place after 1992, the overall average downturn and the modest recovery since 1996 obscure the fact that research spending by some agencies has declined much more than others. Moreover, the agencies with declining of stagnant research budgets also turn out to be the primary funders of certain fields of research (Table A-2). In 1993, for example, DOD provided the majority of federal support of research in electrical engineering (82 percent), mechanical engineering (75 percent), materials engineering (73 percent), and computer science (57 percent). DOE provided the majority of funding for physics research (62 percent) and was the single largest supporter of chemical engineering (42 percent) and chemistry (29 percent). NASA provided the majority of funding for four other fields: aeronautical engineering (81 percent), astronautical engineering (79 percent), astronomy (76 percent), and atmospheric sciences (52 percent).
Issues
It was inevitable that R&D expenditures would be affected by the bipartisan consensus to reduce the budget deficit, and it is not very surprising that agency
4
For full definitions, see NSB (1998, pp. 4–9).
OCR for page 56
TABLE A-2 Top Federal Funders of Research by Field, FY 1993 (percent of total federal funding)
DOD
DOE
NASA
NIH
NSF
USDA
Other
Engineering
Aeronautical
18.3
81.4
0.3
Astronautical
21.3
78.7
Chemical
28.4
42.2
15.8
13.6
Civil
39.4
12.5
48.1
Electrical
82.1
4.9
6.8
6.2
Mechanical
75.4
8.6
7.4
8.6
Metallurgy & materials
73.3
9.6
6.0
11.1
Physical Sciences
Astronomy
2.2
75.8
16.0
6.0
Chemistry
16.6
29.2
16.9
37.3
Physics
18.1
61.9
10.6
9.4
Life Sciences
Biological
3.8
82.4
4.5
9.3
Environmental biology
13.8
34.2
52.0
Agricultural
82.0
18.0
Medical
5.1
2.4
83.5
9.0
Mathematical & computer sciences
Mathematics
27.8
23.7
28.5
20.0
Computer sciences
57.3
12.5
15.4
14.8
Environmental Sciences
Atmospheric
52.3
13.1
34.6
Geological
21.6
17.2
61.2
Oceanography
18.5
23.0
58.5
Note: Percentages greater than 50 percent are in bold to highlight dominant funders.
Source: National Science Foundation (1998b).
R&D budgets would change with circumstances as historic as the dissolution of the Soviet Union. Agency missions change and with them the resources for functions that support the missions. Nevertheless, the trends in agency budgets have raised three concerns:
1.
First, that the nation's capacity and productivity in fundamental longer-range research may be harmed by shorter-term trends, including declining support of research by certain federal mission agencies, especially if there is a trend in private industry to focus on projects with nearer-term payoffs.5
2.
Second, that because of the dependence of certain research fields such as physics, engineering, computer science, and mathematics on agencies with declining budgets, their health could be endangered for reasons unrelated to
5
See Appendix B for an assessment of the evidence on this point.
OCR for page 57
their productivity or opportunities for significant research advances but rather as an unintended consequence of changing agency mission requirements.
3.
Third, that because research grants to university investigators are the principal source of federal funds for the production of highly trained people in those science and engineering disciplines, graduate training could also be curtailed inadvertently.
Funding Trends by Research Field
We have seen that the research budgets of some of the largest R&D agencies have fallen or have not grown significantly. Some of those agencies, notably DOD, DOE, and NASA, provide the majority of funding in most fields of engineering, physical sciences, mathematics, and computer sciences. Some of the fields, for example, physics, computer science, electrical engineering, and materials engineering, are important to innovation in information technologies and to national economic performance generally. This situation raises a set of important questions:
How and to what extent have changes in federal agencies' research budgets affected the funding of fields dependent on them?
Are the changes invariably in the same direction?
Is there evidence of an effort to protect certain performers (e.g. universities) even in fields experiencing declining federal support overall?
Are there cases in which one agency has compensated for reduced support by another agency?
Is there evidence that a mechanism exists for making such adjustments to maintain a balanced research portfolio?
Do changes in graduate student support parallel changes in levels of research funding?
This paper relies on a series of annual surveys conducted by NSF's Division of Science Resources Studies (SRS) to attempt a preliminary test of these questions. The SRS survey of federal funds for R&D includes retrospective reports of agencies' actual obligations by fiscal year. The information is collected in a number of relevant categories that can be cross-tabulated in useful ways for an analysis of trends in federal research funding. For example, federal obligations for research are classified as basic research or applied research in 19 natural science and engineering fields.6 Research performed by universities and colleges
6
The NSF survey also includes obligations for research in the social and behavioral sciences. In addition, the agencies report funding of research ''not elsewhere classified'' for each broad field, such as life sciences and engineering, and for research that cannot be attributed to any broad field (see Table A-3 and footnote 9).
OCR for page 58
but not federally funded research performed in industry or government laboratories can be separately tabulated by field and/or whether it is basic or applied research.7 The ability to consider funding by agency and field of research makes it possible to assess whether the trend in an agency's level of research funding is correlated with trends in the funding of fields of which that agency has been the primary funder.
SRS also conducts an annual survey of university departments that includes reports of the principal source of support for each graduate student in science and engineering, including federally funded mechanisms such as fellowships, traineeships, and research assistantships. This information is reported for most of the research fields included in the federal funds report. Thus, trends in federal support of graduate students in those fields can be related to trends in the overall number of graduate students.
Both surveys—of federal R&D obligations and graduate student support—lag events because they are retrospective. For example, NSF recently released data from the survey of actual obligations during FY 1997 approximately 15 months after the end of the fiscal year. This permits analysis of trends in funding by field and agency during the period of budget cuts—fiscal years 1993 to 1997. A data series of just five years is too short a period to be very sure lasting trends exhibited by individual fields will be, but it is sufficient to derive tentative conclusions with regard to the broad questions posed above.8
Another obvious caveat is that quantitative changes in funding reveal little about the character of the research being cut or increased and therefore little about the qualitative effects of the changes in funding. It may be that a small change in one agency's support would have a far more profound effect on research and training in a field than a more substantial change in funding of research with a different character. The evolving orientation of some research fields is also difficult or impossible to discern from the data, although there is some effort to capture inter- and multidisciplinary research.9
7
Although the nomenclature corresponds to academic disciplines and departments, in NSF's Federal Funds Survey it is applied to R&D performed in industry and government laboratories and by other nonprofit institutions. Obligations for "development" are not reported by field.
8
The surveys ask agencies to estimate future research allocations for seven broad fields for research—e.g., physical sciences, life sciences, math/computer sciences, engineering, etc.—but these estimates are too general to be very useful to this analysis of impacts on specific fields.
9
In the Federal Funds Survey, multidisciplinary or interdisciplinary projects that do not fall within one of the broad fields of science (e.g., engineering, physical sciences, life sciences) are to be reported as "Other science, n.e.c.," where "n.e.c." means ''not elsewhere classified." Multidisciplinary projects that fall within a broad field and single-discipline projects that cannot be classified within one of the listed subfield categories are to be reported as ''Engineering, n.e.c." or "Mathematics & computer sciences, n.e.c.," etc. As shown in Table A-3, the n.e.c. categories are quite large in engineering and environmental science and much smaller in other major fields. They have been growing in most major fields but not very rapidly. All "n.e.c." research combined was 12.8 percent of total federal research in 1993 and 14.1 percent in 1997.
OCR for page 59
There is also a complication with the data reported by NSF on the survey. Beginning with FY 1996, NSF changed its procedures for classifying research obligations by field. The change most affected engineering. The amount classified as "engineering, n.e.c." went from about 20 percent of the total in 1995 to about 40 percent after that year. Mechanical engineering went from 13 percent to 2 percent of engineering research funding. The physical sciences were also affected. The amount classified as "physical sciences, n.e.c" increased from about 9 percent in 1995 to 26 percent in 1997. It appears that this involved moving research previously classified astronomy, chemistry, and physics into the n.e.c. category. Environmental sciences were also affected, except that atmospheric, geology, and ''environmental sciences, n.e.c." research were each apparently reclassified as oceanography. As a result, as we analyze affected fields such as mechanical engineering, we must take into account the extent to which NSF's new classification scheme might affect the results. The change will also affect our ability to ascertain how much NSF might have compensated for cutbacks in the support of fields by other agencies.
The research obligation trends by field are summarized in Table A-3. The findings reported in the following section are for a number of fields related to industrial activity.10
Fields with Declining Support
Fields whose federal support declined from 1993 to 1997 included electrical engineering, mechanical engineering, physics, chemical engineering, chemistry, and geology. Although several of these fields had DOD or DOE as a dominant11 funder in 1993, others (chemistry, chemical engineering, and geology) had quite diversified support.
Electrical Engineering
In 1993 DOD provided 82 percent of the federal funding for electrical engineering. DOD support dropped 40.3 percent between 1993 and 1997 in real terms, which accounted for most of the net drop of 35.7 percent in federal support of the field (see Figure A-1 and Table A-4). In this case there were simultaneous decreases at DOE and NSF. The only increase of significance—90.4 percent—was at the Department of Commerce (DOC) presumably because of the growth of the Advanced Technology Program at the National Institute for Standards and
10
Data for fields not included here (i.e., astronomy, astronautical engineering, agricultural sciences, environmental biology, psychology, and social sciences) are available at www4.nas.edu/pd/step/23ba.nsf, as are tables in current dollars for all of the fields in the survey.
11
"Dominant" funder is defined here as providing 50 percent or more of the federal support for a research field.
OCR for page 61
All Performers
Universities
1993
1997
% Change
1993
1997
% Change
Social Sciences, total
755.3
709.5
-6.1
241.7
201.1
-16.8
Psychology, total
616.3
555.7
-9.8
321.8
288.8
-10.3
Other sciences, n.e.c.
1,267.5
923.3
-27.2
446.8
344.0
-23.0
Note: All performers include federal intramural laboratories, industrial laboratories, universities and colleges, other nonprofit research institutions, national laboratories, and other federally funded research and development centers, state and local governments, foreign performers, and private individuals.
Source: National Science Foundation (1998b).
Figure A-1
Constant dollar trends in federal funding of electrical engineering research, FY 1990-1997.
Source: Tables A-4 and A-5.
OCR for page 62
Technology (NIST), but DOC's program is so small that nearly doubling it only reduced the drop that would have occurred in overall federal support by about 6.4 percentage points.
The results were similar for electrical engineering research conducted at universities and colleges (see Table A-5). The falloff in DOD support accounted for most of the net drop of 31.9 percent in federal funding of university research in electrical engineering in 1997. NSF also reduced its support by 25.4 percent.
TABLE A-4 Federal Obligations for Electrical Engineering Research, All Performers, by Agency, FY 1990–1997 (millions of constant 1998 dollars)
Change, 1993–1997
1990
1991
1992
1993
1994
1995
1996
1997
Amount
Percent
USDA
0.5
0.7
0.8
0.8
0.5
0.5
0.3
0.6
-0.2
-20.5
Commerce
15.3
18.1
23.9
26.9
31.3
60.0
51.2
51.1
24.3
90.4
DOD
659.9
691.0
706.5
809.6
596.8
609.3
533.0
483.3
-326.3
-40.3
DOE
26.0
50.7
44.5
48.2
46.2
54.7
25.8
24.7
-23.5
-48.7
DHHS
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NIH
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Interior
0.9
1.1
0.9
0.2
0.1
0.1
0.1
0.1
-0.1
-52.8
EPA
0.6
1.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NASA
14.2
19.7
17.0
23.4
24.8
25.5
25.4
25.5
2.0
8.7
NSF
61.0
68.4
61.2
67.1
72.7
60.8
44.2
43.0
-24.1
-35.9
All others
11.1
12.2
16.2
10.3
6.8
5.1
9.1
5.6
-4.7
-45.3
TOTAL
789.4
863.1
871.9
986.5
779.2
816.0
689.3
634.0
-352.5
-35.7
Note: Constant collar conversions were made using the GDP deflators in OMB (1998, Table 10.1).
Source: National Science Foundation (1998b).
TABLE A-5 Federal Obligations for University Research in Electrical Engineering, by Agency, FY 1990–1997 (millions of constant 1998 dollars)
Change, 1993–1997
1990
1991
1992
1993
1994
1995
1996
1997
Amount
Percent
USDA
0.0
0.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
DOD
106.4
92.7
104.1
150.6
105.1
130.4
103.9
96.3
-54.3
-36.0
DOE
2.4
2.8
2.2
2.7
1.5
3.2
2.3
2.1
-0.6
-21.9
DHHS
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NIH
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NASA
10.7
10.9
8.0
9.1
9.5
8.0
9.6
8.4
-0.7
-7.5
NSF
51.0
62.0
56.4
56.6
60.3
46.6
41.9
42.3
-14.4
-25.4
TOTAL
170.4
168.8
170.6
219.0
176.4
188.3
157.7
149.1
-69.9
-31.9
Notes: Constant dollar conversions were made using the GDP deflators in OMB (1998, Table 10.1). Federal support for university research is reported for only six agencies: USDA, DOD, DOE, DHHS, NASA, and NSF.
Source: National Science Foundation (1998b).
OCR for page 63
Mechanical Engineering
Federal support of mechanical engineering research followed a pattern very similar to electrical engineering. DOD, which provided 75 percent of all federal funding for mechanical engineering research in 1993, reduced its real level of funding of the field by 52.4 percent in 1997 (Figure A-2 and Table A-6). The next two largest federal funders, DOE and NSF, also reported that they had reduced their levels of support. As a result, net federal support of mechanical engineering research was 50.4 percent less in 1997 than in 1993 in real terms, as reported in the NSF survey. Mechanical engineering, however, is one of the fields affected by NSF's change in classification procedures. If we assume that
Figure A-2
Constant-dollar trends in federal funding of mechanical engineering research, FY 1990–1997.
Source: Tables A-6 and A-7.
OCR for page 88
Figure A-14
Constant-dollar trends in federal funding of oceanography research, FY 1990–1997.
Source: Tables A-30 and A-31.
TABLE A-30 Federal Obligations for Oceanography Research, All Performers, by Agency, FY 1990–1997 (millions of constant 1998 dollars)
Change, 1993–1997
1990
1991
1992
1993
1994
1995
1996
1997
Amount
Percent
USDA
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Commerce
188.5
115.8
160.5
165.1
209.8
102.2
109.3
135.7
-29.4
-17.8
DOD
93.1
102.4
97.3
97.0
86.2
111.7
122.0
120.1
23.1
23.8
DOE
7.4
5.9
7.1
8.4
8.0
8.7
12.1
7.4
-1.0
-11.7
DHHS
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NIH
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Interior
32.6
32.3
41.3
43.0
13.2
13.5
15.7
13.4
-29.6
-68.8
EPA
3.8
0.9
2.2
0.5
0.5
0.0
0.0
0.0
-0.5
-100.0
NASA
81.5
121.4
124.2
120.3
127.1
130.6
127.9
127.7
7.4
6.2
NSF
232.2
89.3
91.6
87.2
86.5
84.1
188.7
203.5
116.3
133.4
All others
2.3
3.3
2.9
1.4
1.2
1.0
1.0
1.1
-0.4
-25.6
TOTAL
641.5
471.3
527.2
523.1
532.6
451.8
576.8
609.0
86.0
16.4
Note: Constant-dollar conversions were made using the GDP deflators in OMB (1998, Table 10.1).
Source: National Science Foundation (1998b).
OCR for page 89
TABLE A-31 Federal Obligations for University Research in Oceanography, by Agency, FY 1990–1997 (millions of constant 1998 dollars)
Change, 1993–1997
1990
1991
1992
1993
1994
1995
1996
1997
Amount
Percent
USDA
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
DOD
55.3
66.6
64.8
64.2
50.1
48.2
50.7
52.8
-6.7
-17.6
DOE
4.1
3.1
4.1
3.9
4.1
4.8
3.6
4.3
0.7
11.7
DHHS
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NIH
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
NASA
10.8
10.7
18.4
14.0
14.5
13.8
11.5
12.2
-0.8
-12.9
NSF
154.3
79.9
84.2
82.2
82.6
54.4
136.1
155.0
76.8
88.6
TOTAL
224.4
160.3
171.5
164.2
151.3
121.1
201.9
224.3
70.0
36.6
Notes: Constant dollar conversions were made using the GDP deflators in OMB (1998, Table 10.1). Federal support for university research is reported for only six agencies: USDA, DOD, DOE, DHHS, NASA, and NSF.
Source: National Science Foundation (1998b).
either about the same or reduced. Also, the size of the increase might be less if spending on university oceanography by the Department of the Interior, which reduced its overall support of oceanography research during the time period, were included.
Trends in Graduate Enrollment
The NSF survey of the numbers of graduate students by source of support publishes most of the data by broad area of science and engineering, such as physical sciences or earth, atmospheric, and ocean sciences, rather than by academic field, such as physics and chemistry or geology and oceanography. The former categories are too general to associate changes in federal research funding with changes in graduate training by field. Table A-32 presents data on sources of federal support of full-time graduate students in science and engineering in all universities and colleges in five fields that correspond to the categories used above: computer science, biological sciences, chemical engineering, electrical engineering, and mechanical engineering.
The changes are mostly in the expected direction (see Table A-33). Where federal support of university research in a field was down sharply between 1993 and 1997 in real terms, graduate enrollments were generally down also. In chemical engineering, for example, federal funding of university research declined by nearly 13 percent between 1993 and 1997 in real terms (from Table A-15). The number of graduate students in chemical engineering also declined during the same period, by 5 percent (the number of graduate students whose main source of support was federal declined more) (Table A-32). In mechanical engineering,
OCR for page 90
TABLE A-32 Full-Time Graduate Students in Science and Engineering, Doctoral Institutions, by Field and Agency of Support: 1990–1997 (number)
Change, 1993–1997
Field and Source of Support
1990
1991
1992
1993
1994
1995
1996
1997
Number
Percent
All S&E Fields
All federally supported
51,706
55,166
57,326
59,469
59,788
58,539
57,094
55,902
-3,567
-6.0
DOD
8,425
8,640
8,753
9,220
8,951
8,834
8,317
8,568
-652
-7.1
DHHS, total
14,522
15,381
16,154
16,701
17,035
17,024
16,817
16,610
-91
-0.5
NIH
13,298
14,081
14,930
15,749
15,843
15,600
15,411
15,181
-568
-3.6
NSF
11,932
12,542
13,223
13,325
13,656
13,447
13,258
13,169
-156
-1.2
USDA
2,629
2,982
3,099
3,202
3,297
3,152
2,902
2,560
-642
-20.0
Other federal
14,198
15,621
16,097
17,021
16,849
16,082
15,800
14,995
-2,026
-11.9
All nonfederally supported
198,824
205,988
215,816
216,827
215,536
210,874
209,138
206,259
-10,568
-4.9
Total, all sources of support
250,530
261,154
273,142
276,296
275,324
269,413
266,232
262,161
-14,135
-5.1
Computer Sciences
All federally supported
2,376
2,501
2,595
2,883
3,007
3,109
3,074
3,114
231
8.0
DOD
1,127
1,125
1,171
1,314
1,311
1,390
1,335
1,411
97
7.4
DHHS, total
68
71
100
103
97
110
113
106
3
2.9
NIH
61
65
97
95
89
92
80
78
-17
-17.9
NSF
818
896
958
1,003
1,040
1,050
1,044
1,076
73
7.3
USDA
14
8
6
7
9
13
13
12
5
71.4
Other federal
349
401
360
456
550
546
569
509
53
11.6
All nonfederally supported
12,522
12,369
13,174
12,944
12,382
12,107
12,559
13,374
430
3.3
Total, all sources of support
14,898
14,870
15,769
15,827
15,389
15,216
15,633
16,488
661
4.2
OCR for page 92
Change, 1993–1997
Field and Source of Support
1990
1991
1992
1993
1994
1995
1996
1997
Number
Percent
Mechanical Engineering
All federally supported
2,494
2,697
2,753
2,952
2,885
2,726
2,546
2,559
-393
-13.3
DOD
735
724
715
777
763
801
709
725
-52
-6.7
DHHS, total
69
71
87
86
79
93
76
109
23
26.7
NIH
68
58
77
70
70
82
63
70
0
0.0
NSF
637
732
786
805
779
734
764
832
27
3.4
USDA
26
29
26
9
10
4
13
12
3
33.3
Other federal
1,027
1,141
1,139
1,275
1,254
1,094
984
881
-394
-30.9
All nonfederally supported
8,095
8,718
9,326
9,176
8,757
8,207
7,891
7,615
-1,561
-17.0
Total, all sources of support
10,589
11,415
12,079
12,128
11,642
10,933
10,437
10,174
-1,954
-16.1
Notes: (1) Ninety-three percent of all full-time graduate students, and 99 percent of federally supported full-time graduate students, were enrolled in doctorate-granting institutions in 1997; (2) Graduate students are reported according to the source of ''the largest amount'' of support received in fall 1997.
Source: NSF (1999, Table 38).
OCR for page 93
TABLE A-33 Changes in Graduate Enrollment and Federal Funding of University Research, Selected Fields, FY 1993–1997 (full-time graduate students in doctorate-granting institutions and federal research in constant dollars)
Number of Graduate Students
Field
All Sources of Funding (%)
Federally Funded (%)
Nonfederally Funded (%)
Federal funding of research in Universities (%)
All S&E fields
-5.1
-6.0
-4.9
2.8
Computer sciences
4.2
8.0
3.3
13.0
Biological sciences
0.9
0.4
1.2
a14.3
Chemical engineering
-4.8
-6.5
-4.1
-12.7
Electrical engineering
-7.0
7.6
-10.7
-31.9
Mechanical engineering
-16.1
-13.3
-17.0
-40.9
a Includes environmental biology because graduate students in environmental biology are included in biological sciences in the survey of graduate students and postdoctorates.
Notes: (1) Ninety-three percent of all full-time graduate students, and 99 percent of federally supported full-time graduate students, were enrolled at doctorate-granting institutions in 1997; (2) Graduate students are reported according to the source of "the largest amount" of support received in fall 1997; (3) Federal support of university research is reported only for six agencies: USDA, DOD, DOE, DHHS, NASA, and NSF.
SOURCES: (1) Graduate students: Table A-29; (2) federal funding: Table A-3 (constant dollars).
federal funding of university research declined by more than 40 percent (from Table A-7). Not surprisingly, then, the number of graduate students funded by federal agencies and by nonfederal sources was down, by 13 and 17 percent, respectively (Table A-32).
Electrical engineering is harder to interpret. Real federal support of academic research in electrical engineering declined by 32 percent between 1993 and 1997 (from Table A-5). Although the overall number of graduate students declined, as might be expected, the number of federally supported graduate students increased (Table A-32).
Federal support of computer science increased between 1993 and 1997 in real terms (the NSF survey reported an increase of nearly 13 percent, but it was probably less for reasons explained earlier) (Table A-27), and the number of federally funded graduate students in computer science also increased, by 8 percent (see Table A-33). The number of nonfederally funded graduate students also increased but by a smaller percentage (3 percent). Since they constituted the vast majority—more than 80 percent—of the graduate students in computer science, overall enrollment went up about 4 percent (Table A-32).
In the biological sciences, even though the level of federal support of university research was up substantially in 1997 compared with 1993 in real terms—more than 14 percent (including environmental biology)—the number of gradu-
OCR for page 94
ate students in biology hardly increased at all—by less than 1 percent. Federally supported graduate students increased even more slowly—by less than 0.5 percent.
The lack of consistent correlations between changes in federal research funding and changes in graduate enrollments is probably attributable to several factors. First, in most fields the federal government supports a fairly small proportion of graduate students. Second, there is bound to be a lag in the effect of federal research funding changes on graduate enrollments as current students work their way through graduate school and before the decisions of prospective students not to enter graduate school in a shrinking field can be felt. The data series is too short to ascertain whether there is a lagged effect. Third, although much federal support of graduate students comes thorough research grant funding in the form of research assistantships, some of it comes through specific fellowship and traineeship programs that are not counted as R&D. DHHS has the largest such programs, and NSF also has substantial fellowship and traineeship programs. If agencies protected education and training programs relative to research grant programs in budget downsizing, the number of federally funded graduate students could increase as research funding remained the same or declined.
Observations
In many fields of science and engineering a single mission agency is the principal source (50 percent or more) of federal support for research. It is natural to assume that, if the research budgets of such agencies are shrinking, the fields they support will be similarly affected. This analysis shows that it is not necessarily true that an agency's level of funding for a particular field goes down or up with the fortunes of the agency's overall research budget. Agencies do not cut or increase programs across the board. They may, and often do, protect some fields from cuts, cut them less, or even let them grow despite a shrinking research budget. Of course, that implies even deeper funding cuts in other fields supported by the shrinking research budget. Conversely, an agency with an increasing budget may hold down or even cut the level of support for one or more fields and give larger increases to other fields. Thus, it is important to assess funding trends for agencies and fields together.
In the period 1993 to 1997 the fields disadvantaged by changes in agency budgets included most fields of engineering (apart from aeronautical and materials engineering), physics, mathematics, and chemistry. The field that benefited most by the changes during this period was medical science. Fields whose funding fortunes did not comport with those of their major sponsor were computer science and materials engineering, which grew despite DOD cuts.
Although university research increased slightly from 1993 to 1997, support by field was also variable. In the 15 research fields examined in this paper, federal spending increased in only six and was lower in nine fields. In many
OCR for page 95
fields university research support fared better than overall research funding—that is, it decreased by a smaller percentage (chemical engineering, electrical engineering, mechanical engineering, physics, and geology) or increased by a larger percentage (metallurgy & materials engineering, biological sciences, atmospheric science and oceanography), but there were exceptions (chemistry, mathematics, and computer science). In two fields (medical sciences and aeronautical engineering), university research support declined while overall support increased.
Although there are instances in which a second agency has "picked up the slack" in a field receiving less support elsewhere, no single agency has explicitly or implicitly played that role, in contrast to the early 1970s when in response to cutbacks at NASA and DOD, NSF was charged by Congress to serve as the balance wheel of the system and the NSF budget was increased to pick up some of the researchers previously funded by NASA and DOD. In contrast, during the 1990s, NSF reduced or held flat its levels of support for several fields being cut by mission agencies, including chemical engineering, electrical engineering, mechanical engineering, physics, mathematics, and geology. NSF did increase its level of support for some fields being cut by DOD—materials engineering and computer sciences. The 1996 changes in how NSF classifies research by field make it difficult to interpret some of the changes from 1993 to 1997, especially in mechanical and materials engineering. Overall, however, NSF funding appears to have followed (or led) aggregate trends, generally boosting funding for fields that prospered and reducing funding for fields that suffered reductions.
It should be acknowledged that, although NSF's research budget has increased faster than inflation, the percentage increases would not be nearly enough to compensate for the magnitude of cutbacks in some fields, given the relatively modest size of NSF's budget. NSF provides support for most fields of science and engineering but typically accounts for 10 to 15 percent or less of total federal research funding in an individual field.
The reverse of stepping up funding to compensate for a drop in another agency's support of an important field would be for an agency under budget pressure to cut back because of increases elsewhere in the federal government. This may have been a factor in DOD's reductions in the biological sciences and medical research.
It is probably too early to see the effects of changes in levels of federal funding of research, even large ones, on graduate enrollments, let alone degree attainment rates. Such effects, if they do appear, will probably be attenuated because federal funding is not the principal source of support for graduate students in most fields.
Finally, what about the argument that having diversified agency research support is preferable to being dependent on a single major source of support because it is less vulnerable to sudden budget changes? This analysis provides little support for that hypothesis. One field with a broad base of federal funding support came out of the downturn in research funding between 1993 and 1997
OCR for page 96
with a larger budget in real terms (oceanography, +16.4 percent), but as explained above, most or all of that increase may be more apparent than real, resulting from changes in NSF classification procedures. The rest had substantially smaller budgets (chemistry, -8.7 percent; chemical engineering, -12.9 percent; geology, -21.1 percent; mathematics, -5.6 percent; and civil engineering, -12.9 percent) (see Figures A-15 and A-16).
Figure A-15
Federal research funding of fields with majority of support from one agency, FY 1993 and FY 1997.
Source: NSF (1998b).
OCR for page 97
Figure A-16.
Federal research funding, of fields with support from multiple agencies, FY 1993 and FY 1997.
Source: NSF (1998b).
References
American Association for the Advancement of Science. 1998. Historical Data on Federal R&D FY 1986-1999. Accessible from the AAAS website: HtmlResAnchor http://www.aaas.org/spp/dspp/rd/rdwwwpg.htm.
National Science Board. 1998. Science & Engineering Indicators--1998. Arlington, VA: National Science Foundation. Accessible from the NSF/SRS website: HtmlResAnchor http://www.nsf.gov/sbe/srs/stats.htm.
National Science Foundation. 1998a. Patterns of R&D Resources: 1998, Early Release Tables. Accessible from NSF/SRS website: HtmlResAnchor http://www.nsf.gov/sbe/srs/srs99402/start.htm.
National Science Foundation. 1998b. Survey of Federal Funds for Research and Development: Fiscal Years 1997, 1998, and 1999. Unpublished tables available from Division of Science Resource Studies, NSF.
National Science Foundation. 1999. Graduate Students and Postdoctorates in Science and Engineering: Fall 1997, Early Release Tables, NSF 99-405. Available at HtmlResAnchor http://www.nsf.gov/sbe/srs/srs99405/start.htm.
Office of Management and Budget. 1998. Historical Tables, Budget of the United States Government: Fiscal Year 1999. Washington, D.C.: U.S. Government Printing Office.
OCR for page 98
This page in the original is blank.
Items in cart [4]
