Adequate financial support for R&D activities at U.S. universities and colleges, as well as excellent research facilities and high-quality research equipment, is essential in enabling U.S. academic researchers to carry out world-class research. Since academic R&D is a significant part of the national R&D enterprise, this section focuses both on the levels and sources of support for R&D activities at U.S. universities and colleges as well as academic R&D facilities and instrumentation.
In 1997, an estimated $23.8 billion was spent on R&D at U.S. academic institutions. Academia's role as an R&D performer increased steadily between 1984, when this sector accounted-as it had for more than a decade-for just 9 percent of all R&D performed in the country, and 1994, when it performed almost 13 percent of all U.S. R&D. (See figure 5-1.) By 1997, the sector's performance share had dipped to just below an estimated 12 percent.
Academic R&D activities are concentrated at the research (basic and applied) end of the R&D spectrum and do not include much development activity. Of 1997 academic R&D expenditures, an estimated 67 percent went for basic research, 25 percent for applied research, and 8 percent for development. (See figure 5-2.) From a national research-as opposed to national R&D-perspective, academic institutions accounted for between 23 and 30 percent of the U.S. total during the past three decades. In terms of basic research alone, the academic sector is the country's largest performer, accounting for between 44 and 53 percent of the national total during the past three decades. (See figure 5-1.)
Average annual R&D growth between 1984 and 1994 (in constant 1992 dollars) was much stronger for the academic sector than for any other R&D-performing sector-5.7 percent, compared to about 4.2 percent for other nonprofit laboratories, 1.5 percent for industrial laboratories, 0.6 percent for federally funded research and development centers (FFRDCs), and zero growth for federal laboratories. Since 1994, this growth has slowed to an estimated 1.6 percent annually; however, this rate is still higher than for any other R&D-performing sector but industry (which grew at an estimated 6.2 percent annually). As a proportion of gross domestic product (GDP), academic R&D rose from 0.23 to 0.30 percent between 1984 and 1997.
The Federal Government continues to provide the majority of funds for academic R&D. In 1997, it accounted for an estimated 60 percent of the funding for R&D performed in academic institutions. Nevertheless, the federal support share is declining fairly steadily, down from 68 percent in 1980 and 71 percent in 1970. (See figure 5-3.) Until the beginning of the 1990s, support from other sectors grew more rapidly than did that from the Federal Government. This trend reversed in the early 1990s, with federal support growing faster than nonfederal through 1995. Federal support is estimated to grow more slowly than nonfederal in both 1996 and 1997. The federal sector primarily supports basic research-71 percent of its 1997 funding went to basic research versus 20 percent to applied. Nonfederal sources provide a larger share of their support for applied research (61 percent for basic and 32 percent for applied research).
Federal support of academic R&D is discussed in detail later in this section; the following summarizes the contributions of other sectors to academic R&D.
Clinical revenues generated by medical school faculty have traditionally been used to support undergraduate and graduate medical education and research at U.S. medical schools. These revenues are also thought to be a major source of support for younger researchers, who often have difficulty obtaining external grants. In a study for the American Association of Medical Colleges Task Force on Medical School Financing (Jones and Sanderson 1996), it is estimated that clinical revenues generated by medical school faculty to support core academic programs totaled $2.4 billion in 1993. The major beneficiary of this support ($816 million) was found to be research, followed by undergraduate medical education ($702 million) and graduate medical education ($594 million). Jones and Sanderson note that hospitals may also provide clinical support for academic missions by applying hospital funds to academic programs and by absorbing academic program expenses that are not otherwise reimbursed. However, changes in the U.S. health care system-particularly the emergence of managed care, the growing consolidation of health care providers, and increased price competition-are believed to be adversely affecting both the level and nature of research at medical schools. For example, two recent studies (Moy et al. 1997; and Campbell, Weissman, and Blumenthal 1997) suggest that faculty members at U.S. medical schools might be conducting less clinical research because of pressure on their institutions to cut costs and raise revenues. They show that in regions where managed care plans are dominant and where there is stiff competition for dollars and patients among hospitals, physicians at academic medical centers report more pressure to take care of patients-and thus conduct fewer human studies, do less clinical research, and publish fewer papers.
The main finding of the Moy study is that medical schools in all markets had comparable rates of growth in NIH awards from 1986 to 1990, but that between 1990 and 1995, medical schools in markets with high managed care penetration had slower growth in the dollar amount and number of awards compared with schools in markets with medium or low managed care penetration. The authors conclude that their results "provide evidence of an inverse relationship between growth in NIH awards during the last decade and managed care penetration among U.S. medical schools," although they do state that it remains to be determined whether the association is causal. One of the findings of the Campbell study is that clinical researchers in less competitive health care markets published more scientific articles than those in more competitive health care markets. Another finding is that a significantly larger proportion of young faculty members had patient care duties in more competitive markets than in less competitive markets. The authors conclude that "increased competitiveness of health care markets seems to hinder the capacity of academic health centers to conduct clinical research and to foster the careers of young clinical faculty."
These findings raise questions as to where the funds for clinical research that might be lost due to the changing health care market are to come from in the future, as well as the patients to participate in clinical research experiments.
Growing industry support of academic R&D and expanding industry-university ties have given rise to two concerns: that universities' commitment to basic research may be undermined, and that free and open disclosure of academic research results may face restrictions. In a chapter in Challenge to the Research University, Wesley M. Cohen and coauthors Richard Florida, Lucien Randazzese, and John Walsh (1998) examine these issues in light of recent research. Key hypotheses and research results are summarized here.
A number of indicators suggest that industry-university research relations have indeed expanded substantially since the mid-1970s. The industry share of academic R&D has more than doubled during that time. In 1990, 1,056 university-industry R&D centers-nearly 60 percent of them established during the 1980s-spent $2.9 billion on R&D. Patenting at the top 100 research universities expanded from 177 awards in 1974 to 1,486 in 1994; 200 offices administered technology transfer and licensing activities in 1990, compared with 25 in 1980. The authors also cite anecdotal evidence of an increase in spinoffs or faculty participation in new firms, along with increasing equity shares held by universities in firms spun off to commercialize academic research outputs.
Different incentives motivate firms and universities to form these partnerships. University initiatives led to the establishment of almost three-quarters of the university-industry research centers-61 percent originating with faculty, 12 percent with administrators.
The authors hypothesize that firms' profit incentive may incline them to control access to results of research they have sponsored and that it may also focus them on applied rather than basic research. This conflicts with academics' priority-the free and open publication and dissemination of their research findings, which is the source of academic eminence and the basis for further scientific inquiry. Thus, widespread industry-university collaborations may induce shifts toward more applied academic research and restricted disclosure of academic research findings. Others have suggested that firms may shift some of their internal fundamental research to academia.
Cohen, Florida, Randazzese, and Walsh provide some evidence for their hypotheses. On the issue of restricted access to research results, 53 percent of a national sample of university-industry research centers allowed firms to request publication delays; 35 percent permitted deleting of information prior to submission for publication (Cohen, Florida, and Goe 1994). For 117 centers whose missions most strongly supported an orientation toward industry needs, these numbers were higher: publication delays, 63 percent; deletion of information, 54 percent. Moreover, study respondents reported restrictions on faculty communications with faculty and staff at the home university (21 percent), with those at other universities (29 percent), and with the general public (42 percent). These numbers are about 15 percentage points higher for centers strongly oriented toward industry needs. Cohen, Florida, Randazzese, and Walsh note, however, that although publication and communications restrictions may be contained in agreements, they are not necessarily always implemented. They also indicate that implementation of such restrictions may undermine key channels through which academic research affects industrial R&D.
Similarly, in a sample of companies supporting academic life science research, 82 percent stipulated that research results could be kept confidential pending a patent application; 47 percent had agreements permitting at least occasional embargo of results beyond the patent application (Blumenthal et al. 1996). In a survey of academic technology managers, 39 percent reported having agreements that placed restrictions on faculty sharing information regarding R&D breakthroughs with departmental or other center faculty. In that study, 79 percent of the technology managers and 53 percent of faculty with experience in interacting with firms indicated that firms had asked for R&D results to be delayed or kept from publication (Rahm 1995). Cohen, Florida, Randazzese, and Walsh note that the existence of spinoff companies raises the same set of questions and speculate that similar pressures may apply to the composition and disclosure of research-the main difference being that they would be generated by the faculty, rather than externally.
The evidence regarding a displacement of basic by applied research is less clear. Several studies have found an empirical association between greater faculty-industry interaction and more applied research (Rahm 1994, Morgan 1993 and 1994); another survey found that stronger center mission focus on improving industry activities was associated with lower shares of center effort going toward basic research (Cohen, Florida, and Goe 1994). However, while acknowledging the difficulty of drawing a boundary between basic and applied research, Cohen, Florida, Randazzese, and Walsh note that university-reported NSF data on the composition of academic R&D fail to reflect a shift away from basic research, which constituted 67 percent of academic R&D during 1980-83 and 66 percent during 1990-93. They point out that industry support may be attracting faculty already inclined toward applied research, rather than inducing others to shift away from basic research.
Patterns of sectoral funding of academic R&D vary depending on the type of academic institution involved. That is, the importance of different funding sources varies for both private and public universities. (See appendix table 5-3.) For all public academic institutions combined, just under 10 percent of R&D funding in 1995-the most recent year for which data are available-came from state and local funds, about 23 percent from institutional funds, and about 54 percent from the Federal Government. Private academic institutions received about 2 percent of funds from state and local governments, 9 percent from institutional sources, and 73 percent from the Federal Government. Both public and private institutions received approximately 7 percent of their respective R&D support from industry in 1995. Over the past two decades, the federal share of support has declined, and the industry and institutional shares have increased, for both public and private institutions.
Most academic R&D is now, and has been historically, concentrated in relatively few of the approximately 3,600 higher education institutions in the United States. In fact, if all such institutions were ranked by their 1995 R&D expenditures, the top 200 institutions would account for about 94 percent of R&D expenditures. In 1995:
This historic concentration of funds, however, has diminished somewhat during the past decade. In 1985, the top 10 institutions received about 19 percent of the funds. The decline in this group's share is approximately matched by the increase in the share of those institutions in the group below the top 100-this group's share increased from 19 to 22 percent of total academic R&D funds. The institutions ranked from 11 to 100 received similar shares in 1995 as in 1985 (between 61 and 62 percent). (See appendix table 5-4.)
The overwhelming share of academic R&D expenditures in 1995 went to the life sciences, which accounted for 55 percent of total academic R&D expenditures, 53 percent of federal academic R&D expenditures, and 57 percent of nonfederal academic R&D expenditures. Within the life sciences, medical sciences accounted for 27 percent of total academic R&D expenditures and biological sciences for 17 percent. The next largest block of total academic R&D expenditures was for engineering-16 percent in 1995. (See appendix table 5-5; for detailed data on expenditures over time by S&E field, see appendix table 5-6.)
Between 1985 and 1995, academic R&D expenditures for all fields combined grew at an average annual rate of 5.2 percent in constant 1992 dollars. (See figure 5-4.) Funding for the social sciences grew fastest during the decade, increasing at an average annual rate of 6.8 percent in constant dollars. Within the social sciences, political science was the fastest growing fine field (8.1 percent) and economics the slowest growing (4.2 percent). Engineering grew second fastest, increasing at an average annual rate of 6.2 percent. Within engineering, aeronautical/astronomical and civil engineering grew the fastest (7.5 percent and 7.4 percent, respectively) and electrical engineering the slowest (5.5 percent). Academic R&D expenditure growth was slowest in the physical sciences, averaging 3.6 percent. Within the physical sciences, physics and chemistry grew the slowest (2.5 percent and 2.9 percent, respectively) and astronomy the fastest (8.8 percent). All other S&E fields averaged rates of annual growth between 4 and 6 percent.
The distribution of federal and nonfederal funding of academic R&D in 1995 varied by field. (See appendix table 5-5.) For example, the Federal Government supported about 78 percent of academic R&D expenditures in both physics and atmospheric sciences, but only 32 percent of academic R&D in economics and 30 percent in the agricultural sciences.
The declining federal share in support of academic R&D is not limited to particular S&E disciplines. Rather, the federally financed fraction of support for each S&E field was lower in 1995 than in 1975. (See appendix table 5-7.) The most dramatic decline occurred in the social sciences (55 percent in 1975 to 39 percent in 1993); the smallest declines were in the computer sciences and environmental sciences (74 to 70 percent and 71 to 67 percent, respectively). The overall decline in federal share also holds for all the reported S&E fine fields except the agricultural sciences (which increased slightly from 29 to 30 percent). Many fields have experienced slight increases in federal share during the first half of the 1990s.
Federal obligations for academic R&D are concentrated in three agencies: the National Institutes of Health, the National Science Foundation, and the Department of Defense. Together, these agencies are estimated to have provided approximately 82 percent of total federal financing of academic R&D in 1997, as follows:
An additional 14 percent of the 1997 obligations for academic R&D are provided by the National Aeronautics and Space Administration (NASA, 6 percent); the Department of Energy (DOE, 5 percent); and the Department of Agriculture (USDA, 3 percent). (See appendix table 5-8.) Federal obligations for academic research are concentrated similarly to those for R&D. (See appendix table 5-9.)There are some differences, however, since some agencies place greater emphasis on development (DOD), while others place greater emphasis on research (NSF).
During the 1990s, NASA's funding of academic R&D increased most rapidly, with an estimated average annual growth rate of 3.1 percent per year in constant 1992 dollars. The next highest rates of growth were experienced by NIH (2.7 percent) and NSF (1.9 percent). Between 1996 and 1997, total federal obligations for federal R&D are estimated to decline in constant dollars. Only NSF (by 3 percent) and DOE (by 0.5 percent) are expected to increase their academic R&D obligations in 1997.
Federal agencies emphasize different S&E fields in their funding of academic research. Several agencies concentrate their funding in one field-the Department of Health and Human Services (HHS) and USDA focus on the life sciences, while DOE concentrates on the physical sciences. Other agencies-NSF, NASA, and DOD-have more diversified funding patterns. (See figure 5-5.) Even though an agency may place a large share of its funds in one field, it may not be an important contributor to that field, particularly if it doesn't spend much on academic research. (See figure 5-6.) NSF is the lead funding agency in the physical sciences (34 percent of total funding), mathematics (53 percent), and the environmental sciences (47 percent). DOD is the lead funding agency in the computer sciences (60 percent) and in engineering (38 percent). HHS is the lead funding agency in the life sciences (85 percent), the social sciences (41 percent), and psychology (86 percent). Within S&E fine fields, other agencies take the leading role-DOE in physics (46 percent), USDA in agricultural sciences (99 percent) and economics (75 percent), and NASA in astronomy (68 percent) and in both aeronautical (60 percent) and astronautical (64 percent) engineering.
Despite fluctuations over the past couple of decades, the number of academic institutions receiving federal support for their R&D activities has increased, rising from 555 in 1975, to 648 in 1985, and to 882 in 1995. (See text table 5-1.) Since most research and doctorate-granting institutions were already receiving federal support in 1975, almost the entire increase has occurred among comprehensive; liberal arts; two-year community, junior, and technical; and professional and other specialized schools. The number of such institutions receiving federal support has just about doubled over the 1975-95 period, rising from 335 in 1975, to 422 in 1985, and to 654 in 1995. These institutions are also receiving a larger share of the reported federal obligations for R&D to universities and colleges now than in the past-11 percent in 1995, compared to 7 percent in 1985.
Recently, legislation has been passed that requires federal agencies to demonstrate the impact of their programs. See "GPRA: Instituting Accountability in Federal Funding of Academic R&D" for a discussion of how this legislation hopes to improve federal planning and management, increase accountability for and assessment of results, and provide better information for congressional and agency decisionmaking. [Skip Text Box]
In response to the Clinton Administration's effort to move toward a government that works better and costs less, Congress passed the Government Performance and Results Act of 1993 (GPRA). GPRA aims to shift the focus of federal agencies away from traditional concerns such as staffing and the level of services provided and toward results. Specifically, GPRA looks to improve federal planning and management, increase accountability for and assessment of results, and provide better information for congressional and agency decisionmaking. To accomplish these and related goals, GPRA requires every federal agency to prepare detailed, multiyear strategic plans; annual performance plans; and annual performance reports. These documents give agencies formal tools with which to set forth goals, prepare plans to meet those goals, and to assess and measure progress and accomplishments on a regular and systematic basis.
GPRA poses a particular challenge for those agencies that must assess the scientific research programs they fund. In fact, the General Accounting Office (GAO) has found that measuring the discrete contribution of a federal initiative to a specific program result is particularly challenging for regulatory programs; scientific research programs; and programs that deliver services to taxpayers through third parties, such as state and local governments (U.S. GAO 1997a). Regarding research programs, GAO points out that the amount of money spent on R&D has been used as the primary indicator of how much research is being performed in a given area, but that such an input indicator does not provide a good indication of the outcomes (results) of the research. In a recent report, GAO notes that:
...experts in research measurement have tried for years to develop indicators that would provide a measure of the results of R&D. However, the very nature of the innovative process makes measuring the performance of science-related projects difficult. For example, a wide range of factors determine if and when a particular R&D project will result in commercial or other benefits. It can also take many years for a research project to achieve results...Experiences from pilot efforts made under the Government Performance and Results Act have reinforced the finding that output measures are highly specific to the management and mission of each federal agency and that no single indicator exists to measure the results of the research (U.S. GAO 1997b).
The Subcommittee on Research of the Committee on Fundamental Science, which operates within the President's Office of Science and Technology Policy, has been working with federal research agencies to establish a broad framework for GPRA implementation in the assessment of fundamental science programs. The subcommittee states that:
The central issue in assessing fundamental science lies in defining the goal against which progress is measured. The Administration's science policy statement, "Science in the National Interest" [Clinton and Gore 1994], establishes that goal as leadership across the frontiers of scientific knowledge. This is the critical measure for assuring that American scientists are expanding the knowledge base at the leading edge...
Leadership evaluation does not entail simplistic numerical ranking of national programs. Our national interest in leadership rests in having our research and educational programs perform at the cutting edge-sometimes in competition, but often in explicit collaboration, with scientists from other nations. This goal is the principal guideline for government stewardship of science in the national interest. It is an enabling or intermediate objective with respect to the over-arching goals of improved health and environment, national security, economic prosperity, and quality of life . . . Science drives progress toward the over-arching national goals over a long time period and only as part of a larger enterprise requiring a complex interplay of science and technological innovation with fiscal, regulatory, intellectual property rights, and trade policies. Consequently, the enabling goal of maintaining broad scientific leadership is that which guides the management and assessment of today's science investments. It provides the principal yardstick for GPRA assessment strategies for fundamental science programs (NSTC 1996).
The subcommittee concludes that retrospective evaluation will have to be the main assessment tool and that it will have to be updated periodically to examine the link between fundamental science and the overarching national goals. A final concern related to GPRA's implementation in an R&D environment is that it may cause science agencies to focus on processes and process issues and to set inflexible process goals. Such an approach is likely to interfere with the conduct of research, which must be flexible and changeable to be effective.
Agencies are still struggling with GPRA's requirements in this arena, puzzling over how to balance the need for planning with the need for flexibility; the need for short-term measures with the reality of accomplishments that will only be realized in the long term. Despite these challenges, GPRA is an important requirement and can be an opportunity for government agencies, Congress, and the university community to better communicate to the public the value of investments in R&D and education.
Total Space. Between 1988-89 and 1996-97, total academic science and engineering research space increased by almost 22 percent, from about 112 million to 136 million net assignable square feet (NASF). (See appendix table 5-12.) Planned construction expenditures for academic research facilities are expected to reach $3.1 billion (in constant dollars) in 1996-97. If this planned funding materializes, it will reverse the recent downward trend that began between 1990-91 and 1992-93. Construction expenditures in constant dollars peaked at around $3.4 billion in 1990-91, dropped to $3.0 billion in 1992-93, and dropped again to $2.8 billion in 1994-95. (See appendix table 5-13.)
New Construction. New construction projects initiated between 1986 and 1995 were expected to produce over 52 million square feet of research space when completed-the equivalent of about 39 percent of estimated existing research space. A significant portion of this new research space likely replaces obsolete or inadequate space rather than actually increases existing space: this is indicated by the fact that the total amount of research space increased by 24 million NASF between 1988-89 and 1996-97, while new construction initiated between 1988-89 and 1994-95 was expected to increase by 43 million NASF. Planned new construction projects initiated in 1996-97 are expected to produce just under 11 million square feet of new research space. (See appendix table 5-12.)
Repair and Renovation. Planned expenditures for major repair/renovation (i.e., projects costing over $100,000) of academic research facilities are expected to reach $1.3 billion (in constant dollars) in 1996-97. These expenditures also increased between 1992-93 and 1994-95, rising from $905 million to $1.1 billion in constant dollars. (See appendix table 5-13.) While expenditures for major repair/renovation increased between 1992-93 and 1994-95, expenditures for smaller S&E research facility repair/renovation projects (those costing less than $100,000) decreased-dropping during this period from $261 million to $135 million in constant dollars. Projects initiated between 1986 and 1995 were expected to result in the repair/renovation of over 55 million square feet of research space. Planned projects initiated in 1996-97 are expected to result in the repair/renovation of an additional 13.7 million square feet of research space. (See appendix table 5-12.)
Repair/renovation expenditures as a proportion of total capital expenditures (construction and repair/renovation) have increased steadily since 1990-91, rising from 25 percent of all capital project spending to 30 percent by 1994-95. Forty-three percent of all research-performing colleges and universities are planning to undertake some type of repair/renovation costing over $100,000 during 1996-97; 29 percent are planning to undertake construction projects during the same period.
Sources of Funds. Since 1986, there have been some shifts in the significance of various funding sources for the construction and repair/renovation of S&E research space. While the relative rankings of these sources have remained fairly constant-with state and local governments providing the largest share of support, followed by institutional funds-the proportions of funding for which they account have changed, sometimes dramatically. Most strikingly, the proportion of funds provided through private donations has declined. In 1986-87, this source accounted for about 20 percent of construction and repair/renovation funding; by 1994-95, however, its share had declined to about 12 percent. This reflects a drop in private donations to public institutions. Also of note, other debt grew from a 0.2 percent share in 1986-87 to account for 5.9 percent of funds in 1994-95; this reflects the increased importance of this funding source to private institutions. During the period, funds from federal sources and from tax-exempt bonds first grew in significance-the former increasing from 6 percent in 1986-87 to about 14 percent in both 1990-91 and 1992-93, and the latter from just below 16 percent to about 21 percent in 1990-91-and then dropped to account for smaller overall shares in 1994-95 (about 8 and 13 percent, respectively). (See appendix table 5-14.)
In general, the major sources of funds for new construction are not the same as those for repair/renovation. About 43 percent of the funds for new construction come from state and local governments, with about 16 percent from institutional funds. The significance of these funding sources is reversed for repair/renovation. About 41 percent of the funds for repair/renovation come from institutional funds, and 25 percent from state and local funds. The proportion of repair/renovation funds from the Federal Government increased from 6 percent in 1988-89 to slightly above 10 percent in 1994-95, while the federal proportion for new construction decreased from 14 to 8 percent during the same period. (See appendix table 5-14.)
Public and private institutions draw upon substantially different sources to fund the construction and repair/renovation of research space. (See figure 5-7.) Public institutions rely primarily on:
The Federal Government share declined from just above 14 percent in 1992-93 to below 7 percent in 1994-95.
Private institutions, for their part, rely primarily on:
A significant shift in the importance of tax-exempt bonds as a funding source for private institutions occurred between 1992-93-when they accounted for about 23 percent of total funding-and 1994-95, when they dropped to only about 10 percent. The decline in the importance of tax-exempt bonds over this period was roughly offset by an increase in the share of other debt from about 4 percent to about 14 percent. (See appendix table 5-14.)
Condition and Adequacy. Reported data suggest little change in the condition of academic S&E research space between 1988 and 1994. (See text table 5-2.) Specifically, about a quarter of this space was rated as suitable for use in the most scientifically sophisticated research; about a third was judged to be effective for most uses, but not most scientifically sophisticated; less than a quarter was reported as needing limited repair/renovation; and about a sixth was said to require major repair/renovation or replacement.
The 1996 survey responses cannot be readily compared to these earlier results because the wording and response choices have been changed. Specifically, the number of response categories has been reduced from five to three: suitable for the most scientifically competitive research; effective for most levels of research, but may need limited repair/renovation; and requires major renovation or replacement to be used effectively. This change essentially resulted in a shifting of about one-third of the space characterized in 1994 as "effective for most uses, but not most scientifically sophisticated," to the new category "suitable for the most scientifically competitive research"; and the other two-thirds to the new category "effective for most levels of research, but may need limited repair/renovation."
Unmet Needs. Determining what universities and colleges need with regard to S&E research space is a complex matter. In order to measure real as opposed to speculative needs, the 1994 facilities survey adopted a new approach to this issue. Faculty respondents were asked to report whether an approved institutional plan existed that included any deferred space needing new construction or repair/renovation. Respondents were then asked to estimate, for each S&E field, the costs of such construction and repair/renovation projects. The 1996 survey expanded on this question by asking institutions to report separately the construction and repair/renovation costs for projects included in such plans, as well as for projects not included.
In 1994, a total of 40 percent of all research-performing universities and colleges had an approved institutional plan that included construction or repair/renovation projects that were either deferred and unfunded. The estimated cost of these projects in constant dollars was $6.2 billion: $4.4 billion for new construction and $1.8 billion for repair/renovation. In 1996, 44 percent of research-performing institutions reported having an approved institutional plan that included construction or repair/renovation projects that were needed but that had to be deferred because funds were not available. These plans cited $7.4 billion of deferred capital project expenditures in constant dollars-$4.6 billion for new construction and $2.8 billion for repair/renovation. This total represents a $1.2 billion increase in deferred capital project costs between 1994 and 1996, the majority for repair/renovation ($970 million) and the remainder in deferred construction costs ($259 million). Another 11 percent of research-performing institutions identified $1.9 billion of needed deferred capital project expenditures that were not included in an institutional plan-$1.0 billion for new construction and $0.9 billion for repair/renovation.
There was little change in the distribution of academic research space across S&E fields between 1988 and 1996. (See appendix table 5-12.) About 90 percent of current academic research space continues to be concentrated in six fields:
The ratio of planned new construction during the 1986-95 period to existing research space differs across S&E fields. More than half of the research space for medical sciences at medical schools and for computer sciences appears to have been built in the 1986-95 period. In contrast, less than 20 percent of the research space for mathematics and psychology appears to have been newly constructed during this period. (See figure 5-8.)
Condition and Adequacy. The condition of academic research space also differs across S&E fields. In 1996, the agricultural sciences reported the largest proportion among all S&E fields-about 24 percent-of research space in need of major repair/renovation or replacement. Other fields with higher than average needs for repair/renovation or replacement are the physical sciences (19 percent of total research space), the environmental sciences (19 percent), and the medical sciences both in universities and colleges (21 percent) and in medical schools (20 percent). In contrast, major repair/renovation or replacement was needed for only 13 percent of the total research space in the social sciences, 12 percent in psychology, 10 percent in mathematics, and less than 8 percent in the computer sciences. No particular trends have emerged as yet with respect to changes over time in repair/renovation needs across S&E fields. (See appendix table 5-15.)
In 1994, 40 percent or more of all institutions surveyed indicated inadequate amounts of research space in engineering, the physical sciences, the biological sciences outside of medical schools, and the medical sciences in medical schools. (See appendix table 5-16.) One-third or less of all institutions surveyed indicated inadequate amounts of S&E research space in the environmental sciences, the agricultural sciences, mathematics, psychology, and the social sciences. In 1996, 40 percent or more of all institutions indicated inadequate amounts of research space in all S&E fields except mathematics. More than half of all institutions indicated inadequate amounts of research space in engineering, the physical sciences, the biological sciences outside of medical schools, the medical sciences (both in and outside of medical schools), and the agricultural sciences. It is unclear how much of the change that occurred over the two periods is due to changes in the survey questionnaire rather than to an increasing inadequacy of research space.
Unmet Needs. Deferred and unfunded needs existed in all S&E fields in 1996. The fields most frequently cited as having an unfunded need for new construction of research facilities as part of an institutional plan were the agricultural sciences (21 percent), engineering (19 percent), the medical sciences in medical schools (14 percent), and the physical sciences (13 percent). (See text table 5-3.) Unfunded need for repair/renovation projects reported in an institutional plan was indicated most strongly in the biological and medical sciences within medical schools (31 and 30 percent, respectively). An additional set of institutions reported deferred capital projects for both new construction and repair/renovation without an institutional plan in all S&E fields, with a larger percentage of institutions in each field reporting a need for repair/renovation projects than for new construction projects.
In four fields, estimated expenditures for needed capital projects (new construction plus repair/renovation) were over $1 billion (including those identified in an institutional plan or not in a plan): the physical sciences ($1.9 billion), engineering ($1.5 billion), the biological sciences outside of medical schools ($1.4 billion), and the medical sciences in medical schools ($1.3 billion). (See appendix table 5-17.)
Expenditures. In 1995, just over $1.2 billion in current fund expenditures was spent for academic research instrumentation. Over 80 percent of these expenditures were concentrated in three fields: the life sciences (38 percent), engineering (23 percent), and the physical sciences (19 percent). (See figure 5-9.)
Between 1985 and 1995, current fund expenditures for academic research instrumentation first increased-growing at an average annual rate of 6.5 percent between 1985 and 1989-then dipped-dropping about 2 percent a year for the next four years-before recovering and growing by 3.6 percent from 1993 to 1994 and by 9.6 percent from 1994 to 1995 (in constant 1992 dollars). There were variations in growth patterns during this period among S&E fields. (See appendix table 5-18.)
Federal Funding. Federal funds for instrumentation are generally received either as part of research grants-thus enabling the research to be performed-or as separate instrumentation grants, depending on the funding policies of the particular federal agencies involved. The importance of federal funding for research instrumentation varies by field. In 1995, the social sciences received about 40 percent of their research equipment funds from the Federal Government. In contrast, federal support accounted for over 60 percent of instrumentation funding in the physical sciences, computer sciences, environmental sciences, psychology, and engineering.
Since 1985, the share of research instrumentation expenditures funded by the Federal Government has declined-although not steadily-dropping from 64 to 59 percent. This overall pattern masks different trends in individual S&E fields. In one field-the environmental sciences-federal support actually rose, albeit very slightly, accounting for just below 68 percent of the field's instrumentation support in 1985 and just above 68 percent in 1995. Two other fields experienced sharp declines in federal support during this decade. The federal share for mathematics instrumentation dropped from 82 to 59 percent, and the share for computer sciences instrumentation dropped from 83 to 62 percent.
R&D Equipment Intensity. R&D equipment intensity is the percentage of total annual R&D expenditures from current funds devoted to research equipment. This proportion has declined since 1986, when research equipment accounted for 7.2 percent of total R&D expenditures. (See appendix table 5-19.) By 1993, R&D equipment intensity had dropped to 5.2 percent; it has since increased-slightly-to 5.6 percent in 1995.
R&D equipment intensity varies across S&E fields. It tends to be higher in the computer sciences (11.3 percent in 1995), physical sciences (10.6 percent), and engineering (8.2 percent); and lower in the social sciences (2.6 percent), psychology (3.3 percent), and life sciences (3.8 percent). This disparity is probably the result of the latter three fields using less equipment and/or less expensive instruments than the former three. Although the recent steady decline in R&D equipment intensity was not felt equally in all S&E fields, the 1986 figure was higher than that for 1995 in every field except mathematics. In that field, research equipment as a proportion of total R&D expenditures rose from 4.5 percent in 1986 to 5.4 percent in 1995. The data indicate, however, that the decline in R&D equipment intensity began to level off or reverse in 1993 for most S&E fields.
Stock, Condition, and Use. By congressional mandate, NSF has monitored academic research instrumentation status and needs since the early 1980s. As of 1993 (the most recent year for which detailed instrumentation data are available), the 318 colleges, universities, and medical schools represented in the survey reported a combined estimated stock of 61,684 instruments, with an estimated aggregate original purchase price of $6.255 billion. These instruments are categorized as shown in text table 5-4; their condition and usage were rated as follows:
Needs. In the 1994 Instrumentation Survey, most (69 percent) of the responding heads of academic departments and research facilities reported that their research instrumentation needs had increased since the last survey in 1992. A slim majority-58 percent-were satisfied with the overall capability of their existing instrumentation to support their faculty's research interests. The remaining 42 percent rated their research instrumentation as inadequate, and estimated the cost of the requisite upgrading at a total of $1.438 billion.
All respondents were asked to list and estimate the combined cost of the three top-priority research instruments costing over $20,000 their faculty most needed. Ten percent of the respondents said they had no immediate needs for additional instruments in this price range. For the others, the total combined cost of these items would be $2.048 billion, of which $942 million would be required for the top-priority item only. The primary reason cited for these top-priority research instruments was to upgrade unit capabilities-i.e., to perform experiments that the unit "cannot do now." The share from departments that reported inadequate overall instrumentation is an estimated cost of $939 million for their top three priority items-or about 65 percent of the $1.438 billion estimated cost to "fix" their units' overall instrumentation needs.