Chapter 5:

Academic Research and Development: Financial and Personnel Resources, Integration With Graduate Education, and Outputs

Highlights

Financial Resources for Academic R&D

In 1997, an estimated $23.8 billion (in current dollars) was spent for research and development (R&D) at U.S. academic institutions ($21.1 billion in constant 1992 dollars). The Federal Government provided $14.2 billion; academic institutions, $4.4 billion; state and local governments, $1.8 billion; and industry and other sources each provided $1.7 billion.
Industrially performed R&D grew faster than academic R&D between 1994 and 1997, and the academic sector's share fell to 12 percent, reversing a decade-long trend of an increasing role for academic performers in total U.S. R&D. Between 1984 and 1994, academia had risen from a 9 percent share to a 13 percent share of total U.S. R&D performance.
The academic sector performs over 50 percent of basic research, continuing to be the largest performer of basic research in the United States. Academic R&D activities are concentrated at the basic research end of the R&D spectrum. Of estimated 1997 academic R&D expenditures, an estimated 67 percent went for basic research, 25 percent for applied research, and 8 percent for development.
The Federal Government continues to provide the majority of funds for academic R&D. It provided an estimated 60 percent of the funding for R&D performed in academic institutions in 1997, down from about 65 percent in the early 1980s. Although nonfederal support increased more rapidly than federal through most of the 1980s, this trend was reversed in the first half of the 1990s. Federal support has grown more slowly than nonfederal in both 1996 and 1997, however.
Federal obligations for academic R&D are concentrated in three agencies: the National Institutes of Health (NIH-57 percent), the National Science Foundation (NSF-15 percent), and the Department of Defense (DOD-10 percent). The National Aeronautics and Space Administration (6 percent), the Department of Energy (5 percent), and the Department of Agriculture (3 percent) provide an additional 14 percent of obligations for academic R&D. Federal agencies emphasize different science and engineering (S&E) fields in their funding of academic research. Several agencies concentrate their funding in one field; others have more diversified funding patterns.
There has been a significant increase in the number of universities and colleges receiving federal R&D support during the past two decades, with almost the entire increase occurring among other than research and doctorate-granting institutions. In 1995, 654 of these institutions received R&D support from the Federal Government, compared to 422 in 1985 and 335 in 1975.
After the Federal Government, the academic institutions performing the R&D provided the second largest share of academic R&D support. The institutional share grew from about 14 percent of academic R&D expenditures in the early 1980s to an estimated 19 percent in 1997. Some of these funds directed by the institutions to research activities derive from federal and state and local government sources, but-since they are not restricted to research and the universities decide how to use them-they are classified as institutional funds.
Industrial R&D support to academic institutions has grown more rapidly than support from all other sources in recent years. In constant dollars, industry-financed academic R&D increased by an estimated average annual rate of 8.1 percent between 1980 and 1997. Industry's share grew from 4 percent to an estimated 7 percent during this period.
Total academic S&E research space increased by almost 22 percent between 1988 and 1996, up from about 112 million to 136 million net assignable square feet. When completed, construction projects initiated between 1986 and 1995 are expected to produce 52 million square feet of new research space, equivalent to about 39 percent of existing space.
In 1996, 55 percent of research-performing institutions reported construction or repair/renovation projects that were needed but had to be deferred because funds were not available. The cost of these deferred projects was $9.3 billion. Sixty percent of the needs reported were for construction and 40 percent were for repair/renovation.
Expenditures for academic research instrumentation in U.S. research universities began increasing recently. This increase follows a pattern of large increases in investment throughout most of the 1980s, followed by a steady decline of about 2 percent a year between 1989 and 1993. Annual research equipment expenditures as a percentage of total R&D expenditures declined from 7.2 percent in 1986 to 5.2 percent in 1993 before rising again to 5.6 percent in 1995.
Computers and data handling equipment represented 19 percent of the number of instruments in the national stock and 30 percent of total aggregate cost. There were an estimated 61,684 instruments with an estimated aggregate original purchase price of $6.255 billion in the stock of research instruments at the 318 colleges, universities, and medical schools represented in the National Survey of Academic Research Instruments and Instrumentation Needs in 1993.

The Academic Doctoral S&E Workforce

The 217,500 academic doctoral scientists and engineers in 1995 represented the largest number ever employed in the academic sector. But employment growth for this highly trained group was stronger in other parts of the economy, and the academic sector's employment share stood at 46 percent-a record low.
Full-time doctoral S&E faculty numbered an estimated 171,400 in 1995, a decline from 173,100 in 1991. Full-time faculty represented 79 percent of academic doctoral S&E employment in 1995, down from 88 percent in 1973. Much of the decline occurred among those with the rank of full professor.
The number of women with S&E doctorates who held academic positions increased to 52,400 in 1995. This represented a new high to 24 percent of total academic employment of doctoral scientists and engineers. Women remained highly concentrated in the life and social sciences and psychology.
Minority employment continued to grow and reached 35,300 in 1995, but stayed at low levels for some groups. The 12,800 members of underrepresented groups-black, Hispanic, Native American, and Alaskan Native-accounted for 6 percent of academic doctoral scientists and engineers, up from 2 percent in 1973. Asian employment in 1995 stood at 22,500, or 10 percent of the total; this was up from 4 percent in 1973.
Women and members of minority groups have tended to enter academic employment in line with or above their proportion of recently awarded S&E doctorates. Among recent Ph.D. recipients in academic employment-doctorates awarded in the preceding three years-women and underrepresented minorities were employed in rough proportion to their share of newly awarded doctorates to U.S. citizens and permanent visa-holders; Asians-many of whom are foreign-born-were represented well in excess of their share of new S&E Ph.D.s.
The progressive aging of the doctoral academic S&E workforce, evident over much of the past two decades, appears to have leveled off. The mean age of full-time doctoral faculty rose from 42.5 years in 1973 to 47.1 years in 1989 and stood at 47.4 years in 1995, suggesting gradual hiring for the system as a whole as faculty retire. However, for young Ph.D.s, this has to be seen in the context of a steep increase in newly awarded doctorates-from about 22,700 in 1989 to 27,800 in 1995.
An estimated 26,900 recent Ph.D. recipients-doctorates awarded in 1992-94-entered academic employment in 1995. But the meaning of academic "employment" has changed for these young doctorate-holders. Fewer than 45 percent had regular faculty appointments, compared with over 75 percent in the early 1970s, while the proportion in postdoctorate positions rose from 13 to 40 percent.
The physical sciences have grown more slowly than other fields in terms of overall doctoral employment-29,300 in 1995-and doctorates in full-time faculty positions. Their doctoral employment share fell from 19 percent in 1973 to 13 percent in 1995. The life sciences, engineering, and psychology gained employment shares.

Work Responsibilities of Academic Doctoral Scientists and Engineers

The academic doctoral S&E research workforce-defined as those whose primary or secondary work responsibility was research-numbered an estimated 153,500 in 1995, up from 80,000 to 90,000 during the 1970s. The highest levels of research participation, so defined, were found in engineering and the environmental sciences; the lowest in mathematics, psychology, and the social sciences.
In 1995, 39 percent of the academic doctoral workforce-85,700-reported having research support from the Federal Government during the week of April 15. This compares with 37 percent in 1993. A sizable fraction of those with federal funding-26 percent-obtained their support from more than one agency.
The number of those reporting teaching as their primary activity has fluctuated around the 100,000 mark since 1985. In contrast, those designating research as primary rose from 56,000 to 83,000 over the period. In 1995, 46 percent of respondents reported teaching as their primary work responsibility, compared with 38 percent who reported research.
Doctoral S&E employment growth in Carnegie research universities was largely confined to those identifying research as their primary activity-from 17,500 in 1973 to 45,900 in 1995. In other types of institutions, the number choosing research grew from 10,300 to 37,100 over the period.

Integration of Research With Graduate Education

In 1995, for the first time in almost two decades, enrollment of full-time S&E graduate students declined slightly. The enrollment decline was irrespective of primary source of support. The numbers of full-time graduate students with primary support from the Federal Government, nonfederal sources, or their own resources (self-support) all declined.
The proportion of full-time graduate students in S&E with a research assistantship as their primary mechanism of support has increased considerably. Research assistantships were the primary support mechanism for 66 percent of the students whose primary source of support was from the Federal Government in 1995, compared to 55 percent in 1980. For students whose primary source was nonfederal, research assistantships rose from 20 percent to 29 percent of the total during this period. The overall number of graduate students with a research assistantship as their primary mechanism of support increased every year between 1985 and 1994 before declining slightly in 1995.
The Federal Government plays a larger role as the primary source of support for some support mechanisms than for others. A majority of traineeships in both private and public institutions (53 and 73 percent, respectively) are financed primarily by the Federal Government, as are 60 percent of the research assistantships in private and 47 percent in public institutions.
NIH and NSF have been the primary source of federal support for full-time S&E graduate students relying on research assistantships as their primary support mechanism. From the early 1970s to the late 1980s, NSF was the federal agency that was the primary source for graduate research assistantships. It was surpassed by NIH in 1993. Between 1972 and 1995, the proportion of federal graduate research assistantships financed primarily by NSF declined from one-third to less than one-quarter, while the proportion financed primarily by NIH increased from one-sixth to one-quarter.
Research assistantships are more frequently identified as a primary mechanism of support in the physical sciences, the environmental sciences, and engineering than in other disciplines. Research assistantships comprise more than 50 percent of the primary support mechanisms for graduate students in astronomy, atmospheric sciences, oceanography, agricultural sciences, chemical engineering, and materials engineering. They account for less than 20 percent in the social sciences, mathematics, and psychology.

Article Outputs From Scientific and Engineering Research

In 1995, about 142,800 scientific and technical articles were published by U.S. authors in a set of journals included in the Science Citation Index (SCI) since 1981. The bulk-71 percent-were by academic authors. Eight percent each had authors affiliated with other major sectors: industry, government, and nonprofit organizations.
Publications by U.S. industrial authors rose strongly in life science fields-clinical medicine, biomedical research, and biology-and constituted nearly half of industry publications; this was up from 19 percent in 1991. From the late 1980s on, industry output in engineering and technology was lower than it had been in preceding years.
Increasingly, scientific collaboration in the United States involves scientists and engineers from different employment sectors. In 1995, just under one-quarter of all academic papers involved such cross-sectoral collaboration-6 percent with industry, 8 percent each with the federal and not-for-profit sectors, 3 percent with federally financed research and development centers, and 2 percent with other sectors. In the other sectors, well over half of their cross-sector collaborations involved academic authors.
Globally, five nations produced more than 60 percent of the 439,000 articles in the SCI set of journals in 1995: the United States (33 percent), Japan (9 percent), the United Kingdom (8 percent), Germany (7 percent), and France (5 percent). No other country's output reached 5 percent of total.
The development or strengthening of national scientific capabilities in several world regions was evident in faster publications output growth elsewhere than in the United States; growth elsewhere accelerated toward the mid-1990s, overshadowing continued growth in U.S. output. This continued a long-term decline in the U.S. share of total article output.
Europe accrued gains in output share-from 32 percent in 1981 to 35 percent in 1995. Asia's share rose from 11 to 15 percent, even though India's output declined by one-third in absolute number of articles over the period.
The number of articles in physics, earth and space sciences, and biomedical research increased the most rapidly-by 63, 36, and 30 percent, respectively-from 1981 to 1995. The output volume of articles in chemistry, clinical medicine, and engineering and technology was little changed; those for mathematics and biology declined.
Great variation marked countries' article outputs per billion U.S. dollars of their estimated 1995 gross domestic product. Israel and some smaller European nations ranked highest, exceeding 30 articles per billion. The United States was in the middle range, with 20 articles. Nations with fast-developing economies had smaller than expected article outputs, reflecting the recent rapidity of their economic strides and suggesting considerable room for further scientific growth.
Countries' science portfolios, as reflected in their published output, show some striking differences. Clinical medicine and biomedical research are heavily emphasized in the article outputs of the United States, United Kingdom, the countries of Northern Europe, several smaller Western European nations, and Chile. Chemistry and physics form a larger than average fraction of the output of France, Germany, Spain, Italy, Eastern Europe, Russia, Mexico, and many Asian countries. Russia, China, Egypt, and Asian countries emphasize engineering and technology.

International Collaboration and Citation of Research Outputs

The globalization of science is reflected in a pervasive trend in scientific publishing toward greater collaboration. In 1995, half of the articles in the SCI journals had multiple authors, and almost 30 percent of these involved international collaboration. This trend affected all fields, and a steadily growing fraction of most nations' papers involved coauthors from different nations. By 1995, article outputs since 1981 had grown by 20 percent, the number of coauthored articles by 80 percent, and the number with international coauthors by 200 percent.
For almost every nation with strong international coauthorship ties, the number of articles involving a U.S. author rose strongly between 1981 and 1995. Concurrently, however, many nations broadened the reach of their international collaborations, causing a diminution of the U.S. share of the world's internationally coauthored articles.
Citation patterns also mirror the global nature of the scientific enterprise, as researchers everywhere extensively cite research outputs from around the world. U.S. scientific and technical articles as a whole are cited by virtually all mature scientific nations in excess of the U.S. output's world share. This holds for chemistry, physics, biomedical research, and clinical medicine. U.S. articles in other fields tend to be cited at or slightly below their world output share.
The number of article citations on U.S. patents increased from 8,600 in 1987 to 47,000 in 1996, and their field distribution shifted strongly toward the life sciences. This rise in number of citations held for all fields and for papers from all sectors, with the fastest growth in citations to biomedical research and clinical medicine.
The number of academic patents, while small, rose more than sevenfold in just over two decades-from about 250 annually in the early 1970s to more than 1,800 in 1995-and the number of academic institutions receiving patents increased from about 73 in the early 1980s to 168 by the mid-1990s. Academic patenting increased more rapidly than all annual U.S. patent awards. Among institutions with patents are a growing number of universities and colleges not traditionally counted among the research universities.
Academic patents are concentrated in fewer utility classes than patents overall; in fact, patents in only three utility classes with presumed biomedical applicability constituted more than a quarter of all academic patents in 1995. Revenue from academic patenting reached $299 million in 1995.