Chapter 5: Academic Research and Development


Financial Resources for Academic R&D

In 2004, U.S. academic institutions spent $42 billion on research and development. Between 1970 and 2004, average annual growth in R&D was stronger for the academic sector than for any other R&D-performing sector except the nonprofit sector.

  • During this period, academic R&D rose from about 0.2% to about 0.4% of the gross domestic product.
  • Academic performers are estimated to account for 54% of U.S. basic research, about 33% of total (basic plus applied) research, and 14% of all R&D estimated to have been conducted in the United States in 2004.

All reported sources of support for academic R&D— federal, industrial, state and local, and institutional— increased fairly continuously in absolute dollar terms between 1972 and 2003, even after adjusting for inflation. However, the long-term trends of a declining share of support from the federal government and an increasing share from industry showed signs of reversing at the end of this period.

  • The federal government provided 62% of academic R&D expenditures in 2003, substantial growth from the 58% share of support provided in 2000. The federal share of support had been in decline since the early 1970s, when it reached a high of 69%.
  • Institutions themselves contributed 19% of funds in 2003, compared with 11% in 1972.
  • Industry's share of academic R&D support grew rapidly during the 1970s and 1980s, fluctuated around 7% of the total during the 1990s, and declined substantially thereafter to 5% in 2003 as a result of absolute constant dollar declines in support in 2002 and 2003.

Between 1973 and 2003, there was a substantial relative shift in the share of academic R&D funds received by different science and engineering fields. However, all gained substantially in terms of absolute dollars, even after adjusting for inflation.

  • The life sciences (59% share in 2003), engineering (15% share), and the computer sciences (3% share) experienced share increases. However, the engineering share declined between 1993 and 2003.
  • The physical sciences (8% share in 2003); earth, atmospheric, and ocean sciences (6% share); social sciences; and psychology (6% combined shares) had share losses.

The historical concentration of academic R&D funds among the top research universities diminished somewhat between the early 1980s and mid-1990s but has remained relatively steady since then. Academic R&D activity is also occurring in a wider set of institutions.

  • The set of institutions in the group below the top 100 academic R&D institutions in funding increased their share of total academic R&D expenditures from 17% to 20% between 1983 and 2003. This was offset by a decline in the top 10 institutions' share from 20% to 17%.
  • The change in the number of institutions supported occurred almost exclusively among higher education institutions classified as Carnegie comprehensive; liberal arts; 2-year community, junior, and technical; or professional and other specialized schools.

In 2003, although about $1.8 billion in current funds was spent on R&D equipment, the share of all annual R&D expenditures spent on research equipment reached a historical low.

  • After reaching a high of just above 7% in 1986, the share of R&D spent on equipment declined by about one-third to 4.5% in 2003.
  • About 81% of equipment expenditures were concentrated in the life sciences (45%), engineering (20%), and the physical sciences (16%).

Research-performing colleges and universities continued to expand their stock of research space in FY 2003 with the largest increase in total research space (11%) since 1988. In addition to the traditional "bricks and mortar" research infrastructure, "cyberinfrastructure" is playing an increasingly important role in the conduct of S&E research .

  • Between 1988 and 2003, little changed in the distribution of research space across S&E fields.
  • Although 71% of university connections to the commodity Internet (Internet1) were at the two lowest speeds, at least 6% of the connections were at 1 gigabit/second or faster.

Top of page.Top of page

Doctoral Scientists and Engineers in Academia

The size of the doctoral academic S&E workforce reached an estimated 258,300 in 2003 but grew more slowly than the number of S&E doctorate holders in other employment sectors. Between 1973 and 2003 in academia, full-time faculty positions increased more slowly than postdoc and other full- and part-time positions, especially at research universities.

  • The academic share of all doctoral S&E employment dropped from 55% in 1973 to 45% in 2003.
  • The share of full-time faculty declined from 87% in the early 1970s to 75% in 2003. Other full-time positions rose to 14% of the total, and postdoc and part-time appointments stood at 6% and 5%, respectively.

The academic doctoral labor force has been aging during the past quarter century.

  • Both the mean and median age increased almost monotonically between 1973 and 2003.
  • In 2003, a growing, albeit small, fraction of employment was made up of individuals age 65 or older (4%), although the share of those 70 years or older declined for the first time since the late 1980s to just below 1%.

The demographic composition of the academic doctoral labor force experienced substantial changes between 1973 and 2003.

  • The number of women in academia increased more than sevenfold between 1973 and 2003, from 10,700 to an estimated 78,500, raising their share from 9% to 30%.
  • Although their numbers are increasing, underrepresented minorities—blacks, Hispanics, and American Indians/Alaska Natives—remain a small percentage of the S&E doctorate holders employed in academia.
  • The number and share of Asians/Pacific Islanders entering the academic S&E doctoral workforce increased substantially between 1973 and 2003.
  • The relative prominence of whites, particularly white males, in the academic S&E doctoral workforce diminished between 1973 and 2003.

Foreign-born scientists and engineers constituted 23% of scientists and engineers with U.S. doctorates in academic employment in 2003. This lower bound estimate of foreign-born doctorate holders excludes doctorates from foreign institutions.

  • The share of foreign-born doctorate holders was more than double that in 1973, when it stood at 11%.
  • Academic employment of foreign-born doctorate holders was highest in the computer sciences and engineering (44% and 40%, respectively), followed by mathematics (33%), the physical sciences (25%), and the life sciences (22%).

As the composition of positions in the academic workforce has changed over the years, a substantial academic researcher pool has developed outside the regular faculty ranks.

  • As the faculty share of the academic workforce has declined, postdocs and others in full-time nonfaculty positions have become an increasing percentage of those doing research at academic institutions. This change was especially pronounced in the 1990s.
  • A long-term upward trend is evident in the number of academically employed S&E doctorate holders whose primary activity is research relative to total academic employment of S&E doctorate holders.

In most fields, the percentage of academic researchers with federal support for their work was lower in 2003 than in the late 1980s.

  • Full-time faculty were less likely to receive federal support (45%) than other full-time doctoral employees (48%). Both of these groups were less frequently supported than postdocs (78%).
  • For each of the three groups mentioned above (full-time faculty, other full-time employees, and postdocs) recent doctorate recipients were less likely to receive federal support than their more established colleagues.

Top of page.Top of page

Outputs of S&E Research: Articles and Patents

The worldwide S&E publications output captured in Science Citation Index and Social Sciences Citation Index grew from approximately 466,000 articles in 1988 to nearly 700,000 in 2003, an increase of 50%.

  • This growth was a result of more articles published per journal and an increase in the number of journals covered by these two databases.

Worldwide growth in article output between 1988 and 2003 was strongest in the European Union (EU)-15, Japan, and the East Asia-4 (China, Singapore, South Korea, and Taiwan).

  • The EU-15 share of world output surpassed that of the United States in 1998, although growth in the EU-15 and also in Japan slowed starting in the mid-1990s.
  • The article output of the East Asia-4 grew more than sevenfold during this period, resulting in its share of world output rising from less than 2% to 8%.

The number of U.S. scientific publications remained essentially flat between 1992 and 2003, causing the U.S. share of world article output to decline from 38% to 30% between 1988 and 2003.

  • The flattening of U.S. output—199,864 articles in 1992, 211,233 articles in 2003—in the face of continuing growth of research inputs represents a trend change from several decades' growth in number of U.S. publications.

The share of publications with authors from multiple countries—an indicator of international collaboration and the globalization of science—grew worldwide and for most countries between 1988 and 2003.

  • In 2003, 20% of all articles had at least one foreign author, up from 8% in 1988.

The increase in international collaboration reflects intensified collaboration among the United States, EU-15, and Japan. It also reflects greater collaboration between these S&E publishing regions and developing countries and an emerging zone of intraregional collaboration centered in East Asia.

  • The share of internationally coauthored articles at least doubled in the United States, the EU-15, and Japan.
  • A pattern of intraregional collaboration emerged in East Asia in the mid-1990s centered in Japan and, increasingly, in China.

The United States has the largest share of all internationally authored papers of any single country, and its researchers collaborate with counterparts in more countries than do the researchers of any other country.

  • U.S.-based authors were represented in 44% of all internationally coauthored articles in 2003 and collaborated with authors in 172 of the 192 countries that had any internationally coauthored articles in 2003.
  • U.S. collaboration with the rest of the world continues to increase, but its relative share of coauthorship on other countries' internationally authored articles has declined as those countries have broadened their international ties.

As measured by the share of collaborative articles, both intrainstitutional collaboration of U.S. sectors and collaboration of these sectors with the rest of the world have increased significantly.

  • The share of U.S. academic articles with at least one non-U.S. address grew from 10% to 24% between 1988 and 2003. The share of U.S. academic articles with nonacademic U.S. authors increased by 6 percentage points during this period, to 30%.

The volume of citations to S&E literature grew more than 60% between 1992 and 2003.

  • The growth in citations was the greatest in the same S&E publishing regions that fueled growth of S&E publications: the EU-15, Japan, and the East Asia-4. The volume of citations to U.S. literature, however, flattened in the late 1990s.

The increase in citation volume in most regions coincided with a growing share of citations to work done outside the author's country, reflecting the growing ease of access to worldwide scientific literature.

  • Citations to literature produced outside the author's home country rose from 42% of all citations in 1992 to 48% in 2003.

The number of scientific articles cited by U.S. patents, an indicator of the linkage between science and technology, rose rapidly until the late 1990s.

  • These increases were heavily centered in academic-authored articles in the fields of biomedical research and clinical medicine.

The growing closeness of basic science and practical applications is also evident in the rising number of U.S. patents issued to U.S. academic institutions.

  • The number of U.S. academic patents quadrupled from approximately 800 in 1988 to more than 3,200 in 2003. The increase in patents was highly concentrated in life sciences applications.

Increases in licensing income and activity suggest growing efforts by universities to commercialize their products and technology.

  • Income from licensing was more than $850 million in FY 2003, more than double the amount in FY 1997, and new licenses and options increased by more than 40% during this period.
National Science Board.