U.S. R&D/GDP Ratio
Rates of Growth Among Sectors
Federal R&D Performance
R&D in Nonprofit Organizations
Recent Growth in Industrial R&D, by Sector, Firm Size, and R&D Intensity
Performance by Geographic Location, Character of Work, and Field of Science
U.S. R&D/GDP Ratio
Growth in R&D expenditures should be examined in the context of the overall growth of the economy, because, as a part of the economy itself, R&D is influenced by many of the same factors. Furthermore, as mentioned earlier, the ratio of R&D expenditures to GDP may be interpreted as a measure of the nations commitment to R&D relative to other endeavors.
A review of U.S. R&D expenditures as a percentage of GDP over time shows an initial low of 1.36 percent in 1953 (when the NSF data series began), rising to its highest peak of 2.88 percent in 1964, followed by a gradual decline to 2.12 percent in 1978. (See figure 4-9 .) From that low in 1978, U.S. R&D expenditures again rose steadily to peak at 2.72 percent in 1985 and did not fall below 2.50 until 1993. In 1994, the rate dropped to 2.40, its lowest point since 1981. Starting in 1994, however, R&D/GDP has been on an upward trend as investments in R&D have outpaced growth of the general economy. As a result, the current ratio of 2.66 for 2000 is the highest the ratio has been since 1985.
The initial drop in the R&D/GDP ratio from its peak in 1964 largely reflects Federal cutbacks in defense and space R&D programs, although gains in energy R&D activities between 1975 and 1979 resulted in a relative stabilization of the ratio between 2.1 and 2.2 percent. Over the entire 196578 period, the annual percentage increase in real R&D was less than the annual percentage increase in real GDP. When real R&D spending decreased during that period, real GDP also fell, but at a lower rate.
The rise in R&D/GDP from 1978 to 1985 was due as much to a slowdown in GDP growth as it was to increased spending on R&D activities. For example, the 1980 and 1982 recessions resulted in a slight decline in real GDP, but there was no corresponding reduction in R&D spending. During previous recessions, changes in funding for R&D tended to match or exceed the adverse movements of the broader economic measures.
The share of defense-related R&D dropped from 31 percent in 1985 to 23 percent in 1991. Commensurate with this change was the sharp fall in the share of federally funded R&D, from 46 percent in 1985 to 37.8 percent in 1991. (See figure 4-4 .) This decline in Federal funding was counterbalanced by increased non-Federal funding, as described earlier in the discussion of industrial trends. Indeed, since the late 1980s, practically all of the rise in the R&D/GDP ratio has resulted from gains in industrial R&D spending.
From 1991 to 1994, the R&D/GDP ratio declined from 2.69 to 2.40. Since then, however, it has risen steadily. Between 1994 and 2000, the R&D supported by industry grew in real terms by 8.6 percent annually, whereas real GDP grew by 4.0 percent, largely explaining the rise in the R&D/GDP ratio to 2.66 in 2000. From 1992 to 2000, the ratio of research alone to GDP has remained at 1.0 percent, while the ratio of development to GDP has varied between 1.5 and 1.6 percent. Within the industrial sector, however, development plays a greater role. In 1999, for example, the ratio of research performance to net sales in industry was 0.8 percent, while the ratio of development to net sales was 2.0 percent.
Rates of Growth Among Sectors
The sectoral shares of U.S. R&D performance have shifted significantly since the early 1980s. (See figure 4-10 for levels of expenditure.) In 1980, industry (including industry-administered FFRDCs) performed 70.3 percent of the nations R&D; the academic sector (including academically administered FFRDCs) accounted for 13.9 percent; the Federal Government accounted for 12.4 percent; and the nonprofit sector (including nonprofit-administered FFRDCs) accounted for 3.3 percent. Industrys defense-related R&D efforts accelerated in the early 1980s, and its share of performance total rose to 73.4 percent in 1985.
From 1985 to 1994, R&D performance grew by only 1.4 percent per year in real terms for all sectors combined. This growth was not evenly balanced across performing sectors, however. R&D performance at universities and colleges (including their FFRDCs) grew by 4.4 percent per year in real terms compared with only 1.0 percent growth for industry (including their FFRDCs), a decline of 0.5 percent per year for Federal intramural performance and growth of 4.0 percent per year for nonprofit organizations (including their FFRDCs).
The 19942000 period witnessed dramatic changes in these growth rates. Total R&D performance, in real terms, averaged 5.8 percent growth per year, which was substantially higher than in the earlier sluggish period. Yet, R&D performance at universities and colleges (including their FFRDCs) grew by only 3.1 percent per year in real terms. Industry (including their FFRDCs) grew at a remarkable rate of 7.0 percent in real terms. Federal intramural performance increased by 0.8 percent per year in real terms. Finally, nonprofit organizations (including their FFRDCs), according to current estimates, increased their R&D by 5.3 percent per year in real terms over the same six-year period. According to preliminary estimates, these shifts in growth have led, in 2000, to academia (including FFRDCs) representing 13.6 percent of total U.S. R&D performance; Federal intramural activities, 7.2 percent; other nonprofit organizations (including FFRDCs), 3.6 percent; and private industry (including FFRDCs), 75.6 percent. (For level of expenditures in 2000, see text table 4-1 .)
Federal R&D Performance
The Federal Government performed $19.1 billion of total U.S. R&D in calendar year 2000, a 2.3 percent rise in real terms from its 1999 level of $18.3 billion. Among the individual agencies, DOD has continued to perform the most intramural R&D; in fact, in FY 2001 it performed more than twice the R&D of the second largest R&D-performing agency, HHS (whose intramural R&D is performed primarily by NIH). (See text table 4-3 .) However, DODs intramural R&D performance has grown by less than 1 percent per year in real terms since FY 1980, reaching a level of $8.6 billion in FY 2001. Furthermore, an undetermined amount of DODs intramural R&D ultimately appears to be contracted out to other extramural performers. NASAs intramural R&D has grown by 1.4 percent per year in real terms since 1980, to $2.5 billion in FY 2001, and HHS intramural performance rose by 4.0 percent to $3.7 billion. Together, these three agencies account for 76.2 percent of the total ($19.4 billion) Federal intramural R&D in FY 2001.
Total R&D performed by industrial, academic, and nonprofit FFRDCs reached $9.3 billion in calendar year 2000, which is essentially the same as its level of $9.0 billion in 1999 after adjusting for inflation. R&D at FFRDCs in 2000 represented 3.5 percent of the national R&D effort, most of which ($5.8 billion in 2000) was accounted for by university- and college-administered FFRDCs.
R&D in Nonprofit Organizations
A recent NSF survey has led to upward revisions in R&D performance estimates for the nonprofit sector (NSF 2001d). Based on a survey of FY 1996 and FY 1997 R&D at nonprofit organizations and on other available data for the past three years, R&D performance by nonprofit organizations is expected to reach $8.8 billion in 2000, reflecting an average annual growth of 5.5 percent, in real terms, since 1990. Such growth, however, varies considerably by source of funding. The average annual real growth in nonprofit intramural R&D over the same period was 8.0 percent for nonprofit R&D supported by nonprofit organizations themselves, 7.1 percent for nonprofit R&D supported by industry, and 3.5 percent for nonprofit R&D supported by the Federal Government.
Like the Federal Government, nonprofit organizations in recent decades have focused largely on medical and health sciences. (See text table 4-6 .) In 1973, only 3.3 percent of all R&D performed by nonprofit organizations was in medical and health sciences, but this share rose dramatically to 60 percent by 1997. In contrast, the agricultural sciences share of intramural nonprofit R&D fell from 21.3 percent in 1973 to 0.3 percent in 1997.
Recent Growth in Industrial R&D, by Sector, Firm Size, and R&D Intensity
R&D performance by private industry reached $199.9 billion in 2000, including $2.6 billion spent by FFRDCs administered by industrial firms. This total represents a 7.1 percent real increase over the 1999 level of $82.8 billion, which, in turn, reflects a smaller, although still noteworthy, real gain of 6.5 percent over 1998. In 2000, R&D performed by industry that was not federally financed rose 8.6 percent in real terms above the 1999 level. Overall, private companies (excluding industry-administered FFRDCs) funded 90.0 percent ($177.6 billion) of their 2000 R&D performance, with the Federal Government funding nearly all the rest ($19.6 billion, or 10 percent of the total).
In recent times, the greatest share of R&D in the United States has been performed by private industry through private industrys own funds. This component of U.S. R&D has grown in importance, from 44 percent of total R&D in 1953, to 49 percent in 1980, to 55 percent in 1990, and 68 percent in 2000. The underlying causes for industrys growing share of R&D financing are complex. In part, the growth may be due to changes in Federal support in areas such as defense and space exploration. Other factors include S&E success stories in specific fields, such as information technology (IT) and biotechnology, in which industry plays a dominant role.
R&D in Manufacturing Versus Nonmanufacturing Industries
Until the 1980s, little attention was paid to R&D conducted by nonmanufacturing companies largely because service-sector R&D activity was negligible compared with the R&D operations of companies classified in manufacturing industries. Before 1983, nonmanufacturing industries accounted for less than 5 percent of the industry R&D total (including industrial FFRDCs), but by 1999 (the most current year for data on industrial sectors), it had reached 36.0 percent. In 1999, nonmanufacturing firms R&D performance totaled $65.9 billion ($60.4 billion in funds provided by companies and other non-Federal sources and $5.5 billion in Federal support).
Beginning with the 1999 cycle, statistics from NSFs Survey of Industrial R&D have been published using the North American Industrial Classification System (NAICS). (See text table 4-7 .) The development of NAICS has been a joint effort of statistical agencies in Canada, Mexico, and the United States. The system replaces the standard industrial classification (SIC) (1980) of Canada, the Mexican Classification of Activities and Products (1994), and SIC (1987) of the United States. NAICS was designed to provide a production-oriented system under which economic units with similar production processes are classified in the same industry. NAICS was developed with special attention to classifications for new and emerging industries, service industries, and industries that produce advanced technologies. NAICS eases comparability of information about the economies of the three North American countries and also increases comparability with the two-digit level of the United Nations International Standard Industrial Classification system (ISIC Revision 3).
Among manufacturers, the new computer and electronic products classification (NAICS 334) includes makers of computers and peripherals, semiconductors, and navigational and electromedical instruments. Among nonmanufacturing industries are information (NAICS 51) and professional, scientific, and technical services (NAICS 54). Information includes publishing (both paper and electronic), broadcasting, and telecommunications. Professional, scientific, and technical services include a variety of industries. Of specific importance for the survey are engineering and scientific R&D services (NSF 2001e).
Following these recent changes in classification, much of the historical data on R&D that had been subdivided according to the previous industrial categories cannot be reclassified into the current industrial categories. As a result, some of trends in the data by industrial category can no longer be observed after 1998 and must be started again, according to different groupings, in 1999. On the other hand, general patterns of change among major sectors are still identifiable. The most striking change in industrial R&D performance during the past two decades is the nonmanufacturing sectors increased prominence.
In 1999, the largest nonmanufacturing industry in the performance of R&D was trade (as it is classified in NAICS), which accounted for 10.7 percent of all industrial R&D performance. This was followed closely by professional, scientific, and technical services, accounting for another 10.4 percent of the total, then information, accounting for 8.4 percent.
Within the manufacturing industrial sector (including industry-administered FFRDCs associated with manufacturing), three subsectors dominate: computer and electronic products, transportation equipment, and chemicals. (See figure 4-11 and text table 4-7 .) Under the new NAICS system of classification, the computer and electronic products sector accounted for the largest amount of R&D performed in 1999 among all industrial sectors$35.9 billion. It accounted for 19.7 percent of all industrial R&D (including industry FFRDCs), as well as 14.7 percent of the entire nations R&D, performed in 1999. Consequently, it exceeded the total amount of R&D performed in 1999 by all universities and colleges and their administered FFRDCs combined (which is only $34.1 billion). For this sector, industrial firms provided $29.9 billion in R&D support; the Federal Government funded the remainder.
Transportation equipment was a close second among the manufacturing sectors in R&D performed in 1999 with $34 billion in R&D, representing 18.6 percent of all industrial R&D (including R&D by industry-administered FFRDCs). Of these expenditures, 29.6 percent was federally funded, primarily for R&D on aerospace products (planes, missiles, and space vehicles). In addition to aerospace products, the sector includes a variety of other forms of transportation equipment, such as motor vehicles, ships, military armored vehicles, locomotives, and smaller vehicles like motorcycles, bicycles, and snowmobiles (U.S. OMB 1997).
Ranking third in R&D is chemicals, with $20.2 billion in 1999, for which less than 1 percent was federally funded. This sector includes the subsectors pharmaceuticals and medicines (61.0 percent of non-Federal R&D funding in the chemical sector); basic chemicals (13.2 percent); resin, synthetic rubber, fibers, and filament (11.1 percent); and other chemicals (14.7 percent).
Although a great deal of R&D in the United States is related in some way to health care services, companies specifically categorized in the health care services sector accounted for only 0.4 percent of all industrial R&D and for only 1.0 percent of all R&D by nonmanufacturing companies. These results illustrate that R&D data disaggregated according to industrial categories (including the distinction between manufacturing and nonmanufacturing industries) may not always reflect the relative proportions of R&D devoted to particular types of scientific or engineering objectives, or to particular fields of science or engineering. The section "Cross-Sector Field-of-Science Classification Analysis" compensates to some extent for this limitation in the data by providing R&D expenditure levels associated with the broadly defined fields of life sciences and chemistry.
As a case in point, firms that perform R&D under contract to other firms are, by definition, in the service sector because the R&D they perform is, in fact, their "product," which is a service as opposed to manufactured goods. However, they often perform R&D under contract with a manufacturer, implying that those same R&D activities would have been classified as R&D in manufacturing if the same research firm were a subsidiary of the manufacturer. This is counterintuitive in that it implies that whether R&D is in manufacturing or in services is determined, in part, not by physical aspects of the R&D actions themselves but by the labels that have been placed on the firms that perform the R&D. Consequently, a growth in measured R&D in services may, in part, "reflect a more general pattern of industrys increasing reliance on outsourcing and contract R&D" (Jankowski 2001b).
R&D Spending by U.S. Corporations
In 1998, the top 20 U.S. corporations in R&D expenditures spent $54.0 billion on R&D. General Motors reported the most R&D in 1998 with $7.9 billion, followed by another company in the motor vehicle sector, Ford Motor Company, with $6.3 billion. (See text table 4-8 .) The rest of the list is dominated by computers, electronic equipment, and software companies, and by pharmaceutical corporations.
Between 1996 and 1998, the total number of publicly held U.S. corporations reporting R&D spending fell from 3,256 to 3,028, although some of this decline is attributable to mergers among existing firms. The decline in the number of firms was not uniform across industrial sectors. For example, the aircraft, guided missiles, and space vehicles sector, which is characterized by relatively large corporations, included exactly 21 corporations in each of the three years. Similarly, the motor vehicles and surface transportation sector went down in number by only 1, from 71 to 70 corporations. This was due to the acquisition of Chrysler Corporation by the German firm Daimler-Benz, which removed Chrysler from the list of U.S. corporations performing R&D (although the R&D it performs within the United States is still collected by NSFs industrial R&D survey and included in this chapters data on U.S. industrial R&D performance). Chrysler was ranked number 12 in U.S. corporations 1997 R&D spending. In contrast, between 1996 and 1998, the number of R&D-performing corporations fell from 1,477 to 1,382 in the information and electronics sector, from 629 to 566 in the medical substances and devices sector, and from 422 to 386 in the basic industries and materials sector (Shepherd and Payson 2001).
Industrial R&D and Firm Size
Industrial manufacturing R&D performers are typically quite different from industrial nonmanufacturing R&D performers; nonmanufacturing R&D performers tend to be smaller firms. (See text table 4-9 .) Approximately 39,000 firms in the United States performed R&D in 1999; of these, 54 percent were in the nonmanufacturing sector. Yet, manufacturers account for 64 percent of total industry R&D performance (including federally funded industry performance). As a share of the nations GDP, on the other hand, manufacturing accounts for less than 20 percent. The main reason for continued dominance in R&D performance is that among manufacturing firms, the largest in terms of number of employees tend to perform a relatively large amount of R&D. Among small R&D-performing firms (fewer than 500 employees) in both the manufacturing and nonmanufacturing sectors, those in the nonmanufacturing sector tend to conduct twice as much R&D per firm as those in the manufacturing sector. However, among large R&D-performing firms (more than 25,000 employees) in both sectors, those in the manufacturing firms tend to conduct more than 10 times as much R&D per firm as those in the nonmanufacturing sector.
Although R&D tends to be performed by large firms in the manufacturing sector and small firms in the nonmanufacturing sector, within each sector there is considerable variation, depending on the type of industry. R&D tends to be conducted primarily by large firms in several industrial sectors: aircraft and missiles, electrical equipment, professional and scientific instruments, transportation equipment (not including aircraft and missiles), and transportation and utilities (which are in the nonmanufacturing sector). On the other hand, in these same sectors much of the economic activity is carried out by large firms to begin with, so the observation that most of the R&D in these sectors is also conducted by large firms is not surprising.
In addition to absolute levels of, and changes in, R&D expenditures, another key indicator of the health of industrial S&T is R&D intensity. R&D is similar to sales, marketing, and general management expenses in that it is discretionary, i.e., a nondirect revenue-producing item that can be trimmed when profits are falling. There seems to be considerable evidence, however, that R&D enjoys a high degree of immunity from belt-tightening endeavors, even when the economy is faltering, because of its crucial role in laying the foundation for future growth and prosperity. Nevertheless, whether industry devotes the right amount of economic resources to R&D has remained an open question. See sidebar, "Does Industry Underinvest in R&D?"
There are numerous ways to measure R&D intensity; the one used most frequently is the ratio of R&D funds to net sales. This statistic provides a way to gauge the relative importance of R&D across industries and firms in the same industry. The industrial sectors with the highest R&D intensities in 1999 were scientific R&D services (32.1 percent), software (16.7 percent), communications equipment (11.6 percent), and computer systems design and related services (11.0 percent). Those with the lowest R&D intensities (less than 0.5 percent) were food, primary metals, broadcasting and telecommunications, and utilities. (See text table 4-10 .) For all industries combined, the ratio of R&D to sales was 2.7 percent in 1999.
Performance by Geographic Location, Character of Work, and Field of Science R&D by Geographic Location
The latest data available on the state distribution of R&D performance are for 1999. These data cover R&D performance by industry, academia, and Federal agencies, along with the federally funded R&D activities of nonprofit institutions. In 1999, total R&D expenditures in the United States were $244.1 billion, of which $231.8 billion could be attributed to expenditures within individual states, with the remainder falling under an undistributed, "other/unknown" category. (See appendix tables 4-21 and 4-22.) The statistics and discussion below refer to state R&D levels in relation to the distributed total of $231.8 billion.
R&D is substantially concentrated in a small number of states. In 1999, California had the highest level of R&D performed within its borders$48.0 billionrepresenting approximately one-fifth of the $231.8 billion U.S. total. The six states with the highest levels of R&D performance, California, Michigan, New York, Texas, Massachusetts, and Pennsylvania (in descending order), accounted for approximately one-half of the entire national effort. (See text table 4-11 .) The top 10 states (the six above-mentioned states plus New Jersey, Illinois, Washington, and Maryland) accounted for approximately two-thirds of the national effort. (See appendix table 4-23.) Californias R&D performance was 2.5 times as large as the R&D performance of the second highest state, Michigan, at $18.8 billion. After Michigan, ranking third was New York, with $14.1 billion, and the lowest of the top 10 states, Maryland, had $8.1 billion in R&D. The 20 highest ranking states in R&D expenditures accounted for 86.0 percent of the U.S. total; the lowest 20 states accounted for 4.5 percent.
States vary widely in the size of their economies because of differences in population, land area, infrastructure, natural resources, and history. Consequently, variation in the R&D expenditure levels of states may simply reflect differences in their economic size or the nature of their R&D efforts. A basic way of controlling for this "size effect" is to measure each states R&D level as a proportion of its gross state product (GSP). (See appendix table 4-23.) Like the term used in reference to the ratio of industrial R&D to sales, the proportion of a states GSP devoted to R&D is referred to as R&D "intensity" or "concentration." Overall, the nations total R&D to GDP ratio in 1999 was 2.63 percent. The top 10 rankings for R&D intensity were, in descending order, New Mexico (6.4 percent), Michigan (6.1 percent), Rhode Island (5.1 percent), Massachusetts (4.6 percent), Maryland (4.6 percent), the District of Columbia (4.5 percent), Washington (4.0 percent), California (3.9 percent), Delaware (3.9 percent), and Idaho (3.8 percent).
States have always varied in terms of the levels and types of industrial operations they contain. Thus, they also vary in the levels of R&D they contain by industrial sector. One measure of such variation among states is the extent to which their industrial R&D is in the manufacturing sector as opposed to the nonmanufacturing sector. Among the top 10 states in 1999 in industrial R&D performance, California, Massachusetts, Ohio, Texas, and Washington all had relatively low shares of R&D in the manufacturing sector (less than 64 percent, which was the national average). Higher levels of R&D in manufacturing, as a percentage of the total, were observed for Illinois, Michigan, New Jersey, New York, Ohio, and Pennsylvania. Among these 10 states, Michigan had the highest ratio of 92 percent, and Texas had the lowest ratio of 40 percent (industrial R&D in the manufacturing sector as a percentage of total industrial R&D). Part of this variation is attributable to differences among states in terms of their relative proportions of manufacturing and nonmanufacturing industries. Michigan, for example, is concentrated in motor vehicle manufacturing, and California devotes a great deal of R&D to software development and agricultural research. In Texas, 25 percent of industrial R&D performance took place in its computer and electronic products sector and another 20 percent in mining and extraction (including drilling for petroleum). Other factors, besides the locations of industrial production, may also play a role. For example, industries tend to perform research near universities that conduct the same type of research, enabling them to benefit from local academic resources.
Trends in National R&D by Character of Work
One traditional way to analyze trends in R&D performance is to examine the amount of funds devoted to basic research, applied research, and development. Admittedly, the traditional categories of basic research, applied research, and development do not always ideally describe the complexity of the relationship between science, technology, and innovation. However, alternative and perhaps more realistic models of the innovation process are probably too complicated to be used in collecting and analyzing comparable and reliable data for policymaking purposes and would not enable time-series analyses. See sidebar, "Choice of Right R&D Taxonomy Is a Historical Concern," later in the chapter. Nonetheless, in spite of these analytical limitations, these categories generally are useful to characterize the relative expected time horizons and types of investments.
The nation spent $47.9 billion on the performance of basic research in 2000, $55.0 billion on applied research and $161.7 billion on development. (See text table 4-1 .) These totals are the result of continuous increases over several years. Namely, since 1980 they reflect a 5.5 percent annual increase, in real terms, for basic research; a 3.9 percent increase for applied research; and a 3.8 percent increase for development. As a share of all 2000 R&D performance expenditures, basic research represented 18.1 percent, applied research represented 20.8 percent, and development represented 61.1 percent. These shares have not changed very much over time. For example, in 1980 basic research accounted for 13.9 percent, applied research accounted for 21.7 percent, and development accounted for 64.3 percent.
Basic Research. In terms of support, the Federal Government has always provided the majority of funds used for basic research. (See figure 4-12 .) However, its share of funding for basic research as a percentage of all funding has fallen substantially, from 70.5 percent in 1980 to 48.7 percent in 2000. This decline in the Federal share of basic research support does not reflect a decline in the actual amount of Federal support, which, in fact, grew 3.5 percent per year in real terms between 1980 and 2000. Rather, it reflects a growing tendency for the funding of basic research to come from other sectors. From 1980 to 2000, industrys self-reported support for basic research grew at the rate of 10.0 percent per year in real terms.
With regard to the performance of basic research in 2000, universities and colleges (excluding FFRDCs) accounted for the largest share with 43.1 percent ($20.7 billion), followed by industry with 32.1 percent ($15.4 billion). Their performance of basic research has undergone, on average, a 4.8 percent real annual increase since 1980. University-administered FFRDCs accounted for another 5.9 percent of total basic research performance in 2000. The dominant role played by universities and colleges in basic research is clearly related to the leading role that universities have in expanding general knowledge of S&E. Along the lines that general knowledge of science is a public good, the Federal Government provided 58.0 percent of the funding for basic research performed by universities and colleges. Non-Federal sources (industry, state and local governments, universities and colleges, and nonprofit organizations) provided the remaining 42.0 percent.
Applied Research. Applied research expenditures total $55.0 billion in 2000 and are performed much more by nonacademic institutions. They have been subject to greater shifts over time because of fluctuations in industrial growth and Federal policy. Applied research experienced a substantial average annual real growth of 7.4 percent between 1980 and 1985, followed by very low growth of 1.1 percent between 1985 and 1994, then rose again to 5.1 percent between 1994 and 2000. Increases in industrial support for applied research explain this recent upturn. Industrial support accounts for 66.1 percent ($36.4 billion) of the 2000 total for applied research and Federal support for 26.3 percent ($14.5 billion).
In the past two decades, Federal support for applied research has been intentionally deemphasized in favor of basic research. Consequently, in 2000 Federal funding for applied research is only 62.0 percent of that for basic research ($14.5 billion versus $23.3 billion, respectively), as reported by research performers.
Most applied research in calendar year 2000 (68.4 percent, or $37.6 billion) was performed by industry. In the same year, most of the nations nonindustrial applied research was performed by universities and colleges and their administered FFRDCs ($8.7 billion) and the Federal Government ($5.8 billion). For Federal intramural applied research (for which data are organized by fiscal year), 24.7 percent in FY 2000 was performed by HHS, 21.8 percent in FY 2000 was performed by DOD, and 11.6 percent was performed by DOC. Total Federal applied research performance has been remarkably level for 34 years, experiencing only a 0.8 percent average annual growth, in real terms, since 1966.
Development. Expenditures on development in calendar year 2000 totaled $161.7 billion, accounting for most of R&D expenditures. Therefore, historical patterns of development expenditures mirror historical patterns of total R&D expenditures. From 1980 to 1985, development grew on average by 7.2 percent per year in real terms as increasingly larger shares of the national R&D effort were directed toward R&D supported by DOD, which tends to be approximately 90 percent development. (See figure 4-13 .) Between 1985 and 1994, on the other hand, development in real terms grew at an average annual rate of only 0.7 percent, from $74.5 billion in 1985 to $103.0 billion in 1994. Between 1994 and 2000, annual growth was back up to 5.9 percent in real terms to $161.7 billion in 2000, of which 79.4 percent was supported by industry and 19.7 percent by the Federal Government.
In terms of performance, industry accounted for 89.2 percent ($144.3 billion) of the nations 2000 development activities, the Federal Government 6.1 percent ($9.8 billion), and all other performers 4.7 percent ($7.6 billion).
Federal Obligations for Research, by Field
Federal obligations for research alone (excluding development) will total $38.7 billion in FY 2001 by preliminary estimates. Life sciences will receive the largest portion of this funding (47.2 percent, or $18.2 billion), most of which will be provided by HHS. (See figure 4-14 .) The next largest field in Federal obligations for research in FY 2001 will be engineering (18.3 percent), followed by physical sciences (11.5 percent), environmental sciences (8.4 percent), and mathematics and computer sciences (6.5 percent). Social sciences, psychology, and all other sciences will account for another 2.6 percent, 1.9 percent, and 3.6 percent, respectively.
In terms of agency contributions to these research efforts, HHS, primarily through NIH, will provide the most (42.8 percent) of all Federal research obligations in FY 2001. The next largest contributor will be NASA (12.2 percent) with substantial funding of research in engineering ($2.2 billion), physical sciences ($0.9 billion), and environmental sciences ($1.1 billion). (See figure 4-14 .) DOE will provide 11.7 percent of research funding, primarily in the fields of engineering, physical sciences, and mathematics and computer sciences. DOD will fund a similar amount of research (11.4 percent of the total), primarily in the areas of engineering and mathematics and computer sciences. NSF will provide 8.2 percent of research funding, with between $0.5 and $0.7 billion contributed to each of the following fields: life sciences, engineering, physical sciences, environmental sciences, and mathematics and computer sciences.
Federal obligations for research have grown at different rates for different fields of S&E, reflecting changes in perceived public interest in those fields, changes in the national resources (e.g., scientists, equipment, and facilities) that have been built up in those fields over time, as well as differences in scientific opportunities across fields. Based on preliminary estimates for FY 2001, the broad field of mathematics and computer sciences has experienced the highest rate of growth in Federal obligations for research, which was 8.3 percent per year in real terms between 1980 and 2001. Life sciences had the second highest rate with 3.9 percent, followed by psychology with 3.2 percent, environmental sciences with 1.3 percent, engineering with 1.2 percent, and physical sciences with 0.6 percent. Research in the social sciences (including anthropology, economics, political sciences, sociology, and other areas) experienced a slight decline of 0.12 percent.
These trends in Federal support for the above-mentioned broad fields of research, however, may not reflect trends for the smaller fields that they contain. For example, with regard to the broad field of mathematics and computer sciences, Federal support for research in mathematics grew by 3.8 percent per year in real terms between FY 1980 and FY 1999, whereas support for research in computer sciences grew by 10.2 percent. During the same period, within life sciences, support for biological and agricultural research grew by 1.7 percent compared with research support for medical sciences, which grew by 4.6 percent. Within the physical sciences, support for astronomy grew by 1.8 percent, whereas support for chemistry declined by 0.23 percent.
Cross-Sector Field-of-Science Classification Analysis
Federal and academic research expenditures are often classified according to the S&E fields they support. However, it may also be useful to classify all R&D activity by specific S&E fields. Such classification, when applied to historical data, would indicate how R&D efforts in various fields of S&E have grown in economic importance over time. This information is potentially useful for science policy analysis and for planning and priority setting.
Classification of academic R&D by field of science is provided in detail in chapter 5. At present, the only additional sector for which there exist extensive data by field is the Federal Government. Industrial R&D, which represents three-fourths of all R&D performed in the United States, is not collected by field of study for three reasons:
- Unlike universities and Federal agencies, most private companies do not have the recordkeeping infrastructure in place to compile such statistics; thus, any efforts on their part to provide this additional information could be significantly burdensome to them.
- Much of the research by private firms is confidential, and the provision of such information to outsiders might compromise that confidentiality.
- Much of the R&D carried out by industry is interdisciplinary, especially at the development stage (e.g., the development of a new vehicle would involve mechanical engineering, electrical engineering, and other fields), which in many cases might make the splitting of R&D by field somewhat arbitrary.
Nonetheless, some analysis by field of study, wherever possible, shed light on overall levels of R&D support for general lines of inquiry. In particular, this problem can be circumvented by grouping fields within standard industrial categories, thereby creating categories of R&D that can be associated both with S&E fields and with related industrial categories. We focus in particular in two broad areas, chemistry (nonmedical) and chemical engineering, and life sciences. For ease in data interpretation, all academic and Federal FY data were converted to calendar year data so that they would be comparable to the data pertaining to industry categories (which are collected and provided on a calendar year basis).
R&D in Chemistry (Nonmedical) and Chemical Engineering. In 1998, R&D in the broad area of chemistry and chemical engineering accounted for approximately $10.3 billion (in constant 1996 dollars). Three categories of R&D were identified in this area. (See figure 4-15 .) The largest of these categories, by far, is company-funded R&D in industrial chemicals and other chemicals (but not pharmaceuticals and medicines). In real terms (constant 1996 dollars), expenditures in this category grew from $6.6 billion in 1985 to $8.8 billion in 1998, although the sector has displayed considerable year-to-year fluctuation between 1996 and 1998 (inclusive). The next two categories were much smaller. Federal obligations for research in chemistry and chemical engineering declined between 1985 and 1998, from $1.2 to $980 million (in constant 1996 dollars). Academic R&D (not federally funded) in chemistry and chemical engineering, the smallest category, grew steadily in real terms, from $237 million in 1985 to $444 million in 1998.
R&D in Life Sciences. The broad life sciences field accounted for $36.5 billion of R&D in 1998 (in constant 1996 dollars). R&D in this area is characterized by strong and fairly continuous real growth in its three largest categories. (See figure 4-16 .) The largest of these three, Federal obligations for research in the life sciences, plus development expenditures by HHS and the Department of Veterans Affairs, rose from $9.3 billion in 1985 to $15.4 billion in 1998 in constant 1996 dollars. Company-funded R&D in pharmaceuticals and medicines grew dramatically in real terms, from $4.7 billion in 1985 to $10.4 billion in 1995 but then declined to $9.3 billion by 1998. In contrast, academic R&D (not federally funded) in life sciences and bioengineering/biomedical engineering grew continuously, from $3.0 billion in 1985 to $6.3 billion in 1998.
With regard to food and other traditional products, however, company-funded R&D in food, beverage, and tobacco products, and development expenditures by USDA, show virtually no real R&D growth. That is, as shown in figure 4-16 , R&D for this combined subcategory grew only from $1.6 to $1.7 billion between 1985 and 1998. Finally, two new categories of industrial R&D in the life sciences, arising from the new NAICS classification system, are company-funded R&D in health care services and company-funded R&D in medical equipment and supplies. In 1998, the former accounted for $566 million in R&D and the latter for $3.3 billion, in constant 1996 dollars.