These trends were partially offset by increased R&D investment by companies classified in the service sector, but overall R&D spending declined, in real terms, during the 1990s. (See appendix table 4-8.) The following sections contain more detailed information about these and other trends in industrial R&D spending.
Information from selected large industrial R&D performers indicate that industry expects some improvement in 1996 (IRI, 1996). For additional information about the current climate for industrial R&D, see "Biggest" Problems Facing Technology Leaders and Industrial Research and Development Planning Environment.
Research and Development in Manufacturing Versus Nonmanufacturing Industries
Probably the most striking trend in industrial R&D performance in recent years is the growing prominence of the service sector. Until fairly recently (the late 1980s), little attention was paid to R&D conducted by nonmanufacturing companies, largely because service sector R&D activity was minuscule compared with the R&D operations of companies classified in manufacturing industries.
Prior to 1983, nonmanufacturing industries accounted for less than 5 percent of the industry total. A decade later, the R&D landscape looked very different. Between 1983 and 1993, there was a steady increase in the proportion of total industrial R&D performed by companies classified in service industries. (See figure 6-10 in Chapter 6, Technology Development and Diffusion.) In 1993, nonmanufacturing firms were responsible for more than one-quarter of all industrial R&D performed in the United States. That year, their expenditures amounted to $25.5 billion in company and other nonfederal funds and $5.7 billion in Federal funds. (See appendix table 4-11.)
Three industry groupings account for the bulk of R&D performed in the service sector:
A sizable portion of the R&D dollars in two of these industries - computer software and communication services - was spent by companies formerly classified in manufacturing industries. Given the growing importance of computer software (relative to hardware) and other information technologies, this shift from manufacturing to nonmanufacturing is not unexpected.
In addition, because the United States invests a relatively large share of its resources in health care, about 13.9 percent of GDP in 1993 (Department of Health and Human Services, 1993), the increasing importance of R&D laboratories in the Nation's industrial R&D portfolio is also predictable. It can be attributed, in large part, to major advancements in research on the human body, the establishment and growth of a variety of medical research facilities, and the maturing and success of the biotechnology industry.
Concurrent with the growing visibility of R&D in the service sector, R&D performed by manufacturing industries lost some of its pre-eminence. Still, the manufacturing sector continues to dominate the R&D spectrum.
In 1993, the six largest manufacturing industries, in terms of companies' own (and other nonfederal) R&D expenditures in the United States were
The share of total nonfederal R&D funds held by five of the six industries fell during the late 1980s and early 1990s. Only the chemicals and allied products industry had an increase in its share of the total - from 15 percent in 1987 to 17 percent in 1993. The machinery industry's decline was particularly acute - from 17 percent of the total to 9 percent. (In 1987, almost all computer manufacturers were included in this group; some of these companies were reclassified into service industries in subsequent years.) Between 1987 and 1993, the chemicals and machinery industries switched places - the chemicals industry became the second largest R&D performing industry (it had been fourth), and the machinery industry dropped from second to fourth. Although the rankings of the other major R&D-performing manufacturing industries stayed the same, the share of the national R&D total held by firms classified in the electrical equipment industry declined from 17 percent to 12 percent, and the transportation equipment industry share dropped from 22 percent to 18 percent.
Research and Development Expenditures by Size of Company
In 1993, 125 companies with more than 25,000 employees spent more than $1 million each on R&D in the United States (SRS, 1995b). Prior to 1990, this group of companies accounted for more than half the nonfederal R&D total. That share has fallen below 50 percent because the R&D expenditures of firms with fewer than 500 employees have been increasing faster than those of companies in the other size-groups.7 Small firms' share of the total increased from 10 percent in 1990 to 15 percent in 1993. (See appendix table 4-9.)
U.S. industrial R&D expenditures are heavily concentrated in a relatively small number of firms. In 1993, the 4 largest R&D performing companies (in terms of nonfederal funds) accounted for 17 percent of the total amount spent; the 20 largest, 33 percent; and the 200 largest, 71 percent. (See appendix table 4-12.)
Over a 10-year period, 1984-94, some major membership changes occurred in the annual list of 100 leading R&D-performing companies according to Standard & Poor's Compustat Services, Inc.8 (See appendix table 4-14.) The four largest R&D-performing companies, however, were the same in both years (although their order changed). That may be one of the few constants revealed by comparing the lists from 1984 and 1994. There were some major changes in rankings among the remaining 96 entries. For instance,
Research and Development Intensity
In addition to absolute levels of and changes in R&D expenditures, another key indication of the health of industrial S&T is R&D intensity. There are a number of ways to measure this indicator, but 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 ratio of R&D dollars to net sales tends to be fairly stable over time, although year-to-year changes of 0.1 to 0.2 percent are not uncommon. Also, there are substantial differences between industries. (See appendix tables 4-13 and 4-14.)
In 1993, nonfederal R&D spending, as a percent of net sales for all R&D-performing companies classified in manufacturing industries, was 3.1 percent. (See appendix table 4-13.) Although this was a 0.2-percent dip from the preceding year's ratio, the change may be a minor fluctuation and not the beginning of a downward trend. The 1993 ratio is very much in line with that recorded for other recent years; for instance, this ratio has been 3.0 percent or higher since 1985. (See figure 4-8.)
The general stability of the R&D/sales ratio in the 1990s indicates that, despite a lack of growth in manufacturing companies' R&D financing, little change has occurred in the level of importance accorded R&D relative to other discretionary spending. That is, roughly the same proportion of companies' income has been devoted to R&D for almost 10 years.9
As mentioned previously, there are significant differences in R&D intensity across industries. (See text table 4-3.) At 12.1 percent, the pharmaceutical industry had the highest, and only double-digit, ratio in 1993. Ten years earlier, it ranked third behind the computer and scientific and mechanical measuring instruments industries.
Federal Research and Development Funds
In 1993, industrial firms spent $22.8 billion in Federal funds on R&D. As mentioned earlier in this chapter, Federal R&D support to industry has been declining steadily since 1987. The aircraft and missiles industry is the leading recipient of Federal R&D funds. Interestingly, this industry used to account for more than two-thirds of all Federal monies spent by companies; the most recent data show this industry accounting for less than one-half in 1993.10 (See appendix table 4-10.)
There has been a major upsurge in company downsizing, mergers, and restructuring in recent years, particularly among defense contractors. Company executives have had to make some tough choices; for example, they have had to decide whether to stick with the defense business as the mainstay of their operations or branch out into civilian product development. (See Defense Conversion: Will It Prove Successful?)
Collaboration Among Firms and with Other Organizations
Cooperative R&D is now an important tool in the development and leveraging of S&T resources. Although data on multifirm and multisector collaborative R&D activities are sparse, there is sufficient evidence to conclude that a major upswing in S&T partnerships has taken place since the early 1980s.11
There is also a growing body of literature that helps identify the reasons for the increase in collaborative R&D efforts, their organizational structure, and their economic and political implications. For example, intra-industry collaboration seems to be a response to the same competitive forces affecting all industries: rising R&D costs and risk in pursuing basic research, shortened product life cycles, increasing multidisciplinary complexity of technologies, and intense foreign competition in domestic and global markets. In addition, high-profile, industry-led cooperative research endeavors are more likely than single-firm efforts to attract support from outside sources, such as Government agencies. This section covers several indicators of cooperative R&D activity.
Industrial Research and Development Consortia
The erosion in U.S. firms' technological leadership and the decline in their share of international markets in the 1970s and 1980s have been blamed on many factors (NSB, 1992), including an overly restrictive antitrust environment, which has, in effect, prohibited U.S. companies from collaborating on most activities, including research. Laws originally enacted to preserve domestic competition had become an impediment to U.S. firms' ability to compete in worldwide markets populated by an increasing number of foreign producers and customers. Therefore, restrictions on multifirm cooperative research relationships were lifted with the passage of the National Cooperative Research Act (NCRA) of 1984. This law was designed to encourage U.S. firms to collaborate on generic, precompetitive research. To gain protection from antitrust litigation, firms engaging in a "joint research venture" (JRV)12 are required by NCRA to register the JRV with the Department of Justice. In 1993, Congress again relaxed restrictions - this time on cooperative production activities - by passing the National Cooperative Research and Production Act, which enables companies to work together in a JRV to apply newly developed technologies.
By the end of 1994, more than 450 JRVs had been registered. The annual number of JRV filings has increased in most years since the law was passed. (See figure 4-9.) Although data on the level of resources invested in these projects are not available, results of two investigations (Link, 1995; Vonortas, 1995) revealed the following:
Federal Cooperative Technology Programs
Federal cooperative technology programs13 have mushroomed in the past 15 years, a trend reflecting major changes in U.S. S&T policy.14 That is, the Federal role in supporting S&T programs has been extended beyond the traditional funding of mission-oriented R&D. It now includes support aimed at speeding the development, application, and commercialization of new technologies in areas likely to contribute to economic growth and other societal needs. Factors shaping this new S&T policy include concern about international industrial competitiveness; a desire to maximize the value of Federal investment in S&T; the declining defense budget; and the military's increasing dependence on civilian-led technologies, especially in the area of electronics.
Since 1980, a series of laws have been enacted to promote Federal-civilian partnerships and to facilitate the transfer of technology15 between sectors. (See text table 4-4 and SBIR Program Expands Support for Small Business R&D.) Terms such as partnership and cost sharing, once rarely heard in government agencies, are now part of the lexicon in most Federal R&D programs. For example, although the dollar amounts are small, the National Science Foundation went from having almost no programs requiring industry cost sharing in 1980 to more than 50 such programs in 1995.
There are now more than 70 Federal cooperative technology programs spread over at least 10 agencies. In fiscal year 1994, Federal agencies spent approximately $2.7 billion on cooperative technology programs. (See text table 4-5.) These cooperative technology programs can be divided into five categories: technology development, industrial problem solving, technology financing, startup assistance, and teaming. The largest of these groups is technology development, with FY 1994 total funding of more than $1 billion. Programs in this group provide assistance to companies to develop or adapt technology and include those that require companies to enter into consortia with other companies, Government agencies, or universities. Examples of these programs are DOD's Manufacturing Technology (MANTECH) program, SEMATECH, the Technology Reinvestment Project (TRP), and NSF's Research Centers programs.
Technology Transfer Activities
Technology transfer activities are now an important mission component of Federal laboratories. (See The National Laboratories - A New Role? later in this chapter.) Of course, some agencies, including the Department of Agriculture's (USDA) Agricultural Research Experiment Stations and the National Aeronautics and Space Administration's (NASA) civilian aeronautics programs, have long shared their research with the private sector. According to most available indicators, Federal efforts to facilitate private sector commercialization of Federal technology have made considerable progress since 1987. For instance,
In addition to direct financial R&D support, the Government has tried other policy instruments to stimulate corporate research spending indirectly. The most notable of these efforts has been to offer tax credits on incremental research and experimentation (R&E) expenditures.16 The credit was first put in place in 1981 and has been renewed six times - most recently, through the end of June 1995.17 Although the computations are complicated, the tax code provides for a 20-percent credit for the amount of a company's qualified R&D that exceeds a certain threshold.18 Since 1986, companies have been allowed to claim a similar credit for basic research grants to universities and other qualifying nonprofit institutions, although the otherwise deductible R&E expenditures are reduced by the amount of the basic research credit.19
The dollar value of R&E tax credits actually received by firms is unknown. Information from the Internal Revenue Service indicates that in any given tax year this dollar value can be 20 to 30 percent less than the amount for which firms file claims - nearly $1.6 billion in 1992, the most recent year for which data are available (OTA, 1995b). This amount has fluctuated since the credit's inception in 1981, but has remained rather steady since 1988. (See figure 4-11 and appendix table 4-16.)
Additionally, as part of the Federal budget process, the U.S. Treasury Department annually calculates estimates of foregone tax revenue (tax expenditures) due to preferential tax provisions, including the R&E tax credit. As one measure of budgetary effect, the Treasury provides outlay-equivalent figures. These allow a comparison of the cost of this tax expenditure with the cost of a direct Federal R&D outlay. Between 1981 and 1994, more than $24 billion was provided to industry through this indirect means of Federal R&D support - an amount equivalent to about 3 percent of direct Federal R&D support. (See appendix table 4-16.)
In general, most of the credit has been claimed by manufacturing firms, which accounted for three-fourths of the $1.6 billion total in 1992. Companies that took the most advantage of the credit were large firms and those that produce pharmaceuticals (17 percent of total), electrical equipment and transportation equipment (14 percent each), and machinery (11 percent). Since 1981, the nonmanufacturing share of claimed credits has risen from less than 20 percent to 24 percent of the total.
The tax credit has received mixed reviews regarding its overall effectiveness. Results of various studies undertaken since the mid-1980s were summarized in a recent literature review and 1995 experts workshop supported by the Office of Technology Assessment (1995b):
8 The ranking is based on the size of R&D expenses. Included are the amount of funds spent on R&D performed outside the company and the amount spent on R&D performed in other countries. Note that the Compustat data base does not contain privately held companies.
9 It is important to note that there were significant increases in the overall R&D/net sales ratio between 1981 and 1982 (from 2.2 percent to 2.6 percent) and between 1984 and 1986 (from 2.7 percent to 3.2 percent). Prior to 1982, company R&D funds, as a percent of net sales, had been in the 2.0-percent range for 20 years.
10 Data from the Aerospace Industries Association (AIA) show a 25-percent decline in sales of military-related hardware between 1990 and 1993 (AIA, 1994).
11 This section covers only domestic alliances. See International Strategic Technology Alliances in this chapter for information on international collaborative R&D activities.
12 A JRV is defined as the formation of a new organization jointly controlled by two or more parent institutions whose purpose is to engage in R&D activities. Members can be from different sectors as well as different countries.
13 Cooperative technology programs are defined here as public-private initiatives involving Government and industry that sponsor the development and use of technology and improved practices to benefit specific companies. The primary goal is to stimulate economic growth. Excluded are the nonstate or nonagency programs of individual institutions, such as Federal laboratories and universities. Also excluded are programs for which economic benefits are secondary or tertiary goals; for instance, programs that have a main purpose to educate or meet agency mission requirements.
14 Most of the information in this section was obtained from Coburn, 1995.
15 Technology transfer can cover a wide spectrum of activities, running the gamut from the exchange of ideas between visiting researchers to contractually structured research collaborations involving the joint use of facilities and equipment.
16 Not all R&D is eligible for such credit, which is limited to expenditures on laboratory or experimental R&D.
17 For a complete history of the tax credit, measurement of its effectiveness, and comparisons with R&D tax provisions enacted in other countries, see Office of Technology Assessment (1995b).
18 The complex base structure for calculating qualified R&D spending was put in place by the Omnibus Reconciliation Act of 1989. With various exceptions, a company's qualifying threshold is the product of a fixed-base percentage multiplied by the average amount of the company's gross receipts for the 4 preceding years. The fixed-base percentage is the ratio of R&E expenses to gross receipts for the increasingly distant 1984-88 period. Special provisions cover start-up firms.
19 In 1992, firms applying for the R&E credit spent about $1 billion on research performed by educational and scientific organizations, of which - after various qualification restrictions - the basic research credit contributed less than $200 million toward the R&E tax credit (OTA, 1995b).
20 Whatever its ultimate impact on R&D spending, the tax credit has certainly influenced spending less than it could have, had it been less subject to erratic legislative treatment. The tax credit has had to be repeatedly (almost annually) renewed, its calculation provisions have changed considerably over the years, and it was even allowed to lapse several times - circumstances that created considerable uncertainty for businesses that would otherwise have planned to take the tax credit.