In the private sector, international R&D collaboration is also on the rise, as is indicated by the growth of formal cooperative partnerships between firms and of overseas R&D activities undertaken under contracts, through subsidiaries, and with the establishment of independent research facilities.45 Although the reasons for this growth are complex, generally it appears that multilateral industrial R&D efforts are a response to the same competitive factors affecting all industries: rising R&D costs and risks in product development, shortened product life cycles, increasing multidisciplinary complexity of technologies, and intense foreign competition in domestic and global markets.
As the numbers have increased, the forms of cooperative activity have changed somewhat. The most prevalent modes of global industrial R&D cooperation in the 1970s were joint ventures and research corporations. In these arrangements, at least two companies share equity investments to form a separate and distinct company; profits and losses are shared according to the equity investment.47 In the second half of the 1980s and continuing into the 1990s, joint nonequity R&D agreements became the most important form of partnership. Under such agreements, two or more companies organize joint R&D activities to reduce costs and minimize risk, while pursuing similar innovations. The participants share technologies but have no joint equity linkages.
Formation of these so-called strategic technology alliances (both equity and nonequity arrangements) are particularly extensive among high-tech firms; data are available for the three advanced technologies listed above.48 Reflecting the general rise in importance of information technologies (IT) across industries and throughout society, growth in IT research alliances far outdistanced the growth related to biotechnology and materials. (See figure 4-23.) Between 1980 and 1994, more than 2,800 IT alliances were reported between firms involved in computer software and hardware, telecommunications, industrial automation, and microelectronics. Internationally, the largest number of such relationships was between U.S. and European firms, although the number of intra-European alliances was also substantial. (See appendix table 4-38.)
Since 1985, U.S. firms generally increased their annual funding of R&D performed outside the country. (See appendix table 4-40.) Indeed, from 1985 to 1993, U.S. firms' investment in overseas R&D increased three times faster than investment in R&D performed domestically (9.3- versus 3.1-percent average annual constant dollar growth). Accounting for the equivalent of about 6 percent of industry's domestic R&D funding in 1985, overseas R&D now amounts to more than 10 percent of U.S. industry's on-shore R&D expenditures. (See figure 4-24.) Additionally, according to data from the Bureau of Economic Analysis (BEA), the majority-owned (i.e., 50 percent or more) foreign affiliate share of U.S. multinational companies' worldwide R&D expenditures increased from 9 percent in 1982 to 13 percent in 1990, where it remained through 1993 (Mataloni, 1995).
R&D investment by U.S. companies and their foreign subsidiaries in the chemicals (including pharmaceuticals and industrial chemicals) industry accounts for the largest share and growth of this foreign-based R&D activity. Indeed, drug companies accounted for almost 30 percent of total 1993 overseas R&D ($9.8 billion), which was equivalent to 17 percent of the pharmaceutical industry's domestically financed R&D. (See figure 4-24.) Of other major R&D-performing manufacturers, recent trends (since 1990) show the overseas R&D investment share of total R&D financing rising considerably for scientific instruments, but declining for machinery and electrical equipment. In each of these cases, however, the funding shifts primarily reflect changes in industry classification of major R&D-performing firms rather than actual reductions in total overseas R&D.
Similarly, the combined total for all nonmanufacturing industries shows substantial increases in foreign R&D activity since 1985, rising from 0.4 to 7.0 percent in 1993. Part of this growth is indicative of increased international R&D financing by firms historically classified as nonmanufacturing industries (particularly computer, data processing, and architectural services). Part of the increase reflects the movement of firms previously classified as manufacturers (for example, office computing companies) to service sector industries (for example, software development).
Most of the U.S. overseas R&D is undertaken in Europe. As indicated by Bureau of Economic Analysis (BEA) data on majority-owned foreign affiliates of nonbank U.S. multinational companies, most of this 1993 R&D total was performed in Europe - primarily Germany (23 percent of the U.S. overseas total), the United Kingdom (15 percent), France (9 percent), and Ireland (6 percent). Collectively, however, the current 70-percent European share of the U.S. total R&D investment abroad is considerably less than the 76-percent share reported as recently as 1991. Since the early 1980s the U.S. R&D investment abroad has shifted more generally away from Europe and Canada and toward Japan and other Asian countries. These shifts reflect, in part, an appreciation for, and a desire to take advantage of, the increase in S&T strengths underway in Asia.50 (See figure 4-25.)
By affiliate industry classification, more than one-half of the 1993 German-based R&D was performed by transportation equipment companies. In the United Kingdom and France, the chemicals industry accounted for the largest share of each countries' respective totals, whereas in Ireland, the machinery industry performed most of this U.S.-funded R&D. Canada accounted for 9 percent of U.S. companies' 1993 R&D performed abroad, and Japan - which receives comparatively small amounts of foreign R&D funds - accounted for 8 percent.51 (See text table 4-10 and appendix table 4-41.) Notably, the U.S. R&D investment in Asian countries other than Japan has grown substantially. U.S. R&D spending in Singapore (primarily in machinery industries) and Indonesia (mostly for petroleum-related research) now surpasses that in many European nations.
Like U.S. firms' overseas R&D funding trends, R&D activity by foreign-owned companies in the United States has increased significantly since the early 1980s. From 1980 to 1993, inflation-adjusted R&D growth from foreign firms (U.S. affiliates with a foreign parent that owns 10 percent or more of the voting equity) averaged 12 percent per year, or more than three times the rate of growth in domestic R&D activities by U.S. companies (3.9 percent).54 Using a more restrictive definition of foreign ownership in which the foreign parent owns 50 percent or more of the voting equity makes little difference in the aggregate trends. U.S.-performed R&D by such firms grew on average by about 12 percent annually from 1980 through 1993. (See figure 4-26.)
Much of this foreign R&D growth occurred since the mid-1980s, just as U.S. firms' domestic R&D investments were beginning to slow. As a result, foreign R&D was equivalent to 12 percent ($14.6 billion) of total industrial R&D performance in the United States in 1993 - or double that of its equivalent 6-percent share in 1985. Majority-owned affiliates accounted for 10 percent ($11.6 billion) of the U.S. 1993 industrial performance total. Alternatively, as a percentage of total foreign and U.S. firms' industrial R&D funding, foreign companies accounted for 15 percent in 1993 (majority-owned affiliates accounted for 12 percent) compared with 9 percent in 1985. Although the R&D flows from Canada and other European countries also increased steadily over the past decade, 80 percent of this foreign funding came from five countries - Japan, Switzerland, the United Kingdom, Germany, and France. Swiss and Japanese firms increased their R&D investment in the United States more rapidly than did companies from the other nations. At least part of the significant expansion of these foreign R&D expenditures in the United States is attributable to several major acquisitions by foreign multinational companies of U.S. firms, particularly of U.S. pharmaceutical and biotechnology firms with large R&D budgets. (See U.S. Research Facilities of Foreign Firms and text table 4-11.)
Foreign-funded research was concentrated in three industries in 1993 - industrial chemicals (funded predominantly by German and Canadian firms), drugs and medicines (mostly from Swiss and British firms), and electrical equipment (one-third of which came from German and Japanese affiliates). These three industries accounted for three-fifths of the $14.6 billion total 1993 foreign R&D investment. Concurrent with gains reported for all domestic U.S. R&D performance, foreign - particularly Japanese - R&D investment in the service sector has also risen considerably. These industries accounted for 8 percent of the 1993 foreign R&D investment total, with most research being funded in computer, data processing, and research and management services. (See text table 4-12 and appendix tables 4-42 and 4-43.)
45 For an extensive review of U.S. globalization trends in trade and investment, including business research and technology development, see recent Office of Technology Assessment reports (1993; 1994).
46 Information in this section is drawn from an extensive data base compiled in the Netherlands (Maastricht Economic Research Institute on Innovation and Technology's Co-Operative Agreements and Technology Indicators data base - MERIT-CATI) on more than 10,000 interfirm cooperative agreements involving thousands of different parent companies. In the CATI data base, only interfirm agreements that contain some arrangements for transferring technology or joint research are collected. The data summarized here extend by 4 years the information presented in Hagedoorn and Schakenraad (1993). These counts are restricted to strategic technology partnerships such as joint ventures for which R&D or technology sharing is a major objective, research corporations, joint R&D pacts, and minority holdings coupled with research contracts.
CATI is a literature-based data base; its key sources are newspapers, journal articles, books, and specialized journals that report on business events. CATI's main drawbacks and limitations are that (1) data are limited to activities publicized by the firm, (2) agreements involving small firms are likely to be underrepresented, (3) reports in the popular press are likely to be incomplete, and (4) it probably reflects a bias because it draws primarily from English-language materials. CATI information should therefore be viewed as indicative and not comprehensive.
47 Joint ventures are companies that have shared R&D as a specific company objective, in addition to production, marketing, and sales. Research corporations are joint R&D ventures with distinctive research programs.
48 For a review of strategic partnering trends in other technologies (chemicals, aviation/defense, automotive, and heavy electrical equipment), see Hagedoorn (1995).
49 Companies consider several factors before undertaking R&D overseas: market access and accommodation of local requirements are but two of these factors. Tax and regulatory policies, as well as the availability of trained researchers and access to new scientific and technological developments in other countries, also influence R&D location decisions.
50 For example, see chapter 6, Technology Development and Competitiveness, and reports on the recent growth in the indigenous S&T capacities of several Asian countries (SRS, 1993; 1995c).
51 These overseas R&D country shares are from the BEA survey on U.S. Direct Investment Abroad, not the NSF data series from which industry-specific shares are taken. The definition used by BEA for R&D expenditures is from the Financial Accounting Standards Board Statement No. 2; these expenditures include all charges for R&D performed for the benefit of the affiliate by the affiliate itself and by others on contract. BEA detail are available for 1982, and annually since 1989. NSF reports a 1993 overseas R&D total of $9.8 billion; BEA estimates overseas R&D expenditures by U.S. companies and their foreign affiliates at $11.0 billion.
52 For countries other than the United States, the data in this section are taken from OECD (1995). The foreign-sourced R&D data for the United States come from an annual survey of Foreign Direct Investment in the United States conducted by BEA. BEA reports that the foreign R&D totals are comparable with the U.S. R&D business data published by NSF. Industry-specific comparisons, however, are limited because of differences in the industry classifications used by the two surveys (Quijano, 1990).
53 These countries are Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, and the United Kingdom. See also OECD (1995) for a discussion of international R&D investment trends.
54 BEA considers all of an investment (including R&D) to be foreign if10 percent or more of the investing U.S.-incorporated firm is foreign owned. (See appendix table 4-38.) Special tabulations were prepared by BEA to reveal R&D expenditures in the United States of those firms in which there is majority foreign ownership - i.e., 50 percent or more. For 1993, the 10-percent foreign ownership threshold results in an estimated $14.6 billion foreign R&D investment total. R&D expenditures of majority-owned U.S. affiliates of foreign companies were $11.6 billion. (See appendix table 4-40.)