Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
The Contribution of R&D to U.S. Economic Development To assess the economic consequences of foreign participation in U.S. R&D, it is essential to understand what activities comprise research and development, how the processes of R&D function, and how readily the outputs of R&D are diffused internationally. At the same time, it is necessary to examine how R&D may affect national economic development and to consider some of the limita- tions on our understanding of this process. The chapter provides a brief overview of these issues. RESEARCH AND DEVELOPMENT: A SIMPLIFIED TAXONOMY The term "research and development" encompasses a range of organized activities directed at the discovery, assimilation, transfer, or application of knowl- edge. The National Science Foundation (NSF) classifies research and develop- ment into three categories: basic research, applied research, and development. Basic research seeks to advance scientific or technical knowledge or understand- ing of a particular phenomenon or subject without specific applications in mind. In contrast, applied research recognizes a specific need and seeks new knowledge or understanding in order to meet that need. NSF defines development as "the systematic use of the knowledge or understanding gained from research directed toward the production of useful materials, devices, systems, or methods, includ- ing design and development of prototypes and processes" (National Science Board, 19939. In practice, the boundaries between these three broad categories of organized R&D activity are often blurred. In addition, the three are interrelated through 29
30 FOREIGN PARTICIPATION IN U.S. RESEARCH AND DEVELOPMENT complex feedback loops. Sometimes the R&D process moves in a linear fashion, from basic research to applied research to development. Basic research may yield new knowledge that can be usefully applied. In addition, it may enable new applications of existing knowledge as well as suggest new directions for applied research and development. Basic research may also nourish and enhance applied R&D in ways that cannot be traced to a discrete piece of new knowledge. Just as often, however, applied research and development provides the impetus for pur- suing new directions in basic research. As noted by Brooks (1994), "Problems arising in industrial development are frequently a rich source of challenging basic science problems which are first picked up with a specific technological problem in mind, but then pursued by a related basic research community well beyond the immediate requirements of the original technological application that motivated them." For example, efforts to understand materials processes and properties critical to the quality and performance of semiconductor devices were largely responsible for the emergence of materials science as a field of academic re- search. The role of serendipity in research and development should not be underes- timated. Often, major advances in knowledge and new applications of existing knowledge are entirely unexpected by those who fund or perform R&D. Basic and applied research directed at the discovery or application of knowledge in a particular field or industry may yield findings that advance fundamental knowl- edge in disciplines unrelated to that of the R&D performer or suggest applica- tions of knowledge that are unrelated to the researcher's original objectives.) THE MULTIPLICITY OF R&D OUTPUTS R&D can have many different outputs, including codified knowledge, know- how or techniques, highly skilled human capital, instrumentation, and technol- ogy.2 However, as the NSF definitions suggest, each type of R&D activity tends to result in particular types of outputs. For example, with the exception of certain fields where direct transfers of knowledge from basic science to technology are frequent (such as in chemistry and molecular biology), basic research yields chiefly new knowledge, new methods, and skilled scientists and engineersout- puts that contribute indirectly to the development and application of technology. By contrast, applied research and development, while creating new knowledge, know-how, and skills, are generally more directly implicated in the generation and application of technology. Another useful way of classifying R&D outputs is to consider who "owns" them. Some R&D outputs are essentially nonproprietary, or public goods. That is, their use by one party does not diminish their value or utility to others, and they can be exploited freely by anyone possessing the requisite technical capa- bilities. Much of R&D conducted to advance explicit government missions, such as national defense or the cure of disease, falls into this category. Most of the
THE CONTRIBUTION OF R&D TO U.S. ECONOMIC DEVELOPMENT 31 products of basic research and generic applied research and development, such as new knowledge, know-how and techniques, skilled scientists and engineers, and generic technology (e.g., standards, metrics, manufacturing practices) can also be classified as nonproprietary. The vast majority of these R&D activities are funded, though not necessarily performed, by the government. The federal government supports roughly 40 percent of applied research and an equal proportion of development work in the United States, most of which is performed by private companies. The govern- ment also funds over two-thirds of U.S. basic research. Virtually all publicly funded basic research and over 80 percent of all U.S. publicly and privately funded basic research is performed by not-for-profit institutions, predominantly univer- sities and colleges (National Science Board, 1993~. Other R&D outputs, such as patents, copyrights, and trade secrets, are pro- prietary in nature. Most proprietary R&D (predominantly applied research and development) tends to be both funded and performed by private companies. U.S. industry supports more than half and performs more than two-thirds of all applied research conducted in the United States. Moreover, industry funds 60 percent and performs nearly 90 percent of all development work in the United States (National Science Board, 1993~. Publicly and privately funded R&D yield both proprietary and nonpropri- etary outputs. Research universities and federal laboratories, although focused primarily on nonproprietary research, also generate intellectual property that they sell or license to private firms. Collaborative R&D involving publicly funded institutions and private companies also can yield proprietary outputs. Moreover, when federal agencies fund R&D performed by private companies in service of national missions, such as defense, these public investments often enable private companies to develop proprietary technology that confers competitive advantages in commercial and noncommercial markets. At the same time, the private R&D investments of companies yield new knowledge, know-how, skills, and generic technology that are often broadly diffused without direct compensation to the it&D-performing company.3 THE POLITICAL-ECONOMIC LOGIC OF PUBLICLY AND PRIVATELY FUNDED R&D As the preceding discussion suggests, the differences between publicly and privately funded R&D and their respective outputs are clearer in theory than in practice. It is nevertheless useful to distinguish between the two types of R&D, since, within the United States at least, each is shaped by a distinct political- economic logic. Privately funded R&D is directed at the generation, assimila- tion, and application of knowledge and technology to advance the economic in- terests of stakeholders in the company making the investment. In market economies, it is generally accepted that, under most circumstances, private com-
32 FOREIGN PARTICIPATION IN U.S. RESEARCH AND DEVELOPMENT panics should be allowed to dispose of the outputs of their R&D investments as they see fit. It is assumed that such firms will pursue their individual economic self-interest with vigilance and, in so doing, collectively advance the interests of the nation as a whole.4 In contrast, publicly funded R&D and its proprietary and nonproprietary out- puts are viewed as public property to be used to advance specific U.S. national interests. For the most part, institutions that conduct or use the outputs of pub- licly funded R&D are subjected to greater public scrutiny than those that conduct or use the outputs of privately funded R&D. This is particularly true with regard to how and where the R&D outputs are used, and how the associated public and private benefits are distributed nationally and internationally. Research universi- ties, federal laboratories, and other not-for-profit institutions that perform the majority of publicly funded basic research do not generally commercialize tech- nology or produce products for commercial markets. This has led some observ- ers to question whether these institutions are adequately equipped to assess, let alone capture the true market value of the intellectual property they generate and manage with public money (U.S. Congress, House, 1989, 19931. R&D, INNOVATION, AND 1!iATIONAL ECONOMIC DEVELOPMENT R&D activities are a critical, yet relatively small, subset of the many comple- mentary activities and capabilities that contribute to technological innovation, which has been defined as "the processes by which firms master and get into practice product for process! designs that are new to them, whether or not they are new to the universe, or even to the nation" (Nelson and Rosenberg, 1993~. These processes integrate multiple functions, including organized R&D, design, production engineering, manufacturing, marketing, and other value-adding ac- tivities in a complex web containing multiple feedback loops (Kline, 1990; Kline and Rosenberg, 1986~.5 Thus, organized R&D activity is not the only source of innovative technology. Much technology, particularly process technology, is generated by other value-adding activities, such as production engineering or manufacturing. Neither R&D capabilities nor the possession of technology is by itself a reli- able indicator of the economic or competitive strength of a company or a nation. Rather, economic and competitive strength are determined by how effectively technology is used and managed in combination with other factors of production, such as labor, capital, and managerial and organizational capabilities. National economic development occurs when these multiple resources are committed and used in a way that causes the value of the economy's output to rise faster than the cost of inputs. This results in profits, however measured, which can be reinvested to the benefit of the nation's citizens. R&D contributes to a nation's economic development in many ways. The multiple outputs of basic research, applied research, and development yield new
THE CONTRIBUTION OF R&D TO U.S. ECONOMIC DEVELOPMENT 33 products, processes, and industries, as well as improvements in existing products and processes, all of which may contribute to economic growth, rising standards of living, and higher quality of life. These outputs can also provide the infrastruc- ture knowledge base, human capital, instrumentation for maintaining and up- grading a nation's capabilities to generate, assimilate, and apply knowledge and skills. Numerous industry analyses indicate that R&D expenditures have yielded high marginal as well as high median rates of return.6 Many private-sector in- vestments in R&D yield significant "spillovers" benefits to society beyond those captured by the individual R&D performer or investor. Therefore, it is generally assumed that the social rate of return (or the return to society as a whole) from private-sector R&D expenditures is substantially higher than the private rate of return to the firms carrying out the R&D. Indeed, estimates of the median social rate of return from private-sector investments in innovations originating from a broad spectrum of industries range from 40 to 99 percent roughly two to four times the estimated median private rate of return on these investments.7 Ultimately, economic returns from R&D investments depend on the comple- mentary assets and competencies of the particular firm or nation. In the case of the individual company, these assets include not only such things as design, pro- duction, and marketing, but also the broader technological and economic infra- structure that supports the firm within a given nation. Similarly, the economic or societal returns a nation may gain from a particular R&D investment depends on whether its innovation system can foster the widespread diffusion and effective use of the outputs generated. These factors are, in turn, influenced by the quality (skill level) of a nation's work force; by the size, wealth, and technological so- phistication of its domestic market; and, increasingly, by the ability of firms within its borders to access markets and technology abroad.8 Different types of R&D make different contributions to economic develop- ment. For example, basic research contributes to technological advance and eco- nomic growth both directly (through the generation and transfer of commercial- izable knowledge or technology) and indirectly (by providing generic knowledge and access to skills, methods, and instruments). Only occasionally do the outputs of basic research have intrinsic economic valued Rather, they feed into other investment processes that yield additional research findings and, at times, inno- vation. Hence, basic and applied research are linked by a complex, recurring cycle of interactions that increase the productivity of both (David et al., 1992; Pavitt, 1991~. For example, as graduate students perform academic research, they develop research skills. Subsequently, many shift from basic to applied work, to which they bring not only knowledge, but also skills, methods, and a web of profes- sional contactsall developed during their basic research training. This carryover from basic research is important, since instrumentation used in that setting is frequently applied in engineering and more applied disciplines, such as clinical
34 FOREIGN PARTICIPATION IN U.S. RESEARCH ED DEVELOPMENT medicine and industrial processes and operations (Brooks, 1994; Nelson and Levin, 1986; Pavitt, 1991; Rosenberg and Nelson, 19941. Applied research and development generate specific proprietary product or process technologies and innovations. This represents their most obvious and substantial contribution to economic development. At the same time, these ac- tivities also yield new knowledge, know-how, skills, and generic technology, some of which are widely, if not freely, diffused throughout a given industrial sector, technological field, or national innovation system. However, most ap- plied R&D takes place in private companies or the nation's defense laboratories, rather than in the more open environment of research universities. It is possible to describe the contributions of different types of R&D activity to a nation's economy as well as to arrive at very rough estimates of the rate of return to society of aggregate R&D investments or specific innovations in par- ticular sectors. It is virtually impossible to anticipate or trace after the fact the aggregate economic impact of a particular R&D investment. The economic con- tribution of a particular R&D activity is conditioned by various market, scientific, and technological forces, and certain types of R&D may have much higher value to society than others at a given time. Yet, any attempt to trace the precise roots of a particular economic benefit or stream of benefits from the customer's needs back through marketing, production, and finally to the germinal R&D is bound to underestimate the importance of seemingly ancillary research and develop- ment and more downstream innovation activities. Add to this the high degree of uncertainty and serendipity involved in R&D and technological innovation gen- erally, and it is virtually impossible to predict which avenues of R&D will yield the greatest returns to society over the long term. HOW THE BENEFITS OF R&D ARE DISTRIBUTED AT THE NATIONAL AND INTERNATIONAL LEVEL Every R&D output has multiple beneficiaries. While this observation is gen- erally accepted with regard to the outputs of nonproprietary or public-goods R&D, its validity for the intellectual property generated by private companies is insuffi- ciently appreciated. In fact, the proprietary outputs of R&D, regardless of where it is performed or how it is funded, yield benefits far beyond those that accrue solely to the individual or institution that owns or controls them. For example, the benefits of proprietary product or process innovations that improve the qual- ity and performance of goods and services, or reduce their costs, are widely dis- tributed within national or global economies. Thus, the benefits associated with a firm's proprietary R&D outputs are shared by the company's customers, suppli- ers, competitors, and the general public (Graham, 19921. Many factors influence how benefits are distributed among various economic actors within a national economy or among national economies. These include the location of R&D activity, the level of competition, and the relative capacity
THE CONTRIBUTION OF R&D TO U.S. ECONOMIC DEVELOPMENT 35 that a firm, region, or nation has to absorb and make use of R&D outputs. These capacities, in turn, depend on the level, intensity, and sophistication of existing R&D activity and on complementary capabilities and assets.~° To understand how each of these factors affects the distribution of benefits, it is useful to consider some of what is currently known about the complex pro- cesses of technology and knowledge transfer. In spite of the many advances in communications and information technology, there continue to be significant barriers to the movement of scientific and technological knowledge across na- tional boundaries and among organizations. To be sure, certain types of R&D output, including highly codified knowledge, can be readily transferred long-dis- tance within a firm or between different R&D players. However, most observers agree that the majority of R&D outputs are transferred most efficiently through face-to-face interactions among those who perform R&D and those who apply its results. Indeed, the transfer of knowledge usually involves human interaction in the form of personal contacts, movements among institutions, and participation in national and international networks (David et al., 1992; Gomory, 1989; Pavitt, 19911. For these reasons, even with the internationalization of industry, R&D ac- tivities still tend to occur in proximity to each other, which allows researchers to draw more efficiently on the work of their counterparts in other institutions. Simi- larly, the economic benefits of R&D activity tend to be much more localized than is commonly assumed. The importance of proximity for capturing R&D outputs has been underscored by recent research on patent licensing and other forms of technology transfer involving research universities and private U.S. companies (Jaffe et al., 1993~. Finally, there is broad consensus among those who study and conduct tech- nology transfer that, in many high-tech sectors, an organization's capacity for absorbing new knowledge and technology depends to a large degree on the level and quality of R&D occurring in that organization. In other words, in order to understand, interpret, and evaluate readily accessible new knowledge generated elsewhere, the recipient organization generally needs to be performing R&D at a level commensurate with that of the organization whose R&D activities it hopes to exploitll (Brooks, 1994~. Or, as Pavitt (1991) notes, "the most effective way to remain plugged in to the scientific network is to be a participant in the research process." IMPLICATIONS FOR THE ASSESSMENT OF FOREIGN PARTICIPATION IN U.S. R&D In earlier decades, the United States occupied a position of global techno- logical and industrial superiority. Many of the issues that today inform debate on foreign access to U.S. R&D and technology were not central. There was little question that Americans would reap most of the economic and technological ben-
36 FOREIGN PARTICIPATION IN U.S. RESEARCH AND DEVELOPMENT efits generated by investment in R&D and other types of technologically innova- tive activity. Through publications, scholarly exchanges, the activities of U.S.- owned multinational companies, and bilateral agreements the United States ex- ported more new knowledge and technology than any other nation. Nevertheless, the American public generally viewed this predominately one-way transfer of technology and know-how as consistent with both the short- and long-term eco- nomic, political, and national security interests of the United States. Moreover, at the time, the research activities of U.S. universities and federal laboratories were not seen as contributing much to the technology strategies and competitive suc- cess of American industry. Hence, there was little concern about the relatively limited efforts of foreign entities to gain increased access to these publicly funded R&D activities. While the United States remains a leader in the generation of new knowledge and technology, its position today is better characterized as first among equals; the gap that once separated the United States from potential competitors has closed. Recent decades have brought increasing convergence in the technologi- cal capabilities of industrialized nations as well as growing cross-penetration of national innovation systems through foreign direct investment and transnational industrial alliances. Other shifts have accompanied this convergence. Changes have taken place in the organization and management of R&D, and new links have been forged between different performers of R&D. These changes have included increased emphasis on R&D as a tool for scanning for and exploiting knowledge generated or applied beyond both institutional and national bound- aries, as well as closer integration of R&D with activities farther downstream in the value-added process of firms (Kash and Rycroft, 1992; Kodarna, 1991; Na- tional Academy of Engineering, 1993; Roberts, 1995a). With these changes have come new questions about the consequences of the continuing net outflow of U.S. technology and know-how and about the growing involvement of foreign nationals in publicly and privately funded U.S.-based R&D. There are both structural and policy reasons that the U.S. innovation sys- tem is more accessible than that of most of its foreign counterparts. Therefore, one concern is that foreign nationals may be taking out more knowledge, know- how, and technology and associated economic benefit activity than they return. Growing foreign involvement in U.S. R&D also heightens concerns about na- tional security. For example, military security might be compromised by the unauthorized transfer of certain highly sensitive knowledge or technology. Secu- rity may also be threatened if the U.S. government is denied timely access to advanced technology that is controlled by foreign-owned firms. CONCLUSION The preceding discussion of R&D activity, technology transfer, and the ways R&D contributes to economic development does not by itself provide clear an-
THE CONTRIBUTION OF R&D TO U.S. ECONOMIC DEVELOPMENT 37 swers to the many questions that have been raised about the consequences of growing foreign participation in the U.S. research system. It does suggest, how- ever, that the task of appropriating the many valuable outputs of U.S.-based R&D activity is significantly more complex and difficult for foreign nationals than is generally assumed. Foreign-owned firms that wish to effectively exploit U.S. technology and R&D outputs must establish a significant, technologically sophisticated presence in the United States to do so. Moreover, as the technological sophistication of foreign companies and their home countries increases, so too does the potential for reciprocal transfers of technology and knowledge into the United States. Fi- nally, this discussion suggests that under most circumstances, any county, in- cluding the United States, should welcome R&D activity within its borders, re- gardless of the nationality of the R&D performer. If a large share of the returns to R&D investments is captured by those proximate to the R&D activity, and these returns are beneficial, clearly it is better to have R&D performed within one's borders than beyond them. The two chapters that follow examine in some detail the causes, scope, and character of foreign involvement in U.S.-based R&D in an attempt to address questions about the costs and benefits to the United States of such participation. Proceeding from a belief that foreign involvement in publicly funded U.S. R&D is governed by a different political-economic logic than is foreign participation in privately funded U.S. R&D, the committee evaluates these two intertwined halves of the nation's R&D enterprise in separate chapters. The distinction between these two types of R&Dis in many instances artificial at least some of the R&D activities of both private companies and not-for-profit institutions are sustained by both public and private monies. Still, the committee believes that separating the two helps clarify and delineate the public-policy issues that are involved. NOTES 1. Summarizing Rosenberg (1990), Brooks (1994) notes that, "[l]aboratory techniques or ana- lytical methods used in basic research, particularly in physics, often find their way either directly, or indirectly via other disciplines, into industrial processes and process controls largely unrelated either to their original use or to the concepts and results of the research for which they were originally devised." 2. Expanding on Nelson's (1992) working definition, Brooks (1994) defines technology "both as 'specific designs and practices' and as 'generic understanding that provides knowledge of how and why things work' . . . and what are the most promising approaches to further advances." 3. For further discussion of the complex relationships between publicly and privately funded R&D and their proprietary and nonproprietary outputs, see Committee on Science, Engineering, and Public Policy (1992); Kash (1989); Mansfield (1986); and Nelson (1989). 4. Indeed, the very concept of intellectual property rights is premised on the assumptions that technological innovation yields significant benefits to society and that without the promise of tempo- rary monopoly rights, individuals and institutions would have insufficient incentive to invest or en- gage in R&D activity and technological innovation more broadly.
38 FOREIGN PARTICIPATION IN U.S. RESEARCH AND DEVELOPMENT 5. Characterizing a representative allocation of effort in the introduction of a new product, a seminal study of the management of technological innovation sponsored by the U.S. Department of Commerce in 1967 (known as the Charpie report) estimated that product conception and the associ- ated generation of primary knowledge (research, advanced development, basic invention) accounted for roughly 5 to 10 percent of the total effort. The remaining effort was devoted to "downstream" activities, including product design and engineering (10 to 20 percent); production layout, tooling, and process design (40 to 60 percent); manufacturing start-up and debugging (5 to 15 percent); and marketing start-up (10 to 20 percent). (See U.S. Department of Commerce, 1967, chart 7, p. 9.) Commenting on the Commerce study, Brooks (1994) notes that since many of the projects launched never get beyond the it&D/product-conception phase, and even a smaller share of total project launches make it all the way to marketing start-up, the 5 to 10 percent estimate probably understates the amount of activity devoted to R&D. Indeed, most companies engaged in R&D also conduct background research unrelated to any particular product. 6. The marginal rate of return is the rate of return from an additional dollar spent on R&D. Evaluating the rate of return on a number of industrial innovations and then calculating the median provides the median rate of return. 7. Mansfield (1986) compares the results of several independent studies, including his own re- search, of the median and marginal rates of return on private investment in particular innovations. He notes that the marginal social rate of return for private-sector investments in R&D is estimated to be in excess of 30 percent. 8. For further discussion of the concept of a national innovation system and the many factors that influence its performance, see Lundvall (1992) and Nelson (1993). 9. The extent to which knowledge is transferred directly from basic to applied science varies according to the economic sector and scientific field. In the development of chemicals and drugs, knowledge gained through basic research frequently results directly in industrially useful technology, including intellectual property. In transport and mechanical technologies, however, the link between basic science and technology is relatively weak (Pavitt, 1991). 10. Many factors play a role in the allocation of benefits among nations, including the quality of interaction and exchange between the various public- and private-sector it&D-performing institu- tions, the quality of the education system, the size and wealth of the domestic market, and other structural and regulatory factors. 11. Clearly, a company does not need to be doing advanced R&D (or any R&D for that matter) to exploit technology developed by somebody else very effectively and profitably. However, effective assimilation and use of a technology that is already developed, or "stabilized" (in effect, "codified"), requires a much lower level of technical sophistication on the part of the acquirer than does the assimilation and use of advanced R&D outputs (i.e., new knowledge, know-how, or technology in the making). In the recent past, the Japanese have proved themselves world leaders in reverse engineer- ing products and commercializing technology developed abroad. The new dimension of the Japanese challenge is their growing ability to access and use the fruits of U.S. basic and long-term applied research to both develop and commercialize new technology more competitively than U.S.-based companies. 12. For further discussion of the relative openness of the U.S. innovation system, see Chapter 3, pp. 70-74, and Chapter 4, pp. 90-91, 124-126.