Future R&D Environments: A Report for the National Institute of Standards and Technology (2002)

Alden S. Bean

This paper examines trends in the organization and management of industrial R&D that may affect the rate and direction of technological innovation over the next 10 years. It builds on a report published by the Industrial Research Institute (IRI) that summarized its members’ expectations of how industrial R&D laboratories would operate in 2008.¹ After reviewing findings and projections of that

report, more recent trends and projections are discussed. An effort is made to identify “game changers”—changes in structures, policies, or practices that may alter innovation processes or outcomes in significant ways. After a discussion of expected changes in structure and practice across different business sectors, the paper concludes with questions and observations that may merit further discussion.

The Industrial Research Institute’s forecast of how an industrial R&D laboratory would operate in 2008 is the starting point for this discussion.² The forecast resulted from several self-study projects carried out by IRI members between 1996 and 1998, topped off by discussions and presentations at their annual meeting in 1998. For the purposes of this paper, it is important to note both the time frame of these studies as well as the makeup of the IRI membership. As for the time frame, the U.S. economy had been expanding rapidly since 1993; the stock market had been rising steadily since 1982, and from 1992 forward it had risen so rapidly that “. . . the price index looks like a rocket taking off through the top of the chart!”³ Industrial R&D funding, which had been flat from 1990 to 1994, began to grow at a double digit rate that carried through 2000.^4,5 As for the makeup of the IRI membership, it represents big company R&D, primarily manufacturing firms, with a heavy representation from the chemical sector. The IRI estimates that its 300 member companies perform as much as 85 percent of the nation’s R&D. While total industrial R&D in the United States grew almost constantly after the mid-1970s (except for a flat period from 1990 to 1993), manufacturing firms’ support for R&D did not grow, in constant dollar terms, from 1986 to 1994.⁶ The contribution to total R&D growth between 1986 and 1993 was due to the rise of R&D in the service/nonmanufacturing sector, which is underrepresented in the IRI membership and is made up of much smaller firms.⁷

Before discussing the 2008 forecast, it is noteworthy that Fusfeld projected that the principal emphasis for industrial R&D in the 1990s and beyond would be as follows:⁸

• Inside the corporation

  • Coordination of teams with decentralized structures,

  • Convergence of technical strategy with corporate strategy,

  • Establishment of global networks of technical units, and

  • Organization of technical intelligence.

• Outside the corporation

  • Alliances with other corporations,

  • Strategic relationships with universities,

  • Active participation with government programs, and

  • Selective use of consortia and cooperative R&D programs.

• Growth areas

  • Leveraging core R&D with global resources and

  • Expanding technical activity and linkages in developing countries.

Fusfeld cited competitive and cost pressures of the early 1990s as the driving forces for these changes and noted those pressures had “changed the practices of industrial research.”

In Appendix E of this report, McGeary provided a summary of the IRI report by Larson and its conclusions.⁹ This discussion expands on McGeary’s summary and selects points that deserve amplification or comment in light of more recent events and new information.

First, the report called attention to six forcing functions, identified by IRI members as drivers that would shape the conduct of industrial R&D and innovation in the decade from 1998 to 2008:

• Information technology,

• Globalization of technology,

• Growing workforce diversity,

• Integration of technology planning and business strategy,

• Partnerships and alliances, and

• Customer power.

These drivers were subsequently combined in various ways to generate five scenarios for change, which were then projected forward for 10 years and used to characterize what industrial R&D might look like by 2008. The scenarios were the following:

1. Cyclical change. This scenario assumes only incremental changes in the way R&D/innovation will be done and that the changes will be cyclical, repeating patterns of change more or less as they occurred in the past. Examples are offered of such patterns, which may take decades to unfold, at GE, DuPont, and Xerox.¹⁰ Hounshell and Smith authored a catalog of R&D management and organization practices at DuPont from 1902 to 1980.¹¹

2. Globalized R&D. This scenario flows from observations that industrial CEOs (surveyed by the Baldridge National Quality Award Foundation) viewed economic globalization to be the single most important trend affecting U.S. companies and concluded that the trend toward the globalization of R&D would continue or perhaps accelerate through 2008.

3. R&D through partnerships. This scenario presumes that technology will become increasingly complex and more expensive to develop, thus forcing companies to rely on partnerships, collaborations, and alliances with other companies, government laboratories, universities, and contract R&D organizations to build the critical mass of skills and know-how needed for innovation.

4. Innovation function absorbs R&D. This scenario assumes that the innovation (new products, processes, services) value proposition will overwhelm the perceived need for a separate R&D function, thus resulting in complete decentralization of R&D into customer-focused business units. This amounts to a complete and permanent victory for the decentralized model of R&D organization, the pros and cons of which are thoroughly described by Hounshell and Smith.¹²

5. Networking counts. This scenario acknowledges the well-documented fact that “invisible colleges” and networks are endemic to science and technology, and that increased recognition of ways to expand and extract value from networks will change traditional R&D. (If Metcalfe’s law and network externalities facilitate the trend toward expanding the scope of “networking” to new content areas, then perhaps security concerns and firewalls will constrain it.)

Larson concludes by stating his belief that the most likely scenario will be some hybrid of the five.¹³ He goes on to describe the four major differences between the laboratories of 1998 and 2008:¹⁴ people, technology intelligence, technical work, and the innovation process.

By identifying information technology and globalization as the most important of the six drivers of change in R&D organization and management, Larson’s Industrial R&D in 2008 invites closer examination of these forces and how they will play out as the scenarios unfold. Will they be “game changers” or evolutionary enhancements? Here are some further thoughts and observations.

Information technology (IT) is a broad category that includes computers, data networks, the Internet, and the World Wide Web but can also include telecommunications. Shiller considers the arrival of the Internet and the World Wide Web beginning in 1993 to have been one of the 12 factors upon which the (irrationally exuberant) bull market of the 1990s was based.¹⁵ On the other hand, Levine et al. propose that the Internet and the Web will change business as usual and work wondrous transformations on business structures and processes.¹⁶ A recent paper by Litan and Rivlin, which analyzes the likely impact of the Internet on the U.S. economy over the next several years, concludes that “there are many reasons to side with the optimists, especially those who claim that widening use of the Internet creates a significant potential for increasing productivity and increasing the standard of living over time.”¹⁷ The paper cites four major findings regarding pathways for these effects:

• Reducing transaction costs in financial services, health care, government, education, and manufacturing;

• Increasing the efficiency of management in automobile manufacturing, trucking operations, and supply chain management;

• Making markets more compatible by lowering information costs and frictions; and

• Increasing consumer choice and convenience.

The Litan and Rivlin paper concludes that the potential of the Internet to contribute to productivity growth is real (although noting that a 2 percent annual growth rate is more likely than the 3 percent experienced between 1995 and 2000); that the gains due to e-commerce are likely to be less important than the reductions in transaction costs in financial and health services, government, and old economy firms in general; that enhanced management efficiencies will increase productivity in product development, supply chain management, and many other aspects of business; and that we should not overestimate the productivity-enhancing effects of the expected improvements in market competitiveness, because much of the cost improvement will be passed along to consumers as price reductions.¹⁸

How important might IT be to the future of R&D? The members of the industrial R&D community also seem to have come down on the side of the optimists. Setting aside the slowdown in demand for IT equipment, the build-out of the IT infrastructure, and the dot-com e-commerce disappointments, there is evidence that the Internet and the World Wide Web are enabling important changes in the way R&D and innovation are managed. If we remember that the Web was invented by the international community of physicists to improve their ability to communicate and collaborate in the design of the particle physics research program carried out at CERN in Switzerland, this should not be a total surprise.

In 1999, Paul Horn, senior vice president for research at IBM, wrote “Information Technology Will Change Everything,” in which he projected key improvements in IT through 2025.¹⁹ He captured the primary impacts of these technical gains in three concepts: (1) “the wired planet” depicts the globalization of the Internet—“everyone and everything” becomes connected to the Internet; (2) “knowledge becomes the new currency”—with an estimated 300 million home pages on the Web, and with new pages being added at the rate of 800,000 a day, how do you find the pages that contain what you need to know; and (3) “augmented reality”—a world in which computers will be able to reason and interact with human beings very much like we interact with one another. Horn concluded his paper by challenging managers to think about three implications of IT growth for their companies:

• What will your brand identity mean in a world where everyone can get what they need on the Internet?

• Are your systems ready for your company to be an Internet company for knowledge management?

• Are your people ready to function in a networked world?

Perhaps because of Horn’s challenging questions, industrial R&D managers have continued to confer about and publish on these issues. Browsing past issues of the Industrial Research Institute’s journal, Research •Technology•Management (RTM), turns up articles on knowledge management, organizational structure, globalization, and intellectual property that provide insights about the ways IT is changing R&D practice. Some of the ways mentioned include the following:

• Leveraging knowledge management toward the goals of total quality management (TQM) and total customer satisfaction;

• Leveraging cross-company resources and permiting rapid integration of new business acquisitions;

• Flattening organizations and defeating command and control management styles; and

• Enabling virtual research teams to transcend geographic, linguistic, and temporal barriers.

The Commerce Department reported that R&D by foreign firms in the United States grew throughout the 1990s as much as R&D by U.S. firms located abroad.²⁰ These trends are expected to continue. The motivation for overseas R&D continues to be the same as it was in the early 1990s: assisting parent companies to meet customer needs, keeping abreast of technological developments, employing foreign scientists and engineers, and cooperating with foreign research laboratories.²¹

The IRI has begun to feature news of international R&D activities in RTM. The January-February 2001 issue detailed current activities in China, Sweden, and Russia, while the March-April 2001 issue covered the Asia-Pacific region. These articles highlighted domestic R&D activities of those countries and also covered the activities of U.S. firms that have laboratories or cooperative programs in those countries. The impression gleaned from these articles is that barriers to international knowledge flow are coming down, thanks to IT; that intellectual property policy regimes are moving in the direction of being inter-

nationalized; that strategic alliances are increasingly popular; and that venture capital support for entrepreneurial activity is increasing throughout the world.

Quantitative data from the National Science Board (NSB) show the globalization of R&D as well as the underlying composition of R&D expenditures.²² Industrial support for R&D has been rising relative to government support throughout the G7 countries, averaging 62 percent of total funding across OECD countries in 1997. R&D funded by foreign sources in these countries doubled between 1981 and 1997, but it still accounted for only 2.5 percent of all R&D funds. While there is substantial variation in industry/government funding ratios across the G7 countries (from a high of 74 percent by industry in Japan to a low of 30 percent in Russia), the performer base is decidedly industrial, ranging from about 70 percent in the United States to about 54 percent in Italy. Academia is the second most important performer of research, with basic research the dominant activity. International cooperative R&D and strategic technical alliances have doubled since 1980, with IT and biotechnology alliances more important than all others combined.²³

In the early 1990s, concerns over R&D workforce diversity referred to management’s need to deal effectively with gender and ethnic differences among scientists and engineers entering the workforce. The issues had to do largely with their enculturation and assimilation into the role structure of laboratories and related functions. Today and for the foreseeable future, industry will be concerned with these same issues, but in an expanding context. It is concerned not just with R&D, but with the full set of roles and behaviors required to make industrial innovation happen rapidly and successfully. The knowledge management emphasis (discussed above), together with the strategic management perspective (discussed below), is driving industry to place a high value on recruiting and retaining people with strong credentials in scientific and engineering fields where explicit knowledge represents a core technical competency; on retaining people whose tacit knowledge is critical to successful innovation based on core technical competitiveness; and on outsourcing its noncore needs. Essentially, a value chain/supply chain management mentality is being applied to the R&D/ innovation management process.

An article by Perry illustrates the retention problem, which peaked during the dot-com craze of the late 1990s but which “won’t be over for a long time because of a shortage of talent,” according to an interviewee at Corning.²⁴ Perry sums up the problem thusly:

As dot-com mania fades, its effect on job mobility lingers on. Star technical performers, once considered organizational loyalists, are more likely to jump ship for competing job offers. Employers are responding to the threat by introducing new policies and procedures designed to retain key personnel. The most visible and expected change is in compensation: Salary, equity participation and bonuses are on the upswing, as is the correlation of pay with performance and the frequency for which compensation is reviewed. But good pay is not enough to keep the best people in a new environment characterized by talent raids by competing organizations. Star performers are demanding, and getting, more opportunities for career advancement, greater participation in leading-edge projects, and entrepreneurial freedom.²⁵

A series of articles has appeared in RTM dealing with these new diversity issues. Demers writes as follows:

• Technology who are implementing e-commerce strategies,

• Information technology who are developing and/or implementing major new systems,

• Junior positions where people are slated for major leadership roles,

• Sales to customers considered vital to the preservation of revenues,

• Positions with unique knowledge of products and services, and

• Positions needed to support a major consolidation or acquisition.²⁶

Other papers deal with defining the specialized roles required to make innovation happen—for example, Markham and Smith on the role of champions in the new product development process²⁷ and Roberts on the critical executive roles required for strategic leadership in technology.²⁸ As innovation continued to absorb more and more of the attention and resources of industrial firms, as compared with traditional R&D laboratory work, large industrial firms have had to deal with “entrepreneurship” in more creative ways, as Perry points out.²⁹ In many cases this has resulted in changes in organizational structure, including the

creation of internal venture capital groups in several firms and the augmentation of the laboratory system to include the equivalent of incubators. (See, for example, the discussion of Xerox’s newly formed Xerox Technology Enterprise division in Lautfy and Belkhir.³⁰) These structural changes are interesting in their own right, but their staffing patterns are also noteworthy because of the academic credentials of the principals. For example, both of the senior managers of Xerox XTE division hold Ph.D.’s in physics and appear to have been very productive scientists. Both have been drawn out of traditional R&D roles into these new innovation management responsibilities.

The National Science Board acknowledges the increasing complexity of the labor market for professional scientists and engineers and notes that changes in the structure of career opportunities in science and engineering (S&E) need to be carefully matched to career expectations and training. Otherwise, the health of scientific research in the United States could be damaged, which could “reduce the ability of industry, academia and the government to perform R&D, transfer knowledge, or perform many of the other functions of scientists in the modern economy.”³¹ Since industry employs such a high percentage of the nation’s S&E professionals (77 percent of B.S. degree holders, 60 percent of M.S. degree holders, and 38 percent of all Ph.D.’s), shifts in the structure of industrial R&D/innovation work could have important implications for S&E education programs.

In retrospect, the recession of 1990 and 1991 seems to have catalyzed serious corporate efforts to rationalize the relationship among business strategies, organizational structures, and operating systems. There was great concern over the Japanese “miracle” in manufacturing, and a flurry of business books appeared that urged corporate reforms. The relationship between business strategy and R&D operations was not a new topic by any means. (See Hounshell and Smith for an account of DuPont’s struggle with the problem between 1902 and 1980.³²) A book by three Arthur D. Little, Inc., consultants drew attention to the technology planning/business strategy issue in a readable and persuasive way.³³ It undoubt-

edly served as a focal point for many corporate discussions and decisions that have reshaped business/technology strategy linkages.

Since the information requirements for developing good business strategies and technology roadmaps are very demanding, and much of the desired information is proprietary and/or unpublished, industry has spent a great deal of time and money on competitive intelligence and data-mining in recent years. Improved databases and software for extracting valued information from data (as discussed above) are making the construction of business/technology portfolios and roadmaps more feasible year by year. As these costs come down, strategy formulation will perhaps become less challenging than strategy implementation and execution.

Surveys in 1992 and repeated in 1999 provide insights into trends in global strategic management of technology practices. Roberts reports on 209 responses received from 400 very large firms surveyed in the United States, Japan, and Europe, and presents a preliminary interpretation of the results.³⁴ Here are some of the observations:

• The integration of corporate technology strategy and overall business strategy depends heavily on the involvement of senior management in formulating and implementing strategy. Trends are as follows:

  • CEOs, R&D VPs, and CTOs are the primary integrators, with marketing VPs and CFOs lagging behind.

  • Japanese and North American firms see themselves as having strongly linked business/technology strategies, while Europe lags. The trend was upward for Japan and North America between 1992 and 1999.

  • 96 percent of Japanese firms placed R&D executives on their boards of directors, compared with 35 percent in Europe and 8 percent in North America.

  • 91 percent of Japanese, 67 percent of European, and 60 percent of North American companies have CTOs or R&D VPs on their executive committees.

• Those firms with the strongest linkages between corporate business/technology strategies are the strongest business performers on the following measures:

  • Overall corporate sales growth rates,

  • Share of 1998 sales from new products and services,

  • Share of sales from new and improved products,

  • Technology leadership, and

  • Perceived R&D timeliness in meeting new product delivery schedules.

• Between 45 and 49 percent of the firms surveyed in the three regions had CEOs with strong technical backgrounds.

• The greater the percentage of the total R&D budget spent on short-term R&D, the better the R&D performance on the following measures:

  • Perceived efficiency,

  • Timeliness,

  • Improved time to market,

  • Meeting market dates for commercialization, and

  • Process implementation.

• Except for the pharmaceuticals industry, spending on research (basic and applied) declined relative to spending on development and support services for both corporate and divisional R&D between 1991 and 1998.

• Globalization of the R&D performer base increased for firms in all three regions.

• External sourcing of technology increased substantially between 1992 and 1998 and was expected to increase even more by 2001:

  • The percentage of firms relying on external sources of technology rose from about 20 percent to about 70 percent by 1998 and is expected to rise above 80 percent by 2001.

  • Central corporate research is the most important source of research, despite the trend toward external technology acquisition.

• Regarding overall technical effectiveness:

  • The greater the technical maturity of a business unit’s key technologies, the lower the performance of the business.

  • The greater the company’s technical leadership position, the stronger its performance.

  • Speed to market is associated with many possible influences, including more short-term R&D.

Roberts also identified improved technology management practices in the following areas as important:

• Use of automation and other nonhuman resources in R&D,

• Use of IT and communications technology, and

• Streamlined R&D organizations.

In his conclusions, Roberts emphasized that

• Strong business/technology strategy linkages are correlated with superior business performance.

• “The most important continuing business change in strategic technology management is the increasing worldwide reliance on external-to-each-company sources of technology.”

• The emphasis on speed to market has “reinforced the disproportionate emphasis on near-term R&D spending at the expense of longer-term strategies.”

Before leaving this topic, it is interesting to note the elements of a presentation by Edelheit at which he spoke on new pathways of innovation at General Electric.³⁵ After noting that GE had been transformed from a company that derived 85 percent of its revenue from manufactured products in 1980 to one that derives 65 percent of its revenues from services today, he noted the following changes in the role of corporate R&D that are going on today:

• Shifts in funding from corporate to business units;

• Project management a major focus;

• Short-term technical support part of mission;

• Synergy across businesses;

• Ensure leadership technology in the businesses;

• R&D time frame shortened;

• “Game changers” still a priority;

• Multigeneration product development;

• Partnerships and joint ventures a way of life;

• Supporting corporate initiatives—for example, six sigma and e-engineering; and

• Human resources strategy.

The overlap between these changes at GE and those reported by Robert is striking and perhaps a sign that these issues will drive corporate technology strategies for some time to come.

Industrial R&D partnerships and alliances appear to be motivated by a variety of corporate concerns and circumstances. The NSB broadly characterizes R&D partnerships as collaborations that allow “individual partners to reduce costs and risks and enabling research ventures that might not have been undertaken otherwise.”³⁶ Moreover, “the underlying theme is that more can be accomplished at lower costs when resources are pooled, especially if organizations can compliment each other in terms of expertise and/or research facilities.” The report goes on to detail special conditions surrounding alliances involving different types of partners—industry to industry, industry-academic, industry-federal laboratories,

academic-federal government—and to briefly describe the impact of international competition on government policy making decisions regarding the formation of R&D consortia.

After reading the results of Roberts’ survey and discussions of the use of R&D alliances in industry by Wyndrum,³⁷ Campbell,³⁸ Brenner and Tao,³⁹ and Miller and Morris,⁴⁰ it is difficult to say whether the increase in external sourcing of R&D should be expected to continue for the foreseeable future or whether it is a more transient phenomenon. To be specific, despite the federal laws enacted since 1980 to promote industry-government partnerships and to promote technology transfer, the number of both corporate research and development agreements (CRADAs), which involve industry/government laboratory collaborations, and industry/industry research joint ventures (RJVs) registered with the Department of Justice seems to have peaked in 1996 and is now in decline.⁴¹ Additionally, federal funding for NIST’s Advanced Technology Program (ATP) has been a political football, having suffered a major cutback in 1996, and is widely expected to be terminated by 2002.

Moving to research collaborations within industry, the pharmaceutical industry is noted for its use of combinations of in-house R&D, research contracts to small external research labs, and merger and acquisition activity to sustain innovation across successive waves of technological change.⁴² Campbell writes:

In addition to small research firms, universities are important R&D collaborators for industrial firms. As noted earlier, universities have overtaken the federal

government as the second most important performer of R&D. A recent paper by Hicks et al. notes that patenting activity in IT and health technologies is growing rapidly in the United States and that university patenting is particularly active.⁴⁴ It further notes that company patents frequently cite publications from “in-state public sector institutions,” perhaps indicating a diffusion of expertise beyond the elite universities.

In the chemical industry, Brenner and Tao report that Air Products and Chemicals has been involved in 42 external R&D partnerships since the mid-1990s. Air Products has saved 2 years of internal effort and hundreds of thousands of dollars in net research expenditures per project and has the potential to generate tens of millions of dollars per year in increased revenues and millions in profits. The company’s partners included industrial consortia, universities, government labs, and unique (often small) research and development companies.⁴⁵

With regard to the telecommunications sector, the complex interplay between the various service and carrier companies across wireless, cable, land lines, and Internet technologies has stimulated major restructuring efforts at large, established telecoms and Internet service providers as well as at smaller newcomers. Wyndrum conducted a comprehensive analysis of technology investment activities at 28 of these firms between 1998 and 2000.⁴⁶ In particular, he studied the motives behind their outright acquisitions and minority equity investment (MEI) activities. Three motives were grouped together and called R&D: acquiring access to (1) technology for new products, (2) technology for new features or functionality of existing products, (3) and skills. The other motive studied was market expansion. Wyndrum found transactions worth more than $300 billion, of which $15 billion was in the form of MEIs and the balance was for direct acquisitions. Upon further analysis, he found that firms that lacked internal R&D spent $274 billion to acquire firms for market expansion, $12.9 billion to acquire R&D expertise, and $3.3 billion to take MEI positions, of which $0.9 billion went for market expansion and $2.3 billion went to acquire R&D capability. On the other hand, he found that firms with strong internal R&D of their own spent $10.2 billion for acquisitions, of which $4.5 billion was for market expansion and $5.7 billion for R&D capability; of the $11.9 billion in MEIs, $5.3 billion was for market expansion and $6.3 billion for R&D expertise.

Finally, Wyndrum notes that the net result of these MEI/acquisition activities was to move firms with little or no internal R&D to a position comparable to

established firms with strong internal R&D (between 8 and 10 percent of revenues) and that companies that already had strong internal R&D maintained or strengthened their R&D/sales ratios as a result of the MEI/acquisition investments. Thus, to the extent that external R&D collaborations evolve into joint business ventures and eventually into merger and acquisition activities, industry restructuring is facilitated.

An article by Germeraad calls attention to other practices in external R&D/ technology development that may be important emerging trends.⁴⁷ His basic point is that firms have done such a good job of integrating their technology strategies and business strategies that they are finding new ways of appropriating returns from their R&D investments. Germeraad describes these methods as follows:

• New avenues for building corporate value. These trends encompass new avenues to building corporate value. These include the changing basis of market capitalization, building companies to flip, off-balance-sheet spin-offs, and licensing-out as a business practice. These market forces are redefining what value means to a company. The basis of market capitalization is changing from a focus on tangible to intangible assets. Companies are no longer being built solely to withstand the tests of time. Instead, some new companies are being built to flip. Spin-offs are going off the balance sheet and licensing-out is emerging as its own business unit. All of these trends are affecting how a CEO explains value to the company’s shareholders and how R&D expenses are explained to the analysts.

• New development processes. Second, new development processes afford tremendous opportunity for substantial returns from the new business models. These trends are “spiral” product development, synchronous portfolio selection and management, iterative integration of product development and portfolio management, and connection-focused new product developments. Product development and portfolio selection are moving from conventional funnel models to spiral development. Acquisitions and partnerships are replacing some of R&D’s traditional role in bringing new products to the company. These trends are both reducing the time to bring new products to market and adding greatly to the complexity of managing R&D.

• Trends that are warping time. The third group of trends relates to “warping time.” These are enabling trends that are improving the productivity of R&D, yet these new capabilities are so far neither robust nor embedded in R&D organizations. They include accessible venture funding, the advent of the image age, and the amalgamation of R&D and general management methods. Readily available venture capital funding is accelerating the development of new business opportunities. The ability to view massive amounts of structured and unstructured

information as interactive pictures is enabling better business decisions at high speed. Collectively, all of the “programs of the year” have, in fact, improved the productivity of R&D organizations and increased the speed and quality of product development. Yet, if not managed carefully, these trends introduce higher risks associated with near-synchronous and/or overlapping project and portfolio activities.

To summarize, economic globalization as well as scientific and technological breakthroughs disrupt traditional sources and means of innovation. To the extent that these forces continue in place over the next 10 years, R&D partnerships and collaborations can be expected to continue, as part of the dynamic that restructures and expands industry.

Chatterji and Davidson credit the total quality management (TQM) movement not only with making U.S. R&D more customer-focused but also with bringing R&D into closer contact with the rest of the business enterprise and making it more business-oriented than in the past.⁴⁸ The TQM movement began in Japan in the 1970s and 1980s and picked up momentum in the United States in the late 1980s and early 1990s, when it began to impact the R&D function. The primary tenets of TQM are as follows:

1. Meet (and if possible, exceed) the customer’s needs and expectations in all areas of contact.

2. Focus on the process that produces a product or an output, not just on the product itself, and in so doing, drive out inefficiencies and waste in all business processes.

3. Incorporate improvements in the organization at all levels, and continuously strive for additional improvement opportunities.

Chatterji and Davidson say that TQM, when applied to R&D, introduced many reforms into R&D management in the mid-1990s, including breaking down the not-invented-here barrier, resulting in increased outsourcing of R&D and acceptance by R&D professionals of performance metrics.

Similarly, the movement to reduce R&D/innovation cycle time is considered to be customer driven.⁴⁹ Customization of the SONY Walkman based on fast

response to customer feedback was so successful that it stimulated changes in U.S. innovation management practices, including concurrent engineering, cross-functional teams, and the widespread acceptance of the product platform concept.⁵⁰

Perhaps the strongest influence of the “closer to the customer” movement in U.S. industrial R&D has been the apparent decline in the importance of the corporate/central R&D lab in the United States. This trend, noted by Roberts,⁵¹ Hicks,⁵² and others in addition to Chatterji and Davidson,⁵³ has resulted in shorter time frames for industrial R&D, as noted earlier. Moreover, as shown by Hicks et al., during the decline of the corporate lab, the rate of industrial patenting slowed relative to university patenting, and in all technologies except health, the linkage to science (as measured by citations to scientific literature) is much stronger in patents issued to universities than in those issued to industry. (In health, the science connection is stronger in both university-owned and government-owned patents, but the rates of change are closer than in other technologies.⁵⁴)

Thus, while the nation is now more dependent on industrial R&D for its technological strength than it has been since before World War II, industrial invention appears to be less science-based than it was 20 years ago. While the pickup in university patenting and its strong linkage to science may suggest that universities may be offsetting the trend in industry, Hicks et al. caution that there is evidence that the surge in university patenting may reflect the pursuit of quantity at the expense of quality.⁵⁵

It is clear from the literature reviewed above that industrial R&D managers believe that important changes are taking place in R&D management practices and structures, and that they expect gains in R&D/innovation efficiency and effectiveness to be realized as a result. Analysts seem to agree that interesting changes are taking place but are at odds about their expected impacts. The crux of the disagreement is over productivity growth effects. Gordon, in a recent paper published by the National Bureau of Economic Research,⁵⁶ is skeptical of future

productivity gains, because the innovations we are counting on for those gains, mainly computer, telecommunications, and Internet technologies, are simply not sufficiently robust to sustain productivity growth rates comparable to those realized from the technologies that we produced in the late 1800s and early 1900s (electricity, internal combustion engines, the chemical and pharmaceutical industries, the entertainment, information, and communication industries, and urban sanitation infrastructures). Gordon regards modern innovations to be primarily substitutes for older technologies, with little upside potential. Similarly, Shiller feels that the Internet pales in comparison with the interstate highway system in its productivity-enhancing potential. He regards the Internet and the World Wide Web as part of the new economy hype that created the stock market bubble of the 1990s and little else.⁵⁷ Schwartz published an article that presented three future scenarios for “recovery”: one is V-shaped, a second is U-shaped, and the third is L- shaped, which is not a recovery at all! As discussed earlier, Litan and Rivlin are more optimistic about the productivity-enhancing potential of the Internet, but cautiously so.⁵⁸ Thus, to the extent that changes in R&D management practices and structures are Internet/Web driven, analysts would probably caution against being too optimistic about these long-term social and economic impacts.

The trends of note are the following:

• A decided shift away from corporate funding toward business unit funding for R&D is under way. While many large corporations (GE is one) have decided to use a hybrid structure for their laboratory system (a combination of corporate and divisional business unit labs), corporate funding of corporate labs has dried up. As a result, industrial R&D is more short-term-oriented and under great pressure to meet the time to market and other immediate needs of the business units that now provide most of their funds.

• The growth of service sector R&D from 5 percent of total industrial R&D in 1983 to 26 percent by 1993 has stabilized at 22 percent. The NSB notes the difficulty of disaggregating these numbers into specific sectors, making it difficult to tell what kind of R&D is being emphasized.⁵⁹ Moreover, the articles by Edelheit⁶⁰ and Campbell,⁶¹ which discuss respectively the transformation of GE into a service company and the expenditure of 20 percent of some pharmaceutical firms’ budgets on external R&D, suggest the need for better data so trends can be better discerned and understood.

• The rise of university patenting propensity relative to that of industry (in fields other than health) reflects the increasing desire of universities to participate in the intellectual property marketplace. The declining number of references to scientific publications in university patents suggests a preference for quantity over quality.

• A trend toward involving professional scientists and engineers in all facets of the technological innovation process, not just R&D, raises questions about the training of young scientists and engineers for entry-level positions in industry.

• The external sourcing of R&D noted by Roberts, as well as the emerging trends in creating/capturing value from R&D results noted by Germeraad, deserves tracking.⁶²,⁶³ The interaction of these activities with existing tax policies and financial accounting standards seem worth watching, as firms continue to find ways to include intellectual capital on the balance sheet.

• Globalization will interact with the Internet and the World Wide Web to create hybrid corporate structures and relationships that we are just beginning to understand. R&D networking will increase as Metcalfe’s law operates within specific scientific and technical communities. Mobility of S&E employees will make knowledge management increasingly difficult, and great pressure will emerge to standardize and enforce intellectual property rights internationally. Universities and industry will struggle more and more with intellectual property rights conflicts as the stakes get higher.