Chapter 3
Adapting to New Challenges

Between World War II and the collapse of the Soviet Union, the U.S. economy operated in an environment characterized by military and economic competition. Federal research and development (R&D) investments, especially through the Department of Defense (DoD), dominated research spending through much of this period (representing about two-thirds of the national total at their high point) and many high-tech advances were defense spinoffs. Product life cycles were measured in years, and allowed ample returns on industrial research investments. Financial markets were relatively patient, and considered a wide range of factors in determining a company's health.

Much has changed, and the capitalizing process must continue to adapt. Commercial interests have replaced military procurement as the driving force of technology; industry now funds two-thirds of the national R&D effort. National competition is giving way to growth in international business and multinational mergers; global outsourcing and supply networks often blur patterns of ownership. Unprecedented mobility of capital and technology bring advanced R&D capabilities to more nations. Investors demand quarterly profit growth, and the marketplace demands shorter product cycles. Both industry and government seek to cut spending and balance budgets.

Chapter 2 identified the factors underlying U.S. strength in capitalizing on investments in science and technology—a diverse portfolio of cutting-edge research across all fields, a favorable environment for capitalization, superior human resources, and effective cooperation across sectors. The U.S. science and engineering enterprise faces new challenges that will require adaptation in all of these areas.

The Environment for Investing in Science and Technology

One result of these changes, especially in industry and federal agencies, is a growing incentive to curtail research whose payoffs are potentially high but whose results cannot be appropriated exclusively by the sponsor. There is support in Congress for basic research, but also a desire for research whose results can be predicted and measured. Agencies are required to emphasize performance measures, accountability, and short-term, practical results. Congress, through the Government Performance and Results Act of 1993, requires federal R&D agencies to submit strategic plans, performance plans, and annual reports demonstrating how



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 34
--> Chapter 3 Adapting to New Challenges Between World War II and the collapse of the Soviet Union, the U.S. economy operated in an environment characterized by military and economic competition. Federal research and development (R&D) investments, especially through the Department of Defense (DoD), dominated research spending through much of this period (representing about two-thirds of the national total at their high point) and many high-tech advances were defense spinoffs. Product life cycles were measured in years, and allowed ample returns on industrial research investments. Financial markets were relatively patient, and considered a wide range of factors in determining a company's health. Much has changed, and the capitalizing process must continue to adapt. Commercial interests have replaced military procurement as the driving force of technology; industry now funds two-thirds of the national R&D effort. National competition is giving way to growth in international business and multinational mergers; global outsourcing and supply networks often blur patterns of ownership. Unprecedented mobility of capital and technology bring advanced R&D capabilities to more nations. Investors demand quarterly profit growth, and the marketplace demands shorter product cycles. Both industry and government seek to cut spending and balance budgets. Chapter 2 identified the factors underlying U.S. strength in capitalizing on investments in science and technology—a diverse portfolio of cutting-edge research across all fields, a favorable environment for capitalization, superior human resources, and effective cooperation across sectors. The U.S. science and engineering enterprise faces new challenges that will require adaptation in all of these areas. The Environment for Investing in Science and Technology One result of these changes, especially in industry and federal agencies, is a growing incentive to curtail research whose payoffs are potentially high but whose results cannot be appropriated exclusively by the sponsor. There is support in Congress for basic research, but also a desire for research whose results can be predicted and measured. Agencies are required to emphasize performance measures, accountability, and short-term, practical results. Congress, through the Government Performance and Results Act of 1993, requires federal R&D agencies to submit strategic plans, performance plans, and annual reports demonstrating how

OCR for page 34
--> their research contributed to those plans. Some in the research community are concerned that this requirement will favor short-term, low-risk research projects whose results are easily measured over long-term, high-risk research whose evaluation is more problematic [see COSEPUP (1999b)]. Within this environment of growing pressure for measurable results and accountability, individual fields face particular challenges. For example, in information technology a recent report by a panel of experts asserts that government agencies are not providing sufficient support to high-risk, long-term research that will lead to future innovations (President's Information Technology Advisory Committee, 1998). Federal funding has been flat and increasingly focuses on short-term problems. These nearer-term investments, such as the Next Generation Internet (NGI) initiative, are aimed at ensuring the robustness of the Internet and are certainly essential.1 The information infrastructure is important to the nation in the same way that its bridges and highways are. Yet, it is just as essential that longer-term, higher-risk research that will contribute to future radical advances also is funded.2 One of the strengths of university research and the traditional central labs of industry has been the vibrancy of doing fundamental and applied research in close proximity. Losses from either category dilute the intellectual climate. During panel discussions and workshops, participants told the working group that universities and industry are doing too little basic research in fields such as networking, catalysis, and semiconductors. The changes in industry are also dramatic. Large companies that traditionally supported rich programs of long-range research have been pressured to cut back to maintain competitiveness (Nelson et al., 1996). Many have reduced their research organizations, have learned to acquire ideas and technology from outside the firm, and have adjusted their sights toward nearer-term goals. Large manufacturers are giving suppliers greater responsibility for engineering and design work, and some medium-size firms that specialize in particular technologies are emerging as the key sources of innovation. The large pharmaceutical companies rely on more agile biotechnology companies for new ideas, and few of the fastest-growing computer hardware or software companies founded in recent years support centralized research facilities.3 Although it would be a mistake to overestimate the scale of long-term basic research that industry performed in the past, this work has produced many 1.   See www.ngi.gov for an overview of NGI. 2.   In an area such as networking, where existing strong products and standards such as TCP/IP exist and are difficult to displace, private firms may be reluctant to fund research on radical new approaches. This makes the federal role in supporting this work even more important. 3.   Dorothy Leonard-Barton and John L. Doyle (1996, p. 181), describe Chaparral Steel, an innovative U.S. minimill, as integrating research with development. They quote the CEO, a former R&D director with a Ph.D. in metallurgy from Massachusetts Institute of Technology, as saying that research laboratories are idea graveyards "not because there are not good ideas there, but because the good ideas are dying there all the time."

OCR for page 34
--> important innovations over the years. In general, the current trend is toward shorter time horizons and greater focus on specific products in industrial R&D. Because they can no longer call on internal labs to answer research questions, many industries have formalized their dependencies on outside sources. Some examples include the following: In the aircraft industry, competitive pressures and defense-related cutbacks in federal R&D funding are forcing a shift in focus from high-risk technologies to demands from airline customers for lower cost of ownership. In recent years, industry has reaped the benefits of past R&D investments in computational fluid dynamics, materials, and computer-integrated design and manufacturing. Long-term research in this industry traditionally has been funded by government, and with tight funding, companies are focusing R&D spending on short-term research and product development. The leading U.S. manufacturer, Boeing, has drawn on its component suppliers for R&D. Suppliers, in turn, are outsourcing more R&D to their subcontractors. In the auto industry, rising development costs are raising overall R&D expenses, but investment in long-term research has fallen. Virtually all R&D is tied to specific product goals, including incremental improvements. To supplement internal R&D, the Big Three automakers rely more on suppliers, cooperating with each other through the United States Council for Automotive Research consortium, and increasing interactions with the government, especially DoE, through the Partnership for a New Generation of Vehicles. They look to universities for short-term engineering needs, long-term research in the physical sciences, and well-trained students. In the chemical industry, the research structure is shifting to accommodate short-term business unit needs rather than longer-term, corporate objectives. Industry is forming more partnerships with both universities and federal laboratories and using these partnerships to leverage their spending on precompetitive research (Council on Competitiveness, 1995). Across the spectrum of industries, major corporations have reduced, sold, or closed their research facilities. During the early and mid 1990s, IBM cut and refocused its research spending (Ziegler, 1997). RCA's Sarnoff Research Center, the source of pioneering research in video, liquid crystals, lasers, and other fields, was spun off to SRI International following GE's acquisition of RCA, and converted to contract research. It is important to recognize that pressures on government for greater accountability and on firms for greater focus and customer orientation are producing many positive results, and are occurring in countries around the world. Decentralized technology strategies are not new developments in U.S. industry. Intel, for example, decided early not to maintain a central research laboratory, choosing instead a strategy that it calls "minimum information" (Moore, 1996). It has thrived by

OCR for page 34
--> making incremental changes to existing technologies, with extremely short times between development and manufacturing, and low research expenses. There are also counterexamples: Hewlett-Packard tripled its R&D budget in the 1990s, and a few software companies, notably Microsoft, are establishing corporate labs. A countervailing tendency is the growth of research funding by private foundations. Although foundations long have played a key role in supporting the U.S. science and engineering enterprise, since World War II their contributions have been dwarfed by the growth of federal government and industry investments. In recent years, vast wealth has been accumulated by individuals and foundations because of the rise in the stock market. The emergence of angel investing, discussed in Chapter 2, is one example of how this wealth is channeled into capitalization. In the future, foundations and wealthy individuals can be expected to play a growing role in directly funding research and supporting universities and research institutions. One example is the Howard Hughes Medical Institute, which provided over $400 million in 1997 for medical research, grants, and special programs (Howard Hughes Medical Institute, 1997). Still, ongoing changes will complicate the task of ensuring that the public and private sectors provide sufficient funding for a diverse national portfolio of science and engineering research. A primary concern is whether investments in long-term research, especially in fields of obvious national importance, will be adequate. Education for the Long term Universities face the challenge of preparing students to be "employable for a lifetime" and of preparing them to enter the current job market upon graduation. These goals are both complementary and competing. It is inevitable that the supply of graduates will not perfectly match employer demand, especially in emerging fields where new or multidisciplinary skills are prized (see Box 3-1). Currently there is heavy demand for skilled employees in the field of information technology. CEOs of leading companies say that worker shortages are preventing the development and marketing of new products, lowering sales, and costing the country hundreds of thousands of jobs (Lerman, 1998).4 Hiring pressure is so strong that the industry is hiring predegree students and junior faculty out of the universities at attractive salaries.5 At the same time, a decrease in university research funding for these fields has reduced the number of graduate students being trained. The federal government and the private sector can both play a major role in helping institutions to meet new demands for trained scientists and engineers. Personnel exchanges can bring nonacademic people to the campus and allow stu- 4.   This is a very complex issue, and the information technology labor market is highly diverse in terms of the training and experience levels needed for different sorts of jobs. 5.   Note that data on median salaries of scientists and engineers by degree level show that engineers and computer scientists receive a relatively low premium for earning a Ph.D. vs. a Master's degree (10 and 14 percent, respectively), whereas life scientists (32 percent) and physical scientists (25 percent) earn a higher premium (National Science Board, 1998).

OCR for page 34
--> BOX 3-1 Challenges for advanced research and education: Networking and bioinformatics Networking During a recent workshop, a panel of experts in networking research described a trend: Little long-term research is being done in computer networking, either in universities, industry, or government labs.a The next wave of technical and systemic challenges to networking may find the field unprepared to deal with them. The era of networking began in the late 1960s, when it became desirable to share computing facilities among users. The problem of how to connect different computers, and eventually different networks, was solved over a period of several decades by a small, informal community of researchers who shared resources and an ethic of openness and cooperation. Many of the groundbreaking advances were funded by the Advanced Research Projects Agency/DoD and NSF, through the universities and a few private firms. The networking industry has grown rapidly during this decade. Job demand has become so strong that faculty with applications-oriented networking expertise are being drawn out of the universities into stimulating and well-paying positions in the private sector, and many students are joining them in industry rather than receiving advanced training. The departure of networking faculty from universities means that students have difficulty finding mentors and acquiring experience in problem selection. Many faculty who do remain are those who study theoretical rather than technological aspects of networking, because theory often is more highly rewarded by the academic world. Compared to theory, practical networking is interdisciplinary, collaborative, and "messy"—that is, the skills required to solve problems of connectivity and communication may extend far beyond engineering to include marketing, politics, certain aspects of mathematics (queuing, logic, probability), and verbal and collaborative skills.b As one panelist put it, "There is too much to fit into one brain." Like the universities, private firms are doing little fundamental systems research. Networking firms are growing so rapidly that they have little interest in or time for long-term research; innovation is driven primarily by product development problems. Companies that need new techniques often acquire a.   National Academy of Engineering (NAE)/COSEPUP Workshop on the Role of Human Capital in Capitalizing on Research, Irvine, Calif., January 20-21, 1998. Panelists included Robert Sproull, Sun Microsystems, Inc. (moderator); David Farber, University of Pennsylvania; Deborah Estrin, University of Southern California; and Brian Reid, Digital Equipment Corp. b.   Although information systems design has always required attention to these issues, they are increasingly relevant to lower-order design tasks.

OCR for page 34
--> smaller firms that already have developed them. Panelists lamented the lack of a strong industry presence not only in performing fundamental networking research, but also in setting the research agenda. Nor is the government any longer taking the lead in supporting research and guiding the agenda, as it once did. The engineering aspects of networking face large challenges in coming years: how to make the transition from a dedicated to a shared infrastructure, how to better meld the networking industry with the telephone industry, how to design optical network systems, how to link embedded processors, and, in general, how to cope with the explosive growth of the industry. Some of these problems require not only engineering experience, but also varying levels of expertise in marketing, consensus building, political science, and urban planning; the installation and linkage of network systems depends on leaders who possess a range of technical, political, and "people" skills. One model that may prove useful for companies and universities involved in networking is the Semiconductor Research Corporation (SRC), an industry partnership that funds university research and student training, described in Chapter 2.c Student internships in industry and mentoring of students by industry researchers are important components of SRC programs. When the SRC was formed in 1982, a shortage of trained researchers threatened the long-term health of the semiconductor industry. Although the situation is not directly analogous to the situation in networking today, SRC illustrates how competitors can come together to create assets important to all. Bioinformatics This new field, which spans mathematics, computer sciences, chemical engineering, the life sciences, and health care, is fueled by federal support for mapping the human genome and the need for mathematical modeling to produce new drugs in the biotechnology and pharmaceutical industries.d. Currently, there is a shortage of qualified people to work in this field. A recent analysis suggests that too few trained students are graduating to meet c.   See SRC's web page at www.src.org. d.   This box is based on a discussion at a workshop organized by NAE and COSEPUP on the Role of Human Capital in Capitalizing on Research, January 12-13, 1998, and a background paper prepared for the workshop (Stephan and Black, 1998). Panelists included Stephen Clark of Amgen, Bernard Palsson of the University of California at San Diego, Paula Stephan of Georgia State University, and Carlos Zemudio of Axiom Biotechnologies.

OCR for page 34
--> the needs of industry, and that junior faculty are leaving universities for challenging, well-paying positions with biotech and pharmaceutical companies. At the same time, many qualified Ph.D.s in the life sciences are taking longer to finish their degrees and are undertaking multiple postdoctoral positions because of a shortage of permanent tenure-track positions (NRC, 1998). What accounts for this inconsistency? Several factors influence the availability of human resources in these fields. One is the research funding system, in which scholars seek to establish themselves as principal investigators (PIs) overseeing their own laboratories. Attracting superior graduate research assistants increases the productivity and quality of research, allowing the PI to secure research funding. In this environment, work in emerging interdisciplinary fields such as bioinformatics may be difficult to initiate. Establishing new educational and research programs in bioinformatics would require collaboration among computer science departments, biology departments, and medical schools. In a funding environment emphasizing research grants to PIs, incentives for such collaboration may be weak.e Further, in the research culture of the life sciences, an M.S. is not seen as an acceptable terminal degree. Research and education in the life sciences therefore may be less responsive to trends in the nonacademic job market than other fields, such as engineering. Finally, it may be difficult to "retool" life scientists to work in bioinformatics by having them take a few computer courses. Some experts argue that students who choose to study biology tend to have a lower level of interest or talent in mathematics. Yet others argue that such retooling can be done. Some workshop participants pointed out that the lack of human resources in this area is understandable, since bioinformatics has emerged only recently. To several panelists, developing new educational programs that impart skills in computing and life sciences, from the undergraduate level onward, is the key long-term challenge for universities and other stakeholders. e.   Several experts who have attempted to secure funding for new centers or programs in bioinformatics reported experiencing difficulties in the peer review process, which they believe partly reflected an inability for some reviewers to consider the context beyond their own disciplines.

OCR for page 34
--> dents to meet nonacademic scientists and engineers, helping both students and faculty understand nonacademic working environments and job opportunities. A basic strength of the university is its disciplinary structure, which allows students to immerse themselves deeply in a well-defined subject. At the same time, the traditional boundaries of a discipline may present a challenge to students who want to investigate emerging fields. To arrange a program in bioinformatics, for example, a student may have to take computer science courses in the school of engineering, mathematics courses in the school of arts and sciences, and biology courses in the life sciences department and the medical school. The structure of financial support for graduate students can affect their ability to investigate emerging fields. A student who is supported by a research assistantship makes a commitment to contribute to a specific program and may lack the ability to pursue broader study. A student supported by a fellowship or direct grant may have more flexibility to study subjects in multiple fields and do research in a less traditional area. Of course, U.S. educational institutions will need to broaden their approaches to ensure that the United States has the human resources needed to capitalize on science and technology advances in the future. This study focuses on research universities because of their central role in advanced research and education. Overall, U.S. research universities are remarkably successful institutions. Their challenge is to maintain traditional strengths as they respond more flexibly to emerging education and training needs. Prior to World War II, the university focused on the codification of applied science and engineering expertise and the development of new fields of inquiry and training in response to industry requirements (Rosenberg and Nelson, 1994). As the university research enterprise grew over the postwar decades, and institutions came to rely on federal funding to maintain their excellence, filling the ranks of the professoriate became a key task for advanced science and engineering education. COSEPUP's (1995) report on graduate education in science and engineering called on educators to put greater emphasis on training students for nonacademic careers and suggested that greater diversification in federal funding mechanisms could contribute. Since the release of that report, federal agencies have developed new initiatives that move in this direction.6 Continuity is a strength of the U.S. research university, and it probably would be impossible to eliminate supply-and demand mismatches in science and engineering labor markets. Nevertheless, the current shortage of talent in bioinformatics and the career difficulties being experienced by young life scientists should indicate to universities and federal agencies that a coordinated response is required. 6.   One example is the National Science Foundation (NSF) Integrative Graduate Education and Research Training program. The National Institutes of Health supports a wide range of training grants.

OCR for page 34
--> Partnerships to Leverage Capabilities and Resources As science and engineering become more complex and multidisciplinary, more skills, more teamwork, and more people are required to perform and capitalize on research. As described in Chapter 2, the past two decades have seen an explosion of research collaboration and partnerships between industry, universities, and government. For example, in 1994, there were 1,000 university-industry research centers (UIRCs) on more than 200 university campuses (Cohen et al, 1994). And yet there are barriers to more efficient functioning of such partnerships (Government-University-Industry Research Roundtable, 1999). For example, graduate students and even faculty may know little about the changing role of senior professionals in industry. Individual scientists and engineers must be able to lay out program goals, identify several options, and plan ways to capitalize on their research. They need to know how to network, negotiate, and manage a partnership with other researchers. Differences in expectations and culture can challenge students seeking careers and faculty seeking partnerships in industry. In a discussion on catalysis, several U.S. academics stated that it was easier for them to work with European-based companies than with U.S.-based companies (see Appendix A). Likewise, researchers at several U.S. companies reported greater success in structuring collaboration with foreign or second-tier U.S. universities than with the U.S. universities leading in catalysis research. Universities and industry have different views on intellectual property rights (IPRs). Patents allow inventors a period to exploit their innovation in exchange for publication. Companies often seek exclusive rights in order to capitalize on their investment; some universities now seek control of rights as well. As in other areas of capitalization, the situation is complex and varies significantly by field. Universities differ in their attitudes toward faculty who wish to hold equity in start-up companies. An important, unresolved question is the extent to which current IPR restrictions may be inhibiting the development and application of new knowledge and, conversely, the extent to which the pursuit of profits may inhibit the progress of basic research. In short, universities are challenged to develop partnership modes that promote effective interaction with industry and complement their primary missions: education and the creation of new knowledge. In this task, the diversity of approaches among universities can be a strength of the U.S. system; not every institution needs to emulate Massachusetts Institute of Technology or Stanford. The benefits of breaking down institutional barriers can be seen in the high-performing industry networks: Silicon Valley; Research Triangle Park, North Carolina; the Route 128 complex outside Boston; the textile firms of northern Italy; and the industrial centers of Japan. Note that the most successful of these networks depend on the proximity of competing firms. The zero-sum depiction, in which institutions gain at the expense of others, does not seem to be accurate; in these cases, joint gains are realized and advantages shared, even under conditions of fierce competition (Fountain, 1998). The growth of partnerships between sectors over the past two decades repre-

OCR for page 34
--> sents a significant adaptation of the U.S. science and engineering enterprise aimed at improving capitalization. Much anecdotal evidence attests to the value of collaboration, but evaluating the effectiveness of individual programs and approaches is inherently difficult. Although SRC is considered to be very successful and has a significant track record, discussion at an NAE/COSEPUP workshop revealed that sustained efforts are required on the part of the member companies to extract maximum value. Cases of university-to-industry knowledge transfer that clearly contribute to specific products are limited (Randazzese, 1996). In the broader context, a focused effort may be required to codify lessons and highlight the best practices of collaboration (Mowery, 1998). Utilizing such lessons, it may be possible to widen the scope of partnership activities. For example, the community of researchers who work on cognition and learning believe that they have generated a number of significant insights that could be applied to K-12 education, where the United States faces serious challenges (see Appendix A). Incorporating these insights into new teaching approaches, testing them, and encouraging their adoption in schools are activities that require focused and extensive effort. Several institutions are doing this work, but not on a large scale. In addition, new barriers to collaboration are emerging and, as discussed earlier, old ones may be reemerging in new forms as the perspectives of stakeholders change. Continued efforts to reduce these barriers could deliver significant benefits. Finally, as partnerships grow and change, it will be important to maintain realistic expectations about what they can and cannot do. Programs such as SRC, NSF's Science and Technology Centers and ERCs and others show that collaboration can encourage companies to fund areas of fundamental, long-term research that they would not support by themselves. However, workshop discussions during the study revealed that partnerships cannot be expected to replace the federal government as the primary funder of fundamental research in most fields. Changes in the Capitalization Environment The globalization of economic activity is straining old international relationships and demanding new trade and ownership policies. Concerns about national security must be balanced against the development of new kinds of alliances. For example, U.S. computer companies must seek exemptions from old trade laws to reimport components that they send to their Asian factories for assembly. U.S. auto companies must step nimbly around traditional import limits to sell what are mostly foreign-made cars as domestics (Brown, 1998). Many changes alter the ways firms must do business. The highest cost—and risk—for research-based firms is in development and commercialization. Historically, they could count on long-term research from their own central labs and, in defense, a ready first customer in the federal government. Today, companies must add new options. The corporations with their functional specialization have given way to smaller, leaner organizations in which team-based structures cross functional lines, transcend hierarchical chains of command, and focus on core functions while contracting with outside firms for other tasks. For example, DuPont has

OCR for page 34
--> doubled its spending on external R&D in the past three years, entering more than 30 cooperative research and development agreements and 6 Advanced Technology Program (ATP) grants; it expects more revenue growth overseas than at home (Guschi, 1996). To increase access to markets and expertise, U.S. firms increasingly set up facilities abroad. Similarly, foreign direct investment in R&D by foreign enterprises is the most rapidly growing segment of U.S. R&D. Japan-based firms alone have 98 R&D labs in North America.) To compete, firms need preferred partners, new ways to interact with universities, government, and other companies, focused communication with Washington and their state capitals, and good corporate knowledge of what they really have to offer. At home, the climate for capitalizing on research is richer for the availability of private financing. A role for the limited-partnership venture capital firm emerged in the late 1960s, enabled by a favorable economy, stock market, and tax policy. A key reform came in 1979, when the Department of Labor changed the "prudent man rule" to allow pension funds to invest in venture capital funds. This change, plus liberal tax changes at about the same time, gave rise to the modern era of venture financing. Venture activities in the biotech field boomed in the 1980s; software and communications technologies dominated in the 1990s. The total amount of money invested by venture capitalists is small compared to other sources of finance for technology development, but the venture capital industry plays a significant role in the creation of new firms. Individual changes may seem small, but the cumulative power of the capitalizing environment is great. This has been demonstrated dramatically since the post-World War II years, when it was predicted that open markets, growing demand, and free access to technical knowledge would close the gap between the strong U.S. economy and the economies of Japan, the United Kingdom, Germany, and France. At the time of a recent study (Patel and Pavitt, 1994), Japan and Germany had moved ahead, but the United Kingdom and France had fallen behind. Similarly, Taiwan, Korea, and Singapore had leapt ahead from very backward conditions, whereas Brazil, Mexico, and India had made less progress. Patel and Pavitt concluded that such differences in technology diffusion sprang from cultural, managerial, and institutional differences: the climate of capitalizing. Since the study was conducted, some of these conditions have changed appreciably, illustrating the dynamic quality of the capitalizing climate. Today, the capitalizing environment appears to be quite favorable. Another National Research Council study explores these complex trends in greater detail (STEP, 1999). Although the primary current challenge in this area is to "not mess up a good thing," the study clearly shows the importance of a favorable capitalizing environment, and the speed with which conditions can change. Heightened recognition of these points on the part of the science, engineering, and policy communities can help the nation to maintain and improve this environment in the future.