Definitions of R&D

R&D. According to international guidelines for conducting research and development surveys, R&D, also called research and experimental development, comprises creative work "undertaken on a systematic basis to increase the stock of knowledge—including knowledge of man, culture, and society—and the use of this stock of knowledge to devise new applications" (OECD 2002, p. 30).

Basic research. The objective of basic research is to gain more comprehensive knowledge or understanding of the subject under study without specific applications in mind. Although basic research may not have specific applications as its goal, it can be directed in fields of present or potential interest. This is often the case with basic research performed by industry or mission-driven federal agencies.

Applied research. The objective of applied research is to gain knowledge or understanding to meet a specific, recognized need. In industry, applied research includes investigations to discover new scientific knowledge that has specific commercial objectives with respect to products, processes, or services.

Development. Development is the systematic use of the knowledge or understanding gained from research directed toward the production of useful materials, devices, systems, or methods, including the design and development of prototypes and processes.

R&D plant. R&D plant includes the acquisition of, construction of, major repairs to, or alterations in structures, works, equipment, facilities, or land for use in R&D activities. U.S. statistics include separate tabulations for R&D plant (NSF/SRS 2007b), which are not generally available in comparable international R&D statistics.

Budget authority. Budget authority is the authority provided by federal law to incur financial obligations that will result in outlays.

Obligations. Federal obligations represent the dollar amounts for orders placed, contracts and grants awarded, services received, and similar transactions during a given period, regardless of when funds were appropriated or payment was required.

Outlays. Federal outlays represent the dollar amounts for checks issued and cash payments made during a given period, regardless of when funds were appropriated or obligated.

For an annotated compilation of definitions of R&D by U.S. statistical agencies, tax statutes, accounting bodies, and other official sources, see NSF/SRS (2006b).

Unmeasured R&D

The estimates of U.S. R&D presented in this volume are derived from surveys of organizations that have historically performed the vast majority of R&D in the United States. However, to evaluate U.S. R&D performance over time and in comparison with other countries, it is necessary to gauge how much R&D is going unmeasured in the United States. The following are indicators of unmeasured R&D performance in the United States:

• To reduce cost and respondent burden, U.S. industrial R&D estimates are derived from a survey of R&Dperforming companies with five or more employees. There are no estimates of R&D performance for companies with fewer than five employees.
• The activity of individuals performing R&D on their own time (and not under the auspices of a corporation, university, or other organization) is similarly not included in official U.S. R&D statistics.
• Social science R&D is excluded from U.S. industrial R&D statistics, and R&D in the humanities is excluded from U.S. academic R&D statistics. Other countries include both in their national statistics, making their national R&D expenditures relatively larger when compared with those of the United States.
• R&D performed by state and local governments in the United States is not currently estimated for national statistics. A new survey of state R&D is currently being collected by NSF and the Census Bureau.

Although NSF estimates the R&D performance of nonprofit organizations, a nonprofit R&D survey has not been fielded since 1998.

Recent Developments in Innovation-Related Metrics

This sidebar reports on recent or ongoing initiatives aimed at advancing innovation-related measures. As noted earlier, a distinction is made between R&D and the subsequent implementation or commercialization of the resulting knowledge.

NSF Workshop: Advancing Measures of Innovation
NSF held a workshop focused on innovation metrics during the summer of 2006, "Advancing Measures of Innovation: Knowledge Flows, Business Metrics, and Measurement Strategies." The workshop was driven by several considerations, including the challenge by Dr. John H. Marburger III, the president's S&T adviser, for better data, models, and tools for understanding the U.S. S&E enterprise (Marburger 2005a, b). A number of strategies for data development were discussed at the workshop: survey-based methods, data linking and data integration, nonsurvey-based methods (such as mining of administrative data), and using case studies and qualitative data. The sense of the workshop was that these diverse strategies are not mutually exclusive and can be pursued productively in parallel or in combination. For workshop presentations and a summary report, see NSF/SRS (2006a).The OECD's Blue Sky Forum, which followed the NSF workshop, discussed the development of new and better indicators of science, technology, and innovation and developed a synthesis of findings toward an agenda for the next decade. For more information about the Blue Sky Forum, see OECD (2006a).

Federal Initiatives Supporting New Metrics
Science of Science and Innovation Policy (SciSIP) is an NSF research initiative started in the fall of 2006. The initiative is expected to develop the foundations of an evidence-based platform from which policymakers and researchers may assess the nation's S&E enterprise, improve their understanding of its dynamics, and predict its outcomes. The research, data collection, and community development components of SciSIP's activities will: (1) develop theories of creative processes and their transformation into social and economic outcomes; (2) improve and expand science metrics, datasets, and analytical tools; and (3) develop a community of experts on SciSIP. Additional information is available at NSF/SBE (2007).

In addition to the OSTP interagency taskforce described on page 4-11, the Department of Commerce (DOC) established the Measuring Innovation in the 21st Century Economy Advisory Committee to "study metrics on effectiveness of innovation in various businesses and sectors, and work to identify which data can be used to develop a broader measure of innovation's impact on the economy." The committee held its first public meeting in February 2007. See DOC (2007) for further details.

Lastly, the America COMPETES Act (Public Law 110–69) enacted in the summer of 2007 establishes, among other measures, a President's Council on Innovation and Competitiveness. In addition to policy monitoring and advice, the Council's duties include "developing a process for using metrics to assess the impact of existing and proposed policies and rules that affect innovation capabilities in the United States" as well as "developing metrics for measuring the progress of the Federal government with respect to improving conditions for innovation, including through talent development, investment, and infrastructure development. . . ." For the complete text of the America Competes Act, see Library of Congress (2007).

The BEA/NSF R&D Satellite Account: R&D and Economic Growth

Satellite accounts are supplementary estimates of the GDP and related measures that provide greater detail or alternative measurement concepts without changing the core accounts. In particular, the purpose of the R&D satellite account is to consider R&D as an economic investment or capital (i.e., capitalizing R&D). This is an ongoing project involving NSF's Division of Science Resources Statistics, the agency responsible for official U.S. statistics on R&D expenditures, and the Bureau of Economic Analysis (BEA), the agency responsible for the U.S. national economic accounts. This activity is one of several interagency efforts aimed at improved measures of intangibles and their economic role (Jorgensen, Landefeld, and Nordhaus [2006]; Okubo et al. [2006]). Current plans call for the incorporation of R&D capital into the National Income and Product Accounts' core accounts in 2013, based on the concepts developed in the satellite account.

Measuring R&D as capital investment recognizes its long-term benefits much as investments in physical assets such as highways and machinery. As a newly recognized component of investment, R&D has a direct impact on GDP because business expenditures for R&D become part of economic output, instead of being treated as an expense. According to these estimates, capitalizing R&D increases the level of GDP in current dollars by an average of 2.5% per year from 1959 to 2002 (Okubo et al. 2006). In terms of GDP growth, R&D capital would account for about 4.5% of real GDP growth during that same period. During the more recent period 1995–2002, R&D investment would account for about 6.5% of growth. By comparison, according to BEA, business investment in commercial and all other types of buildings accounted for slightly more than 2% of real GDP growth between 1959 and 2002.

Further research topics include the measurement of the overall impact, both direct and indirect, of R&D activity on productivity. The indirect effects of R&D activity on productivity include spillovers that accrue when the benefits to the economy as a whole are larger than the benefits to the private owners of R&D. Additional research topics include the incorporation of international R&D flows and several methodological improvements. For more information, see BEA (2007a).

Industry Classification

As a result of classification conventions, interpretation of industry-level R&D data is not always straightforward. Initially, each company sampled in NSF's Survey of Industrial Research and Development is assigned to a single industry according to payroll data for the company,* and each is requested to report its R&D expenditures for the entire company. These expenditures are assigned to the previously classified single industry. This classification scheme reasonably categorizes most companies into industries closely aligned with their primary business activities. However, for diversified companies that perform R&D in support of a variety of industries, any single assigned industry is only partly correct. And in some cases, the industry assigned based on payroll data is not directly related to a company's R&D activities.

It is important to assess the relationships between industries as well as the business structure within industries when analyzing R&D data. For example, most of the federally funded R&D reported in the navigational, measuring, electromedical, and control instruments industry is performed by large defense contractors that also produce aerospace products. And investigations of survey microdata revealed that most of the R&D classified into the trade industry represents the activities of manufacturing firms that have integrated their supply chains and brought their warehousing, sales, and marketing efforts in-house. Consequently, beginning with the 2004 cycle of the survey, the assigned industry classification of companies in selected industries (such as wholesale trade) and also companies that most influence the overall R&D performance estimates is subjected to manual review and potential reclassification. Wherever possible, this report includes industry-level data that results from this new method of industry classification.†

Trends in R&D for Industrial Research Institute Members

For more than 20 years, the Industrial Research Institute (IRI), a nonprofit association of more than 200 leading R&D-performing industrial companies, has surveyed its U.S.-based members on their intentions for the coming year with respect to R&D expenditures, focus of R&D, R&D personnel, and other items. Because IRI member companies carry out a large amount of industrial R&D in the United States, the results from these surveys help identify broad trends in corporate R&D strategies. Dr. Jules J. Duga, a senior analyst at the Battelle Memorial Institute in Columbus, Ohio, notes (in a personal communication) that the IRI survey

. . . provides a reasonable overview of the actions that are being taken by industry. Although the internal analysis of IRI survey results does not delve deeply into the driving forces for the stated planning, the overall results are certainly a reflection of industrial response to markets, federal actions, and approaches for the most effective means for acquiring technological assets. Although there have been changes in the type of membership pattern that is represented within IRI, and there are similar changes in the character of the respondents, the IRI survey provides a long-term envelope of planning and practices as applied to R&D, and results in there being the raw material for qualitative and semi-quantitative longitudinal studies that well serve the objectives of industrial science policy analyses. One of the major characteristics of the IRI survey is that for all intents and purposes the questionnaire has maintained the same format for many years, thus permitting the development of a long-term analytical framework with a minimum of disruptions. The analysis of the responses to individual questions, as well as the introduction of a so-called "sea change" indicator, provides a series of snapshots of postures. Over the past few years, efforts have been directed toward viewing clusters of responses to questions that have internal conceptual linkages. Such an approach has provided a means for developing broader pictures of the driving forces and action items that are influencing industrial R&D strategy.

The most recent survey, administered during the summer of 2006, suggests that many companies continue to shift the focus of their R&D spending away from directed basic research and the support of existing business to new business projects (IRI 2007). This reported shift in R&D priorities also is reflected in how responding companies intend to spend their R&D budgets. IRI survey respondents reported the following plans for 2007:

• Increase total company expenditures on R&D
• Increase hiring of new graduates
• Increase outsourcing of R&D to other companies
• Increase outsourcing for university R&D and federal laboratories
• Increase participation in alliances and joint R&D ventures
• Increase licensing of technology to and from other companies
• Increase acquisition of technological capabilities through mergers and acquisitions

Overall, these strategic moves are consistent with responses suggesting increased R&D budgets. Responding companies are increasing R&D spending to support existing lines of business as well as new business projects and are leveraging their R&D spending through joint R&D ventures and grants/contracts for university R&D. (For more information, see the section entitled "Technology Linkages.")

R&D Expenses of Public Corporations

Most firms that make significant investments in R&D track their R&D expenses separately in their accounting records and financial statements. The annual reports of public corporations often include data on these R&D expenses. In 2004, the 25 public corporations with the largest reported worldwide R&D expenses spent $127.3 billion on R&D. The three companies that topped the list were automobile manufacturers. Ford Motor Company, DaimlerChrysler, and Toyota, together with the other four automobile manufacturers on the list, reported spending $41.0 billion on R&D (32.5% of the total for the top 25) (table 4-6). There are 10 companies in the information and communications technologies (ICT) sector that spent a total of $49.4 billion (38.8% of the total). The remaining eight companies include six pharmaceutical manufacturers and two diversified consumer product-oriented manufacturers. As Hira and Goldstein (2005) point out, although four of the five top leaders in R&D in 2004 were automobile manufacturers, which is a marked difference compared with 2000, when the top four spenders were the telecommunications giants Ericsson, Lucent, Motorola, and Nortel, "automakers face an uncertain near-term outlook because of pressures from an increasing cost structure and the need to achieve shorter product life cycles to meet rapidly changing consumer preferences."

The top 25 companies are headquartered in seven different countries, with nine headquartered in the United States. However, the location of a company's headquarters is not necessarily the location of all its R&D activities. Most of the companies on this list have manufacturing and research facilities in multiple countries around the world. (For more information, see the section entitled "R&D by Multinational Corporations.")

Overall, R&D spending for the top 25 increased 4.0% in 2004 compared with 2003. Sales for the group as a whole increased 6.8%; sales increased in the 6%–8% range for the automobile and pharmaceutical manufacturers and ICT companies in the group, and more than 11% for the consumer product manufacturers. R&D expenditures increased for the manufacturers (pharmaceutical, 6.9%; automobile, 6.3%; and consumer products, 10.4%). However, the ICT companies, representing the sector with traditionally the highest R&D intensity, reported only a 0.1% increase.

It should be noted that a recent change in accounting standards by the Financial Accounting Standards Board (FASB) may result in discontinuities in companies' reported R&D expenses, making it more difficult to evaluate R&D spending trends from publicly available financial data. By 2004, most large companies began following the guidelines of FASB's Statement of Financial Accounting Standards, "Accounting for Stock-Based Compensation," which requires companies to expense the fair value of all stock-based compensation.* Many high-technology companies have historically compensated their R&D employees with stock options and stock awards. This stock-based compensation may not have been reported as company expenses before these new guidelines. For example, according to Hira and Goldstein (2005), "Microsoft's R&D spending decreased 20.5% in 2004 despite an increase in R&D employees. According to its U.S. Securities and Exchange Commission filings, the decrease was "‘due to lower stock-based compensation expense' [because] in 2003 the company began offering its employees stock-based compensation in lieu of options. This affected its R&D accounting significantly. . . ." For information on how many of the largest U.S.-based corporations intended to adjust their R&D strategies and spending, see sidebar, "Trends in R&D for Industrial Research Institute Members."

Tracking R&D: Gap Between Performer-and Source-Reported Expenditures

In some OECD countries, including the United States, total government R&D support figures reported by government agencies differ from those reported by performers of R&D work. Consistent with international guidance and standards, most countries' national R&D expenditure totals and time series are based primarily on data reported by performers (OECD 2002). Although funding and performing series may be expected to differ for many reasons, such as different bases used for reporting government obligations (fiscal year) and performance expenditures (calendar year), the gap between the two U.S. R&D series has widened during the past decade or more.

During the mid-1980s, performer-reported federal R&D in the United States exceeded federal reports of funding by $3–$4 billion annually (5%–10% of the government total). This pattern reversed itself toward the end of the decade; in 1989, the government-reported R&D total exceeded performer reports by $1 billion. For FY 2005, federal agencies reported obligating $109 billion in total R&D to all R&D performers ($44 billion to the business sector), compared with $94 billion in federal funding reported by the performers of R&D ($23 billion by businesses). Hence, overall industrywide estimates equal approximately a 50% paper "loss" of federally reported 2005 R&D support (figure 4-9). The difference in federal R&D totals was primarily in DOD development funding of industry.

Several investigations into the possible causes for the data gap produced insights into the issue, but a conclusive explanation has been elusive. According to a General Accounting Office (GAO 2001) investigation, "Because the gap is the result of comparing two dissimilar types of financial data [federal obligations and performer expenditures], it does not necessarily reflect poor quality data, nor does it reflect whether performers are receiving or spending all the federal R&D funds obligated to them. Thus, even if the data collection and reporting issues were addressed, a gap would still exist." Echoing this assessment, the National Research Council (2005a) notes that comparing federal outlays for R&D (as opposed to obligations) to performer expenditures results in a smaller discrepancy. In FY 2005, federal agencies reported total R&D outlays of $103 billion.

Federal R&D Infrastructure

The U.S. government invests substantial resources not only in R&D, but also in the facilities and instrumentation required by researchers to tackle problems at the frontier of S&T. In FY 2007, federal agencies are expected to obligate more than $3.5 billion for R&D plant, capital equipment, and facilities for use in R&D. Two agencies, NASA and DOE, account for more than two-thirds of all federal R&D plant obligations in FY 2007. Some examples of research infrastructure made possible through federal funding include:

• Supercomputing resources. As of November 2006, 6 of the top 10 supercomputers in the world were located in U.S. FFRDCs or government laboratories (TOP500 Supercomputer Sites 2007). The Terascale Simulation Facility at Lawrence Livermore National Laboratory houses two of the world's fastest supercomputers: BlueGene/L, ranked fastest in the world, and ASC Purple, ranked number four. These powerful computers support DOE's research on the safety and reliability of the nation's nuclear arsenal. The federal supercomputing resources are also used for nondefense purposes such as research on climate change and bioinformatics. For example, the DOE Joint Genome Institute leveraged the computing resources and research capabilities of multiple federal laboratories to contribute to the sequencing of the human genome. For more information, see DOE (2007).
• Hubble Space Telescope. Launched in 1990 and upgraded during four subsequent servicing missions, NASA's Hubble Space Telescope revolutionized astronomy by providing deep, clear views of the universe without the distorting effects of the Earth's atmosphere. Among its many highlights, Hubble was the first optical telescope to provide convincing proof of a black hole. More than 6,300 published scientific papers have been based on its data. At the time of its launch, the Hubble Space Telescope cost $1.5 billion. More details are available at NASA (2007).
• Antarctic research stations. NSF funds and manages the U.S. Antarctic Program, which coordinates almost all U.S. science on the continent, including research carried out by other federal agencies. The unique Antarctic environment has proven to be a boon to many fields of study. For example, astronomers and astrophysicists have benefited from the excellent optical properties of the atmosphere at the South Pole (resulting from its high elevation, low temperature, and low humidity) and from the extremely clear, thick, and homogeneous ice that makes neutrino detection possible. For additional information, see NSF/OPP (2007).
• Highly Infectious Diseases Laboratories. DOD and HHS (through both NIH and the Centers for Disease Control and Prevention) currently operate several laboratories that facilitate research on pathogens that require the highest levels of safety precaution, such as Ebola, viral hemorrhagic fevers, monkeypox, and avian influenza. DHS also plans to operate two such labs.

Many of the laboratories funded by the federal government provide scientists and engineers with tools and facilities that otherwise would not exist. For example, capabilities in DOE user facilities include particle and nuclear physics accelerators, synchrotron light sources, neutron scattering facilities, genome sequencing, supercomputers, and high-speed computer networks. By itself, DOE's Office of Science oversees facilities used by more than 20,000 non-DOE researchers each year in a range of scientific disciplines (figure 4-10). User facilities are one channel for collaborating and diffusing knowledge and technologies (see "Technology Transfer Metrics" later in this chapter).

Federal R&D Initiatives

The 2008 budget targets R&D priority areas often involving the expertise of multiple federal agencies (OMB 2007). To improve the efficiency and effectiveness of federal R&D investments in these areas, the administration continues to encourage strategic coordination among stakeholder agencies. Priorities detailed in the Administration's FY 2008 budget include:

• American Competitive Initiative (ACI). The ACI invests in basic research areas that advance knowledge and technologies used by scientists in nearly every field through DOC's National Institute of Standards and Technology, DOE's Office of Science, and NSF. For FY 2008, the second year of ACI, President Bush proposes $11.4 billion for these three agencies. For an overview of the initiative, see OSTP (2006).
• Climate Change. The Climate Change Science Program (CCSP) is focused on improving decisionmaking on climate change science issues. This program has an FY 2008 R&D budget of $1.5 billion, of which the National Aeronautics and Space Administration accounts for 56%. More information is available at CCSP (2007) and Climate Change Technology Program (2007).
• Combating Terrorism. This area supports the president's strategy for homeland security by harnessing federal R&D programs that could help to deter, prevent, or mitigate terrorist acts. The FY 2008 budget provides support for capabilities in several areas including detection and imaging, cargo screening, biometric systems, and critical medical countermeasures. For an overview of homeland-security related R&D, see Knezo (2006).
• Hydrogen Fuel. The Hydrogen Fuel Initiative seeks to support R&D aimed at developing and improving technologies for producing, distributing, and using hydrogen to power automobiles. DOE will continue to lead this initiative. The 2008 budget completes the president's 5-year, $1.2 billion commitment announced in his 2003 State of the Union address, but work will continue on the many technical challenges that remain. For more details, see Interagency Working Group on Hydrogen and Fuel Cells (2007).
• Nanotechnology. The National Nanotechnology Initiative (NNI) supports basic and applied research on materials, devices, and systems that exploit the fundamentally distinct properties of matter at the atomic and molecular levels. The FY 2008 budget provides $1.4 billion for NNI R&D, three-fourths of which is allocated to NSF, DOD, and DOE. For more information, see NNI (2007).
• Networking and Information Technology. The multiagency Networking and Information Technology Research and Development (NITRD) program aims to leverage agency research efforts in advanced networking and information technologies. The FY 2008 budget provides $3.1 billion for NITRD R&D, including about $1 billion each to DOD and NSF. Additional information is available at NITRD (2007).

Comparing International R&D Expenditures

Comparisons of international R&D statistics are hampered by the lack of R&D-specific exchange rates. If countries do not share a common currency, some conversion must be made to compare their R&D expenditures. Two approaches are commonly used to facilitate international R&D comparisons: (1) normalize national R&D expenditures by dividing by GDP, which circumvents the problem of currency conversion; and (2) convert all foreign-denominated expenditures to a single currency, which results in indicators of absolute effort. The first method is a straightforward calculation that permits only gross national comparisons of R&D intensity. The second method permits absolute-level comparisons and analyses of countries' sector- and field-specific R&D, but it entails choosing an appropriate method of currency conversion.

Because no widely accepted R&D-specific exchange rates exist, the choice is between market exchange rates (MERs) and purchasing power parities (PPPs). These rates are the only series consistently compiled and available for a large number of countries over an extended period of time.

MERs. At their best, MERs represent the relative value of currencies for goods and services that are traded across borders; that is, MERs measure a currency's relative international buying power. However, MERs may not accurately reflect the true cost of goods or services that are not traded internationally. In addition, fluctuations in MERs as a result of currency speculation, political events such as wars or boycotts, and official currency intervention, which have little or nothing to do with changes in the relative prices of internationally traded goods, greatly reduce their statistical utility.

PPPs. PPPs were developed because of the shortcomings of MERs described above (Ward 1985). PPPs take into account the cost differences across countries of buying a similar "market basket" of goods and services in numerous expenditure categories, including nontradables. The PPP basket is therefore assumed to be representative of total GDP across countries.

Although the goods and services included in the market basket used to calculate PPP rates differ from the major components of R&D costs (fixed assets as well as wages of scientists, engineers, and support personnel), they still result in a more suitable domestic price converter than one based on foreign trade flows. Exchange rate movements bear little relationship to changes in the cost of domestically performed R&D. The adoption of the euro as the common currency for many European countries provides a useful example: although Germany and Portugal now share a common currency, the real costs of most goods and services are substantially less in Portugal. PPPs are therefore the preferred international standard for calculating cross-country R&D comparisons wherever possible and are used in all official R&D tabulations of OECD.*

Because MERs tend to understate the domestic purchasing power of developing countries' currencies, PPPs can produce substantially larger R&D estimates than MERs do for these countries. For example, China's 2005 R&D expenditures are $30 billion using MERs but are $115 billion using PPPs. Appendix table 4-2 shows the relative difference between MERs and PPPs for a number of countries.

Although PPPs are available for developing countries such as India and China, there are several reasons why they may be less useful for converting R&D expenditures than in more developed countries:

• It is difficult or impossible to assess the quality of PPPs for some countries, most notably China. Although PPP estimates for OECD countries are quite reliable, PPP estimates for developing countries are often rough approximations. The latter estimates are based on extrapolations of numbers published by the United Nations International Comparison Program and by Professors Robert Summers and Alan Heston of the University of Pennsylvania and their colleagues.
• The composition of the market basket used to calculate PPPs likely differs substantially between developing and developed countries. The structural differences in the economies of developing and developed countries, as well as disparities in income, may result in a market basket of goods and services in a developing country that is quite different from the market basket of a developed country, particularly as far as these baskets relate to the various costs of R&D.
• R&D performance in developing countries often is concentrated geographically in the most advanced cities and regions in terms of infrastructure and level of educated workforce. The costs of goods and services in these areas can be substantially greater than for the country as a whole.

R&D in the ICT Sector

Information and communications technologies (ICTs) play an increasingly important role in the economies of OECD member countries. Both the production and use of these technologies contribute to output and productivity growth. Compared with other industries, ICT industries are among the most R&D intensive, with their products and services embodying increasingly complex technology. Because R&D data are often unavailable for detailed industries, for the purpose of this analysis, ICT industries include the following International Standard Industrial Classification categories:

• Manufacturing industries: 30 (office, accounting, and computer machinery), 32 (radio, television, and communications equipment), and 33 (instruments, watches, and clocks)
• Services industries: 64 (post and communications) and 72 (computer software and related activities) (OECD 2002)

The ICT sector accounted for more than one-quarter of total business R&D in 11 of the 19 OECD countries shown in (figure 4-25), and more than half of total business R&D in Finland, Ireland, and South Korea. ICT industries accounted for 37% of the business R&D in the United States and 32% of Japanese business R&D. Of the other G-7 countries, Canada comes closest to matching the ICT R&D concentration of the United States and Japan (figure 4-25).

Government Funding Mechanisms for Academic Research

Because U.S. universities generally do not maintain data on departmental research, U.S. totals are understated relative to the R&D effort reported for other countries. The national totals for Europe, Canada, and Japan include the research component of general university fund (GUF) block grants provided by all levels of government to the academic sector. These funds can support departmental R&D programs that are not separately budgeted. GUF is not equivalent to basic research. The U.S. federal government does not provide research support through a GUF equivalent, preferring instead to support specific, separately budgeted R&D projects, usually to address the objectives of the federal agencies that provide the R&D funds. However, some state government funding probably does support departmental research at public universities in the United States.

The treatment of GUF is one of the major areas of difficulty in making international R&D comparisons. In many countries, governments support academic research primarily through large block grants that are used at the discretion of each individual higher education institution to cover administrative, teaching, and research costs. Only the R&D component of GUF is included in national R&D statistics, but problems arise in identifying the amount of the R&D component and the objective of the research. Government GUF support is in addition to support provided in the form of earmarked, directed, or project-specific grants and contracts (funds for which can be assigned to specific socioeconomic categories). In the United States, the federal government (although not necessarily state governments) is much more directly involved in choosing which academic research projects are supported than are national governments in Europe and elsewhere. In each of the European G-7 countries, GUF accounts for 50% or more of total government R&D to universities, and in Canada it accounts for roughly 45% of government academic R&D support. These data indicate not only relative international funding priorities, but also funding mechanisms and philosophies regarding the best methods for financing academic research.

Foreign Direct Investment in R&D

Foreign direct investment (FDI) refers to the ownership of productive assets outside the home country by multinational corporations (MNCs). More specifically, the Bureau of Economic Analysis (BEA) defines direct investment as ownership or control of 10% or more of the voting securities of a business in another country (BEA 1995). A company located in one country but owned or controlled by a parent company in another country is known as an affiliate. Affiliate data used in this section are for majority-owned affiliates, i.e., those in which the ownership stake of parent companies is more than 50%. Statistics on R&D by affiliates of foreign companies in the United States and by foreign affiliates of U.S. MNCs and their parent companies are part of operations data obtained from BEA's Survey of Foreign Direct Investment in the United States (FDIUS) and BEA's Survey of U.S. Direct Investment Abroad (USDIA), respectively. Operations data exclude depository institutions and are on a fiscal-year basis.

Global R&D supports a range of objectives, from technology adaptation to the development of new products or services (Kumar 2001; Niosi 1999). The location decision for global R&D sites is driven by market- and science-based factors, including cost considerations, the investment climate, the pull of large markets, and the search for location-specific expertise (von Zedtwitz and Gassmann 2002). Furthermore, the relative importance of these factors is likely to vary depending on the industry, the technology objectives of the overseas activity, and host country characteristics relative to those of home countries. For example, in a recent study examining motives to locate R&D overseas, Thursby and Thursby (2006) report that the size of output markets and the quality of R&D personnel are the top "attractors" for FDI R&D in emerging markets, whereas the activities associated with strong research universities remain a key factor for R&D in the home market or in overseas developed economies. Barriers or challenges include managing and coordinating knowledge on a global scale and intellectual property protection.

Linking MNC Data From International Investment and Industrial R&D Surveys

An ongoing data development project aims to integrate the statistical information from the BEA's international investment surveys with the NSF/Census Survey of Industrial Research and Development. Such data sharing among federal statistical agencies has been facilitated by the Confidential Information Protection and Statistical Efficiency Act of 2002. Combining technological and investment data from these separate but complementary sources will facilitate a better assessment of globalization trends in R&D and technological innovation. The initial methodological study (completed in 2005) demonstrated not only the feasibility of such a linkage, but also its utility.

A combined preliminary dataset provided information for the first time on R&D expenditures by U.S. and foreign MNCs by character of work (basic research, applied research, development). The study also has produced tangible benefits for the participating agencies, including improvements in survey sampling and the quality of reported data. As a result of these promising initial results, the three participating agencies are considering future work in this area. For more information, see NSF/SRS (2007e) and Census Bureau et al. (2005).

A Window Into Open or Collaborative Innovation

Industrial innovation is increasingly global and performed collaboratively, requiring partners, resources, and ideas outside the company and national boundaries (Chesbrough, Vanhaverbeke, and West 2006; OECD 2006c). Knowledge may be generated internally, codeveloped, or acquired from a variety of private and public sources, then further developed for a specific market. Often, to successfully enter the marketplace ahead of competitors, an invention or new organizational method requires a new business model (Chesbrough 2007), as well as complementary assets such as manufacturing, marketing, or distribution capabilities. The latter may also be developed internally, acquired, or outsourced (Howells 2006; Teece 1986). The following excerpts from publications provide a flavor of some of the current industry thinking and activities in this area.

Harvard Business Review
Connect and Develop: Inside Proctor & Gamble's New Model for Innovation

As we studied outside sources of innovation, we estimated that for every P&G [Procter & Gamble] researcher there were 200 scientists or engineers elsewhere in the world who were just as good—a total of perhaps 1.5 million people whose talents we could potentially use. But tapping into the creative thinking of inventors and others on the outside would require massive operational changes. We needed to move the company's attitude from resistance to innovations "not invented here" to enthusiasm for those ‘proudly found elsewhere.' And we needed to change how we defined, and perceived, our R&D organization—from 7,500 people inside to 7,500 plus 1.5 million outside, with a permeable boundary between them. (Huston and Sakkab 2006)

Business Week
Crowdsourcing: Milk the Masses for Inspiration

Business model innovation is happening at a lightning clip. First there was outsourcing, then open-sourcing, and now crowdsourcing. . . . Crowdsourcing often produces a wealth of ideas, and companies need effective filters to pick the gems. Consider IBM's innovation jam, a two-part brainstorming session launched in July [2006] designed to tap the collective minds of employees, family members, and customers to target potential areas for innovation. CEO Sam Palmisano will put $100 million into promising ideas. (Hempel 2006)

Chemical & Engineering News
Start-Up Firm NineSigma Uses Internet To Match Industrial Clients With Inventive Partners

In his 28 years at Procter & Gamble, Paul Stiros says he never doubted the wisdom behind connecting R&D to customer needs. As president and chief executive officer of privately held NineSigma, Stiros heads a firm committed to helping corporations acquire technical innovations that will quickly bring tomorrow's star products to market. . . . Competing firms such as InnoCentive and YourEncore also help corporations get research help outside the usual channels. InnoCentive posts specific problems for corporate customers on the Internet and pays a bounty for solutions. YourEncore connects technology and product development needs of member companies with retirees who have scientific backgrounds. (American Chemical Society 2006)

Boeing
YourEncore and Your Retirement

Boeing partnered in August 2003 with YourEncore Inc. to provide Boeing retirees with scientific and engineering skills [and] challenging and rewarding project opportunities in various industries, including aerospace, chemical, communications, pharmaceutical and consumer products. Retirees can contribute their expertise to major companies on high-level projects while networking among peers and gaining experience in new industries. . . . "YourEncore is an ideal opportunity for Boeing retirees to stay intellectually engaged on a part-time basis to the degree the retiree wishes and get fairly compensated," said Dick Paul, Boeing Phantom Works* vice president, strategic development and analysis. "Boeing retirees can join YourEncore and consult either back at Boeing or with other member companies in varied industries." (Sopranos 2004)

Major Federal Legislation Related to Cooperative R&D and Technology Tranfer

Stevenson-Wydler Technology Innovation Act (1980). Required federal laboratories to facilitate the transfer of federally owned and originated technology to state and local governments and the private sector.

Bayh-Dole University and Small Business Patent Act (1980). Permitted government grantees and contractors to retain title to federally funded inventions and encouraged universities to license inventions to industry. The act is designed to foster interactions between academia and the business community.

Small Business Innovation Development Act (1982). Established the Small Business Innovation Research (SBIR) program within the major federal R&D agencies to increase government funding of research that has commercialization potential within small high-technology companies.

National Cooperative Research Act (1984). Encouraged U.S. firms to collaborate on generic, precompetitive research by establishing a rule of reason for evaluating the antitrust implications of research joint ventures. The act was amended in 1993 by the National Cooperative Research and Production Act, which let companies collaborate on production and research activities.

Federal Technology Transfer Act (1986). Amended the Stevenson-Wydler Technology Innovation Act to authorize cooperative R&D agreements (CRADAs) between federal laboratories and other entities, including other federal agencies, state or local governments, universities and other nonprofit organizations, and industrial companies.

Omnibus Trade and Competitiveness Act (1988). Established the Competitiveness Policy Council to develop recommendations for national strategies and specific policies to enhance industrial competitiveness. The act created the Advanced Technology Program and the Manufacturing Technology Centers within NIST to help U.S. companies become more competitive.

National Competitiveness Technology Transfer Act (1989). Amended the Stevenson-Wydler Act to allow government-owned, contractor-operated laboratories to enter into CRADAs.

National Cooperative Research and Production Act (1993). Relaxed restrictions on cooperative production activities, enabling research joint venture participants to work together in the application of technologies they jointly acquire.

Technology Transfer Commercialization Act (2000). Amended the Stevenson-Wydler Act and the Bayh-Dole Act to improve the ability of government agencies to monitor and license federally owned inventions.