[6] See OECD (2001) for discussion of classifying economic activities according to degree of "knowledge intensity."

[7] In designating these high-technology manufacturing industries, OECD took into account both the R&D done directly by firms and R&D embedded in purchased inputs (indirect R&D) for 13 countries: the United States, Japan, Germany, France, the UK, Canada, Italy, Spain, Sweden, Denmark, Finland, Norway, and Ireland. Direct intensities were calculated as the ratio of R&D expenditure to output (production) in 22 industrial sectors. Each sector was weighted according to its share of the total output among the 13 countries, using purchasing power parities as exchange rates. Indirect intensities were calculated using the technical coefficients of industries on the basis of input-output matrices. OECD then assumed that, for a given type of input and for all groups of products, the proportions of R&D expenditure embodied in value added remained constant. The input-output coefficients were then multiplied by the direct R&D intensities. For further details concerning the methodology used, see OECD (2001). It should be noted that several nonmanufacturing industries have equal or greater R&D intensities. For additional perspectives on OECD's methodology, see Godin B. 2004. The new economy: What the concept owes to the OECD. Research Policy 33:679–90.

[8] Data are extracted from the Global Insight World Industry Service database, which provides information for 70 countries that account for more than 97% of global economic activity. The Global Insight data on international country activity within the service and manufacturing industries are expressed in 2000 constant dollars. Constant dollar data for foreign countries are calculated by deflating industry data valued in each country's nominal currency.

[9] Compared with the extensive data available for the manufacturing industries, national data that track activity in many rapidly growing service sectors are limited in the level of industry aggregation and types of data collected. For example, export and import data are currently not available for many services.

[10] Gross revenue includes inputs or supplies purchased from other industries or services. Knowledge-intensive service and high-technology manufacturing industry data are expressed in 2000 constant dollars. Constant-dollar data for foreign countries is calculated by deflating nominal domestic currency with a sector-specific price index constructed for that country, then converting the result to U.S. dollars based on average annual market exchange rates.

[11] Asia is defined in this section as consisting of China, India, Indonesia, Japan, Malaysia, the Philippines, Singapore, South Korea, Taiwan, and Thailand. China includes Hong Kong.

[12] One of the earliest quantitative analyses of R&D was done in 1955 by R.H. Ewell, supported by the National Science Foundation. This study showed a definite correlation between research and productivity. Also see Godin B. 2004. The obsession for competitiveness and its impact on statistics: The construction of high-technology indicators. Research Policy 33:1217–29.

[13] This conclusion is derived from an examination of weighted U.S. data from the Bureau of Labor Statistics Occupational Employment Survey concerning average annual pay during the period 1997–2001.

[14] Global Insight's data show that U.S. high-technology industry manufacturers' share of value added to total output was 20% higher than the share of all other U.S. manufacturing industries.

[15] This conclusion is derived from an examination of weighted U.S. data from the Bureau of Labor Statistics Occupational Employment Survey on average annual pay from 1997–2001.

[16] Europe's success in growing its aerospace industry and China's efforts to develop a semiconductor industry are two examples.

[17] In February 1996, the Telecommunications Act became U.S. law. This Act was the first major telecommunications reform in more than 60 years. It facilitated competition between cable companies and telephone companies and may have contributed to increased U.S. manufacturing activity in both the communications and computer hardware industries.

[18] In 1999, the State Department's responsibilities under the International Traffic in Arms Regulation were expanded to include research activity formerly covered under the Commerce Department's export regulations. The transfer placed scientific satellites, related data, and certain computer components and software on the U.S. Munitions List. Related research activities and the country of origin of researchers working on related research activities also became subject to many of the same regulations controlling exports of sensitive products.

[19] Like the United States, other national governments usually have strong ties to their aerospace industries, often supporting and funding R&D and serving as major customers.

[20] Unlike the previous section that examined data on industry manufacturing value added (domestic content), the value of exports reported in this section reflects the final value of industry shipments exported, not just the value resulting from domestic production. Exported shipments will, therefore, often include the value of purchased foreign inputs.

[21] EU exports exclude intra-EU exports.

[22] The U.S. trade balance is affected by many other factors including currency fluctuations, differing fiscal and monetary policies, and export subsidies between the United States and its trading partners.