From the July 1994 SURVEY OF CURRENT BUSINESS: Foreign Direct Investment in the United States: 1992 Benchmark Survey Results
By William J. Zeile
U.S. affiliates of foreign companies accounted for a slightly smaller share of the U.S. economy in 1992 than in 1991./1/ According to preliminary results of BEA's latest benchmark survey of foreign direct investment in the United States (see the box on page 158), the affiliate share of gross domestic product (GDP) of all nonbank U.S. businesses was 5.8 percent in 1992; the comparable figure for 1991 was 6.0 percent (table 1 and chart 1)./2/ Although the gross product of affiliates in current dollars grew 3.2 percent in 1992, in constant dollars, the gross product of affiliates was essentially flat in 1992, compared with a growth rate of about 3 percent for all U.S. businesses; in 198791, growth in the real gross product of affiliates was higher than that of all U.S. businesses (chart 2)./3/
The decrease in the affiliate share of GDP in 1992 was the first since 1985 and followed several years of rapid growth. It primarily reflected a sharp drop in new investment activity. Despite the decrease, affiliate operations accounted for a substantially larger share of U.S. economic activity in 1992 than in 1987, the year of the last benchmark survey.
The following are other highlights of the survey for 1992:
- The growth in total assets of affiliates slowed substantially from previous years; the slowdown was accompanied by a large drop in affiliate expenditures for new plant and equipment.
- Employment by nonbank affiliates declined 3 percent, the first decrease since at least 1977, when BEA began to collect annual data on affiliate operations. Increases in employment resulting from new investments were much smaller in 1992 than in 1991, and they were more than offset by decreases in employment resulting from sales and liquidations of foreign ownership interests.
- For the third consecutive year, the after-tax net income of nonbank affiliates was negative. Affiliates reported losses of $20 billion, which includes special charges taken against earnings in order to conform to new accounting standards for post-employment and post-retirement benefits and for deferred income taxes. Operating profits on a national income accounting basis were positive after having been negative in 1991.
- As in earlier years, more than one-half of the gross product of nonbank affiliates was in manufacturing. In 198792, the affiliate share of all-U.S.-business gross product in manufacturing increased from 10 percent to 14 percent.
- Affiliates with ultimate beneficial owners (UBO's) in the United Kingdom and Japan accounted for the largest shares of total affiliate gross product21 percent and 16 percent, respectively./4/ Canadian-owned affiliates ranked third, with a share of 13 percent; before 1991, their share was higher than that of Japanese-owned affiliates.
- The affiliates' share of total U.S. expenditures on research and development (R&D), at about 13 percent, was much higher than their share of all-U.S.-business GDP. Their higher share of R&D reflects the typically large size of affiliates and their tendency to be concentrated in research-intensive industries. The number of affiliate employees engaged in R&D was 104,000, or about 2 percent of affiliate employment.
- For all industries combined, the research intensity of affiliates that performed R&D was about the same as that of all R&D-performing U.S. companies; however, in many individual industries, the research intensity of affiliates was substantially lower than that of all R&D-performing companies.
- Affiliates tended to be more highly unionized than all U.S. companies. Employees covered by collective bargaining agreements accounted for one-fifth of total employment by U.S. affiliates, compared with one-eighth for all U.S. businesses. In manufacturing, however, the union-employment shares for affiliates and all U.S. businesses were much closer (24 percent and 21 percent, respectively).
- Goods shipped by U.S. affiliates accounted for 22 percent of total U.S. merchandise exports. For two major product categoriesfood products and petroleum productsaffiliates accounted for more than one-half of total exports.
- Goods shipped to U.S. affiliates accounted for 34 percent of total U.S. merchandise imports; affiliates accounted for about one-half or more of the imports of beverages and tobacco, chemicals, road vehicles and parts, and metal manufactures. More than two-thirds of the imports by affiliates were goods for resale without further manufacture by the affiliates, reflecting the large share of imports that was accounted for by wholesale trade affiliates.
This article first discusses changes in affiliate employment in 1992. It then briefly examines affiliate net income and operating profits in 1991 and 1992. Next, the article reviews changes in the share of the U.S. economy accounted for by U.S. affiliates since 1987, the year of the last benchmark survey. It then analyzes information on research and development, employment, and merchandise trade from the 1992 benchmark survey that either has not been available before or has not been available since the last benchmark survey. Finally, it summarizes selected data for majority-owned affiliates. In the discussion, information from outside sources, including press reports on specific companies, is used to assist in the analysis and interpretation of the survey results.
Employment in 1992
Employment by nonbank U.S. affiliates decreased 166,000 in 1992 to 4,705,000 after increasing 137,000 in 1991. The decrease was the first since at least 1977, when BEA began collecting annual data on U.S. affiliate operations (chart 3). The decrease can be attributed mainly to the fact that increases in employment due to new foreign investments were more than offset by decreases in employment due to sales and liquidations of affiliate businesses. New investments added only 100,000 employees in 1992, compared with 291,000 in 1991 and 482,000 in 1990 (table 2). Sales and liquidations of affiliate businesses reduced employment by 293,000. Other changes had little net effect on employment: Decreases due to cutbacks in existing operations of affiliates were roughly balanced by increases due to expansions of existing operations.
By industry of affiliate, employment decreased in every major industry (table 3). In retail tradea very labor-intensive industryemployment dropped 90,000, amounting to more than one-half of the total decrease in affiliate employment in 1992. This drop was more than accounted for by the liquidation of the ownership interest of a Canadian investor in a nationwide U.S. retail chain. Affiliate employment decreased 33,000 in "other industries," mainly in construction and transportation; a large part of the decrease in transportation was due to the liquidation of a Canadian interest in a U.S. railroad company.
By country of UBO, the largest decreases in employment were by affiliates with UBO's in Canada (136,000) and the United Kingdom (116,000). The largest increases were by affiliates with UBO's in the United States (62,000), Switzerland (16,000), and Japan (9,000)./5/
Affiliate employment decreased in two-thirds of the States. The largest decreases were in California (39,000), New York (32,000), Florida (16,000), and New Jersey (13,000); in each State, a sizable portion of the decrease was accounted for by the liquidation of the Canadian interest in a nationwide retail chain. The only States with substantial increases in affiliate employment were Texas (11,000) and North Carolina (10,000).
Net Income in 1992
Affiliates reported losses in 1992 of $20 billion, almost twice as large as their losses in 1991 (table 4); prior to 1990, the net income of affiliates had been positive. These figures are after taxes, and they include capital gains, income from investments, and other nonoperating income. However, the "profit-type return" of affiliatesan economic accounting measure of the profits generated from production (see footnote to table 4)was positive in 1992 ($2 billion) after being negative in 1991 (-$2 billion). Much of the drop in net income in 1992 was due to one-time adjustments to earnings made by many affiliates to conform with new accounting standards for post-employment and post-retirement benefits and for deferred income taxes. The net effect of these adjustments was to reduce net income by a substantial amount. However, the adjustments had no effect on the profit-type-return measure.
By major industry, affiliate net income in 1992 was negative in every industry except nonbank finance and insurance. In three industriespetroleum, manufacturing, and wholesale tradeaffiliates incurred losses despite having a positive profit-type return, and in a fourth"other industries"affiliates incurred losses on a net-income basis that were many times larger than their losses in terms of profit-type return.
In manufacturing, net income dropped $5 billion in 1992 despite a $2 billion increase in profit-type return. Almost all of the decrease was accounted for by affiliates in chemicals; it mainly reflected the one-time accounting adjustments described earlier for post-employment and post-retirement benefits. In "other industries," net income dropped $3 billion, mainly reflecting one-time accounting adjustments for deferred income taxes by affiliates in the communications industry.
In some industries, the negative net income of affiliates reflected continuing losses from current operations (that is, negative profit-type return). Operating losses were particularly large for affiliates in real estate and services; most of the losses in services were by affiliates in the hotel and motion picture industries.
Share of the U.S. Economy, 198792
This section discusses changes in the share of the U.S. economy accounted for by nonbank U.S. affiliates since 1987, the year of the last benchmark survey. The changes are discussed in terms of two measures of economic activity: Gross product (an economic accounting measure of production) and employment. Unlike the data on gross product, the data on U.S.-affiliate employment are available by industry of sales as well as by industry of affiliate (see the box on this page). Because the affiliate employment data classified by industry of sales are roughly comparable to the all-U.S.-businesses employment data classified by industry of establishment, they can be used to calculate affiliate shares of the U.S. economy at a greater level of industry detail than is appropriate using the gross product data, which are available only by industry of affiliate./6/ Data on affiliate employment are also collected by State; thus, affiliates' share of all-U.S.-business employment in each State can be computed.
Whether measured in terms of gross product or employment, affiliates' share of the U.S. economy has increased substantially since 1987. Much of the increase was the result of acquisitions by foreigners of existing U.S. companies.
The U.S.-affiliate share of the gross product of all nonbank U.S. businesses increased steadily from 4.5 percent in 1987 to 6.0 in 1991 and then declined to 5.8 percent in 1992./7/ In 198287, the share had edged up from 4.3 percent to 4.5 percent.
Most of the increase in share occurred in 198789, when new investment activity was strong. Annual outlays by foreign investors to acquire and establish U.S. business enterprises averaged $61 billion in 198789, compared with $19 billion in 198286. After continuing to increase in 1990 and 1991, the affiliate share of all-U.S.-business gross product declined in 1992, when investment outlaysat $15 billionwere lower than at any time since 1984./8/
By industry.In 198792, the affiliate share of all-U.S.-business gross product increased substantially in manufacturing, services, and "other industries," but decreased substantially in nonbank finance (table 5)./9/
In manufacturing, which consistently accounted for more than one-half of total affiliate gross product, the affiliate share of all-U.S.-business GDP increased from 10.5 percent in 1987 to 14.2 percent in 1991, dipping slightly in 1992.
In services, the affiliate share increased from 0.9 percent in 1987 to 1.8 percent in 1992. Underlying this increase were rapid increases in affiliate gross product in such industries as hotels, business services, and motion pictures.
In "other industries," the affiliate share increased from 1.2 percent to 2.2 percent, reflecting large increases in affiliate gross product in mining, transportation, and communications and public utilities.
In nonbank finance, the affiliate share of all-U.S.-business gross product increased from 16.5 percent in 1987 to 18.7 percent in 1989 and then dropped to 6.1 percent in 1990. The large drop in 1990 was due to a reduction, to below 10 percent, in a foreign investor's minority stake in a large U.S. firm and, to a lesser extent, to a sizable reduction in operations by another minority-owned firm.
By country.In 1992, affiliates with UBO's in the United Kingdom and Japan accounted for the largest shares of total affiliate gross product21.5 percent and 16.2 percent, respectively (table 6 and chart 4). The share for British-owned affiliates changed little in 198792, but the share for Japanese-owned affiliates increased by almost one-halffrom 11.1 percent in 1987. The share for French-owned affiliates also increased substantiallyfrom 5.3 percent to 7.3 percent. In contrast, the share for Canadian-owned affiliates dropped from 18.4 percent to 12.5 percent; about one-half of this decrease occurred in 1992.
The share of all-U.S.-business employment accounted for by affiliates increased every year in 198791rising from 3.7 percent to 5.3 percentand then declined to 5.1 percent in 1992 (table 7).
By industry.In 198792, the affiliate shares of all-U.S.-business employment increased substantially in mining, insurance, transportation, and manufacturing.
In mining, the major industry in which the affiliate share has consistently been highest, the share increased through 1990, when it peaked at 12.7 percent, and then declined to 12.3 percent by 1992.
In insurance, the share increased from 3.7 percent in 1987 to 6.5 percent in 1991 and then dropped slightly in 1992.
In transportation, the share increased rapidly in 198790, from 2.7 percent to 6.1 percent, but leveled off in 1991 and declined to 5.1 percent in 1992. The increase in 198790 was largely due to a few major acquisitions by foreign companies (including acquisitions of minority interests in U.S. airlines); the decline in 1992 was mainly the result of the liquidation of a Canadian investor's interest in a U.S. railroad.
In manufacturing, the share increased steadily, from 7.7 percent in 1987 to 11.6 percent in 1991 and remained at 11.6 percent in 1992. Within manufacturing, the largest increases in affiliate shares in 198792 were in rubber and plastics products (from 6.6 percent to 14.8 percent) and in stone, clay, and glass products (from 14.5 percent to 21.0 percent). The increase in rubber and plastics products was mainly due to two large acquisitions in tire manufacturingone in 1988 by a Japanese tire manufacturer and the other in 1990 by the U.S. affiliate of a French tire manufacturer. The increase in stone, clay, and glass products also reflected substantial acquisition activity, particularly in 198890.
Affiliate shares also increased substantially in chemicals, electric and electronic equipment, machinery, and motor vehicles. The share in motor vehicles increased from 6.6 percent in 1987 to 12.2 percent in 1991 and then declined to 11.1 percent in 1992. The increase in 198791 largely reflected expansions in the operations of Japanese-owned affiliates; the decline in 1992 was partly due to the liquidation of a Canadian-owned affiliate and to reductions resulting from the consolidation of a Swedish-owned affiliate.
By State.The share of all-U.S.-business employment accounted for by affiliates in each of the years 198792 was highest in Delaware, followed by Hawaii and South Carolina (table 8). The affiliate share in Delaware was about 14 percent in 198891, but dropped to less than 12 percent in 1992; most of this drop was due to cutbacks in employment at a large minority-owned affiliate. In Hawaii, the affiliate share increased rapidly in 198791from 7.2 percent to 12.5 percentand then declined slightly in 1992; more than two-thirds of affiliate employment was by Japanese-owned affiliates. In South Carolina, the affiliate share increased every yearfrom 6.5 percent in 1987 to 8.8 percent in 1992.
Expanded Information from the Benchmark Survey
The 1992 benchmark survey provides data on expenditures on research and development (R&D) performed by U.S. affiliates, whether financed by the affiliates themselves or by others./10/ These data, which were collected for the first time since the 1974 benchmark survey, are comparable to the data on R&D performed by all U.S. companies that are published by the National Science Foundation (NSF). In addition, information was collected on the R&D employment of affiliates (that is, the number of scientists, engineers, and other employees engaged in R&D); such data had been previously collected only in the 1980 benchmark survey.
The 1992 benchmark survey also provides the following data that are regularly collected in benchmark surveys but not in annual surveys: The number of affiliate employees covered by collective bargaining agreements, U.S. merchandise exports and imports of affiliates by product and by country of destination or origin, and merchandise imports of affiliates by intended use.
The following discussion presents some findings from these new data, particularly as they relate to similar data for all U.S. businesses.
Research and development
In 1992, expenditures on R&D performed by U.S. affiliates totaled $14 billion, about 13 percent of the NSF estimate of R&D performed by all U.S. businesses (table 9). The R&D employment of affiliates was 104,000, about 15 percent of the R&D employment of all U.S. businesses.
Of the total R&D performed by affiliates, nearly all95 percentwas financed by the affiliates themselves; only 1 percent was financed by the Federal Government and only 4 percent was financed by other private companies for which affiliates performed R&D under contract. In contrast, 27 percent of the R&D performed by all U.S. businesses was financed by the Federal Government. U.S. affiliates accounted for 17 percent of the privately funded R&D performed by all U.S. businesses, but for less than 1 percent of the federally funded R&D. The low affiliate share of federally funded R&D may reflect the fact that much of this research is military related and therefore generally off limits to foreign-owned companies.
The share of R&D accounted for by affiliates was much higher than their share of all-U.S.-business GDP (6 percent). The higher share in R&D partly reflects the fact that U.S. affiliates tend to be large companies, which perform most of the R&D in the United States. It also reflects the tendency for affiliates to be more concentrated in research-intensive industries, such as chemicals.
Table 10 presents industry detail on the privately funded R&D performed by affiliates and on the R&D employment of affiliates. Also shown, by industry, are two measures of the research intensity of R&D-performing affiliates and of all R&D-performing U.S. companies: Privately funded R&D as a percentage of sales and R&D employment as a percentage of total employment./11/
For affiliates, the two measures of research intensity were highest in two service industries: Computer and data processing services (for which the R&D-expenditures measure was 13 percent and the R&D-employment measure was 20 percent) and accounting, research, and management services (for which the two measures were 35 percent and 18 percent, respectively). Within manufacturing, affiliates in drugs had the highest research intensity.
For all industries combined, the research intensity of R&D-performing affiliates was very similar to that of all R&D-performing U.S. companies: It was marginally lower on the basis of the expenditures measure and marginally higher on the basis of the employment measure. The similarity between the measures at the all-industries level, however, appears largely to reflect offsetting industry-mix and within-industry effects: Affiliates tend to be concentrated in high-research-intensity industries, but for most industries for which comparable data are available, their research intensity was lower than that of all U.S. companies. Affiliates had a lower research intensity in 13 of the 22 industries with comparable data on R&D expenditures and in 8 of the 13 industries with comparable data on R&D employment. The most substantial differences were in computer and office equipment, electronic components, transportation equipment, and instruments.
In a few industries, the research intensity of affiliates was higher than that of all U.S. companies. The largest differences were in drugs and in audio, video, and communications equipment.
The finding that U.S. affiliates generally have a lower research intensity than all R&D-performing U.S. companies is not as strong as might be expected, given the well-known tendency for large multinational corporations to locate most of their R&D near company headquarters in the country of ownership. In the case of foreign direct investment in the United States, the advantages of locating R&D near company headquarters may be partly offset by advantages stemming from proximity to U.S. research centers and access to the large U.S. pool of scientists and engineers.
In 1992, 20 percent of the total employment of U.S. affiliates was covered by collective bargaining agreements, compared with a share of 13 percent for all U.S. businesses (table 11). This difference partly reflects industry-mix effects; for example, manufacturing (an industry with relatively high unionization) accounted for nearly one-half of the employment of affiliates but for only one-fifth of the employment of all U.S. businesses. However, even on a disaggregated industry basis, the union-employment share for U.S. affiliates exceeded that for all U.S. businesses in most industries for which comparable data are available. The higher shares for affiliates can probably be attributed to the tendency for foreign direct investment to be confined to large-scale enterprises, which generally have higher rates of unionization than small businesses.
Among industries, the union-employment share for affiliates was particularly high relative to that of all U.S. businesses in retail trade (20 percent, compared with 7 percent) and construction (33 percent, compared with 21 percent). In manufacturing, the union-employment share for affiliates was much closer to that of all U.S. businesses (24 percent, compared with 21 percent). In nondurables manufacturing, the shares were almost identical.
Within manufacturing, the union-employment share for affiliates varied considerably. The share was highest in primary metals (52 percent) and lowest in instruments (8 percent).
U.S. affiliates continued to account for a significant share of total U.S. merchandise trade in 1992. U.S. merchandise exports shipped by affiliates were $101 billion, or 22 percent of the U.S. total; U.S. merchandise imports shipped to affiliates were $182 billion, or 34 percent of the U.S. total.
By product.In 1992, U.S. affiliates accounted for about one-half of U.S. exports of food products and for about three-fourths of U.S. exports of petroleum and products (table 12). In contrast, they accounted for less than 15 percent of U.S. exports of machinery, road vehicles and parts, and other transport equipment.
The share of food exports accounted for by affiliates in 1992 was nearly identical to the share in 1987. In both years, nearly all of the exports were by wholesale trade affiliates (table 13), mainly by those with UBO's in Japan, France, and Switzerland.
The share of petroleum exports accounted for by affiliates was somewhat larger in 1992 than in 1987. About two-thirds of affiliates' petroleum exports in 1992 were by affiliates in wholesale trade, mainly affiliates of Japan's large general trading companies (the sogo shosha) and French-owned affiliates. Most of the remainder was by affiliates specializing in petroleum wholesale trade (which is classified as part of the petroleum industry).
On the import side, U.S. affiliates accounted for about 60 percent of U.S. imports of beverages and tobacco in 1992 and for about one-half of U.S. imports of chemicals, road vehicles and parts, and metal manufactures. For beverages and tobacco, the affiliate share was substantially higher in 1992 than in 1987; for road vehicles and parts, however, it was substantially lower.
Although product detail for trade by U.S. affiliates is not collected annually, the drop in the affiliate share for road vehicles and parts appears to reflect a steady declinefrom $50 billion in 1987 to $35 billion in 1992in imports by wholesale trade affiliates specializing in motor vehicles and equipment./12/ These affiliates tend to function as the primary distribution channel for finished vehicles produced in their home countries. The decline in their imports may partly reflect the substitution of production by foreign-owned auto plants in the United States for production overseas; in 198792, sales by U.S. affiliates in motor vehicles and equipment manufacturing increased from $6 billion to $16 billion.
Imports by intended use.As in 1987, more than two-thirds of the imports by U.S. affiliates in 1992 were goods for resale without further processing, assembly, or manufacture by the affiliates. In the case of imports by wholesale trade affiliates, the share of goods for resale without further manufacture was more than 90 percent. Most of the remaining imports by affiliates were goods for further manufacture by the affiliates; as would be expected, these imports were mainly by manufacturing affiliates. In 1992, 70 percent of the imports by manufacturing affiliates were goods for further manufacture.
Imports of capital equipment accounted for only 1 percent of affiliate imports. They accounted for 2 percent of the imports by manufacturing affiliates and for 4 percent of the imports by affiliates in "other" industries. In the latter group, most of the capital-equipment imports were by affiliates in business services or transportation.
By country of destination or origin.In 1992, U.S. affiliates accounted for two-thirds of total U.S. exports to Japan and for more than one-third of U.S. exports to China (table 14). For most of the other major trading-partner countries, the share of U.S. exports accounted for by affiliates ranged from 10 to 20 percent.
More than 80 percent of affiliate exports to Japan were by Japanese-owned affiliates, mainly wholesale trade affiliates. Japanese-owned affiliates also accounted for more than one-half of U.S.-affiliate exports to both Malaysia and Taiwan.
On the import side, U.S. affiliates accounted for more than three-fourths of total U.S. imports from Japan and Switzerland in 1992 and for more than one-half of U.S. imports from Sweden, Germany, the Netherlands, and Venezuela. Most of the U.S.-affiliate imports from each of these countries were by affiliates with UBO's in that country. In the case of Japan, Sweden, and Germany, these imports were mainly by wholesale trade affiliates functioning as distribution channels for manufactures produced in the investing country.
Japanese-owned affiliates accounted for more than 97 percent of U.S.-affiliate imports from Japan. They also accounted for about 60 percent of U.S.-affiliate imports from Singapore and about 40 percent of U.S.-affiliate imports from Taiwan and Thailand.
Majority-Owned U.S. Affiliates
The estimates presented thus far have covered the operations of all U.S. nonbank affiliatesthat is, all U.S. nonbank companies that are owned 10 percent or more by a foreign direct investor. This section presents estimates for nonbank majority-owned U.S. affiliates (MOUSA's), which are affiliates owned more than 50 percent by foreign direct investors. It also examines industries in which minority-owned U.S. affiliates account for a sizable portion of the data for all nonbank U.S. affiliates.
Table 15 shows estimates of gross product, total assets, sales, and employment for MOUSA's and gives their shares of the affiliate totals for these items. Most of the MOUSA shares are high because most U.S. affiliates are majority owned. Altogether, MOUSA's accounted for about four-fifths or more of the gross product, total assets, sales, and employment of all nonbank U.S. affiliates.
The following paragraphs briefly discuss MOUSA shares of gross product of all nonbank affiliates by major industry, area, and country. Except where noted, the distributions of MOUSA shares of total assets, sales, and employment tend to be similar to those of gross product.
In manufacturing, MOUSA's accounted for 83 percent of the gross product of all U.S. manufacturing affiliates. The share was highest in food and kindred products (99 percent) and lowest in primary and fabricated metals (70 percent).
Excluding manufacturing, the share of gross product accounted for by MOUSA's was highest in wholesale trade (96 percent). It was lowest in nonbank finance (36 percent); in this industry, however, MOUSA's accounted for much larger shares of total assets and sales (86 percent and 81 percent, respectively). The MOUSA shares of total assets, sales, and employment were lowest in "other industries."
By area, the share of gross product for MOUSA's was highest for affiliates with UBO's in Europe (88 percent) and lowest for those with UBO's in the United States. By major country, MOUSA's with UBO's in the United Kingdom had the highest share (93 percent); MOUSA's with UBO's in Australia had the lowest share (41 percent).
Although MOUSA's accounted for a dominant share of the data for all affiliates in most industries, there were a few industries in which minority-owned affiliates were important. Table 16 identifies, at a more detailed level of aggregation, the specific industries in which minority-owned affiliates accounted for sizable sharesat least 30 percentof the gross product, total assets, sales, or employment of all nonbank U.S. affiliates. Minority-owned affiliates accounted for particularly large sharesmore than one-halfof the gross product of affiliates in three industries: Primary ferrous metals, transportation, and communication and public utilities. Their share of gross product was just under one-half in computer and data processing services.
- A U.S. affiliate is a U.S. business enterprise in which there is foreign direct investmentthat is, in which a single foreign person owns or controls, directly or indirectly, 10 percent or more of the voting securities of an incorporated U.S. business enterprise or an equivalent interest in an unincorporated U.S. business enterprise. An affiliate is called a U.S. affiliate to denote that it is located in the United States; in this article, "affiliate" and "U.S. affiliate" are used interchangeably. "Person" is broadly defined to include any individual, corporation, branch, partnership, associated group, association, estate, trust, or other organization and any government (including any corporation, institution, or other entity or instrumentality of a government).
- Although the benchmark survey covered both bank and nonbank affiliates, only data for nonbank affiliates are presented in this article; thus, the data for 1992 are consistent with the data from BEA's annual surveys for other years, which cover only nonbank affiliates.
- The data used to estimate gross product of affiliates are reported to BEA in current dollars. Price indexes specifically designed for deflating production by affiliates are unavailable; however, rough estimates of affiliate gross product in constant dollars were constructed for 198791 by applying industry-level deflators for all U.S. businesses, weighted to take into account the industry mix of affiliate production. Industry-level deflators for all U.S. businesses are not yet available for 1992, so affiliate gross product was deflated by the implicit price deflator for nonfarm U.S. businesses, less housing.
- The UBO is that person, proceeding up a U.S. affiliate's ownership chain, beginning with and including the foreign parent, that is not owned more than 50 percent by another person. The foreign parent is the first foreign person in the affiliate's ownership chain. Unlike the foreign parent, the UBO of an affiliate may be located in the United States. The UBO of each U.S. affiliate is identified to ascertain the person that ultimately owns or controls and that, therefore, ultimately derives the benefits from owning or controlling the U.S. affiliate.
- The definition of foreign direct investment in the United States is based on whether a U.S. company has a foreign parent rather than on the location of the UBO. Thus, while all U.S. affiliates have a foreign parent, some may have a UBO that is located in the United States.
- Data on the manufacturing establishments of U.S. affiliates can be used to calculate affiliate shares of U.S. economic activity in each of the detailed manufacturing industries defined at the four-digit level of the Standard Industrial Classification. These establishment-level datathe result of a joint project of BEA and the Bureau of the Censusare currently available for the years 198790. The data for 1990 are discussed in "Characteristics of Foreign-Owned U.S. Manufacturing Establishments," SURVEY OF CURRENT BUSINESS 74 (January 1994): 3459. Data for 1991 are scheduled for publication this fall.
- The U.S.-affiliate gross product estimates for 198891 presented in this article are revised from those previously published. The revised estimates incorporate improved estimates of net interest paid, a major component of gross product. The improved estimates of net interest paid, in turn, reflect information on interest payments and receipts reported in the 1992 benchmark survey. (Such payments and receipts are reported only in benchmark survey years and must be estimated for other years.) Tables presenting the revised gross product estimates for 1991 (and preliminary estimates for 1992) disaggregated by industry and component and by industry and country of UBO appear at the end of this article. Tables presenting the same detail for the revised 198890 estimates are available from BEA on request; for information, call (202) 6069893.
- Data on investment outlays appear in "U.S. Business Enterprises Acquired or Established by Foreign Direct Investors in 1993," SURVEY 74 (May 1994): 50.
- At levels of industry disaggregation more detailed than those shown in table 5, the data used to compute the shares of gross product are not strictly comparable. The data on the GDP of all U.S. businesses are on an establishment, or plant, basis, whereas the data on gross product of affiliates are on an enterprise, or company, basis. On an enterprise basis, all of the gross product of an affiliate is assigned to its major industry, even though the affiliate may have establishments operating in a number of secondary industries.
- These data supplement the data collected in previous annual and benchmark surveys on the total expenditures on R&D financed by U.S. affiliates (whether performed by the affiliates or by others), which is the measure recommended by the Financial Accounting Standards Board for use by companies in accounting for the costs of R&D.
- Comparisons between the research-intensity measures for R&D-performing U.S. affiliates and all R&D-performing U.S. companies should be viewed as approximate, because the data used to construct them are from different sources and for different years. The data for affiliates are from the 1992 benchmark survey, and the data for all R&D-performing U.S. companies are estimates based on a sample survey of industrial firms in 1991 (with the 1991 sample itself being a subset of a larger probability sample selected for 1987). Differences between the measures at the industry level may also reflect potential differences in the industry classification of individual companies according to sales (the basis used for classifying affiliates) or payroll (the basis used for classifying R&D-performing companies in the surveys conducted by the Census Bureau for NSF).
- Product data on the merchandise trade of affiliates are collected only in benchmark survey years, but data on trade by industry of affiliate are collected annually. The annual data for 197792 show that imports by wholesale trade affiliates specializing in motor vehicles and equipment peaked in 1987 and declined thereafter.