Gross product of U.S. affiliates of foreign companies, 1977-87.

Lowe, Jeffrey H.

Gross Product of U.S. Affiliates of Foreign Companies, 1977-87

THIS article presents estimates of gross product (value added) of nonbank U.S. affiliates of foreign companies--the affiliates' contribution to U.S. gross domestic product (GDP)--for 1977-87.(1) Gross product is an economic accounting measure of production. For an individual business, it can be defined as sales plus inventory change, less purchases from other businesses. Thus, it measures value added by the business. It can also be defined as the sum of income from current production plus certain nonfactor charges. For affiliates, the major types of income are employee compensation, profit-type return, and net interest; nonfactor charges are indirect business taxes and capital consumption allowances. The estimates presented in this article were prepared by summing these items.

Estimates of affiliate gross product are useful in measuring the size and economic impact of affiliates on the U.S. economy as a whole and on individual U.S. industries. Although sales by affiliates can also be used to measure this impact, gross product is a preferable measure for some purposes. Gross product indicates the extent to which affiliates' sales result from their own production rather than from production that originates elsewhere, whereas sales data do not distinguish between these two sources of production. In addition, gross product estimates measure the value added to the economy by affiliates in a specific time period. In contrast, sales in a given period may represent production of earlier periods, that is, out of inventory.

The gross product estimates, while useful measures of U.S. GDP attributable to firms in which there is foreign direct investment, are subject to several limitations or qualifications. Movements in affiliate gross product reflect acquisitions of existing U.S. businesses, as well as the establishment of new affiliates and changes in production by existing affiliates. Thus, an increase in affiliate gross product may not represent an increase in U.S. GDP; rather, it may simply represent a shift in the ownership or control of productive resources that would have contributed to GDP in any event.(2) Furthermore, because the estimates are in current dollars, they reflect changes in prices as well as changes in real output. Finally, it should be emphasized that not all of the factors of production that generate affiliate gross product are foreign owned. The largest share of affiliate gross product is accounted for by employee compensation, almost all of which accrues to U.S. workers, and some of the profit-type return of affiliates that are not wholly owned by foreign direct investors accrues to U.S. owners.

The remainder of this article is divided into three sections. The first reviews the growth and distribution from 1977 to 1987 of U.S. affiliate gross product by industry of affiliate, by country of ultimate beneficial owner (UBO), and by component.(3) The second compares the level, growth, and composition of affiliate gross product with those of all-U.S.-business gross product, as measured in the national income and product accounts (NIPA's). The third illustrates how gross product data, together with other data on U.S. affiliates' operations, can be used to analyze the structure of affiliates' production. A technical note at the end of the article discusses data sources, estimation procedures, and conceptual differences between the components of U.S. affiliate and NIPA gross product.

Growth and Distribution of U.S. Affiliate Gross Product, 1977-87

Overview

Gross product of U.S. affiliates grew from $35.2 billion in 1977 to $151.9 billion in 1987 (table 1). The average annual growth rate during this period was 16 percent. Affiliate gross product grew much more rapidly during 1977-81, although from a smaller base, than during 1981-87--an average annual rate of 29 percent, compared with 7 percent. The faster growth in the earlier period may have reflected several factors. First, during that period, U.S. companies were being acquired by foreigners at a rapid pace. After slowing in 1982-83, the pace and the size of acquisitions picked up again in 1984. However, after 1981, disinvestment increased, as some of the acquisitions made earlier proved unprofitable and as foreign parents sold off unwanted divisions of recently acquired affiliates.(4)

Second, growth in affiliate gross product slowed considerably in 1982 because of the worldwide economic recession. Slack demand led to sharp declines in production by existing affiliates, and slow recovery overseas limited foreigners' ability to make new investments.

Third, inflation rates in the United States were higher during 1977-81 than after 1981. (As noted earlier, the estimates are in current dollars and thus reflect price changes as well as changes in real output.)

Finally, growth in affiliate gross product may have been affected by fluctuations in the value of the dollar vis-a-vis foreign currencies. During 1977-80, depreciation of the dollar encouraged new investment in the United States by making it cheaper for foreigners to produce and invest here. When the dollar appreciated during 1981-85, these activities became relatively more expensive, and new U.S. investment may have been dampened.

By industry

The pattern of rapid growth of affiliate gross product in 1977-81, and of much slower growth in 1981-87, was widespread by industry. For example, in manufacturing--which accounted for nearly 50 percent of the affiliate total throughout 1977-87--gross product grew at an average annual rate of 30 percent in 1977-81, compared with 8 percent in 1981-87, about the same rates as those for all industries combined. In petroleum, a 29-percent growth rate was followed by a negative 2-percent rate. All other industries combined grew at a 30-percent rate in 1977-81 and a 12-percent rate in 1981-87.

In manufacturing, growth in gross product throughout 1977-87 was at an average annual rate of 16 percent. Within manufacturing, the most rapid growth was in "other manufacturing" and chemicals.(5)

In "other manufacturing," growth was particularly strong in motor vehicles and equipment. However, most U.S. affiliates of large foreign automobile manufacturers are classified in motor vehicle and equipment wholesale trade and not in motor vehicle and equipment manufacturing, because a majority of their sales result from the wholesale distribution of imported cars rather than from their sales of cars manufactured in the United States. For analytical purposes, it is useful to combine these two segments of the auto industry and examine them together. In the tables, the data for the combined industries are shown in the addenda, under the heading of "total motor vehicles and equipment."

Most of the growth in total motor vehicles and equipment occurred between 1977 and 1985. Surging demand for fuel-efficient imported vehicles induced foreign auto companies--mainly from Japan and Germany--to expand their U.S. wholesale operations. Fears of U.S. trade protectionism may have also encouraged them to produce in the United States rather than to supply U.S. markets entirely from abroad. Some increases in production from affiliates of Japanese UBO's may have resulted from Japan's institution of a voluntary export restraint program for motor vehicles in 1981. In addition, a French UBO's acquisition in 1979 of a U.S. automobile manufacturer boosted affiliate production.

In 1986-87, gross product in total motor, vehicles and equipment declined. The French UBO's automobile manufacturer proved unprofitable and was sold to a U.S. company in 1987. That same year, a German UBO closed its U.S. production facilities following several years of poor sales. In addition, gross product declined in 1986-87, when wholesalers were forced to raise prices for imported vehicles, because of dollar depreciation. These higher prices dampened demand. Although several joint ventures between Japanese and U.S. companies to produce cars in the United States were launched during 1986-87, they did not make substantial contributions to gross product in those years, because they had not become fully operational. Since 1987, most of these ventures have become operational, and their gross product has probably increased.

In chemicals, gross product rose at an average annual rate of 17 percent in 1977-87. Growth was very rapid in 1977-81; however, much of it occurred in 1981, when gross product more than doubled because a Canadian UBO acquired a minority interest in a major producer of industrial chemicals and synthetics. The rate of growth slowed in 1981-87, largely because increased affiliate production resulting from several acquisitions in 1985-86 was mostly offset by the disinvestment of a large agricultural chemicals affiliate that repurchased the minority equity interest held by its German UBO.

In petroleum, gross product grew at an average annual rate of 9 percent in 1977-87. During 1977-81, gross product of petroleum affiliates increased at the same rate as that of all affiliates. The increase mainly reflected rising crude oil prices and stepped-up production in Alaska. However, crude oil prices began to fall in 1982; in 1986 alone, they fell by one-half. As a result of the price collapse, gross product in petroleum declined in 1981-87, and these affiliates' share of total affiliate gross product fell from 22 percent in 1981 to 12 percent in 1987.(6)

In finance (except banking), gross product of affiliates grew at an average annual rate of 39 percent. These affiliates accounted for a small, but growing, share of affiliate gross product. Their faster-than-average growth mirrored the faster growth of this industry in the U.S. economy as a whole. Increased consolidation and globalization and a surge in the varieties of financial instruments available made it essential for successful competitors in this industry to have access to large amounts of capital. Foreign investors were willing to supply this capital in return for minority ownership interests.(7)

By country of UBO

Gross product of affiliates with European UBO's grew at a 14-percent average annual rate in 1977-87 (table 2). These affiliates accounted for 69 percent of total affiliate gross product in 1977, but their share fell to 60 percent by 1987, because of their slower-than-average growth over the period. Gross product of affiliates with UBO's in Africa, Asia, and Pacific had faster-than-average growth, particularly in 1981-87; thus, their share of the total increased from 9 percent to 16 percent. Although the gross product of affiliates with Canadian UBO's also grew faster than average, much of the growth occurred in 1981 and resulted from a single transaction--the previously mentioned purchase of the minority interest in a major producer of industrial chemicals.

Among affiliates with European UBO's, growth rates varied by country. Growth was relatively rapid for affiliates with UBO's in Germany, Switzerland, the United Kingdom, and "other" Europe; it was relatively slow for affiliates with UBO's in France and the Netherlands. The differences in growth rates mostly reflected differences in the distribution of gross product by industry. For example, affiliates with French UBO's were concentrated in manufacturing industries--such as paper, transportation equipment, and stone, clay, and glass--that were among those most affected by recession-related layoffs and financial losses in the early 1980's; their gross product did not exceed the 1981 level until 1986. Growth among affiliates with Netherlands UBO's was particularly slow; it partly reflected the concentration of their investment in petroleum. (As noted earlier, gross product in petroleum declined during 1981-87.) Affiliates with German and Swiss UBO's, in contrast, were concentrated in industries--such as industrial chemicals and drug manufacturing--that grew relatively quickly. Increases in gross product of affiliates with UBO's in the United Kingdom probably reflected the large number and size of acquisitions by these UBO's. In "other" Europe, much of the growth reflected the reclassification of a finance affiliate's UBO to Belgium from Kuwait in 1986.

Gross product of affiliates with UBO's in Africa, Asia, and Pacific increased every year and exhibited the fastest growth among the major areas with a substantial amount of gross product.(8) Compared with other areas, growth was strong during 1981-87. Affiliates with Japanese UBO's accounted for most of the gross product in this area. In the early 1980's, these UBO's rapidly expanded their wholesale trade operations in the United States, particularly in motor vehicles and equipment and in electrical goods. More recently, growth has mainly resulted from the acquisition of minority interests in several large finance companies and the startup or expansion of manufacturing facilities. The rapid growth of affiliates with UBO's in countries other than Japan partly reflected a number of large acquisitions by Australian UBO's and the establishment of wholesale trade operations by investors from the newly industrialized countries in Asia, particularly South Korea.

By component

The distribution of U.S. affiliate gross product by component is presented for major industries in table 3. The components whose shares of total affiliate gross product grew from 1977 to 1987 were employee compensation and capital consumption allowances. The shares of the other three components--profit-type return, net interest, and indirect business taxes--declined. As discussed below, these changes in shares may have reflected changes in the industry composition of total affiliate gross product, variations in general economic conditions, and other factors. Each factor is discussed in relation to the component it most directly affects. A given factor, however, also affects the shares of other components, because a higher (lower) share for one component necessarily means a lower (higher) share for other components.

Employee compensation.--The share of total gross product accounted for by employee compensation (EC) increased from 53 percent in 1977 to 62 percent in 1987. This increase in share partly reflected the relatively faster growth in gross product of affiliates in labor-intensive industries. For example, in 1987, EC accounted for 105 percent and 80 percent of total gross product in finance (except banking) and services, respectively(9). These industries grew much faster than the average for all industries combined in 1977-87. In contrast, the much more capital-intensive petroleum industry--which had an EC share of only 26 percent in 1987--grew more slowly than average. The increased EC share may also have reflected the increased concentration of affiliates in certain high-wage industries, such as manufacturing.

Capital consumption allowances.--The share of total gross product accounted for by capital consumption allowances (CCA)--a measure of depreciation--increased from 9 percent in 1977 to 12 percent in 1987. Most of the increase occurred after 1981 and may have reflected the availability of accelerated depreciation methods for calculating income taxes under the Economic Recovery Tax Act of 1981. Although CCA for affiliates are computed on the basis of book depreciation, rather than tax depreciation, the 1981 Act may have encouraged new investment in depreciable assets, thus yielding higher CCA for affiliates. The increased CCA share may have also reflected stepped-up investment in assets that have relatively short service lives, such as computers.

Profit-type return.--The share of gross product accounted for by profit-type return (PTR) declined from 18 percent in 1977 to 9 percent in 1987. This component is more sensitive to changes in general economic conditions than other components. Although generally trending downward, the PTR share of total gross product fluctuated considerably during 1977-87. It averaged about 15 percent in 1978-81 but declined sharply to 7 percent in 1982, when the economic recession caused profits to drop. Manufacturing affiliates--particularly those in nonelectrical machinery, transportation equipment, primary metals, and stone, clay, and glass products--suffered large losses. The profits of petroleum affiliates declined slightly, as crude oil prices began to fall from their 1981 peak.

After 1982, production and profits began to recover. By 1984, the share of gross product accounted for by PTR grew to 13 percent. After 1984, however, the PTR share declined. In each year, sharp decreases in the PTR of different industries accounted for the overall decline. In 1985, manufacturing affiliates' profits decreased. In 1986, petroleum affiliates' PTR declined because of the steep drop in crude oil prices. In 1987, profits in retail trade and finance were down; the decline in retail trade may have reflected the increased debt burden and higher interest expenses associated with leveraged buyouts of several U.S. retailers. (Retail trade was one of the few industries in which the net interest share of gross product increased from 1982 to 1987.) The decline in finance affiliates' PTR probably reflected the sharp decline in stock prices and the divestiture of several affiliates in that year.

Indirect business taxes.--The share of gross product accounted for by indirect business taxes (IBT) declined from 14 percent in 1977 to 12 percent in 1987. This decline partly reflected slower growth in two industries--food manufacturing and petroleum--in which IBT accounted for a large share of gross product. In food manufacturing, the large share mainly reflected excise taxes on alcoholic beverages; production grew slowly, partly because of shifting tastes away from distilled liquors. In petroleum, growth was slow for the reasons discussed earlier.

Net interest.--The share of gross product accounted for by net interest was roughly the same in 1977 and 1987--6 percent and 5 percent, respectively. However, it was as high as 9 percent in 1982. The increase from 1977 to 1982 probably reflected rising interest rates. Following 1982, the net interest share generally declined through 1987. The decline probably reflected falling interest rates and a slight increase in the portion of affiliate operations that was financed with funds from their foreign parent groups. (These funds tend to cost less than externally borrowed funds.) By industry, the net interest share was by far the largest in real estate, where affiliate assets tend to be heavily leveraged.

Comparison With All-U.S.-Business Gross Product

This section examines the U.S. affiliate share of all-U.S.-business gross product and how it has changed since 1977. In addition, distributions of affiliate and all-U.S.-business gross product by component are compared. Certain adjustments were made to the all-U.S.-business data, which are from the national income and product accounts (NIPA's), to make them more comparable to the U.S. affiliate data.(10) Overall, therefore, the affiliate gross product estimates are conceptually consistent with the NIPA estimates. However, it is important to note that the affiliate data are on an enterprise, or company, basis, while those for all U.S. businesses are on an establishment, or plant, basis. Thus, the two sets of data are not strictly comparable at a detailed industry level. Because the sources of data for affiliate and NIPA estimates differ, differences in timing, valuation, and industry classification could also significantly hamper detailed industry comparisons. Despite these limitations, analyses for major industries probably are not significantly affected, and comparisons of the two data sets can provide a picture of the relative shares of all-U.S.-business gross product accounted for by affiliates in the major industries.

U.S. affiliates accounted for 4.3 percent of all-U.S.-business gross product in 1987 (table 4), up from 2.3 percent in 1977. Nearly all of the increase, however, occurred during 1977-81, when growth in affiliate production mainly reflected the rapid pace of acquisitions of U.S. businesses by foreigners. From a relatively small base, affiliate gross product grew during this period at an average annual rate of 29 percent, compared with about 11 percent for all U.S. businesses; thus, the affiliate share of all-U.S.-business gross product rose. Since 1981, however, both affiliate and all-U.S.-business growth have slowed to about the same 7-percent average annual rate, and the affiliate share of all-U.S.-business gross product has remained constant.

By major industry

Despite the increase in the affiliate share of all-U.S.-business gross product since 1977, the share in 1987 remained relatively small. In four industries that accounted for over 60 percent of the all-U.S.-business gross product in 1987--retail trade, real estate, services, and "other industries"--the affiliate share ranged from only 1 percent to 3 percent.(11) In only one major industry, manufacturing, did the affiliate share exceed 10 percent.

In retail trade and services, much of the all-U.S.-business gross product is accounted for by small businesses, such as proprietorships, which usually do not attract foreign investment. In real estate, despite the widely publicized foreign investment in some expensive "trophy" properties--mainly urban office buildings--most investments by foreigners tend to be fairly small; in addition, the vast majority of U.S. commercial properties remain domestically owned. In "other industries," the low affiliate share partly reflects restrictions on foreign investment in some segments of these industries, especially in transportation, communications, and public utilities. Additionally, like retail trade and services, much of the remainder of this industry group consists of small businesses that do not attract foreign investment.

Affiliate shares in manufacturing and finance (except banking) increased sharply from 1977 to 1987--from 5.0 percent to 10.5 percent in manufacturing and from 2.2 percent to 9.4 percent in finance (except banking).(12) In manufacturing, as in all industries combined, virtually all of the increase in share occurred before 1982. Although, for the reasons stated earlier, exact comparisons of affiliate data with all-U.S.-business data are inappropriate at a detailed industry level, affiliate shares probably increased in most manufacturing subindustries. The increase appears to have been particularly large in chemicals.(13)

In chemicals, the increase in the affiliate share reflected several factors. Rather than exporting to the United States, foreigners may have preferred establishing production facilities here, partly because of the availability of raw material feedstocks, such as petroleum. In addition, foreign pharmaceutical companies may have found it easier to obtain U.S. Federal Government approval of new products by producing them here rather than abroad. Before 1977, foreign chemical manufacturers--mostly European--gained a share of U.S. production mainly by establishing operations in the United States. Since then, they have expanded their U.S. presence primarily through acquisitions of existing companies. Much of this expansion reflected a single acquisition, mentioned earlier, in 1981--that of a minority interest in a major producer of industrial chemicals and synthetics by a Canadian UBO. Since 1982, growth in the affiliate share has slowed partly because numerous acquisitions have been largely offset by the divestiture, mentioned earlier, of the minority interest in the German-owned agricultural chemicals affiliate.

In finance (except banking), most of the increase in the affiliate share of all-U.S.-business gross product resulted from the foreign acquisitions of minority interests in large U.S. finance companies mentioned earlier.

By component

In 1977, the distributions of the components of affiliate and all-U.S.-business gross product were similar and only differed significantly for employee compensation and indirect business taxes (table 5).(14) Although both the affiliate and all-U.S.-business distributions changed between 1977 and 1987, the pattern of change differed mainly for employee compensation and net interest.

The employee compensation share of affiliate gross product increased sharply--from 53 percent to 62 percent--in 1977-87, even though for all U.S. businesses, it increased only slightly, from 58 percent to 59 percent. The share increase for affiliates occurred because, compared with all U.S. businesses, affiliates have become increasingly concentrated in industries--such as manufacturing, finance (except banking), and insurance--in which compensation per employee (CPE) is higher than average, and relatively less concentrated in industries, such as services and retail trade, in which CPE is lower than average. Furthermore, affiliate CPE tends to be higher than all-U.S.-business CPE in the high-CPE industries and to be lower than all-U.S.-business CPE in the low-CPE industries. The affiliate share may also have increased because foreign investors focused their more recent acquisition efforts on large companies, which tend to pay above-average compensation. For example, in finance (except banking), most of the affiliate gross product is accounted for by major securities firms, which generally have very high levels of compensation. Moreover, foreign parents may be shifting more of their higher paid positions, such as those involving financial management and marketing, from abroad to their U.S. affiliates.

The net interest share of affiliate gross product decreased slightly--from 6 percent in 1977 to 5 percent in 1987. The share for all U.S. businesses increased from 4 percent to 6 percent. The different pattern may reflect two factors. First, between 1977 and 1987, affiliates had become relatively more concentrated than all U.S. businesses in certain industries--particularly finance (except banking) and insurance--in which the net interest share of gross product is usually very small or negative, second, although the degree of leverage has increased both for affiliates and all U.S. businesses since 1977, affiliates' interest payments may have been held down by an increase in the portion of borrowed funds that are from their foreign parent groups; these funds are often supplied at interest rates below those charged by financial intermediaries.

Structure of Affiliate Production

The estimates of U.S. affiliate gross product, together with other information on U.S. affiliates' operations, can be used to analyze how affiliates structure their production (table 6). Data on gross product, sales, and inventory change can be used to derive estimates of affiliates' purchases from outside suppliers (i.e., as sales minus gross product plus inventory change). These estimates, together with the data on sales and gross product, can in turn be used to gauge the extent to which affiliates' sales result from their own production (as measured by gross product) or from the production of others (as measured by purchases). In addition, by subtracting affiliates' imports from their total purchases, the portion of total purchases that is from U.S. businesses can be estimated. By summing affiliates' gross product and purchases in the United States, an estimate of the local (U.S.) content of U.S. affiliates' sales can be made; this estimate includes both the affiliates' own production and the production of other U.S. businesses that is used as inputs into the affiliates' production.

The remainder of this section briefly discusses some of these estimates by industry of affiliate to illustrate a few uses of the gross product data.(15) One possible extension of the analysis presented here would be to compare these data by industry to similar data for all U.S. businesses to determine whether affiliates and all U.S. businesses structure their production differently.

The extent to which affiliate sales are provided by the affiliates' own production, rather than by production originating elsewhere, is indicated by the ratio of gross product to sales.(16) This ratio indicates the degree of vertical integration of affiliates; the higher the ratio, the higher the degree of integration. For all industries, the ratio increased from 18 percent in 1977 to 21 percent in 1987. (Consequently, the portion of affiliate sales derived from the production of others declined.) The increase suggests that production in the United States may have become a somewhat more important way for foreign companies to serve the U.S. market during this period. However, the ratio has remained roughly constant at between 21 percent and 22 percent since 1983, perhaps indicating that the degree of vertical integration of affiliates has stabilized or that there have been offsetting industry mix effects.

By industry, the ratio of manufacturing affiliates, which accounted for nearly one-half of affiliate gross product, increased slightly, from 32 percent in 1977 to 33 percent in 1987. Within manufacturing, however, there were larger, mostly offsetting changes. The ratios of affiliates in chemicals and "other manufacturing" increased, while the ratios of affiliates in foods, in primary and fabricated metals, and in machinery decreased. In total motor vehicles and equipment (defined earlier as the sum of motor vehicles and equipment manufacturing and wholesale trade), the ratio increased from 6 percent to 9 percent. In wholesale trade, where affiliates mainly distribute, without adding significantly to their value, goods produced by others, the ratio increased from 5 percent to 7 percent, but it remained lower than in any other industry. Its increase may reflect the fact that some affiliates classified in wholesale trade--particularly in motor vehicles and equipment--have expanded into manufacturing and have increased the extent to which their sales resulted from their own production.

If sales by affiliates do not result from their own production, they must result from the production of others, as shown by total purchases by affiliates. This measure can be derived by subtracting affiliate gross product from affiliate sales and adding inventory change.(17) The ratio of imports to total purchases by affiliates indicates the extent to which purchases of goods and services used by the affiliate are provided by imports. For all industries, imports as a percentage of total purchases declined from 27 percent in 1977 to 24 percent in 1987; however, the decline was not continuous. From 1979 to 1983, the import content dropped steadily, mostly because the price (and volume) of imports shipped to petroleum affiliates declined sharply. In 1983-87, however, the import content rose, perhaps in response to the relatively high value of the U.S. dollar, particularly through 1985, which made it cheaper for affiliates to import. (In 1987, the import content rose slightly from 1986, although the dollar declined sharply.)

By industry, the sharp decline in petroleum affiliates' imports-to-total-purchases ratio, from 33 percent in 1977 to 16 percent in 1987, was partly offset by an increase in the ratio in wholesale trade, from 34 percent to 41 percent. The ratio for manufacturing affiliates was unchanged at 16 percent. Within manufacturing, declines in the ratios for food, machinery, and "other manufacturing" affiliates were offset by an increase in the ratio for chemical affiliates. In the total motor vehicles and equipment industry, the ratio increased from 56 percent to 64 percent. The very high, and rising, ratio in 1977-87 probably reflected the significant reliance by these affiliates on imports both of goods for resale without additional processing and of components to be used in subsequent production.(18)

Inputs to production that are not imported by affiliates are purchased domestically. By adding domestic purchases to the gross product of affiliates and by comparing the sum to affiliate sales, an estimate of the ratio of "local content" to affiliate sales can be derived.(19) Over time, this ratio usually moves inversely to the ratio of imports to total purchases. For all industries, the ratio increased slightly, from 79 percent in 1977 to 81 percent in 1987. However, the 1987 ratio reflects a decline since 1983, when local content was about 85 percent; in recent years, affiliates, like all U.S. businesses, have apparently increased their reliance on imported inputs.

By industry, a large increase in the ratio of local content to sales by petroleum affiliates and a small increase by manufacturing affiliates were partly offset by a large decline in wholesale trade and a smaller decline in retail trade. In petroleum, the ratio rose from 78 percent in 1977 to 88 percent in 1987, because of a slowdown in the use of imports as an input to production. Within manufacturing, the small increase--from 90 to 91 percent--reflected offsetting changes. Increases in foods, machinery, and "other manufacturing" were offset by declines in chemicals and in primary and fabricated metals. In the total motor vehicles and equipment industry, the ratio of local content to sales declined from 48 percent to 41 percent. However, the ratio probably increased in 1988-89, because several manufacturing joint ventures between Japanese and U.S. companies increased U.S. affiliate production during these years. In addition, some foreign parts manufacturers that previously exported goods to the United States have located production facilities here to be closer to their U.S. customers.

Technical Note

Data sources

For all years except 1980 and 1987, U.S. affiliate gross product estimates were based on universe estimates derived from sample data from BEA's Annual Survey of Foreign Direct Investment in the United States. For 1980 and 1987, the estimates were based on universe data from BEA's Benchmark Survey of Foreign Direct Investment in the United States.

Estimates of 1987 all-U.S.-business gross product were obtained from table 6.1, GNP by Industry, in the national income and product accounts (NIPA's) tables in the July 1988 SURVEY OF CURRENT BUSINESS. Estimates for 1977 and 1981 were obtained from The National Income and Product Accounts of the United States, 1929-82: Statistical Tables.(20)

Estimation

Although most of the data required to obtain affiliate gross product were collected in the BEA surveys mentioned above, several data items had to be estimated for some or all of the years. Capital gains and losses had to be estimated for 1977-79, because, for those years, data on them were not collected in the annual surveys. (Profit-type return (PTR) is measured before capital gains and losses.)

An inventory valuation adjustment (IVA) was estimated for all years and applied to affiliate PTR. The IVA is defined as the excess of the replacement cost of inventories used up over their historical cost. In the NIPA's, the IVA is calculated from information on inventory book values, accounting methods for valuing inventories, and price changes. Because this information is not available for U.S. affiliates, affiliate IVA was estimated.

Except for the benchmark survey years of 1980 and 1987, when data on monetary interest paid and received were collected, it was necessary to estimate these items in order to calculate the net interest component of gross product. In addition, for all years, it was necessary to estimate imputed interest paid and received.

Differences in U.S. affiliate and NIPA gross product components

U.S. affiliate and NIPA gross product components are compared in table 7. In general, the U.S. affiliate gross product components are conceptually consistent with the corresponding NIPA components. The net effect of the conceptual differences is about 2 percent of all-U.S.-business GDP. These differences include bad debt, business transfer payments, subsidies, and depreciation of expenditures for mining exploration, shafts, and wells.(21) In addition, both profit-type return and capital consumption allowances (CCA) reflect a conceptual difference in the measure of depreciation; however, its effects are offsetting and do not affect total gross product. NIPA estimates of CCA are, for the most part, based on Federal income tax returns; therefore, valuation of these charges reflects tax accounting practices under Internal Revenue Service regulations.(22) Affiliate depreciation charges, in contrast, are drawn from accounting records on which annual reports are based, which usually do not conform to tax regulations. [Tabular Data 1 to 7 Omitted]

(1)A U.S. affiliate is a U.S. business enterprise in which a single foreign person owns or controls, directly or indirectly, 10 percent or more of the voting securities of an incorporated business enterprise or the equivalent interest in an unincorporated business enterprise. (2)Because data on U.S. affiliates are reported to BEA on a consolidated basis, it is not possible to isolate increases in gross product due to acquisitions from increases due to other factors. When a U.S. business enterprise is acquired by an existing U.S. affiliate, data for the acquired entity are consolidated with those of the existing affiliate and cannot be separately identified. It should be noted that although the primary effect of the acquisition of an existing business enterprise is merely a shift in ownership, secondary effects on U.S. GDP may occur. For example, some or all of any funds that were brought into the United States from abroad and transferred to the previous owners may be used for investment in the United States, or the new owners may utilize resources more or less efficiently than the previous ones. Data needed to gauge such secondary effects are unavailable. (3)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. (4)The pattern of rapid growth during 1977-81 followed by slower growth from 1981-87 is also reflected in other measures of foreign direct investment in the United States. For example, sales by affiliates grew at an average annual rate of 27 percent in 1977-81 and 6 percent in 1981-87. The respective growth rates for assets were 30 percent and 15 percent; for employment, 19 percent and 5 percent; and, for the foreign direct investment position in the United States, 33 percent and 16 percent. (5)Industries in "other manufacturing" are textile products and apparel; lumber, wood, furniture, and fixtures; paper and allied products; printing and publishing; rubber and plastics products; stone; clay, and glass products; transportation equipment; instruments and related products; and manufacturing industries not elsewhere classified. (6)The acquisitions of the remaining shares of a petroleum affiliate by a Netherlands UBO in 1985 and those of a different petroleum affiliate by a British UBO in 1987 did not by themselves increase gross product. Because the data are not adjusted for percentage of foreign ownership, the gross product of these affiliates was already included in the data before the acquisitions of the remaining shares. (7)The growth of affiliate gross product in this industry would have been even larger, but a South African UBO and a Middle Eastern UBO each sold minority interests in large affiliates to U.S. buyers in 1987. (8)Affiliates with UBO's in the Middle East and the United States grew somewhat faster in 1977-87, but they accounted for only 1-3 percent of total affiliate gross product in any single year. (9)The employee compensation share in finance (except banking) could exceed 100 percent because the share of another component--net interest (paid)--was negative. (That is, interest received was larger than interest paid.) (10)Specifically, gross product originating in banks, government and government enterprises, and private households; imputed GDP of owner-occupied farm and nonfarm housing; rental income of persons; business transfer payments; subsidies; and the statistical discrepancy were excluded from the all-U.S.-business data. (11)"Other industries" consists of agriculture, forestry, and fishing; mining; construction; transportation; and communication and public utilities. (12)In this section, unlike elsewhere in this article, manufacturing includes petroleum refining and coal products, and petroleum is not shown as a separate major industry. Instead, in order to be consistent with the industry classification of the all-U.S.-business data, affiliate gross product in the various petroleum subindustries is distributed among the other major industries. Thus, in table 4, manufacturing includes petroleum refining and coal products, wholesale trade includes petroleum wholesale trade, retail trade includes gasoline service stations, and so on. (13)This statement is based upon an examination of the two measures of affiliate operations--employment and sales--that are available on an industry-of-sales basis, which approximates an establishment-based classification. The employment data were collected on this basis only for the 2 years--1980 and 1987--for which BEA conducted benchmark surveys of foreign direct investment. By either measure, chemicals had a higher initial share, a faster growth rate, and a higher share in 1987 than any other major manufacturing industry. (14)Conceptual differences between U.S. affiliate and all-U.S.-business gross product components, that is, NIPA components, are discussed in the technical note. (15)Data by country of UBO will not be presented in this section because differences among countries in the ratios shown in table 6 mainly reflect variations in the industry mix of affiliates of the UBO's in those countries. (16)Because, as mentioned earlier, affiliate sales can come out of inventory (which may have resulted from affiliate production) or production may be added to inventory, the extent to which affiliate sales are provided by the affiliates' own production is measured in table 6 by comparing gross product to the sum of sales and inventory change, rather than to sales alone. However, because inventory change tends to be very small compared to either gross product or sales, the ratio is referred to in this section as the "gross product to sales" ratio. (17)Affiliate inventory data were not available for yearend 1976; thus, it was necessary to estimate the inventory change for 1977. (18)Additional data on trade of U.S. affiliates in 1987 can be found in Foreign Direct Investment in the United States: 1987 Benchmark Survey, Preliminary Results. This publication can be ordered from the U.S. Government Printing Office, Washington, DC (GPO Stock No. 003-010-00188-7, price $5.00). (19)These estimates should be used with caution, because the calculation of "local content" is subject to several qualifications. First, merchandise imports are reported on a "shipped" basis, that is, on the basis of when, where, and to whom the goods were physically shipped. Most affiliates keep their sales data on a "charged" basis, that is, on the basis of when, where, and to whom the goods were charged. Thus, the derived data on purchases are on a "charged" basis and are not completely comparable to the import data. Second, local purchases are overstated to the extent that purchases from domestic suppliers include imports. Third, local purchases are overstated because they include purchases of services from foreigners, which were not reported separately and thus could not be subtracted from total purchases in deriving local purchases. (20)BEA is currently incorporating several improvements into its estimates of GNP by industry; revisions will be available back to 1977. However, most of the improvements relate to the constant-dollar estimates that are published in NIPA table 6.2, GNP by Industry in Constant Dollars. For additional information on these improvements, see Survey 69 (June 1989): 2. (21)A comparison of the components was made for 1983 data and is available as part of the supplementary table package discussed in the box below. (22)Two measures of depreciation, or capital consumption, are used in the NIPA's: (1) CCA and (2) CCA with capital consumption adjustment (CCAdj). In contrast to the tax-return-based CCA measure, CCA with CCAdj is based on the use of uniform service lives, straight-line depreciation, and current replacement cost. Because CCA with CCAdj is not available by industry in the NIPA's, CCA is used in the GDP estimates in tables 4 and 5.