U.S. affiliates of foreign companies: operations in 1988.

Howenstine, Ned G.

U.S. Affiliates of Foreign Companies: Operations in 1988

Primarily as a result of acquisitions of U.S. companies by foreign investors, growth in the operations of nonbank U.S. affiliates of foreign companies accelerated in 1988. According to preliminary results of BEA's latest annual survey of foreign direct investment in the United States, affiliates' employment, assets, and sales each grew faster than in 1987.(1)

The following are the highlights for 1988:

* Employment by U.S. affiliates increased 458,000, or 14 percent, in 1988, to 3,682,000, after a 10-percent increase in 1987 (table 1). Growth rates in both years were substantially higher than in the previous 5 years, when employment grew at an average annual rate of 4 percent and the highest rate in any year was 7 percent (in 1984).

* The share of all-U.S.-business employment accounted for by affiliates was 4.1 percent in 1988, up from 3.7 percent in 1987.

* Total assets of U.S. affiliates were $1,147 billion, up $204 billion. By industry of affiliate, the largest increases were in nonbank finance ($65 billion), manufacturing ($58 billion), and insurance ($30 billion). By country of ultimate beneficial owner (UBO), affiliates with UBO's in Japan had by far the largest increase - $74 billion.(2) The next largest increase - $35 billion - was for affiliates with UBO's in the United Kingdom.

* The gross book value of U.S. affiliates' property, plant, and equipment increased $53 billion, to $406 billion. Commercial property owned by affiliates increased $10 billion, to $100 billion. Affiliates with UBO's in Japan accounted for over one-fourth of the increase in commercial property.

* Sales by U.S. affiliates increased $109 billion, to $853 billion. Sales of goods increased $83 billion; sales of services, $19 billion; and investment income, $6 billion.

* Net income of U.S. affiliates increased $4 billion, to $12 billion. A shift from capital losses to capital gains accounted for a substantial portion of the increase. Net income before capital gains or losses increased $1 billion, to $9 billion.

* U.S. merchandise exports shipped by affiliates increased $12 billion, to $60 billion, and U.S. merchandise imports shipped to affiliates increased $6 billion, to $150 billion. By industry of affiliate, substantial increases in exports occurred in wholesale trade - particularly in wholesaling of machinery and farm products - and in manufacturing - particularly in chemicals and machinery manufacturing. The increase in imports was primarily in machinery wholesale trade. Affiliates with UBO's in Japan accounted for more than one-third of the increase in exports and for more than one-half of the increase in imports.

* In 1988, affiliates with UBO's in the United Kingdom had the largest employment. Affiliates with UBO's in Japan had the largest total assets and sales.

* New estimates presented in this article indicate that nonbank majority-owned U.S. affiliates (MOUSA's) accounted for 73 percent of the total assets and 82 percent of the employment of all nonbank U.S. affiliates. MOUSA's share of the employment of all nonbank U.S. businesses was 3.4 percent (compared with 4.1 percent for all U.S. affiliates).

The 1988 estimates cover the universe of nonbank U.S. affiliates. They update similar estimates for 1977-87 from BEA's annual and benchmark surveys. The 1988 estimates were derived by combining data reported by a sample of affiliates in the annual survey with BEA estimates of data for affiliates not in the sample. The technical note describes the method used to estimate data for affiliates not in the sample.

The estimates for 1987 are revised from those published last year. In general, the revisions were small: Total assets and sales were each revised up 2 percent, and employment was revised up 1 percent.

The remainder of this article reviews changes in affiliate operations in 1988 in more detail, discusses the share of the U.S. economy accounted for by U.S. affiliates, and introduces new estimates covering majority-owned U.S. affiliates.

Employment in 1988

Although the accompanying tables present a number of key items on U.S.-affiliate operations, this section discusses affiliate operations based on only one item - employment. Employment was chosen because changes in it are not directly affected by inflation and thus tend to correspond more closely than other available items to changes in real economic activity.

Employment by affiliates increased 458,000 in 1988, to 3,682,000. Most of the increase reflected acquisitions of existing U.S. businesses by foreigners and, thus, probably does not represent an increase in total U.S. employment.

By industry

By industry of affiliate, the increase in employment in manufacturing (220,000) was by far the largest (table 2). Within manufacturing, increases were substantial in machinery and "other manufacturing." In the latter, the largest increases were in printing and publishing and in rubber products. A 123,000 increase in employment in retail trade was next largest; almost three-fourths of this increase was in general merchandise stores.

These changes in affiliate employment refer to data classified by "industry of affiliate" (see the first seven columns of table 2). For this classification, an affiliate's primary industry - that is, the industry that accounts for the largest portion of its sales - is determined, and all data are shown in that industry even if the affiliate has activities in secondary industries. This classification is also used in most of the other tables that present data by industry in this article.

For two items - sales and employment - data are also available by "industry of sales." On this basis, sales and the associated employment in secondary industries are shown in those industries rather than in the affiliate's primary industry.(3) Data classified by industry of sales are preferable to those classified by industry of affiliate for analyzing the various activities in which diversified affiliates are engaged. The last four columns of table 2 show data on affiliate employment by industry of sales.

The pattern of changes in employment by industry of sales and that by industry of affiliate may differ because affiliates often have substantial employment in secondary industries and because changes in their employment in those industries may not parallel the changes in their primary industries. Affiliates classified in petroleum, manufacturing, and wholesale trade have particularly large portions of their employment in secondary industries (table 3). Those classified in petroleum have substantial employment in manufacturing; those classified in manufacturing have substantial employment in retail and wholesale trade and in services; and those classified in wholesale trade have substantial employment in manufacturing. [Tabular Data Omitted]

The pattern of changes in employment on the two bases may also differ when an affiliate's industry classification changes. When classified by industry of affiliate, all employees are shifted from the old to the new industry. In contrast, when employment is classified by industry of sales, no such shift occurs. In the latter case, changes in employment for a given industry reflect only actual changes in the affiliate's employment in the industry, regardless of the affiliate's industry classification.

The remainder of the discussion of changes in employment by industry is in terms of data classified by industry of sales. In some cases, the larger changes noted below reflect employment in secondary industries of companies acquired by foreign investors in 1988.

By industry of sales, the largest increases in employment were in manufacturing - 188,000 - and retail trade - 127,000. In manufacturing, all major industries except chemicals had substantial increases (employment in chemicals increased moderately). The largest increase in manufacturing - 66,000 - was in "other manufacturing." Within this category, increases were substantial in both printing and publishing (19,000) and rubber products (16,000). In printing and publishing, employment was boosted by several acquisitions. In the two largest, a publication and information services company was acquired by a British company and an encyclopedia publisher by the U.S. affiliate of a French company. In rubber products, employment was boosted by the acquisition of a large U.S. tiremaker by the U.S. affiliate of a Japanese company.

In machinery manufacturing, employment increased 46,000. The increase was evenly divided between nonelectrical and electrical machinery and resulted from several acquisitions in each industry. In nonelectrical machinery, the largest acquisitions were of a computer equipment manufacturer by a Netherlands company and a manufacturer of hydraulic cranes by the U.S. affiliate of a British company. In electrical machinery, the largest acquisitions were of a manufacturer of flight simulators by a Canadian company and a manufacturer of lighting fixtures by the U.S. affiliate of a British company.

In primary and fabricated metals manufacturing, employment increased 35,000. Most of the increase was in fabricated metals and resulted from the acquisitions of a metal can maker by the U.S. affiliate of a French company and an air conditioner and furnace manufacturer by a Japanese company.

In food and kindred products, employment increased 29,000. A substantial portion of the increase reflected the acquisitions of a brewer by an Australian company, a bakery by a Netherlands company, and a maker of sweeteners by the U.S. affiliate of a British company.

In retail trade, employment increased 127,000. The increase reflected the acquisition of a major U.S. department store chain by the U.S. affiliate of a Canadian company. The acquisition of a clothing and sporting goods chain by a German company and the opening of new stores by a Belgian-owned food store chain also boosted employment.

In services, employment increased 45,000. The acquisition of a car rental agency by a Swedish firm contributed to the increase.

In insurance, employment increased 31,000, largely as a result of the acquisitions of property and casualty insurers by a Netherlands company and by the U.S. affiliate of a British company.

In "other industries," employment increased 27,000. A substantial portion of this increase - 23,000 - was in transportation and reflected the acquisition of a U.S. railroad by a Canadian company.

By country

By country of UBO, over two-thirds of the increase in employment was attributable to affiliates with UBO's in three countries - Canada (122,000), Japan (98,000), and the United Kingdom (87,000) (table 4). In each case, the increases resulted mainly from acquisitions. For Canada, the largest acquisitions were in retail trade and transportation. For Japan, the largest acquisitions were in rubber products manufacturing and finance (except banking). For the United Kingdom, they were in insurance and in printing and publishing. [Tabular Data Omitted]

At the end of 1988, affiliates with British and Canadian UBO's had by far the largest employment - 735,000 and 715,000, respectively. Employment by affiliates with Japanese UBO's - at 401,000 - was third largest, and employment by affiliates with German UBO's - at 377,000 - was fourth.

Share of the U.S. Economy

Two measures - employment and total assets - are used to gauge the share of the U.S. economy accounted for by U.S. affiliates.(4) First, the size of U.S. affiliates is compared with that of all U.S. businesses, in total and by industry, in terms of employment. The comparisons by industry use affiliate employment data classified by industry of sales, which correspond more closely to the data classified by industry of establishment used for the all-U.S.-business employment data than do the data classified by industry of affiliate. Second, affiliates' shares in manufacturing are analyzed in terms of total assets. Assets are classified by industry of enterprise for both U.S. affiliates and all U.S. businesses.

Employment

In 1988, U.S. affiliate employment accounted for 4.1 percent of the 89,583,000 employees of all nonbank U.S. businesses (table 5). The affiliate share was up from 3.7 percent in 1987. The increase reflected the strong growth in affiliate employment in 1988, which, as discussed earlier, largely reflected acquisitions of U.S. companies by foreign investors.(5) [Tabular Data Omitted]

By industry, affiliate shares of all-U.S.-business employment were highest in mining (9.7 percent) and manufacturing (8.5 percent).(6) Within manufacturing, the affiliate share was highest in chemicals and in petroleum and coal products - over 25 percent in each case.(7)

Among the major industries, the sharpest increase in the affiliate share in 1988 was in insurance. The affiliate share in this industry increased over 30 percent - from 3.9 percent to 5.2 percent. The increase mainly reflected the previously mentioned acquisitions of property and casualty insurers by a Netherlands company and by the U.S. affiliate of a British company. After insurance, the sharpest increases in affiliate shares were in transportation, retail trade, and finance (except banking).

In manufacturing, the affiliate share of employment increased from 7.7 percent to 8.5 percent. Within manufacturing, the sharpest increases in shares were in fabricated metal products - from 4.1 percent to 6.3 percent - and rubber and plastics products - from 6.9 percent to 9.4 percent.

Assets

For manufacturing as a whole, U.S. affiliates' share of the total assets of all U.S. businesses increased to 14.7 percent in 1988 from 13.2 percent in 1987 (table 6).(8) In both years, affiliates' shares of manufacturing assets were substantially higher than their shares of manufacturing employment for two main reasons. First, affiliates are more concentrated than all U.S. businesses in capital-intensive industries, such as chemicals and petroleum and coal products, that have relatively low employment-to-assets ratios. Second, much of the growth in foreign direct investment in recent years has been through acquisitions, and when a company is acquired, whether by foreign or U.S. buyers, its assets are often revalued to reflect the new, generally higher, value implicit in the acquisition price. Consequently, the portion of assets that has been recently revalued is probably higher for affiliates than for all U.S. businesses.(9) [Tabular Data Omitted]

Within manufacturing, the sharpest increases in affiliate shares of all-U.S.-business assets were in the same industries - fabricated metal products and rubber and plastics products - in which increases in affiliate shares of employment were sharpest.(10)

Majority-Owned U.S.

Affiliates

Estimates presented thus far in this article cover the entire operations of U.S. affiliates, that is, U.S. companies owned 10 percent or more by foreigners. A 10-percent-or-more-ownership interest is considered evidence that a foreign owner has a lasting interest in and a degree of influence over the management of the U.S. affiliate sufficient to constitute direct investment.

For some purposes, it may be of interest to analyze separately U.S. affiliates that are controlled by foreign investors, rather than merely influenced by them. One measure of control is majority ownership (that is, ownership of more than 50 percent of an affiliate's equity). Although control can sometimes be achieved with smaller ownership shares, particularly if the remaining shares are spread among many owners, majority ownership usually means the affiliates are unambiguously controlled by their foreign owners.

BEA has developed estimates of selected items - total assets; gross property, plant, and equipment; sales; and employment - for nonbank majority-owned U.S. affiliates (MOUSA's). These estimates, along with the percent of the nonbank affiliate totals accounted for by MOUSA's, are shown in table 7. With few exceptions, the shares accounted for by MOUSA's are high because most U.S. affiliates are majority owned. The following discusses MOUSA shares of all nonbank affiliates by industry, by area, and by country for total assets and employment; the distributions of MOUSA shares of gross property, plant, and equipment and sales are similar, but not identical, to those for assets and employment. [Tabular Data Omitted]

In 1988, MOUSA's accounted for 73 percent of the total assets and 82 percent of the employment of all nonbank U.S. affiliates. MOUSA shares of the affiliate totals were high in most industries for both items. Shares were highest - 93 percent of assets and 91 percent of employment - in wholesale trade - and lowest - 55 percent of assets and 28 percent of employment - in finance (except banking). In the latter, the shares were relatively low because foreign investors held minority interests in a few large U.S. companies in that industry.

In manufacturing, MOUSA's accounted for 80 percent of the assets and 83 percent of the employment of all U.S. affiliates. Within manufacturing, the MOUSA shares of the affiliate totals were highest - 97 percent of assets and 99 percent of employment - in food manufacturing and lowest - less than 70 percent of both items - in chemicals. The low shares in chemicals reflected the minority interest of a Canadian investor in a major U.S. chemical company.

By area, the MOUSA shares were highest - 85 percent of assets and 89 percent of employment - for affiliates with European UBO's and lowest for affiliates with UBO's in the United States.(11)

By country, the patterns for the two items differed. For total assets, the MOUSA share was highest (95 percent) for affiliates with Swiss UBO's and lowest (60 percent) for affiliates with Japanese UBO's. The share was low for Japanese UBO'S because they had minority interests in a few affiliates in nonbank finance that had particularly large assets. If nonbank finance is excluded, the MOUSA share of the assets of all affiliates with Japanese UBO's is significantly higher - over 85 percent. For employment, the MOUSA share was highest (94 percent) for affiliates with British UBO's and lowest (70 percent) for affiliates with Australian UBO's.

The previous section of this article discussed the shares of all-U.S.-business employment and assets accounted for by all nonbank U.S. affiliates - whether majority or minority owned. If only MOUSA's, rather than all affiliates, are used for the calculation of the shares, the affiliate share of all-U.S.-business employment would be 3.4 percent (compared with 4.1 percent for all affiliates), and their share of all-U.S.-business assets in manufacturing would be 9.6 percent (compared with 14.7 percent).

As noted earlier, the employment and assets of MOUSA's in some industries are significantly lower than those of all affiliates. Thus, the MOUSA share of all-U.S.-business employment or assets in these industries is significantly lower than the share for all U.S. affiliates. The industries are finance (except banking); mining; agriculture, forestry, and fishing; and transportation. Within manufacturing, MOUSA shares are significantly lower in primary metals; stone, clay, and glass; paper and allied products; and chemicals.

The estimates for MOUSA's presented here are based on data reported to BEA on a fully consolidated basis. The estimates were computed using the foreign investor's ownership share of the top U.S. company in the consolidation, even though its share of other companies included in the consolidation may be lower.

Technical Note

The universe estimates of U.S. affiliates' operations in 1988 cover nonbank U.S. affiliates that had assets, sales, or net income greater than $1 million; these were the affiliates that had to complete a 1987 benchmark survey report. The estimates for 1988 were prepared based on data reported by a sample of U.S. affiliates in BEA's annual survey of foreign direct investment in the United States. Nonbank U.S. affiliates that had assets, sales, or net income greater than $10 million had to report in the annual survey on either a long or a less detailed short form. Long forms, which collected data on all items covered by the survey, were filed by nonbank affiliates with total assets, sales, or net income greater than $20 million. Short forms, which collected data only for selected items, were filed by nonbank affiliates with total assets, sales, or net income greater than $10 million but for which all three items were $20 million or less. In terms of total assets, sales, and employment, these affiliates accounted for about 1 percent of the universe. BEA estimated items that appeared only on the long form for affiliates that reported on the short form so that financial and operating data for all affiliates could be presented in the same detail.

To adjust the sample to universe coverage, estimates were made for affiliates that were below the exemption level for reporting in the annual survey or that were required to report but either did not file or filed too late to meet BEA's publication schedule. Estimates were derived based on data reported by the affiliate in BEA surveys for earlier years or, if the affiliate was new, based on data reported in BEA's 1987 or 1988 surveys of U.S. businesses newly acquired or established by foreign direct investors. The estimated data accounts for 18 percent of the universe for total assets, 20 percent for sales, and 28 percent for employment.

Table : Table 16. - Employment and Gross Property, Plant, and Equipment of Nonbank U.S. Affiliates, by State, 1987

 Thousands of Millions of
 employees dollars
 Gross property,
 plant, and
 equipment
 Total Of which: Of
 Manufac- which
 turing(1) Total Com-
 mercial
 proper-
 ty(2)
 Total 3,224.3 1,315.4 353,278 89,919

New England:

 Connecticut 57.0 25.2 3,092 1,216
 Maine 20.6 7.1 1,549 305
 Massachusetts 92.5 32.5 5,214 2,476
 New Hampshire 18.9 7.7 736 245
 Rhode Island 10.5 6.7 605 123
 Vermont 6.0 1.6 382 42

Midwest:

Delaware 36 .8 12.0 3,432 505

District of

 Columbia 7.0 .2 1,655 1,533
 Maryland 53.8 19.1 3,124 1,221
 New Jersey 173.3 70.8 11,458 3,130
 New York 302.8 77.6 23,069 13,292
 Pennsylvania 166.9 88.0 10,898 2,054

Great Lakes:

 Illinois 168.2 65.5 12,920 3,674
 Indiana 67.4 40.7 4,183 502
 Michigan 96.3 51.7 7,640 932
 Ohio 132.9 71.7 10,622 1,925
 Wisconsin 54.9 27.4 2,803 443

Plains:

 Iowa 20.4 10.7 1,663 254
 Kansas 19.9 7.8 2,350 204
 Minnesota 40.2 16.6 4,344 1,241
 Missouri 50.2 21.5 4,233 824
 Nebraska 7.6 2.7 459 99
 North Dakota 2.7 1.0 1,295 62
 South Dakota 1.9 .9 378 (D)

Southeast:

 Alabama 35.4 21.7 4,011 163
 Arkansas 21.3 11.2 1,289 252
 Florida 124.0 30.6 9,574 5,105
 Georgia 122.4 55.8 9,059 3,092
 Kentucky 38.5 21.1 4,557 641
 Louisiana 51.3 15.8 14,292 1,320
 Mississippi 17.5 11.2 2,425 191
 North Carolina 134.1 74.8 9,727 1,509
 South Carolina 75.7 37.7 6,182 732
 Tennessee 82.2 51.9 5,604 780
 Virginia 79.7 30.5 6,808 2,029
 West Virginia 25.4 14.5 5,060 78

Southwest:

 Arizona 43.5 14.2 4,103 1,432
 New Mexico 14.1 3.0 2,751 210
 Oklahoma 27.2 5.6 5,088 676
 Texas 210.1 72.1 41,591 9,736

Rocky Mountains:

 Colorado 28.3 9.5 4,487 1,971
 Idaho 4.0 1.0 395 33
 Montana 3.7 1.2 1,684 92
 Utah 12.6 4.6 2,610 147
 Wyoming 4.3 .9 2,962 35

Far West:

 California 334.9 126.3 44,275 17,848
 Nevada 10.5 .4 1,606 441
 Oregon 21.0 8.7 1,812 581
 Washington 40.6 13.3 3,588 1,153
Alaska 7.5 2.6 18,42 (D)
Hawaii 27.3 1.0 3,474 2,848
Puerto Rico 12.8 7.3 558 43

Other U.S.

[areas.sup.3] 4.0 .2 15,019 202
[Foreign.sup.4] 1.6 .3 2,165 33

(D) Suppressed to avoid disclosure of data of individual companies

(1.) Manufacturing employees are employees on the payroll of manufacturing plants. Total affiliate manufacturing employment in this table differs from the total when classified by industry of sales, which is shown in tables 2 and 3. In this table, total manufacturing employment consists only of employees on the payroll of manufacturing plants. whereas in tables 2 and 3, it includes some nonmanufacturing employees (see footnote 3 in the text). Also, total manufacturing employment in this table includes, but in tables 2 and 3 excludes, petroleum refining employees. Affiliates' manufacturing employment in this table is defined to be consistent with data on total U.S. manufacturing employment by State to which these data might be compared.

(2.) See footnote 1 to table 8.

(3.) Consists of the Virgin Islands, Guam, American Samoa, U.S. offshore oil and gas sites, and all other outlying U.S. areas.

(4.) For employment, consists of employees of U.S. affiliates working abroad. For gross property, plant, and equipment, consists primarily of movable fixed assets temporarily located outside the United States and of any foreign assets, including mineral rights carried on the U.S. affiliates' books.

NOTE. - The estimates for 1987 are revised. Estimates for 1988 will be available in the fall; see the box on the first page of this article.

(1.) A U.S. affiliate is a U.S. business enterprise in which there is foreign direct investment - that is, in which a single foreign person owns or controls, directly or indirectly, 10 percent or more of the voting securities if the enterprise is unincorporated. An affiliate is called a U.S. affiliate to denote that it is located in the United States.

(2.) 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.

(3.) Affiliate employment classified by industry of sales should approximate that classified by industry of establishment (plant), because an affiliate that has an establishment in an industry usually also has sales in that industry. However, if one establishment of an affiliate provides all of its output to another establishment of the affiliate, the affiliate will not have sales in the industry of the first establishment. For example, if an affiliate operates both a metal mine and a metal manufacturing plant and if the entire output of the mine is used by the manufacturing plant, all of the affiliate's sales will be in metal manufacturing and none in metal mining. When the mining employees are distributed by industry of sales, they would be classified in manufacturing, even though the industry of the establishment is mining.

(4.) Another measure of the share of the U.S. economy accounted for by U.S. affiliates - U.S. affiliates' contribution to U.S. gross domestic product (GDP) - is presented in "Gross Product of U.S. Affiliates of Foreign Companies, 1977-87" in the June 1990 issue of the Survey of Current Business. According to that article, affiliates accounted for 4.1 percent of all-U.S.-business gross product in 1987. (The share was calculated using affiliate gross product estimates based on preliminary 1987 data published last year, rather than on the revised data published here. Because the revisions are small, the affiliate share of all-U.S.-business gross product in 1987 should change little from the 4.1 percent cited above.) In general, U.S. affiliates' shares of all-U.S.-business employment tend to correspond closely to their share of all-U.S.-business gross product both over time and by industry.

(5.) The data on employment by all nonbank U.S. businesses are from table 6.6B of the "National Income and Product Accounts Tables" in this issue. The total used here is equal to employment of private industries less the employment of banks and private households.

(6.) In table 5, petroleum is not shown as a separate major industry. Instead, to be consistent with the all-U.S.-business data, affiliate employment in the various petroleum subindustries is distributed among the other major industries. Thus, in table 5, manufacturing includes petroleum and coal products, wholesale trade includes petroleum wholesale trade, retail trade includes gasoline service stations, and so on.

(7.) If calculated from the data shown in table 5, the affiliate share in petroleum and coal products would be significantly higher than 25 percent. However, the share so calculated is significantly overstated because affiliate employment in this industry includes a substantial number of nonmanufacturing employees. The 25-percent share cited in the text is based on a rough adjustment to exclude these employees.

In petroleum and coal products, affiliate employment is largely accounted for by integrated petroleum companies that are involved in all phases of the petroleum industry, including the extraction and refining of crude oil and the marketing of gasoline and other petroleum products. When classified by industry of sales, all of the employment of the integrated companies in any of these activities is included under petroleum and coal products manufacturing, even though they may be engaged in nonmanufacturing activities. In contrast, in the all-U.S.-business data, which are classified by industry of establishment, the employment of integrated companies is distributed among the activities of the companies' individual establishments; thus, only employees in the companies' manufacturing establishments are included in petroleum and coal products manufacturing.

The adjustment of affiliate employment in petroleum and coal products also slightly reduces the affiliate share of U.S. manufacturing employment as a whole - to 8.4 percent in 1988. As part of the adjustment, the employees subtracted from petroleum and coal products should be added to other petroleum-related subindustries. Although it is likely that most of the employees would be added to retail trade (gasoline service stations) or mining (oil and gas extraction), information on the number of employees that should be added to each subindustry is not available.

(8.) The comparisons based on total assets cover only manufacturing because comparable data on total assets of all U.S. businesses in other industries are not available. Comparisons based on sales are also shown in table 6, but are not discussed. All-U.S.-business total assets and sales are from the Census Bureau's Quarterly Financial Report for Manufacturing, Mining, and Trade Corporations (QFR). Comparisons for mining and trade are not appropriate because the QFR data for these industries cover only corporations with assets over $25 million. The exclusion of unincorporated businesses and small corporations from the QFR mining and trade data means that a significant portion of the all-U.S.-business activity in these industries is missing.

(9.) A comparison of affiliates' share of all-U.S.-business sales with their share of all-U.S.-business assets may indicate the importance of differences in valuation. Comparisons based on sales, unlike those based on assets, are not distorted by differences in valuation because sales are generally valued at current prices.

For manufacturing as a whole, U.S. affiliates' share of all-U.S.-business sales in 1988 was 12.2 percent - lower than their 14.7-percent share of assets, but higher than their 8.5-percent share of employment. These percentages indicate that some of the difference between the asset- and employment-based shares may be due to differences in asset valuation.

(10.) Partly because of differences in valuation, the affiliate shares based on total assets were significantly higher than those based on employment in several industries - particularly stone, clay, and glass; primary metals; rubber and plastics products; printing and publishing; and fabricated metal products. In some of these industries - such as stone, clay, and glass - the higher affiliate share for assets may also reflect a higher degree of diversification for affiliates than for other U.S. businesses. In such industries, an enterprise-based classification, like that used for the comparisons of total assets, would result in higher affiliate shares than an activity-based classification, like the industry-of-sales and industry-of-establishment classifications used for the comparisons of employment, because data for secondary activities are included in the enterprise-based classification.

(11.) A U.S. affiliate with a UBO in the United States has a foreign parent (that is, a foreign persons owns at least 10 percent of it), but its ultimate owner is located in the United States. See the definition of UBO in footnote 2. Affiliates with U.S. UBO'S are included because the definition of direct investment is based on whether a U.S. company has a foreign parent rather than on the location of the UBO.