1982 benchmark survey of U.S. direct investment abroad.

Whichard, Obie G. ; Shea, Michael A.

1982 Benchmark Survey of U.S. Direct Investment Abroad

THE results of BEA's 1982 benchmark survey, or census, of U.S. direct investment abroad have just ben released. The survey covered 2,245 U.S. parent companies and their 18,339 foreign affiliates (table 1). Of the parents, 2,110 were nonbank parents of nonbank affiliates; of the affiliates, 17,213 were nonbank affiliates of nonbank parents. For such nonbank parents and affiliates, highlights of the survey are:

* Nonbank U.S. multinational companies (MNC's)--consisting of both the nonbank U.S. parent companies and their nonbank foreign affiliates--had aggregated total assets of $3,493 billion, aggregated sales of $3,284 billion, and 25 million employees. U.S. parents accounted for 78 percent of the assets, 72 percent of the sales, and 74 percent of the employment of the MNC's as a whole.

* Affiliates accounted for 27 percent of the assets, 30 percent of the sales, and 31 percent of the employment of MNC's with U.S. parents in manufacturing. For MNC's with U.S. parents in services, affiliates accounted for much smaller shares--15 percent of assets, 16 percent of sales, and 13 percent of employment.

* Affiliates had assets of $751 billion (chart 7), of which net property, plant, and equipment was $228 billion. They had sales of $936 billion and 6.6 million employees (chart 8). Employee compensation was $112 billion, and net income was $31 billion. U.S. exports shipped to the affiliates were $57 billion, and U.S. imports shipped by the affiliates were $51 billion.

* The operations of affiliates were centered in developed countries, which accounted for 68 percent of the assets, 74 percent of the sales, and 67 percent of the employment of all affiliates. Among individual countries, affiliates in Canada and the United Kingdom had the largest assets, $110 billion and $107 billion, respectively.

* By industry, assets of foreign affiliates were largest--$266 billion--in manufacturing. Within manufacturing, they were largest in chemicals ($57 billion) and transportation equipment ($56 billion).

* Of the 17,213 nonbank affiliates of nonbank parents, 14,475--or 84 percent--were majority--owned foreign affiliates (MOFA's). MOFA's accounted for 77 percent of the assets, 78 percent of the sales, and 76 percent of the employment of all affiliates of nonbank parents.

* Sales by MOFA's were $730.2 billion, of which $663.9 billion, or 91 percent, were of goods and $66.3 billion, or 9 percent, were of services. By Destination, the largest portion of total sales--65 percent--was to customers in the affiliates' countries of location. Sales to other foreign countries accounted for 24 percent, and sales to the United States for 11 percent, of the total.

* For U.S. parents of MOFA's compensation per hour of production workers in manufacturing was $14.01; for MOFA's, it was $7.27. Compensation per hour was $9.92 for affiliates in developed countries and $2.86 for affiliates in developing countries.

* Parents of MOFA's spent $37.6 billion on research and development (R&D) and employed 578,000 scientists and engineers. The MOFA's spent $3.6 billion on R&D, 9 percent of the MNC total, and employed 88,500 scientists and engieners, 13 percent of the total.

This article first describes the benchmark survey. Then, to illustrate the types and uses of the data collected in the survey, it discusses various aspects of the operations of nonbank foreign affiliates of nonbank U.S. parents and their U.S. parents.

The Benchmark Survey

The 1982 benchmark survey was a census, intended to cover the universe of U.S. direct investment abroad. U.S. direct investment abroad exists when one U.S. person owns, directly or indirectly, 10 percent or more of a foreign business enterprise. Reports covering 2,245 U.S. parent companies--both bank and nonbank--and their 18,339 foreign affiliates were filed. To reduce the reporting burden, the rules for the survey exempted foreign affiliates for which assets, sales, and net income were each less than $3 million from being reported; 15,311 affiliates although large in number, accounted for little of the universe value, so that coverage in terms of value was virtually complete. U.S. persons having direct investments abroad were required to report on the benchmark survey under the International Investment Survey Act of 1976.

Three related types of data were collected in the 1982 survey: (1) Foreign affiliate financial and operating data, (2) U.S. parent financial and operating data, and (3) direct investment position and balance of payments data. Financial and operating data include balance sheets; income statements; property, plant, and equipment; employment; employee compensation; U.S. merchandise trade; sales; technology; taxes; and, for foreign affiliates, external financial position. The direct investment position and balance of payments data cover positions and transactions between foreign affiliates and their U.S. parents; thus, they are the intersection of the financial and operating data of the foreign affiliates with those of their U.S. parents. Balance of payments data include data on capital flows between U.S. parents and their foreign affiliates, receipts of income and of royalties and license fees by U.S. parents from their foreign affiliates, and other services transactions between parents and affiliates.

The 1982 survey differed from the last benchmark survey, which covered 1977, in two important ways. First, to reduce the burden on reporters, data were collected on reporters, data were collected on a fiscal-year basis, rather than the calendar-year basis used in the past; for reporting purposes, a reporter's 1982 fiscal year was the fiscal year having an ending date in calendar year 1982. Second, the 1982 benchmark survey collected more complete and detailed information on services than previous surveys. The improvements in this area are outlined in the box, "Services Data in the Benchmark Survey."

The data presented in this article are limited to the U.S. parent and foreign affiliate financial and operating data; future articles will discuss the direct investment position and balance of payments data. The financial and operating data provide comprehensive cross-section information on the activities of foreign affiliates and their U.S. parents in 1982. In addition, they provide the beginning of a time series that will be updated by a new annual sample survey covering selected data items. Previously, financial and operating data were collected only in benchmark surveys taken at irregular intervals. Data from the new annual survey will be expanded to universe estimates; thus, published estimates for 1983 forward will be comparable to the 1982 benchmark survey data. A SURVEY article on the 1983 results of the annual survey is planned for early 1986.

Because much more detailed data were collected for nonbanks than for banks, the remainder of the article is further limited to data for nonbank foreign affiliates of nonbank U.S. parents and their U.S. parents. (Most of the information on banks required by the U.S. Government was reported to other Government agencies.) Nonbank affiliates of nonbank U.S. parents accounted for 94 percent of the number, and 97 percent of the employment, of all foreign affiliates in 1982; because assets per affiliate were much lower for nonbank affiliates than for bank affiliates, however, they accounted for a much smaller share--only 56 percent--of the assets of all affiliates (table 1). Similarly, nonbank parents accounted for 94 percent of the number and 96 percent of the employment, but for only 73 percent of the assets, of all U.S. parents. For convenience, in the remainder of this article, "MNC's," "parents," and "affiliates" refer only to nonbanks, even if the term "nonbank" is not specifically used.

The cost detailed affiliate financial and operating data in the survey were obtained for MOFa's. (A MOFA is an affiliate in which the combined direct and indirect ownership interest of all U.S. parents exceeded 50 percent.) The later sections of this article present selected data that are available only for MOFA's.

Data on All Nonbank MNC's

Domestic and foreign operations

Table 2 shows the division of total MNC assets, sales, and employment between U.S. parents and their foreign affiliates. Both parents and affiliates are classified by industry of parent, so that data for both the U.S. parent and foreign affiliates of a given MNC appear opposite the same industry.

MNC assets are shown on an aggregated basis--that is, parent and affiliate assets have simply been added together. This sum contains duplication from intercompany positions between parents and their affiliates and between affiliates of the same parent. MNC sales are also on an aggregated basis; they include intercompany sales as well as sales outside the MNC. Data needed to derive consolidated--that is, unduplicated--assets and sales of MNC's were not collected in the benchmark survey.

MNC aggregated assets were $3,493 billion, of which $751 billion, or 22 percent, were affiliate assets. Aggregated sales were $3,284 billion, of which $936 billion, or 28 percent, were affiliate sales. MNC's had 25 million employees; 7 million, or 26 percent, of the total were employed by affiliates. Thus, regardless of the measure used, roughly one-fourth of total MNC operations were located abroad.

Affiliates accounted for 27 percent of the assets, and for 31 percent of the employment, of MNC's with U.S. parents in manufacturing. These shares are slightly higher than the average for all MNC's and may reflect the tendency for manufacturers to establish operations abroad in order to realize labor and production cost advantages and to avoid tariff and other barriers to exports from the United States.

MNC's in services, in contrast, have not diversified internationally as much as those in manufacturing: Affiliates accounted for only 15 percent of the assets of these MNC's. MNC's in transportation, communication, and public utilities had the lowest share--5 percent--of their assets abroad. This low share may reflect restrictions on foreign investment in these industries in many countries. MNC's in finance (except banking) also had a low share--8 percent--of their assets abroad.

Country distribution of affiliate

operations

Affiliates in developed countries accounted for $511 billion, or 68 percent, of the assets of all affiliates (table 3). Affiliates in developing countries accounted for $223 billion, or 30 percent. The remainder was accounted for by affiliates classified in "international," that is, by affiliates that have operations spanning more than one country and that are engaged in petroleum shipping, other water transportation, or oil and gas drilling. Affiliates in developed countries accounted for similarly large proportions of sales and employment--74 percent and 67 percent, respectively.

Among individual countries, assets were largest in Canada ($110 billion) and the United Kingdom ($107 billion). In both countries, the distribution of assets among industries was similar to the global distribution: one-third in manufacturing, one-fourth in petroleum, and a very small proportion in services. The large assets in these countries may be explained by several factors, including the similarity of their cultural, political, and economic institutions to those of the United States; the large size of their economies; and, for Canada, proximity to the United States.

Japanese and German affiliates had the next largest assets, about $50 billion each. In both countries, manufacturing accounted for approximately one-half of the assets.

In developing countries, assets were larget in the Netherlands Antilles ($38 billion) and Bermuda ($26 billion). In each country, finance affiliates accounted for a large share of the total. In the Netherlands Antilles, finance affiliates had been established to borrow funds abroad and relend them to their U.S. parents; the associated claims against the parents accounted for most of the assets. Under a tax treaty between the United States and the Netherlands Antilles, interest payments to these affiliates, unlike those to foreigners in countries with which there was no such treaty, were exempt from a 30-percent U.S. withholding tax. (The tax has since been abolished.)

In Bermuda, several U.S. MNC's have established affiliates to provide financial services to the entire MNC. These affiliates may serve as financial intermediaries between different parts of the MNC (for example, by financing trade between the parent and its affiliates or among affiliats), hold equity in other affiliates, borrow funds from outside lenders that are subsequently reloaned within the MNC, or hold and invest undistributed profits of the MNC.

Affiliate assets were also large in Brazil ($24 billion), Mexico ($17 billion), and Saudi Arabia ($11 billion). In Brazil and Mexico, assets were mainly in manufacturing. In Saudi Arabia, they were concentrated in petroleum.

Industry distribution of parent and

affiliate operations

U.S. parents had $2,742 billion of assets and 18.7 million employees. Manufacturing accounted for the largest shares of both the assets and employment of all U.S. parents--$1,018 billion, or 37 percent, of the assets and 10.5 million, or 56 percent, of the employees. Within manufacturing, assets were largest--$191 billion and $179 billion, respectively--in transportation equipment and chemicals. Next to manufacturing companies, insurance companies had the largest assets, $570 billion, or 21 percent, of the total. Petroleum companies accounted for $487 billion, or 18 percent.

For foreign affiliates, as for U.S. parents, assets were largest--$266 billion--in manufacturing (table 4). Within manufacturing, they were largest in chemicals ($57 billion) and transportation equipment ($56 billion). Manufacturing affiliates, with 35 percent of the assets, had 43 percent of the property, plant, and equipment and 67 percent of the employment of all affiliates.

Petroleum affiliates had $195 billion, or 26 percent, of total affiliate assets. These assets were distributed among several petroleum subindustries--a reflection of U.S. companies' global participation in all phases of the industry. Assets were largest in crude petroleum extraction without refining. Because of the capital-intensive nature of most petroleum activities, petroleum affiliates accounted for a much larger share of the property, plant, and equipment than of the employment of all affiliates--39 percent as compared with 6 percent.

Finance affiliates had $82 billion, or 11 percent, of total affiliate assets. In wholesale trade, the only other industry group with over $50 billion of assets, affiliates had $58 billion, or 8 percent, of the all-affiliate total.

Comparison of affiliate and parent

industries

Table 5 shows assets of foreign affiliates disaggregated by industry of U.S. parent and cross-classified by industry of the affiliate itself. The distribution by industry of U.S. parent (first column of table 5) differed significantly from that by industry of the affiliate itself (first row of table 5). In particular, when disaggregated by industry of U.S. parent, assets of affiliates were more concentrated in manufacturing and petroleum, and less concentrated in finance (except banking), insurance, and real estate and in trade, than they were when disaggregated by industry of affiliate.

Affiliates classified in the same industries as their parents accounted for from 48 to 79 percent of the assets of all affiliates of U.S. parents in the six major industries shown in the table. The percentage was highest in finance (except banking), insurance, and real estate and lowest in "other industries." Generally, affiliates in industries that complemented or supported operations in their parent's industry accounted for most of the assets in industries other than that of the parent. Also, because the parents often operated in several industries, affiliates in industries other than that of the parent sometimes were in industries in which the parent operated domestically as a secondary activity.

Petroleum affiliates accounted for 73 percent of the assets of all affiliates of petroleum parents. Nearly one-half of the remaining assets were

Classification of MNC's and their

sales by industry

Most of the data collected in the benchmark survey were classified by industry of enterprise; each parent and affiliate was assigned to a single industry based on the industry in which the parent's or affiliate's sales were largest. However, because parents and affiliates may conduct operations in several industries, the diversity of their operations can be seen only if their data are disaggregated according to the various industries in which they operate. In the bechmark survey, sales were disaggregated by industry for use in classifying parents and affiliates by industry. This information can also be used to show sales disaggregated by industry of sales (as opposed to the industry of the parent or affiliate).

For most major industries, and for both U.S. parents and foreign affiliates, the distribution of sales by industry of enterprise (either parent or affiliate) was similar to that by industry of sales (table 6). The similarity was particularly great for affiliates, which tended to be more highly specialized than their parents. Higher specialization, in turn, partly reflected the lower level of consolidation permitted for affiliates than for parents. In the benchmark survey, affiliates in a given country could be consolidated only if they were in the same industry or were integral parts of the same business operation. (Consolidation across country lines was uniformly prohibited.) In contrast, U.S. parents were defined as fully consolidated domestic enterprises, and domestic subsidiaries required to be consolidated with the parents were not limited as the industry classification.

The data by industry of enterprise tend to understate the importance of operations in services, because many parents and, to a lesser extent, affiliates in goods-producing industries also have service activities. For both parents and affiliates, the understatement was particularly large in computer and data processing services, which are often provided by the manufacturers of computer and data processing equipment rather than by specialized service firms. For parents, sales in finance (excpet banking) were also much larger by industry of sales than by industry of enterprise. Parents classified in transportation equipment and retail trade that had their own facilities for extending credit to customers largely accounted for the difference.

Size of parents and affiliates

Table 7 shows, by industry, the number of U.S. parents and foreign affiliates, and their respective assets, in each of several asset size classes, together with the assets per parent or affiliate. In each of the industries shown, average assets were much higher for the parents than for the affiliates. In addition, average assets varied a good deal among industries. For parents, average assets ranged from $3.4 billion in petroleum to $0.3 billion in wholesale trade and in services. For affiliates, the range was from $151 million in transportation equipment manufacturing to $15 million in services.

For both U.S. parents and foreign affiliates, particularly the former, the size distribution of assets was highly skewed; a relatively small number of parents and affiliates accounted for a disproportionate share of assets. The 427 parents with assets of at least $1 billion, for example, accounted for only 20 percent of the number, but for almost 90 percent of the assets, of all parents. The 49 largest parents--those with assets of at least $10 billion--accounted for only 2 percent of the number, but for 47 percent of the assets, of all parents. In contrast, parents in the lowest size class, under $10 million, accounted for 11 percent of the number, but for a negligible share of the assets, of all parents.

The results for affiliates were similar. Two-thirds of the assets of affiliates were accounted for by 8 percent of the affiliates--those having assets over $100 million. Affiliates with assets of from $1 billion to just under $10 billion accounted for a negligible fraction of the number, but for 23 percent of the assets, of all affiliates. (No affiliate had assets as large as $10 billion.) In contrast, affiliates with under $10 million in assets accounted for over one-half of the number of affiliates, but for only 5 percent of the assets.

The distribution of assets was also highly skewed in most industries. For U.S. parents, in no industry did the lowest size class account for as much as 1 percent of the assets. In every industry except wholesale trade, the two highest size classes accounted for over one-half of the assets.

For affiliates, the shares of assets accounted for by the two highest size classes in which affiliates fell were genrally in the range of from 40 to nearly 100 percent. In contrast, the share accounted for by the lowest size class was usually 10 percent or less. Two exceptions to this pattern were in services and wholesale trade, where more of the assets were accounted for by affiliates in the smaller classes.

Data on Nonbank MOFA's

U.S. parents' percentage ownership of

affiliates

Although U.S. direct investment abroad is considered to exist when a U.S. person owns 10 percent or more of a foreign business enterprise, actual control of an enterprise may require that the U.S. parents' ownership exceed 50 percent. Nonbank affiliates in this group--majority-owned foreign affiliates (MOFA's)--constituted 84 percent, or 14.475, of the 17,213 nonbank affiliates of nonbank parents in 1982 (table 8). MOFA's accounted for 77 percent of the assets, 78 percent of the sales, and 76 percent of the emlpoyment of all nonbank affiliates.

MOFA's accounted for 76 percent of affiliate assets in developed countries, and for 81 percent in developing countries. MOFA's ahres of total affiliate assets varied widely among countries. Several countries, including both developed and developing countries, have restrictions on, or strong national sentiments against, majority ownership by foreign investors.

The MOFA share of assets was highest--99 percent--in the Bahamas, where most of the assets were in finance. It was lowest--16 percent--in South Korea, followed by India, Israel, Japan, and Mexico. As of 1982, several of these countries restricted foreign direct investment. In South Korea, for example, government approval of investments was required, and only a few industries, mainly those using advanced technology, were open to majority ownership by foreigners. In India, policies discouraged foreign ownership shares in excess of 40 percent and prohibited foreign investment in certain service industries and in industries considered to have been adequately developed by domestic investors.

In Mexico, new foreign direct investment was authorized up to a maximum of 49 percent of a company's capital stock; foreign acquisition of more than 25 percent of the capital stock, or 49 percent of the fixed assets, of an existing company required government approval.

In Japn, majority ownership was prohibited in certain industries, including forestry, petroleum refining, mining, and leather goods. In most industries, majority ownership by foreign investors was not specifically prohibited, but policies and regulations encouraged Japanese control. Foreign majority-owned business enterprises did not have equal access to government-subsidized R&D or financial credits, and procurement policies generally favored locally owned firms. In some cases, factors other than government policy may have influenced the decision to have only a minority interest in affiliates. For example, U.S. interests in several large minority-owned automotive affiliates may have been acquired more to transfer technology and facilitate trade than to gain control; apparently, these objectives could be accomplished with minority ownership.

Sales by MOFA's

In the benchmark survey, MOFA sales were disaggregated by destination, by affiliation of customer, and according to whether the sale was of goods or of services. The disaggregation by destination suggests that U.S. companies have established majority-owned operations abroad primarily to service local and other foreign markets (often those in the same region as, or that share membership in a common-market type of arrangement with, the country of the affiliate). Of MOFA sales of $730.2 billion, 65 percent was to customers in the affiliates' countries of location, 24 percent to customers in other foreign countries, and 11 percent to customers in the United States (tables 9-11).

Only in selected industries and countries did sales to the United States account for a sizable fraction of total affiliate sales. Among industries, sales to the United States were relatively high, around 40 percent of the total, in three natural-resource-based industries--agriculture, forestry, and fishing; crude petroleum extraction; and metal mining--and in electronic components and accessories. These sales were also relatively high in finance (except bankin), largely due to U.S. parents' interest payments to finance affiliates in the Netherlands Antilles (as noted above) and elsewhere that were established to provide services to their U.S. parents.

Among countries, sales to the United States were relatively high in several countries--including Egypt, Ecuador, Indonesia,Norway, and Trinidad-Tobago--in which the sales were largely by petroleum affiliates. Sales to the United States were also relatively high in the Netherlands Antilles and in several countries in "other Asia and Pacific," where affiliates produced electronic components and accessories.

Sales to foreign countries other than the affiliates' countries of location were relatively high for affiliates in several European countries; these sales were mainly to other European countries. Sales to "other" foreign countries were also high in several countries in which affiliates were active in petroleum and in a few Latin American countries in which the sales were largely by finance affiliates and in the form of investment income.

Approximately 91 percent, or $3.4 billion, of MOFA sales were of goods, and 9 percent, or $66.4 billion, were of services (tables 9 and 10). Sales of services tended to be more concentrated in the country of the affiliate than were sales of goods: 78 percent for services versus 64 percent for goods. In many industries, from 90 to 100 percent of services were sold locally. These high percentages probably reflected the fact that many services can be efficiency supplied only by an entity with a local presence. In some instances, the very nature of the service is local (for example, a hotel). In other instances, a service-industry affiliate may have been established locally in response to prohibitions on the provision of services by foreign entities.

In most industries, affiliates' sales were highly concentrated either in goods or in services. This patern reflects the tendency, discussed earlier, for affiliates to be highly specialized. There were, however, a few exceptions. In office and computing machine manufacturing, for example, 16 percent of sales were services, mainly computer and data processing services associated with the use and maintenance of the equipment manufactured by the affiliates. Equipment rental affiliates had 31 percent of their sales in goods; some of them may have offered customers the option of purchasing the equipment.

The distribution of sales between goods and services did not vary much among countries. Goods constituted 90 percent or more of sales in every developed country and in most developing countries (table 10). Developing countries where the proportion was lower included several oil-producing countries to which affiliates provided petroleum services, Caribbean countries in which a sizable proportion of sales wereby finance affiliates, and a few countries in "other Asia and Pacific." In "international," 100 percent of sales were of services, largely transportation.

Nearly 80 percent of MOFA sales--$570.4 billion out of $730.2 billion--were to unaffiliated customers (table 11). (An "unaffiliated" customer is a customer other than an affiliate's U.S. parent or another foreign affiliate of the parent.) Sales to unaffiliated customers accounted for almost all-94 percent--of affiliates' sales in their countries of location and for 61 percent of their sales to "other" foreign countries. In contrast, sales to unaffiliated customers accounted for only 17 percent of affiliates' sales to the United States. Thus, to a large extent, sales to the United States tended to be restricted to trade with the U.S. parent.

Compensation of production workers

in manufacturing

For MOFA's with manufacturing operations, detailed information was collected on the number of hours worked by, and compensation paid to, production workers. Compensation includes wages, salaries, contributions to pension plans, and other fringe benefits. The data for affiliates were classified by industry of MOFA. The data for U.S. parents of MOFA's were classified by industry of sales. This classification was used for parents, even though data for affiliates were classified by industry of MOFA, because parents generally had much more diversified operations than their affiliates.

In interpreting the hourly compensation figures, it should be noted that they provide a measure of labor cost per unit of time worked. Due to variations in labor productivity, they do not measure labor costs per unit of output; nor, for several reasons, do they reliably measure the purchasing power of working incomes.

For U.S. parents of MOFA's, hourly compensation of production workers was $14.01, on average (table 12). For MOFA's in developed countries, it was $9.91. Hourly compensation averaged $12.17 in Canada, $11.49 in Germany, and $8.12 in the United Kingdom. For MOFA's in developing countries, it averaged $2.68.

Affiliate compensation rates varied more widely among countries than among industries. Among countries, averae compensation per hour ranged from $13.00 in Luxembourg to $1.04 in the Philippines. Among industries, the rates ranged from $9.83 in machinery to $4.42 in electric and electronic equipment. The relatively low rate in the latter industry reflects its characteristically semiskilled assembly work.

Although U.S. direct investments in manufacturing may sometimes be motivated by lower compensation rates abroad, there is no fixed relationship between compensation rates and the number of production workers employed by MOFA's. While lower compensation rates may be associated with lower production costs, they are not invariably so, because higher rates may be associated wih higher worker productivity. Also, higher compensation rates tend to be associated with higher per capita incomes and, thus, a larger local market for the goods manufactured by affiliates.

R&D activities

Table 13 presents data on R&D expenditures and employees of MOFA's and their parents, classified by industry. Parents spent $37.6 billion on R&D and had 578,000 scientists and enginners engaged in R&D activities. MOFA's spent $3.6 billion on R&D, 9 percent of the MNC total, and had 88,500 scientists and engineers, 13 percent of the total.

Manufacturing parents and affiliates accounted for the largest shares of R&D expenditures (80 percent each) and R&D employment (87 percent for parents and 86 percent for affiliates). PArents' expenditures were largest in chemicals ($6.4 billion); affiliates' expenditures were largest in transportation equipment ($0.9 billion) and chemicals ($0.8 billion).

U.S. parnets devoted more of their sales dollar and staff to R&D than did their foreign affiliates. R&D expenditures per $1,000 of sales were $17 for parents and $5 for affiliates. The number of scientists and engineers per 1,000 employees was 32 for parents and 18 for affiliates. For both aprents and affiliates, R&D intensity was highest in manufacturing, although it varied from industry to industry within manufacturing. For U.S. parents, R&D expenditures per $1,000 of sales were highest in electronic components, computers, and drugs. For affiliates, they were highest in "other" transportation equipment and in radio, television, and communication equipment.

R&D intensity was relatively low for parents and affiliates in petroleum, possibly due more to dupplication of their sales than to a lesser commitment to R&D. (Frequently petroleum is resold several times within an MNC as it moves from affiliates engaged in extracting the crude oil to those engaged in marketing and distributing the refined product.)

The expenditures data in table 13 measure R&D for the parent's or affiliate's benefit, that is, R&D performed for the MNC, either by itself or by others on contract. Data were also collected on R&D performed by the parent or affiliate, both for itself and for others on contract. Parents performed a higher percentage of R&D for themselves, as well as relatively more R&D for others on contract, than their affiliates. Parents performed 98 percent, and affiliates 84 percent, of their R&D for themselves. Total R&D performed by parents was $55.3 billion, including $16.2 billion for the Fedeal Government and $2.1 billion for others on contract. Total R&D performed by affiliates was $3.9 billion, including $0.8 billion performed for others on contract.