Search

Examining detailed data for the home-country operations in Swedish multinationals during the period 1986-94, this paper shows that there are signs of very notable structural changes in the home-country operations of these corporations. It also shows that the effects vary according to economic conditions in the home country. In the 1980s, when the Swedish economy was characterized by high taxes, high inflation rates, and a tight labour market, relatively attractive jobs within the multinational corporations (MNCs) were relocated from Swedish plants to foreign affiliates. In the 1990s, by contrast, when the financial crisis had necessitated a host of micro- and macroeconomic reforms, the location decision of the MNCs were more favourable for the Swedish economy. New jobs created by the multinationals were found in activities with high productivity and wages. Thus, home-country effects of foreign direct investment seem, to a large extent, to be determined by the home countries' economic environment. Copyright 2000 by Oxford University Press.

This paper examines spillover effects of the activities of multinational firms (MNCs). Such effects are most likely to be found in host countries, where the operations of foreign multinationals may influence local firms in the MNC’s own industry as well as firms in other industries. There is no comprehensive evidence on the exact nature or magnitude of these effects, however, although it is suggested that host country spillovers vary systematically between countries and industries. In particular, the positive effects of foreign investment are likely to increase with the level of local capability and competition. The spillovers to the home countries of MNCs are often more difficult to identify, for various reasons. Earlier studies suggest that the effects are generally positive, but the increasing international division of labour within multinationals complicates the analysis. The impact on the home country is likely to depend on what activities these firms concentrate on at home.

The authors compare the relation between foreign affiliate production and parent employment in U.S. manufacturing multinationals with that in Swedish firms. U.S. multinationals appear to have allocated some of their more labor-intensive operations selling in world markets to affiliates in developing countries, reducing the labor intensity in their home production. Swedish multinationals produce relatively little in developing countries and most of that has been for sale within host countries with import-substituting trade regimes. The great majority of Swedish affiliate production is in high-income countries, the United States and Europe, and is associated with more employment, particularly blue-collar employment, in the parent companies. The small Swedish-owned production that does take place in developing countries is also associated with more white-collar employment at home. The effects on white-collar employment within the Swedish firms have grown smaller and weaker over time. Copyright 1997 by Royal Economic Society.

The debate on the role of government policies for economic performance has, in recent years, turned from discussing the choice between free markets and government intervention to asking what types of intervention are good or bad. One reason is that almost all governments, irrespective of their political orientation, have chosen to play an active role in their economy (see, for example, Bardhan, 1990). However, the definition of successful intervention is still disputed, although an important lesson from the recent experience of several Asian economies seems to be that governments should make use of market forces in their efforts to influence the direction and character of economic growth: markets and competition need to be retained to discourage wasteful use of resources and to encourage learning and technical advances.

In Chapter 13, we examined aggregated data on the technology imports of US affiliates in thirty-three host countries, and found some weak support for the hypotheses proposed by Wang and Blomstrom (1992). Our results showed that the affiliates’ technology imports were positively related to the income level of the host country and (crude proxies for) the competitive pressure in the host economy, but negatively related to the level of distortions and various host country performance requirements. In this chapter, we use more detailed data from a single host country, Mexico, to analyze how the technology imports of foreign firms are related to various industry characteristics. We are particularly interested in the hypotheses that market rivalry and availability of skilled labour may encourage the multinationals to bring more technology to their foreign operations.1

This is a new analysis of recent changes in important Japanese institutions. It addresses the origin, development, and recent adaptation of core institutions, including financial institutions, corporate governance, lifetime employment, and the amakudari system. After four decades of rapid economic growth in Japan, the 1990s saw the country enter a prolonged period of economic stagnation. Policy reforms were initially half-hearted, and businesses were slow to restructure as the global economy changed. The lagging economy has been impervious to aggressive fiscal stimulus measures and has been plagued by ongoing price deflation for years. Japan's struggle has called into question the ability of the country's economic institutions, originally designed to support factor accumulation and rapid development, to adapt to the new economic environment of the twenty-first century. This book discusses both historical and international comparisons including Meiji Japan, and recent economic and financial reforms in Korea, Scandinavia, Switzerland, and New Zealand, placing the current institutional changes in perspective. The contributors argue that, contrary to conventional wisdom that Japanese institutions have remained relatively rigid, there has been significant institutional change over the last decade.

This Paper suggests that the use of investment incentives focusing exclusively on foreign firms - although motivated in some cases from a theoretical point of view - is generally not an efficient way to raise national welfare. The main reason is that the strongest theoretical motive for financial subsidies to inward FDI – spillovers of foreign technology and skills to local industry – is not an automatic consequence of foreign investment. The potential spillover benefits are realized only if local firms have the ability and motivation to invest in absorbing foreign technologies and skills. To motivate subsidization of foreign investment, it is therefore necessary, at the same time, to support learning and investment in local firms as well.

A team of notable editors and contributors present nine papers offering a comprehensive assessment of the long economic stagnation in Japan and the long-term challenges ahead.

After a review of the literature, this paper concludes that there is potential for significant “spillover effects” from FDI into host countries. However, it identifies some limitations of this potential to do with the stock of human capital, the interest in local firms of promoting skills transfer and the competition environment. The authors suggest comparing conditions and effects between regions, particularly between East Asia and Latin America where transfer in the former has been more consistent than in the latter. They propose, further, that an analysis of the type of FDI flowing to different regions and countries could provide clues to the potential for maximising the gains to local skills accumulation. Finally, studies are needed which examine the nature of skills provided by FDI, and ways in which training institutions, business schools, for example, can complement in-service training by firms in FDI host countries ...

In this volume emanating from the National Bureau of Economic Research's programme in international economics, leading economists address recent developments in three important areas. The first section of the book focuses on international comparisons of output and prices, and includes papers that present new measures of product market integration, new methodology to infer relative factor price changes from quantitative data, and an ongoing capital stock measurement project. The next section features articles on international trade, including such significant issues as deterring child labour exploitation in developing countries, exchange rate regimes, and mapping US comparative advantage across various factors. The book concludes with research on multinational corporations and includes a discussion of the long-debated issue of whether growth of production abroad substitutes for or is complementary to production growth at home. The papers in the volume are dedicated to Robert E. Lipsey, who, for more than a half century at the NBER contributed significantly to the broad field of empirical international economics.

The existence of spillover efficiency benefits to host country economies from inward foreign direct investment (FDI) is well documented in the literature.1 The determinants of the size and scope of the spillover benefits have also been studied, but they are not as clearly and consistently documented as the existence and magnitude of the relevant externalities.

This paper examines how inward and outward foreign direct investment (FDI) have influenced the restructuring of the Japanese economy and can be expected to continue to do so in the future. We find that outward investment has helped Japanese firms to sustain foreign market shares and contributed to the restructuring of the Japanese economy away from older industries. By shifting from exporting to affiliate production, there has been a geographical reallocation of the activities of Japanese firms, particularly those of multinational manufacturing firms. However, Japanese outward FDI is still not very large relative to the Japanese economy, despite the rapid growth since the mid-1980s, and there is still scope for significant increase when compared with the levels of most other OECD countries. Inward FDI will presumably have an even stronger impact on the restructuring of the Japanese economy. Although the stock of inward foreign direct investment is still very small, there are important changes under way. Deregulation has opened up much of the industrial and service sectors to foreign multinationals.

This paper examines how inward and outward foreign direct investment (FDI) have influenced the restructuring of the Japanese economy and can be expected to continue to do so in the future. We find that outward investment has helped Japanese firms to sustain foreign market shares and contributed to the restructuring of the Japanese economy away from older industries. By shifting from exporting to affiliate production, there has been a geographical reallocation of the activities of Japanese firms, particularly those of multinational manufacturing firms. However, Japanese outward FDI is still not very large relative to the Japanese economy, despite the rapid growth since the mid-1980s, and there is still scope for significant increase when compared with the levels of most other OECD countries. Inward FDI will presumably have an even stronger impact on the restructuring of the Japanese economy. Although the stock of inward foreign direct investment is still very small, there are important changes under way. Deregulation has opened up much of the industrial and service sectors to foreign multinationals.

Within Japanese multinational firms, parent exports from Japan to a foreign region are positively related to production in that region by affiliates of that parent, given the parent's home production in Japan and the region's size and income level. This relationship is similar to that found for Swedish and U.S. multinationals in parallel studies. A Japanese parent's worldwide exports tend to be larger, relative to its output, the larger the firm's overseas production. In this respect also, Japanese firms resembled U.S. multinationals. A Japanese parent's employment, given the level of its production, tends to be higher, the greater the production abroad by the firm's foreign affiliates. Japanese firms' behavior in this respect is similar to that of Swedish firms, but contrasts with that of U.S. firms. U.S. firms appear to reduce employment at home, relative to production, by allocating labor-intensive parts of their production to affiliates in developing countries. Swedish firms seem to allocate the more capital-intensive parts of their production to their foreign affiliates, mostly in high-wage countries. We conclude that in Japanese firms and ancillary employment at home to service foreign operations outweighs any allocation of labor-intensive production to developing countries.

Within Japanese multinational firms, parent exports from Japan to a foreign region are positively related to production in that region by affiliates of that parent, given the parent's home production in Japan and the region's size and income level. This relationship is similar to that found for Swedish and U.S. multinationals in parallel studies. A Japanese parent's worldwide exports tend to be larger, relative to its output, the larger the firm's overseas production. In this respect also, Japanese firms resembled U.S. multinationals. A Japanese parent's employment, given the level of its production, tends to be higher, the greater the production abroad by the firm's foreign affiliates. Japanese firms' behavior in this respect is similar to that of Swedish firms, but contrasts with that of U.S. firms. U.S. firms appear to reduce employment at home, relative to production, by allocating labor-intensive parts of their production to affiliates in developing countries. Swedish firms seem to allocate the more capital-intensive parts of their production to their foreign affiliates, mostly in high-wage countries. We conclude that in Japanese firms and ancillary employment at home to service foreign operations outweighs any allocation of labor-intensive production to developing countries.

The predominant view in the literature on foreign direct investment is that various types of spillover may provide important benefits for the countries that host foreign multinational corporations. For example, numerous case studies have shown that the technology and productivity of local firms may improve as foreign firms enter the market and demonstrate new products and technologies, provide technical assistance to their local suppliers and customers, and train workers and managers who are later employed by local firms. There are also reports that the competitive pressure exerted by foreign affiliates has forced local firms to operate more efficiently and introduce new technologies earlier than would otherwise have been the case (see Chapter 8 for a review of the literature).

Since the 1960s the developing countries have had very different experiences regarding income and productivity growth, and the extent to which they have converged on developed countries. Some, such as the Asian newly-industrialized countries (NICs), clearly are in a process of rapid convergence, whereas others, such as most countries in Africa, show no sign of convergence. This indicates that the realization of the potentiality for productivity catch-up simply because of backwardness depends strongly on another set of causes, some of which are internal and others external to the countries themselves (see Abramovitz, 1986).

The role of technical progress as a key to economic growth is widely recognized today. New technologies can be developed domestically through investments in research and development (R&D), they can be imported in various ways from abroad, or generated through some combination of the two. Since R&D activities are normally very costly, many countries rely heavily on imports of modern technology from abroad. In most cases they must turn to multinational corporations, who have become the most important actors in the generation of technology.

Multinational firms may initiate affiliate activities abroad in two different ways; either by building a new establishment (greenfield investment) or by taking over an already existing firm (acquisition).The two methods can be expected to yield different costs and benefits for the host economy. Some argue, for example, that acquisitions have few positive effects on productive capacity, employment or market concentration, and that foreign purchases of local firms should therefore be prevented. Others are less pessimistic and point to possible long- run effects on the host economy, such as improvements in technology and management practices.

International trade has always been an important aspect of economic development and there has been an increasing emphasis on trade as a mechanism for promoting economic growth. In almost every year since the end of the 1940s, the volume of international trade has grown faster than the volume of world production and, as a result, the degree of interdependence of the world economy has increased markedly. A large share of this rapid growth of international trade has been accomplished under the control of multinational corporations, and a good proportion of the MNCs’ exports and imports consist of intra-firm or intra-corporate trade.

The fact that multinational companies do not undertake all their research and development activities at home has become a matter of recent concern, both in the home and host countries of these companies. Some empirical work has been done on the question of why MNCs decentralize their RD Lall, 1979a; Hakansson, 1980; and Hirschey and Caves, 1981).1 The study by Hakansson (1980) is one of the few that deal with the determinants of the R&D intensity of MNC affiliates, but it is confined to subsidiaries that carry out R&D, while foreign affiliates that do not are excluded.2 Furthermore, the analysis does not take into account the characteristics of the MNC to which the affiliates belong.

This paper examines the effects on technology transfer and spillovers deriving from ownership sharing of foreign multinational affiliates. More specifically, we try to answer two questions, using unpublished Indonesian micro data. First, do establishments with minority and majority ownership differ in terms of productivity levels? Second, does the degree of spillover differ with the degree of ownership in the Foreign Direct Investment (FDI)? Our results show that foreign establishments have comparable high levels of labour productivity and that domestic establishments benefit from spillovers. The degree of foreign ownership affects neither the level of labour productivity in foreign establishments, nor the degree of spillovers.

Economic theory provides two approaches to studying the effects of foreign direct investment (FDI) on host countries. One is rooted in the standard theory of international trade and dates back to MacDougall (1960).This is a partial equilibrium comparative-static approach intended to examine how marginal increments in investment from abroad are distributed. The main prediction of the model is that inflows of foreign capital — whether in the form of FDI or portfolio capital — will raise the marginal product of labor and reduce the marginal product of capital in the host country. In addition, MacDougall suggests that FDI may be connected to other potentially important benefits

This paper deals with the investment effects of regional integration agreements and discusses how such arrangements may affect inward and outward foreign direct investment (FDI) flows in the integrating region. After setting up a conceptual framework for the analysis, we provide three studies focusing on different kinds of regional integration: North-North integration (Canada joining CUSFTA), North-South integration (Mexico’s accession to NAFTA), and South-South integration (MERCOSUR). The main conclusion of the study is that the responses to an integration agreement largely depend on the environmental change brought about by the agreement and the locational advantages of the participating countries and industries. Moreover, the findings suggest that the most positive impact on FDI has occurred when regional integration agreements have coincided with domestic liberalization and macroeconomic stabilization in the member countries.

This paper deals with the investment effects of regional integration agreements and discusses how such arrangements may affect inward and outward foreign direct investment flows in the integrating region. After setting up a conceptual framework for the analysis, we provide three studies focusing on different kinds of regional integration: North-North integration (Canada joining CUSFTA), North-South integration (Mexico's accession to NAFTA), and South-South integration (MERCOSUR). The main conclusion of the study is that the responses to an integration agreement largely depend on the environmental change brought about by the agreement and the locational advantages of the participating countries and industries. Moreover, the findings suggest that the most positive impact on FDI has occurred when regional integration agreements have coincided with domestic liberalization and macroeconomic stabilization in the member countries.

Internationalized production, that is, production by multinational firms outside their home countries has increased over the last two decades, but it was still, in 1990, only about 7 percent of world output. The share was higher, at 15 percent in 'industry,' including manufacturing, trade, construction, and public utilities, but it was negligible in 'services,' which are about 60 percent of world output. Given all the attention that 'globalization' has received from scholars, international organizations, and the press, these numbers are a reminder of how large a proportion of economic activity is confined to single geographical locations and home country ownership. Internationalization of production is clearly growing in importance, but the vast majority of production is still carried out by national producers within their own borders.

Internationalized production, that is, production by multinational firms outside their home countries has increased over the last two decades, but it was still, in 1990, only about 7 percent of world output. The share was higher, at 15 percent in 'industry,' including manufacturing, trade, construction, and public utilities, but it was negligible in 'services,' which are about 60 percent of world output. Given all the attention that 'globalization' has received from scholars, international organizations, and the press, these numbers are a reminder of how large a proportion of economic activity is confined to single geographical locations and home country ownership. Internationalization of production is clearly growing in importance, but the vast majority of production is still carried out by national producers within their own borders.

This paper examines two broad issues related to foreign investment by Swedish multinational corporations (MNCs): first, the effects of outward foreign direct investment on domestic investment, exports and employment, and second, the effects on the domestic economy from the increasing division of labour between the parents and foreign affiliates of Swedish MNCs. The paper summarizes and synthesizes the existing empirical evidence on these matters (much of which has hitherto only been available in Swedish) and discusses some possible long-run effects that have received little attention in the literature.

Growth and structural transformation of the manufacturing sector in developing countries are generally considered to be the result of the expansion of the "modem" (large-scale) sector relative to the "traditional" (small-scale) sector. Examining the sources of labor productivity growth in Mexican manufacturing, however, does not provide support for such a conclusion. Although we find that labor productivity levels vary almost in direct relation to establishment size, labor productivity growth shows no systematic variation by size class. In fact, small establishments have had the same rate of labor productivity growth as larger ones, partly because of the "excise-effect" (i.e. the exiting of low-productivity, small plants). Moreover, most of the variation in labor productivity across plant class sizes is found to be due to differences in capital intensity. The variation in TFP levels across size classes tends to be small. Thus, our results remove some justification of the policy measures that favor large firms in developing countries.

Analyses of international competitiveness and comparative advantage focus on the characteristics and behaviour of countries. They generally assign the responsibility for changes in countries’ competitiveness to macroeconomic developments and for changes in comparative advantage to changes in factor abundance and factor prices, to industry productivity developments, or to economies of scale in production. There is also another strand of literature that attributes changes in competitiveness to more ‘structural’ developments, in the sense that they are more deeply imbedded and long term, and not subject to manipulation by macroeconomic policy. These include changes in the aggregate productivity of the country, its workers, and its firms relative to those of its competitors. Recent discussions of US trade problems have emphasised factors of the second type, in particular worker skills or motivation, or the innovativeness, inventiveness, management abilities, and technological capabilities of US firms, all or some of which have supposedly declined.

This paper examines whether rivalry in host country markets may force multinational films to increase the technology transfer to their foreign affiliates. Such technology flows should be interesting from the perspective of the host country and its firms, since they would increase the potential for "spillovers". Using detailed (unpublished) industry data from Mexican manufacturing industry we find that indicators for local competition are positively related to the technology imports of foreign owned affiliates. The effects appear to be strong in consumer goods industries, which suggest that foreign multinationals are especially sensitive to the local market environment when barriers to entry in the form of complex technology or high capital requirements are relatively low.

Among developing countries, there was no gross relationship between real income per capita in 1960 and subsequent growth in per capita income. However, once other significant influences, such as education, changes in labor force participation rates, inflows of foreign investment, price structures, and fixed investment ratios are taken into account, the lower the 1960 income level, the faster the income growth. This "conditional" convergence was particularly strong among the poorest half of the developing countries, contradicting the idea of a "convergence club" confined to relatively well-off countries. Inflows of direct investment were an important influence on growth rates for higher income developing countries, but not for lower income ones. For the latter group, secondary education, changes in labor force participation rates, and initial distance behind the United States were all major factors.

This paper reviews the empirical evidence on the very different conclusions that can be drawn about productivity spillovers of foreign direct investment. It explains the concept of host country spillover benefits, describes the various forms these benefits can take, both within and between industries, and summarizes the evidence regarding the relative magnitudes of the various forms of spillovers. Moreover, the paper discusses host country policy measures which can accelerate both the BC affiliates' technology imports and the diffusion of their technology in the host economies.

This paper reviews the empirical evidence on the very different conclusions that can be drawn about productivity spillovers of foreign direct investment. It explains the concept of host country spillover benefits, describes the various forms these benefits can take, both within and between industries, and summarizes the evidence regarding the relative magnitudes of the various forms of spillovers. Moreover, the paper discusses host country policy measures which can accelerate both the BC affiliates' technology imports and the diffusion of their technology in the host economies.

This paper compares U.S.-owned affiliates with other firms in developing countries with respect to the shifts in sales from home to export markets in response to the debt crisis of the early 1980s. The U.S. affiliates in heavily indebted countries increased their exports and the share of their production exported more rapidly than other firms did after 1982, while affiliates in less indebted countries did neither. However, a large part of the shift in sales by affiliates in the heavily indebted countries involved sharp reductions in local sales, often larger than the growth in exports.

This paper compares U.S.-owned affiliates with other firms in developing countries with respect to the shifts in sales from home to export markets in response to the debt crisis of the early 1980s. The U.S. affiliates in heavily indebted countries increased their exports and the share of their production exported more rapidly than other firms did after 1982, while affiliates in less indebted countries did neither. However, a large part of the shift in sales by affiliates in the heavily indebted countries involved sharp reductions in local sales, often larger than the growth in exports.

This paper summarises some of the literature on the links between FDI and the transfer and diffusion of technology. We argue that the positive effects of FDI postulated in much of the recent debate are not automatic, that the effects of FDI will vary depending on the host country's characteristics and policies, and that there is a role for economic policy in maximising the potential benefits of FDI. Many developing countries have traditionally relied on a combination of various fiscal incentives and performance and technology transfer requirements to attract foreign multinational firms and to control their operations. However, these measures may not be sufficient to generate significant knowledge spillovers if the majority of local firms employ technologies that are very different from those used by foreigners. The studies reviewed in the paper suggest two additional areas for host country policy. Firstly, policies to support local technological capability and labour skills may facilitate spillovers of technology from foreign MNCs. The reason is not only that the local industry's ability to absorb foreign technology improves, but also that a more skilled local labour force reduces the costs of intra-firm technology transfer within the MNC, which is likely to encourage affiliates to import "more" technology from their parents. Secondly, policies to ensure that the foreign affiliates operate in a competitive environment appear to be essential. Foreign MNCs that are protected by trade or entry barriers can afford to employ obsolete technologies and still generate significant profits, without generating much diffusion of valuable knowledge and skills to local firms. Foreign MNCs facing national or international competition, by contrast, must continuously adjust their operations and technologies to changing market conditions, which creates a greater potential for spillovers to local industry

The standard literature on multinationals emphasizes the importance of firm size in explaining foreign direct investment. A typical statement, for example by Caves, is that "(T)he relation between direct investment and firm size, with other factors held constant, has been established statistically" (1974, p. 280). However, descriptions of the role of firm size are often ambiguous. As pointed out by Hufbauer (1975) and Swedenborg (1979), such descriptions do not make a clear distinction between the probability or likelihood that a firm would be a foreign investor and the extent of such investment or, as expressed by Swedenborg, a firm's propensity to operate abroad, the share of its production carried on outside its home country. A study by Horst (1972), which has become the basis for most statements about the importance of firm size, dealt only with the probability that a firm would be a foreign investor, or more specifically, the probability that it would meet the criteria for being considered a multinational firm according to the definition used in the Harvard Studies, including that of having a certain number of overseas operations; cf. Vernon (197 1).

This paper examines whether the relative performance of firms within Mexican manufacturing industries varies systematically with the presence of foreign subsidiaries. It also analyzes how foreign entry influences the technological structure in host country industries. The findings suggest that (1) foreign presence in an industry is positively correlated with structural efficiency and (2) foreign entry is related to structural changes only in the "modern" part of the industries. The most important source of spillover efficiency is found to be in the competitive pressure induced by the foreign firms. I. INTRODUCTION A WIDELY held view among economists studying foreign investment is that one of the important contributions to the host country is likely to stem from external effects or spillovers. The spillovers can be of different kinds. The foreign firms may influence the productivity and growth of the domestically owned firm; they may change the nature and evolution of concentration; they may alter financing, marketing, and technological and managerial practices in the industries that they enter, et cetera. Because of great methodological difficulties in investigating these effects empirically in addition to lack of data, few empirical investigations appear in the literature. Three available econometric studies deal with the influence of foreign investment on the technical efficiency of host country firms. One focuses on Australia (Caves [1974]), one on Canada (Globerman [1979]), and one on Mexico (Blomstrom and Persson [1983]). All three find some support for the spillover benefit hypothesis, but none of the studies analyzes the nature of spillover efficiency in depth. The present paper is an attempt to identify through what mechanisms the spillover efficiency of foreign investment takes place by analyzing the effects of foreign investment on the productive efficiency of the industrial structure in Mexico. We will examine whether the relative performance of firms within an industry varies systematically with the presence of foreign subsidiaries. We will also analyze the impact of foreign investment on structural change, that is how foreign entry influences the technological structure in host country industries.

This paper examines how foreign direct investment influences market concentration in Mexican manufacturing industries. The results suggest that MNC presence has a positive influence on concentration independently of the industrial variables which are commonly thought to determine it. Several tentative explanations, including the MNCs' market conduct are offered, although no definite conclusions can be drawn. Other variables that are found to be significant indetermining market concentration are market size, economies of scale, capital intensity and advertising intensity.

The participation of US service industry firms in Latin America markets for services consists mainly of the activities of US-owned affiliates operating in Latin America and very little of direct exports of services from the United States. The important policy issues, thus, involve barriers to the establishment and operation of affiliates in host countries rather than trade barriers. The main distinctive feature of the service industries, as compared with goods industries, is the similarity in physical capital intensity and human capital intensity among parents, affiliates in developed countries, and affiliates in Latin America. The data suggest that, because many services are difficult to trade, service industry affiliates are less likely than manufacturing industry affiliates to be part of a worldwide allocation of production by multinationals to take advantage of differences in factor prices. US shares in the Latin American service sector are also found to be very small overall, and not likely to reach the levels in manufacturing or petroleum in the foreseeable future.