|
|
|
|
Journal of International Relations and Development
Subsidiary Exports: A Conceptual Framework and Empirical Evidence
by Matija Rojec
*
Introduction
IN THE CONTEXT OF GLOBALISATION AND INTEGRATION PROCESSES WITHIN THE EUROPEAN UNION (EU), THE EXPORT ORIENTATION OF FOREIGN SUBSIDIARIES IS INCREASINGLY COMING TO THE FORE OF HOST COUNTRIES' CONSIDERATIONS RELATING TO INWARD FOREIGN DIRECT INVESTMENT (FDI). Inward FDI is an obvious vehicle for increasing a host country's export competitiveness. The objective of the article is, by reviewing the existing literature on FDI and trade and, more specifically, on foreign subsidiaries' export propensity, to identify the possible determinants of and factors relating to foreign subsidiaries' export propensity. This will be done through an analysis of the existing theoretical and empirical findings on FDI and trade and foreign subsidiaries' export propensity, including new trends towards integrated international production. In defining the determinants of and factors relating to foreign subsidiaries' export propensity, we will broadly distinguish between the following groups of variables: (i) investing firm; (ii) industry; (iii) foreign subsidiary; (iv) home country; and (v) host country variables.
The article comprises six sections. The first deals with FDI and trade relations in international trade theory and the second with export-oriented FDI in FDI theory. The third section discusses export-oriented FDI in the context of new trends in international investment, while the fourth summarises the empirical evidence on export-oriented FDI. The fifth section analyses the determinants and factors relating to the export propensity of foreign subsidiaries while the conclusion summarises the main findings of the article.
Foreign direct investment and trade relations in international trade theory
THE LINK BETWEEN TRADE AND FDI WAS FIRST RECOGNISED BY MUNDELL (1957) WHO ARGUED THAT (UNDER CERTAIN RESTRICTIVE ASSUMPTIONS) THE FREE MOVEMENT OF CAPITAL (AND LABOUR) WOULD PRODUCE THE SAME RESULTS AS THE FREE MOVEMENT OF GOODS AND SERVICES, I.E. THAT FREE MOVEMENT OF THE FACTORS OF PRODUCTION IS A SUBSTITUTE FOR TRADE. Mundell's capital movements were, however, far removed from the FDI type where ownership-specific advantages are more significant than the transfer of capital as such. FDI and activities of multinational enterprises (MNEs) were clearly brought into international trade theory with Vernon's (1966) product life cycle theory. Here, FDI and trade are no longer clear-cut substitutes, as for Mundell. The character of FDI-trade relations depends on the phase of the cycle; the closer to the end of the cycle we come the more FDI and trade become complements. Vernon's presumption is that MNEs exploit cross-country differences in factor costs or technologies and in thereby contribute to the international specialisation of production and, consequently, intensify international trade. For the new theories of trade (Krugman 1983; Helpman and Krugman 1985), which introduced product differentiation and economies of scale into the issue of MNEs in international trade, the substitutive or complementary character of FDI-trade relations is not a priori defined but depends on a number of additional presumptions (vertical or horizontal FDI, income level of a host country, type of intra-firm transactions, etc.).
Export-Oriented Foreign Direct Investment: A Theoretical Perspective
THE DISCUSSION OF FDI-TRADE RELATIONS HAS OBVIOUSLY DEVELOPED TO THE STAGE WHERE THERE IS NO LONGER ANY SIMPLE PRO-TRADE (FDI AS A COMPLEMENT TO TRADE) OR ANTI-TRADE (FDI AS A SUBSTITUTE TO TRADE) FDI HYPOTHESIS. There is increasing evidence, both theoretical and empirical, that FDI and trade can be either complements or substitutes. Because MNEs are likely to affect the allocation of value-added activity both within a country and between countries, it is to be expected that MNEs and their affiliates will have a distinct impact on the structure of trade of both home and host countries whenever they operate in sectors with different trade propensities from their single-country or indigenous competitors. On balance, the figures suggest that FDI tends to concentrate in trade-intensive sectors and to engage in more trade in those sectors than do single-country or indigenous firms (Dunning 1993:401-4). The difference in both groups of firms is most pronounced in open and smaller industrialised (or industrialising) economies. While foreign firms do play a positive role in assisting the restructuring of the production of developing countries towards tradable commodities, there are important regional and industrial differences. The impact of MNEs' activities on the trade structure of home and host countries 'will depend on the interaction between the configuration of ownership, location and internalisation (OLI) advantages facing firms and the environment, system and policies (ESP) configuration facing countries'1 1 (Dunning 1993:385-6).
Two streams of FDI theory are especially relevant to the analysis of export-oriented FDI and export propensity of foreign subsidiaries. The first are the so-called developmental approaches to FDI, 2 which are particularly appropriate to the situation where the home and host country are at different levels of development. The second stream involves the analysis of motives for foreign production and the main types of foreign production which are the main determinants of FDI export orientation and the export propensity of foreign subsidiaries.
Developmental Approach
The first developmental approach relevant to export-oriented FDI is Vernon's (1966) product life cycle theory. Here, the market orientation of FDI depends on the phase of the product cycle the investing firm's product is in. In the early phase, FDI will be local market oriented. In a later stage, when a product becomes standardised and mass production prevails, cost considerations in the context of increased competition will pressure an MNE to relocate its production to less-advanced countries with comparative advantages in manual labour. Thus, FDI in the later phase of the product cycle will be export oriented, motivated predominantly by cheap labour considerations. Compared to Vernon's product cycle model, later approaches with a focus on trade replacement have gone much further. Today we know that final-product trade between home and host countries may be displaced, but there may be new trade in intermediate products and in other complementary products (see Cantwell and Bellak in this special issue).
One kind of application of the product life cycle theory at a national level was made by Ozawa (1979; 1992b) who described Japanese investment in Asia as a result of industrial upgrading in Japan itself. As the Japanese economy has gradually advanced towards skill-intensive sectors, labour-intensive production processes and environmentally sensitive industries have moved to South-east Asia. The shortage of unskilled labour in Japan has led to investment in labour-intensive production in the economies of Southeast Asia where labour costs were low (Meyer 1998:81).
Later on, Ozawa introduced the notion of the comparative-advantage-augmenting type of FDI. In the framework of outward-looking, export-based development in a host country, MNEs which have a global perspective will structure their FDI in each particular country in accordance with the pattern of relative factor endowments of that country. This kind of FDI, motivated by factor cost differentials and the production of exports, helps the host country to develop its comparative advantages (Ozawa 1992a).
Another developmental approach to FDI is the structural change FDI introduced by Meyer (1995). Because of cost pressure, enterprises in industries threatened by a loss of competitiveness are faced with the need to restructure. One possible way of restructuring is to move production facilities abroad. 3 By doing this, enterprises continue to utilise their existing industry-specific assets, but swap the expensive home country labour force with the cheaper one in a host country (Meyer 1995; 1998). As in the case of other developmental approaches to FDI, structural change FDI is centred around factor cost differentials, i.e. around export-oriented FDI.
Different variations of the developmental approach to FDI share two main characteristics. The first is the conclusion that export-oriented FDI basically happens when an investing firm (country) begins to lose its comparative advantage or competitive edge in a particular area of production. Here, production will be relocated to locations with a comparative advantage in that particular production. This will, as a rule, involve a simple export-platform type of FDI motivated by cheap unskilled labour (and partly also by the relocation of pollution-intensive activities). The second characteristic of the developmental approaches to FDI is that they devote almost all, if not all, their attention to export-oriented FDI. It seems widely accepted that it is the export-oriented FDI that is the most important for host countries. This perception is also reflected in host countries' FDI policies, where export-oriented FDI, in principle, is given special attention. The reasons are obvious. It is export-oriented FDI which is bound to make a positive contribution to the restructuring of host economies along the lines of their comparative advantages and, thus, to upgrading their export competitiveness. It is export-oriented FDI catering for larger international markets which is more growth inducing.
The Main Types of Foreign Direct Investment
From a host country development point of view, not all types of export-oriented FDI are the same. Simple export-platform type of FDI motivated mostly by cheap unskilled labour, efficiency-seeking FDI, or internationally (or even globally) integrated production offer different opportunities for a host country's development. This involves an analysis of the motives of foreign production and of the main types of foreign production. The motives for and types of foreign production are the main determinants of FDI export orientation and also of various types of export-oriented FDI. What are the main types of FDI from the point of view of foreign subsidiaries' export orientation, and how does the label of export-oriented FDI fit into various categorisations of FDI? The two most widely adopted categorisations are those of business management literature (Behrman 1972; Porter 1986; Dunning 1993) and new international trade theory (Caves 1971; Markusen 1995; Lankes and Venables 1996a).
In the business management literature, the most commonly used categorisation is that developed by Dunning, who distinguishes between natural resource-seeking, market-seeking, efficiency-seeking and strategic asset-seeking FDI (strategies). Of the four FDI types, the market-seeking one should be regarded as a "non-export" type, although the motivation of an investor is not necessarily to supply goods and services only to a host country market but also to the markets of adjacent countries in a region. Resource-seeking FDI is typical export-oriented FDI. There are three main subtypes of resource-seeking FDI, i.e. the one seeking physical resources motivated by cost minimisation and supply security, the one seeking a supply of cheap and motivated skilled or semi-skilled labour to produce labour-intensive intermediate or final products for exports, and the one prompted by the need of firms to acquire technological capability, management or marketing expertise and organisational skills. Efficiency-seeking FDI is another type of export-oriented FDI. The motivation for efficiency-seeking FDI is to gain from the common governance of geographically dispersed activities (economies of scale and scope and of risk diversification). The intention is to take advantage of different factor endowments (factor cost differences), cultures, institutional arrangements, economic systems and policies, and market structures by concentrating production in a limited number of locations to supply multiple markets. Efficiency-seeking FDI is of two main kinds. The first takes advantage of differences in the availability and cost of traditional factor endowments in various countries. The second takes place in countries with broadly similar economic structures and income levels and is designed to take advantage of economies of scale and scope, and of differences in consumer tastes and supply capabilities. The strategic asset-seeking FDI, where MNEs engage in FDI to promote their long-term strategic objectives, is also predominantly export-oriented FDI. Like the efficiency-seeking MNE, the strategic asset acquirer aims to capitalise on the benefits of the common ownership of diversified activities and capabilities, or of similar activities and capabilities in diverse economic and potential environments (Dunning 1993:56-63).
New international trade theorists distinguish between horizontal and vertical FDI, based on the distinction between market access and factor cost motivation. This distinction is reflected in the organisation of foreign production and has important implications for foreign subsidiaries' export orientation. Horizontal FDI consists of the duplication of the entire production process, except for the headquarters' activities, in several countries. Each foreign affiliate produces primarily for the local market, and there is little trade between units in separate countries. Vertical FDI consists of the geographical separation of different stages of the value-added chain with forward and backward integration. It involves international specialisation since production facilities in subsidiaries supplement rather than replicate those in the parent company. This stimulates cross-border (intra-firm) trade and results in a relatively high propensity of subsidiaries to export (Caves 1971; 1982; Andersson and Fredriksson 1996:250-2; UNCTAD 1996:123-5).
A number of authors have made attempts to reclassify the above types of FDI according to their export orientation and, thus, to identify types which are directly and explicitly related to export orientation. Thus, Narula (1996) distinguishes between trade-substituting FDI, if it goes to import-substituting activities that supply the domestic market; trade-promoting, if its aim is to supply other markets; trade-complementing, if it is directed towards rationalised production and back-up facilities in export markets; or trade-diverting, if it aims to exploit unfilled quotas under preferential trade agreements. Meyer (1998) classifies foreign subsidiaries into market-oriented aimed at local or regional markets, factor price-oriented aimed at intra-firm and extra-regional sales, and a combination of both, i.e. market and factor price-oriented ones. Lankes and Venables (1996a) differentiate between local or regional suppliers, exporters from the region and distributor firms with small number of employees carrying out import distribution. In export-oriented operations, Papanastassiou and Pearce (1992) distinguish between rationalised product subsidiaries and world (or regional) product mandate subsidiaries. The role of rationalised product subsidiaries is to produce a product or component/part, or to carry out a particular stage of a vertically integrated production process (standardised products and production methods; limited development implications). World (or regional) product mandate subsidiaries are operations that utilise a much wider range of the potentially available distinctive characteristics of host countries in a much more intensely creative way. They allow a much more entrepreneurial scope for management to work with local marketing, engineering and technological personnel to develop distinctive and innovative new products that can be sold throughout a wide part of the MNE group's world-wide markets (better development implications for a host country). According to Michalet (1997), there are two prime strategies that motivate most FDI. In most cases, a firm will decide to set up facilities abroad either so that it can serve the local market directly (market-seeking), or because the particular country possesses the right combination of production factors needed to manufacture a certain product. In this latter case (sourcing), production may be primarily destined for exports. Éltet_ and Sass (1998) distinguish between export-oriented firms, which they further divide into assembler export-oriented firms and domestically-based export-oriented firms, and domestically-oriented or non export-oriented firms which include those that export to adjacent markets.
The above survey of literature reveals not only the distinction between local/regional market and export-oriented FDI but also demonstrates that there are various kinds of export-oriented FDI. The latter is of great significance from a host country's development point of view. Here, the main distinction seems to be the one between simple assembly, export platform, more or less enclave type (for instance, Dunning's resource-seeking FDI motivated by cost minimisation of physical resources, or cheap unskilled and semi-skilled labour, and efficiency-seeking FDI motivated by cost differences in traditional factors of production; Papanastassiou's and Pearce's rationalised product subsidiaries; and Éltet_'s and Sass' assembler export-oriented firms) and integrated international (global) production type (for instance, Dunning's resource-seeking FDI prompted by the need of firms to acquire technological capability, management or marketing expertise and organisational skills, efficiency-seeking FDI designed to take advantage of economies of scale and scope, and of differences in consumer tastes and supply capabilities; Papanastassiou's and Pearce's regional or world product mandate subsidiaries, and Éltet_'s and Sass' domestically-based export-oriented firms). In spite of the different terms used for both types of subsidiaries, the contents are very similar and may best be described by Papanastassiou and Pearce (1992:2-3), where rationalised product subsidiaries produce standardised products by standardised production methods with limited development implications for host countries, while world (regional) product mandate subsidiaries utilise a much wider range of the potentially available distinctive characteristics of host countries in a much more deeply creative way and, therefore, have better development implications for host countries.
To conclude, as far as the character of FDI-trade relations is concerned, theory basically distinguishes between two types of FDI/investing firm strategy/subsidiary position. The first is market-seeking (horizontal) FDI with stand-alone subsidiaries, duplicating the production process of headquarters and established for the procurement of local or/and adjacent regional markets. The second is export-oriented (natural resource-seeking, efficiency-seeking, strategic asset-seeking, vertical, sourcing, factor cost differences-seeking) FDI with vertically integrated subsidiaries, which is determined by differentials in factor endowments, usually among countries at various stages of development (assembly, rationalised product subsidiaries), or by the advantage of economies of scale and scope, and of differences in consumer tastes and supply capabilities (integrated international production, regional or world product mandate subsidiaries), usually in countries with broadly similar economic structures and income levels.
Export-Oriented Foreign Direct Investment and New Trends in International Investment
THERE ARE A NUMBER OF PROCESSES DEMONSTRATING THAT EXPORT-ORIENTED FDI IS GAINING IN IMPORTANCE. These interlinked processes are seen in emerging integrated international production, corresponding changes in organisational structures of MNEs and in the nature and role of subsidiaries, and changes in the policy framework for trade and FDI characterised by overall liberalisation.
FDI is becoming increasingly associated with integrated international production within the multinational corporate systems of firms pursuing complex integration strategies, under which firms disperse their activities regionally or globally across production sites from which to serve regional or international markets. These trends reduce the importance of classical market-seeking FDI and increasingly lead to the servicing of regional/international markets based on integrated international production, which means more export-oriented FDI due to: (i) intra-firm trade arising from the international division of labour in the framework of integrated production system, 4 (ii) servicing of the regional/international market from one point of final production (UNCTAD 1996:74). 5 Integrated international production strategies of MNEs are characterised by a closer interconnection between FDI and trade. Because of the extension of specialisation within MNE systems, the scope for functional and product specialisation among countries is increased, and there are more opportunities for trade in accordance with comparative advantage. At the same time, these new interrelationships increase the scope for altering and upgrading the comparative advantage of countries, particularly of developing countries, 6 because production capacity can increasingly be located wherever the necessary capabilities are found (UNCTAD 1996:120).
Trends towards integrated international production have also brought about important changes in the organisational structure of MNEs and in the nature and role of subsidiaries. A predominant notion of headquarters as a node to which foreign activities are connected (collection of stand-alone affiliates) is not one characteristic of integrated international production. Instead, what appears to be more frequent is that where, in the framework of complex international strategies, headquarters and affiliates are integrated into international production systems (organisation of activities between multiple countries) which transform global inputs into outputs for global markets (Andersson and Fredriksson 1996:249-250; UNCTAD 1996:97-8). 7
Changes in the policy framework have been of crucial importance for more export-oriented integrated international production. The ability and freedom of firms to exercise choice with respect to the location of their production and, hence, the investment and trade they undertake, has increased significantly in recent years (UNCTAD 1996). The major policy changes have been liberalisation of FDI and trade policies, enabling MNEs to consider the advantages offered by a growing number of sites. This has been possible only due to the pronounced general orientation towards outward-looking, export-oriented development concepts which are a necessary framework for any FDI and trade liberalisation (Agosin and Prieto 1993; Islam 1995). Another important trend in the international policy framework that is very conducive to export-oriented FDI and the creation of integrated international production systems is the increasing importance of regional economic integrations.
Trends towards more integrated international production and, thus, export-oriented FDI will strengthen further in the future. More precisely, economic liberalisation will increase offshore investment and competitive pressures will lead to increased overseas sourcing. Global strategies will have a dramatic impact on the organisation of firms. The key to success will be the leverage of corporate capabilities around the world in support of operations in each country so that the company as a whole is larger than the sum of its parts. This requires a series of organisational initiatives in order to develop a global strategy to its full potential and to secure effective implementation of such strategies at both the national and cross-border levels. The central control of strategic elements will allow for the leveraging of capabilities and company-wide use of best practices as a result of learning (Czinkota and Ronkainen 1997:839-40).
Summary of Empirical Evidence on Export-Oriented Foreign Direct Investment
WHAT ARE THE MAIN FINDINGS OF EMPIRICAL STUDIES ON FDI-TRADE RELATIONS, AND ESPECIALLY ON THE EXPORT ORIENTATION OF FDI? As Dunning would put it, the relations depend on the configuration of OLI (ownership, location and internalisation) considerations in firms and ESP conditions in countries.
A priori, there is no presumption whatsoever that MNEs as a group of firms will affect the extent or direction of trade in one way or another. De facto, however, the evidence strongly suggests that the combination of the actions of MNEs and the influence of governments on these actions has, over the past 20 to 30 years, both increased the level and restructured the composition of world trade to the general benefit of the participating countries (Dunning 1993:407-8).
There are at least three levels of empirical analysis of FDI and trade and/or FDI export orientation. The first level is an empirical analysis of the motivation and strategies of foreign investors. This empirical analysis does not leave much room for doubt. Tapping the local/regional market is still by far the most important motive of foreign investors (Root 1990; Michalet 1997; Kearney 1998; for countries of Central and Eastern Europe see Gatling 1993; EBRD 1994; Lankes and Venables 1996a; Éltet_ and Sass 1998; Meyer 1998). Lankes and Venables (1996a) further demonstrated that the export performance of foreign subsidiaries is decisively determined by the basic motivation for FDI. The obvious and strong domination of the market-seeking motive is, however, coupled with the increasing importance and frequency of export-oriented FDI (UNCTAD 1996). It seems that MNEs are less interested in either big markets without conditions for world competitive production, or in pure export platforms, than they used to be. Instead, they are ever more seeking sites that offer both market access and conditions for world competitive production FDI (Michalet 1997:4-11).
The second level analyses the impact of FDI on home and predominantly host country exports. In spite of the domination of the market-seeking motive, it is obvious that FDI makes a positive contribution to international trade. In other words, FDI and international trade are, in general, mutually supportive and, together, they play the central role in the ongoing integration of the world economy (WTO 1996:73). MNEs and their subsidiaries generally enjoy a larger share of home or host country exports and imports than they do of output (Dunning 1993:386). Outward FDI in most analyses proves to have a positive impact on home country exports (Dunning 1993:397-8; WTO 1996:53-4). Earlier studies on the export performance of foreign subsidiaries (in developing countries) (Reuber et al. 1973; Lall and Streeten 1977) have not been very optimistic, but there is mounting evidence that foreign subsidiaries play a significant role in promoting exports from host countries (Blomstroem 1990; UNCTC 1992; Islam 1995; Andersson and Fredriksson 1996). Generally speaking, the available empirical evidence suggests that FDI and host country exports are complementary, and that a weaker yet still positive relationship exists between FDI and host country imports. In other words, on average, FDI seems to have a bigger positive impact on host country exports than on host country imports (WTO 1996:54-5). Data for countries of Central and Eastern Europe (CEE) strongly suggest the positive impact of inward FDI on their exports (Éltet_ 1998; WIIW 1998; Rojec 2000).
Finally, the third level of empirical analysis deals with the export propensity of foreign subsidiaries versus indigenous firms. Here, the findings are often diverse. Lall and Streeten (1977) claimed that the export propensity of foreign subsidiaries in developing countries is lower than that of indigenous firms. Koo (1993) found the same situation in Korea. It seems that the import-substitution development concept importantly helps to determine such a situation. Two comprehensive reviews of studies dealing with the comparison of the export propensity of foreign subsidiaries and indigenous firms have been done by Kumar and Dunning. Most of the empirical studies reviewed by Kumar (1987:370-1) demonstrate the relatively poor performance of foreign firms compared to their local counterparts or, at best, no significant difference in the export performance between the two groups of firms was found. Even among export-oriented firms, foreign firms frequently exhibit poorer export performance than domestic ones. Taken as a whole, studies from Dunning's (1993:404-7) review reveal that foreign subsidiaries generally have a higher propensity to export than do indigenous firms, but that this is not as great or as widespread as some commentators have suggested once one normalises for firm-specific and industry-specific characteristics (differences in the industrial distribution of foreign subsidiaries and indigenous firms). Comparisons of the export propensity of foreign and domestic firms in CEE countries do not leave much space for doubt. Foreign firms are much more export-oriented than domestic ones (UNCTAD 1995:14; Hooley at al. 1996; Lankes and Venables 1996a; WIIW 1998). According to Éltet_ (1998), this is not only due to the tendency of FDI to go into export-intensive sectors, but even much more because of the superior export propensity of foreign firms within individual industries.
Determinants of and Factors Relating to the Export Propensity of Foreign Subsidiaries
THE AIM OF THIS SECTION IS TO IDENTIFY THE POSSIBLE DETERMINANTS OF AND FACTORS RELATING TO THE EXPORT PROPENSITY OF FOREIGN SUBSIDIARIES. In spite of the many studies on the subject, those tackling the issue in a more systematic, model-based way, are rare: (i) Andersson and Fredriksson (1996) seek to explain variations in the export propensity of subsidiaries of Swedish MNEs in manufacturing through differences in organisational forms of international production, where the main distinction is between horizontal and vertical integration; (ii) Lankes and Venables (1996a) who analysed FDI in CEE countries distinguish between local supply, export-supply and distribution FDI projects and found a number of characteristics related to the export propensity of foreign subsidiaries; (iii) a study on the impact of inward FDI on Brazilian trading patterns (UNCTAD 1983) found that the effect of foreign ownership on exports was positive but insignificant; and (iv) Kumar (1990) who found that there are no significant differences in the role of capital, skill intensity, product differentiation, competitive structure and size of firm in explaining the export performance of the two groups of foreign and domestic firms in India.
The most commonly used and obvious variables to be tested as far as export propensity is concerned are the type of ownership, i.e. foreign versus domestic, foreign investors motivation, i.e. market-seeking versus factor cost differences and/or efficiency-seeking FDI, and the degree of multinationality. Other variables are rarely taken into account. Most analyses try to normalise for sectoral differences between foreign and domestic firms, but they rarely do this systematically. Dunning (1993:406) suggests that at least the following variables should also be included: (i) industry characteristics; (ii) the psychical distance between trading countries; (iii) the extent of vertical integration of firms; (iv) the size of firms; (v) the age of foreign affiliates; and (vi) the degree of market concentration. A list of possible variables which determine or at least correlate with the export propensity of foreign subsidiaries is, however, much broader. Later, we will try to shortly elaborate a number of them. We will broadly distinguish between the following groups of variables: (i) investing firm; (ii) industry; (iii) foreign subsidiary; (iv) home country; and (v) host country variables. In doing so, one should be aware that the classification of some variables is not really clear in the sense that they can be attributed to more than one group. For example, a free-trade agreement or investment protection agreement between a home and host country could be classified as a home country as well as a host country variable.
Investing Firm Variables
The most frequently cited investing firm variables include the internationalisation strategy of the investing firm, its degree of multinationality, its size, and its R&D (research and development) intensity, as explained below.
INVESTING FIRM INTERNATIONALISATION STRATEGY. Whether an investing firm applies a horizontal or vertical internationalisation strategy will be the main determinant of a foreign subsidiary's export performance (see, for instance, Dunning 1993; Andersson and Fredriksson 1996; Lankes and Venables 1996a), being high in the case of a vertical and low in the case of a horizontal strategy. 8
DEGREE OF MULTINATIONALITY. It is commonly argued that a higher degree of multinationality leads to more trade, including exports of foreign subsidiaries. Andersson and Fredriksson (1996), however, claim that there is good reason to discard this conventional perspective. A high degree of multinationality (including the number of countries in which an MNE has subsidiaries) may signal horizontal rather than vertical integration, since vertical integration is likely to involve a smaller number of producing affiliates relative to horizontal integration. A high degree of multinationality in horizontal FDI would have a negative impact on the exports of foreign subsidiaries since subsidiaries in the case of horizontal integration substitute for, rather than complement, activities at home.
SIZE OF THE INVESTING FIRM. The size of the investing firm is also expected to influence trade performance (Dunning 1993). On a priori grounds, one would expect a positive correlation between size and trade performance, but there is no empirical evidence in this direction.
R&D INTENSITY OF THE INVESTING FIRM. Andersson and Fredriksson (1996) claim that investment in R&D is important for the internationalisation of horizontally integrated firms, for which foreign production is directly motivated by the internalisation of intangible assets. R&D may also have considerable importance in the case of vertically integrated MNEs, but in that case R&D in the parent is likely to motivate exports from home as a means to channel embodied knowledge, rather than giving rise to extensive exports from subsidiaries. Results of their model confirm the negative impact of investing firm R&D intensity on the propensity of foreign subsidiaries to export.
INDUSTRY VARIABLES
The type of activity in which MNEs are engaged and the nature of activities being undertaken by the subsidiaries importantly co-determine foreign trade transactions, including the export propensity of foreign subsidiaries, associated with FDI. Therefore, any analysis of foreign subsidiaries' export propensity, or comparisons of foreign and domestic firms' export propensity should take into account sectoral differences, i.e. assess to what extent above or under the average export propensity of foreign subsidiaries is related to their concentration in above- or under-average export-intensive industries. A higher export propensity of foreign subsidiaries might well be primarily due to their concentration in trade-intensive sectors (allocation/sectoral effect) and only less, if at all, due to their higher export propensity within the same industry. 9
Theoretical considerations and empirical evidence suggest that MNEs or their subsidiaries tend to concentrate in trade-intensive sectors and to engage in more trade in those sectors than do single-country or indigenous firms (Dunning 1993:401-4). On the other hand, developmental approaches to FDI would say that vertical/factor cost advantage seeking/efficiency seeking FDI will go into industries in which host countries have comparative advantages. Therefore, the (higher) export propensity of foreign subsidiaries is supposed to be importantly influenced by the fact that FDI concentrates in (i) trade-intensive sectors/industries, (ii) in globalised industries characterised by a high degree of market concentration and intra-firm trade, and (iii) often in industries in which a host country has a comparative advantage or in which that comparative advantage can be developed. What are these sectors/industries? Thus, Makhija et al. (1997:693-4) distinguish between multi-domestic industry, multi-domestic transitional industry, simple global industry and integrated global industry, the latter being specifically characterised by export-oriented foreign subsidiaries. Integrated global industries, characterised by the complex co-ordination of geographically dispersed value-added activities, are industrial chemicals, fertilisers and pesticides, resins and plastics, engines and turbines, specialised industrial machinery, office and computing machinery, radio and telecommunications equipment, shipbuilding, railroad, aircraft, photographic and optical goods.
Éltet_ (1998) who analysed the impact of FDI on foreign trade of Czech Republic, Hungary, Slovakia and Slovenia, distinguishes between export-oriented industries, which may be the target of efficiency-seeking foreign investors, and industries with a characteristically medium or low export intensity, oriented towards the domestic and regional market (with exports directed to neighbouring countries), which may attract market-seeking foreign investors. The former group includes electrical and office machinery, radio and TV sets, motor vehicles, precision instruments and furniture manufacturing. The latter group includes food, beverages, tobacco, wood, paper, printing, minerals, basic and fabricated metals, and chemicals. There is also a third group, consisting of industries that are export-oriented, but mainly and traditionally dominated by outward processing activity; this groups includes textiles and clothing, and leather. Éltet_'s analysis suggests two main conclusions. The first is that FDI in the analysed CEE countries tends to concentrate in export-oriented industries. The second is that the four CEE countries examined have clear comparative advantages in labour-intensive goods, and rapidly falling comparative disadvantages in human-capital-intensive goods. In all four countries, the production of human-capital-intensive goods is dominated by efficiency-seeking foreign subsidiaries (Éltet_ 1998:9-18).
Gatling (1993:8-15) in her analysis of foreign investors' motivation in CEE countries found that motives are industry-specific. Establishing a market share was especially important to consumer product companies, a large regional market was the primary reason for investing only in those sectors where economies of scale are important (for instance, car or baby food manufacturing), while low-cost sourcing was a primary motive in energy and power generating sectors. Similarly Stern (1995) argues that, in CEE countries, FDI as a substitute to trade tends to go predominantly to the consumer goods sector, food sector and services, while FDI as a complement to trade (which would happen at a later stage) tends to go predominantly into heavier manufacturing industries.
FOREIGN SUBSIDIARY VARIABLES
Since the final aim of the article is to provide a theoretical and empirical foundation for developing a hypothesis on the determinants of the export propensity of foreign subsidiaries (in the manufacturing industries of CEE countries in transition), the analysis of foreign subsidiary variables is central. Theoretical and empirical analyses offer the following foreign subsidiary variables, which are supposed to be related with their export propensity.
TYPE OF OWNERSHIP: DOMESTIC VERSUS FOREIGN. The main issue and final aim of most studies analysing the export propensity of foreign subsidiaries is to determine to what extent is foreign subsidiaries' higher (or lower), compared to indigenous firms, export propensity due to the factor of "foreign ownership" itself (Lall and Streeten 1977; UNCTAD 1983; Kumar 1990; Dunning 1993). To test this, one should ideally normalise for all other differences between foreign and domestic firms.
HIGHER DEGREE OF VERTICAL INTEGRATION WITHIN THE MNE, RESULTING AND REFLECTED IN A HIGHER INTENSITY OF INTRA-FIRM TRADE, 10 is considered to be a characteristic of efficiency-seeking FDI, that is of export-oriented FDI and, thus, as one of the main determinants of the export propensity of foreign subsidiaries. Dunning (1993:400) claims that investments designed to protect or advance the global competitiveness of investing companies (i.e. efficiency-seeking FDI) are likely to increase intra-firm trade in both intermediate and final products. The authors of the UNCTAD World Investment Report 1996 are of the same opinion. The volume of intra-firm flows tends to increase, and the direction as well as nature of intra-firm flows and their geographical spread tend to change with the complexity of corporate integration. With the growing importance of efficiency-seeking FDI by firms pursuing complex regional or global strategies, a multiplicity of linkages develops, and intra-firm trade becomes more important (UNCTAD 1996:103). According to Lankes and Venables (1996a:346), export-oriented (vertical) FDI is more integrated in the parent firm (e.g. in their sales orientation and their control mode). The reason for this is that these projects are an integral part of MNEs' production network, and as such security of supply becomes of great importance.
The empirical evidence predominantly confirms the link between export propensity and vertical integration and/or intra-firm intensity. Andersson and Fredriksson (1996) found a significant positive impact of vertical integration on the export propensity of foreign subsidiaries of Swedish MNEs. The study of UNCTAD (1983) came to the same result for subsidiaries of foreign firms in Brazil. Lankes and Venables (1996a) in their analysis of foreign subsidiaries in CEE countries confirmed that export-supply type foreign subsidiaries transfer a much higher proportion of their output within the corporation. Éltet_ and Sass (1998), in analysing the export behaviour of foreign subsidiaries in Hungary, found that assembly export-oriented subsidiaries have a much higher share of intra-firm trade (more than 80 percent of imports of inputs was intra-firm) than domestically-based export-oriented and non-export-oriented subsidiaries (about 30 percent of imports of inputs was intra-firm).
THE LEVEL OF FOREIGN OWNERSHIP (EQUITY SHARE). It is widely accepted that foreign investors in export-oriented FDI, in principle, insist more on wholly or high majority ownership than in the case of market-seeking FDI. Lall and Streeten (1977) claim that foreign investors are ready to establish export-oriented subsidiaries only in those developing countries which do not limit the level of foreign ownership and that they appear to be particularly reluctant to invest heavily in export-oriented facilities in developing countries where there are strong moves towards local participation. In the study of Lankes and Venables (1996a), full foreign ownership is more frequent in export-supply than in the local-supply type of foreign subsidiaries in CEE countries. In other words, wholly foreign-owned foreign subsidiaries tend to be more export oriented, and to have more of their output transferred within the firm. Éltet_ and Sass (1998), in analysing foreign subsidiaries in Hungary, came to very similar conclusions.
TYPE OF ENTRY MODE: GREENFIELD VERSUS ACQUISITION. The suggestion here is that greenfield FDI is a more common entry mode for export-oriented than for market-seeking FDI. In the Hungarian case, Éltet_ and Sass (1998) found that the share of greenfield FDI was 30 percent for non export-oriented subsidiaries, 55 percent for domestically-based export-oriented subsidiaries and 60 percent for assembly export-oriented foreign subsidiaries.
SIZE OF INVESTMENT/SUBSIDIARY is supposed to be positively correlated with the export propensity of a foreign subsidiary. According to Andersson and Fredriksson (1996), this is because size is indicative of economies of scale at the plant level, and associated with an international specialisation of production and greater exports from subsidiaries. The empirical evidence is not conclusive in this case. The size of Swedish subsidiaries abroad has a significant positive impact on their export propensity (Andersson and Fredriksson 1996), and the activity of export-oriented foreign subsidiaries in Hungary seems to require greater capital than that of market-oriented subsidiaries (Éltet_ and Sass 1998). However, there is almost no difference in the size of local supply and export-supply type foreign subsidiaries in CEE countries from the sample of Lankes and Venables (1996a), and the size of foreign subsidiaries did not explain differences in the export propensity of foreign and domestic firms in India (Kumar 1990).
CAPITAL INTENSITY VERSUS LOW-COST UNSKILLED OR SEMI-SKILLED LABOUR VERSUS SKILLED LABOUR. Is export-oriented FDI concentrated (i) in capital-intensive industries, where economies of scale and scope lead to efficiency-seeking FDI, (ii) in traditional labour-intensive industries, where factor cost differentials stimulate the relocation of labour-intensive production, or parts of production to low labour cost locations, or (iii) in industries characterised by intensive use of skilled labour in the framework of factor cost differences FDI and/or efficiency-seeking FDI (integrated international production), or not? The three levers of export-oriented FDI are to a certain extent alternative to each other. 11 Each of the three propositions could be tested in one or the other direction, depending on other factors. This is clearly reflected in the existing studies on the subject that confirm and deny a positive correlation between each of the three variables and the export propensity of foreign subsidiaries:
- EXPORT-ORIENTED FDI IN CAPITAL-INTENSIVE INDUSTRIES. The correlation between capital intensity and the export performance of foreign subsidiaries in Brazil was found to be positive and statistically significant (UNCTAD 1983). On a similar front, Ozawa (1972) spoke of Japanese FDI in the early 1970s, when Japan began to restructure its economy away from pollution-prone industries and firms in heavy and chemical industries were encouraged to invest abroad. Contrary to that, Kumar (1990) could not explain the export performance of foreign and domestic firms in India by capital intensity.
- EXPORT-ORIENTED FDI IN UNSKILLED/SEMI-SKILLED LABOUR-INTENSIVE INDUSTRIES. Low-cost unskilled or semi-skilled labour has traditionally been considered as the major motivating factor for export-oriented FDI based on factor cost differences. Thus, the product life cycle theory argues that MNEs locate standardised stages of production in low-income countries for the purpose of production for world markets. Empirical evidence of a somewhat older date (Ozawa 1972; Riedel 1975; Hood and Young 1979) confirms that low labour costs are probably the major determinant of export-oriented FDI in developing countries. In literature of more recent times, the role of cheap labour as a motivation for export-oriented FDI is very much reduced, if not eliminated. Thus, the European Commission (1994) is of the opinion that the popular view of FDI being essentially motivated by the search for lower labour costs is far away from the reality. Lankes and Venables (1996a) claim that most surveys on the motivation of foreign investors in CEE countries explicitly play down the role of cheap labour.
- EXPORT-ORIENTED FDI IN SKILLED LABOUR-INTENSIVE INDUSTRIES. Parallel to the decreasing role of cheap unskilled or semi-skilled labour, the importance of skilled labour as the motivation for export-oriented FDI is on the increase. With the shift towards advanced, flexible production systems and the need to assure quality and reliability (such as just-in-time delivery), growing importance is attached to factors such as infrastructure, educational standards and skill levels (the European Commission 1994). Kravis and Lipsey (1982) 12 found that affiliates of the United States' (US) origin tend to export from countries with high wages, suggesting that the cost is more than offset by other influences, skill of the labour force being one of them. According to Lankes and Venables (1996a), in export-supply FDI projects skilled labour costs are emerging as a more important consideration than in local supply FDI projects. 13 Éltet_ and Sass (1998) produced similar results for export-oriented FDI in Hungary: workforce qualifications are much more important than low labour costs. However, the results of Meyer's (1998) survey of foreign subsidiaries in CEE countries are quite the opposite: for factor-cost oriented FDI low labour costs are more important than qualifications.
How do we reconcile these two aspects of labour, low costs and qualification, as being a motivation for export-oriented FDI? The solution obviously lies in the different types of export-oriented FDI. While low costs of labour are more important for simple factor-cost-oriented FDI, qualification is more important for efficiency-seeking FDI or integrated international production. This issue was addressed by Papanastassiou and Pearce (1992) who distinguished between rationalised product subsidiaries and world (or regional) product mandate subsidiaries. For the former, a key attraction are low labour costs, while the latter are predominantly attracted by the availability of personnel with particular types of original creative abilities needed to generate a subsidiary's distinctive competitive potential. Wage rates are expected to be negatively related to export propensity in the case of rationalised product subsidiaries, and positively in the case of world product mandate subsidiaries. In world product mandate subsidiaries, the wage rate is considered an indicator of quality (with cost having a much lower priority). 14 Therefore, one could not say that a low or high wage rate is positively correlated with the export propensity of foreign subsidiaries. A low and/or high wage rate could both be positively correlated with the high export propensity of foreign subsidiaries. However, a positive correlation of a low wage rate with a high export propensity is an indicator of a rationalised product subsidiary, while a positive correlation of a high wage rate with a high export propensity is an indicator of world product mandate subsidiaries.
THE SCOPE OF VALUE ADDED. One of the differences between stand-alone subsidiaries in horizontal (market-seeking) integration and vertically integrated (export-oriented) subsidiaries is the scope of activities/functions performed by each type of subsidiary. Stand-alone subsidiaries are, in general, active in all functions in the vertical chain (from research to distribution), while the activities of subsidiaries in vertical integration are confined to processing and assembling of imported components, which are then exported for further processing and marketing (UNCTAD 1996). A subsidiary in vertical integration may, thus, have less scope for its own value-added activities than stand-alone ventures. This would lead to a prediction of a negative correlation between a subsidiary's export propensity and its scope for value added.
IMPORT PROPENSITY. A positive correlation between the export and import propensity of foreign subsidiaries is somehow a priori. In a system of MNE integrated international production, a vertically integrated subsidiary produces and exports what it is most efficient in and imports all it needs from subsidiaries that are more efficient in other segments. Vertical internationalisation with efficiency-seeking FDI has strengthened the international division of labour with the increasing exports and imports, in particularly intra-firm (including inter-affiliate flows), of those involved (UNCTAD 1996). In a sample of foreign subsidiaries in various developing countries analysed by Reuber et al. (1973), import substitutive subsidiaries bought 60 percent of inputs locally, while export-oriented ones bought only 40 percent. Rojec (1994) also found a positive correlation between export and import propensity in a sample of foreign subsidiaries in Slovenia. Half of the subsidiaries also claimed that the fact that a product was destined for exports was an important reason for the use of imported instead of local inputs. In foreign subsidiaries in Hungary, the share of imported inputs was 40-45 percent in the case of non-export-oriented subsidiaries, 60-70 percent in assembly export-oriented subsidiaries and less than 20 percent in domestically-based export-oriented subsidiaries (Éltet_ and Sass 1998). 15 Similar results are reported by Lankes and Venables (1996b).
PRODUCTION COST CONSIDERATIONS. Export-oriented foreign subsidiaries attach greater importance to production cost considerations. According to Lankes and Venables (1996a), the most striking difference between local-supply and export-supply types of foreign subsidiaries in CEE countries, as far as the motivation of foreign investors is concerned, is the importance attached to production costs by export suppliers.
HOME COUNTRY VARIABLES
There are likely to be variations in the extent and pattern of trade transactions associated with FDI according to the home countries involved (Dunning 1993). Even more, this was the very basis of Kojima's (1987) theory of trade- and anti-trade-oriented FDI. 16 However, empirical evidence has not really confirmed Kojima's theoretical postulate that Japanese FDI is something specific; the particular behaviour of Japanese foreign investors was rather a reflection of a specific phase in Japanese economic development than being a lasting characteristic (Lee 1987). Also, there is no doubt that EU-based MNEs behaved specifically when the single European market was announced in 1987, since their own "backyard" increased tremendously, and that today they probably do behave specifically in CEE countries which are candidates for EU membership. But this is again a transitory phenomenon, or a consequence of a specific situation. Once MNEs restructure their FDI according to the new situation the reasons for such specific behaviour will disappear. Reich (1998) argues that high intra-firm exports from parent companies to their subsidiaries are characteristic for German and especially Japanese MNEs, but not for US ones. To conclude, there is no doubt that home country variables co-determine MNEs behaviour regarding the export propensity of their foreign subsidiaries, but one can hardly develop this relationship into a testable proposition.
Host Country Variables (The ESP Paradigm)
If there are likely to be variations in the extent and pattern of trade transactions associated with FDI according to the home countries involved, it is even more likely that there are variations according to the host countries involved. For instance, data support the view that differences in the trade propensities of foreign and domestic firms are the most pronounced in open smaller industrialised (or industrialising) economies and in resource-based economies where MNEs are primarily engaged in export-generating activities (Dunning 1993:401-4). The influence of FDI on a host country's trade will crucially depend on its environment/system/policies (ESP) configuration. 17 There are three host country variables that are especially relevant for the export propensity of foreign subsidiaries. They are host country (regional) market size, level of development and the policy environment. In CEE countries, one additional variable is included, i.e. the scope and stage of transition reforms.
HOST COUNTRY (REGIONAL) MARKET SIZE. A large (and growing) host country market is obviously a major motivation for horizontal, market-seeking FDI. Small host countries do not possess this basic attractiveness for market-seeking FDI unless they are an integral part of a free-trade area or a customs union. However, is it reasonable on that basis to expect that compared to large countries FDI in small countries is more of an export-oriented type? In other words, is the size of a host country negatively correlated with the export propensity of foreign subsidiaries, i.e. do foreign subsidiaries in small host countries demonstrate a higher export propensity? Kravis and Lipsey (1982, cited in Andersson and Fredriksson 1996: 253) do not agree. They found that host country market size is positively related to exports from subsidiaries of US firms, which is explained by market size allowing for greater economies of scale (specialised operations). 18 Michalet (1997) favoured the proposition that a large host country market is a major attraction for market-seeking FDI, but did not necessarily favour the proposition that foreign subsidiaries in small host countries are more export oriented. Major FDI recipient countries are attractive as investment locations under both the market-seeking and cost-reduction strategies. MNEs are less interested in either big markets without conditions for world competitive production, or in pure export platforms, than they used to be. Instead, they are increasingly looking for sites that offer both market access and conditions for world competitive production. Andersson and Fredriksson (1996) found a statistically significant negative correlation between host country market size and the export propensity of foreign subsidiaries of Swedish MNEs. This is because the need for product differentiation and after-sales services, arising from segmentation of national markets in terms of consumer preferences, is positively correlated with market size, favouring horizontal investments. Papanastassiou and Pearce (1992) also suggest a negative relationship between host country market size and foreign subsidiaries' export orientations; the larger the market the more likely it is that foreign subsidiaries would take on a predominantly import-substituting role, with a primary focus on the local market. For Éltet_ (1998), a small host country market means that significant exports may arise not only from efficiency-seeking but also from market-seeking FDI that embraces the neighbouring countries. To conclude, although the literature on the subject is not fully unanimous, the overall findings do suggest that host country market size is negatively correlated with the export propensity of foreign subsidiaries.
HOST COUNTRY DEVELOPMENT LEVEL. The transactions of MNEs are likely to vary according to the stage of development of the countries in which they operate (Dunning 1993). But what is the direction of the correlation between a host country's development level and the export propensity of foreign subsidiaries? Is a higher development level correlated with horizontal rather than vertical FDI, or vice versa? Brouthers et al. (1996) suggest that MNEs generally tend to use FDI in advanced industrial nations for market access and penetration (market-seeking), but they tend to use FDI in developing countries to gain resource advantages that can be exploited in export markets (factor cost advantage-seeking). Opposite to that, for Helpman and Krugman (1985), an increasing discrepancy in factor costs, i.e. a falling host country development level, would favour less trade-oriented subsidiaries. Andersson and Fredriksson (1996) found a statistically significant positive impact of host country income level on the export propensity of foreign subsidiaries of Swedish MNEs. This is explained by vertical investments being particularly dependent on the quality of infrastructure, productivity of the labour force and political stability, i.e. factors that are generally associated with a high development level. Papanastassiou and Pearce (1992) claim that foreign investors are keener to locate world (or regional) product mandate subsidiaries in developed and rationalised product subsidiaries in developing countries. Empirical evidence seems to favour a positive correlation between host country development level and foreign subsidiaries' export propensity.
SCOPE AND STAGE OF TRANSITION REFORMS IN CEE COUNTRIES is a specific aspect of host country variables which relates to the level of development as well as to the systemic and policy issues. According to Lankes and Venables (1996a), FDI projects in CEE countries that are more advanced in transition are more likely to be export oriented, more integrated into foreign parents' multinational production process and more likely to exploit the comparative advantage of a host economy. Econometric analyses provide strong support for the proposition that progress in transition is associated with a changing mix of project types, away from projects that are designed to serve local markets, and towards export-supply projects. This is because export-supply FDI projects are an integral part of parent firms' production network, and as such security of supply becomes greatly important. Whereas horizontal investments tend to replicate activities, vertical investments tend to involve relocation, leaving the firm vulnerable to supply disruption.
POLICY ENVIRONMENT IN A HOST COUNTRY. Two aspects of the policy environment in host countries are generally considered when discussing export-oriented FDI. The first is the general policy environment conceptualised in the export-oriented, outward-looking versus import-substitutive, inward-looking development/policy concept and the second is the policy of international economic relations, composed of the foreign trade regime, economic integration policy and FDI policy. To successfully attract FDI, an appropriate policy environment is much more important to export-oriented than market-seeking FDI. While a large local market could persuade a foreign investor to establish a stand-alone subsidiary to service the local market even in an sub-optimal policy environment, he will never threaten the operations of the MNE as a whole by establishing a vertically integrated subsidiary in such an environment. The general policy environment in host countries is much more important for vertically integrated export-oriented than for stand-alone horizontal, market-seeking FDI.
The impact of a host country's general policy environment on export-oriented FDI should be understood in the broadest sense possible. (i) The first in this regard is a liberal economic policy concept. The main trend in developing countries' policies in the last decade or so is a trend towards export promotion, liberalisation of imports, privatisation and in general a greater reliance on market forces in economic decisions. All of these have created a more congenial environment for the operations of MNEs in developing countries, which might bring a stimulating effect on export-oriented industrial collaboration (Islam 1995). (ii) The second is political and economic stability, and a favourable and friendly attitude to FDI. Lall and Streeten (1977) explain the high export propensity of foreign subsidiaries in Asia in the 1970s by the fact that foreign investors are prepared to invest in export-oriented capacities only in those developing countries that are politically stable, which have a favourable attitude to FDI and in which there are no trade union pressures nor restrictions on the share of foreign equity. In the case of local market-seeking FDI, foreign investors are ready to accept higher risks. (iii) The third is the export-oriented versus import-substitutive development concept and the policy environment. Empirical evidence clearly demonstrates that countries with an export-oriented development concept have had more success in attracting export-oriented FDI than countries with an import-substitutive concept. FDI in countries with high levels of import protection tends to be less export-oriented than FDI in countries with a low level of protection (IMF 1985; UNCTAD 1996). The theoretical validity of the proposition that a country adopting an export promotion strategy would attract more export-oriented FDI rests on the inefficiencies of the import substitution model, the neutrality of the export promotion strategy between export and import sectors and the efficient allocation of resources based on a comparative advantage (Bhagwati 1978).
As for policy of international economic relations, a liberal foreign trade regime (import liberalisation) and integration into free-trade areas or customs unions (free access to foreign markets) are crucial stimulators for foreign investors to invest in export-oriented facilities in a country. (i) In order for efficiency-seeking (or rationalised) foreign production to take place, cross-border markets must be both well developed and open (Dunning 1993). Barry and Bradley (1997) claimed that liberalised trading arrangements are clearly a prerequisite for substantial FDI inflows in small economies, WTO (1996) suggests that a low level of import protection can be a strong magnet for export-oriented FDI, and Éltet_ (1998) says that a liberalised trade system is a prerequisite for export-oriented FDI. (ii) The lowering of barriers to trade especially facilitates vertical FDI (Andersson and Fredriksson 1996). That is why efficiency-seeking FDI flourishes in regionally integrated markets (Dunning 1993). A free-trade area or customs union gives firms the opportunity to serve an integrated market from one or a few production sites, and thereby to reap the benefits of scale economies. This can have a pronounced impact on investment flows, especially while firms are restructuring their production activities (WTO 1996). (iii) A special case of economic integration, with a particularly high influence on the export propensity of foreign subsidiaries, is the EU. Here, integration is very deep and comprehensive, a single market really creates the situation of a large domestic market. In the late 1980s, firms restructured their activities in Europe in response to the EC-1992 programme. This led primarily to an increase of vertically integrated investments. The removal of trade barriers and harmonisation of standards also improve the opportunities for exports from horizontally integrated firms (UNCTC 1993). Éltet_ and Sass (1998) found that in the case of Hungary export-oriented foreign subsidiaries attached greater importance to the association agreement with EU and facilitation of exports to EU (tax reductions, customs-free zones, tax and tariff concessions) than market-seeking FDI. Andersson and Fredriksson (1996) hypothesised in their model that Swedish subsidiaries in EU should have a relatively high export propensity due to regional liberalisation facilitating intra-EU trade. This impact should arise both because of improved prospects for vertical integration characterised by a concentration of production in a limited number of locations, and because of a greater tendency to export from horizontally integrated subsidiaries. Indeed, they found a statistically positive influence of being an EU-based subsidiary on its export propensity.
Conclusions
THE DISCUSSION OF FDI-TRADE RELATIONS HAS DEVELOPED TO THE STAGE WHEN THERE IS NO LONGER A SIMPLE PRO-TRADE (FDI AS A COMPLEMENT TO TRADE) OR ANTI-TRADE (FDI AS A SUBSTITUTE TO TRADE) FDI HYPOTHESIS. There is increasing recognition that FDI and trade can be either complements or substitutes. Two streams of FDI theory are especially relevant for analysing the export-oriented FDI and export propensity of foreign subsidiaries. The first is developmental approaches to FDI, and the second is the analysis of the motives and of the main types of foreign production. The main conclusion of the developmental approach to FDI, which devotes almost all, if not all, its attention to export-oriented FDI, is that export-oriented FDI basically happens when an investing firm (country) begins to lose its comparative advantage or competitive edge in a particular area of production. Here, production will be relocated to locations with a comparative advantage in that particular production. This will, as a rule, be simple export-platform type of FDI motivated by cheap unskilled labour (and partly also by the relocation of pollution-intensive activities). The analysis of the main types of foreign production, in business management literature and new international trade theory, goes beyond the simple export platform type of FDI and basically distinguishes between two types of FDI/investing firm strategy/subsidiary positions. The first is market-seeking (horizontal) FDI established for the procurement of a local or/and adjacent regional market. The second is export-oriented (natural resource-seeking, efficiency-seeking, strategic asset-seeking, vertical, sourcing, factor cost differences-seeking) FDI with vertically integrated subsidiaries, which is determined by differentials in factor endowments (assembly, rationalised product subsidiaries), or by advantages in economies of scale and scope, and of differences in consumer tastes and supply capabilities (integrated international production, regional or world product mandate subsidiaries).
The analysis provides number of propositions for potential empirical research on the determinants of and factors relating to the export propensity of foreign subsidiaries. In the formulation and testing of propositions for regression analysis, two aspects should be specifically taken into account. The first is the distinction between (i) determinants of export propensity like investing firm internationalisation strategy (horizontal versus vertical), type of ownership (foreign subsidiaries versus indigenous firms), low-cost unskilled or semi-skilled labour, skilled labour, industry characteristics, host country market size, host country development level, scope and stage of transition reforms, policy environment in a host country — where the direction of causality is obvious and clear, i.e. that higher or lower values of a certain factor lead to higher export propensity, and (ii) factors relating to export propensity — like degree of multinationality, size of the investing firm, higher degree of vertical integration within an MNE, level of foreign ownership (equity share), type of entry mode (greenfield versus acquisition), size of investment/subsidiary, scope of value added, import propensity, production cost consideration — where the direction of causality it is not a priori. The latter factors could be treated as a cause of higher or lower export propensity, but they might be only a consequence of higher export propensity, i.e. patterns which are typical once a subsidiary is export oriented. The second aspect is the fact that, except perhaps for investing firm internationalisation strategy (horizontal versus vertical strategy), theoretical considerations and particularly the empirical evidence are not unanimous, and for some factors they are far from being unanimous.
References
AGOSIN, Manuel R. and Francisco J. PRIETO (1993) Trade and Foreign Direct Investment Policies: Pieces of a New Strategic Approach to Development? Transnational Corporations 2(2), 63-86.
ANDERSSON, Thomas and Torbjoern FREDRIKSSON (1996) International Organization of Production and Variation in Exports from Affiliates. Journal of International Business Studies 27(2), 249-63.
BARRY, Frank and John BRADLEY (1997) FDI and Trade: The Irish Host Country Experience. The Economic Journal 107(445), November, 1789-811.
BEHRMAN, Jack N. (1972) The Role of International Companies in Latin America: Autos and Petrochemicals. Lexington, MA: Lexington Books.
BHAGWATI, Jagdish (1978) Foreign Trade Regimes and Economic Development: Anatomy and Consequences of Exchange Control Regimes. Cambridge, MA: Balinger.
BIRKINSHAW, Julian and Neil HOOD, eds (1998) Multinational Corporate Evolution and Subsidiary Development. London: Macmillan.
BLOMSTROEM, Magnus (1990) Transnational Corporations and Manufacturing Exports from Developing Countries. New York: United Nations.
BROUTHERS, Lance Eliot, Steve WERNER and Timothy J. WILKINSON (1996) The Aggregate Impact of Firms' FDI Strategies on the Trade Balances of Host Countries. Journal of International Business Studies 27(2), 359-73.
CAVES, Richard E. (1971) International Corporations: The Industrial Economics of Foreign Investment. Economica 38(2), 1-27.
CAVES, Richard E. (1982) Multinational Enterprise and Economic Analysis. Cambridge: Cambridge University Press.
CZINKOTA, Michael R. and Ilkka A. RONKAINEN (1997) International Business and Trade in the Next Decade: Report from a Delphi Study. Journal of International Business Studies 28(4), 827-44.
DUNNING, John H. (1993) Multinational Enterprises and the Global Economy. Wokingham: Addison-Wesley.
EBRD (1994) Transition Report. London: European Bank for Reconstruction and Development.
ÉLTET_, Andrea (1998) "The Impact of FDI on the Foreign Trade of Four Smaller CEE Countries". Budapest: Institute for World Economics, Hungarian Academy of Sciences, Working Papers, No. 96.
ÉLTET_, Andrea and Magdolna SASS (1998) "Motivations and Behaviour by Hungary's Foreign Investors in Relations to Exports". Budapest: Institute for World Economics, Hungarian Academy of Sciences, Working Papers, No. 88.
EUROPEAN COMMISSION (1994) "Trade and Investment". Brussels: Directorate General for External Economic Relations, Discussion Paper (December).
FDI Confidence Index (1998) 1(2), December. A.T. Kearney, Global Business Policy Council.
GATLING, Rene (1993) "Foreign Investment in Eastern Europe". London: The Economist Intelligence Unit.
GRAY, H. Peter, ed. (1992) Transnational Corporations and International Trade and Payments. UN Library on Transnational Corporations. London: Routledge..
HELPMAN, Elhanan and Paul R. KRUGMAN (1985) Market Structure and Foreign Trade: Increasing Returns, Imperfect Competition, and the International Economy. Cambridge, MA: MIT Press.
HOOD, Neil and Stephen YOUNG (1979) The Economics of Multinational Enterprise. London: Longman.
HOOLEY, Graham, David SHIPLEY, John FAHY, Tony COX, Jozsef BERACS and Krisztina KOLOS (1996) Foreign Direct Investment in Hungary: Resource Acquisition and Domestic Competitive Advantage. Journal of International Business Studies 27(4), 683-709.
IMF (1985) "Foreign Private Investment in Developing Countries". Washington, DC: International Monetary Fund, Occasional Paper No. 33.
ISLAM, Azizul (1995) Collaboration with Transnational Corporations and Manufactured Exports from Developing Countries: Potentials and Policy Implications. RIS (Research & Information System) Digest 11(3/4), 1-21.
KOJIMA, Kiyoshi (1978) Direct Foreign Investment. London: Croom Helm.
KOJIMA, Kiyoshi (1987) A Model of Trade-Oriented Direct Foreign Investment. In Seiji NAYA, Vinju VICHIT-VADAKAN and Udom KERDPIBULE (eds) Direct Foreign Investment and Export Promotion: Policies and Experiences in Asia, 159—83. Kuala Lumpur/Honolulu: Southeast Asian Central Banks Research and Training Centre/East-West Resource Systems Institute.
KOO, Bohn-Young (1993) Foreign Investment and Economic Performance in Korea. In Sanjaya LALL (ed.) Transnational Corporations and Economic Development, 288-313. The United Nations Library on Transnational Corporations. London and New York: Routledge.
KRUGMAN, Paul (1983) The 'New Theories' of International Trade and the Multinational Enterprises. In David B. AUDRETSCH and Charles P. KINDLEBERGER (eds) The Multinational Corporation in the 1980s. Cambridge, MA: MIT Press.
KUMAR, Nagesh (1987) Foreign Investment and Export Orientation: The case of India. In Seiji NAYA, Vinju VICHIT-VADAKAN and Udom KERDPIBULE (eds) Direct Foreign Investment and Export Promotion: Policies and Experiences in Asia, 357-81. Kuala Lumpur/Honolulu: Southeast Asian Central Banks' Research and Training Centre/East-West Resource Systems Institute.
KUMAR, Nagesh (1990) Multinational Enterprises in India. London: Routledge.
LALL, Sanjaya and Paul STREETEN (1977) Foreign Investment, Transnationals and Developing Countries. London: Macmillan.
LANKES, Hans-Peter and Anthony J. VENABLES (1996a) Foreign Direct Investment in Economic Transition: The Changing Pattern of Investments. Economics of Transition 4(2), 331-47.
LANKES, Hans-Peter and Anthony J. VENABLES (1996b) "Foreign Direct Investment in Eastern Europe and the Former Soviet Union: Results from a Survey of Investors". London: EBRD, Working Paper.
LEE, Chung H (1987) Is There Anything Unique about Japanese Direct Foreign Investment. In Seiji NAYA, Vinju VICHIT-VADAKAN and Udom KERDPIBULE (eds) Direct Foreign Investment and Export Promotion: Policies and Experiences in Asia, 185-200. Kuala Lumpur/Honolulu: Southeast Asian Central Banks' Research and Training Centre/East-West Resource Systems Institute.
MAKHIJA, Mona V., Kwangsoo KIM and Sandra D. WILLIAMSON (1997) Measuring Globalization of Industries Using a National Industry Approach: Empirical Evidence across Five Countries and over Time. Journal of International Business Studies 28(4), 679-711.
MARKUSEN, James R. (1995) The Boundaries of Multinational Enterprises and the Theory of International Trade. Journal of Economic Perspectives 9(2), 169-89.
MEYER, Klaus (1998) Direct Investment in Economies in Transition. Cheltenham: Edward Elgar.
MEYER, Klaus E. (1995) "Direct Foreign Investment, Structural Change and Development: Can the East Asian Experience be Replicated in East Central Europe?". London: London Business School, CIS-Middle Europe Centre, Discussion Paper Series, No. 16 (January).
MICHALET, Charles-Albert (1997) "Strategies of Multinationals and Competition for Foreign Direct Investment". Washington, DC: Foreign Investment Advisory Service, The World Bank, Foreign Investment Advisory Service Occasional Paper 10.
MUNDELL, Robert A. (1957) International Trade and Factor Mobility. American Economic Review 47(3), 321-35.
NARULA, Rajnesh (1996) Multinational Investment and Economic Structure. London and New York: Routledge.
OZAWA, Terutomo (1972) Multinationalism — Japanese Style. Columbia Journal of World Business 7(6), 33-42.
OZAWA, Terutomo (1979) International Investment and Industrial Structures: New Theoretical Implications from the Japanese Experiences. Oxford Economic Papers 31(1), 72-92.
OZAWA, Terutomo (1992a) Foreign Direct Investment and Economic Development. Transnational Corporations 1(1), 27-54.
OZAWA, Terutomo (1992b) "Japanese MNCs as Potential Partners in Eastern Europe's Economic Reconstruction". Copenhagen: Institute of International Economics and Management, Working Paper 2-92.
PAPANASTASSIOU, Marina and Robert PEARCE (1992) Motivation and Market Strategies of US Foreign Direct Investments: an Analysis of Host Country Determinants. In Proceedings of the 18th Annual Conference of EIBA (European International Business Association), the University of Reading, 13-15 December, 1-22. Reading: EIBA.
PORTER, Michael E, ed. (1986) Competition in Global Industries. Boston: Harvard Business School Press.
REICH, Simon (1998) Globalisation and Changing Patterns of Foreign Direct Investment: The Lessons for Central and Eastern Europe. Journal of International Relations and Development 1(1-2), 106-19.
REUBER, Grant L., H. CROOKELL, M. EMERSON and G. GELLENS-HAMMONS (1973) Private Foreign Investment in Development. Oxford: Clarendon Press.
RIEDEL, James (1975) The Nature and Determinants of Export-oriented Direct Foreign Investment in a Developing Country: A Case Study of Taiwan. Weltwirtschaftlisches Archiv 111(3), 505-28.
ROJEC, Matija (1994) Tuje investicije v slovenski razvoj [Foreign Investment in Slovenian Development]. Ljubljana: Znanstveno in publicisti_no sredi__e.
ROJEC, Matija (2000) Restructuring and Efficiency Upgrading with FDI. In Gabor HUNYA (ed.) Integration through Foreign Direct Investment, 130-49. Cheltenham, UK, Northampton, USA: Edward Elgar in association with the Vienna Institute for International Economic Studies (WIIW).
ROOT, Franklin (1990) International investment and trade, 6th edition. Cincinnati, OH: Southwestern.
STERN, Richard E (1995) "Foreign Direct Investment, Exports, and East-West Integration: Theory and Practice". Laxenburg: International Institute for Applied Systems Analysis, mimeo.
UNCTAD (1983) The Role of Transnational Enterprises in Latin American Integration Efforts: Who Integrates for Whom and for What Benefits?, TO/B/C/7/38. Geneva: UNCTAD.
UNCTAD (1995) Foreign Direct Investment in Central and Eastern Europe, TD/B/ITNC/Misc.2. Geneva: UNCTAD.
UNCTAD (1996) World Investment Report 1996. New York and Geneva: United Nations.
UNCTC (1992) World Investment Report: Transnational Corporations as Engines of Growth. New York: United Nations Centre on Transnational Corporations.
UNCTC (1993) World Investment Report: Transnational Corporations and Integrated International Production. New York: United Nations Centre on Transnational Corporations.
VERNON, Raymond (1996) International Investment and international trade in the product life cycle. Quarterly Journal of Economics 80(2), 190-207.
WHITE, R. E. and T. A. POYNTER (1984) Strategies for Foreign-owned Subsidiaries in Canada. Business Quarterly 48(4), 59-69.
WIIW (1998) Database on Foreign Investment Enterprises in Central European Manufacturing — Austria, the Czech Republic, Hungary, Slovakia, Slovenia: 1993-1996. Vienna: Vienna Institute for International Economic Studies.
WTO (1996) Annual Report 1996. Geneva: World Trade Organization.
Endnotes
Note *: Matija Rojec is Assistant Professor at the University of Ljubljana, Faculty of Social Sciences, and Senior Research Fellow at the Institute of Macroeconomic Analysis and Development, and the Centre of International Relations, Faculty of Social Sciences.
This research was undertaken with support from the European Union's Phare ACE Programme 1997 in the framework of Phare-ACE research project Impact of foreign direct investment on the international competitiveness of CEEC manufacturing and EU enlargement (Project number: P97-8112-R). The content of this article is the sole responsibility of the author and it in no way represents the views of the Commission or its services.
Note 1: For more on the ESP paradigm, see Dunning (1993:270-1).Back.
Note 2: For an overview of these approaches, see Meyer (1998:80-6).Back.
Note 3: Another possibility is to move into different, less labour-intensive industry at home. Why would an MNE move within the same industry abroad than in another industry at home? There are two reasons: (i) depreciation of existing industry-specific intangible capital and loss of rents arising thereof; (ii) loss of wealth in the sale of its industry- or firm-specific physical capital. By making FDI in the same industry and by transferring its second-hand machinery to its affiliate, the firm internalises the market and avoids the loss of wealth (Meyer 1995; Lee 1987:196).Back.
Note 4: In the 1983-1993 period, for parent firms the share of intra-firm exports in total exports increased from 33.8 percent to 44.4 percent, and the share of intra-firm imports in total imports from 37.9 percent to 48.6 percent. The corresponding increases in the intra-firm trade of foreign affiliates were from 55.2 percent to 64.0 percent in the case of exports and from 82.8 percent to 85.5 percent in the case of imports. Intra-firm sales are, moreover, no longer confined mainly to flows between parent firms and affiliates in one or both directions; inter affiliate flows (exports of affiliates to one another) within MNE systems assume greater importance (UNCTAD 1996:103-4).Back.
Note 5: As put by Gray (1992), the two main characteristics of modern-era FDI are: (i) the growth of production sharing (production sharing exists when an MNE plans the production of value-added activities in different countries according to relative differences in production costs) and of global planning of production and strategy and, therefore of intra-firm trade; and (ii) a substantial amount of intra-industry (cross hauling) FDI.Back.
Note 6: Papanastassiou and Pearce (1992) argue that as host countries get richer the export orientation of foreign subsidiaries may become relatively less dependent on the pure cost effectiveness of the local labour force, and more responsive to elements of skill and creativity, which may instead generate competitiveness through the innovative development of differentiated products. This means that the global perspectives of MNEs, as implemented through the roles played by their subsidiaries in individual host countries and the way their transformation responds to the challenges and opportunities of an interdependent world, can boost host countries' understanding of their potential in the global economy.Back.
Note 7: For an overview of articles on changes in MNE organisational structures and on subsidiaries' position/role and determinants in an MNE framework, see Andersson and Fredriksson (1996:249-50). See also White and Poynter (1984) and Birkinshaw and Hood (1998).Back.
Note 8: This has a broader impact on analysing the export propensity of foreign subsidiaries because one can hypothesise that it is precisely 'the horizontality' versus 'verticality' which co-determine a number of other specificities of local/regional market versus export-oriented FDI.Back.
Note 9: Thus, the authors of the study by UNCTAD (1983) found that the overall import propensity of the affiliates of foreign-owned firms was 77 percent higher than of Brazilian-owned firms, but that, after controlling for industrial characteristics and market structure, the former group of firms imported only 4 percent more of their sales than the latter firms.Back.
Note 10: A certain degree of caution might be necessary here, since empirical evidence on the relations between vertical integration and intra-firm trade is not fully conclusive (Dunning 1993:408-10).Back.
Note 11: It is basically the unskilled or semi-skilled labour intensity that does not go along with capital intensity or skilled labour intensity, while capital and skilled labour intensity frequently go along with each other.Back.
Note 12: KRAVIS, Irving B. and Robert E. LIPSEY (1982) The Location of Overseas Production and Production for Export by U.S. Multinational Firms. Journal of International Economics 12, 201-23; quoted in Andersson and Fredriksson (1996:253).Back.
Note 13: In their sample of foreign subsidiaries in CEE countries, the share of skilled labour in total labour costs was 58.9 percent for local supply subsidiaries and 69.2 percent for export-supply subsidiaries.Back.
Note 14: Since firms need to take productivity into account in determining the role of subsidiaries in particular countries, the wage rate is only one side of the coin. Therefore, Papanastassiou and Pearce (1992) also took wage bill productivity (sales in US $ per US $ of wage paid) into consideration. They postulated a positive relationship between wage bill productivity and the export performance of rationalised product subsidiaries, because highly cost efficient labour is generally assumed to be a key factor in attracting these very strongly export-focused subsidiaries. A positive sign for world product mandate subsidiaries is also possible. It would follow from local inputs of skilled labour that lead to successful exports of high value-added differentiated products. Nevertheless, pursuit of this type of cost effectiveness reflected in wage bill productivity is much less of a priority for world product mandate subsidiaries than for rationalised product subsidiaries.Back.
Note 15: A 'lack of adequate local suppliers' was an important or very important problem for 58 percent of the export-oriented foreign subsidiaries, but only for 43 percent of the non-export-oriented subsidiaries. However, the regression analysis of Kopint (Kopint (1997) Business Activity Report (2), Budapest; quoted in Éltet_ 1998:11), based on an empirical survey of Hungarian manufacturers, showed that majority foreign ownership does not increase the probability of greater imports, in general. There was no significant difference in the import propensity of two firms with the same parameters, one of which was in majority foreign ownership.Back.
Note 16: According to Kojima (1978, 1987), if FDI occurs in an industry in which an investing country has a comparative disadvantage and a host country a comparative advantage, this is trade oriented FDI which does not substitute but complements international trade, and vice versa if FDI occurs in an industry in which an investing country has a comparative advantage. Kojima claimed that Japanese FDI were trade oriented and US FDI anti trade oriented.Back.
Note 17: For more on the ESP paradigm, see Dunning (1993:270-1).Back.
Note 18: Andersson and Fredriksson (1996:253) say this argument cannot be taken for granted because, under conditions of extensive trade liberalisation, economies of scale do not require a large domestic market.Back.
April 2000