Economic Determinants of Foreign Direct Investment by Multinational Firms: A New Institutional Analysis

Transcription

1 Economic Determinants of Foreign Direct Investment by Multinational Firms: A New Institutional Analysis Hikari Ishido Faculty of Law and Economics, Chiba University October

2 Preface This book aims at providing both a theoretical framework for economic determinants of foreign direct investment (FDI) by multinational firms (MNFs), and an empirical application of the framework to the empirical cases of the manufacturing operations. The theoretical section of this research draws on new institutional economics which views existing institutional or organizational settings as endogenous variables, and stresses that the robust and hence path-dependent nature of FDI should be brought to the fore of the analysis of FDI determination, given market imperfections represented by under-utilization of firms proprietary assets, possibilities of efficiency improvement over time, and irreversibility of investment. These considerations of market imperfection necessitate the distinction between initial FDI and subsequent or incremental FDI. The methodological framework for the empirical part of this monograph draws on a new institutional perspective which views firms governance structure as endogenous variables. A firm-level case study approach is adopted and a cost analysis is undertaken in which the extent of comparative advantage generated from the firms production operation is quantitatively estimated. The result of the cost analysis shows that apart from the market-mediated benefit of cheap factor costs, specialized firms and those producing more standardized products both gain higher non-market benefits through agglomerated and continuous production in one location. These benefits reinforce the robustness of their FDI operations irrespective of the fluctuation in macroeconomic parameters. Pro-FDI industrial policy should be formulated with due consideration to these institutional characteristics. I convey my sincere gratitude to Prof. Machiko Nissanke, who has always provided me with timely and well-focused guidance during my research period spent at the School of Oriental and African Studies, University of London. From the beginning of my research she was helpful in directing me along the right track. I also thank the Institute of Developing Economies of the Japan External Trade Organization for its financial support which enabled this study. My family has been a constant source of encouragement. Various forms of support and prayers from my family remain my debts. 2

3 CONTENTS Chapter 1 Introduction 1.1 The Rise of Foreign Direct Investment 1.2 Objective and Methodology of the Study 1.3 Structure of the Research Chapter 2 Economic Determinants of Foreign Direct Investment: A Literature Survey 2.1 Definition and Characteristics of FDI and MNF 2.2 Existing Theoretical Arguments Related to the Determinants of FDI Macroeconomic Approaches International Trade Theories The Theory of the Firm Other Theoretical Contributions 2.3 Conclusions Chapter 3 A Microeconomic Approach to the Choice of Foreign Direct Investment 3.1 Introduction 3.2 Theoretical Concepts for the Model Analysis Introduction Market-Mediated Factors Non-Market-Mediated Factors 3.3 The Models Introduction Model 1: Exporting Model 2: Downstream Vertical FDI Model 3: Full-Set FDI Model 4: Horizontal FDI Model 5: Exporting to Country B with a Rival Firm Model 6a: Licensing Model 6b: Licensing of firm-specific assets with external synergy effect Model 6c: Self-Licensing of Firm-Specific Assets with Internal Synergy Effect Model 7: Full-set FDI with a Rival Firm 3.4 Simulation of the models Simulation of Models 1 through Simulation of Models 5, 6b and 7: Choice of the Most Profitable Mode of Operation 3.5 Implications of the Modelling Analysis in the Context of FDI by Japanese MNFs in Malaysia Appendix 3A Simulation and Interpretation of the Models 1-4 Chapter 4 A New Institutional Perspective on the Determination of Foreign Direct Investment 4.1 Introduction 4.2 Research issues 4.3 Research hypotheses and research methodology A classification of economic transactions Research hypotheses and corresponding proxy measures Synergy effect and institutional coherence Cost Analysis as a Methodological Framework 4.4 Research design for empirical analyses Country analysis Overview of FDI in Malaysia by electronic and electrical MNFs Choice of case study firms Appendix 4A A formal representation of synergy effect Appendix 4B List of participants to World Semiconductor Trade Statistics Chapter 5 A Country-Level Analysis of Locational Advantages for Foreign Direct Investment 3

4 5.1 Introduction 5.2 Analysis of FDI inflows in Malaysia 5.3 Malaysia s locational advantages Industrial master plans Dominance of electrical and electronic sectors in Malaysia Policy incentives for FDI Costs of FDI in Malaysia Impact of Asian financial crisis revisited 5.4 Conclusions Appendix 5A A formalisation of Malaysian Government s Development Strategy through attracting FDI Chapter 6 Determinants of Foreign Direct Investment: A Firm-Level Approach 6.1 Introduction 6.2 Existing surveys on determinants of FDI by Japanese manufacturing MNFs 6.3 Case study 1: Firm X Overview of Firm X s operation Firm X in Malaysia A cost analysis Results and interpretation of the cost analysis The impact of Asian Financial Crisis in Conclusions of the case study of Firm X 6.4 Case study 2: Firm Y Importance of institutional difference Institutional characteristics of Firm Y s headquarters operation Firm Y s operation in Malaysia A cost analysis 6.5 Case study 3: Firm Z Importance of product characteristics Overview of Firm Z s headquarters Production of air conditioners by Firm Z s headquarters in Malaysia A Cost analysis Results and interpretation of the cost analysis 6.6 Conclusions Chapter 7 Summary and Conclusions 7.1 Summary of Each Chapter and Conclusions 7.2 Prospects for Future Research Appendix A model of investment determination with synergy effect 4

5 Chapter 1 Introduction 1.1 The Rise of Foreign Direct Investment The world economy is globalizing, and economic activities increasingly take place in a supranational dimension (Table 1.1). Industrial products, once manufactured in stand-alone factories, are now manufactured with visible materials, physical assets and invisible technical know-how, and these inputs are sourced from nearly every part of the globe (Dicken, 1998). Factories themselves are also frequently located outside of their home economies, even though these factories are viewed as internal organisations of a single business entity (Dunning, 1992). This type of global economic activity, which stretches across borders, has been labelled foreign direct investment (FDI). The scale of FDI in Asian economies has increased relative to economies in other parts of the world (Table 1.2). The performance of FDI undertaken by multinational firms (MNFs) as supranational entities is therefore one of the key phenomena of economic globalisation, and is a timely and important topic for research. In the 1950s, the countries of Association of South East Asian Nations (ASEAN) adopted industrialisation policies in pursuit of rapid economic development. Both import substitution and export-oriented industrialisation policy measures were adopted by national governments in ASEAN. Since the late 1980s, these economies experienced a period of economic take-off, with high growth rates of sometimes more than 10 percent per annum (Table 1.3). This rapid economic growth was sustained, to a large degree, by international capital inflows, employment generation and technology transfer, all of which were facilitated by the FDI surge into these ASEAN economies undertaken by MNFs. As with many of its neighbours, Malaysia, the case country in the present study, has enjoyed FDI-driven economic development over the past three decades. 5

6 Portfolio investment inflows and bank lending to Asian countries affected by the so-called Asian financial crisis of 1997, are two other important types of capital flows (Table 1.4). It is notable that, with the exception of Indonesia, FDI inflows were positive during and after the Crisis period, while portfolio investment and bank lending exhibited net outflows. The unexpected occurrence of the Crisis in mid-1997, triggered by the sharp devaluation of Thai baht, caused a net outflow of portfolio investment from the Thailand as well as from other ASEAN economies including Malaysia. However, FDI flows largely stayed positive. MNFs, as foreign direct investors in ASEAN countries, have also been streamlining their production operation in response to the changing economic circumstances following the Crisis and free trade negotiations involving the ASEAN region. However, the difference in growth rates and sustainability of FDI as opposed to portfolio investment and bank lending raises an interesting question as to the factors behind the performance of FDI as opposed to other types of capital flows. A systematic theoretical and applied investigation into the factors contributing to these differences is one clear reason for further research into FDI. Another reason to carry out research into FDI has to do with the performance of FDI on the firm making the investment, as the main objective of FDI by MNFs is to capture benefits in cost terms exemplified by the existence of cheap labour force in the ASEAN economies. However, foreign governments often seek different benefits from FDI, including technology transfer, skill building of the labour force or other benefits. 6

7 Table 1.1 Selected Indicators of FDI and International Transaction, (US$ billion, percent) Value at current prices (US$ billion) Annual growth rate (Percent) Item FDI inflows FDI outflows FDI inward stock 719 1, FDI outward stock 568 1, Cross-border M&As Sales of foreign affiliates 2,465 5,467 17, Gross product of foreign affiliates 565 1,420 3, Total asset of foreign affiliates 1,888 5,744 30, Export of foreign affiliates 637 1,166 3, Employment of foreign affiliates 17,45 23,721 54, (thousands) 4 0 GDP at factor cost 10,61 21,475 36, Gross fixed capital formation 2,236 4,501 7, Royalties and fees receipts Exports of goods and non-factor services 2,124 4,381 9,

8 - not available. Source: UNCTAD (1998, 1999, 2000a, 2001, 2003). 8

9 Table 1.2 FDI Inflows in Selected Asian Economies, (US$ million) Host economy (Annual average) China 3,105 11,156 27,515 33,787 35,849 40,180 44,236 43,751 40,400 40,772 46,878 52,743 53,505 Korea, Republic of ,357 2,308 3,088 5,215 10,340 10,186 3,683 2,941 3,752 Taiwan Province of China 1, ,375 1,470 1,864 2, ,926 4,928 4,109 1, Indonesia 746 1,777 2,004 2,109 4,500 6,194 4, ,550-2, Malaysia 1,605 5,183 5,006 4,581 5,816 7,296 6,513 2,700 3,532 5, ,203 2,474 Philippines , Singapore 3,592 2,351 5,016 8,550 7,206 8,984 8,085 5,493 6,984 6,390 15,038 5,730 11,409 Thailand 1,325 2,116 1,726 1,343 2,000 2,405 3,732 7,449 6,078 2,448 3,813 1,068 1,802 Memorandum: United States 49,088 18,885 43,534 45,095 58,772 76,453 90, , ,97 281,11 159,46 62,870 29, Japan 556 2, ,224 3,268 12,741 8,187 6,241 9,239 6,324 Western Europe (Average) 3,693 4,769 4,660 4,357 6,821 5,553 6,381 15,189 26,962 35,176 18,441 19,012 15,512 North America (Average) 27,337 11,831 24,151 26,786 34,798 41,426 49, ,06 172,22 93,474 41,950 18,176 98, Africa (Average) Latin America and Caribbean (Average) ,151 1,477 2,080 2,757 2,154 2,203 1,284 1,243 Asia a (Average) ,113 1,319 1,465 1,739 1,890 1,992 2,078 2,989 2,330 1,966 2,232 9

10 The Pacific (Average) Central and Eastern Europe (Average) ,229 1,501 1,659 1,816 1,388 1,644 1,104 a Excluding Japan. Source: Calculated from UNCTAD (1998, 2000a, 2001, 2004). 10

11 Table 1.3 GDP Growth Rates of Selected Asian Economies, (percent) Country average China South Korea Taiwan Indonesia Malaysia Philippines Thailand Source: Calculated from IMF, International Financial Statistics, various years, Institute of Developing Economies, Warudo Torendo (World Trend), various issues. Table 1.4 Portfolio investment inflows and bank lending to crisis-affected Asian countries a, (US$ billion dollars) Net portfolio investment Bank loans and other a Referring to Indonesia, Malaysia, Philippines, Republic of Korea and Thailand. Source: IMF, International Capital Markets (Washington DC, September 2000). 11

12 As Tejima (1998) points out, MNFs aim to construct the most efficient international production network driven by the motivation of profit, whereas host countries crave for FDI for the full-set of production facilities, which become a full package within their own territories. In other words, MNFs shift, in certain economic circumstances, only their labour-intensive and therefore low-value-added production processes to foreign economies, in spite of host governments policies designed to attain economic development through the establishment of all-encompassing domestic industries. It is the right of MNFs to decide whether to undertake FDI or not. Depending on the policy circumstances, once FDI has been undertaken, a decision on the type of the MNFs operations to shift to foreign economies rests with the MNFs themselves. For example, Japanese MNFs shifted much of their production facilities abroad, mainly to the neighbouring East Asian (including ASEAN) economies, after the appreciation of the Yen in the wake of the Plaza Accord in Unlike official development assistance, the decisions of MNFs regarding FDI behaviour has been motivated primarily by their profit seeking objectives obtained through cost reduction by FDI in ASEAN economies. The nature of FDI undertaken by MNFs and its effect on an Asian country s economic development in the face of globalisation are other important theoretical and empirical research. 1.2 Objective and Methodology of the Study The overarching objective of the present monograph is to make a fresh enquiry into the nature of FDI determination, from both theoretical and empirical perspectives. FDI has customarily been treated in the macroeconomic framework as one form of international capital flow. Aggregate figures have therefore been employed for 12

13 conventional empirical investigations of FDI. Since FDI represents both the flow of monetary assets (i.e. capital in a simple sense) and technological/managerial assets possessed by individual business firms (i.e. firm-specific assets ), conventional investigations into FDI fail to account for firm-specific assets that make up a critical part of FDI. The standard neoclassical framework has treated firm behaviour in the microeconomic branch of theory, and has led to a disjointed approach to explaining FDI, since FDI is considered macroeconomic in statistics. Since the actual phenomena of FDI stand on the crossroad of international trade theory, industrial economics and new institutional economics, an attempt is made in this research to synthesise these three rather disjointed branches of economic theory. The presumption here is that both industrial (or product) and institutional (or organisational) characteristics significantly influence firms FDI behaviour as an international economic phenomenon. A microeconomic or firm-level investigation into FDI determination from this synthesised analytical perspective is therefore deemed to shed new light on the relevant theoretical fields. Perfect market assumptions underlie the analytical foundations of the conventional neoclassical theory of firm behaviour. Empirically, however, firms are known to engage in production activities under market imperfection. They operate in their value-adding activities with incomplete knowledge of what constitutes the optimal set of corporate decisions. In general, imperfect information, arising from economic agents bounded rationality in perception, calculation and action renders market functioning imperfect. In other words, price signals do not reflect the true opportunity costs of the raw materials, factors of production and final products/services involved. The market-entry mode of FDI, too, may be chosen as a response to market imperfection, which would make the causes and effects of FDI very different from 13

14 under the conventional theories of FDI. The present research therefore employs an exploration of FDI based on the existence of market imperfections. In formulating a theoretical foundation for empirical research into FDI determination, an extension of the neoclassical framework is made in relation to the other two types of market entry modes, exporting and licensing. In this context, firm-level analyses of FDI are conducted empirically, in which three firms operating in Malaysia s electronic and electrical industries are closely studied with a view to identifying factors influencing the choice of FDI. The present research employs Dunning s (1992) so-called eclectic framework as a useful taxonomy of FDI determinants according to the source of comparative advantages conducive to the choice of FDI. More specifically, the ownership-specific advantage, locational advantage and internalisation advantage are considered pertinent to a firm s FDI decision. With due consideration to this eclectic framework, an attempt is made to identify sources of comparative advantage which account for MNFs decisions to engage in FDI. MNFs motivations for undertaking FDI are also influenced by host economies industrial policy formulation regarding FDI. International free trade and investment regimes/negotiations involving the ASEAN region, including the concept of ASEAN Free Trade Area (AFTA) and Asia Pacific Economic Cooperation (APEC), should also be put under the scope of the analysis. It is therefore essential to be concerned with host governments historical and current policy prescription in the international context of trade and investment liberalisation, before undertaking the firm-level study. In this light, a country analysis or Malaysia s policy environment in favour of attracting FDI, precedes the firm-level analyses. 1.3 Structure of the Research 14

15 The structure of this research is as follows. Chapter 2 overviews the literature addressing economic determinants of FDI for the analytical setting of the present study. Based on this setting, Chapter 3 proposes a set of microeconomic models with a view to capturing economic determinants of FDI in the empirical investigation. These models are devised to address the main issue of how to coherently treat economic motivations of FDI by MNFs, with reference to but not limited in scope to the neoclassical framework of international trade theories, industrial economics and new institutional economics. In Chapter 4, a methodological framework for undertaking cost analyses is provided. Counter-factual arguments are made in both the cost analyses, which attempt to measure the comparative advantage of the FDI decision, and related complementary analyses which support the views drawn from the cost analyses. Analytical concepts are introduced for the operationalisation of these theoretical bases. Chapter 5 is dedicated to a country-level case study. The chapter first analyses FDI inflow in the case-study country, Malaysia. Then the discussion of the country s locational advantages in attracting FDI follows, with a particular reference to its industrial policy framework. Chapters 6 undertakes empirical studies of the FDI determination made by electrical/electronic MNFs based in Malaysia. In this chapter, a cost analysis is developed and undertaken. In the cost analysis, the degree of comparative advantage generated from each case study MNF s production operation in the country is quantitatively estimated. Chapter 7 summarises the main findings of this research and discusses future research directions. It is expected that this monograph contributes to shedding some light, in both theoretical and empirical terms, on the essential role of market imperfection in the determination of FDI from a new institutional perspective. 15

16 Chapter 2 Economic Determinants of Foreign Direct Investment: A Literature Survey 2.1 Definition and Characteristics of FDI and MNF The theoretical literature on the phenomenon of FDI has several strands, with various analytical ramifications and corresponding emphasis points. In the present chapter the mainstream economic theories on the determinants of FDI, as distinct from its impacts, are explored. The central question of the present research concerns what determines whether a firm decides to directly establish production facilities abroad rather than export its product or licence its knowledge to overseas affiliated firms (or subsidiaries). Macro- or micro-based theories that have aspects pertaining to this question are investigated selectively in the present chapter. The present chapter first furnishes the arguments on the efficacy and bearing of macroeconomic view on determinants of FDI, which is the traditional viewpoint. However, microeconomic considerations, which are exclusively focused on individual decision-making agents, namely firms, governments and individuals, are also an important consideration in the determinants of FDI. The latter part of the present chapter therefore focuses on microeconomic theories, under the headings of international trade theories and the theory of the firm. Before reviewing the relevant literature, it is first important to clearly define the concepts of FDI and MNF. International capital flows in general can be categorized into official development assistance (ODA) flows and international private capital flows. FDI falls under international private capital flows, together with portfolio investment flows and commercial bank loans. The criterion used to distinguish FDI from portfolio investments and commercial bank loans is that FDI is chosen when an investor or an economic agent based in one country (or home country) acquires an asset in another country (or host country) with an intention to manage, or directly control that asset. In the case of portfolio investment and commercial bank loans, in contrast, 16

17 investing agents do not have the intention to manage, or directly control assets, when it acquires the foreign asset or disburse loans (WTO, 1996). Similarly, one of the most commonly accepted legal definitions of FDI by the IMF is the investment that reflects the objective of obtaining a lasting interest by a resident entity in one economy in an enterprise in another economy (IMF, 1993). Both of these definitions indicate that in the conduct of FDI, as the term itself alludes to, investors are interested in directly controlling the acquired asset. For practical purposes, a minimum of 10 % ownership and some management role are required as a threshold for FDI (IMF, 1993). 1 There are several axes along which FDI is categorised. FDI can first be categorised by the organisational structure of the investment, with a distinction between FDI that is vertically integrated (which can be further subcategorised into backward-integrated and forward integrated) and FDI that is horizontally integrated. Capital participation is another important categorisation, e.g., green-field FDI, referring to an FDI from scratch 2, mergers and acquisitions (M&A) and joint ventures 3, as well as special participation types such as keiretsu relationships in Japan. 4 In terms of financial arrangements, FDI comprises three types: (1) new equity from the parent company in the home country to the subsidiary in the host country; (2) reinvested profits of the subsidiary; and (3) long and short term net loans from the parent to the subsidiary (WTO, 1996). Although these definitions and categorisations of FDI are idiosyncratic and overlap considerably 5, the straightforward and conventional definition of FDI, e.g., a source of funds to create new and additional productive assets obtained by an economic 1 There is no special reason for the choice of the threshold level, i.e., 10%, to distinguish FDI from portfolio investment. The threshold adopted by the IMF is merely for practical purposes. 2 This has been the major form of FDI undertaken in ASEAN economies since the 1970s by U.S and Japanese MNFs. 3 Both of these capital participation modes have been prevalent in the US and European markets in the twentieth century. 4 Keiretsu are a unique type of business conglomerate prevalent in Japan, which can be viewed as one type of FDI if formed across national boundaries. 5 For practical difficulties in the measurement of FDI, see South Centre (1997). 17

18 agent with the intention of directly controlling those acquired assets, 6 is sufficient for the present study. Further scrutiny of definitions and categorisations are beyond the scope of the present research. What is notable here is that FDI can be recognized as the investment made outside the home country of the firm, but inside the firm 7 (Dunning, 1992), and FDI consists of a package of assets and intermediate products, such as capital, technology, management skills, access to markets and entrepreneurship 8. This view effectively brings up the need for due consideration to MNFs internal organisations, which is highlighted in the subsequent empirical portion of the present research (i.e., Chapter 6). 2.2 Existing Theoretical Arguments Related to the Determinants of FDI Logically, a salient inquiry into determinants of FDI should consider why a particular firm has decided to undertake a project that is foreign, direct and investment. This question, however, has not been central in strands of existing research. Nevertheless, determinants of FDI have implicitly or explicitly been touched upon in various, albeit idiosyncratic, approaches with a significant difference in weight of alleged importance for different possible determinants. There have thus been various ways of categorising economic theories as to determinants of FDI (see, e.g., Casson, 1987; Dunning, 1992; Economic Planning Agency, 1990; Hara, 1992; Amano, 1986 and Horaguchi, 1992). The following section reviews economic theories on determinants of FDI, and groups these theories into four categories: (1) macroeconomic approaches; (2) international trade theories; (3) the theory of the firm; and (4) other theoretical contributions. These headings are used for the sake of analytical convenience, and there 6 Internationalisation of firm production and FDI are therefore seen as loosely equivalent to each other. 7 Dunning (1992) also points out that in the case of portfolio investment, as opposed to FDI, specific assets and intermediate products are transferred between two independent economic agents through markets. 8 These assets are generally referred to as firm-specific assets. The main characteristics of firm-specific assets are discussed in Chapter 3. 18

19 is a considerable overlap among these Macroeconomic Approaches FDI exhibits several properties of international financial flows. Although FDI should be distinguished from portfolio investment, FDI by definition entails financial flows, and in this context, determinants of FDI have been discussed within the framework of macro-based international finance. Aliber s (1970, 1971, 1983) arguments represent the orthodox macroeconomic approach to the determinants of FDI. He focuses on why MNFs finance their foreign assets in their domestic currencies, and explains the choice of FDI in terms of monopolistic advantages, i.e. the ability of MNFs from countries with strong currencies to raise capital more cheaply in foreign markets compared to competitors from countries with weak currencies. 9 He attributes the MNF s motivation of FDI to the structural failure, or market imperfections, of international financial markets, thereby allowing the MNF in a stronger financial position with the acquisition of foreign assets. Aliber s theory addresses why countries, not MNFs, might shift their international investment status over time, and adopts the realistic assumption of market imperfection in the international financial market mechanism. Assuming a perfect market mechanism, it should not matter in which currency economic agents hold their assets, since the exchange rate across any currency would reflect at any given moment the costs and future profitability of holding assets in that particular currency. Market perfection would therefore make any sort of asset portfolios equally profitable (or unprofitable) through perfect arbitrage. This presumption is at odds with reality. Aliber s theory is useful in explaining U.S. FDI in Europe, in the form of M&A, in the 1950s and 1960s 9 If the holder of financial assets in a currency has the chance of getting foreign exchange gains, the currency is called a strong currency. If, on the other hand, a currency bears the risk of depreciation relative to other currencies, it is called a weak currency. Therefore, assets in strong currencies can expect higher returns on investment. 19

20 (Buckley and Casson, 1976). However, in Aliber s theory, it is difficult to distinguish FDI from portfolio investment. 10 His theory explains FDI in terms of higher returns on investments, but it cannot explain why the particular form of FDI was preferred to portfolio investment. Moreover, it does not explain the two-directional nature of FDI between two countries: FDI flows from country A to country B are empirically observed in parallel with FDI flows from country B to country A, but the approach fails to explain this phenomenon, as only firms from the country with the stronger currency should undertake FDI. Similarly, this approach is unable to explain FDI flows within the same currency areas (Buckley and Casson, 1976). These limitations are primarily due to the very fact that his analysis falls under the theoretical framework of macro-based international finance, which only views FDI as an aggregate phenomenon. This could be misleading since it overcasts the role of the firms, or MNFs as the actual players of FDI, i.e., the reality is that MNFs headquartered in country A (B) undertake FDI, by establishing their affiliated firms in country B (A). Tobin (1969) first introduced the concept of what is often referred to as Tobin s q, defined as the ratio of the financial market valuation of reproducible real capital assets to the replacement cost of these assets. According to his approach, if the value that shareholders place on capital assets is higher than the opportunity cost of the assets (meaning Tobin s q is greater than unity), then the decision should be made to invest in the capital assets. In spite of the difficulty of actually measuring the level of q, this can be considered as a macroeconomic version of the determinants of FDI, since FDI is one type of capital investment. Although Tobin s approach primarily focuses on shareholders aggregate expectations on the firms future profitability based on 10 Empirically, portfolio investments are characterized by large year-to-year fluctuations, or short-term fluctuations, as compared to FDI (UNCTAD, 1998). This implies that determinants of FDI should be distinguished from those of portfolio investments. 20

21 macroeconomic fundamentals, it seems to have similarity with the analysis of economic determinants of FDI in that individual firms are focused on as units of the analysis. However, macroeconomic approaches in general fail to explain FDI flows between two nations with similar macroeconomic fundamentals, such as interest rates and inflation rates. Just as the phenomenon of intra-industry trade cannot be encapsulated in macro-based analyses, the empirically observed phenomenon of intra-industry FDI, denoting, for instance, FDI in the U.S. electronic industry by Japanese electronic MNFs or vice versa, cannot be captured by the macroeconomic framework or theories of international finance. The bi-directional nature of FDI seems to be beyond the analytical capacity that the macro-based framework possesses. Another way of stressing the limitation of the macroeconomic approach is that it stays away from non-financial aspect of FDI, namely the transfer of resources specific to respective firms, or firm-specific assets. As per the definition of FDI, the MNF transfers its capital in the form of tangible assets and intangible assets 11 which are unique or specific to the MNF. Furthermore, these assets are directly controlled and utilized by the MNF for its production operations. Although macro conditions are surely one element of consideration, a firm s specific tangible and intangible assets are arguably just as important. In sum, the governing view of macroeconomic theory that FDI is a form of financial flow neglects the transfer of firm-specific assets International Trade Theories FDI is to a large extent a phenomenon of international economics, which has since its inception been dominated by the theory of determinants of international trade between nation states. One of the most basic concepts in international trade is the 11 These include specific designs of physical capital assets (such as machines), technological know-how or patents, managerial skills in the form of written manuals on management for intra-firm use, and customer information. 21

22 principle of comparative advantage, first introduced by classical economist David Ricardo. The core purpose of the comparative advantage concept is to explain geographical differences in production and trade in terms of difference in productivity of two factor inputs, i.e., labour and capital. It expressly features the role of technological capability possessed by countries to utilize factor inputs for production of tradable goods. The notion of comparative advantage could be useful in explaining FDI, as greater factor productivity in a certain country could lead MNFs to invest in this country to achieve production advantages. However, the theory of comparative advantage takes the technological difference between countries as exogenously given and hence does not explain why such differences arise. Extensions of the theory have led to the concept of dynamic comparative advantage, which denotes acquired comparative advantage, through accumulation of learning experiences of value-adding production over time. However, the unit of analysis in these extensions is the country as opposed to MNFs, which constitutes another limitation to this approach. Notwithstanding the empirical importance of the role which FDI plays in the global economy, international trade theory precludes the phenomenon of FDI undertaken by MNFs. Heckscher (1949), Ohlin (1933, 1967), Stolper and Samuelson (1941), and Samuelson (1948, 1953) formulated and expounded the notion of comparative advantage within the realm of international trade theory by constructing a general equilibrium model with two factors, two countries and two goods, which is the so-called Heckscher-Ohlin-Stolper-Samuelson (HOSS) model. The HOSS approach, as the precursor to several derivative theories, is based upon the factor endowments of countries. A country which is relatively capital abundant will export those goods which use capital intensively, and a country which is relatively labour abundant will export those goods which intensively use labour as a factor of production. In the process, factor 22

23 prices tend to be equalised. In sum, the fundamental economic notion of substitutability under the scarce resource assumption is expressly incorporated. Four major criticisms though could be raised against international trade theory when considering determinants of FDI. First, the possibility of factor movements and subsequent intra-firm trade are excluded from the arguments in the HOSS theory. FDI as international capital movement is thus precluded from analysis, in spite of its increasing empirical importance. In other words, the emergence of MNFs is assumed away under the HOSS orthodoxy. Within this framework, all trade is assumed to be conducted at arm s length market prices between independent firms whose value-adding activities take place within their national boundaries. As it explicitly assumes factors of production to be fixed and immobile across national boundaries, the HOSS theory is marked by the implicit assumption that there is no need for geographical consideration, since transport costs are assumed to be zero. The theoretical consequence of these assumptions is to assume away FDI. Thus, the first criticism against the factor immobility assumption is that it precludes outright the possibility of FDI, or the emergence of MNFs. Second, agents of international trade within the HOSS framework are assumed to be nation states. However, MNFs possessing their production facilities outside their home countries are actual agents of trade and production. In this light the theory precludes the idea that FDI is a prime mover of intra-firm trade. Third, the international trade theory based on the HOSS framework remains space-less, at least at its highest conceptual level, in spite of its explicit concern with trade between geographical regions. This is at odds with real world observation, since geographical distance actually imposes a cost on movement of materials, finished products, or human resources, as well as less tangible factors such as managerial 23

24 know-how, technological skills and other information 12. In response to this criticism, so-called location theorists have been trying to redress this neoclassical assumption of the space-less production activity. The location theory approach explicitly addresses where/how firms locate their production facilities, and when firms decide to start operating abroad, these firms can be viewed as MNFs. In this light, location theories are relevant to determinants of FDI. Most location theories are based on the so called least-cost approach (Dicken, 1992, 1998). As the presage to this approach, Dicken s concern was to identify the optimal location for an individual firm, operating a single plant, given certain assumptions. Dicken regarded transport costs as the initial determinant of industrial location. In this context, the primary concern was to minimise transport costs incurred in collecting the necessary materials and transporting finished products to the market 13. Although transport cost cannot be realistically taken as a firm s primary consideration, since production costs usually well exceed transport costs, Dicken s geographical concern merits mention in its empirical focus on cost minimisation as the basic theoretical principle of economics. Krugman (1980, 1983, 1991, 1993, 1995), followed by Fujita, Krugman and Venables (1999), pioneered the construction of formal microeconomic models addressing the points Dicken raised, under the label economics of geography. This approach, exemplified by Krugman (1981), consists of a simple yet rigorous model, derived from industrial organisation theory, addressing the main question of why and how geographical concentration of manufacturing takes place in one country and concentration of agriculture takes place in another. This is done by explicitly 12 Each of these factors differs in its degree of sensitivity to geographical distance. Dicken (1992) points out that some materials and products in manufacturing industry are costly to move in relation to their value while others are relatively less costly to move. Whatever the case, he continues, the explanation of the geographical distribution of economic activity at the global scale must incorporate the role of geographical distance and transport costs. 13 For instance, heavy materials, the bulk of which is reduced in the process of manufacture, would tend to lead to a production location close to the source of the material. In contrast, manufacturing processes which add to the weight or bulk of the materials would tend to be located at the market (Dicken, 1992). 24

25 introducing the assumptions of increasing returns to scale 14 together with transport costs incurred by manufacturing firms. Stressing the role of pecuniary externality in the model construction, Krugman argues that once the concentration of manufacturing production has started, it will increasingly become more economical for potential manufacturers to base their production facilities near the already concentrated area because it will be cheaper to buy factor inputs, raw materials and intermediate products which the concentrated area provides with lower transport costs. Thus, Krugman s theory of economic geography explicitly employs MNFs as agents of FDI and addresses the issue of why firms pursuing profit maximisation decide to establish production facilities abroad. Although his argument intentionally omits the role of technological spillover effects 15 through FDI by MNFs and focuses exclusively on pecuniary externalities between firms and not inside firms, his models contribute significantly to the understanding of the interplay between economies of agglomeration and FDI decision. The fourth criticism against the HOSS theory in conjunction with FDI determinants states that although the theory recognises the existence of a number of production factors, it actually employs only two, i.e., capital and labour, mainly for analytical simplicity. Based on this two-factor postulate, a country with an abundant supply of capital (labour) will inevitably import labour (capital) intensive products. However, much of the actual international trade, particularly in the manufacturing sector, occurs between countries with similar factor endowments. Leontief (1953) points out in this connection that the most capital intensive country, the U.S., exports labour intensive products, which seems to be paradoxical in the light of HOSS theory. 14 Arthur (1986), in the context of complexity analysis, argues that increasing (and not diminishing) returns to scale constitutes major part of the reality in varieties of empirically observed economic processes over time, including the agglomeration of industrial locations. 15 The technological spillover effect is considered under the original concept synergy in the modelling analysis of Chapter 3. 25

26 Gruber, Mehta and Vernon (1967) gave the solution to this paradox, insisting on the necessity of taking into consideration another, third factor input, viz., firm-specific assets such as technological know-how and management skills. One interpretation of their argument then is that U.S. actually exports technology-intensive products, with the technological assets being crudely subsumed in labour assets. Technological assets, notably in the case of the computer software industry, are created by human labour, and physical capital assets as embodiments of such technological labour become only a secondary concern. Another way of stressing this fourth criticism is to note that created technological assets have the public good characteristics, viz., non-excludability and non-rivalry (Kaul et al., 1999). The HOSS theory as a two-factor input framework consequently lacks this perspective, since neither factor of production has such characteristics. This poses serious limitations to the applicability of the framework to the investigation of economic determinants of FDI by MNFs. Considering the fact that MNFs own firm-specific assets, of which technological assets constitute a part, it is legitimate to introduce a third factor input. 16 And, in view of the second criticism, the third factor input should be firm-specific instead of country-specific. The analytical concept of firm-specific assets therefore becomes relevant in the study of firms production and investment. Other strands of international trade outside of the HOSS framework have attempted a partial analysis on the location of FDI and the subsequent patterns of trade. Vernon s (1966) major contribution in its initial form was to introduce an explicitly locational dimension, applying the notion of product life-cycle. His argument was the first dynamic interpretation of the determinants of FDI 17 (Dunning, 1992) with a 16 Firm-specific assets as the third factor input are introduced in the models constructed in Chapter Dicken (1992) termed MNFs dynamic foreign operations as global shift of production. 26

27 particular emphasis on the technological standardization process. Vernon s starting point was the assumption that each phase of the product life-cycle has important locational implications in terms of both demand and production process itself. Vernon explains where the manufacturing of new products would take place by drawing on elements of location theory, especially the notion of external economies 18. The non-standardised character of the new product and its production process requires communication between producers (including company researchers and engineers) and customers, leading to optimal location for the production facility in the home country, while overseas demand is met by exports. Unlike the static assumption made by the HOSS theory, this initial stage does not last indefinitely. Vernon suggests that firms manufacturing the new product would gradually shift their production facilities from the home country to the overseas market either because it can reduce production and distribution costs, or because there exists a threat 19 to their market position. In this consideration of dynamics, Vernon s so-called product cycle theory is a partial break-through in the international trade framework. According to Vernon, it then follows that the first foreign production will take place in countries characterised by the existence of high-income domestic markets. New foreign plants would come to serve their national markets and thus displace exports from the home country. As the new foreign plants achieve production cost advantages, caused by scale economies, the firm may begin to export to third country markets, and even to the home country 20. Finally, as the production process becomes totally standardised or requires little further elaboration in an engineering sense, production facilities may be shifted to countries with lower costs, mainly developing countries. 18 Weber (1909) termed this as agglomeration economies. 19 Such a threat might derive from the local government attempting to reduce imports through trade barriers (Dicken, 1998). 20 The export back to the home country is often referred to as boomerang effect. 27

28 Akamatsu (1956, 1961, 1962) uses the term flying geese pattern to describe this shift in the factor input intensity of the production process and the subsequent shift in the geographical location of production. Vernon s and Akamatsu s analytical focus is placed on the dynamic nature of both the production process and consequent locational shift in production activities 21. Therefore these approaches are relevant not just to the static pattern, but to the dynamic or sequential process, of the growth of MNFs and FDI. Buckley and Casson (1976) acutely point to four key assumptions, both explicit and implicit, on which these theories stand: (1) difference in demand according to income; (2) existence of communication costs both within the firm and between the firm and the market, which increase in proportion to geographical distance; (3) technological and marketing predictability of change in product specifications; (4) market imperfections, especially in technological know-how. 22 These four assumptions make the product-cycle theory highly applicable to the empirical observation. From the perspective of FDI determination, there are two criticisms 23 made against the product cycle theory. First, it has no actual time dimension, i.e., the question of when in the actual flow of time a firm will establish its production facility abroad is somewhat vaguely treated (Buckley and Casson, 1976; Horaguchi, 1992). The theory can therefore be validated on an ad-hoc basis, but not systematically. Second, the theory does not explain why an MNF needs to establish its production facilities abroad in the first place, instead of licensing out its technology/products which, apart from exporting and FDI, could be deemed a feasible third option by MNFs (South Centre, 21 Vernon s (1974, 1979, 1983, 1986, 1987a, 1987b) later emphasis is more on firms desires to maintain an oligopolistic market structure by erecting to entry (Buckley and Casson, 1976). This point is examined in Chapter The empirical validity of these assumptions is studied in Chapter Regarding criticism against the efficacy of FDI on host economies, Lewis and Sappington (1991) argue, cost-reducing technological change leads the firm to produce the input itself more often. The firm s profit depends on whether the subcontractor s skills are idiosyncratic or transferable. In the latter case, technological progress can even be detrimental to the firm and to society as a whole. 28