Abstract:
As Foreign Direct Investment (FDI) plays an important role in international economics, there are many theories were developed to explain the drivers of FDI. This paper presents a review of theoretical models of FDI concluding (1) early studies of determinants of FDI, (2) neoclassical trade theory, (3) ownership advantages, (4) aggregate variables, (5) the ownership, location and internationalization advantage framework, (6) horizontal and vertical FDI model, (7) the knowledge-capital model, (8) diversified FDI and risk diversification models, and (9) policy variables. The relevant determinants of FDI are derived from each theory. Besides, empirical studies demonstrated the vital role of these determinants in reality. The paper indicates that there is variety of theoretical explaining FDI. Therefore, the analysis of drivers of FDI should not be based on a single theoretical model. Instead, it should be more broadly explained by a combination of determinants from several theoretical models such as agglomeration economics, market size, cost and risk variables, protection, and policy factors.
Keywords: Theoretical model, Foreign Direct Investment, FDI determinants.
The growth of international business and FDI led to extensive research on the phenomenon of multinational enterprises (MNEs) and FDI. There is not only one single theory related to FDI, but a variety of theoretical models attempting to explain FDI and the location decision of multinational firms. Early empirical research was mainly conducted in form of field study with only limited theoretical foundation. A theory of FDI was developed independently based on a trade theory perspective. Until the 1960s, descriptive analysis dominated. While econometric analysis started to emerge in the next decade. Various combinations of research setups are possible in studies based on secondary statistics. FDI of a country or a group of countries (these can be developing countries, developed countries or both) can be analyzed using cross-section, time-series or panel data in an aggregated or disaggregated form. Determinants can be macroeconomic factors, microeconomic factors or a combination of both. In this paper, author focus on reviewing the development of general theoretical models and empirical studies of drivers of FDI.
2. Theoretical Models and Summary of Empirical Studies of Determinants of FDI
2.1. Early Studies of Determinants of FDI
In early empirical studies, typically based on questionnaires, companies were asked to identify their reasons for the initial investment decision. Major contributors analyzing FDI in general were Robinson (1961), Behrman (1962), Basi (1966), Kolde (1968), Wilkins (1970) and Forsyth (1972), Brash (1966), Deane (1970), Forsyth (1972) Andrews (1972). Variety of factors such as marketing factors, trade barriers, costs factors and investment climate were these studies looked at. The consensus is that marketing factors, in particular market size, market growth and maintaining market share, but also dissatisfaction with existing market arrangements, were the main determinants of FDI. However, cost factors, especially the availability of labor and raw materials, lower production costs and financial inducements by the government, were seen as equally important in some of the studies. Political stability was the most important determinant of FDI found by Basi, while foreign exchange stability and a positive attitude to foreign investment were other notable factors. Wilkins (1970) found local competitive threat and lower costs the predominant reasons for foreign investment.
2.2. Determinants of FDI according to the Neoclassical Trade Theory
The first theoretical attempt to explain FDI was based on the Heckscher-Ohlin model of the neoclassical trade theory where FDI was seen as part of international capital trade. The Heckscher–Ohlin model was based on a 2×2×2 general equilibrium framework with two countries (home and foreign), two factors of production (usually capital and labour) and two goods, assuming perfectly competitive goods and factor markets, identical constant returns to scale production functions, zero transport costs and factor endowments that are such as to exclude specialization. According to Jasay (1960), MacDougall (1960) and Kemp (1964), capital was expected to move to the country with higher capital returns. However, countries could manipulate capital returns and capital flows by imposing taxes on internationally mobile capital to enhance their welfare. Firms from countries with currencies with less fluctuation in value (harder currencies) could borrow money in countries with softer currencies at a lower interest rate than host country firms due to their lower risk structure. Foreign firms could therefore capitalize the same stream of expected earnings at a higher rate than host country firms, giving them a reason to invest in the host country.
Neoclassical approach was criticized for its limited ability to explain FDI flows by Hymer (1976) and Kindleberger (1969). They argued that the assumption of perfect competition in neoclassical theory could not explain FDI, which - in their view - needed structural market imperfections to flourish.
2.3. Ownership Advantages as Determinants of FDI
Regression analysis became a popular approach from the 1970s onwards and was based on firm-level data or on the foreign share in the domestic industry.
Numerous empirical studies such as those of Dunning and Buckley (1977), Swedenborg (1979), Lall (1980), Saunders (1982), Owen (1982), Blomstrom and Lipsey (1986) have substantiated the theoretical belief that ownership advantages are significant determinants of FDI, showing that factors such as Reseache and Development (R&D) and advertising expenditure, managerial resources, technology, capital intensity, labour skills, firm size, scale economies and experience had an effect on FDI or MNE activity.
2.4. Aggregate Variables as Determinants of FDI
While one approach is to view FDI as related to monopolistic advantage, another is to test the effect of aggregate variables such as market size and market growth on FDI - even without relating those variables to specific theoretical models. Scaperlanda and Mauer (1969) finding a significant relationship between Gross National Product and investment but no significant effects of market growth and trade barriers on FDI. In contrast to Scaperlanda and Mauer, empirical studies such as those of Goldberg (1972), Lunn (1980), Davidson (1980) showed that market size, market growth and trade barriers could potentially be important determinants of FDI and should thus be incorporated into the theoretical models explaining FDI.
2.5. Determinants of FDI in the OLI Framework
The classical model for determinants of FDI begins from the earlier research work of Dunning (1973, 1981) which provide a comprehensive analysis based on ownership, location and the internationalization (OLI) paradigm.
Dunning (1993) describes three main types of FDI based on the motive behind the investment from the perspective of the investing firm. The first type of FDI is called market-seeking FDI, whose aim is to serve local and regional markets. It is also called horizontal FDI, as it involves replication of production facilities in the host country. A second type of FDI is called resource-seeking: when firms invest abroad to obtain resources not available in the home country, such as natural resources, raw materials, or low-cost labor. The third type of FDI, called efficiency-seeking, takes place when the firm can gain from the common governance of geographically dispersed activities in the presence of economies of scale and scope.
Following the studies of Dunning, there have been numerous studies analyzing factors of FDI linked to ownership, location and internationalization advantages in different countries such as those of Santiago (1987), Lall and Siddharthan (1982), Ray (1989), Wheeler and Mody (1992), Kogut and Chang (1991), Drake and Caves (1992), Culem (1988), Barrell and Pain (1996), Biswas (2002), Love and Lage Hidalgo (2000), Milner and Pentecost (1996). The reasons vary in different host countries. For example, Barrell and Pain found the Gross National Product (GNP) level and growth, R&D expenditure, production costs in the USA relative to overseas and profits to positively affect outward FDI, while exchange rate appreciation was a reason of a postponement of the investment.
Schneider and Frey (1985) indicated that a model mixing political and economic determinants such as GNP per capita, real GNP growth, inflation, balance of payments deficit, wage costs, skilled workforce, political instability, government ideology, and bilateral and multilateral aid explained FDI in 54 less developed countries better than purely economic or political models.
Study of Cheng and Kwan (2000) strengthens the argument that agglomeration effects (positive externalities generated by localization of industry) matter when they looked at Chinese FDI. They indicated farther evidence for the self-reinforcing or “positive feedback” effect of FDI. Lagged FDI, large regional market, good infrastructure, preferential policy and wage costs were found to be significant, while education was not.
In summary, empirical studies testing the OLI framework have found FDI to be determined by a combination of ownership advantages, market size and characteristics, factor costs, transport costs, protection and other factors including regime type, infrastructure, property rights and industrial disputes.
2.6. Determinants of Horizontal FDI according to the Proximity-Concentration Hypothesis and Vertical FDI according to the Factor-Proportions Hypothesis
The new trade theory is built on the industrial organization models, including OLI and internationalization theory, and followed the tradition of microeconomic theoretical models of Hymer, and Kindleberger and Caves. This theory proposed an alternative framework for analyzing FDI and Multi-National Enterprise (MNE) activity. It combined the ownership and location advantages with technology and country characteristics. Knowledge capital was considered as the ownership advantages, while location advantages consisted of country size and moderate to high costs for horizontal firms, and low trade costs, stages of production with differing factor intensities and countries differences in relative factor endowments for vertical firms. Internalization advantages only augment owing to the joint-input property of knowledge capital.
Some studies following this direction were those of Helpman (1984, 1985), Ethier (1986), Markusen (1987a), Horstmann and Markusen (1992), Dixit and Grossman (1982), Sanyal and Jones (1982), Deardorff (2001) and Jones and Kierzkowski (1990, 2001, 2005), Brainard (1993a, b, 1997), Eaton and Tamura (1994) and Ekholm (1998).
In summary, the proximity–concentration hypothesis was vigorous when tested in empirical studies. Market size, transport costs and trade barriers increased FDI, while factor were only concerned in some cases. The results substantiated the idea that MNEs were firms with ownership advantages.
2.7. Determinants of FDI according to the Horizontal FDI, Vertical FDI and Knowledge-Capital Model
The two viewpoints that explained vertical firms and horizontal firms independently were integrated in a knowledge-capital model by several studies of Markusen and al. They built the model with one good with constant returns to scale and a second one with plant- and firm-level scale economies which allows for differences in relatives endowments, country size, high and low cost of transport, and optional FDI ban. Different types of firms could exist. Only national firms existed in both countries if trade costs were high and FDI was prohibited. Trade liberalization (with FDI remaining prohibited) did not influence much in the outcome, as national companies still existed over most of the parameter space. As FDI was allowed, horizontal MNEs entered over much of the parameter space, while the existence of vertical MNEs when factor endowments were different could be explained by trade and FDI liberalization, existence of no MNEs was the result of equalized prices.
Markusen argued that vertical MNEs were less common than horizontal MNEs, it only existed for some host economies in some industries. Combining those new trade theories and internalization theory, Markusen (2002) referred to the models of Horstmann and Markusen (1987b), Ethier and Markusen (1996) and Markusen (2001) that explained how firms entered foreign markets, adding licensing as the third choice (in addition to exporting and FDI). These models were built basing on game theory, information theory and the theory of contracts with concepts such as asymmetric information, moral hazard, and incomplete or unenforceable contracts.
Empirical evidence has been divided into whether to support the horizontal FDI model, vertical FDI model (Braconieret al. (2002), Hanson et al.s (2001)) or knowledge-capital model (Markusen (1997), Carret and al. (1998)). While there was strong support for the idea that market size and transport costs determined FDI, the idea that factor endowments were significant determinants (which would substantiate the vertical FDI model) remained disputed.
2.8. Determinants of FDI according to the Diversified FDI and Risk Diversification Models
Hanson et al. (2001) argued that research should focus on the choice between production - and distribution-oriented (wholesale) FDI and analyze the use of foreign affiliates as export platforms and outsourcing by MNEs to their affiliates. However, the idea of different FDI types was not original. Dunning (1980) had already mentioned and distinguished between six types of international production (resource-based, import substituting manufacturing, export platform manufacturing, trade and distribution, ancillary services and miscellaneous) and differences in their determinants.
Based on Hanson et al.s findings, Ekholm et al. (2003) derived a theoretical model explaining export-platform FDI, while Grossman and Helpman (2002a, b) modeled international outsourcing. Grossman and Helpman (2002a, b) analyzed an MNEs decision between vertical integration and outsourcing (vertical specialization). He found that MNEs were more likely to choose outsourcing in following cases: specialized firms had a productivity or cost advantage, there was an efficiency improvement of the search technology, industry and economy were large and the number of specialized firms was high, or there was a great substitutability between the specialized goods.
While the horizontal and vertical MNEs were explained well by using the transaction-cost approach or the knowledge-capital model, diversified MNEs could not be explained basing on this approach, as it occurred owing to firms wanting to spread business risk. The idea of establishing MNEs to spread risk was a development of Rugmans (1975, 1977) risk diversification hypothesis. Rugman argued that the MNEs objective of locating overseas was to enjoy product and factor market diversification and maximize their profits. Producing abroad was considered as a diversification from solely producing domestically, while producing different products abroad was considered as a double diversification - a diversification in product and location.
Evidence supporting the hypothesis that firms wanted to diversify geographically to reduce risk was presented by Hughes et al. (1975), Michel and Shaked (1986), Miller and Pras (1980) and Thompson (1985). Terrorism is another risk factor that could affect FDI. Enders and Sandler (1996) found that terrorism in Spain and Greece led to a persistent and significant negative influence on FDI stocks and flows. The effect of labor disputes on FDI from Korea was analyzed by Tcha (1998). Taking geographical and product diversification into account, Kim et al. (1993) showed that it was possible for MNEs to have high return-low risk profiles when they diversified both geographically and on a product basis. Since FDI could also be seen as a diversification of real assets by MNEs, exchange rates (reflecting market risk for foreign affiliates) should affect FDI flows. Cushman (1988), Caves (1989), Klein and Rosengren (1994), Dewenter (1995) found the relationships between FDI and exchange rate.
In summary, empirical studies showed that risk factors including market-based risk, exchange rate and interest rate, could determine FDI and should thus be incorporated into the theoretical models explaining FDI.
2.9. Policy Variables as Determinants of FDI
In addition to the models discussed above, FDI can be seen as a game with two players, MNE and host government, or as a contest between two or more host countries competing for FDI. Different stages of the investment decision process are affected by government policies and incentives. Governments can influence the firms choice between domestic production, licensing or FDI, the firms location choice (from a region or pool of investors to a specific location), the firms choice between green field investment (i.e. the construction of new factories or offices) or the acquisition of an existing firm (including joint ventures), the firms choice to stay or to pull out after the investment is made and the firms choice to stay or to expand. Hence, the investment decision-making process is very complex and not simply a decision between investing and not investing.
In practice, MNEs and host countries bargain over numerous issues, including taxes, subsidies, financing arrangements, use of expatriates, training, local employment, local input, export conditions and capital repatriation, i.e. factors related to areas of government intervention in FDI.
Bargaining strength is important in the negotiating process and is influenced by information asymmetries, competitive structure of the economy, market size and projected market growth, political stability, the level of infrastructure and the endowment of natural resources, but also the competition from other host governments.
In general, one can distinguish between three main types of investment incentives: fiscal incentives (profit-based, capital investment-based, labor-based, sales-based, value-added-based, import- or export-based and incentives based on particular expenses), financial incentives (government grants, government credits at subsidized rates, government equity participation and government insurance at preferential rates) and other incentives (subsidized dedicated infrastructure, subsidized services, market preferences and preferential treatment on foreign exchange). Hence, considering the many potential combinations of policy and non-policy determinants and their effects, many areas are still open for theoretical and empirical research despite the research that has been undertaken since the late 1980s.
In summary, policy variables such as corporate tax rates, tax concessions, tariffs and other fiscal and financial investment incentives had a significant effect on FDI in a number of studies such as those of Asiedu and Villamil 2000; Wei 2000; Asiedu 2006; Ting & Tang 2010; and should thus be considered as potentially important determinants of FDI. In general, the effect that tax policy had on FDI was small compared with that of other factors (including market size and growth, basic infrastructure, political stability, cost and availability of factors of production).
Fiscal incentives should be seen as adding explanatory power to models that explain FDI using non-policy variables, rather than replacing them. Tax policy cannot compensate for a negative investment climate, though fiscal incentives can promote investment in a favorable investment climate.
3. Conclusions
There is not one single theory of FDI, but a variety of theoretical models attempting to explain FDI and the location decision of MNEs. While the neoclassical model, which explained international capital trade due to differences in returns on capital, was heavily criticized because of its assumption of perfect competition, Dunnings OLI framework proved to be a better approach of explaining FDI as linked to MNEs, which were seen as firms with market power. His model combined ownership, location and internalization advantages as determinants of FDI after they were previously discussed in separate theories. An alternative framework for analysing FDI and MNE activity, combining ownership and location advantages with technology and country characteristics and explaining both horizontal and vertical FDI, was offered by the new trade theory. Horizontal FDI, for instance, was explained using the proximity–concentration hypothesis, while vertical FDI was explained using the factor-proportions hypothesis. This area of research was complemented by Markusens knowledge-capital model that allowed for both FDI forms as special cases. These models could be modified to explain other FDI forms such as export-platform FDI, wholesale FDI and outsourcing. An additional type of MNEs, diversified MNEs, was explained by the risk diversification hypothesis with firms seen as risk averse and trying to spread business risk. FDI could also be viewed as a game with two players, MNE and host government, and a contest between two or more host countries competing for FDI with a variety of policy, fiscal, financial and other investment incentives influencing the FDI location. Hence, the different approaches do not necessarily replace each other, but explain different aspects of the same phenomenon.
Since there are a variety of theoretical models explaining FDI, there are many factors that were experimented with in empirical studies to determine which factors influence FDI. R&D and advertising expenditure, skill and technology intensity, the existence of multiplant enterprises and firm size were important ownership advantages in a number of studies while, in another area of research, aggregate variables (such as market size, growth and trade barriers) had an effect on FDI. A combination of ownership advantages, location advantages (including market size and characteristics, factor costs, transport costs and protection) and other factors (such as political regime and infrastructure quality) had explanatory power when analyzed under the OLI framework. The proximity–concentration hypothesis was also robust, as FDI could be explained by market size, transport costs and protection and agglomeration economics such as R&D and advertising intensity or corporate scale economies in general. Studies that looked at the horizontal FDI, vertical FDI and knowledge-capital model and their determinants found market size and characteristics (in particular a countrys skilled labor endowment) and transport costs and protection to be important factors explaining FDI. However, the horizontal FDI model explained overall FDI better than the vertical FDI model, while the knowledge-capital model had the same explanatory power as the horizontal FDI model. Risk factors (such as market risk, the exchange rate and the interest rate) affected the location of MNEs, as did policy variables (such as corporate tax rates and tax concessions and tariffs and other fiscal and financial investment incentives). Hence, the empirical evidence strengthens the idea that the different approaches do not necessarily replace each other, as every theoretical model found some support through regression analysis. Therefore, FDI should not be explained by single theories but more broadly by a combination of ownership advantages or agglomeration economics, market size and characteristics, cost factors, transport costs, protection and risk factors, and policy variables. Many empirical studies have already taken that approach, even when focusing on specific theories or aspects of FDI.
REFERENCES:
1. Asiedu, E., & A, Villamil. (2000), “Discount Factors and Thresholds: Foreign Investment when Enforcement is Imperfect”, Macroeconomic Dynamics, 4(1): 1-21.
2. Dunning, J.H. (1973) The determinants of international production, Oxford Economic Papers, Vol. 25.
3. Dunning, J.H. (1981) International Production and Multinational Enterprise, Allen & Unwin, London.
4. Hanson, G.H., Mataloni, R.J. and Slaughter, M.J. (2001) “Expansion Strategies of U.S. Multinational Firms. NBER Working Paper 8433. Cambridge, MA: National Bureau of Economic Research”.
5. Hymer, S.H. (1960), “The International Operations of National Firms: A Study of Direct Investment”, Ph.D thesis MIT press: Cambridge, MA.
6. Markusen, J.R (2001), “Contracts, Intellectual Property Rights, and Multinational Investment in Developing Countries”. Journal of Economics 53: 189-204.
CÁC MÔ HÌNH LÝ THUYẾT VỀ CÁC YẾU TỐ QUYẾT ĐỊNH ĐẦU TƯ TRỰC TIẾP NƯỚC NGOÀI
ThS. NCS Nguyễn Thu Hằng
Khoa Kinh tế - Vận tải, Trường Đại học Công nghệ Giao thông Vận tải
Tóm tắt:
Đầu tư trực tiếp nước ngoài (FDI) đóng vai trò quan trọng trong kinh tế quốc tế. Các nhà nghiên cứu đã phát triển nhiều mô hình lý thuyết để giải thích các yếu tố quyết định đến FDI. Bài báo này trình bày tóm tắt các mô hình lý thuyết về FDI gồm: (1) nghiên cứu ban đầu về các yếu tố quyết định FDI, (2) lý thuyết thương mại tân cổ điển, (3) lợi thế quyền sở hữu, (4) các biến kết hợp, (5) quyền sở hữu, vị trí địa lý và lợi thế quốc tế hóa, (6) mô hình FDI ngang và FDI dọc, (7) mô hình vốn tri thức, (8) mô hình đa dạng hóa rủi ro và (9) mô hình biến chính sách. Mỗi lý thuyết đưa ra một số biến ảnh hưởng tới FDI. Bên cạnh đó, các nghiên cứu thực nghiệm đã chứng minh vai trò quan trọng của các biến này trong thực tế. Do có rất nhiều lý thuyết về FDI, việc phân tích động lực của FDI không nên dựa trên một mô hình lý thuyết duy nhất. Động lực của FDI nên được phân tích dựa trên việc kết hợp rộng rãi nhiều mô hình lý thuyết như kinh tế tập trung, quy mô thị trường, chi phí, các biến rủi ro, yếu tố bảo hộ và biến chính sách.
Từ khóa: Mô hình lý thuyết, đầu tư trực tiếp nước ngoài, các yếu tố quyết định FDI.
