Mainstream internationalisation theories
Mainstream Internationalisation Theories
Instead of looking at the global strategy of the MNE from the viewpoint of
management science, marketing, and decision theory, it is necessary to consider
more explicitly the economics of the foreign investment decision.
International business activity is not a recent phenomenon. However, the great majority of foreign investment until the late 1940s was in the form of portfolio capital, which international capital theory explained as the flow of capital among countries in the pursuit of higher returns. After World War II, the volume of foreign direct investment (FDI) grew tremendously and was increasingly directed away from primary goods and towards knowledge-based products that could be produced in developed countries. Neoclassical economic theory, with its assumption of perfect markets and internationally immobile factors of production, could not easily accommodate this post-war boom in FDI. So, beginning with the publication of the product cycle theory by Raymond Vernon (1966) and Stephen Hymer’s dissertation (written in 1960 and published 1976), an outpouring of literature has focused on extending the theoretical foundations of the concept of foreign direct investment.
The objective of this chapter is to provide a review of the mainstream literature on internationalisation. Given my research problem, the focus is on theories that consider transnational expansion at the firm level. Among others, the investment development path (IDP) concept and Ozawa’s “tandom growth” treatment of the flying geese metaphor are popular frameworks for considering FDI. They are not included, however, since their research setting is that of the economy as a whole.
Despite considerable disciplinary diversity, a mainstream internationalisation construct with three major approaches can be identified: Theories of the MNE, Internationalisation Process Models, and Network-based Approaches to Internationalisation. The first of these, Theories of the MNE, is outlined in Section 2.1. Since these MNE theories have been criticised on the grounds that they may explain the existence of the international firm but not how the firm got there, Section 2.2 reviews Internationalisation Process Models, which more explicitly focus on the dynamic process of internationalisation. Section 2.3 examines leading network-based approaches to internationalisation. The chapter concludes with a summary of the points that are most applicable to my thesis and an assessment of the limitations of the mainstream internationalisation literature.
Theories of the MNE
This section presents the economics-based literature on MNEs, beginning with Hymer’s seminal work. Following a review in Sections 2.1.2 and 2.1.3 of Internalisation Theory and Dunning’s OLI framework, Section 2.1.4 focuses on theorisations specific to developing-country MNEs.
Monopolistic Advantage Theory
Hymer’s (1960) work represented a major departure from the standard orthodox theory of international trade and capital movements. The standard neoclassical trade theory of Heckscher and Ohlin, for example, carried restrictive assumptions about the immobility of factors of production and identical production functions across national boundaries. And in the neoclassical financial theory of portfolio flows, multinational enterprises had been viewed simply as arbitrageurs of capital in response to changes in interest rate differentials.
Hymer argued that explanations for why firms engage in international production should be based on an analysis of the MNE from an industrial organisation perspective. According to Hymer (1976), Kindleberger (1969), and Caves (1971), MNEs emerged because of “market imperfections.” These imperfections were “structural” in nature and resulted from the control of ownership advantages, such as special access to inputs, scale economies, gathered managerial expertise, proprietary technology, and product differentiation (Kalfadellis and Gray: 2003: 3). The result of these barriers to entry was a divergence from perfect competition in the final product market. MNEs would seek to internalise these ownership advantages by establishing monopolistic-type advantages through the vertical integration of the potential licensee (Hymer 1976). Internalising operations could lead to gains such as cost reductions, product quality improvements, and innovation. For Hymer, though, “…the firm internalises or supersedes the market…” (1976: 48) primarily because, by internalising international economic activity, the MNE has an opportunity to further advance its monopolistic advantage. In short, it is the pursuit by firms of market power and monopolistic advantages in a foreign market that largely drives the international expansion of domestic firms.
A criticism raised in the 1970s about Monopolistic Advantage theory was that it did not differentiate between imperfections brought about by market structure (i.e., the number and size of enterprises on both the demand and supply sides) and those associated with transaction costs. By not doing so, Buckley and Casson (1976) and others argued Hymer had failed to incorporate the insights of Coase’s (1937) concept of market failure.
Coase’s theory of the firm contended that, contrary to the classical understanding in which price mechanisms optimally coordinate markets, market failure can occur as costs associated with the price mechanism develop (such as finding buyers and sellers, and the costs involved with negotiating, coordinating, monitoring, and enforcing contracts, and costs associated with government regulations and taxes). The operation of markets is therefore not costless, and the firm is an organising unit that supplants the price mechanism. Domestic firms would prefer to use internal prices in the face of excessive costs in the outside market. Firms therefore seek to avoid these costs by internalising them wherever the market is non-existent or when it is cheaper for the firm to undertake the activity internally rather than via the market mechanism.
To Coase, markets and firms were alternative methods for organising economic exchanges. The choice between the two depended on whether a firm evaluated the transaction costs of an exchange to be lower if carried out within the firm than through the market. Where the costs of such transactions are lower when carried out within the firm than through the market, the activity will be internalised under the firm’s ownership and control.
The concept of transaction costs was more fully developed by Williamson (1975) and Chandler (1977). Transaction cost theory extended Coase’s work by substituting a conception of contractual man for neoclassical theory’s economic man. Its starts with the assumption that markets are the “natural” mechanism of economic organisation (Williamson 1975: 21), and that market failures lead to the replacement of certain market relations by internalising these relationships within a firm. The deficiencies of the market system are seen to be rooted in “bounded rationality” (i.e., the lack of perfect knowledge which means that agents cannot foresee all possible circumstances to incorporate in the contract) and “opportunism” (i.e., agents make decisions based on self-interest, thus making the contract difficult to enforce).
Drawing upon Coase’s (1937) theory of the firm and Williamson’s (1975) and Chandler’s (1977) transaction cost theory, Buckley and Casson (1976) argued that these same insights can be applied to the global arena to explain the growth of MNEs. Accordingly, Buckley and Casson explained international expansion as occurring whenever a market imperfection exists and a firm can gain strategic benefits by internalising a market across national boundaries and exploiting the advantage this gives it in competition with others. This results in the growth of the firm. Just as a firm may increase its efficiency through internalising transactions, the vertical integration of global operations may lead to economies and efficiencies. These include long-term contracts through more efficient governance structures, the chance to exploit tax differentials and foreign exchange controls, better quality control, and R&D benefits.
Brown (1976) also combined insights from Coase’s theory with transaction cost theory and applied it to international expansion. He put particular emphasis on the point that there are higher market transaction costs and more expenses associated with internal organisation abroad than in the domestic environment. Teece (1983) added the insight that internalisation can also be advantageous when vertically-integrated firms need to secure their supply of intermediate goods.
So, whereas transaction cost theory aims to explain the existence of the firm, the aim of internalisation theory is to explain its multi-plant operation over space (Casson 1982). And whereas Hymer argued that it is the pursuit of market power that drives MNE growth, Buckley and Casson (1976) argued that once transaction costs are internalised they do not necessarily lead to an increase in “rent” by the MNE. However, they can result in savings for the MNE, and it is this potential cost minimisation that provides the impetus for MNEs to expand their operations via the internalisation of transaction costs.
Internalisation theory has been a dominant construct in the last quarter century of international business literature in relation to the growth of the MNE and FDI. However, it does have weaknesses. For instance, internalisation’s inherent intangibility makes it difficult to empirically test (Kalfadellis and Gray 2003: 10). Buckley, describing internalisation as “a concept in search of a theory” (Buckley 1983: 42), argued that a theory needs to do more than assert firms will internalise when the cost of using markets or contractual agreements is higher than that of organising it within the firm; it needs to explain why there were differences in costs between market and intra-firm organisation (Hennart 1986: 791).
It has also been seen as overly-preoccupied with the costs of organising transactions in markets, leading it to under-appreciate other relevant costs, especially those associated with managing firms across borders (Demsetz 1988). An argument has been made that it does not sufficiently distinguish between a firm’s willingness and its capability to become more international (Dunning 1993). These types of limitations led Calvet (1981), among others, to question whether the assertion that firms expand overseas because they can internalise transactions within their hierarchies (just as they do within a domestic context) is a full enough explanation. Calvet argued instead for a theory of transnational expansion that explicitly included both the multinational-the “foreign”-character of the activity as well as the internalisation of transactions within a single firm.
Dunning’s OLI Paradigm
A third landmark development in MNE theory was Dunning’s OLI paradigm, sometimes referred to as the eclectic paradigm. Countering Rugman’s (1982; 1985) claim that internalisation is a general all-encompassing theory which can explain FDI, Dunning (1980; 1988; 1993; 1995; 2000) acknowledged the importance of internalisation theory but argued that “set[ting] out to explain the growth of international production as a market replacing activity” (Dunning 1988: 24) explains only part of the FDI phenomenon. Dunning argued that a full explanation required the integration of the insights from three strands of economic theory – industrial organisation, international trade theory, and internalisation theory – into a general theoretical framework. Each dimension on its own was insufficient to explain the multinational firm’s engagement in foreign production.
According to Dunning, a firm must perceive certain advantageous conditions before it engages in cross-border investment. These advantages are rationally considered within the firm’s decision-making process. The first relates to ownership (O) advantages, which, following Hymer, refer to assets or resources capable of generating a future income stream that could compensate for the higher costs of operating abroad. Ownership advantages are endogenous to the firm and refer to intangible assets and/or property rights. These O advantages give the firm a competitive edge vis à vis other firms. The second factor is internalisation (I) advantages, which encourage a firm to internalise operations for production via foreign direct investment rather than through exporting or licensing to a local producer. In other words, the firm must perceive the benefits of internalising of operations to be greater than the need to utilize markets. If a firm perceives it has sufficient O and I advantages, then it will examine a third set of conditions, location (L) advantages. Choosing a foreign location is one of the key decisions made by a firm since the financial and human capital invested must generally be long-term in nature. Drawing upon the insights of location theory, Dunning’s “L” advantages were considered to be external to the firm and determine which host country is selected for expansion. (A fourth condition later added by Dunning  asserted that a firm’s international investment activities must harmonize with its long-term management strategy.)
In the eclectic paradigm, all three of these conditions must exist for FDI to occur. If a firm only perceives it has ownership advantages, then it would be likely to license abroad. If it also perceives internalisation advantages, then it would be likely to exploit its O advantages through exporting. It is only when location advantages are also perceived that the firm may consider FDI (Dunning 1993: 196).
Dunning’s OLI paradigm has been welcomed for its conceptual richness-it integrates many partial approaches to the subject and therefore addresses a larger number of the factors considered in the decision to internationalise-and it has withstood some empirical testing (Dunning 1979, 1983, 1988). However, it has also frequently been criticised, particularly on definitional grounds. For example, Rugman and Dunning had a long-running public debate over whether Dunning’s concepts of ownership and location advantages were already encompassed in the theory of internalisation (Parry 1985). In a similar vein, Buckley (1988) suggested that considering ownership advantages as a separate category results in double counting as the “O” advantage of Dunning’s OLI triumvirate is already accounted for by “I” (internalisation advantages) since the firm seeks to carry out a strategic move by internalising the market and thus exploits this advantage in competition with other firms.
Responding to definitional criticisms, Dunning (1995) argued that, in contrast to how they are conceived in internalisation theory, ownership advantages are endogenous rather than exogenous variables already belonging to the firm. Accordingly, he stressed a definitional division between ownership advantages, which are already possessed by firms, and internalisation advantages, which result from the firm’s exploitation of market imperfections.
The electric paradigm has become a leading conceptualisation for FDI, and as such there now many variants within the approach. For example, another eclectic framework that is pertinent to my thesis concentrates on understanding how a firm chooses among various entry modes. In comparison to Dunning’s OLI paradigm, the framework by Hill et al. (1990) emphasised the control of resources, resource commitment, and the dissemination risks of entry. They argued that firms rationally weigh different entry modes with the need to control their foreign operation. The amount of control a firm can exercise varies from minimal in the case of licensing to maximally high in wholly-owned subsidiaries. A firm also weighs the resource commitment that is involved with the different entry modes, and the risk that its firm-specific advantages could be disseminated or expropriated by a partner. As discussed in Chapter 5, the latter danger was frequently highlighted by my interviewees as an influence on their internationalisation decisions.
Though eclectic models such as those by Hill et al. and Dunning have a dominant place in the MNE and FDI literature, they do have significant shortcomings. Some critics find the emphasis on the initial phase of internationalisation makes them unhelpful. Others have argued that inadequate attention was given to the insight that firms make cross-border investments not just to reap benefits from existing ownership advantages but to create new ones, such as acquiring knowledge in new markets or access to resources. Also, the broadness of the eclectic decision-making framework has made it difficult to formulate operationally testable theories of foreign direct investment processes, especially given the heterogeneity of firms. Various proxy measures have been employed as a means for measuring internalisation, but the validity of proxies in general has been contested (Kalfadellis and Gray 2003: 11). Similarly, ranking the large variety of strategic alternatives the firm can choose among is methodologically problematic.
Two other criticisms of the eclectic decision-making paradigm have been particularly acute and are of specific concern given the subject of this dissertation. The first is that they principally focus on relatively large firms from developed countries. Dunning’s OLI paradigm, in common with the other theories of the MNE reviewed in Section 2.2, was developed primarily in response to the experiences of post-war expansion by developed-country multinationals. Transnational firms from developing countries, it has been argued, require a different approach (Lall 1983a; Wells 1983a; Khan 1986a; Yeung 2004). For instance, as they are frequently much smaller than developed country MNEs, their transnational investment choices may be more “chunky” in nature, in the sense that certain costs that are incurred in international activity will loom relatively larger for small firms than big ones.
Second, the eclectic framework has been criticised for its lack of dynamism. While it is not true that Dunning’s OLI model has no dynamic dimension, Buckley (1985: 18), for example, argued that it does adequately consider the deployment of advantages over time. Both of these shortcomings are apparent when eclectic frameworks are applied to the phenomenon of Singaporean SME transnational expansion into China.
Developing-Country MNE Theories
A dramatic growth in outward FDI flowing from developing countries has occurred over the last three decades. Prior to the 1980s, more than 90 per cent of global FDI originated from developed countries. Since the early 1990s, though, the share of outward FDI from developing countries has rapidly grown; it was over 14 per cent in 2006 (WorldBank 2008). Moreover, aggregate figures conceal the relative intensity of developing-country FDI flows from, and into, certain countries and regions The bulk of this outward FDI-some 67 per cent-has originated from South, East, and Southeast Asian countries (WorldBank 2008). Though the availability and quality of FDI data has been problematic-an important point which is discussed in Chapter 6-it is clear that China has received a particularly large percentage share of FDI originating from developing countries.
A number of researchers have argued that MNEs originating in developing countries possess distinctive characteristics in comparison to their counterparts from developed countries (Lall 1983a; Wells 1983a; Khan 1986a; Yeung 1996). One obvious difference is that they are generally much smaller, which may make locational advantages and the internalisation of transactions costs less plausible explanations for internationalisation (Wells 1983a). Though still dwarfed by the number of theoretical and empirical studies investigating developed-country MNEs, research into these “unconventional” MNEs (Giddy and Young 1982) has by now developed into a large body of literature that can be divided into two categories: “first-wave” and “second-wave” literature.
The so-called “first-wave literature” emerged in the late 1970s and was primarily concerned with the cost advantages of developing-country firms in comparison with their competitors from developed countries. Two strands of literature dominate. One is based on Wells’s (1983) application of the product cycle concept (originally associated with Vernon’s seminal article ) to the situations found in developing countries. The second dominant strand of “first wave” literature is associated with Lall (1983).
Wells contended that an understanding of developing country transnational firms could be undertaken by applying Vernon’s concept of the product cycle (1966), which explained changes in production locations as a reaction to different stages in a product’s life cycle. Vernon’s argument was that a new product had to be produced in the home country since it was unstandardised and thus production needed to be monitored close to the product’s source of innovation and markets. As the product matured and became standardized, producers would increasingly become concerned about production costs and seek cheaper production sites elsewhere. Thus, Vernon’s model suggested that locations of production moved from developed countries to less developed ones as products went through their life cycle over time. This would then explain investment flows from developed- to less developed-countries, and flows among less-developed countries.
The uniqueness of Well’s approach lies in his application of the product cycle concept to explain the emergence of developing-country transnational firms. Wells suggested that the markets and characteristics of developing countries influence local firms to innovate in ways that are more suited to the development conditions found in their country. In particular, he pointed to the smaller size of the markets and relative abundance of cheap labour in developing countries as key influences on local firms. Wells suggested that firms developing in this kind of environment could build their initial advantages from “descale manufacturing,” a process of adapting technologies from developed countries to suit less developed markets by reducing scale, replacing machinery with manual labour, and relying on local inputs. The cost advantages to be derived from descale manufacturing would constitute a very important ownership advantage, and, to exploit these costs advantages, developing country firms would concentrate on serving the price-sensitive market instead of the specialty markets dominated by firms with the resources for massive marketing.
This kind of low-cost, low-price competitive strategy would largely confine the transnational expansion of developing country firms to those markets of other developing countries at or below the host country’s economic status. Changes over time in investment flows would occur as this cost advantage was gradually undercut by the catch up of local firms or affiliates of advanced-country multinationals. Wells’s model has been influential, though it does seemingly suggest a rather pessimistic future for developing-country transnational firms (Wells 1983 and Aggarwal 1984).
Taking a different approach, Lall (1983) argued that the smaller size of production in developing countries was “not by itself evidence of a descaling advantage” (1983: 11). He did not share Wells’s pessimism over the sustainability of developing-country firms, asserting instead that such firms could generate their own sustainable proprietary assets to be exploited successfully in transnational operations. Lall saw the development of these proprietary assets as entailing different innovations than those used by multinationals from developed countries; for instance, they would come from widely diffused technologies and from a special knowledge of developing-country markets. They would be sustained, Lall contended, by the localisation of technical change and the irreversibility of such change. So, developing-country firms could develop products more suitable to developing-country markets, and innovations could be localised around techniques more relevant to developing-country market conditions (such as cheap labour).
Thus, according to Lall the ownership advantages of developing-country transnational firms come about not because of their ability to descale manufacturing technologies to smaller markets, but rather are derived from their greater knowledge of operations and conditions in developing-country markets (see also Kimura 2007). Such advantages would not necessarily be eroded over time, as suggested by Wells, since firms could engage in R&D and continued learning.
Challenging these models by Wells and Lall is the so-called “second-wave” literature that emerged in the early 1990s. This new strand was a response to the apparent changes that were seen to characterize more recent developing-country transnationals. For instance, it was observed that they were investing in markets farther away from home, in some cases in highly competitive markets such as the United States and European Union, and in new sectors, some of which did not depend on labour-intensive techniques. Moreover, the ownership-specific advantages of the newer transnational firms had changed. No longer did they seem primarily dependent on small-scale, labour-intensive technology, low-price, and low-cost operations. Now, they appeared to also derive ownership advantages from their ability to accumulate technological capabilities and to improve their production efficiency (Dunning 2000).
This last observation in particular encouraged second-wave theorists to apply the concept of technological accumulation to try to understand the more recent transnational expansion of developing-country firms (e.g., Dunning 2000; Ulgado et al. 1994). The result was a model that proposes that over time technological accumulation can lead to a more sophisticated structure of outward investment. This gradually comes about, it was argued, as firms accumulate technological expertise and experience in foreign markets. Although their technological capabilities are not based on frontier technology, developing-country firms are believed to innovate and accumulate technological skills that will be appropriate to the environment of developing-country markets. Thus, a firm’s initial outward investment, which is originally centred on resource-based and simple manufacturing activities in markets close to home, changes to focus on more sophisticated manufacturing activities, eventually even to research-intensive and differentiated products. Through this path, second-wave theorists suggested, firms can enhance their technological capabilities over time, which will improve their ownership advantages, and, eventually, allow them to catch up with competitors from developed countries.
A variant within the second-wave approach was proposed by van Hoesel (1997). He argued that firms from developing countries begin their technological accumulation process by gradually climbing the value-added ladder, from shop floor production operations upward to other value-added functions such as marketing or R&D activities. They need to do this, according to van Hoesel, because developing countries are latecomers to the industrialisation process and therefore their firms do not have significant proprietary innovations (in some respects, van Hoesel’s approach is similar to the Late Industrialisation framework, reviewed in Section 2.3.3). The ownership advantages of developing-country firms are therefore seen to lay initially in the lower value-added production units, with international expansion largely a function of the incremental accumulation of technology that moves the firm up to more sophisticated operations. This incremental technological accumulation process is also held to determine the organisational form of the firm, with early investment forays typified by lower-risk and less-committed forms, such as sales representatives and joint ventures with local partners, and later investment characterised by more complex forms, such as wholly owned subsidiaries or acquisitions of local firms.
Despite the valuable insights provided by both the first- and second-wave literature, it has generated criticism on methodological, empirical, and theoretical grounds. From a methodological point of view, Ulgado et al. (1994: 125) raised the important point that most of these studies of investment by developing-country firms consist mainly of macro-level considerations at the expense of micro-level studies of organisational, operational, and managerial workings. These aggregate analyses often fail to reveal the detailed dynamism of the internationalisation process and the other aspects of business organisation, such as the cultural, political, and social context. Moreover, the FDI from some countries is heavily concentrated in particular markets or industries, and this may lead to research bias. For example, van Hoesel acknowledged that, as his study was of Korean and Taiwanese MNEs in the electronics industry, his conclusions might not be applicable to other developing country MNEs (1997: 239). In fact, it should be more pointed out more generally that the availability and quality of FDI data from developing countries is limited and therefore conclusions drawn from it may not be reliable. In short, more studies at the firm level are called for to provide insights on the internationalisation behaviour of MNEs from developing countries.
Section 2.1 has reviewed a number of conventional economics-based theories of FDI. They share the perspective that FDI is motivated by a firm’s desire to exploit its proprietary advantages abroad. These advantages are seen as transferable from country to country within a firm, but transferred only with difficulty between firms. While the proprietary advantages from developed-countries are derived from frontier technologies and sophisticated management and marketing, those for investors from developing-countries are embodied in imported technologies that have been localised through imitation and adaptation. These theorisations, however, are often criticized for their rather aggregated analyses and for their emphasis on explaining the structure of MNEs as opposed to the process by which firms internationalise. The following section reviews models that explicitly concentrate on the dynamics of transnational expansion.
Internationalisation Process Models
Internationalisation process theorising began with the early studies carried out in the 1970s by a group of Scandinavian scholars. Unlike the economics-based theories reviewed in Section 2.2 which accept the neoclassical economic model of rational agents exhibiting optimizing behaviour as a core assumption, the so-called Scandinavian School is rooted in the behavioural theory of the firm (Cyert and March 1963; Hosseini 2005: 528-9). The behavioural dimension is the assumption that “learning” takes place in response to limited cognitive capabilities in a complex and uncertain environment. Accordingly, internationalisation process models attribute the timing of market entry, its structural form, and its development over time as functions of the increasing commitment of managers to foreign markets. The process behind this increasing commitment is not (neoclassical) rational executive decision-making but an incremental learning trajectory that is human- and history-dependent.
A variety of internationalisation process models can be found in the literature. These have often been divided into two groups (Andersen 1993). The first group is the so-called “innovation-related learning models (I-models). These are based on Roger’s (1962: 81-86) stages of adaptation process. Like internationalisation process models in general, they consider internationalisation to develop out of a learning process that occurs in fixed and sequential stages. I-models explain this learning occurs as innovations and new ideas are adopted by the firm at each subsequent stage (e.g., Czinkota and Johnston 1981).
The second group is the so-called Uppsala, or U-model, which was based on work by Johanson and Wiedersheim-Paul (1975) and Johanson and Vahlne (1977; 1990). Like the I-models, the process of internationalisation is characterised by incremental change and development, but for U-models, firms sequentially internationalise as they acquire knowledge and experience in foreign markets. Johanson and Wiedersheim-Paul (1975) distinguished five successive stages, each representing a higher degree of international involvement. At first, the firm focuses on the domestic market. This first stage of no export activity is followed by a second stage characterised by occasional export sales. The third stage involves systematic export through independent activities or agents. In the fourth stage, the firm establishes an overseas sales subsidiary, while in the fifth and final stage an overseas production or manufacturing unit is established. The successive increase in equity participation from stage one to stage five reflects the firm’s increasing commitment to international expansion. They argued that the incremental nature of international involvement is reflected in both the firm’s choice of market and the entry mode it selects.
In regard to how a firm chooses which foreign market to enter, Johanson and Vahlne (1977, 1990) hypothesised that the most significant factor explaining increased international involvement by a firm was its accumulation of experiential knowledge of foreign markets. This learning-by-doing reduced the “psychic distance” between the firm and the foreign market. Psychic distance was defined as the factors-such as language, culture, political system, or level of industrial development-preventing or hampering the flow of information between a firm and its markets (Johanson and Wiedersheim-Paul 1975; Johanson and Vahlne 1977, 1990; O’Grady and Lane 1996). Firms are expected to begin their international expansion in nearby markets and expand to countries with greater levels of psychic distance once they have accumulated a certain amount of international experience (Ahokangas 1998).
In regard to entry mode, internationalisation process models conceptualise each stage of international involvement as having its own characteristic mode. Entry modes are ranked from lowest to highest in terms of commitment, risk, control and profit potential, and include indirect or direct exporting, licensing, joint ventures, or direct investments in marketing or production subsidiaries. A firm starts out at indirect exports and, with each successive stage, increases its commitment, risk, control, and profit potential as it progresses to direct investment. A firm’s method of serving a foreign market may also change over time, as the firm gains international experience (Welch and Luostarinen 1998).
Buckley specifically considered small firms in his internationalisation process model (1989). Like all such models, his mode of explanation described FDI as an evolutionary process. Prior to making a foreign investment, small firms would export their products to the target country either directly or indirectly. Learning acquired at each new stage explained how firms are able to proceed first to increasing exports abroad, then through a number of intermediate states, such as using agents and establishing a sales subsidiary, to eventually setting up full production in the foreign market (Ahokangas 1998: Section 2.1).
Both the theoretical foundations and the empirical validity of Internationalisation process models have been criticized. Critics of the model’s theoretical underpinnings included Turnball (1987), who argued that the internationalisation process model is too deterministic. In addition, Anderson (1993) argued that the model lacked explanatory power because it could not explain how the process took place, what its dimensions might be, and how movements from one stage to the next were initiated. While empirical evidence that internationalisation was carried out in stages was found in some studies (Fina and Rugman 1996), it was not found in others (Millington and Bayliss 1990). McKiernan found evidence instead that firms often leapfrogged stages (1992).
Some of these shortcomings were addressed by Johanson and Vahlne (1990), who suggested three conditions why a firm’s transnational expansion may deviate from the stages predicted in internationalisation process models. First, firms with large resources may be able to leapfrog some stages. Second, when market conditions are stable and homogenous, relevant market knowledge need not be gained solely from direct experience. Third, when firms have considerable experience from markets with similar conditions, they may be able to generalise such knowledge to other markets without having to go through all the stages.
Casson (1993) argued a similar point. Although he generally agreed with the Scandinavian School’s view of incremental internationalisation, he stressed the necessity to clarify the assumption of similarity (or difference) between home market and foreign one, and between each foreign market. He argued that the advantage from incremental internationalisation was greatest when foreign markets were believed to be significantly different from the home one, but similar to each other. Under such circumstances, the internationalising firm can achieve economies of scope in knowledge.
Despite these criticisms, various insights from the Scandinavian literature remain widely accepted. One is the idea that the firm’s internationalisation is a learning process and is incremental in nature. Firms do not enter all major markets at once, so as to benefit from the lessons learned in the early stages of expansion and to avoid costly mistakes. And a firm’s incremental commitment is reflected not only in the sequence of the markets entered but in the sequence of entry mode choices. A second insight from the internationalisation process literature that remains useful-particularly from the perspective of this thesis-is the suggestion that a firm’s international market selection is influenced by its familiarity with foreign markets. Firms initially expand into markets with close “psychic distance,” and move farther as they accumulate their experience.
In summary, whereas the theories of the MNE reviewed in Section 2.1 tend to explain the existence of multinational corporations as institutional forms for organising international production, Internationalisation Process Models focus on managerial aspects of internationalisation. (This difference in concentration, however, in some ways belies the considerable “domestication” within the realm of economics that has occurred as attributes of behavioural economics are applied to Internationalisation Process Models [Hosseini 2005: 528-30].) Though the models differ in the emphasis they put on learning and technological accumulation, and the process is variously described as cyclic, stage-based, or evolutionary, the core premise of these models is that internationalisation is incremental by nature, as the firm accumulates knowledge about foreign markets.
Given this focus, it is not surprising that Internationalisation Process Models are frequently chosen to study the internationalisation experience of smaller firms, which are endowed with relatively limited resources, as well as those firms that are new to international expansion (Buckley 1989).
Network-based Approaches to Internationalisation
The importance of non-market arrangements between organizations has been increasingly recognised over the past three decades. As early as 1972 Richardson explained how, under certain circumstances, cooperation between distinct organizational units can be more advantageous than a market exchange (Richardson 1972). Since then, research into inter-firm cooperation-often referred to as “network analysis”-has grown into a complex, often conflicting corpus of literature addressing many topics and drawing upon widely differing definitions. In many respects, the concept of the “network” has become a dominant cultural and academic paradigm, yet a critically important question in the network debate remains utterly unresolved. To what extent should a network be considered as: 1) a new way of looking at existing phenomena (Williamson 1985; Johanson and Mattsson 1993); or 2) a particular form of organising economic activity, in contradistinction to other potential forms (Powell 1991; Grabher 1993)?
The first view of networks was dominant in much of the literature until more recently. For instance, as reviewed in the following two sections, networks are treated as an organisational aspect of (all) firms and (all) relationships between firms in transaction cost theory and the internationalisation process models.
More recent analyses, however, have moved toward the second view. This can be seen in the increasingly common treatment of a firm’s interactions with other partners in a network as a type of ownership advantage (Dunning 1995). According to these newer analyses, there are two main ownership advantages that are generated by network relationships. First, networks may increase revenue by binding competitors together as allies (Porter and Fuller 1986), and by facilitating coordination for complementary resources and capabilities (Contractor and Lorange 1988). Second, networks may enable costs to be reduced through economies of scale and scope brought about by joint marketing, research and development, and production (Contractor and Lorange 1988). Costs may also be reduced through the generation of transaction cost savings that develop through the existence and maintenance of trust among network members. In this type of cost reduction, transaction costs are minimised through the sharing of valuable information, by the spreading out of financial risks in risky projects (Contractor and Lorange 1988), and by gaining certain kinds of knowledge that could not be transferred by other methods, such as licensing (Fina and Rugman 1996: 200-01).
While the above approach views network ties as valuable assets, networking has also started to be considered an important organisational capability. Networking has been defined as the concerted effort to develop relationships or ties with the firm’s environment to gain access to or control of scarce resources that have long-term profitability implications for the firm (Zahra et al.1999: 42). In a broader sense, networking capabilities are organisational skills through which firms can gain competitive advantages because they can acquire access to the desired resources and capabilities that are complementary to their own (Pananond 2007: 434; Ebers 1997 Introduction). The relationships or ties generated do not have to be formalised through governance structures, and they may include a variety of players, from suppliers to the state.
The suggestion that a firm can develop ownership advantages through its relationships with others was similarly stressed by Dunning (1995). In an attempt to reappraise explanations of MNE activities in the age of alliance capitalism, Dunning argued that the traditional assumption that the capabilities of individual firms ended at their formal boundaries was no longer acceptable. Instead, according to Dunning, the ownership advantages of any particular firm were closely related to the exchange of skills, learning experiences, exchange of knowledge, and sharing of finance between firms in a network. Others have gone further, arguing that networking bestows “social capital,” which could become an important basis for ownership advantage (e.g., Zahra et al. 1999). Social capital boosters assert that “social capital” is as necessary as: “financial capital” (e.g., cash in hand, reserves in banks, lines of credit), and “human capital” (e.g., intelligence, skills, and experience).
Yet the schism within network literature between the two approaches remains evident, and has led to two divergent theorisations: one which views a network as a form of governance structure, and one which focuses on networks as relationships. An overview of the first approach is reviewed in the following section 2.3.1 as it is conventionally considered to be part of (or least domesticated by) the economics discipline. (The second vein-“networks as relationships”-is considered in Chapter 3 as its roots are in the sociology discipline and it is arguably one of the antecedents of the Chinese Capitalism literature.) Section 2.3.1 then considers the how a network perspective has been incorporated into internationalisation process models. The influential Late Industrialisation approach is reviewed in Section 2.3.3.
Networks as Governance Structures
This vein of literature defines networks as a particular way of organising exchange relationships among firms. Viewing transaction cost minimisation as the main determinate of governance structure, this perspective has its theoretical roots in transaction cost analysis as elaborated by Williamson (1985).
Neoclassical economics assumes a perfectly competitive environment in which the price system perfectly orders the market, where there is perfect information and individual honesty, and where transactions, therefore, are costless. Transaction cost analysis breaks out of these narrow confines by recognizing that this ideal situation cannot exist and transaction costs in the market are inevitable (Williamson 1985; Chandler 1977). Market failures lead to the replacement of certain market relations by relationships within a firm. The deficiencies of the market system are seen to be rooted in two human behavioural characteristics-bounded rationality, and opportunism-and two situational factors-complexity, and the number of transaction partners.
The first behavioural factor, “bounded rationality,” refers to the limitations individuals face in handling information as well as to the difficulties inherent in communicating information. The extent to which bounded rationality factors into any particular case depends on the first of the situational factors Williamson identified: complexity. When a situation is not very complex, it is only a matter of time before the proper allocation of factors is found and so bounded rationality will play only a minor role. However, in complex situations, bounded rationality is likely to an important constraint. The second behavioural factor that leads to market failure is “opportunism,” which Williamson defined as “self-interest seeking with guile” (Williamson 1985: 47). An example of opportunism is the strategic manipulation of information, which causes contacts between economic actors to incur high monitoring costs. The degree to which opportunism is likely to be present (and thus to lead to market failure) depends on the second of the situational factors Williamson identified: the number of potential transaction partners available. Williamson argued that opportunism is more likely in situations involving a small number of potential transaction partners, and less likely when the number of potential partners is large enough that opportunistic behaviour is likely to be revealed and punished.
One way to reduce the impact of market imperfections is through internal organisation. Since under conditions of uncertainty organisations cannot develop a comprehensive plan of decisions in advance, they must deal with uncertainty and complexity sequentially. The flexible nature of internal contracts allows for adaptive steps when new situations emerge. Another advantage of internal organisation is the development of common language and communication norms that help minimize the problems inherent in interpersonal relations. Furthermore, due to converging expectations of economic actors within the organisation, uncertainty about the future actions of partners is reduced.
Three features inherent to internal organisation minimize the tendency toward opportunistic behaviour within firms. The first is the reduced chance of subgroups within the organisation to make gains at the expense of the whole, since every loss of the enterprise (at least in the long run) also affects each subgroup. Second, it is easier to monitor the performance of members of the organisation than that of outside actors. This is due to the full use of formal information and the potential use of informal information within the firm. Third, dispute-settling mechanisms within firms usually work more smoothly than those in the market.
Though there are benefits to internalisation-better flow of information and innovation, easier decision-making on resource allocation, cheaper access to resources-Williamson (1985) also pointed out that the advantages of hierarchical organisation decrease as the scale of the firm increases due to the costs of “delegation.” For example, as firm size grows, the limitations of bounded rationality will increasingly be faced by the firm as information loss due to internal friction occurs. Also, since leaders in larger organisations are likely to be (or be thought of as) beyond the control of lower-level members, there is an increased risk of opportunistic behaviour in the former group. Another counteracting factor is that willingness of low-level participants to cooperate decreases with the size of the hierarchy, due to a lack of consciousness of belonging to the same group. This can cause free-rider behaviour, again limiting the effectiveness of internal organisation. In short, the more complex the structure of an enterprise grows, the higher the internal transaction costs are likely to become.
To avoid the possibility of growing so large that the firm engenders the very market failures it seeks to avoid through internalisation, there are clearly circumstances when transactions are best externalized through the market. These include situations which have a low complexity of environment, a large number of potential partners, and where similar levels of information are available for all partners.
Networks are employed by firms in their attempt to balance the benefits and drawbacks of internalisations/externalisations, and can be used in both horizontal and vertical coordination. Firms choose among different institutional arrangements because of variables such as relation-specific investments and relational contractual mechanisms such as trust. This recognition-that institutional arrangements are not guided by an invisible hand into some “normal” and beneficent path-is a core aspect of the New Institutional Economics (e.g., North 1991).
As is true throughout network analysis, there are a variety of viewpoints within the networks as governance structure approach. Some scholars treat networks as an intermediate or hybrid form governance structure, something that lies between the market and the firm (Williamson 1985). Others have argued that networks are a unique form of organising relationships that are neither market nor firm (Powell 1990). Another group, mainly coming from the business studies framework, has paid special attention to more formalised inter-firm alliances, such as strategic alliances (Contrator and Lorange 1988). They all share, however, the economics discipline’s assumption that the main determinant of their structure is cost minimisation.
Though the “networks as governance structures” approach has become dominant in network analysis, it does have many critics. Nohria famously dismissed it by arguing that according to “a network perspective, all organizations can be characterized as networks and indeed are properly understood only in these terms. So to say an organization has a network form is a tautology” (Nohria 1992: 12). Another criticism is that analysing the organisation of a firm in this way leads to temporal reductionism, since it treats alliances as occurring in context divorced from history (Gulati 1998). By taking each transaction as the unit of analysis, it ignores the possibility that a network governance structure could be a result of prior relationships and interactions between partners. On a related point, this overarching way of viewing organisations is also divorced from the social and political-economic context, which leaves little room for explanations of how and why organisations from different economies and cultures may differ in their network structures.
Networks and Internationalisation Process Models
The concept of the network has also been explicitly incorporated into internationalisation process models. This has been a fairly easy fit since, with their roots in behavioural theory, firms are conceived of as embedded actors in business networks in these models (Johanson and Vahlne 1990; Johanson and Mattsson 1993; McAuley 1999).
Drawing primarily on the Uppsala model of gradual learning and the accumulation of market knowledge, “network internationalisation process models” consider this learning and knowledge to develop out of the firm’s internations with networks. First, a distinction is made between a firm’s “extension” of its networks-investing in networks that are new to the firm-and its “penetration” into networks-deepening its resources and relationships in networks in which the firm already is embedded. And the concept of “integration” is added to refer to the co-ordination of different national networks. Then a firm’s position in the network is considered from either a “firm-to-firm” or a macro level.
At the micro level, firms are viewed as interdependent in terms of both complementary as well as competitive relationships, and both are crucial elements of the internationalization process. At the macro level, these models argue that a firm’s direct (partnerships with others in the network) and indirect (relationships with firms not in the network) relations need to be taken into account. The dynamics of internationalisation occur because an internationalising firm’s networks are at first primarily domestic, but it gradually extends its business relationships to networks in foreign countries. It does so through the establishment of new relationships in host country networks (international extension), through the deepening of relationships in those networks (penetration), and by developing linkages between networks in different foreign markets (international integration) (Ruzzier et al. 2007). The internationalising firm exploits its networks relationships to minimise the need for it to develop market knowledge on its own and to minimise its need to make adjustments to the foreign market (Johanson and Mattsson 1993).
This attempt to explain the process rather than the structure or existence of international firms is an important strength of these internationalization process models. Another key conclusion from the perspective of this thesis is its argument that firms can be influenced to internationalise by the internationalisation of other firms (Ruzzier et al. 2007). This pressure is brought about because of the organisation of the modern industrial system, in which a large number of industries or types of markets are made up of firms that are deeply interdependent because the goods and services they distribute, production, and/or use are quite specialised (Buckley and Ghauri 1993; Andersen 1993). Certain industries and markets would at any one time face greater pressure than others, but given the increasing globalisation of industries and markets, the trend would be towards increasing levels of internationalisation. This important point is discussed in Chapter 5 in conjunction with my research results.
Networks and Late Industrialisation
A third leading mainstream approach that incorporates networks and is useful from the perspective of this thesis for explaining FDI draws upon the concept of “late industrialisation” (Amsden 1989; Wade 1990). It begins with the observation that the first industrial revolution in the United Kingdom at the end of the 18th century, and the second industrial revolution 100 years later in the United States and Germany both were characterised by the invention of new products and innovations to productiive processes. In contrast, the industrialisation of countries in the 20th century was marked by neither invention nor innovation. Instead, these late industrialisers-such as South Korea, Japan, and Taiwan-emphasised “learning” and used borrowed technology to increase their economies and transform their productive structures.
The Late Industrialisation approach, as its name suggests, argues that countries which industrialised in the 20th century shared a common pattern and similar institutions which led to their successful economic development, the most important of which were those associated with highly interventionist industrial policies. Amsden defined national economic development as the “process of moving from a set of assets based on primary products, exploited by unskilled labor, to a set of assets based on knowledge, exploited by skilled labor” (2001: 2). Learning in these countries took place, initially at least, via low wages and significant state assistance. In reference to the second factor, Amsden wrote: “….not only has Korea not gotten relative prices right, it has deliberately gotten them ‘wrong.'” (Amsden 1989: 139). The state’s industrial policies included a wide variety of subsidies and other forms of government support (Amsden 1989: 139). Amsden argued that not only had the Korean and Taiwanese states actively manipulated trade and exchange rates, the allocation of finance, as well as the price structure of the domestic economy, but they had also actively developed a large public enterprise sector and influenced the structure of private investment (1989; 2001). Most importantly, these states combined their commitment to govern the economy with a capacity to enforce the principle of “reciprocity,” which required those industries that received state assistance to meet strict performance standards (see also Dore 1986). So, in contrast to conventional economic orthodoxy that prescribes letting prices guide markets, Amsden argued that interventionist policies by strong states that produced “wrong” prices could be a positive condition for industrialisation.
The result in these late-industrialisers was an environment in which gradual increases in the quality of existing products and the efficiency with which they were produced could take place, where borrowed technology could be adapted to local conditions and applied to the “shop floor.” Amsden argued that one of the most significant aspects of this process was that it did not occur in an atomized manner: it was facilitated and promoted by the vital force of the many diversified business groups and subcontracting networks in these counties that had substantial forward and backward linkages (Amsden 1989: 6-7). The networking of these highly linked business groups and subcontractor circles was seen as an important source of ownership advantage for these firms and helped promote late industrialisation.
There are several reasons advanced for why Asian business groupings and subcontracting networks translate into ownership advantages. The dominant organizing principle is that they may substitute for weak market institutions in developing economies (e.g., Leff 1978). They do this, for example, by augmenting internal capital markets, intermediating in labour and product markets, and helping to minimize opportunistic behaviour.
While the late industrialisation framework shares with other approaches the recognition that many factors drive economic development-e.g., institutional, economic, geographical, and cultural factors-its main contribution to development economics is its inclusion of the state as one of the most significant of these factors. It was through the interventionist policies of the developmental state that the late industrialisers were able to move into the more mature markets of the innovators and to successfully challenge the productivity of long-established innovator based on their lower wages and intense efforts to raise productivity.
The Late Industrialisation approach has many advantages over previously reviewed theorizations of internationalization. First, it is intertemporal in nature, recognizing the structural break in the growth process and political economies of many developing countries in the last century (Katz 1994). Second, it incorporates political-economic factors within a cross-comparative approach. Third, it considers (and adds to) theorizations on the “Developmental State,” which has become an important topic in development economics. All of these reasons suggest it might be a useful framework for this thesis. In particular, the point it raises about the necessity for interventionist policies to be combined with governmental capacity to discipline industry will be discussed in Chapter 7’s assessment of PAP-state developmental policies.
Critics of the Late Industrialisation approach have taken issue with its assumption that “states are autonomous and the most appropriate units of analysis” (Hamilton 1990) and with its dichotomizing of strong versus weak states (Evans 1995). It has also been argued that the global political and economic situation faced by states now trying to develop has dramatically changed and therefore the theory is no longer applicable. These changes include the propensities to save and invest in new capital, and a shift in the norms of entrepreneurial behaviour (Katz 1994). On a related point, Japan’s lingering “lost decade” and the Asian financial crisis of 1997/8 also implicitly challenge this framework. Therefore, “it is by no means obvious that the same model that could help us to understand the development process and the technological and innovative performance of countries of ‘late industrialization’ in the 1960s and 1970s could be used equally well to examine these same aspects in the 1980s” (Katz 1994, introduction).
Another shortcoming of the Late Industrialisation approach is that it has been heavily based on the experiences of firms from Hong Kong, Taiwan, and South Korea, and considerably less consideration has been given to Southeast Asian firms. Indonesian firms received some attention (Amsden 2001) and Amsden specifically mentions Southeast Asia in a 2006 article (Amsden 2006), but the region has been relatively ignored. Although enterprises from Southeast Asia may be expected to share some similarities with their East Asian counterparts, it has been argued that they fundamentally differ from the latter, especially in their heavier reliance on opportunistic and risk-taking behaviour (Lim 1996a). Moreover, whereas outward FDI flows from Hong Kong, Taiwan, and South Korea have been somewhat limited to a few industries, such as consumer electronics and automobiles, investment outflows from Southeast Asia have tended to be much more diversified, which suggests that a different process underlies these countries’ articulation into the global economy. Other important differences between East and Southeast Asia in terms of their historical backgrounds, initial conditions, resource endowments, and political economic environment have also been highlighted (see, for instance, Jomo et al. 1997: 8-26).
For all these reasons, it may be imprudent to assume that literature drawn mainly from studies of MNEs from the late industrialised East Asian countries can be considered a u