The Role of Strategic Focus of the MNC on Entry Mode
Choice
Applying the Bartlett and Ghoshal typology to CEE Entry Mode Choice
Master Thesis International Business and Management
Weeda, C.B. s1404296
s1404296@student.rug.nl
Under supervision of:
Dr. D. Dikova
Word Count: 11.974
Rijks Universiteit Groningen
Faculty of Economics
WSN Building, Landleven 5
Groningen, 9747AD
Abstract
Table of Contents
1: Introduction……….3
2: Literature Review………5
2.1 Strategic Focus of the MNE………..5
2.2 Firm-specific Assets………..9 2.3 Institutions………..……….14 2.4 Cultural Distance……….17 3: Methodology………..22 3.1 Data Selection..………....22 3.2 Variables………..24 3.3 Data Analysis………..……….26 4: Results………27
5: Discussion and Limitations…..………..31
1. Introduction
A multinational enterprise (MNE) investing abroad faces many important decisions, on many dimensions. One of these is the decision whether the firm will take over an existing business (acquisition), or set up a business from scratch (greenfield). This establishment mode choice (Cho and Padmanabhan, 1995) has become increasingly more important for the competitive advantage of an MNE (Barkema, Baum and Mannix, 2002).
Similarly, and equally important, a firm faces an entry mode choice: whether the ownership of this subsidiary should be shared (in a joint venture, JV) or kept in ownership wholly (a whole ownership, WO). This choice has a large influence on the risks, profits and future of the foreign subsidiary.
For instance, a joint venture may allow fast transfer of tacit knowledge between partners, but bears large risks of a mismatch of corporate cultures and goals, as well as the risk of opportunistic behavior. A wholly-owned subsidiary bears no such culture-clash risks but faces longer establishment times and may not gain swift access to local business networks.
The amount of literature on factors influencing entry mode choice is abundant (Zhao, Luo and Suh, 2004), and a lot of research has been done on determining the various influences on the degree of ownership, differentiating between wholly owned subsidiaries and joint ventures (Gatignon and Anderson, 1988; Kim and Hwang, 1992; Delios and Beamish, 1999; Luo, 2001a).
The strategic focus of the firm as developed by Bartlett and Ghoshal (1989) has only been linked to establishment mode (Harzing, 2002). In the context of entry mode decision literature, the effects of global concentration, global synergies and global strategic motivations have been examined (Kim and Hwang, 1992). However, the strategic focus of an MNE determines the manner in which internationalization activities are pursued, as well as the entry mode that is decided to be used pursuing these activities. It seems likely that firms which follow a strictly global or multidomestic strategy are affected differently by asset specificity, large cultural distance or underdeveloped institutions. These factors may have a smaller or more pronounced influence on the entry mode chosen by the MNE depending on its strategic focus. By accounting for the strategic focus of the MNE in entry mode choice, a more complete picture of the diverse predictors of entry mode choice is presented—this allows us to determine the likely entry mode choice that is conditional on the strategic focus of the MNE.
The influence of strategic focus of the firm can exist in various ways, which need to be examined. Therefore, the main research question of this research is:
“What are the determinants of entry mode choices in CEE transition economies?”
This research question will be answered using the following specific research questions: 1) Is entry mode choice influenced by asset specificity, the institutional environment and cultural distance?
2) Is entry mode choice influenced by the strategic focus of the firm?
3) Is the influence of asset specificity, cultural distance and the institutional environment on entry mode choice moderated by the strategic focus of the firm?
The remainder of this research is structured as follows. First we will present a literature review in which not only the relevant theoretical arguments are presented, but earlier conducted research on the relation between each variable and entry mode choice will be presented as well, followed by the expected relation in this research in the form of a hypothesis. Afterwards, the used methodology of this research will be discussed and its results presented, which are again discussed and put in broader perspective in the conclusion, where the limitations of this research and directions of future research will be discussed as well.
2. Literature Review
Entry mode choice depends on many different variables, and no single theory has been able to provide a complete picture on the rationale behind the entry mode choice (Datta, Hermann and Rasheed, 2002). Although there has been some research where different theories were combined into a more general model (Dunning, 1981; Hill, Hwang and Kim, 1990), even these eclectic theories are not able to provide all the key insights on the entry mode choice. Rather, they research several piece of a large puzzle, where many different variables following from many different theory-based arguments are tested for their significance and influence on other variables.
Likewise, this research will combine four different fields of theory, of which some have been researched more extensively than others; transaction cost economics (TCE) (Williamson, 1975), new institutional economics (North, 1990), cultural distance (Kogut and Singh, 1988) and MNE typology (Bartlett and Ghoshal, 1989). A framework of the proposed hypotheses can be found in Figure 1.
2.1 Strategic focus of the MNE
subsidiaries (Hout, Porter and Rudden, 1982). This strategic decision has been captured in the integration - responsiveness (I-R) paradigm (Prahalad and Doz, 1987; Doz and Prahalad, 1991), which has been researched extensively and found to be a solid framework for the analysis of MNE strategic choices (Roth and Morrison, 1990; Taggart, 1998). In this two-dimensional grid the need for integration and the need for responsiveness determine the position of the company and the strategy that would best fit this position (see Figure 2). Different pressures for global strategic coordination, such as the importance of multinational customers, the investment and technology intensity and universal needs, as well as pressures for local responsiveness, such as differences in customer needs and distribution channels, market structure and host government demands, influence the position of the MNE on the grid (Prahalad and Doz, 1987). The framework has been discussed by Bartlett and Ghoshal (1989), who expanded the framework into a typology of the MNE based on the characteristics of the integration – responsiveness paradigm. The I-R framework and the discussion of Bartlett and Ghoshal (1989) have contributed to the categorization of the strategic focus of the MNE into the global-multidomestic dimension, as the pressures for global strategic coordination or local responsiveness vary among firms.
The strategic focus of an MNE comprises the strategy it has determined for entering foreign countries when investing abroad. The goals set for a subsidiary and the way production, marketing and research and development are to be organized have a large influence on the degree of control the parent firm will wish to hold over its subsidiary.
Ghoshal, 1989). From this strategy economies of scale can be obtained which conditional to a large extent on the availability of non-location bound firm-specific advantages (NLB-FSA) (Rugman and Verbeke, 1992).
There are several caveats in pursuing a global strategy which demand a high degree of control over the subsidiaries of the MNE (Kim and Hwang, 1992). When an MNE increases in size, it becomes confronted with other global firms across several countries and industries. This global competition creates a global competitive interdependence, where the strategy and actions of a subsidiary in one country may affect its subsidiaries in other countries (Watson, 1982). Pursuing a global strategy may therefore favor a high degree of control over the subsidiaries that are set up abroad, not only so that the actions of one subsidiary may not create negative side-effects to other subsidiaries larger than the initial gains of the action, but also so that in case of competitive moves other subsidiaries may come to help in different countries and markets.1 In these situations, shared control over the subsidiary could lead to managerial problems, the questioning of strategic choices and the procrastination of their implementation (Kemp, 1999).
Additionally, the global firm has the necessity to optimize the economies of scale it achieves, creating a value chain of which at each stage the added value is optimized (Hout et al., 1982). The smooth running of this value chain is paramount for the MNE, as it constitutes its main firm-specific advantage. Therefore, each subsidiary has to perform its part in the value chain, whether it be the high-volume production of certain components or the local assembly of a final product, as it is centrally coordinated, up to the point where the interdependent network of subsidiaries is appointed specific tasks of what to produce when, and for which price. Such an obsequious role will not likely be accepted by any contract- or joint venture partner, which can prove to be a possible conflict as no deviations from the imposed decisions can be allowed (Hill et al., 1990).
1
A multidomestic firm, on the other end of the I-R dimension, feels pressure to use its strategy to fulfill local needs and demands by creating products that are tailored to the local market. This type of firm finds that there are profound differences between countries in the product markets it serves, and bases its competitive strategy on the various competition of each country and market (Hill et al., 1990). The multidomestic firm combines the firm-specific advantages like inter-subsidiary learning with location-bound advantages such as market specificities and imperfections. It responds to the complex and fast-changing local market with local responsiveness (Morrison and Roth, 1992), pursuing adjustments in products, marketing, management style, research and development and relations with local stakeholders, due to the local differences in customer preferences, habits, culture, regulations and the way business is conducted (Roth and Morrison, 1990).
So that a multidomestic strategy may successfully be pursued, each subsidiary will need to be controlled and coordinated to a very limited extent. Each subsidiary may have its own marketing, production and research and development facilities, which allows the local responsiveness on which its competitive advantage depends to be built. By allowing the locally responsive subsidiaries to take initiative in seeking out new business partners and opportunities (Birkinshaw, 1996), long-term relations with local customers, suppliers, government and industry partners can be forged. This in turn may lead to previously unattainable opportunities, or the avoidance of problematic situations within the local context (Ghoshal and Nohria, 1989).
The limited amount of control that is typical for subsidiaries with high local responsiveness not only saves costs as centralized coordination is no longer necessary. A tight central coordination by the MNE headquarters will not provide the best structure for a multidomestic firm because of the need for flexible subsidiaries—subsidiaries need this flexibility to respond swiftly and promptly to changes in the local environment, be it consumer preferences or new government regulations affecting the legitimacy of the local production unit. (Bartlett and Ghoshal, 1989).
partner when setting up a joint venture. The transfer of tacit knowledge is a time-consuming process, therefore there is a need of a reliable and long-term relation between the local partner and the MNE subsidiary. A joint venture provides such a relation, in which the country-specific advantages (CSAs), to which the local partner holds access, are combined with the non-location bound firm-specific advantages (NLB-FSAs) the MNE holds which is created by the international network of the MNE.
The relation between the strategic focus of the MNE and the degree of control (ownership) it chooses has been researched to some degree. Several conceptual frameworks for entry mode choice analysis considered global and multidomestic strategies of the MNE as predictors (Hill, Hwang and Kim, 1990), but only a few researches have tested empirically the relationship between corporate strategy and preferred entry modes in international markets.
Kim and Hwang (1992) used a survey that yielded 89 questionnaires on foreign direct investments undertaken after 1980. Dividing the influence of an MNE’s strategic focus into three categories (global strategic variables, global synergies and global strategic motivations), they found that a global strategy led to the preference of a JV or WOS over licensing. Although no further distinction was made between JV and WOS, this research showed that the entry mode choice framework should be expanded with a strategic focus variable, as results showed that a global strategic focus of the MNE leads to a higher degree of control of the subsidiary.
Although empirical research on the influence of strategic focus of the MNE is scarce, theoretical frameworks and MNE typologies seem to agree to a very large extent on the sign of the relation between strategic focus and entry mode choice. We therefore propose:
Hypothesis 1: Global firms will have a higher probability to choose a WO entry mode than multidomestic firms
2.2 Firm-specific assets
In his seminal article in 1937 Coase introduced the principle of marginalism, asking the basic question whether internalizing an additional exchange transaction within the firm is manageable (profitable) or not. This idea was expanded upon by Williamson (1975), his transaction cost economics (TCE) specifying that each type of transaction should take place in the most efficient governance structure. Transaction costs are defined, according to Williamson (1975, 1985) as the costs that are involved in finding a contract partner, drafting a contract, and ensuring that the goals that are specified in advance will be met. These costs depend on three characteristics: the degree of asset specificity, the uncertainty and complexity, and the frequency of the transaction. Two assumptions are made on the behavior of the contract partner: he/she behaves opportunistically and has bounded rationality. (Williamson, 1975)
The extent and source of asset specificity, however, largely determine the transaction costs that come with writing a contingent contract. Firm-specific assets can be hard to share or codify, as in the case of capital assets or tacit knowledge: these assets can only be shared through training or created through research and development, and therefore are exposed to little risk of appropriation as their use and specifics can be bound by contracts or protected legally (e.g. by use of patents). On the other hand, explicit knowledge-intensive assets can be more easily imitated and learned by contracting partners, as their nature makes it harder to specify their use and ownership rights through the use of contracts. The MNE bringing in these assets to the joint venture are exposed to risk of losing their firm-specific competitive advantage, as contract partners may have access to critical information and core technologies (Geringer and Hebert, 1989).
Another threat that comes from specifying a contract is the chance for a lock-in to occur between two partners. For instance, if transaction-specific assets are purchased by one of the partners, the other partner gains leverage to negotiate for new contract terms or change the agreement. The dependent partner who has invested in assets that are specialized for only one or a few partners or uses (Williamson, 1981) would in this situation be better off by either replacing specific assets by general purpose assets, to decrease the bargaining power a partner holds, or to internalize the function of the partner. Even though a lock-in could still occur with the employees of the firm, legitimate authority, performance monitoring and a diverse range of incentives not available for contract partners (Anderson and Gatignon, 1986), as well as use of regulation and control can be used within the MNE to ensure opportunistic behavior will not occur (Williamson, 1985).
Unfortunately, Hennart (1991) was not able to find any significant relation between asset specificity and the propensity to set up any kind of subsidiary using logit regression in his paper on Japanese subsidiaries in the US.
Focusing on service firms, Erramilli and Rao (1993) used survey-data to determine the entry mode choice using TCE variables. Their total sample size of 381 entry mode choices confirmed earlier findings of higher asset specificity increasing the propensity to establish a WOS.
Using logit analyses, Padmanabhan and Cho (1996) found significant results for several TCE variables on entry mode choice. Their data on Japanese firms setting up subsidiaries showed a positive relation between the R&D intensity of the firm2 and a WOS as entry mode choice.
Rajan and Pangarkar (2000) conducted a survey on Singapore-based firms setting up foreign subsidiaries and found several significant results using their sample of 103 questionnaires. They found that if firm-specific assets were transferred, there was a tendency for higher control entry modes.
In two more recent articles, the relation between asset specificity and entry mode choice was not confirmed as the asset specificity was not found to be a significant influence on entry mode choice (Chen and Hennart, 2002; Lu, 2002). However, R&D intensity as a proxy for asset specificity has been found significant in various research (Brouthers, 2002).
Following the earlier research of Gatignon and Anderson (1988), Erramilli and Rao (1993) and Padmanabhan and Cho (1996), we therefore propose the following:
Hypothesis 2: Higher asset specificity of the MNE will lead to a higher probability of a WO entry mode
For an MNE with a global strategic focus, its subsidiaries are set up as copies of the parent company in order to strengthen the pursuit of economies of scale, centralized coordination and control as strategic advantage (Harzing, 1999). Firm-specific
2
advantages (FSAs) in the global firm are therefore usually non-location bound (NLB), since these “can be transferred abroad at low marginal costs and used effectively in foreign operations without substantial adaptation” (Rugman and Verbeke, 1992: 763). These NLB-FSAs can be exploited globally and support the strategic advantages of the global firm by leading to economies of scale. Additionally, NLB-FSAs can help the MNE negate market imperfections in the host country market (Rugman, 1981). Since the internalization choice of the MNE depends on the difference between the internalization costs and the transaction costs, we can already say that internalization costs seem to be rather low due to the low adaptation and transfer costs.
Additionally, it can be added that transaction costs for foreign firms are rather high — starting a foreign operation brings a disadvantage to the foreign firm in comparison to the local firms, which is typically referred to as liability of foreignness. The liability of foreignness implies that the foreign firm is at disadvantage to the local firms as it has no established relations with supplies, customers, and other stakeholders, and therefore needs to invest time and money in building a business network. However, despite the liability of foreignness, MNEs do decide in favor of investment because they hold certain advantages in comparison to host-country firms (Dunning, 1981; Hill and Kim, 1988). These advantages are derived form firm-specific assets, or NLB-FSAs. The quasi-rents that can be gained from holding these FSAs are likely to be high, and therefore need protection against potential partners who are self-interest seeking with guile. If the nature of the NLB-FSA is tacit in nature as well, the contract has difficulties to articulate all contingencies, while when internalized the MNE learns to utilize its tacit knowledge more efficiently and thus strengthen the NLB-FSA (Kim and Hwang, 1992). Therefore, a global firm with NLB-FSAs is less likely to establish a joint venture.
NLB-FSAs, LB-FSAs have high costs involved in transferring the assets and adapting them for use in other countries or regions. It can therefore be said, that these LB-FSAs consist largely of local knowledge combined with the NLB-FSAs that form the basis of the existence of the multidomestic firm. This local knowledge can only be gained from entrusting local firms by sharing knowledge, or, as Caves (1996) put it, location-specific capabilities depend on the skill and routines of the (local) employees of the firm. However, setting up a joint venture in order to strengthen the location-bound FSA may prove to be difficult for a multidomestic MNE with high asset specificity, as the contract between the two parties will be unable to articulate all contingencies. This may expose the assets to quite some risk as the contract partner can behave opportunistically as it gains access to them. Therefore, a multidomestic MNE wishing to set up a foreign subsidiary will choose to protect its firm-specific assets. As full protection against opportunistic behavior is impossible, the possibility that the MNE will set up a joint venture decreases.
Hypothesis 3: Asset specificity negatively moderates the likelihood of multidomestic firms to establish shared ownership entry modes
2.3 Institutions
Firms planning to invest in Central and Eastern Europe (CEE) find that this region is in a transition stage from central-plan coordination to a market economy, resulting in an unusually unstable institutional framework (Swaan, 1997). Earlier-mentioned regulative structure, or, as North (1990) calls it, the market economy’s formal rules of the game, needs to be provided by a country’s institutions in order to have efficient markets, since they lower uncertainty and provide means for interactions between firms. Developing these formal rules is an important step for economies in transition (Clague, 1997), as they were non-existent when the central-plan coordination system crumbled. Therefore, several reforms need to be executed after the fall of the central-plan coordination in areas such as financial markets and banks, trade and fiscal regulations, as well as property laws. In CEE countries these institutions have been established, albeit slowly (Meyer, 2001), but there are still high transaction costs in these countries because of high uncertainty, stemming from the lack of experience with and knowledge of market economies (Peng, 2003). Additionally, in a planned economy many organizations relied upon networks formed with other organizations, when no reliable institutions existed. These networks are still used during the institutional build-up-phase, which creates a situation where part of the coordination of institutional changes lie within the network-based coordination mechanisms (Meyer, 2001). As these networks are hard to enter for any foreign company, and the direction of coordination can not be predicted, the existence of such networks in CEE countries creates additional uncertainty and therefore increases transaction costs.
In a system with rapidly changing institutions, there may be large differences between several institutional reforms, or between legislation and law enforcement. Also, future changes in institutions can not be foreseen, making it hard for MNEs to determine the future prospects of their subsidiaries and any decision on the continuation of their foreign investment.
face high costs of setting up a wholly owned subsidiary. The unfamiliarity with local environment and institutions in transition may increase uncertainty and overall investment risk, which can be somewhat mitigated by possible involvement of a local partner who is in possession of specific knowledge how to operate in the local environment at relatively lower transaction costs (Beamish and Banks, 1987).
The influence of institutions on entry mode choice has not been researched as extensively as TCT variables. However, many researchers have included country risk in their empirical model, a variable that is related to institutional advancement since country risk falls as institutional advancement (and therefore regulations such as property protection) increases.
Findings of country risk’s influence on entry mode choice are mixed:
Kim and Hwang (1992) found in their research that firms with wholly owned subsidiaries tend to choose countries with a low country risk.
Erramilli and Rao (1993) on the other hand, found only a significant influence of country risk on entry mode when the firm needed to transfer highly specific assets; in that case a higher control entry mode was preferred.
This conditional relation was generalized further by Agarwal (1994), who found that as the country has higher uncertainty (higher country risk), a higher mode of control is preferred over a lower-control entry mode such as a joint venture.
Brouthers (2002) found the opposite result in his more recent research: as investment risk increased joint ventures were preferred over wholly owned entry modes.
As results are mixed, we follow the more recent and specific results. We therefore propose:
Hypothesis 4: Low institutional advancement of the host country will lead to a lower probability of a WO entry mode
Firms with a global strategic focus are highly dependent on their non-location bound firm-specific advantages (NLB-FSAs). This advantage is gained from their global production and common routines and standards, which can be imposed on each market (Boisot and Child, 1999) and lowers the need to rely on network-building and negotiations with local suppliers, competitors and governments, allowing the firm to keep the transaction at arms length (Peng, 2003). If investment risk increases due to an uncertain institutional environment or low institutional advancement, the amount of local knowledge necessary to allow the firm to conduct its business successfully is limited, as the global firm is able to perform most of its functions independently of the local environment. The local institutional environment may impede the level of income due to barriers of repatriation of income or large currency fluctuations, which show that the global MNE is still impacted by an uncertain local environment. In such situations, local knowledge or a local benefactor, both obtained by sharing ownership, will most likely not be able to influence the institutional environment in such a way that these restrictions will no longer be upheld. Therefore, the global MNE has no specific need for a local partner by setting up a joint venture, and will not alter its entry mode choice.
(Peng, 2000) for the local knowledge necessary for effective customization of the production and marketing process.
However, following the argument by Dikova and Van Witteloostuijn (2007), the working relationship between the multidomestic MNE and its local partner requires development to create firm-specific knowledge and headquarter-subsidiary communication paths (Anderson and Gatignon, 1986). Within the local joint venture a firm-specific asset is thus created, that needs to be safeguarded from free-riding by the local partner. If the institutional environment is not able to guarantee protection of the intellectual rights of the multidomestic MNE, the firm will be willing to trade some of its freedom of local responsiveness for a control mechanism by increasing the ownership over the subsidiary. Considering that the risk of piracy of intellectual assets property outweighs the costs of a loss of access to local business networks, we propose:
Hypothesis 5: Low institutional advancement of the host country negatively moderates the likelihood of the multidomestic firm to set up shared ownership entry modes
2.4 Cultural Distance
Additional to the influence of strategic focus of the MNE, the degree of asset specificity, and the institutional environment of the host country, the home country has influence on the degree of ownership as well, or more specifically: the difference between home and host country (Puxty, 1979). In order to chart the relation between cultural differences and entry mode choice systematically, Kogut and Singh found a way to quantify the cultural distance and used it to research its influence on several variables (Kogut and Singh, 1989).
increases, the costs of gaining information for monitoring and evaluating purposes increase (Jones and Hill, 1988). If cultural distance is lower, the information necessary for foreign operation is easier to obtain and interpret, while such information may be hard to find and put to use if cultural distance is high. A lack of information can result in difficulties with evaluating performance and setting clear goals for the subsidiary to attain (Gatignon and Anderson, 1988). These costs of acquiring information, which can be regarded as costs pertaining to the internalization of the subsidiary, rise with cultural distance, as uncertainty increases along with it.
Entry mode choice can influence the degree of uncertainty that arises with increasing cultural distance. In a joint venture, the local partner can be assigned to manage relations with employees, suppliers, buyers and government (Franko, 1971), which greatly decreases the probability of any large problems or miscomprehension with a possible negative effect (for instance a ban on in- or exporting certain products after insulting the local government due to not following informal rules).The local partner can transfer part of its work force to the joint venture, thereby transferring a part of its culture as well (Hofstede, 2001). Although a joint venture may negate the costs of external conflict that arise from cultural differences, if the cultures of both partners are not comparable, problems within the joint venture may arise as a result (Brown, Rugman and Verbeke, 1989). Such problems, possibly leading to the failure of the joint venture, are hard to circumvent. However, as the foreign subsidiary can not function without acquiring information on the host country’s culture, the costs of such internal conflicts can be assumed to be smaller than the costs of acquiring information using a wholly owned subsidiary. Therefore, a joint venture bears lower cultural risk than a WOS, and problems stemming from cultural differences can be resolved with a joint venture, although at the cost of control and ownership (Kogut and Singh, 1988).
The amount of literature on cultural distance, not only in relation to entry and establishment mode, is quite large. Recent undertakings to map the influence of various variables based on TCE on entry mode choice included cultural distance as well as asset specificity (Zhao, Luo and Suh, 2004).
Empirical research on the relation between cultural distance and entry mode choice has been rather large, although a singular distinction between joint ventures and wholly-owned subsidiaries is often complemented by an additional dimension (such as establishment mode).
Erramilli (1991), and later in cooperation with Rao (Erramilli and Rao, 1993) condoned a mail survey to US service firms, receiving a total of 175 useable responses. Using this data he found that the probability of a service firm favoring a shared-control mode of entry increased as cultural distance increased, although they were not able to find an interaction between asset specificity and cultural distance as they initially hypothesized (Erramilli and Rao, 1993).
The same relation between entry mode choice and cultural distance was found by Agarwal (1994). He found that cultural distance increases the probability to enter in a joint venture, as dealing with a single contract partner seems less costly than internalizing and having to deal with many employees, suppliers, customers and governments. Using 1992/1993 data on 1519 cases of foreign direct investment by 402 Japanese firms, Padmanabhan and Cho (1999) researched the role of decision-specific experience on entry mode choice, finding that establishment and entry mode experience significantly and largely contributed to the outcome of these two choices. Cultural distance, acting as a control variable for this research, was not found to be significant.
Studies on the influence of cultural distance on entry mode choice in China (Zhao and Zhu, 1998) and Singapore (Rajan and Pangarkar, 2000) showed no significant influence of an increase in cultural distance on either the probability for joint ventures or wholly-owned subsidiaries as cultural.
higher-equity modes of entry. Their research among 231 Dutch, German, British and US firms showed that although in culturally distant countries with low country investment risk joint ventures were still the preferred entry mode, if country investment risk is high wholly owned modes of entry are preferred.
Following these results, we propose:
Hypothesis 6: Higher cultural distance will lead to a higher probability of a WO entry mode
For an MNE with a global strategic focus, a high cultural distance does not necessarily pose a problem. Following the reasoning by Orr and Levitt (2004) on high tech partnerships, if relatively simple subsidiaries are set up by global MNEs that do not depend heavily on institutions or “other informal practices and protocols in the local host environment”, cultural distance does not generate external friction because of a lack of intense cross-cultural interaction. Since the global MNE is mostly independent of these institutions and informal practices, it is not prone to increased transaction and opportunity costs that result from this external friction (Kogut and Singh, 1988). The lack of friction is not only true for aforementioned external friction, but also for internal friction, which arises when managers and staff from different cultures work together within a joint venture (Killing, 1983). The global firm will usually set up a wholly owned subsidiary which implies that close coordination by managers or staff from different countries and cultures will be rare. Therefore, there will be only few occasions where cultural differences will lead to internal friction, resulting in a low risk of increased costs due to this friction and therefore a low necessity to change the preference of wholly owned subsidiaries.
communication difficulties (Browne, Rugman and Verbeke, 1989), uncertainty (Lin & Germain, 1998) and misinterpretation (Orr and Levitt, 2004). Summarized, cultural dissimilarities can lead to friction through cross-cultural interaction. For the multidomestic MNE, the joint venture with local partners will be built upon reciprocal interdependence (Thompson, 1967) as the local partner will work on the marketing and production of the locally adapted product. These complex interactions require coordination that can be hampered by miscommunication or inefficiencies, leading to higher coordination costs as a result of internal friction (Galbraith, 1973). Also, misinterpretation of a culturally distant partner’s actions can lead to unjustified (mis)trust, which can lead to principal-agent problems and the costs that arise from them (Woodcock & Geringer, 1991).
Apart from internal friction, the multidomestic MNE faces external forms of friction to a larger extent than the global MNE due to the local responsiveness it seeks through customizing its production and marketing within the host country. The uncertainty of the unknown environment and the unfamiliarity of the trading protocols will lead to increased transaction costs when establishing trade relations or setting up contracts (Orr and Levitt, 2004). The uncertainty of the business and institutional environment for the expatriate can also lead to higher opportunity costs (Eriksson , 1997) as sub-optimal strategies are chosen and the results of these strategies are unforeseen. Although these costs can be mitigated by setting up a joint venture with a local partner, within an unfamiliar environment opportunistic behavior is hard to detect and prevent as the local partner is at an advantage due to the country-specific knowledge it has.
In occasions of great cultural distance which can¬ possibly lead to both internal and external friction and amounting costs, the multidomestic MNE may choose to internalize its operations rather than partner up with a local firm to achieve optimum local responsiveness. We therefore propose:
3. Methodology
3.1 Data Selection and Testing
The data used for this research comes from several primary and secondary sources. It mostly contains data from the personal database of Dikova (2006), which consists of data gathered through a survey in the form of a questionnaire. The survey was conducted among firms from the EU that had at least a ten percent ownership in a subsidiary in Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia or Slovenia. The survey contained questions regarding the host country, the entry and establishment mode, as well as several characteristics of both the firm and the subsidiary.
In order to supplement the data derived from this survey, as well as to establish the robustness of the primary data, data has been extracted from several other sources. The World Bank’s Governance Indicators data (Kaufmann, Kraay and Mastruzzi, 2005) was used to add objective data on the institutional environment of the host country.
Also, the cultural indicators of Hofstede (2001) were derived from his publications and personal websites.
Taking a random sample from the total population of EU firms setting up subsidiaries in CEE countries and comparing the number of employees and total sales (of which the data were derived from the online database Amadeus) with the observed values from the sample, we found no statistical difference between the means of the random sample and the surveyed sample on both accounts.
Additionally, by comparing the perceived institutions as derived from the dataset as collected by Dikova (2006) with the data collected from the World Bank Governance Indicators (Kaufmann et al., 2005), we find no statistical difference between the means of the perceived and the objective measure of institutional environment.
employees was compared to the number of employees of a firm according to the data in Amadeus. Similarly, annual sales provided in the survey was compared to reported total sales in Amadeus, which was available for 56 of the 142 firms. For all other firms operating revenue was obtained, so to keep an as low as possible difference between the underlying accounting theory on reported total sales and the numbers obtained from Amadeus, while at the same time keeping a high enough sample population for the reliability tests.
For the number of employees, an almost perfect correlation (correlation = 0.927, significance = 0.000) was found, while the difference between measured and obtained number of employees was not statistically significant (t = 0.676, df = 123, significance = 0.5). The amount of annual sales reported had a slightly lower correlation (correlation = 0.553, significance = 0.000), but the difference between measured and obtained amount of sales does not differ statistically significant (t = 1.143, df = 123, significance = 0.255). We can therefore conclude that the reported and obtained data can be treated as equal, and that there is no reason to doubt the information reported in the survey.
In order to control for common method variance in the data, Harman’s one-factor test was performed on all variables based on the primary data.3 The assumption of this test is that if a significant amount of common method variance exists, either a single factor will come out, or one factor will explain the majority of the variance (Podsakoff and Organ, 1986). The test showed a total of four factors, of which the largest explained a mere 20.6% of the variance. We can therefore conclude that the primary data does not suffer from common-method variance.
Additionally, the significance of the dummy interaction variables has been tested using an F-test, which showed a significant result at the 1% level. We can therefore assume that the dummy interaction variables add to the model’s R2, and can therefore safely be added to the model.
3
3.2 Variables
The dependent variable of this research is the percentage of ownership the MNE has in its foreign subsidiary. Due to the chosen statistical method of analysis (see 3.3), the variable has been converted into a dummy variable, taking the value of 1 if the firm holds whole ownership over the subsidiary, or a 0 if the subsidiary is a joint venture. Following existing literature, an ownership percentage of 10 percent or higher in a foreign enterprise will be seen as a foreign direct investment (see, for instance, Padamanabhan and Cho, 1999). Of the initial 193 observations included in the data, 134 had whole ownership.
The four independent variables as proposed in the hypotheses are the strategic focus of the firm, the asset specificity of the firm, the institutional environment of the host country, and the cultural distance between home and host country.
First, the strategic focus of the firm is derived from the survey by Dikova (2006), which contained four questions adapted from Harzing (2000; 2002). These questions measured the focus of the international competition in the industry of the subsidiary. Using cluster analysis, the results showed a clear distinction along the lines of pursuing economies of scale or local adaptation, which using a t-test was found to be statistically significant as well. This allowed the creation of a division between a global focus and a multidomestic focus of the MNE’s strategy. A dummy variable is used for this variable, where a 1 refers to a global strategic focus and a 0 to a multidomestic strategic focus. The results of this cluster analysis can be found in Table 1.
Secondly, asset specificity was measured by Dikova (2006) through a question in the survey what the perceived amount of annual sales spent on Research and Development (R&D) was, where a 1 denoted very low, and a 5 denoted very high amounts.
number means a more advanced institution. Scores are available for the years 1996, 1998, 2000 and 2002, from whence the scores became available on a yearly basis. In order to circumvent any problems with the odd-year observations from the dataset, these were treated as if one year later (so a 2001 observation would get a 2002 score). Any observations from before 1996 were linked to those of 1996. In order to create a more linear variable, the scores of this variable were adjusted to a range of 0 to 5.
Lastly, cultural distance was calculated using the formula Kogut and Singh (1988) developed for measuring the cultural distance between home and host country, using data collected on the four cultural dimensions as researched and categorized by Hofstede (1980).
As addition to the independent variables, several interaction variables have been created. By multiplying the strategic focus variable with each of the other independent variables, three additional independent variables have been created which can measure whether the sign and significance of an independent variable may be altered by the strategic focus of the MNE.
To control for influences on the outcome that are not accounted for by the independent variables, but might have a significant influence on the entry mode choice, six control variables have been added to the model. First, in order to control for intervention by the host-country’s regulations, restrictions on entry mode has been controlled for, since regulations that impede the inflow of wholly owned subsidiaries are quite common; forcing foreigners to cooperate with local partners, thereby allowing at least part of the extracted rents to flow back into the host country’s economy. Secondly, firm size of the mother-company is controlled for, following similar research (Brouthers, 2002; Nakos and Brouthers, 2002), since larger firms have more resources at their disposal to support their subsidiaries (Slangen, 2005). On the other hand, if the relative size of the subsidiary is too small, the subsidiary is of less importance and may therefore receive less support and attention from the parent firm (Hennart et al., 1998). Also, the propensity towards shared ownership decreases as the size of foreign assets relative to total assets increases (Agarwal, 1994). Therefore, relative firm size is added as a third control variable.
controlled for by adding a variable with the number of countries a firm has invested in. Regional international experience is to replace country-specific experience, due to the transitional nature of CEE. This influence is controlled for using two control variables: a dummy variable that measures whether the firm has invested in the region before, and one that measures the extent of these investments by measuring the number of CEE countries previously invested in. In this way, the relative similarity of these countries is modeled in addition to a linear learning variable.
3.3 Data Analysis
In order to test abovementioned variables for their influence on entry mode choice, a binomial logistic regression was performed. This type of analysis is chosen for several reasons. First of all, binomial logistic regression analysis is well suited for the dichotomous dependent variable. Also, the different types of measurements used for the independent and control variables (being continuous, dichotomous and scale-type) can be implemented simultaneously without any problems (Hair, Anderson, Tatham and Black, 1995). In order to check for any unwanted correlation between the variables, a bivariate correlation test has been performed, of which the results can be found in Table 2.
In the binomial logistic regression analysis, the influence of each variable on the likelihood that the entry mode of the subsidiary is wholly owned is estimated and quantified in a coefficient and a significance. The model can be written algebraically as:
Z Z e e Y P + = 1 ) (
where P is the chance of Y, the independent variable, being 1 (wholly owned), and Z is the combined linear model of all independent and control variables:
where β0 is the constant, β1 ... β11 are the coefficients, and X1 ….. X9 are the
independent and control variables.
For the analysis three models were created. Abovementioned model with all independent and control variables comprises the first model. The second and third model were created by splitting the data into a part of MNE’s focusing on a global strategy (such as economies of scale), and those MNE’s that were focusing strategically on local responsiveness. For these two models, the interaction variables were omitted, allowing a comparison between the two models on the influence of the independent on entry mode choice.
4. Results
Table 2 provides the mean and standard deviation of each variable, as well as the correlation between each variable. Most of the significant correlations can be explained rather easily, as they are transformations or multiplications of other variables and are therefore correlated. The significant correlation between cultural distance and institutional advancement can be explained as both are country-dependent variables and thus correlate per country rather than differ for each observation. Combining the correlation data with the Variance Inflation Factors (VIF) for the linear regression performed as shown in Table 3, can give us an indication on the existence of multicollinearity within the dataset. Apart from the score for strategic focus of the firm (VIF=7.326) and the score for asset specificity times strategic focus of the firm (VIF=6.865), all VIF-scores are below 5. Considering that the threshold value for the Variance Inflation Factor is 10, we can assume that no problems regarding multicollinearity are present in the dataset.
the control variables are included) has been performed on both subgroups after dividing the dataset in a global strategic focus and a multidomestic strategic focus group.
The fit of the model as represented in the R2 statistic, or in order to convey a more suitable statistic the Nagelkerke R2, is 0.281, which means that of all the model’s variance 28.1% is explained by the variables in the model. Since the binomial logistic regression is strictly about probabilities, the percentage of the outcomes (whole or shared ownership) correctly predicted based on the information contained in the variables is far more important. In the binomial logistic regression with all independent variables and control variables, 73.4% of the observations are correctly predicted based on the available information: this is a significant improvement compared to the 50% chance of correctly predicting the outcome based on utilizing random chance.
All hypotheses from Section 2 are tested in the binomial logistic regression, since the coefficients and their (non-)significance will show the (non-)existence of any relationship or influence between (several) variables and the probability of whole or shared ownership.
Hypothesis 1, regarding the influence of the predominant strategic focus of the multinational firm, predicts that more global strategy-oriented firms have a higher probability to have full ownership over their subsidiary than multidomestic strategy-oriented firms. As can be seen in Table 4, the coefficient for the strategic focus of the firm is positive (3.154) and significant (α < 5%), indicating that our predictions were correct and global strategy-oriented firms have a higher probability to have full ownership over their foreign subsidiary than multidomestic strategy-oriented firms.
For the binomial logistic regression of each different strategic focus as subgroup, we find that asset specificity is significant for the multidomestic strategy-focused group (α < 5%), but not for the global strategy-focused group. The positive sign (0.590) shows that for firms pursuing a multidomestic strategy higher asset specificity increases the probability that they will choose whole ownership over their subsidiary. Concerning Hypothesis 3, the moderator variable, i.e. the interaction variable between the strategic focus dummy variable and the asset specificity variable, is estimated to be significantly different from 0 as well (α < 5%). This implies that for firms pursuing a global or a multidomestic strategy, there are different effects of asset specificity on the probabilities of the outcomes. This is in line with hypothesis 3, which stated that the likelihood of a joint venture as entry mode would be negatively moderated by the degree of asset specificity. As the coefficient estimate is significant and has a negative sign (-0.866), we find that as asset specificity increases all firms have a higher probability to have whole ownership, but that for firms with a global strategic focus there is a negative effect that outweighs the positive effect. Therefore, firms pursuing a multidomestic strategy will have a higher incentive for setting up a wholly owned subsidiary in case of asset specificity than firms pursuing a global strategy, which is in line with our arguments. In section 5 this will be discussed in more detail.
Hypothesis 4, concerning the influence of the degree of institutional advancement of the host country on the probability of whole or shared ownership, is not supported as the coefficient estimate in the binomial logistic regression is non-significant at the α = 10% level. In the binomial logistic regression for both the multidomestic strategy-focused and global strategy-focused subgroup the regression coefficients for this variable are non-significant at the α = 10% level. Additionally, the interaction variable between the dummy variable of strategic focus of the firm and the institutional advancement of the host country is non-significant (again, at the α = 10% level). Therefore, we can not support Hypothesis 5.
country on the probability of whole- or shared ownership. The binomial logistic regressions after dividing the dataset in firms pursuing a global or a multidomestic strategy shows different signs for the global (-0.283) and the multidomestic (0.182) subgroup. However, neither is significant at the α = 10% level. Contrary to expectations arising from these results, the interaction variable of the strategic focus dummy variable and the cultural distance variable is mildly significant (α < 10%) and thereby shows that a difference exists between the probability firms with a global strategic focus and firms with a multidomestic strategic focus in choosing a wholly owned or shared ownership subsidiary. The sign of the coefficient estimate (-0.620) is in line with the prediction made in hypothesis 7, suggesting that a firm pursuing a global strategy will have a lower probability of choosing a wholly owned subsidiary than a firm pursuing a multidomestic strategy in case of higher cultural distance between the home and host country.
With regard to the control variables we find some mixed results. Government restrictions on the degree of ownership allowed for foreign owners over their subsidiary are significant (a < 5% and a < 1%) and negative (-1.137 and -1.653) for both the model containing the entire dataset and the model containing only firms with a global strategic focus. This sign is in line with expectations as government places restrictions on ownership in transitional economies, which is used to limit the outflow of profits to the home countries of the multinational firms in order for the local economy to profit from the economic gains of the MNE. For the multidomestic strategy-focused firm, the coefficient estimate was not significantly different from zero at the a = 10% level. Again, this seems in line with expectations as a multidomestic firm has a lower probability to have whole ownership over its subsidiary and therefore has less chance to face the government-imposed restriction. Firm size is non-significant at the a = 10% level for all three models. However, relative size of the subsidiary is significant (a < 0.1%, a < 10% and a < 1%) and negative (-2.556, -2.477 and -2.459) for all three models. None of the international and CEE experience control variables are significant in any of the models, nor is the previous CEE entry dummy variable.
5. Discussion and Limitations
This research aims to provide an overview of the various influences on entry mode choice, combining various theories into an eclectic model in order to answer the following research question:
What are the determinants of entry mode choices in CEE transition economies?
This research question will be answered using the following specific research questions: 1) Is entry mode choice influenced by asset specificity, the institutional environment and cultural distance?
2) Is entry mode choice influenced by the strategic focus of the firm?
3) Is the influence of asset specificity, cultural distance and the institutional environment on entry mode choice moderated by the strategic focus of the firm?
degree of ownership it wishes to hold over its subsidiary. Apparently, the higher transaction costs arising from uncertainty due to the transitional nature of the CEE economy does not provide enough incentive to increase or decrease its ownership. This may also be the result of several effects canceling each other out due to underspecification of the model, for instance the more general increased country risk because of low institutional development, that may increase transaction costs (Peng, 2003). Therefore, it increases the probability for a wholly owned subsidiary, while similarly the underdeveloped institutional environment depends highly on informal networks that are inaccessible without a local partner (Meyer, 2001), which may lower the probability for a wholly owned subsidiary. Lastly, considering cultural distance, we find no evidence to conclude that the cultural distance between home- and host country has any influence on the entry mode choice; a result similar to Padmanabhan and Cho (1999). The non-significant coefficient estimate shows that multinational firms are apparently unaffected in their degree of ownership as the costs of obtaining information on the host country, as well as the degree of perceived uncertainty, increase. The insignificance of the cultural distance variable can also be explained by the lack of a single influence of cultural distance on the degree of ownership firms choose over their subsidiaries, as the effect is perhaps limited to global firms who generally prefer whole ownership, while multinational firms who prefer a joint venture have no additional incentive to share ownership due to cultural distance. If so, such a situation will be presented in the discussion of research question 3). Concluding the first research question, we find that asset specificity has a clear positive effect on the probability of whole ownership, while the institutional environment and cultural distance seem to have no influence on the probability of a firm to have whole ownership. This might be due to, for instance, underspecification of the model as important influences have not been controlled for or relevant variables have not been included in the model.
chance to have a whole ownership than a firm pursuing a multidomestic strategic focus. This result corresponds with earlier theoretical works such as those by Prahalad and Doz (1987) and Bartlett and Ghoshal (1989), and the research and theory by Kim and Hwang (1992) and Luo (2001b). Apparently, an MNE with a global strategic focus wishes to avoid problems arising from shared ownership (Kemp, 1999) and benefit from a wholly owned international network that increases competition on a global level (Kim and Hwang, 1992). At the same time, a multidomestic firm values location-specific information and experience highly, in order to adapt quickly to a complex local market (Morrison and Roth, 1992), which is most efficiently done with a local partner and is enabled through the decentralized coordination and the freedom to build up long-term (shared ownership) relations with local stakeholders (Ghoshal and Nohria, 1989). In conclusion, we find the strategic focus of the multinational enterprise to have an influence on the degree of ownership a firm chooses, as global firms have a significantly higher probability to choose a wholly owned subsidiary.
Concerning the other moderating influence of strategic focus of the firm, namely on cultural distance, this result is in line with theoretical arguments such as the need for flexibility of the MNE to withdraw from the culturally distant country if the need arises (Kim and Hwang, 1992). Although this argument is applicable to multidomestic firms as well, the multidomestic firm is affected more by the cultural distance than the global firm, as the former can operate mostly independent of the cultural distance (Kogut and Singh, 1988), while the latter has more cross-cultural interaction which can lead to internal and external friction (Orr and Levitt, 2004).
Several limitations are applicable to this research. First of all, the research sample of 169 observations that have no missing variables could be improved by a larger-scale research, although the data has been proven to be representative (Dikova and Van Witteloostuijn, 2007). Additionally, the lag between filling in the questionnaire and the actual foreign direct investment might influence the results and allow biases as data needs to be recalled or found in sources that are old and therefore perhaps unreliable or misinterpreted. A survey focusing on more recent FDI might therefore increase the reliability of the dataset. Lastly, as this research focuses on Western MNEs investing in a selected (although rather exhaustive) set of CEE countries, the results obtained in this research can not be easily generalized to other countries and other flows of foreign direct investment. Therefore, this research functions well as an exploratory research, instead of a research from which general results and strong empirical proof are to be extracted.
6. Conclusion
This thesis presents the theoretical arguments and the methodological approach used for this research to investigate the significance of several theories on predicting the entry mode choice, as well as the influence of the strategic focus of the MNE on entry mode choice. Using 169 observations of various Western European firms setting up a
between whole ownership and joint venture. This shows that firms base their degree of ownership decision mainly on the presence of strategically important assets that need to be protected from interest-seekers with guile. Additionally, the influence of the firm’s strategic focus firm does not only have a direct relation with the entry mode choice, but also interacts with asset specificity and cultural distance. This shows a clear distinction in preferences as the multidomestic firm has a far lower probability of having a wholly owned subsidiary than the global firm, but also the willingness of multidomestic firms to give up their local partnership in case of high cultural distance or asset specificity
compared to global firms.
The significance of this research lies in the combination of several theories that are tested in a single dataset, as well as analyzed for their interaction with the strategic focus of the firm variable – a previously only sparingly used variable even though its theoretical contribution to the entry mode choice can not be discarded easily. Earlier findings on the variables used in this research are frequently isolated or driven by use of a single theory; in this research three well-known and widely accredited theories drive the combination of variables to provide an eclectic and comprehensive framework for the entry mode choice. The contribution of this research to the field of entry mode choice is therefore significant, as it gives a prominent role to the strategic focus of the firm by testing it as a stand-alone variable, as well as for its moderating influence on other variables. The results of this research can therefore provide a promising basis for future research to further explore the moderating influence of several other variables, as well as the interaction between some of the used variables, for instance between cultural distance and asset specificity
following Erramilli and Rao (1993).
The significance of this research is not limited to academic fields, however. For firms that are exploring a possible investment in a foreign subsidiary, this research provides an overview of and insight into the entry mode choice of other firms, combined with several important properties of these companies. Using these, managers may be better informed in assessing the important influences of several properties of the MNE and its strategic focus, thereby being able to make a well-argued and informed decision.
be as decisive as hoped for. Future research could be done with a larger dataset,
providing more reliable coefficients that adhere to the rule of thumb of 50 observations per independent variable in a logistic regression analysis. Additionally, the practical value of this research would be improved if it could be combined with an analysis of subsidiary performance, most importantly on the difference of performance between firms choosing the predicted entry mode and those that deviate from this prediction. Lastly, there are undoubtedly several (lesser known and researched) theories and
References
Agarwal, S. (1994), ´Socio-cultural distance and the choice of joint ventures: A contingency perspective”, Journal of International Marketing, 2: 63-81.
Anand, J., and Delios, A. (1997), “Location specificity and the transferability of
downstream assets to foreign subsidiaries”, Journal of International Business Studies, 28: 579-603.
Anderson, E. and Gatignon, H. (1986), “Modes of foreign entry: A transaction cost analysis and propositions”, Journal of International Business Studies, 17: 1-26.
Barkema, H.G., Baum, J.A.C., and Mannix, E.A. (2002), “Management challenges in a new time”, Academy of Management Journal, 45: 916-930.
Bartlett, C.A., and Ghoshal, S. (1989), Managing Across Borders, Harvard Business School Press: Boston, MA.
Beamish, P.W., and Banks, J.C. (1987), “Equity joint ventures and the theory of the multinational enterprise”, Journal of International Business Studies, 18: 1-16.
Birkinshaw, J. (1996), “How multinational subsidiary mandates are gained and lost”, Journal of International Business Studies, 27: 467-495.
Boisot, M., and Child, J. (1999), “Organizations as adaptive systems in complex environments: the case of China”, Organization Science, 10: 237-252.
Brooks, C. (2002), Introductory Econometrics for Finance, Cambridge University Press: Cambridge.
Brouthers, K.D. (2002), “Institutional, cultural and transaction cost influences on entry mode choice and performance”, Journal of International Business Studies, 33: 203-221. Brouthers, K.D., and Brouthers, L.E. (2001), “Explaining the national cultural distance paradox”, Journal of International Business Studies, 32: 177-189.
Brouthers, K.D., and Nakos, G. (2004), “SME entry mode choice and performance: A transaction cost perspective”, Entrepreneurship Theory and Practice, 28: 229-249. Brown, L.T., Rugman, A.M. and Verbeke, A. (1989), “Japanese joint ventures with Western multinationals: Synthesizing the economic and cultural explanations of failure”, Asia Pacific Journal of Management 6: 225-242.
Caves, R.E., and Mehra, S. (1986), “Entry of foreign multinationals into U.S.
manufacturing industries”, in Porter, M. (ed.), Competition in Global Industries, Harvard Business School: Boston.
Chen, S.S., and Hennart, J-F. (2002), “Japanese investors’ choice of joint ventures vs wholly-owned subsidiaries in the US: The role of market barriers and firm capabilities”, Journal of International Business Studies, 33: 1-18.
Cho, K.R., and Padmanabhan, P. (1995), “Acquisition versus new venture: The choice of foreign establishment mode by Japanese firms”, Journal of International Management, 1: 255-285.
Clague, C., ed. (1997), Institutions and Economic Development, John Hopkins University Press: Baltimore, MD.
Coase, R. (1937), “The nature of the firm”, Economica, 4: 386-405.
Datta, D.K., Herrmann, P., and Rasheed, A.A. (2002), “Choice of foreign market entry mode: Critical review and future directions”, in Hitt, M.E., and Cheng, J.L.C. (eds.), Advances in International Management, 14, Elsevier Science, 2002, 87-155.
Delios, A., and Beamish, P.W. (1999), “Ownership strategy of Japanese firms:
transnational, institutional, and experience influences”, Strategic Management Journal, 20: 915-933.
Dikova, D. (2006), Studies on Foreign Direct Investment in Central and Eastern Europe: Establishment Modes, Performance and Strategies of Western European Enterprises, Labyrint Publications: Ridderkerk.
Dikova, D., and Van Witteloostuijn, A. (2007), “Foreign direct investment mode choice: entry and establishment modes in transition economies”, Journal of International
Business Studies, 38: 1013-1033.
Doz, Y., and Prahalad, C.L. (1991), “Managing DMNC’s: A search for a new paradigm”, Strategic Management Journal, 12: 145-164.
Dunning, J.H. (1981), International Production and the Multinational Enterprise, George Allen & Unwin: London.
Erramilli, M.K., and Rao, C.P. (1993), “Service firms international entry mode choice: A modified transaction-cost analysis approach”, Journal of Marketing, 57: 19-38.
Erramilli, M.K., Agarwal, S., and Kim, S. (1997), “Are firm-specific advantages location-specific too?”, Journal of International Business Studies, 28: 735-757. Franko, L.G. (1971), Joint Venture Survival in Multinational Corporations, Praeger Publishers: New York.
Galbraith, J.R. (1973), Designing Complex Organizations, Addison-Wesley Longman Publishing: Boston.
Gatignon, H., and Anderson, E. (1988), “The multinational corporation’s degree of control over foreign subsidiaries: An empirical test of a transaction cost explanation”, Journal of Law, Economics and Organization 4: 305-336.
Geringer, J.M., and Hebert, L. (1989), “Control and performance of international joint ventures, Journal of International Business Studies, 20: 235-254.
Ghoshal, S., and Nohria, N. (1989), “Internal differentiation within multinational corporations”, Strategic Management Journal, 10: 323-337.
Granovetter, M. (1985), “Economic action and social structure: The problem of embeddedness”, American Journal of Sociology, 91: 481-510.
Hair, J.F., Anderson, R.E., Liao-Tatham, R.L., and Black, W.C. (1995), Multivariate Data Analysis, Prentice Hall: Upper Saddle River, NJ.
Harzing, A.W.K. (1999), Managing the Multinationals: An International Study of Control Mechanisms, Edward Elgar: Cheltenham.
Harzing, A.W.K. (2000), “An empirical analysis and extension of the Bartlett and Ghoshal typology of multinational companies”, Journal of International Business Studies, 31: 101-120.
Harzing, A.W.K. (2002), “Acquisitions versus greenfield investments: international strategy and management of entry modes”, Strategic Management Journal, 23: 211-227. Hennart, J-F. (1991), “The transaction costs theory of joint ventures: An empirical study of Japanese subsidiaries in the United States”, Management Science, 37: 483-497.
Hill, C.W.L., and Kim, W.C. (1988), “Searching for a dynamic theory of the
multinational enterprise: A transaction cost model”, Strategic Management Journal, 9: 93-104.
Hill, C.W.L., Hwang, P., and Kim, W.C. (1990), “An eclectic theory of the choice of international entry mode”, Strategic Management Journal, 11: 117-128.
Hofstede, G. (1980), Culture’s Consequences: International Differences in Work-related Values, Sage Publications: Beverly Hills.
Hofstede, G. (2001), Culture’s Consequences: Comparing Values, Behaviors, Institutions and Organizations Across Nations, Sage Publications: Thousands Oaks, CA.
Hout, T., Porter, M.E., and Rudden, E. (1982), “How global companies win out”, Harvard Business Review, 60: 98-108.
Johanson, J. and Wiedersheim-Paul, F. (1975), “The internationalization of the firm: Four Swedish case studies”, The Journal of Management Studies, 12: 305-322.
Jones, G.J., and Hill, C.W.L. (1988), “Transaction cost analysis of strategy-structure choice”, Strategic Management Journal, 9: 159-172.
Kaufmann, D., Kraay, A., and Mastruzzi, M. (2005), “Governance matters IV:
Governance indicators for 1996, 1998, 2000, and 2002”, World Bank Policy Research Working Paper No. 3630, Washington, DC.
Kemp, R.G.M. (1999), Managing Interdependence for Joint Venture Success, Datawyse: Maastricht.
Killing, J.P. (1983), Strategies for Joint Venture Success, Routledge & Kegan Paul: London.
Kim, C.W., and Hwang, P. (1992), “Global strategy and multinationals’ entry mode choice”, Journal of International Business Studies, 23: 29-53.
Kogut, B., and Singh, H. (1988), “The effect of national culture on the choice of entry mode”, Journal of International Business Studies, 9: 319-332.
