The role of culture in foreign market entry modes : investigating the relationship between cultural distance and entry mode choices for Dutch MNEs

(1)

Amsterdam Business School

The Role of Culture in Foreign Market Entry Modes: Investigating the

Relationship between Cultural Distance and Entry Mode Choices for Dutch

MNEs.

By: Sean Hickman Student Number: 6077277 Submission Date: 11-08-2014

MSc. in Business Studies – International Management Track First supervisor: Stephan von Delft

(2)

Abstract

This study investigated entry mode decisions by Dutch firms engaging in foreign direct investment (FDI). In particular, it studied a potential inverted U-shaped relationship between entry mode choice and cultural distance, as opposed to the traditional linear relationship in which these variables are thought to interact. The major rationale behind the study was the so far inconsistent results on the matter. Building on a previous study that found this type of non-linear association between the variables for Japanese firms, this study failed to replicate those findings. The results do show a significant interaction between entry mode and cultural distance, but this interaction appears to be linear for Dutch firms. Furthermore, other factors such as firm performance, industry, and host country development did not significantly influence entry mode choice, but firm size did have a significant effect.

(3)

4.6 Descriptive Statistics and Analysis for 5th Hypothesis………...40 5. Discussion……….…41 6. Limitations……...……….…...45 6.1 Sample……….……….……45 6.2 Measures……….…….45 6.3 Analysis………...46 7. Future Research………..………47 8. Conclusion………48 9. Bibliography………...……….51 10. Appendices……….56 10.1 Appendix A………56 10.2 Appendix B………60

(5)

1. Introduction

In an era of growing globalization many companies are increasing their international business activities, but while this was traditionally done through the means of imports and exports, there is an increasing trend of directly investing abroad, as foreign direct investment (FDI) rates have increased drastically throughout the world in recent decades (Bende-Nabende, 2002). However, the way in which investment abroad takes place varies significantly per company and country, as a range of factors contribute to companies’ entry modes into foreign markets differing from case to case. Pan & Tse (2000) distinguish equity-based from non-equity-based foreign market entry modes. Non-equity-based entry modes are those traditional ones such as exports and contractual agreements, while equity-based entry modes are examples of more contemporary means of international business activities in the form of foreign direct investment. Pan & Tse (2000) also identify four main and general forms of FDI activities, namely greenfield investments, acquisitions, joint ventures, and the acquiring of shares. Investment abroad by means of greenfields and acquisitions lead to wholly owned subsidiaries (WOS) of a company that have no association with or dependence on any other firm, while collaborating with local firms lead to equity joint ventures in which two or more firms have a stake in the newly established subsidiary.

The choice of entry mode has been extensively studied in international business research, with its effect on performance being regarded as one of the foremost reasons why. Brouthers (2002) shows that the choice of entry mode indeed has implications for firm performance, which makes it essential for companies to carefully make a choice as to what entry mode will be used when entering foreign markets, as firms will evidently want to choose the entry mode which will yield the highest performance.

(6)

There are a number of factors which likely play a role in the choice of entry mode, yet the extensive empirical research on the topic has yet to provide us with concrete information as to which factors are of greatest influence in determining entry mode choice. According to Kogut & Singh (1988), firms will likely face cultural challenges resulting from entering less known markets when determining FDI decisions and entry mode choices, and as such culture is considered a major factor that will influence companies’ entry mode decisions. International business research on entry modes has therefore also emphasized culture as a potential influencer, yet results on this show no clear consensus (Wang & Schaan, 2008; Morschett, 2010). This lack of consensus amongst researchers on the topic is highly problematic, as internationalizing companies and their managers are unaware of the consequences of their entry mode choices, which will likely prevent them from yielding the greatest potential performance from their international business activities. The cost of doing business abroad is undoubtedly affected by the culture of those where the foreign business is taking place, as international business activities inherently involves interactions with local populations, and these interactions are likely to be affected by their culture.

Without properly assessing the cultural implications of doing business in certain countries, even successful firms with proven track records in one country might miserably fail in another country, which further emphasizes the need to establish the extent to which cultural factors have an effect on international business performance. A noteworthy and recent example of where a successful firm turned out to be much less successful in a foreign market is the case of Home Depot in China (Gao, 2013). Since its foundation in the United States in the late 70s, the company grew out to be the leader in the US home improvement and DIY market. The business model which proved to be so successful in the American market turned out to be a failure in the Chinese market, which is mostly attributed to the cultural differences that Home Depot failed to foresee, which in this particular case generally came down to the differences in the Chinese

(7)

home improvement market and the lack of experience with the DIY concept. As such, after years of losses, Home Depot decided to shut down all its remaining stores in China by 2012 and fully exited the market (Gao, 2013). The company’s unsuccessful venture in China was not unique however, with Mattel Inc. and Best Buy, two major American retail companies, having faced similar problems in the past, which had resulted in both companies withdrawing from the Chinese market. These cases prove that a replication of a successful strategy in a home country will not necessarily lead to that strategy being as successful in other host countries. Firm characteristics will undoubtedly have an effect on performance, with companies such as Home Depot having initially become so successful due to their unique business models, yet it is a common and erroneous assumption that these successful models can be replicated in other countries. External factors clearly also play a role in the extent to which a company’s performance is successful, and when entering foreign markets culture is likely to be one of the most influential external factors.

It is therefore of interest to both companies and their managers to carefully assess and analyze the different modes of entering foreign markets, so that the eventual chosen entry mode best reflects the interest of the firm whilst simultaneously addressing the cultural difference with the country/market they are entering. This research on culture is most commonly focused on the concept of cultural distance, developed by Geert Hofstede (1980). He found that people from different countries have different cultures by measuring four distinct attributes or values, later adding a fifth, which are work-related and therefore have an influence on business. These attributes, or dimensions, include power distance, individuality versus collectivism, masculinity versus femininity, uncertainty avoidance, and long term versus short term orientation. According to Hofstede, these attributes differ per country, and the differences between these dimensions give rise to the concept of cultural distance. Depending on its characteristics and norms and values of its population, countries score on a scale of 1 to 120 for each dimension.

(8)

The differences between each dimension for two given countries together make up an average cultural distance between the two. Since this concept was developed by Hofstede, it has been widely used in international business research to study cultural differences in general, but has also most frequently been used to investigate the effect of culture on performance and entry modes (Barkema, 1996; Cho & Padmanabhan, 2005; Wang & Schaan, 2008; Morschett, 2010; Chang et al., 2012).

High cultural distance will likely result in major differences in the way business activities are run which will in turn have implications for firm performance. In order to deal with cultural differences, the choice of entry mode in relation to cultural distance has frequently been studied by international business research, with some noteworthy early research providing evidence for cultural distance affecting the entry mode choice (Kogut, 1988; Agarwal and Ramaswani, 1992) Therefore, it is not only important for MNEs to carefully decide where to locate their FDI activity, but it is also crucial to determine which entry mode to choose as different modes might be more or less successful depending on the cultural distance between the home and host country.

Because previous studies have shown a positive association between entry mode and performance (e.g. Brouthers, 2002; Chen, 2003), further emphasizing the need for companies to be cautious when selecting entry modes for their investments abroad, firms will want to know what entry modes yield the highest performance in countries with a high cultural distance when compared to countries with a low cultural distance. Different entry modes lead to different ways in which the subsidiary is managed, while foreign management is likely perceived more positively or negatively depending on the culture of the country. Some countries have stronger institutional regulations which will likely also affect the decision of entry mode. A recurring assumption is that the management of WOS will be more costly in a culturally distant country, and companies will be more geared towards opting for joint ventures (Canabal, 2008).

(9)

However, numerous studies on the topic have also questioned the relationship between cultural distance and entry modes, showing no direct association (Harzing, 2003, Tihanyi and Russell, 2005; Morschett et al., 2010; Slangen and van Tulder, 2009; Chang, 2012). This has ultimately lead to international business research on this popular topic being contradictory and therefore inconclusive.

More recently, Wang & Schaan (2008) attributed this inconsistency to the recurring notion of there being a linear association between cultural distance and entry mode choice, and their study revealed a U-shaped relationship between the two variables instead. It is therefore of interest to MNEs to know whether this is indeed the case, as this will likely solve the cultural distance paradox and help them choose an appropriate entry mode.

Because the choice of entry mode is considered an integral part of the internationalization process, and because at least part of the existing literature suggest the choice of entry mode will have an impact on performance, this report will delve deeper into the topic of how cultural distance may affect firm’s choices of entry modes. The report will build upon Wang & Schaan’s work (2008) which points towards the possibility of a U-shaped relationship between cultural distance and the choice of WOS when entering foreign markets, as was shown by their Japanese examples. With no other studies having further investigated this association, this assumption still lacks considerable evidence and because of the importance of the issue at stake this notion should be further investigated. Showing an association between the two variables for countries other than Japan would further help support this claim. Whether their findings can be replicated using a different home country is of particular value to firms and their managers all around the world, as it well help depict a more general trend of the association between cultural distance and entry mode, and help them chose most appropriately. As an industrialized and developed European country with a significant amount of outward FDI flows, the Netherlands represents

(10)

an interesting example to investigate how the foreign activities by MNEs from this country are affected by cultural distance when determining foreign entry modes.

Following Wang & Schaan’s (2008) study, the report will also focus on WOS and joint ventures as they are the most recurring forms of equity-based entry modes and FDI. This leads to the following research question for this study:

Is there a U-shaped relationship between cultural distance and the choice of wholly owned subsidiary for Dutch MNEs?

(11)

2. Theoretical Framework

A vast amount of literature in international business research has been dedicated to the internationalization process, the different types of FDI activities, motives behind FDI and entry mode decisions for foreign markets, and external factors such as culture and performance. The different studies vary widely both in terms of entry choices being investigated and in terms of possible determinants thereof, making it difficult to draw overall conclusions. This section will highlight the main theoretical background and terminology of the study by firstly discussing the internationalization process and the role entry mode decisions plays in it. Secondly, several theories attempting to explain the choice of entry mode will be elaborated on, such as the transaction-cost theory and eclectic paradigm. Consequently, the impact of entry modes on firm performance will be addressed, which will in turn be followed by a discussion on the concept of culture and its perceived effect on FDI decisions and entry mode choices. The discussion on culture will lead to a definition of cultural distance concept, and a review on the literature which have addressed the relationship between this cultural distance and entry mode choice.

2.1 Internationalization Process

The concept of internationalization suffers from a considerable lack of consensus as to how it is defined, but generally speaking it is considered to be the outward movement in a firm’s international operations (Turnbull, 1987). Anderson (1997) labels internationalization as a strategy process of “adapting exchange transaction modality to international markets”, which differs from other types of strategy processes on two main dimensions. Firstly, it deals with the question of which international markets to invest in, and secondly it deals with what strategy should be used for entering these markets. There are several different theories and approaches dealing with the internationalization process, all of which attempt to explain the motives and ways in which firms increase their involvement in foreign markets. Because so many different

(12)

theories exist, it would be extremely challenging to discuss all these different approaches. Yet some of these are rather significant and recurring theories on the topic of entry mode choices, and will therefore be elaborated on, as well as linking them to the concept of culture. These include the transaction cost approach, the Uppsala model, and the ownership, location and internalization (OLI) framework.

2.1.1 Transaction-cost Approach

Economic theory of FDI activities has been mostly centered on industry- and firm-level variables instead of country-level variables (Kogut & Singh, 1988). The most commonly used theory in explaining the motives behind certain entry modes in the literature is the transaction-cost theory, centered on the firm as the unit of analysis, as the traditional theory for entry mode studies (Anderson, 1997; Makino & Neupert, 2000; Chen, 2003; Canabal, 2008; Morschett, 2010). According to this theory, firms need to create competencies that will minimize costs and inefficiencies related to entering and operating in foreign markets (Williamson, 1979). This comes down to specific assets, the frequency of economic exchange and uncertainty surrounding the exchange of resources between buyer and seller characterizing the transactions between them (Anderson, 1997). Examples of assets typically include the likes of firm proprietary knowledge, products, processes, brand name, product differentiation and marketing skills (Chen, 2002). Following this theory, companies entering foreign markets are expected to choose entry modes which will reduce the transaction costs the most, and one which will provide the best return on investment. Although this approach greatly revolves around firm assets and competencies, culture will likely also have an influence within it. The actual process of entering foreign markets will bring about transaction costs, which will ultimately depend on when, where and how they enter these markets. Culturally different countries will arguably provide higher transaction costs and a lower return on investment, which will likely lead to

(13)

firms sticking to culturally similar countries instead of culturally different countries once they start internationalizing.

2.1.2 The Uppsala Internationalization Model

The internationalization process of firms is often associated with the work of Johanson & Wiedersheim-Paul (1975) and Johanson & Vahlne (1975). In their model the entering of foreign markets is a gradual process connected to time and experience. The internationalization process is divided into four successive stages, each representing a higher degree of international involvement. These stages consist of the following (Anderson, 1993):

1) No regular export activities

2) Exports through independent representatives or agents 3) Establishment of an overseas subsidiary

4) Establishment of overseas production/manufacturing units

This approach is also known as the entry mode as a chain of establishment, with increased market knowledge linked to increased market commitment and vice versa (Anderson, 1997). The underlying idea is that firms first gain experience in their home market before moving abroad, and when they do they enter markets which are either culturally or geographically similar, before moving to more distant countries. Culture therefore has at least a partial significance in this approach, with the overarching idea that culturally similar markets are entered at an earlier stage of the internationalization process and culturally different ones at a later stage. In this model, the outcome of the events of each stage is the input for the next stage and internationalizing firms strictly stick to these steps in the model in order to ensure successful ventures in foreign markets.

Although being a crucial and recurring theory in entry mode studies, the chain of establishment theory fails to include cooperative modes of entry choices in its model, presenting a significant

(14)

weakness as it therefore does not help to explain the motives behind choosing a joint venture over a wholly owned subsidiary. Moreover, the idea of a gradual and step-by-step process of internationalization is arguably an outdated concept, with numerous factors such as globalization and increases in technology playing a role in undermining this notion, as it is nowadays conceivable for firms to be international from the outset without the experience or market knowledge that this theory suggests it should have.

2.1.3 The Eclectic Paradigm (OLI Model)

Meanwhile, the Ownership, Location and Internalization (OLI) framework is also frequently used to assess entry mode choices (Agarwal and Ramaswani, 1992; Canabal, 2008). This theory, also known as the eclectic paradigm, asserts that entry mode choices are dependent on three factors: ownership advantages (firm-specific assets such as size and multinational experience and skills such as the ability to develop differentiated products), location advantages (the attractiveness of a specific country in terms of its market potential, investment risks, and institutional attributes) and internalization advantages (related to the transaction-cost approach) (Dunning, 1993; Anderson, 1997). Ownership advantages are influenced by control, costs, and benefits of inter-firm relationships; location advantages are about resource commitment, costs, and availability; while internalization advantages concern the reduction of transaction and coordination costs (Canabal, 2008). This approach is therefore a multi-theoretical framework as it comprises theories based on trade, resources, and transaction-costs. According to the OLI framework all these factors will have an influence on entry mode decisions, with all three categories of advantages being present when firms embark on FDI missions, as opposed to for example licensing or exporting which would only provide one and two of the categories of advantages respectively (Dunning, 1993). Agarwal and Ramaswani (1992) find that there are significant interactions between the three variables and thereby influence entry mode choice, with large and multinational firms more likely to opt for sole or joint ventures than smaller and

(15)

less multinational firms. Of the advantage categories location is arguably the most relevant in terms of cultural differences, with culturally similar countries providing location advantages and culturally distant countries providing location disadvantages. However, the location category of the OLI framework is the only one of the three which deals with external factors such as culture, with the other two categories focusing much more on internal and firm-level factors. But although these other two categories will likely be insufficient in explaining which foreign markets firms should invest in, firm-level factors and competencies will arguably have an influence on what the best mode is for entering these markets.

2.1.4 The Organizational Capability Perspective

A less well-known but nevertheless important perspective explaining entry mode choices is the organizational capability perspective, which can either be seen as an alternative to or complementary to the transaction-cost approach, and is based on the concept of bounded rationality (Anderson, 1997). According to this approach, the firm is a bundle of static and transferable resources which are transformed into capabilities through firm-specific processes (Madhok, 1997). Intangible resources such as skills and competencies are seen to be either specific and embedded or generic and nonembedded, with a high embedded-to-generic knowhow ratio tending to lead to a preference for internalization and collaboration. This perspective clearly takes a firm-level approach without taking into account external factors such as culture to determine when and where to internationalize, but as is the case with the eclectic paradigm model, it is possible that this perspective will help explain how particular firms internationalize.

Besides these major approaches and perspectives of the internationalization process, other factors such as market attractiveness, the competitive situation, control, risks, and institutional theories have also all been used to a certain degree to explain the nature behind internationalizing firms (Brouthers, 2002; Chen, 2003; Canabal, 2008; Morschett, 2010).

(16)

2.2 Entry Mode Choices

With the internationalization process dealing not only with what foreign markets to enter, but also with the different manners in which these markets can be entered, it is apparent that entry mode decisions are crucial for internationalizing companies and inherently linked to their internationalizing strategies. The choice of what entry mode to choose is widely regarded as one of the most critical decisions in the internationalization process (e.g. Kogut & Singh, 1988; Anderson, 1997; Morschett, 2010) and has therefore been extensively studied in international business research. An entry mode is defined by Sharma & Erramilli (2004) as “a structural agreement that allows a firm to implement its product market strategy in a host country either by carrying out only the marketing operations, or both production and marketing operations there by itself or in partnership with others”.

In the most general sense, entry modes can be divided into equity-based modes of entry and non-equity-based modes of entry. Within the latter we can distinguish between contractual agreements and within the former distinguish wholly owned operations from equity joint ventures (Pan & Tse, 2000). In this particular study we are only concerned with equity-based modes of entry as these, unlike the non-equity-based modes, reflect direct investments in foreign markets and will therefore be more likely to be affected by factors such as cultural differences, as they arguably carry more risks with them . Pan & Tse (2000) also assert that when moving from non-equity-based modes to equity-based modes of entry, there is a continuum along which resource commitment, risk exposure, control and profit potential increase in levels. As we move along this continuum, cultural differences will arguably become more important, as a larger commitment and presence will more susceptible to cultural challenges than a lower commitment or presence.

Centered around the equity-based forms of FDI, Kogut & Singh (1988) identify three main kinds of foreign market entry modes: acquisitions, joint ventures, and greenfield investments.

(17)

The first refers to the purchase of sufficient stock in already existing companies that will amount to at least some form of control. Meanwhile, a joint venture is considered the pooling of assets in a common or separate organization by two or more firms who share joint ownership and control over the use of these assets (Kogut & Singh, 1988). Finally, a greenfield investment is a start-up in new facilities, and can be either wholly owned or a joint venture. For simplification purposes, and following the example of Kogut & Singh, this report will classify start-up investments which are wholly-owned under greenfields, and those involving shared ownership under joint ventures.

2.3 Entry Modes’ Impact on Performance

The importance of entry mode decisions is partially due to the idea that entry modes have a considerable impact on firm performance (Brouthers, 2002; Chen, 2003; Canabal, 2008). Although much literature on the topic is also contradictory or failed to show a noteworthy association between the two variables, there is growing evidence that there is indeed a relationship between the two, which ultimately explains why studies on entry mode decisions are so recurring. Brouthers (2002) found that firms using entry modes predicted by transaction cost theory significantly performed better, both in financial and non-financial terms, than firms whose entry mode choice were not predicted by the theory. Using the example of Chinese firms, Chen (2003) also found evidence for the transaction cost explanation of entry mode decisions significantly predicts future performance. This furthermore suggests that the choice of entry mode is one which has to be taken with utmost care, as it will have major implications for the future of the firm. Performance could also be affected more indirectly by entry mode choice, as for example, there will likely be managerial implications for these decisions, as different cultures are likely to adopt different managerial styles, which will undoubtedly play a role in determining what form of entry mode will lead to a greater success. Taking into account the main models which try to explain entry mode decisions, it is clear that the relation between

(18)

entry mode choice and performance is a complex one which affected by internal factors such as firm competencies, knowledge and experience, and external factors such as market size, industry type, and culture.

2.4 Culture & Cultural Distance

One of the most recurring topics in explaining entry modes is that of culture, cultural difference and/or cultural distance between countries (Barkema, 1996; Harzing, 2003; López-Duarte, 2010). In general the distinction is made between culturally similar countries and culturally different countries, each of which would be associated with a different entry mode decision. But the effect of culture in the international business world is much more wide-ranging than solely on entry mode decisions, and has been studied in a variety of contexts. Zeng et al. (2013) find that MNE learning abilities are eroded by cultural differences, leading to incorrect inferences and learning. As such, the authors find that a low level of experience in dissimilar countries is more likely to lead to subsidiary mortality. This correlation is apparently weaker when the firm has a prior presence across different cultures, stronger if its international expansions was at a fast pace, and negative once the firm has accumulated a lot of experience in the particular host culture. Shane (1994) also points to the differences in national perceptions of transaction costs, which ultimately imply that cultural variables have at least some explanatory power in FDI motivations. Meanwhile, cultural differences are also thought to be influential in affecting individual and employee perceptions of ethical issues, with these perceptions altering significantly per culture (Armstrong & Sweeney, 1994).

Within studies of culture in international business studies, cultural distance remains an important recurring concept. It is generally regarded as uncertainty arising from informal host environments, as opposed to those arising from formal host environment such as political risk (Slangen & van Tulder, 2009; López-Duarte, 2010; Samiee, 2013). As such the concept of cultural distance is more subjective than formal risks, and therefore less easy to measure. This

(19)

could help explain the somewhat contradictory evidence as to whether cultural distance indeed has or has no effect on entry mode choice. Although often considered a separate explanation or theory, it arguably both influences and is influenced by factors such as transactions costs and ownership, location and internalization advantages.

According to Hostede (1984), culture refers to the “collective programming of the mind which distinguishes the members of one category of people from those of another”. The category used by Hofstede is national culture, implying the existence of cultural differences between different countries. As such, cultural distance is the difference between two national cultures, which in entry mode studies comes down to the difference between home and host country.

In studying this concept of cultural distance, Hofstede’s cultural dimensions are widely regarded as the traditional tool in assessing this distance (e.g. Kogut, 1988; Samiee, 2013). Hofstede’s dimensions include power distance, individuality versus collectivism, masculinity versus femininity, uncertainty avoidance, and long term versus short term orientation (Hofstede, 1994). These dimensions together help make up a quantifiable assessment of the cultural distance between two countries, and is therefore the point of reference in almost all the literature dealing with cultural distance.

2.5 Effect of Cultural Distance on Entry Mode Choice

The effects of cultural distance on the choice of entry mode is a recurring topic, with the general idea that larger cultural distance will lead to greater transaction costs and less likelihood of success, yet the empirical evidence suggests there is little agreement on an existing a link between the two concepts, as parts of the literature suggest there is indeed a positive association between the two (Kogut & Singh, 1988; Agarwal and Ramaswani, 1992; Chen, 2003) while other studies refute this by either showing negative association or no association at all (Harzing, 2003, Tihanyi and Russell, 2005; Morschett et al., 2010: Slangen and van Tulder, 2009; Chang,

(20)

2012). Meanwhile, some studies such as that of Cho & Padmanabhan (2005) claim there is an association when taking into account moderating effects, in this case decision-specific experience, while Hennart & Larimo (1998) claim that although cultural characteristics of home countries do not influence entry mode choices, cultural distance between countries does.

The correlation between the two principles is more complex than cultural distance simply predicting the choice of entry mode, with this choice likely having an effect on the way the country and its culture is perceived. As such, if lower levels of equity are associated with more dissimilar cultures, firms opting for joint ventures over solely owned ventures will be perceived as having their FDI activity taking place in culturally distant countries. The interaction between cultural distance and entry mode was most famously researched by Kogut and Singh (1988), where they found strong empirical evidence for entry mode selection being influenced by cultural factors, with for example the management of multinational corporations likely being influenced by a dominant country culture, which would have further implications for the choice of foreign market to invest in and with what entry mode.

The studies that have shown an association mostly suggest that the likelihood of using cooperative entry modes such as joint ventures seems to increase with higher cultural distance (e.g. Kogut, 1988; Chen, 2003), which would in turn suggest that the chance of starting a wholly owned subsidiary (WOS) would increase with lower cultural distance. Slangen and van Tulder (2009) also point towards the notion that in uncertain external environments, cooperative entry modes such as joint ventures are more likely to occur than WOS. Samiee (2013) supports this notion by also showing how cultural distance plays a role in deciding whether to opt for a partnership or a sole ownership when entering a foreign market. However, Chang et al. (2012) suggest the complete opposite, claiming that joint ventures with firms in countries with higher cultural distance will dramatically increase the risks of MNEs, and therefore WOS are more likely to be set up.

(21)

Nevertheless, the before mentioned studies which provided empirical evidence for a link between cultural distance and entry mode choice, suggesting important implications for MNEs when they wish to enter foreign markets, remain noteworthy and cannot be ignored.

Other studies on the effect of cultural distance include that of Barkema & Bell (1996) in which they discuss how cultural distance between countries may affect the learning process of the firm, which in turn will have an effect on their choice of entry modes when deciding to enter foreign markets in the future. Meanwhile, Tihanyi & Russell (2005) explore the effects of cultural distance on international diversification and conclude that there is no strong association between the two, while Slangen and van Tulder (2009) find that neither informal risks such as cultural distance, nor formal ones such as political risks have an impact on entry mode. Moreover, Harzing (2003) and Chang (2012) openly criticize the existing studies on the effect of cultural distance on entry modes, questioning whether there is indeed an association and calling for deeper research into the subject, which could help solve the apparent paradox of the contradictory results on this topic. The inconsistent results of the association between cultural distance and entry modes is arguably partly due to the conceptual and methodological inadequacies of the cultural distance construct (Shenkar, 2001).

In an attempt to resolve this paradox, Wang & Schaan (2008) studied potential non-linear relationships between both cultural distance and entry modes on the one hand and cultural distance and performance on the other, both suggesting non-linear U-shaped relationships. The use of JV is said to increase as cultural distance increases when cultural distance is low, yet the use of WOS increases when cultural distance is high (Wang & Schaan, 2008). Because these are rather new and unforeseen findings, the authors recommend future studies should replicate this study using FDIs from other host countries, as this would shed light on whether there is a more general trend for non-linear relationships, or whether this is exclusively an attribute of Japanese FDI, which was in their study.

(22)

The existing literature clearly pays a great deal of attention to the role of cultural distance in entry mode choice, but because there is an apparent lack of real consensus as to whether there is an actual association between the two, with existing studies presenting these rather inconclusive and mixed results, this study will use these inconsistent finding as the main rationale behind the research. The recent findings by Wang & Schaan (2008), suggesting a U-shaped relationship between the two variables but without any other studies supporting this claim, will provide further motivation for the study.

(23)

3. Methodology & Hypothesis Formulation

3.1 Sample

Data on cultural distance and modes of entry were gathered through secondary data found in relevant databases, most of which were found through the Bureau van Dijk (BvD), which provides a range of company information and business intelligence, many of which are co-published with renowned information providers. As such, the Zephyr database for international mergers and acquisitions was used to gather data on the number of acquisitions and joint ventures taken place by Dutch companies to determine how often each of the entry modes are chosen and the countries in which they take place. The Zephyr database is one of the most comprehensive of its kind in deal information and is regularly updated, making it a very reliable source. It not only contains information about mergers and acquisitions, but also on private equity and venture capital deals. It also contains information about targeted and rumored mergers and acquisitions. In order to increase the validity of the study, targeted and rumored deals were left out of the sample, as only concluded deals will be used in the study. A sample of 299 cases was chosen, representing 150 acquisitions and 149 joint ventures taken place in 65 different countries and across 20 different industries. The majority of the cases in the sample were firms who originate from the Netherlands, and a small number of them are firms currently listed in the Netherlands but with their origins in another country. All the cases in the sample were from the years 2005 to 2014. By using this latest and recent data, the reliability of the study should also improve, as it looks at deals from a time period in which the world has already become significantly interconnected and one in which cultural differences have arguably come smaller do to this global and interconnected nature of the world. The cultural distance between certain countries will have likely become less profound in recent years, making it more obvious to only select recent cases of entry mode decisions.

(24)

Data on cultural distance between the Netherlands and foreign countries whose markets are entered by Dutch companies were gathered from the Hofstede Centre, which aims to provide high quality information in terms of the association between culture and management. The Centre includes a number of tools to find information on this link, with the most important of these, the country comparison tool. This tool gives information about each of the different dimensions of business cultures as asserted by Hofstede, and compares each of these dimensions to those of a different country. A composite index was formed based on the deviations along each of the four cultural dimensions used in Kogut and Singh’s study (1988), and complemented by a fifth and sixth dimension now also available at the Hofstede Centre, namely pragmatism and indulgence. For a number of countries used in the study, data was missing on these extra dimensions, with some of them lacking data on either pragmatism or indulgence, and some lacking information on both. In this case the composite index for that country excluded the dimensions for which data was missing. The deviations along each of the six dimensions were corrected for differences in variance and then averaged to obtain a sole index for cultural distance for each country. The final cultural distance scores ranged from 0,35 (very similar culture) between the Netherlands and Sweden to 4,98 (very dissimilar culture) between the Netherlands and Slovakia.

3.2 Hypotheses Formulation

The first hypothesis will explicitly look into the relationship between entry mode choice and cultural distance, and following the study of Wang & Schaan (2008), it will investigate the possibility of a non-linear relationship between the two variables, H1: There is a non-linear

(inverted U-shape) relationship between entry mode choice and cultural distance.

Because the choice of entry mode might be affected by other factors and because the link between entry mode and cultural distance might also be moderated by such factors, the study will also look into some of these factors more concretely. Therefore, the other hypotheses will

(25)

look into the effect of other variables on entry mode choice. The control variables will be elaborated on further on, but will be explicitly mentioned in the next section as these control variables will function as independent variables for the remaining hypotheses. Because the choice of entry mode might differ from industry to industry, the second hypothesis will look into the relationship between entry mode decisions and type of industry. Previous studies suggest there might be a relationship between the type of industry and entry mode choice (Tihanyi et al., 2005; Zhao et al., 2004; Brouthers & Brouthers, 2003). Because the services sector is arguably harder to enter independently, the choice of joint ventures is seen as a more logical step when entering foreign markets, leading to the following hypothesis, H2: There is a

significant relationship between the choice of entry mode and industry.

The next hypothesis will look into the frequently studied relationship between entry mode and performance (Brouthers, 2002; Chen, 2003; Canabal, 2008). Previous studies have focused mostly on the effect of entry mode on performance, but in this case the vice-versa effect, that of performance on entry mode choice, will be investigated. The rationale behind this association is that a wholly owned subsidiary arguably presents more risks than a joint venture when entering a foreign market, with higher performing firms being able to withstand greater risks. This leads to the following hypothesis, H3: There is a significant relationship between the

choice of entry mode and firm performance.

Meanwhile, with larger firms generally believed to favor wholly owned subsidiaries over joint ventures (Agarwal & Ramaswami, 1992; Zaheer & Zaheer, 1997), the fourth hypothesis will look into the relationship between firm size and entry mode choice, H4: There is a significant

relationship between the choice of entry mode and firm size.

Finally, following the idea that country development and therefore country risk will influence FDI decisions (Brouthers & Brouthers, 2001), the last hypothesis will look into this relationship

(26)

and assume that acquisitions are more prevalent in more developed countries, H5: There is a

significant relationship between the choice of entry mode and country development.

3.3 Measures

The study will include a similar design as one of the most academically relevant articles on this topic, namely that of Kogut and Singh (1988). To identify the relationship between the two variables, cultural distance will act as the independent variable, as this is what is thought to predict and affect the dependent variable, which in this case is entry mode choice. Because of the lack of data on greenfield investments, the study will focus solely on acquisitions as wholly owned subsidiary. Therefore the study will simply investigate the choice of acquisitions over joint ventures and how these are related to either high or low cultural distance. Following Wang & Schaan’s (2008) distinction between wholly owned subsidiaries and joint ventures, a joint venture is defined as a subsidiary in which 2 or more partners each hold between 5 and 95 percent of its equity, while subsidiaries are considered wholly owned subsidiaries when one partner holds at least 95 percent of equity.

The independent variable cultural distance, as previously mentioned, will be calculated through the Kogut & Singh formula (1988), and like Wang & Schaan’s (2008) study it will exclude the original fifth dimension (long term orientation) as data on this is mostly missing. It will however include another fifth and sixth dimension (pragmatism and indulgence) on which data is available for most countries.

3.3.1. Control Variables

Because there are a number of factors which could also contribute to one entry mode decision over another, a few control variables will be used in this study, including two different firm-level variables, one host country variable and one industry variable.

(27)

The firm-level control variables are firm size and performance. Firm size is measured by the number of employees the parent firm had in the year the acquisition or joint venture took place, and is an important control variable as it is generally thought that larger firms will be more prone to choosing wholly owned subsidiaries over joint ventures (Agarwal & Ramaswami, 1992; Zaheer & Zaheer, 1997). Information on firm size will be gathered from the Orbis database, accessed through the Bureau van Dijk (DvB) database. The data of number of employees ranged from 2 to over 100,000, making it too large of a range for significant statistical analysis. Therefore, firm size will be divided into five different groups each representing a different category of firm size. These categories will be 1) very small (up to 100 employees), 2) small (between 101 and 1,000 employees), 3) medium (between 1,001 and 10,000 employees), 4) large (between 10,001 and 50,000 employees), and 5) very large (more than 50,000 employees). Meanwhile, overall firm performance is likely also to have an influence on what entry modes are chosen, with better performing firms arguably more inclined to choose higher levels of equity and therefore wholly owned subsidiaries over joint ventures. Because the manner in which performance should be measured is often contested (Chen, 2002; Madhok, 1997; Tihanyi et al., 2005), the study will look at what is generally considered the most important performance indicator, namely profitability of a firm in the year of the acquisition or joint venture, to see how performance might moderate the interaction between entry mode and cultural distance. In this case profitability is measured by the net income of the firm in that particular year. This data is also collected from the Orbis database, and is measured in US dollars, and corrected for the exchange rate of the particular year. Because the performance range is also so high, these figures will also be recoded into separate groups, five of which represent negative numbers and 5 of which will be positive. As such, net incomes below -$10 million will be assigned a -5 and those above $10 million will be assigned a 5. Negative and positive incomes between $1 million and $10 million will become -4 and 4

(28)

respectively. Negative and positive incomes between $100,000 and $1 million will become -3 and 3 respectively. Negative and positive incomes between $10,000 and $100,000 million are assigned -2 and 2 respectively, while negative and positive incomes between 0 and $10,000 will be assigned -1 and 1 respectively.

Moreover, the country-level control variable that will be used is the level of development of a country. Wang & Schaan (2008) note that there is a significant correlation between country development and country risk, which in turn will likely affect firm’s FDI decisions and subsidiary operations (Brouthers & Brouthers, 2001). This variable will be measured using the latest Human Development Index as established by the United Nations, and will split countries into four separate groups (very high development, high development, medium development, and low development)

Finally, the last control variable will be industry, and will be measured by separating all the companies in the sample into twenty distinct industry categories, namely Automotive, Chemicals, Consultancy/Recruitment, Energy, Electronics, Financial Services, Food/Drinks, IT, Machinery, Media, Metals, Pharmaceuticals, Plastics, Real Estate/Property Development, Shipping/Port Operations, (Tele)communications, Textiles, Transport, Other Manufacturers and Other Services, each of which will be divided into two separate groups (manufacturers and service). The type of industry is important as their characteristics could have a moderating effect in the relationship between entry mode choice and cultural distance. Tihanyi et al. (2005) and Brouthers & Brouthers (2003) assert that some industries, particularly the distinction between service and manufacturing, are like to choose different entry modes.

For each of the different hypotheses will deal with distinct control variables. As such, for the second hypothesis dealing with industry, the control variables will be cultural distance, HDI, size, and performance. For the third hypothesis looking at the effect of performance on entry mode, the control variables will be cultural distance, industry type, HDI, and size.

(29)

3.4 Analytical Approach

To identify the relationship between the independent variable, cultural distance, and the dependent variable, entry mode choice, a binary logistic regression analysis will be used. A logistic regression is especially useful for a study of this kind as it has been broadly used in many different fields in the past (Wang & Schaan, 2008). Also, a mix of continuous, discrete, and dichotomous variables can be used in such an analysis, while not making assumptions about distributions of independent variables (Tabachnick & Fidell, 2007). This is particularly helpful when looking at the effect of a number of different types of independent variables, such as industry which will be dichotomous and HDI and size which will be ordinal variables.

The binary regression will be especially helpful due to the fact that cultural distance is measured quantitatively as a real value, while entry mode choice is measured qualitatively and categorically distributed into two categories (acquisition and joint ventures). The choice of entry mode was therefore coded by assigning the value 0 to acquisitions and assigning the value 1 to joint ventures.

Once the values of cultural distance and entry mode choices are identified, the binomial logistic regression should help determine whether there is a causal relationship between the two variables. In order to identify the possibility of a non-linear relationship, the approach taken will be similar to other studies looking into U-shaped relationships, namely by adding the quadratic function of cultural distance to the regression analysis. A significant negative coefficient of the quadratic function would imply an inverted U-shaped relationship. This approach has been widely used in non-linear regression studies (Wang & Schaan, 2008; Baruch et al., 2014). The relationship between both variables will be tested controlling for firm, country and industry variables, as these could also play a role in the choice of entry mode.

(30)

The binary logistic regression will also be used for the remaining hypotheses, as these also deal with independent variables which are thought to predict the dichotomous outcome of either an acquisition or a joint venture. Because industry is a categorical value with 20 possible outcomes, a simple binary logistic regression is not plausible. Instead, sectorial dummies are used to divide the industries in two main groups, mainly services on the one hand, and the extractive and manufacturing industries on the other. These dummies are required to gain a better picture on whether there are entry mode patterns in the different industries, which could potentially moderate the effect of cultural distance. The industries are therefore recoded into binary values of 0 and 1, in which 0 made up the manufacturing and extraction services (including the Automotive, Chemicals, Energy, Electronics, Food/Drinks, Machinery, Metals, Pharmaceuticals, Plastics, Textiles and Other Manufacturers industries), and 1 made up the services industries (Consultancy/Recruitment, Financial Services, IT, Media, Real Estate/Property Development, Shipping/Port Operations, (Tele)communications, Transport and Other Services industries). This is due to the fact that it is not possible to include a categorical variable of more than two outcomes in the binary regression, therefore making it essential to divide the 20 different industries in just 2 separate ones which can be included in the analysis.

(31)

4. Results

4.1 Descriptive statistics for main hypothesis

Before showing the results of the regression analysis, some descriptive statistics will be presented. First of all, a missing values analysis was conducted to identify how many cases in the sample have missing values and will therefore be omitted from the analysis. The following table shows the data on missing values:

Variable Summary

Missing

Valid N Mean Std. Deviation N Percent

Cultural Distance 22 7,4% 277 2,0195 ,95933

Recoded Performance 7 2,3% 292 1,9212 2,53443

Size 5 1,7% 294 3,04 1,417

The table shows that of the sample of 299 cases, 22 cases are missing values for cultural distance, 7 are missing values for performance and 5 are missing values for size. The 22 missing values for the cultural distance variable were missing because data on these countries were not available through the Hofstede Center. As such, the cases which involved either acquisitions or joint ventures in the Ukraine, Kazakhstan, Gabon, Cyprus, Oman, Bahrain, Uzbekistan, Tunisia and Georgia were left out of the analysis as a cultural distance index could not be formed between these countries and the Netherlands.

Meanwhile, information on performance and size, gathered through the Orbis database, were missing for 7 and 5 cases respectively. To make sure that all control variables and independent variables of the remaining hypotheses had data, cases with missing values on these were also left out of the analysis.

(32)

The following tables shows the descriptive statistics on the frequency of each entry mode:

Entry Mode

Frequency Percent Valid Percent

Cumulative Percent

Valid Acquisition 150 50,2 50,2 50,2

Joint Venture 149 49,8 49,8 100,0

Total 299 100,0 100,0

This shows that entry modes of all the cases in the sample are roughly equally divided between acquisitions (150 cases) and joint ventures (149 cases). The next table shows the statistics on the cultural distance index for each of the entry modes (acquisitions and joint ventures):

Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

Cultural Distance 277 ,35 4,98 2,0195 ,95933

Acquisitions 144 ,35 4,98 1,6874 ,81641

Joint Ventures 133 ,35 3,83 2,3791 ,97550

The table shows that of the 277 remaining cases for which data was available, 144 were acquisitions and 133 were joint ventures, which essentially means that more of the missing cases for which cultural dimension scores were missing were joint ventures and less were acquisitions. The table shows that the cultural distance index ranged from 0,35 to 4,98 for all countries used in the study, with a mean cultural distance of 2,01. The cultural distance index range for acquisitions was the same, but the mean was considerably lower at 1,68. Meanwhile,

(33)

the cultural distance range of joint ventures was smaller, with a lower maximum index of 3,83. However, this did not prevent the mean from ending up being higher than the overall mean, at 2,38. These initial figures would suggest that the use of joint ventures seems to be more prevalent at higher indexes of cultural distance.

In order to show how many acquisitions and joint ventures took place in countries with different levels of cultural distance, these values on cultural distance were grouped into separate clusters each representing half a point of cultural distance (0 to 0,5, 0,5 to 1, 1 to 1,5, etc.). This lead to the following frequency table for acquisitions:

Acquisitions

Cumulative Percent Cultural Distance 0-0,50 5 1,7 3,5 3,5 0,50-1,00 11 3,7 7,6 11,1 1,00-1,50 61 20,4 42,4 53,5 1,50-2,00 42 14,0 29,2 82,6 2,00-2,50 4 1,3 2,8 85,4 2,50-3,00 5 1,7 3,5 88,9 3,00-3,50 6 2,0 4,2 93,1 3,50-4,00 8 2,7 5,6 98,6 4,00-4,50 1 ,3 ,7 99,3 4,50-5,00 1 ,3 ,7 100,0 Total 144 96,0 100,0 Missing System 6 4,0 Total 150 100,0

(34)

The table shows that most acquisitions take place in countries with lower cultural distance, with a 119 of them taking place in countries with cultural distance indexes of up to 2,00 (representing almost 40% of the cases), with a cultural distance range of 1 to 1,5 being the cluster where most acquisitions took place (61 cases, i.e. 20,4% of the sample).

This can be compared to the frequency of joint ventures per cultural distance cluster group, through the following table:

Joint Ventures

Cumulative Percent Cultural Distance 0-0,50 3 1,0 2,3 2,3 0,50-1,00 2 ,7 1,5 3,8 1,00-1,50 26 8,7 19,5 23,3 1,50-2,00 24 8,0 18,0 41,4 2,00-2,50 22 7,4 16,5 57,9 2,50-3,00 9 3,0 6,8 64,7 3,00-3,50 17 5,7 12,8 77,4 3,50-4,00 30 10,0 22,6 100,0 Total 133 89,3 100,0 Missing System 16 10,7 Total 149 100,0

This table in general shows that the use of joint ventures seems to be more spread out across different cultural distance index clusters. What also stands out is that the cultural distance only runs till 4,00, suggesting no joint ventures took place in countries where the cultural distance

(35)

between the Netherlands and the country where the FDI was taking place was greater than 4. However, it is clear that less joint ventures took place in culturally similar countries, with only 18,4% of joint ventures taking place in countries with cultural distances of up to 2, as compared to the almost 40% of acquisitions for the same level of cultural distance. Moreover, the cluster that was most prevalent for joint ventures was the one representing cultural distance indexes between 3 and 3,5, which could imply that firms value joint ventures over acquisitions wherever there is high cultural distance.

4.2 Regression Analysis

After combining the data on missing values, the binary logistic regression with the choice of entry mode as a dependent variable, and with cultural distance, industry (sectorial dummies thereof), firm size, firm performance, and HDI of the country whose foreign market is being entered, as independent variables, provided the following case processing summary.

Case Processing Summary

Unweighted Casesa N Percent

Selected Cases Included in

Analysis 272 91,0 Missing Cases 27 9,0 Total 299 100,0 Unselected Cases 0 ,0 Total 299 100,0

The sample therefore had 27 (9,0%) missing cases, which represented the missing values for cultural distance, firm size and firm performance. This left a total of 272 cases (91,0% of the sample) in which data for all variables was available and these remaining cases were therefore

(36)

all included in the analysis. Without the missing cases the sample size remains large enough for the analysis to be reliable.

In order to test the non-linearity of the relationship between cultural distance and entry mode choice, two separate models were constructed, one with simply cultural distance as an independent variable to test the extent of the linear relationship, and another model which included the quadratic form of cultural distance to test non-linearity and how this compares to the linear model.

The following table shows the results of the binary logistic regression for the first model, used to test the effect of cultural distance, while controlling for industry, country HDI, firm size and firm performance on the choice of entry mode.

B S.E. Wald df Sig. Exp(B)

Step 1a Cultural Distance ,745 ,179 17,294 1 ,000 2,106

Recoded Industry -,368 ,270 1,853 1 ,173 ,692

HDI -,509 ,279 3,331 1 ,068 ,601

Size -,325 ,107 9,271 1 ,002 ,722

Recoded Performance ,007 ,056 ,014 1 ,906 1,007

Constant 1,438 1,300 1,224 1 ,268 4,213

The table shows that increasing indexes of cultural distance are associated with a 0,745 more likelihood of choosing joint ventures over acquisitions, with a significance of 0,0. This suggests that cultural distance is indeed a significant predictor of entry mode choice, and its predictive capacity is highly significant. This initial regression analysis suggest a potential linear relationship between the two variables.

(37)

In order to test whether the non-linear model is stronger than the linear model, a second regression analysis was made including the quadratic function of cultural distance. The following table shows the results thereof:

B S.E. Wald df Sig. Exp(B)

Step 1a _{Cultural Distance}

1,894 ,845 5,027 1 ,025 6,645 Recoded Industry _-,347 _,271 _1,634 ₁ _,201 _,707 HDI _-,453 _,279 _2,639 ₁ _,104 _,636 Size _-,327 _,107 _9,288 ₁ _,002 _,721 Recoded Performance _,008 _,056 _,021 ₁ _,884 _1,008 CD squared _-,236 _,167 _1,997 ₁ _,158 _,790 Constant _,067 _1,636 _,002 ₁ _,967 _1,070

This table shows that the beta coefficient for cultural distance squared is negative, compared to the positive beta coefficient of cultural distance. This could potentially suggest that there is indeed an inverted U-shaped relationship between the two variables, but with a significance of just 0,158 it does not prove to be statistically significant. With the quadratic form of cultural distance included in the second model, the linear association between entry mode choice and cultural distance also decreases in significance somewhat, yet is still highly significant at a value of 0,25.

These results essentially mean that the linear model of the association of both variables is a better predictor than the non-linear quadratic model, which means the first hypothesis is not supported. There does not seem to be a significant U-shaped relationship between the choice of entry mode and cultural distance. However, the results do point towards a linear relationship between these variables.

(38)

Related to the second hypothesis, the following table shows the frequencies of acquisitions and joint ventures for each industry:

Entry Mode Total Acquisition Joint Venture Industry: Automotive 5 3 8 Media 6 5 11 Metals 1 2 3 Pharmaceuticals/Medical Equipment 10 1 11 Plastics 2 3 5

Real Estate/Property Development 4 14 18

Shipping/Port Operations 3 14 17 (Tele)communication 6 2 8 Textile 1 3 4 Transport 4 5 9 Other Manufacturers 11 7 18 Chemicals 11 7 18 Other Services 7 11 18 Consultancy/Recruitment 7 6 13 Energy (petroleum/gas) 4 14 18 Electronics 16 10 26 Financial Services 18 16 34 Food/Drinks 22 16 38 IT 8 5 13 Machinery 4 5 9 Total 150 149 299

There does not seem to be anything out of the ordinary in the data in this table, except for some noteworthy differences in the number of each entry mode in the Energy, Real Estate/Property Development and Shipping/Port Operations industries in favor of joint ventures and the Pharmaceuticals industry in favor of acquisitions. The next table shows the frequencies of each entry mode after they were recoded into either manufacturers or service.

(39)

Entry Mode

Total Acquisition Joint Venture

Manufacturers Service

0 87 71 158

1 ₆₃ ₇₈ ₁₄₁

Total 150 149 299

The table shows that out of the sample of 299 cases, 158 were in the manufacturers industry while 141 of the cases were in the service industry. The difference in the number of each entry mode does not seem to differ too much per industry. Firms in the manufacturers industry chose more acquisitions than joint ventures, while firms in the service industry chose more joint ventures, but these differences do not appear to be significant.

Furthermore, the results in the regression analysis do not point towards a significant relationship between entry mode choice and industry, while controlling for the other variables. Therefore, H2 is not supported.

4.4 Descriptive statistics and analysis for 3rd_hypothesis

With firm performance measured in net income and divided into groups ranged from -5 to +5, the following table shows the frequencies of acquisitions and joint ventures for each cluster of firm performance:

(40)

Entry Mode Total Acquisition Joint Venture Performance -5,00 5 1 6 -4,00 3 6 9 -3,00 9 6 15 -2,00 5 8 13 -1,00 5 5 10 ,00 1 0 1 1,00 8 22 30 2,00 28 25 53 3,00 47 36 83 4,00 22 15 37 5,00 16 19 35 Total 149 143 292

The table does not show anything remarkable, but there generally seem to be more entries for better performing firms. Meanwhile, the regression analysis also failed to show a significant relationship between entry mode and performance, which means H3 is not supported.

4.5 Descriptive statistics and analysis for 4th_hypothesis

The following table shows the frequencies of each entry mode in relation to firm size:

Entry Mode

Size 1 29 34 63 2 ₁₀ ₂₈ ₃₈ 3 ₄₂ ₃₅ ₇₇ 4 ₃₅ ₂₀ ₅₅ 5 ₃₃ ₂₈ ₆₁ Total 149 145 294

At first sight the descriptive statistics don’t seem to point towards anything significant, but the regression analysis shows that the relationship between both variables, when controlled for the

(41)

remaining variables, is statistically significant. As such, larger firms are -,325 as likely to choose joint ventures over acquisitions, with a significance of 0,02. This means, as was expected, that larger firms are more likely to favor acquisitions. H4 is therefore supported.

4.6 Descriptive statistics and analysis for 5th_hypothesis

The following table shows the frequencies of acquisitions and entry mode choice for each of the four categories of HDI

Entry Mode

HDI Low Development 3 1 4

Medium Development 3 11 14

High Development 19 63 82

Very High Development 125 74 199

Total 150 149 299

The general pattern of this data suggests that FDI activities by Dutch companies are more likely to take place in higher developed countries, with only 18 of the 299 cases taking place in low or medium developed countries. Two thirds of cases in the sample were FDI by Dutch firms in very highly developed countries, and these investments are considerably more in the form of acquisitions than joint ventures, which have 125 and 74 of the entries respectively. Meanwhile, the regression analysis suggest the relationship between HDI and entry mode is close to being significant at 0,68, suggesting that in higher developed countries firms less likely to opt for joint ventures. This would be in line with was is generally conceived about the relationship between entry mode choice and host country development. However, although being close, the relationship is not statistically significant, therefore H5 is not supported.

(42)

5. Discussion

While previous studies on the relationship between cultural distance and entry mode decisions have generally provided inconclusive and inconsistent results, it is often still believed that cultural distance will indeed have an effect on the choice of entry mode. Wang & Schaan (2008) blamed these inconsistencies on the fixation of studying this relationship in a linear fashion, and their study showed that there was an inverted U-shaped relationship between the two variables for Japanese firms entering foreign market. The regression analysis in this study doesn’t however provide evidence that there is also a U-shaped relationship between these variables for Dutch firms. The binary logistic regression analysis with entry mode as the dependent variable and cultural distance as the independent variable, whilst controlling for industry, firm size, firm performance, and host country development, did show a significant relationship between the two variables. In order for this association to have an inverted U-shaped pattern, the quadratic function of cultural distance in the regression analysis had to have a negative beta coefficient that was statistically significant. Although the coefficient was indeed negative, it failed to be statistically significant. With the linear model being stronger than the non-linear model, the results of the study point towards an actual linear relationship between cultural distance and entry mode choice, with more joint ventures taking place in culturally dissimilar countries.

These findings would go hand in hand with much of the previous studies that have found a positive relationship between these variables, such as those by Kogut and Singh (1988), Agarwal and Ramaswani (1992) and Chen (2003). Contradictive results on this matter and the fact that Wang & Schaan (2008) had found a non-linear relationship between the variables for Japanese firms was sufficient rationale to look into the topic more in depth. The findings of Wang & Schaan could not be replicated with a similar study for Dutch FDI decisions, and it

The role of culture in foreign market entry modes : investigating the relationship between cultural distance and entry mode choices for Dutch MNEs

Amsterdam Business School