Corps de l’article

Considerable research has been conducted on the determinants of foreign entry mode by multinational enterprises (MNEs) (Brouthers and Hennart, 2007). However, findings have often been inconsistent and even contradictory (Slangen and Hennart, 2007; Tihanyi, Griffith and Russell, 2005). One of the main tensions within this field is whether national cultural factors (Erramilli, 1996; Kogut and Singh, 1988) or institutional dimensions of countries (Brouthers, 2002; Delios and Beamish, 1999) matter most to foreign location strategy. Culture and country risk are the two most common sources of external uncertainty facing the MNE (Brouthers and Hennart, 2007; Hitt, Franklin and Zhu, 2006). On the one hand, cultural factors are deemed to be important because the beliefs and attitudes of managers are shaped by national cultures, and this influences patterns of behaviour and decision making (Erramilli, 1996). On the other hand, scholars of institutional theory (IT) have argued that entrants adjust their governance mode choice to the specific transaction costs in different locations (Meyer, 2001).

Scholars have also recently highlighted the possibility of differences in international strategy according to the level of economic development of the host country. Developing economies have held a growing importance in the world economy (Hoskisson, Eden, Lau and Wright, 2000) and MNE investment in developing countries has increased. Despite this, the growing interdependence between countries has not resulted in universal cultures (Leung, Bhagat, Buchan, Erez and Gibson, 2005). Furthermore, developing countries represent volatile institutional environments (Meyer, 2001) as the formal ‘rules of the game’ are modified to support a new market economy (Dikova and van Witteloostuijn, 2007). Given the heterogeneity within these countries, applying theoretical approaches born out of scholarly study of international strategy in developed countries becomes challenging (Wright et al., 2005).

To contribute to our understanding of international strategy across locations in both developed and developing countries, we study both the effects of culture and institutions on new foreign expansion investments made by MNEs. Given recent trends in globalization, technological development, the growth of the knowledge-based economy, and the emergence of new countries as sources of supply and demand (Archibugi and Iammarino, 2002; Audretsch and Thurik, 2001), we believe it is necessary to reassess well-established arguments using insights from recent foreign expansion decisions in both developed and developing countries (Wright et al., 2005). Thus we focus on MNEs expanding into developed and developing countries in the same period of time.

Our sample consists of 544 entries made by all active MNEs listed on the AEX index (the Amsterdam Exchange Index is the main stock exchange in The Netherlands) during the five years between 2004 and 2008 inclusive. Given that foreign expansion may also be determined by a range of country and firm characteristics (Erramilli, 1996; Hollensen, 2001; Pan and Tse, 2000), we control for these effects. We test the impacts of cultural distance and host country institutional quality on the choice for majority/full control vs. minority control. These different levels of commitment require different levels of direct foreign investment (Davis, Desai and Francis, 2000) and were accurately reported in our data sources. For cultural distance we use Kogut and Singh’s (1988) measure based on Hofstede’s dimensions of national culture. For institutional quality, we use the mean of Kaufmann’s six governance indicators (Kaufmann, Kraay and Mastruzzi, 2006).

We find that, when expanding into developed countries, the external uncertainty that arises through cultural distance dominates MNE considerations for expansion strategy; institutional quality is not relevant. For these types of locations, increasing cultural distance leads to a greater propensity for avoiding controlling stakes. When expanding into less developed emerging and transition economies on the other hand, cultural distance becomes irrelevant and the effects of institutional quality become stronger. For these types of locations, increasing institutional quality leads to a greater propensity for pursuing controlling stakes.

The contribution of this study is twofold. Firstly, we highlight the differential effects of cultural distance and institutional quality on foreign expansion choices in modern-day MNEs originating in a highly-developed, open economy. We advance our understanding of how these variables determine international strategy across locations in developed vs. developing countries. Cultural distance and institutional quality are sources of uncertainty that impact decisions relating to the control of new operations in foreign locations. Our contribution is to show how the degree of development and industrialization of the location influences how the MNE views the relevance of these sources of uncertainty. Secondly, we contribute to the debate on whether internal or external sources of uncertainty matter most when considering cultural distance during international strategy formulation (Shenkar, 2001). Our study suggests that resolving this debate can be achieved by considering the role that the economic development of a location plays in amplifying the relevance of cultural distance.

Theoretical background

Entry modes can be classified in two broad categories: full-control and shared-control. Full-control modes imply a controlling stake and include Greenfield investments or cross border acquisition, whereas joint venture entails sharing of control and ownership (Herrmann and Datta, 2002). A joint venture is (a) a vehicle by which to share complementary but distinct knowledge which otherwise could not be shared or (b) a way to influence the competitive positioning of the firm by coordinating a limited set of activities (Kogut, 1988). Different entry modes involve different resource deployment patterns and levels of risk and control. Full-control entry modes are typically more sensitive to environmental uncertainties and political instabilities (Henisz, 2000). Resource commitment is usually greater in full-control entry modes and often involves deployment of assets which cannot easily be redeployed without a substantial cost (Hill, Hwang and Kim, 1990). Choosing the right entry mode has a profound effect on firm foreign performance and survival (Davis, Desai and Francis, 2000).

The study of foreign entry mode decisions has largely been based on the theory of transaction-cost economics (TCE) (Coase, 1937; Gatignon and Anderson, 1988; Hennart, 1991). According to TCE, MNEs arise in order to internalize intermediate product markets to reduce transaction costs that would be present in the market (Buckley and Casson, 1976; Williamson, 1975, 1981). This becomes particularly important in entry mode choice, as each entry mode requires a certain amount of resource commitment, i.e., the amount of assets, tangible and intangible, that cannot be redeployed to alternative use without cost (Hill, Hwang and Kim, 1990). Transaction costs include those of finding and negotiating with an appropriate partner, and the costs of monitoring the performance of the partner firm (e.g., Brouthers, 2002; Gatignon and Anderson, 1988). TCE predicts that the choice between full and partial ownership in foreign expansion of the MNE depends on the cost and benefits of full ownership (acquisition and Greenfield) opposed to shared ownership (joint venture) (Hennart, 1991).

International strategy scholars recently have emphasized an ‘extended’ transaction cost model for explaining foreign entry mode investment (Brouthers, 2002). In this extended view, two sets of arguments are put forward, one relating to the managerial cost and uncertainty evaluation of national cultural factors (Brouthers and Brouthers, 2000) and one concerning the contextual conditions that might present exchange risks, i.e., institutional factors (Delios and Beamish, 1999). This extended perspective argues that TCE should be combined with insights from institutional theory (North, 1990) and include the central role of the host country institutional framework on enabling or constraining MNE behaviour (Dikova and van Witteloostuijn, 2007).

Cultural distance and foreign expansion

The influence of national culture on international strategy has been stressed by many authors in the field of international business (Erramilli, 1996; Hitt et al., 2006; Hollensen, 2001; Kogut and Singh, 1988; Tihanyi et al., 2005). Culture is seen as a ‘collective programming of the mind’, a way of distinguishing the members of one group or category of people from ‘others’ (Hofstede, 1997). The term ‘culture’ can thus apply to organizations, occupations and professions, age groups etc., as much as it can to nations. Erramilli (1996) argued that beliefs and attitudes of managers, and consequently the patterns of decision making, are shaped by national cultures. Brouthers (2002) argued that national cultural context variables influence managerial cost and uncertainty evaluations in foreign markets. Thus firms are better able to manage employees in culturally similar countries (Hitt et al., 2006).

The term ‘cultural distance’ has been used to indicate the difference in culture between two countries (Brouthers, 2002; Kogut and Singh, 1988). Differences in national cultures can be portrayed as differing organizational and administrative practices, as well as employee expectations (Kogut and Singh, 1988; Leung et al., 2005). Transaction cost logic can readily be applied to the construct of cultural distance since firms incur a nontrivial expense when they set up and operate their businesses on foreign soil. Various studies have supported the argument that national culture influences the choices made by individuals within organizations. Schneider and De Meyer (1991), for example, showed how managers from different national cultures respond to strategic issues in different ways. Newman and Nollen (1996) showed how the fit between management practices and national culture can impact the performance of overseas work units of the MNE. Salk and Brannen (2000) provided evidence of the effect of national culture on the patterns of relationships within multinational joint-venture teams. Chen, Chen, and Meindl (1998) theorized that culture influences the mechanism by which behavioral cooperation within the organization takes place, for instance, that more individualistic cultures require goal interdependence between actors, rather than goal sharing. Similarly, Kirkman and Shapiro (1997) argued that cultural values influence resistance to self-managed (empowered) work-teams in overseas subsidiaries, power distance being among the variables that may lead to resistance to self-management, and individualism hypothesized to influence resistance to working in teams.

A large body of literature examines the impact of cultural distance on entry modes, diversification and performance (for a recent meta-analysis see Tihanyi et al., 2005). However, results have not been consistent or conclusive (Shenkar, 2001). In their meta-analysis, for instance, Tihanyi et al. (2005) found a negative association between cultural distance and equity involvement in host country operations, although this finding was not statistically significant. One group of scholars argues that external uncertainty arising through cultural distance means that internationalizing firms are more likely to utilize local partners through minority control arrangements. According to this line of literature, local partners have a better understanding of the local culture and are better able to manage external issues with the local labour force and relationships with suppliers, buyers and governments (Kogut and Singh, 1988). Acquisitions are less attractive because when an existing foreign management team must be integrated into the acquiring firm’s organization, post-acquisition costs caused by problems in the ‘organizational fit’ in terms of administrative practices, cultural practices and personal characteristics are substantial (Hitt et al., 2006; Kogut and Singh 1988). Researchers in this tradition have therefore emphasized joint ventures as a way of reducing transaction costs during internationalization (Balakrishnan and Koza, 1993). Another group of scholars, on the other hand, provide argument and evidence that internal uncertainty arises through cultural distance and this encourages the internationalizing firm to take a controlling stake in new foreign operations in order to understand and deal with culturally distant agents (Anand and Delios, 1997; Pan, 1996; Shenkar, 2001). In this view the MNE acts to internalize transactions in order to overcome issues relating to the divergent characteristics and practices of individuals and organizations in different foreign locations. Hence we posit two competing hypotheses:

H1a: The greater the cultural distance between the home-country and the country of entry, the more likely an internationalizing firm will choose a minority control entry mode over a majority or full-control stake.

H1alt: The greater the cultural distance between the home-country and the country of entry, the more likely an internationalizing firm will choose a majority or full-control stake.

Institutional quality and foreign expansion

Institutions provide the structure in which business transactions occur (North, 1990). This structure commonly takes the form of regulations, but also includes normative (social obligations) and cognitive (collective constructions of social reality) aspects (Hoffman and Ventresca, 1999; Scott, 1995). In this sense, institutions set “the rules of the game” through coercive, mimetic and normative mechanisms (DiMaggio and Powell, 1983; Rodriguez, Uhlenbruck and Eden, 2005) and even determine the level of stability and order in societies (North, 1990). They provide the conditions that protect or undermine property rights and increase or decrease resource commitment risk (Brouthers, 2002).

To function and to prosper, societies need the guidelines set forth by institutions so that human behavior and economic outcomes will be more effective and efficient. In most industrialized societies, the costliness of transacting is minimized by the functioning, quality institutions that have been put in place to constrain actors (individuals and organizations) from acting opportunistically (Williamson, 1985). In societies that do not have strong legal systems, that is, property rights and contract enforcement are not easily upheld in a court of law, then transacting becomes both more costly and riskier. Not only may property rights and contracts be devalued in such societies, but also a weak economic appropriability regime means that knowledge (e.g., patents, trademarks, brand names) cannot be lawfully protected and suppliers, buyers, rivals, and joint venture partners can appropriate the economic value (Teece, 1986; Williamson, 1996). Institutions influence the comparative efficiency of governance structures, and, consistent with the objective to minimize transaction costs, the firm’s ownership structure choice will vary with the need to safeguard its assets and minimize risks in differing institutional environments across its international locations. In environments marked by weak institutions, firms often choose low-control entry modes, such as equity joint ventures, thus shifting the risk to others, or even avoiding ownership in order to retain flexibility against environmental changes (Gatignon and Anderson, 1988).

Researchers also have shown that host country institutions influence the choice of entry mode (Henisz and Macher, 2004). Firms’ international strategic moves are affected by institutional constraints (e.g., regulatory limitations, immature legal safeguards) and institutional infrastructure (Hitt et al., 2006). Recent research has shown how institutional voids, i.e., a lack of strong institutions – or, put another way, poorly functioning institutions - can predict the prevalence of family owned firms in a country (Chakrabarty, 2009).

In some countries, the institutional context may create a situation where the entry mode choice predicted by transaction cost theory may not be the best choice. Firms may “face pressures to adopt designs that are within the subset of socio-politically legitimated designs” instead of choosing for a mode of entry which is based on the transaction cost design (Brouthers, 2002). Scholars have therefore argued that entrants adjust their mode choice to the specific transaction costs in different institutional frameworks (Meyer, 2001). In particular, political and economic risks in the host country institutional environment are likely to have the highest impact on the entry mode choice (Delios and Beamish, 1999). In addition, MNEs can benefit from spillover effects. They can ‘free ride’ on the reputation of the partner and learn ways of dealing with the host country government and other aspects of the institutional infrastructure (Yiu and Makino, 2002). Hence:

H2: The lower the institutional quality in the host country, the more likely that an internationalizing firm will choose minority control entry mode over a majority or full-control stake.

Differences between developed and developing economies

These three hypotheses represent a mixture of TCE and IT perspectives. Choosing for minority control partner involvement under conditions of high cultural difference is a way of avoiding unnecessary transaction costs. Choosing for minority control partner involvement under conditions of low institutional quality reflects reaction towards volatility. As argued by recent scholars, little is known about how these alternative sets of argument apply in the case of MNEs from developed countries expanding into both developed and less developed economies over the same period of time (Hoskisson et al., 2000; Wright et al., 2005).

We argue that there will be differences in terms of how cultural distance and institutional quality influence entry mode depending on the overall level of development or industrialization of the location in which the MNE is investing. In our analysis we use OECD membership to capture this. The current thirty-four member-countries of the OECD represent the world’s most economically powerful; their membership attests to the benefits of belonging to a free market system that fosters economic and industrial development (Convention on the Organisation for Economic Co-operation and Development, 2009). Levels of industrialization are higher in OECD countries than in non-OECD countries, in part, because of government subsidies for industrial activity and technological development (Ford and Suyker, 1990). Along with the ties it maintains with a handful of the world’s strongest emerging economies (e.g., Brazil, Russia, India, and China), the OECD group of countries accounts for almost 80% of global world trade and investment (OECD, 2011).

Governance institutions, such as political, legal, and social systems, tend to be more stable and reliable in developed countries than in developing countries. Quality institutions inspire confidence in the ability of a government to monitor and enforce codes of conduct and laws, thus allowing individuals to assume that the business and legal systems are trustworthy and able to protect commercial transactions (Globerman and Shapiro, 2003). Consequently, we argue that when a MNE from a developed country invests in an OECD country, cultural distance will be more of a concern for investment risk considerations and entry strategy than will the host country’s institutional context since the latter is assumed to be at a similar, high-quality par of the home country. When entering a less developed economy, however, institutional quality becomes more paramount and will override unease that may arise due to cultural distances. Hence:

H3a: Cultural distance will increase the likelihood of majority control entry modes when a developed country firm expands into another developed country.

H3b: Institutional quality will increase the likelihood of majority control entry modes when a developed country firm expands into a developing country.

Methodology

Sample

We collected foreign expansion data for all companies listed on the Dutch AEX index for the 5 year period 2004-2008 inclusive. There were 24 MNEs active on this listing over this period and these were all included in the data collection and subsequent analysis. At the time of data collection these were the most recent full years reported. The main benefit of using multiple years was to eliminate any spurious effects attributed to single year cross-sectional surveys. We chose the AEX-listed companies for three reasons: (1) they are generally large MNEs with widespread international operations, including developed and developing countries; (2) AEX-listed firms publish audited annual reports from which foreign expansion moves are clearly reported; (3) we were able to control for a single home-country. The annual reports were obtained from the corporate websites for these MNEs. Prominent researchers in the field of MNE international strategy and foreign expansion have previously used company reports as the principal data source (e.g., Gatignon and Anderson, 1988). Evidence of new entries into foreign markets was found by a qualitative search on the English version of the company report using keywords such as: acquisition, Greenfield, joint venture, start-up, investment, and partnership. We were able to identify acquisitions which arose out of earlier joint ventures within the period of analysis and these were excluded from the final data in order to avoid confusing initial expansion with subsequent expansion.

We took steps to enhance the robustness and reliability of our data. We performed an inter-rater reliability test on roughly one-third of all cases during a pilot phase. Three coders each coded between 30 and 40 foreign entry moves. Each coder’s results were then cross-checked by one of the other coders. In most cases we reached immediate agreement. We found that, in a very small number of cases, the interpretation of mode was potentially ambiguous. In these cases we used press releases from the companies concerned in order to make a clear identification of the entry mode. We continued to use press releases to check ambiguous cases for the full sample. A small number of observations remained unclear and were not included in the final sample. We also found around fifty references to new acquisitions and joint ventures in The Netherlands (i.e., the home country). These were not included in the final sample. The initial sample consisted of 572 observations (reduced to 544 observations in subsequent regression models) in 66 countries, with the yearly activity increasing over the five years as follows: 76 cases in 2004 (13.3%), 86 in 2005 (15.0%), 135 in 2006 (23.6%), 135 in 2007 (23.6%), and 140 (24.5%) in 2008.

Dependent variable

In line with our theory, we captured two entry modes: full / majority control and minority control. Where the equity stake was explicitly reported, we treated majority control ownership as 51% or greater stake[1]. Any unclear or ambiguous expansion modes which could not be clarified by reference to press releases were left out. Minority equity investments in start-ups involving more than one partner were coded as minority control partner involvement. We coded the entries as 1 for majority control and 0 for minority control.

Independent variables

Firstly, in order to test H1, we calculated the cultural distance (CD) of host-countries from the Netherlands using Kogut and Singh’s (1988) formula. We therefore rely on Hofstede’s (1980) dimensions of power distance (PDI), individualism (IDV), masculinity (MAS), and uncertainty avoidance (UAI) as the underlying components of national culture. A number of criticisms of Hofstede’s work have been published in recent years. Critical concerns include the number and choice of questionnaire items used by Hofstede in his original study, of the Western bias in Hofstede’s questionnaire, and of misrepresentation through devaluing and overvaluing (Ailon, 2008). A central criticism is that the ‘collective programming of the mind’ argument espoused by Hofstede gives little credit to the individual as an active social agent (Ailon, 2008). However, Hofstede’s measures continue to be used in the field of international business study. Many large scale studies have supported, rather than contradicted the original conclusions although there have been calls for extending beyond Hofstede and exploring new dimensions of culture (Kirkman et al., 2006). The Schwartz-based measures provide an alternative framework (Schwartz, 1994); these measures are based on more recent data and obtained through a purposefully chosen research design. Drogendijk and Slangen (2006) analyzed the different ways of measuring national culture, i.e., Hofstede (1980, 1984) or Schwartz (1994), and the influence on entry mode decisions of MNEs. These authors concluded that the explanatory power of the Hofstede and Schwartz-based measures are comparable, providing support for our use of Hofstede dimensions.

Secondly, in order to test H2, we calculated a measure of host country institutional quality (IQ) using the arithmetic mean of Kaufmann’s six institutional variables for the host country (Kaufmann et al., 2006). We used the data from 2006 as this was the most recent available and closest to the timing of the foreign expansion moves in our data. As expected, we observed extremely high correlations between the six variables (0.77<r<0.98, p<0.001) justifying the use of the mean as a measure of overall institutional quality. Cronbach’s alpha for these six items was 0.72. A high score on this variable indicates a stable institutional environment in the host country, while a low score indicates a more unstable and volatile institutional environment. Kaufman’s six dimensions are: voice and accountability (the extent to which a country’s citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and free media); political stability and absence of violence (the likelihood that the government will be destabilized or overthrown by unconstitutional or violent means, including political violence and terrorism); government effectiveness (quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies); regulatory quality (ability of the government to formulate and implement sound policies and regulations that permits and promotes private sector development); rule of law (the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, the police, and the courts, as well as the likelihood of crime and violence); control of corruption (extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as “capture” of the state by elites and private interests).

Finally, in order to test H3, we split the data into two sub-samples: entries into OECD countries and entries into non-OECD countries. We ran two models, one for the OECD sub-sample, and one for the non-OECD sub-sample. We used a logistic regression to test the model in each case.

Control variables

Foreign expansion may also be determined by a range of host-country and firm characteristics (Brouthers et al., 2002; Erramilli, 1996; Hollensen, 2001; Pan and Tse, 2000). At host-country level there are several variables that could play a role. A controlling stake can be expected when the market size of the host-country is high, and therefore strategically important to the MNE (Agarwal, 1994). We controlled for market size using host country population (in millions) from the year 2006 (obtained from Thomson DataStream, log transformed). Economic development and human development (life expectancy, literacy, educational attainment) may also encourage wholly owned entry modes in order to protect accumulated knowledge and technology. As expected, we observed a high correlation between economic development (GDP per capita in the year 2006), human development (using the Human Development Index - data for this index was obtained from the United Nations Development Program (hdr.undp.org)) and institutional quality (0.829<r<0.866; p<0.001). We included economic development (GDP per capita) as control variable. This has been used in prior studies as an indicator of market potential and where market potential is greater, the MNE is more likely to make larger resources commitments (Agarwal, 1994). The final country-level control we included in our analysis refers to the legislation that host country governments in developing countries have in place that encourage or discourage foreign direct investment. When these laws require shared ownership with a local partner, they express a host government’s desire to facilitate catch-up, to encourage knowledge and skill spill-over from investing foreign firms from developed countries, and often to protect the interests of the state (La Porta et al., 1999). We operationalized the degree to which host countries legislate vis-à-vis foreign direct investment by using the mean of item 6.12 from the Global Competitiveness Report published by the World Economic Forum across the years of the study as a proxy for government intent. This item (effect of business rules on FDI) is captured on the World Economic Forum Executive Opinion Survey for each year and stated as “To what extent do rules governing foreign direct investment (FDI) encourage or discourage it?” This item ranged from 1 – 7, a higher value indicating rules encouraging FDI.

At the firm level there are several characteristics that could influence the mode of entry, including the size of the parent firm and the degree of internationalization of the parent firm (Erramilli, 1996; Hollensen, 2001). Firm size may influence the MNE’s international strategy, for example, in terms of the resources available to establish a wholly-owned subsidiary. Degree of internationalization potentially may impact foreign entry strategy through firm experience effects, particularly relating to subjectivity, monitoring and assessing inputs in new markets. We controlled for both of these factors, with size taken as number of employees, log transformed, and degree of internationalization as the ratio of foreign sales to total sales.

Results

Appendix 1 shows the foreign expansion activity to be across a wide range of locations. The three most active locations were in each of the triad regions: USA (107 cases or 18.7%), China (48 cases or 8.4%) and Germany (37 cases or 6.5%). Table 1 shows the descriptive information for the dataset. The 66 countries captured in our data represent a sufficiently wide range of cultures and institutional environments to test the hypotheses. In terms of institutional quality, we have a sufficient range, although we notice a bias towards more developed countries. 21 of the countries were classified as OECD countries and 45 as non-OECD. For the entries into OECD countries, 79% of the sub-sample was majority control and 21.0% was minority control. For the entries into non-OECD countries, 59.4% of the sub-sample was majority control and 40.6% was minority control. In terms of country size, we also note a sufficient distribution, with a mean size of 247 million people.

Table 1

Descriptive information (n=551-572)

Descriptive information (n=551-572)

-> Voir la liste des tableaux

Tables 2 and 3 show the inter-correlations between the variables of interest for OECD and non-OECD countries respectively. We should exercise caution when interpreting the correlations with entry mode as this variable is a dichotomous variable. However, the correlations do show that as the sample MNEs expanded abroad into larger markets, they also encountered weaker institutional environments and greater cultural distances. Institutional quality is negatively associated with population in the non-OECD partition (r=-0.40, p<0.001), reflecting the presence of emerging economies such as Brazil, China, India, Russia and Turkey within the top locations for foreign expansion of the MNEs in that part of the sample.

Table 2

Correlation matrix – OECD countries (n=310-328)

Correlation matrix – OECD countries (n=310-328)

***p<0.001, **p<0.01, *p<0.05, +p<0.1

-> Voir la liste des tableaux

Table 3

Correlation matrix – non-OECD countries (n=234-244)

Correlation matrix – non-OECD countries (n=234-244)

***p<0.001, **p<0.01, *p<0.05, +p<0.1

-> Voir la liste des tableaux

Logistic regression results are shown in Table 4. In the full sample (Model 1) we find no support for H1, H1alt, or H2. However, we do find some support for H1 over H1alt, but only for OECD countries (Model 2). In terms of institutional quality (H2) we find no support in the full sample or in OECD countries. However, the hypothesis for institutional quality (H2) is supported for non-OECD countries (Model 3). In this model we find no support for either H1 or H1alt. Cultural distance does not matter to Dutch MNEs expanding into these locations whatsoever. Finally, we find broad support for H3a and H3b. We see marked differences between OECD and non-OECD countries. In Model 2 (OECD countries), increasing cultural distance from the Netherlands leads to an increased likelihood of pursuing minority controlled local partner involvement over a controlling stake. In sum, cultural distance is only statistically significant (and increasing the likelihood of minority control entry modes) when the sampled Dutch firms entered other OECD countries and institutional quality of the host country is statistically significant (and increasing the likelihood of majority control entry modes) when entering non-OECD countries.

Table 4

Logistic regression results (beta coefficients and robust standard errors)

Logistic regression results (beta coefficients and robust standard errors)

***p<0.001, **p<0.01, *p<0.05, +p<0.1 Dependent variable: 1=Majority control, 0=Minority control

-> Voir la liste des tableaux

Discussion

Cultural differences and institutional quality are central concerns facing internationalizing firms as they formulate and implement international strategy (Hitt et al., 2006; Kogut and Singh, 1988; Globerman and Shapiro, 2003; Meyer, 2001). In the current study we analyze the strategy of MNEs in terms of how they deal with these concerns when entering locations with contrasting levels of economic development. We believe this approach is important because – and as we note from our data - large MNEs from developed, open economies do enter economically dissimilar locations in the same period of time. Consistent with prior literature we find cultural similarity between home- and host-country has an impact on levels of commitment and control. However, we find this only applies when the developed country MNE enters another developed country. Of course, being similar in terms of economic and industrial development does not necessarily mean two locations are culturally similar. It is in this economically similar scenario where the argument applies that firms from culturally similar countries will perceive less cost and lower investment risk. To this extent, our findings reinforce TCE arguments relating to external uncertainty and entry mode choice rather than arguments relating to internal uncertainty (Shenkar, 2001). Nevertheless, our results suggest that there are limits to this logic. We find cultural distance to be irrelevant when the Dutch MNEs enter non-OECD countries. This suggests the degree of development and industrialization of the location influences how the MNE views the relevance of sources of uncertainty.

Institutional quality, on the other hand, is only relevant when FDI from a developed country MNE takes place in non-OECD locations. Thus IT is more meaningful in these contexts to explain and predict the institutional implications of foreign locational strategies relative to majority and minority control entry modes. However, there are limits and nuances here as well. Following IT, we argued that institutional quality is more likely to encourage majority control over minority control partnering. We find evidence to support this for non-OECD countries. Table 5 shows the logistic regression result for non-OECD countries with the individual Kaufman indicators broken out. This indicates political stability and government effectiveness to be particularly salient. The greater the political stability, the more likely the MNE will seek control of operations in the foreign location. On the other hand, the greater the government effectiveness, the more likely the MNE will seek partnering. Here it appears the Dutch MNEs are particularly sensitive and averse to making large commitments in a politically turbulent environment. A relatively effective government (and civil service) in a non-OECD country, on the other hand, means that any government policies towards partnering are seen by the MNE as reliable. Overall, this suggests that the developed country MNE needs to make trade-offs amongst a number of dimensions of institutional quality when choosing a governance structure under conditions of institutional risk.

Table 5

Logistic regression results (beta coefficients and robust standard errors)

Logistic regression results (beta coefficients and robust standard errors)

***p<0.001, **p<0.01, *p<0.05, +p<0.1

Dependent variable: 1=Majority control, 0=Minority control

-> Voir la liste des tableaux

There are a number of implications that can be drawn from the current study. Overall, because we find that the state of development of the host country matters in terms of which theory (TCE or IT) predominates, we suggest theory on location strategy of internationalizing MNEs needs to integrate and synthesize arguments in a more subtle way than has been attempted to date. TCE appears to be more relevant when MNEs from OECD countries enter similarly developed economies, and IT more relevant when the same group considers locational strategies in developing economies. The implication of this is that, despite both of these variables being part of the extended transaction cost model of entry modes (Brouthers, 2002), managers and researchers should not treat cultural and institutional variables with the same degree of importance when analyzing foreign expansion of MNEs from developed countries.

Future theory on location strategy of internationalizing MNEs needs to incorporate the notion that MNEs are capable of entering multiple countries in the same period of time, and that, as globalization and development of the world’s economies progress, these host countries are likely to be in various stages of development. As shown in our data, a single MNE will find many locations attractive in a given time period, and will allocate resources to those locations accordingly in this time period. A principal concern for theory development is how to structure and interpret arguments to predict how a MNE will enter multiple locations that do not share the same profile in terms of state of industrialization and development. We suggest that when the strategic leaders of a MNE consider the pattern of potential locations and make decisions regarding the associated types of commitments that would need to be made in order to make those commitments viable and successful, different types of logic apply. Thus new theory on MNE location strategy should be based on the tenets of both transaction costs incurred due to distance and contextual problems due to institutional weaknesses.

The present research is based on choices of entry modes by large Dutch AEX-listed companies. This sample frame limits the generalizability of the findings to non-Dutch firms (especially those from less developed economies) and to small and medium sized enterprises (SMEs). Because we gathered our data from secondary sources, we cannot be certain that we included all market entries of the firms. A potential risk is that we only gathered information on the biggest entries and small, exploratory market entries may have been excluded. We did not control for financial size and sub-national location of the market entries. We did not consider financial and non-financial outcomes of the entry mode decisions, nor did we control for the motives for entry as these were not always clear in our data sources. Furthermore, during our project we observed that large MNEs can choose multiple entry modes in the same host country at more or less the same point in time, and that ownership levels of one subsidiary can change over time. The performance of earlier joint venture or shared-ownership operations may determine increasing equity commitment; this was not picked up in our study. This suggests that country level factors are not the decisive factor in all cases. Other factors may dominate, such as firm-, or even product-market level factors that encourage the MNE to choose multiple entry modes in the same host country.

Future research could address these concerns by broadening the sample frame to MNEs originating in countries with lower institutional quality, and to internationalizing SMEs. Future research also could include various types of distance and extend the dataset built in the current study to capture the performance of MNE foreign expansion in developed and emerging countries in subsequent years. Previous research showed that choice of entry mode does matter to performance (Brouthers, 2002); firms that base their entry mode decision on the extended transaction cost model tend to perform better than others. In addition, building a dataset of entry modes over a long time-frame, and capturing the motives for foreign expansion moves, would enable researchers to assess the influence of firm-specific recent experience effects in developing countries. Furthermore, including different systems for measuring national culture into the dataset could enable a comprehensive assessment of comparability and consistency of cultural measurement systems as applied to firms’ foreign expansion. This could be a possible future application for the present dataset. While we present an interesting outcome in terms of direct comparison of the potential cultural and institutional explanations for the ownership and control preferences of large MNEs from a developed, open economy, these additional avenues would provide further insights into multinational firms’ foreign expansion and location strategies in an ever-globalizing world, one where developing and emerging economies play an increasingly important role.