Faculty of Management and Organization Msc International Business and Management

(1)

By

DANIELLE STRIJKER

University of Groningen

Faculty of Management and Organization Msc International Business and Management

First supervisor: Dr. A. Saka-Helmhout Second supervisor: Drs. R. W. de Vries

5

^th

of July 2006

Dobbenwal 59

9407 AD Assen

The Netherlands

+31 6 52411252

d.strijker@student.rug.nl

(2)

INDEX

1. Introduction

1.1 Problem Statement 3

1.2 Research Objective 5

1.3 Significance of the Research 7

1.4 Preview 7

2. Literature Review

2.1 Organizational Learning 8 2.2 International Strategy 10 2.3 Cultural Discrepancies 14 2.4 Institutional Discrepancies 16

2.5 MNE Performance 18

3. Methodology

3.1 Sample and Data 21

3.2 Definitions 22

3.3 Measures 23

4. Results

4.1 Sample Characteristics 26 4.2 Multiple Regression Analysis 28

5. Discussion 32

6. Conclusion 35

References 36

(3)

1. INTRODUCTION

With developed world markets becoming increasingly saturated, MNEs (multinational enterprises) have turned to emerging economies such as Eastern Europe, India, Indonesia, Brazil, China, and Mexico as key locations for future growth (London and Hart, 2004). Moreover, to optimally compete in this global world, western firms internationalize into emerging markets to have access to their cheap labour markets (Brenton et al., 1999). Overall, these developments have resulted in increased internationalization by firms. Internationalization itself is seen as a learning process (Tsang, 2002): firms learn from their foreign experience. Operating in multiple locations offers increased opportunities for learning and knowledge acquisition (Hitt et al., 1997; Zaheer, 1995). A factor that reinforces organizational learning is a dynamic environment in which the firm operates. A dynamic environment allows the firm more opportunities for exploration and exploitation based on experiential knowledge (March, 1991). It is well established in the literature that CEE (Central and Eastern Europe) has such a dynamic environment (e.g. Bevan et al., 2004; Brouthers et al., 1998; Meyer, 2001). After the break down of the socialist system in CEE, there was a striking increase in the complexity and dynamism of the environment (Lang and Steger, 2002). Based on theories in the IB (international business) literature it seems that the CEE region should offer quite a number of opportunities affecting organizational learning. However, in the current IB literature the determinants of organizational learning are not yet well established (Sapienza et al., 2005) and few scholars (e.g. Luo et al., 1999) focused on organizational learning by MNEs that expand into emerging markets within the international business field.

These issues are interesting to investigate because organizational learning is seen as an important source of competitive advantage (Levitt and March, 1988). The main focus of this dissertation is on organizational learning; its determinants and its impact on performance. This study makes four contributions. Three constructs are investigated that might impact organizational learning in emerging markets. The first construct is strategy, which is expected to impact organizational learning in emerging markets because it is well established in the literature that strategic flexibility is important to overcome uncertainties in transition economies (Steensma et al., 2005).

Secondly, cultural discrepancies are expected to impact organizational learning.

(4)

between cultural values in Western Europe and CEE. Acquiring knowledge of a host country’s culture enhances the MNE’s organizational learning up to a certain point (Barkema and Vermeulen, 1998). Thirdly, institutional discrepancies between the home and host country institutions are expected to impact organizational learning.

CEE countries show characteristics of the former socialist economies which implies their institutional framework is in transition. Whereas in Western Europe, represented in this study by The Netherlands, market-based institutions are established (Meyer, 2001). Institutional upheaval in emerging markets may inhibit organizational learning (Newman, 2000). In turn, it is expected that organizational learning impacts MNE performance. Luo et al. (1999) have found that there is a linear relation between different types of experience and different measures of performance.

1.2 Problem Statement

Organizational learning has widespread attention in the management literature (Cyert and March, 1963; Fiol and Lyles, 1985; Levitt and March, 1988), though in the IB field there are still gaps to fill in. Classical authors, the ones mentioned above, described organizational learning thoroughly but did not relate it to the IB field. Few scholars (Luo et al., 1999; Barkema and Vermeulen, 1998) focused on organizational learning in an emerging market context. It is well established in the IB literature that emerging markets are a perfect environment with the potential to boost organizational learning because of its dynamism that encourages learning (e.g. March, 1991;

Uhlenbruck et al. 2003). Steensma et al. (2005) argue that transition economies are marked by long periods of rapid and discontinuous change which create incentives for continued learning, thus potentially lead to an enhancement of organizational learning.

These authors focused mainly on mode of entry decisions in the Hungarian context instead of organizational learning covering the whole CEE region. Luo et al. (1999) were among the first IB scholars to address organizational learning in emerging markets, though exclusively was focused on China that has different characteristics than CEE. In line with these authors, this dissertation focuses on organizational learning and at the same time covers the CEE region.

Another aspect that lacks attention in the IB literature is an investigation of the

determinants of organizational learning. Few scholars (Sapienza et al., 2005; Tsang,

2002) indicated the antecedents of organizational learning. Uhlenbruck et al. (2003)

did recognize the importance of firms to pursue strategic flexibility to respond to

(5)

changing environmental conditions in order to learn, but did not provide empirical support and focused on post-socialist firm learning instead of western MNE learning from post-socialist countries. Especially in emerging markets where environments are dynamic, an MNE should be able to respond to complexity and change, for instance by adopting the transnational model suggested by Bartlett and Ghoshal (2002). But as London and Hart (2004) pointed out in their work, it might be necessary for MNEs that expand into the limited developed region CEE to move beyond the transnational model. Bartlett and Ghoshal’s article has widespread attention in the IB theory, though they based their research on nine case studies while this study tries to formulate testable hypotheses.

Furthermore, the impact of cultural discrepancies on organizational learning is emphasized few times in the IB literature. Barkema et al. (1996) indicated that MNEs learn from their previous foreign experience when expanding into culturally diverse markets. This is in line with the view provided by Johanson and Vahlne (1977) that MNEs move to more cultural distant markets when they are to a greater extent expanded. In culturally distant markets, firms often cooperate with local firms in order to speed the organizational learning process (Barkema and Vermeulen, 1997;

Gatignon and Anderson, 1988). Above articles focused on culture in relation to the mode of entry decision while covering cultural distance issues and the general assumption was that organizational learning took place because of riskier modes of entry being chosen. By contrast, the present study is conceptualized by focusing on the impacts of cultural distance on MNE learning, the more culturally diverse markets an MNE entered through international expansion instead of specific modes of entry.

The last aspect that lacks attention in the IB theory is the impact of institutions on organizational learning. It is well established in the IB literature that CEE countries have relatively weak institutions (Bevan et al., 2004; Meyer, 2001; Meyer and Peng, 2005) compared to market-based institutions in the West. Newman (2000) argued that an organization’s capacity for learning and transformation depends on the institutional context but he did not provide empirical support. He states that without legitimating institutional characteristics a business climate results in a lack of norms and values.

Other studies focused on the impact of institutions on mode of entry decision in CEE

(Meyer, 2001; Nakos and Brouthers, 2002; Steensma et al., 2005). This study fills in

the gap in the IB literature by indicating whether limited legitimacy in CEE

(6)

Furthermore a link is provided with firm performance. Many authors indicated that there is a link between internationalization and firm performance (e.g. Contractor et al., 2003; Ruigrok and Wagner, 2003; Thomas, 2006; Zahra et al., 2000).

Organizational learning theory argues that knowledge and experience are important predictors of firm performance; the degree to which firms acquire knowledge through experience determines their success (Fiol and Lyles, 1985). Luo et al. (1999) investigated the learning component in transition economies at subsidiary level performance by looking at experience and by taking China as case study. An important part of learning for MNEs is from spill over across subsidiaries (Bartlett and Ghoshal, 1989), thus this dissertation focuses on the headquarters’ performance instead of subsidiary level performance, and how it is impacted by learning from host markets.

Moreover, it integrates research about MNE’s internationalization and firm performance. In summary, the present study fills in the gaps of past IB research by indicating factors that impact organizational learning and how it impacts on firm performance at MNE level.

1.3 Research Objective

The aim of this study is to fill in gaps from past IB research. More precisely, it is investigated what impact international strategy, cultural discrepancies and institutional discrepancies have on organizational learning by MNEs that expand into CEE and in turn how it impacts MNE performance.

Bartlett and Ghoshal (2002) indicated that firms are adopting more and more complex strategies to deal with the ever changing multidimensional demands in the international business environment. Markets in transition, like CEE, require strategic flexibility by MNEs in order to respond quickly to changing competitive conditions and thereby develop or maintain competitive advantage (Hitt et al., 1997). Fiol and Lyles (1985) pointed it our more precisely by indicating that one of the contextual factors that affect the probability that learning will occur is a strategy that allows flexibility. In line with these authors, this dissertation investigates whether pursuing a certain international strategy impacts organizational learning by firms expanding into CEE. Furthermore, it is well established in the literature that cultural discrepancies cause problems when operating across borders (Hofstede, 1980; Barkema and Vermeulen, 1997), while on the other hand stimulate learning (Barkema et al., 1996).

This dissertation investigates the impact of cultural discrepancies in terms of cultural

(7)

distance on organizational learning. Another stream in the IB literature that is of importance in this study for determining organizational learning concerns institutions.

A transition environment is different than the one in developed markets with respect to institutions (Peng and Heath, 1996), which might complicate organizational learning in emerging markets (Newman, 2000). Institutions in transition result in transaction costs for MNEs because they are exposed to unclear regulatory frameworks, underdeveloped court systems and corruption (Meyer, 2001). In comparison to these studies, this dissertation investigates the impact of weak institutions on organizational learning in CEE. These three explanatory variables were chosen because they play a substantial role in the IB literature.

Furthermore, organizational learning is linked to MNE performance. IB scholars often made a link between internationalization and performance (e.g. Gerpott and Jakopin, 2005; Sullivan, 1994; Zahra et al., 2000). Few scholars (e.g. Ruigrok and Wagner, 2003) addressed the link between organizational learning and firm performance. However, internationalization and organizational learning can be interpreted as a similar concept in this study. But still opposing views exist in the IB literature with respect to the issues mentioned above. This study has linked organizational learning with MNE performance. Above discussion leads to the formulation of the following research question:

What impacts have international strategy, cultural discrepancies and institutional discrepancies on organizational learning and how does learning impact MNE performance?

In order to answer above research question fully, some sub questions were formulated:

o What is the impact of MNEs’ international strategy on organizational learning?

o What is the impact of cultural discrepancies on organizational learning?

o What is the impact of institutions discrepancies on organizational learning?

o How does organizational learning impact MNE performance?

1.4 Significance of the Research

The practical relevance of this study is that insights into organizational learning

in CEE can be valuable to western MNE managers. Organizational learning is

(8)

advantage. Learning enables organizations to build an understanding and interpretation of their environment (Fiol and Lyles, 1985) that in turn might lead to better performance (Levitt and March, 1988). Gieskes and Van der Heijden (2004) add to this insight that organizational learning is a prerequisite for successful innovation. In turn, it may enhance MNE performance (Luo et al., 1999).

The theoretical contribution of this study is that the framework adds value to the existing IB literature by presenting an overview of important elements in internationalization theory with respect to organizational learning by investigating the impact of international strategy, cultural discrepancies and institutional discrepancies on organizational learning. Each concept is drawn on separately in the IB literature.

And, in turn, it sheds light on the impact of organizational learning on MNE performance.

Preview

The present thesis is organised along the following lines. This chapter one reflects the problem statement, research objective and significance of the research.

Literature in chapter two presents the development of our hypotheses. Chapter three covers the methodology. Results covering data analysis are presented in chapter four.

Chapter five covers the discussion as well as the conclusion and limitations of this

study.

(9)

2. LITERATURE REVIEW

This chapter presents the development of the hypotheses. Firstly, there is a discussion about the concept organizational learning, i.e. how this concept is embedded in the management and the IB literature. Then will be continued with discussing the independent variables relating to organizational learning, i.e.

international strategy, cultural discrepancies and institutional discrepancies. Lastly, the impact of organizational learning on MNE performance is discussed.

2.1 Organizational Learning

Cyert and March (1963) were the first who conducted research on the behavioural theory of the firm and created the concept organizational learning. In their view, organizational learning is an adaptive process in which solutions for problems are adapted to experiences that are made within the organization. Levitt and March (1988) describe organizational learning from the perspective of experience. From a knowledge based point of view, Penrose (1959) stated that experience is a prime source of learning. Further studies in the IB field have built on their insights.

Experiential learning theory describes learning as the process of transforming experience into new knowledge (Kolb, 1984). Internationalization is seen as enhancing organizational learning or international experience (Kobrin, 1991). Barkema and Vermeulen (1998) add to these studies that learning is fostered by diversity in experience that can be accomplished through international expansion. Operating in a diverse environment increases the variety of events and ideas to which an MNE is exposed, leading to a more extensive knowledge base and stronger technological capabilities. This is also what Hitt et al. (1997) and Eriksson et al. (2000) put forward by arguing that operating in multiple locations offers increased opportunities for learning and experiential knowledge. Zahra et al. (2000) found that organizational learning enables a firm to develop new knowledge. One means by which firms learn is to move into new foreign markets in which they are exposed to, and then assimilate different types of knowledge. Thus, the higher the diversity of foreign markets the MNE enters the greater the opportunity for organizational learning.

Quite a number of scholars have investigated knowledge transfer issues (e.g.

Inkpen and Tsang, 2005; Kogut and Zander, 2003; Kostova, 1999). They mainly

(10)

socialist firms that learn organizational practices from their western parent are regularly addressed in IB research. This is not of relevance in this study because the focus is on learning by headquarters, e.g. learning peculiarities of an uncertain host environment, achieved by diverse expansions. Learning is reinforced by a dynamic environment that allows the MNE more opportunities for exploration and exploitation based on experiential knowledge (Butler, 1995; March, 1991).

Among the classical scholars in the management field, Fiol and Lyles (1985) were the first that formulated four contextual factors that affected the probability that learning will occur. They provided a relevant overview for organizational learning in general but did not take into account the foreign, uncertain environment. The present study focuses on the building blocks of organizational learning, i.e. its determinants that are explored individually in a more general way in the IB field but not yet integrated into a comprehensive piece with respect to organizational learning.

While organizational learning has received widespread attention in the

management and organization literature, limited research is conducted so far on the

determinants of organizational learning in the international business field. This

literature review discusses the impact of international strategy, cultural discrepancies,

and institutional discrepancies on organizational learning (see figure 1). Moreover, the

present study investigates the impact of organizational learning on MNE performance.

(11)

FIGURE 1

Impact of international strategy, cultural discrepancies and institutional discrepancies on organizational learning by MNEs’ expansion into CEE

2.2 International Strategy

Internationalization strategies refer to the way MNEs fashion relations between headquarters, subsidiaries and the diverse markets and institutional contexts in which they operate (Harzing and Sorge, 2003). Most of the international management literature assumes the existence of different types of MNEs which are, according to Harzing (2000), terms as international, global, multidomestic and transnational organizations. Furthermore, she argues that if meaningful typologies can be discovered they can then be used in a predictive way. The main structuring dimensions of internationalization strategy are, according to Bartlett and Ghoshal (1989), the interdependence of subsidiaries and headquarters, i.e. global integration, and responsiveness to local markets and other situational specificities. Along these dimensions they provided an extensive typology, the ones mentioned above, of MNEs’

international strategy that has widespread attention in the IB literature. MNEs are International Strategy

Global & Transnational

Cultural Discrepancies Hofstede’s dimensions

Institutional Discrepancies

Regulatory

Organizational Learning Expansion

MNE Performance ROA

Internationalization

Emerging Markets

Central and Eastern Europe

(12)

expected to pursue an international strategy that enhances organizational learning, such that competitive advantage is being created. What strategy is expected to be the best option is discussed in the following paragraphs.

Already decades ago Fiol and Lyles (1985) indicated that one of the four

¹

contextual factors that affect the probability that learning will occur is a strategy that allows flexibility. This has a circular relationship with learning in that it creates and reinforces learning (Fiol and Lyles, 1985). These scholars argue that the firm’s strategic posture partially determines its learning capacity. To address the uncertainty of transition economies, MNEs entering these markets look for ways to overcome limitations in the business environment (London and Hart, 2004). London and Hart (2004) found that success depends on developing a deep understanding of the local environment. MNEs operating or having the intention to operate in CEE must recognize that social contracts and social institutions dominate in order to establish legitimacy. Some scholars (e.g. London and Hart, 2004; Wright et al., 2005) already raised the question whether an MNE’s global strategy can be extended and adapted to an emerging economy without facing problems. London and Hart (2004) found that MNEs investing in low-income markets cannot rely on a strategy that is based on overcoming limitations in the business environment. In addition, they put forward that strategies in emerging economies must recognize that social contracts and social institutions dominate, and that social performance matters. Knowing the local environment and recognizing social contracts requires organizational learning to enhance the MNE’s competitiveness. The point of view these scholars present calls for a flexible strategy pursued by a MNE in order to anticipate to environmental conditions.

In order to indicate which international strategy fits best to MNE’s learning by expansion into CEE, a brief overview is given of each typology according to Barlett and Ghoshal (1989; 2002).

1 Other three contextual factors are: corporate culture conducive to learning, an organizational structure that allows both innovativeness and new insights, and the environment (see Lyles, 1985).

(13)

FIGURE 2

Integration and Responsiveness Framework

Strong

Global Integration

Weak

Weak

Strong

Local Responsiveness

Source: Bartlett and Ghoshal (1989)

The international typology is characterized by centralization of sources of core competencies. The role of foreign subsidiaries is to adapt and leverage parent company competencies. Knowledge is developed at the centre and transferred to foreign units. There is weak global integration as well as weak local responsiveness (see figure 2). The global typology is characterized by centralization of assets and capabilities and is globally scaled. These kinds of MNEs treat the world as a whole.

The role of foreign subsidiaries is to implement the parent company’s strategy and they are dependent on the headquarters. Knowledge is developed and retained at the centre. The multidomestic MNE is decentralized and nationally self-sufficient. The subsidiary is not or hardly dependent on headquarters and operates very much as a stand-alone company. The role of foreign subsidiaries is to sense and exploit local opportunities. Knowledge is developed and retained within each unit. The transnational typology has dispersed and interdependent assets and resources. It has differentiated and specialized subsidiary roles. Worldwide learning takes place through by joint development and sharing of knowledge (Bartlett and Ghoshal, 1989).

Headquarters and subsidiaries are all dependent on each other. An MNE should pursue a strategy that responds to various changing environmental conditions (Bartlett and Ghoshal, 2002). The environmental changes in transition economies, such as the ones in CEE, result in new conditions to which the MNE has to react in new and different ways (Uhlenbruck et al., 2003). What kind of international strategy fits such a

Global MNE Transnational MNE

International MNE Multinational MNE

(14)

turbulent environment? The answer to this question is formulated in the next paragraph where differences between typologies are discussed in terms of learning by expansion.

An appropriate international strategy is expected to enhance organizational learning because the firm can better cope with the increased environmental complexities and thus MNEs expand more easily. The first strategy that should enhance learning of an MNE that expands into CEE is a transnational strategy. The transnational strategy is an MNE typology that is an appropriate strategy in a situation that requires MNEs to respond to national differences and complexity. This is definitely the case in CEE. Research on Hungarian firms has shown that firms improved their capacity to learn if organizational flexibility was promoted (Steensma et al., 2005; Lyles and Salk, 1996). An MNE is confronted with a relatively complex environment compared to the one in the west due to the transition phase taking place in CEE. This might imply that products are necessarily adapted or modified to the local market which is accomplished by local responsiveness or that different control mechanisms need to be pursued. Expansion into emerging markets means bridging the formal (e.g. legal contracts) and informal (e.g. social contracts) economy (London and Hart, 2004). High interdependence by which the transnational MNE is characterized means that the headquarter and subsidiary are strongly interdependent. These kinds of firms recognize that environmental demands and opportunities vary widely from country to country. The home country may be the most critical environment for some divisions, but not necessarily for all. The broader range of customer preferences, competitor behaviour, government demands, and technological stimuli, in a market such as CEE, can trigger learning and innovation within the organization (Bartlett and Ghoshal, 1989). The MNE learns in cooperation with a local partner, or individually, the peculiarities of the local environment. Uncertain yet is whether this proposition holds in a CEE context, because London and Hart (2004) suggested MNEs should move beyond the transnational strategy in emerging markets. Learning through expansion is expected to be encouraged by pursuing the transnational strategy because of the flexibility, strong local responsiveness and strong global integration, by which the strategy is characterized.

The second strategy that is proposed to enhance MNE learning by expansion

into CEE is the global typology. One of the reasons why Western MNEs expand into

CEE is because of the relatively cheap labour market in CEE (Brenton et al., 1999).

(15)

The global MNEs’ strategic requirement is to manufacture standardized products in a cost-efficient way (Harzing, 2000), thus there is no need to be locally responsive. The role of offshore subsidiaries is limited to sales and service, although local assembly plants may be dictated by economic or political pressures (Bartlett and Ghoshal, 1989). However this is not the case in every single country because there are examples of R&D centres in Hungary and Czech Republic. An MNE pursuing this strategy is not locally responsive, and is highly integrated. The MNE knows what is going on in unfamiliar environments and hence a global strategy enhances organizational learning because of the fewer uncertainties the MNE faces due to experience gained. The multidomestic strategy is inappropriate because it does not lend itself to a reverse of knowledge due to the low interdependence between the MNE and its subsidiaries. An international strategy was not included in this study because many authors did not include that type in their research and it has not received much empirical support (Harzing, 2000). Above discussion of international strategies and organizational learning leads to the first hypothesis:

H

_1a

: A transnational strategy, pursued by an MNE, is positively associated with learning

H

_1b

: A global strategy, pursued by an MNE, is positively associated with learning

2.3 Cultural Discrepancies

Hofstede’s (1980) definition of culture is one of the most popular in the IB

field. He argues that culture consists of shared mental programs that control

individuals’ response to their environment. Culture refers to the social context within

which humans live (Gabriel, 1992). Moreover, the general assumption is that culture is

shared by members of a particular group. MNEs belong to the group that have a

particular cultural background embedded in their home country’s culture, whereas the

host country into which they established subsidiaries can have a totally dissimilar

cultural background. This suggests that MNEs might encounter cultural discrepancies

when establishing subsidiaries in a foreign environment, in theory this is a so-called

cultural distance. Cultural distance is defined by Luostarinen (1980) as the sum of

factors creating, on the one hand, a need for knowledge, and on the other hand,

(16)

country. MNEs that expand encounter dissimilar cultures, i.e. expand into cultural distant counties, but adjust to foreign cultures (Barkema et al., 1996). Also other points of view are established in cultural theory that are based on the importance of experience an MNE has acquired during expansions and might enhance cultural adaptation due to learning taking place. This is discussed in the next paragraph.

Cross-cultural interaction, i.e. contact with people from different cultures, requires new interpretations and understandings. Valuing different cultures goes beyond knowing the differences there are between cultures. Building relationships within the host culture is generated by coming into contact with others in the host culture (Kayes and Yamazaki, 2005). By coming into contact with the host culture, MNEs gain experience. Kolb (1984) pointed out that experience is transformed into knowledge which is the process of learning. Kogut and Singh (1988) suggested in their article that as firms increasingly gain experience during internationalization, the less problems they face with respect to foreign cultures. Barkema et al. (1996) stated that firms learn from their previous experience when expanding into cultural space. They conclude by discussing that least is benefited from earlier expansions in blocks that are more proximate to the home country. Hence, expected is that an MNE learns from relatively unfamiliar host country cultures. Previous experience with expansions in the same country, even further strengthens this relationship (Barkema et al., 1996). In the following paragraph cultural values and Hofstede’s index are discussed.

Hofstede, together with some other scholars (Kolman et al., 2003), conducted a survey in four

²

CEE countries in order to estimate cultural values in this region. Their findings show that there are substantial differences between the value orientation in The Netherlands and CEE. Bhagat et al. (2002) argue that interaction among societies characterized by dissimilar cultural patterns can be fraught with problems. Firstly, Hofstede’s index will be discussed briefly. Hofstede (1983) is one of the most respected scholars that invented five cultural dimensions that have often been cited in cross-cultural IB research. Namely, power distance, individualism versus collectivism, masculinity versus feminity, uncertainty avoidance and long-term versus short term orientation. These dimensions are quantitative measures of cultural dimensions. The benefit of such dimensions is the ability to construct indexes of ‘cultural distances’

2 These countries are: Poland, Czech Republic, Hungary and Slovakia.

(17)

between countries. As was already argued in the first paragraph, cultural distance impacts organizational learning.

Kolman et al. (2003) found that CEE countries exhibit a relatively large power distance compared with Western European countries

³

. The Netherlands is far more individualistic than CEE countries: The Netherlands score 80 on this dimension whereas CEE countries score on average 60 (Kolman et al., 2003). CEE countries score high on uncertainty avoidance, while The Netherlands scores moderately. Also the scores on the masculinity index vary widely: The Netherlands has a relatively low score of 14 whereas CEE countries score relatively high (varying scores from 81 in Czech Republic to 127 in Slovakia). The long-term dimension score varies from low to high scores in CEE countries, The Netherlands scores 44 on this dimension. These figures indicate that cultural values do differ between The Netherlands and CEE, and that cross-cultural differences even exist across CEE. Thus, cultural discrepancies exist between The Netherlands and CEE countries. Expected is that a higher cultural distance, following the arguments mentioned above, enhances organizational learning.

In conclusion, above discussion leads to the formulation of the third hypothesis.

H

₂

: A relatively high cultural distance between the home and the host country is positively associated with organizational learning

2.4 Institutional Discrepancies

Institutional theory focuses on the role of the political, social, and economic systems surrounding firms in shaping their behaviour (North, 1990). Scott (2001) argues that institutions consist of regulative, normative and cognitive structures and activities that provide stability and meaning to social behaviour. He invented three pillars that are interpreted as the building blocks of the institutional context in a country. The regulative pillar concerns rule setting in a country such as law and treaties, in other words it is the area of governance. The normative pillar emphasizes values and norms that guide social behaviour. The cognitive pillar encompasses the frames through which meaning is made and the rules that constitute the nature of reality. This study adopts the regulative pillar. The regulative pillar emphasizes conformity to rules. A regulative view would ascertain whether the organization is

(18)

legally established and whether it is acting in accord with relevant laws and regulation.

However, during institutional change the institutional stability is threatened (Scott, 2001).

In CEE, costs associated with institutional frameworks vary less with the characteristics of the firms, but more across countries and industries (Meyer and Peng, 2005). Even among emerging economies, CEE is special owing to the transition from central planning to market competition. Furthermore, Meyer and Peng (2005) argue that the introduction of new and coherent formal systems of governance is a time consuming process in CEE. Before the arrival of these systems, it is likely that informal mechanisms of control, monitoring and contract enforcement are being used during the transition phase. Institutional upheaval in CEE means that a rapid and pervasive change in the norms and values result in a fundamental change in a society’s political system, its legal and regulatory frameworks, its economic system, and its financial infrastructure. However, when institution-level change is too extreme, when the underlying values, ideologies, and norms in society are in question, and when the economic and political systems are in disarray, past experience has little value as a guide for future action (Newman, 2000). Due to a lack of legitimacy in CEE countries, MNEs need to establish legitimacy, i.e. creating stability themselves.

Institutions reduce transaction costs by reducing uncertainty and establishing a stable structure, i.e. establishing legitimacy, to facilitate interactions. In post-socialist CEE these rules were not in place when the socialist system disintegrated (Meyer, 2001). Gradually, CEE institutions move towards more market based institutions as Meyer (2001) has shown in his study. The next paragraph continues with the impact of institutions on organizational learning.

Newman (2000) argued that an organization’s capacity for learning and

transformation depend on the institutional context. He states that institutional upheaval

inhibits organizational learning. Institutional upheaval creates confusion and

uncertainty and produces a business climate that lacks norms, values, templates, and

models about appropriate strategies, structures, and systems. Without these

legitimating institutional characteristics, organizational change is likely to be episodic,

ineffectual, temporary, and misguided (Newman, 2000). These institutional pressures

in CEE motivate MNEs to pursue activities that will increase their legitimacy such that

they are in agreement with the norms in the local business environment and thus face

stability. One way MNEs can achieve this is through participation in

(19)

interorganizational relationships (Barringer and Harrison, 2000). With deep, simultaneous and repeated institutional upheavals, organizational learning and the search for the appropriate organizational ‘path’ by firms may become impossible (Wright et al., 2005). It is difficult to learn from experience during periods of significant institutional change, because past experience is no longer an appropriate guide for future action (Weick, 1979). This implies that a lack of legitimacy in CEE leads to limited exploration of knowledge that results in learning. In the absence of new formed institutions in CEE, a firm’s capacity to learn is diminished, because there are no frameworks or schemas to help organize, interpret, and react to new information. Learning is inhibited because organizations do not know what actions lead to what outcomes (Newman, 2000). In the next paragraph the study of Meyer is addressed that focused on institutions in the CEE region.

Meyer (2001) indicated in his study that institutions in CEE are not yet the same as the market-based institutions, i.e. the ones that exist in The Netherlands. He argues that progress in reform brings the institutional framework of CEE economies closer to that of Western European economies. This convergence of institutions reduces psychic distance. Diversity is still important, because if there is no diversity at all there is nothing new MNEs gain experience from. However, as was indicated in some studies (e.g. Newman, 2000; Wright et al., 2005) too much institutional change may limit exploration because of the lack of legitimacy in CEE. In the case of CEE, institutional discrepancies between home and host country institutions hamper learning by MNEs due to a lack of legitimacy. Capturing above arguments leads to the following formulation of the fourth hypothesis:

H

₃

: A large distance between home and host country institutions is negatively associated with learning by expansion.

2.5 MNE Performance

In the past decades an impressive amount of research was conducted about firm performance by IB scholars. Because this study conceptualizes learning as exploration through expansion, studies that link internationalization to performance are reviewed.

Some scholars argue that internationalization influences negatively firm performance,

others take an opposite view. Organizational learning theory argues that knowledge

(20)

Hitt et al. (1997) indicated that managerial experience with complex environments provides organizations with indispensable knowledge for maintaining superior performance at high degrees of internationalization. These scholars found a positive linear relationship between internationalization and performance. They state that international expansion can have a direct effect on performance, by allowing a firm to take advantage of substantial opportunities in international markets. Furthermore, they argue that international expansion connects the firm with important constituencies in diverse markets such that MNEs can profit from networks or obtain key resources. In the beginning of the internationalization process firms initially face a disadvantage relative to local firms because of their lack of knowledge and experience of the local market (Johanson and Vahlne, 1977). Over time, firms learn to operate in foreign environments. Experience in foreign markets allows firms to overcome the costs related to the liability of foreignness and to reap the benefits of internationalization (Thomas, 2006). Contractor et al. (2003) contributed to the firm performance issue in IB literature by making a three-stage theory of international expansion. They found that at an early stage of internationalization there may be a diminution in performance because of the initial learning costs, cultural and foreign market inexperience, an insufficient scale of global operations, and, in general, the liability of foreignness.

Greater multinationality, thus once a firm is to a larger extent expanded, is associated in the study of Contractor et al. (2003) as augmenting performance. However, as they state, there are always some firms that may ‘over internationalize’ due to expansion into too many nations. This occurs in the latest stage of the model. However, as Warner and Ruigrok (2003) indicated, this is not confirmed that the costs of expansion inevitably outweigh the potential benefits beyond a certain point. This study adapts to the point of view by Wagner and Ruigrok (2003). In the following paragraph the possible direction of the relation between learning by expansion and performance is discussed.

There are different types of relations visible: positive linear, negative linear, a U-shaped relationship, and an inverted U-shaped relationship

⁴

. In this study a positive linear relationship is expected, briefly its definition is discussed. The definition of a positive linear relationship is quite straightforward: a firm’s performance improves with greater international expansion (Contractor et al., 2003). Further geographical

4 See Contractor et al. (2003) for details about these relationships

(21)

scale makes possible efficiencies that improve performance indicators. The more a firm has internationally expanded, the lesser become the incremental costs of expansion, hence the more benefits a firm faces. The following is proposed based on above discussion. The more an MNE has learned by expansion, the more its performance is positively influenced. The hypothesis is as follows.

H

4

: Organizational learning is positively associated with MNE performance.

(22)

3. METHODOLOGY

This study is a deductive, quantitative research. It presents a theoretical model for organizational learning explained by the MNE’s international strategy, cultural discrepancies, and institutional discrepancies. According to Yin (2002), four measures are taken to establish the quality of any empirical research. This study has dealt with construct validity by justifying the choices of measurements of the explanatory and dependent variables. Internal validity is dealt with by including control variables that might cause organizational learning. Various industries are included in the sample to enhance the external validity of the study. However, The Netherlands represents the west which might imply limited external validity. Procedures that are followed are documented to increase the reliability of this study.

3.1 Sample and Data

A database is constructed based on a sample from Dutch MNEs across various industries. The Netherlands is chosen because of its relatively small size and thus its need to internationalize. Firms are selected from the AMADEUS database compiled by Bureau van Dijk. All firms are selected based on the criterion of having a foreign subsidiary owned by at least 51 percent in the CEE region. The practical reason to take this percentage is that if the subsidiary is owned by less than 51 percent, it might imply that the largest owner is non-Dutch. Countries in the sample include Poland, Czech Republic, Slovakia, Hungary, Romania and Turkey. Of these countries a cultural index was available, thus other CEE countries were excluded. The result is a dataset of 658 firms. However, most firms are not ‘true’ Dutch firms. They are established due to tax reasons or else in the Netherlands. Other problems that are encountered are that in some cases the firm did not at all expand into CEE or no annual reports were available. In the end it has resulted in a sample of 72 Dutch MNEs. In a study by Hadley and Wilson (2003), regression analysis was conducted based on 58 observations. Thus based on their study, 72 MNEs in this study’s sample seems a good number. The MNEs in the sample are active in a wide variety of industries: in manufacturing, construction, retail and wholesale trade, transportation and warehousing, finance, and other industries.

The variables were put in an Excel database before exporting it into a SPSS

database. The variables cultural discrepancies and institutional discrepancies,

(23)

corruption indexes of the year 2004, needed to calculate in excel with the formula provided in section 3.3. Also return on assets was calculated based on financial data available concerning the year 2004 (formula see section 3.3). In SPSS the variable strategy was transformed into a dummy variable (1 = transnational strategy and 2 = global strategy). Columns of organizational learning, i.e. geographical diversity, were inserted, as well as firm size, economic development and industry type. Industry type, according to the NAICS 2002 industry classification provided by AMADEUS, was transformed into a dummy variable (1=manufacturing, 2=construction, 3=retail and wholesale trade, 4=transportation, 5=information, 6=finance and insurance, 7=professional, scientific and technical services, 8=administrative and support, 9=real estate, rental and leasing). In the next section the definitions of the variables are given.

3.2 Definitions

Independent variables. The first explanatory variable in this study is

international strategy by Bartlett and Ghoshal’s typology. A proxy was chosen to measure an MNE’s international strategy. A global strategy is characterized by strong global integration and weak local responsiveness. The transnational strategy has strong global integration as well as strong local responsiveness. Headquarters and subsidiaries are all dependent on each other. The second explanatory variable is cultural discrepancies. Hofstede’s cultural dimensions were used to measure the cultural distance. Cultural discrepancies were thus defined in this study as the cultural distance that exists between The Netherlands and CEE countries. The third explanatory variable is institutional discrepancies. Host country institutions might differ from the home country institutions due to the transition taking place from post- socialist towards market based institutions (Meyer, 2001), which leads to a lack of legitimacy (Newman, 2000).

Dependent variables. The dependent as well as the explanatory variable of

firm performance is organizational learning. Organizational learning is something that

is intangible and thus quite complicated to define. It is a concept with multiple

interpretations in the literature. Some scholars interpret organizational learning as

knowledge transfer, e.g. transfer of organizational practices, taking place between

MNEs and their subsidiaries. Others see the concept as knowledge exploration by an

MNE. Another interpretation, and the one adopted here, is that organizational learning

(24)

definition is about acquiring knowledge about uncertain environments which is accomplished by foreign expansion. The other dependent variable in this study, firm performance, is a concept that has been extensively used in IB research. Experience in complex and multiple environments provide firms with knowledge for maintaining superior performance at high degrees of international diversity (Hitt et al., 1997).

Control variables. Some control variables that are expected to influence the

relationships among the concepts were included in this study. The first control variable was MNE size. Tsang (2002) found in his study that firm size and learning intent were significantly correlated. Larger firms have greater resources and thus can take higher risks (Gatignon and Anderson, 1988). The number of employees is used to measure MNE size. This was also used as a proxy in the study of Dhanaraj and Beamish (2003). Another control variable is the MNE’s age. MNEs that have been in operation a longer period would be expected to have more international experience. An MNE’s age is measured by the years between data collection and its first international expansion. The second control variable, for both organizational learning and MNE performance, is the country’s level of economic development. Some CEE countries might be further in the transition process towards a market economy than other countries due to their level of economic development (see table A3). This can imply that they are more similar to Western firms with as a result that the environment is less dynamic than countries more dissimilar than The Netherlands. If the latter is true, it is easier for the Dutch MNEs to apply past knowledge and experience and thus the learning effect is smaller (Levitt and March, 1988). This proxy is measured by GDP.

This control variable is used for MNE performance. The World Bank (2006) states that GDP is a main indicator of economic development. Though this control variable is related to the proxy of institutional discrepancies, corruption indexes, it does not supersede the institutional determinant of expansion. The last control variable is industry type because the need for experiential knowledge may depend upon the type of industry in which MNE operates (Eriksson et al., 1997). Also this variable is used to control for MNE performance. This variable was coded into a dummy variable (see section 3.1).

3.3 Measures

International strategy is something that is hard to measure by secondary data.

Studies (e.g. Harzing, 2000; Harzing and Sorge, 2003 London and Hart, 2004) in the

(25)

IB field relied on primary data acquisition. However, this study measures international strategy based on empirical definitions existing in the literature. This study focuses on an empirical definition of the local responsiveness measure. Harzing and Sorge (2003) measured local responsiveness by percentage of local R&D and local production. In other words if there were R&D activities in the subsidiaries and whether the subsidiary had its own products modified for local markets, then the MNE pursued a transnational strategy. Thus, local R&D and/or local production was the proxy for local responsiveness. The transnational MNE has strong local responsiveness. By contrast, the global firm has a weak local responsiveness, thus R&D is kept at the centre and products are not modified for the local market.

Cultural distance between The Netherlands and CEE was measured by Kogut and Singh’s (1988) cultural distance formula. This index was used in many other studies (e.g. Agarwal & Ramaswami, 1992; Barkema et al., 1996; Barkema and Vermeulen, 1998) in the IB field. It is an arithmetic average of the deviations of each country from the index of The Netherlands along Hofstede’s cultural dimensions.

Cultural distance between the country of origin, The Netherlands, and CEE countries was calculated in the following way (see formula 1):

Formula 1: Cultural distance CD

_j

= ∑

²_i=j

((I

_ij

- I

_iu

)

²

/V

_i

)/5

Where I

_ij

stands for the index for the ith cultural dimension and jth country, V

_i

is the variance of the index of the ith dimension, u indicates The Netherlands, and CD

_j

is the cultural difference of the jth country from The Netherlands (Kogut and Singh, 1988).

Institutions were measured by the CPI (corruption perceptions index). The same measurement was used by Steensma et al. (2005) to measure institutions in Hungary. However, the present study is not looking at absolute values but at the difference between The Netherlands and the CEE countries. The Netherlands has a relative high score of 8.7 (ranges between 10 – highly clean, and 0 – highly corrupt), whereas Turkey has the lowest score of 3.2 (see table appendix A). Institutions in The Netherlands are relatively strong compared to CEE and taking the difference between the home and host countries will measure the institutional distance. The formula (see formula 2) is similar to the one of cultural distance, namely:

Formula 2: Institutional distance = ID

j

= ∑

²i=j

((I

ij

- I

iu

)

²

/V

i

)/2

Organizational learning was measured by expansion diversity in this study. In

the IB field expansion is predominantly measured by mode of entry decisions (e.g.

(26)

internationalization (Sullivan, 1994). The DOI refers to the extent to which a firm depends on foreign markets for customers, factors of production, and the capacity to create value, and to the geographical dispersion of such internationalization (Sanders et al., 1998). In particular, the geographical dispersion is relevant in this study because it is in line with the conceptualisation of learning in this study. It reflects what many scholars (e.g. Barkema and Vermeulen, 1998; Eriksson et al., 2000; Hitt et al., 1997) already stated that experience can be acquired by international diversity through MNEs’ expansion process. The more geographically dispersed an MNE, the more an MNE learns through expansion. DOI was measured by what Sullivan (1994) indicated as a valid measure, namely the number of foreign countries in which the MNEs are active.

MNE performance is measured by return on assets (ROA). Also the measure net assets turnover was used as an indicator, however, for the analysis it did not make any difference in outcomes, thus ROA is the indicator of MNE performance in this study. This has been a valid measure method in previous studies (Hitt et al., 1997;

Sullivan, 1994). The formula (Eitemann et al., 2004) looks as follows:

ROA = net profit / total assets

(27)

4. RESULTS

This chapter presents the data analysis. In the first paragraph the sample characteristics are discussed. This is necessary to check whether the regression analysis is a valid one.

4.1 Sample Characteristics

Various steps needed to be taken before the data could be analysed in order to be sure of the reliability of the model. To start with, the sample characteristics were analysed to see whether the variables were normally distributed, which is a necessary condition for conducting linear regression analysis. Table 1 (see below) shows an overview of the sample characteristics.

TABLE 1 Descriptive Statistics

Minimum Maximum Mean

Standard

Deviation Skewness Kurtosis

Statistic Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error

Strategy 1,00 2,00 1,6667 ,47471 -,722 ,283 -1,521 ,559

Cultural distance 21,00 41,00 28,4408 4,57304 ,497 ,285 ,098 ,563

Institutional

distance 3,90 6,50 4,8327 ,47642 ,705 ,283 1,483 ,559

Learning by

expansion 3,00 130,00 21,2917 21,51150 2,704 ,283 9,501 ,559

MNE performance -21,14 21,19 5,4100 5,85883 -,904 ,283 5,620 ,559

Years of foreign

experience 7,00 152,00 57,9706 38,52625 ,790 ,403 -,341 ,788

Economic

development 41,10 242,30 143,3456 41,67104 ,154 ,283 ,354 ,559

MNE size 30,00 236860,00 21799,29 52762,69 3,306 ,283 10,147 ,559

Industry type 1 9 3,1667 2,31346 0,879 0,283 -,333 ,599

The skewness and kurtosis statistics show the normality of the observed variables. The table shows that two variables are problematic for conducting regression analysis.

These variables are learning by expansion, and MNE size. The other variables are

normally distributed. The high skewness and kurtosis statistics indicate that further

analysis would be insignificant, because the values should be between -1 and 1. These

high values are caused by extreme values of variables. In order to conduct a reliable

analysis both variables that are not normally distributed were transformed into a

(28)

logarithm in order to approach normality. Several transformation formulas can be used in transformation depending on the best transformation (Rummel, 1970). The formula used in this study is:

X

j

= log X

i

Where X

j

is the transformed variable into a logarithm. And where X

i

is the variable that needs to approach normality. Table 2 shows the results of the logarithm of learning by expansion and MNE size. It appears that the logarithm is normally distributed (see appendix B for histogram). Hence, at this point the sample is reliable and so far there are no restrictions to use them for regression analysis.

TABLE 2

Descriptive Statistics after Transformation

Minimum Maximum Mean

Std.

Deviation Skewness Kurtosis

Statistic Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error

Strategy 1,00 2,00 1,6667 ,47471 -,722 ,283 -1,521 ,559

Cultural distance 21,00 41,00 28,441 4,57304 ,497 ,285 ,098 ,563 Institutional

distance 3,90 6,50 4,8327 ,47642 ,705 ,283 1,483 ,559

Log learning by

expansion ,48 2,11 1,1702 ,36781 ,182 ,283 -,237 ,559

MNE

performance -21,14 21,19 5,4100 5,85883 -,904 ,283 5,620 ,559

Years of foreign

experience 7,00 152,00 57,970 38,52625 ,790 ,403 -,341 ,788

Log MNE size 1,48 5,37 3,4656 ,92896 ,191 ,283 -,656 ,559

Economic

development 41,10 242,30 143,35 41,67104 ,154 ,283 ,354 ,559

Industry type 1 9 3,1667 2,31346 ,879 ,283 -,333 ,599

The next step, to see whether the model fits the conditions to conduct regression

analysis, is to check them for multicollinearity. Multicollinearity is the undesirable

situation when one independent variable is a linear function of other independent

variables. The results of the multicollinearity can be seen in table 3 (see below).

(29)

TABLE 3

Multicollinearity Matrix

A multicollinearity higher than 0.7 or lower than -0.7 is unacceptable (Anderson et al., 2005). The highest multicollinearity relationship is between cultural distance and economic development, namely -0.729. The other statistically significant relationships are acceptable because they are all lower than 0.7 or higher than -0.7. According to Farrar and Glauber (1967), the problem of high multicollinearity between cultural distance and economic development can be tackled in two ways. One is to drop economic development from the model. The other is to enhance the information on economic development, because multicollinearity is a signal that the information might be too few. This study chose the first option: eliminate economic development from the model, because more information on GNP, which was the proxy of economic development, is simply not available. And with eliminating economic development, there still remain two other control variables. The next paragraph shows the multiple regression outcomes.

4.2 Multiple Regression Analysis

Finally, at this point a multiple regression analysis is conducted. The results are shown in table 4.

Strategy

Cultural distance

Instit.

distance

Log Learning by expansion

Economic

developm. Log MNE size

Yrs of

for.exp Industry

Strategy Pearson Correlation 1 -,159 ,171 -,140 ,180 -,108 -,017 ,103

Sig. (2-tailed) (,184) (,151) (,240) (,131) (,369) (,884) (,391)

Cultural Pearson Correlation -,159 1 -,068 ,351 -,729 ,238 ,002 -,084

distance Sig. (2-tailed) (,184) (,575) (,003) (,000) (,046) (,988) (,487)

Institutional Pearson Correlation ,171 -,068 1 -,061 ,099 -,108 -,075 -,031

distance Sig. (2-tailed) (,151) (,575) (,611) (,408) (,368) (,532) (,795)

Log learning Pearson Correlation -,140 ,351 -,061 1 -,202 ,555 -,166 -,002

by expansion Sig. (2-tailed) (,240) (,003) (,611) (,088) (,000) (,164) (,988)

Economic Pearson Correlation ,180 -,729 ,099 -,202 1 -,128 -,029 ,018

development Sig. (2-tailed) (,131) (,000) (,408) (,088) (,284) (,809) (,883)

Log MNE Pearson Correlation -,108 ,238 -,108 ,555 -,128 1 -,115 -,110

size Sig. (2-tailed) (,369) (,046) (,368) (,000) (,284) (,338) (,356)

Years of Pearson Correlation -,017 ,002 -,075 -,166 -,029 -,115 1 -,007

foreign exp. Sig. (2-tailed) (,884) (,988) (,532) (,164) (,809) (,338) (,967)

Industry Pearson Correlation ,103 -,084 -,031 -,002 ,018 -,110 -,007 1

Sig. (2-tailed) (,391) (,487) (,795) (,988) (,883) (,356) (,967)

(30)

TABLE 4

Multiple Regression Outcomes

Unstandardized Coefficients

Standardized

Coefficients t Sig.

Model B Std. Error Beta

Constant ,174 ,526 ,331 ,742

Strategy -,049 ,078 -,063 -,662 ,536

Cultural distance ,019 ,008 ,229 2,244 ,028**

Institutional

distance -,017 ,085 -,020 -,198 ,844

Log MNE size ,196 ,041 ,493 4,812 ,000*

Years of foreign

experience -,002 ,001 -,112 -1,145 ,265

Industry ,012 ,016 ,074 ,743 ,460

** Significant at P< 0,05

* Significant at P< 0.01

In the appendix, the model summary and the ANOVA table results are attached (see appendix C). Briefly, these results reveal that there is quite a strong relationship between the independent variables and learning by expansion, namely an R-value of 0.621. Furthermore, the ANOVA table has shown that from a statistical perspective the model is accepted, it is significant at a p-level smaller than 0.01. This implies that the variation explained by the model is not due to chance. However, the next paragraph discusses the most important findings that are detracted from table 4. In the discussion chapter, the results are more thoroughly discussed and linked to the literature.

Strategy has a negative regression coefficient of -0.063. The negative sign means that the transnational or global strategy negatively impact learning, thus they are not appropriate for MNEs expansion into CEE. Furthermore the relationship is highly insignificant at a value of 0.536. Thus pursuing a multidomestic or transnational strategy does not contribute to more learning by expansion. In summary, hypothesis one is not supported.

Cultural distance, or cultural discrepancies, has a positive regression

coefficient of 0.229. The relationship between cultural discrepancies and learning by

expansion is significant (a value of 0.028) at a p-level of 0.05. This implies that

(31)

hypothesis two is supported, which suggests that MNEs should expand into cultural distant countries to enhance their learning and thus strengthen their competitiveness.

The variable institutional discrepancies have a negative regression coefficient of -0.020. This means that MNEs should avoid entering host countries with a different institutional framework than they the one they are familiar with because it is negatively associated with learning. Institutional discrepancies and learning by expansion are highly insignificant (0.844). In other words, the third hypothesis is not supported.

Log MNE size has a positive regression coefficient of 0.493 and is highly significant (0.000) at a p-value of 0.01. This means that this control variable is a very good predictor of organizational learning by expansion.

MNEs’ age has a negative correlation coefficient of -0.112. It is an insignificant relation because the value of 0.265 is relatively far above the p-level.

This means that this control variable does not influence MNE’s learning by expansion.

Industry type has a positive value of 0.074. It is an insignificant relation of 0.460. This means the type of industry does not impact organizational learning as control variable.

So far only the first part of the model was discussed. In the next section the impact of learning on the MNE’s performance is discussed.

The results of the linear regression analysis are shown in table 5 (see below).

TABLE 5

Linear Regression Results

* Significant at P < 0.01

Unstandardized Coefficients

Standardized

Coefficients t Sig.

Model B Std. Error Beta

Constant 4,815 3,692 1,304 ,197

Log experience

by learning ,998 1,866 ,063 ,467 ,594

Economic

development ,013 ,016 ,095 ,808 ,422

Industry -,783 ,291 -,309 -2,696 ,009*

(32)

The results in table 5 show that the regression coefficient between learning by

expansion and ROA is positive but has a value of only 0.063. Surprisingly it is a

highly insignificant relation (a value of 0.594). This fact is strengthened by the

ANOVA results (see appendix C) having an R value of 0.324. This implies that there

are many other factors that impact MNE performance. Moreover, from a statistical

point of view, the model is not accepted (p = 0.055). One of which is control variable

used in this study, industry type, that has quite a surprising value of -0.309, thus it

negatively contributes to organizational learning. It is significant at a p-level of 0.01 (a

value of 0.009). The other control variable, economic development, has a positive

value of 0.095 but is insignificant (0.422). However, the fourth hypothesis is not

supported.

(33)

5. DISCUSSION

In this chapter the most important findings are summarized and discussed. In addition, a link is provided to the literature. This study has tried to complement existing IB literature with respect to organizational learning by focusing on MNEs that expand into emerging markets, namely the CEE region. Previous studies did not provide an overview of determinants of organizational learning and did not focus on learning that took place in emerging markets like CEE countries that are recognized by IB scholars as a more complicated environmental context (Meyer, 2001; Peng and Heath, 1996). Several analyses were conducted before arriving at the suggestion that cultural distance and MNE size have a significant affect on organizational learning.

Also industry type impacts significantly MNE performance. In turn, it was found that organizational learning does not have a significant impact on MNE performance. In the next paragraph, these as well as other results are discussed and linked to what other studies found in the international business field.