Entry and establishment mode choice
Firm-, industry- and country-specific predictors
Author:
Caroline Cerwick
Supervisor:
Dr. W. Biemans
Co-assessor:
Dr. Mr. H. A. Ritsema
August 2010
University of Groningen
Faculty of Economics and Business
Abstract
This thesis analyses the internationalization patterns of German based manufacturing MNEs. The aim is to integrate both streams of FDI mode research, the entry mode choice (i.e. the choice between a wholly owned subsidiary and a joint venture) and the establishment mode choice (i.e. the choice between an acquisition and a greenfield investment), in one single analysis. Focusing on shared elements of both streams of research, a conceptual synthesis is created. Moreover, I extend traditional transaction cost logic by additionally consulting the institutional theory as well as the organizational learning perspective. Those three theoretical perspectives vary regarding the key determinants of FDI mode choices. The transaction cost theory focuses mainly on economic considerations, the institutional theory emphasizes the importance of the influence of both institutional forces and decision makers´ cognitive constraints and the organizational learning perspective highlights the impact of the level of MNEs´ experience on their internationalization patterns. Using a data sample comprising 378 foreign investments made by 62 German based firms within the time period from 1978 to 2010, the influence of seven firm-, industry- and country-specific predictors on the dual entry-establishment mode choice is analysed. I demonstrate that the level of economic development in the target country influences the dual entry-establishment mode choice. In addition, international experience, R&D intensity, cultural distance, market growth and country risk are found to influence the establishment mode choice.
Preface
This thesis forms the conclusion of my Master in International Business & Management at the
University of Groningen. During the past five months, I have studied the entry and establishment
mode choice applied in the case of German manufacturing FDI.
To start with, I would like to say a few words in order to reflect the process of writing. I have postponed the writing itself several times to fully enjoy my last months in Groningen and simply due to a lack of self-discipline. However, once I have started writing my thesis my interest in this field of research constantly grew. I was ambitious to find out more about my topic and I started to enjoy gathering information and data.
The realization of this thesis would not have been possible without the support of a number of people. First of all, I would like to thank my thesis supervisor Dr. Wim Biemans for his guidance and constructive comments. Moreover, I would like to thank Charles Belfor, my friend and flatmate, for his technical support and critical remarks. Finally, my special thanks go to my father Hansjörg and my mother Erika, who have supported me in many ways. All the advice and support truly helped to conquer the challenges, which I faced in the process of writing, and enabled me to complete this assignment.
I hope you will enjoy reading my master thesis.
Caroline Cerwick
TABLE OF CONTENTS
LIST OF TABLES AND FIGURES …... 06
ABBREVIATIONS …... 07
1. INTRODUCTION …... 08
1.1 FDI mode choices – a critical issue …... 08
1.2 Why focus on German FDI? …... 08
1.3 Problem statement …... 09
1.4 Contribution …... 10
1.5 Thesis structure …... 11
2. FDI MODE CHOICES …... 12
2.1 Entry mode choice …...…...12
2.1.1 Joint venture …... 13
2.1.2 Wholly owned subsidiary …... 13
2.1.3 The choice between joint venture and wholly owned subsidiary …... 13
2.2 Establishment mode choice …... 15
2.2.1 Greenfield …... 15
2.2.2 Acquisition …... 15
2.2.3 The choice between greenfield investment and acquisition …... 16
2.3 Combining insight from both streams of research …... 16
3. THEORETICAL BACKGROUND …... 19
3.1 Transaction cost theory …... 19
3.2 Institutional theory …... 20
3.3 Organizational learning perspective …... 21
3.4 Why use an integrating framework? …... 22
4. HYPOTHESES DEVELOPMENT …... 23
4.1 Firm-specific predictors …... 23
4.1.1 International experience …... 23
4.1.2 Firm size …... 25
4.2 Industry-specific predictors …... 26
4.2.1 Research and development intensity …... 27
4.3 Country-level predictors …... 28 4.3.1 Cultural distance …... 28 4.3.2 Market growth …... 30 4.3.3 Economic development …... 32 4.3.4 Country risk …... 33 5. VARIABLE CONSTRUCTION …... 36 5.1 Dependent variables …... 36 5.2 Independent variables …... 36 5.2.1 Firm-level variables …... 36 5.2.2 Industry-level variables …... 37 5.2.3 Country-level variables …... 37 6. METHODOLOGY …... 39
6.1 Method of data collection …... 39
6.2 Data sample …... 40
6.2.1 Sample construction …... 40
6.3 Empirical analysis …... 41
6.3.1 Preliminary analysis of the data …... 41
6.3.1.1 Outliers …... 42
6.3.1.2 Correlation ... 42
6.3.2 Regression analysis …... 42
7. EMPIRICAL RESULTS …... 44
7.1 Results binomial logistic regression analysis …... 44
7.1.1 Results binomial logistic regression analysis – entry mode choice …... 44
7.1.1.1 Firm-specific predictors …... 45
7.1.1.2 Industry-specific predictors …... 45
7.1.1.3 Country-specific predictors …... 45
7.1.1.4 Aggregate binomial logistic regression results …... 46
7.1.2 Results binomial logistic regression analysis – establishment mode choice ….. 46
7.1.2.1 Firm-specific predictors …... 47
7.1.2.2 Industry-specific predictors …... 47
7.1.2.3 Country-specific predictors …... 47
7.1.2.4 Aggregate binomial logistic regression results …... 47
7.2 Discussion of the results …... 48
7.2.1 Entry mode choice …... 48
7.2.1.1 Firm-specific predictors …... 49
7.2.1.2 Industry specific predictors …... 49
7.2.1.3 Country-specific predictors …... 50
7.2.1.4 Summary of the results for the entry mode choice …... 50
7.2.2 Establishment mode choice …... 51
7.2.2.1 Firm-specific predictors …... 52
7.2.2.2 Industry specific predictors …... 52
7.2.2.3 Country-specific predictors …... 52
7.2.2.4 Summary of the results for the establishment mode choice …... 52
7.2.3 Dual entry-establishment mode choice …... 53
8. CONCLUSIONS …... 54
7.1 Conclusion …... 54
7.3 Limitations and suggestions for future research... 56
REFERENCES …... 58
LIST OF TABLES AND FIGURES
List of tables
Table 1 Overview hypotheses …... 35
Table 2 Sample composition …... 41
Table 3 Binomial logistic regression analysis results – model 1, 2 and 3... 44
Table 4 Binomial logistic regression analysis results – model 3, 4 and 5 …... 46
Table 5 Overview of the hypothesized and empirical signs for the entry mode choice …... …... 48
Table 6 Overview of the hypothesized and empirical signs for the establishment mode choice …... 51
Table7 Firm-, industry- and country-level predictors …... 54
List of tables appendix Table I Sample distribution of firms across two-digit SIC industries ... 71
Table II Geographic distribution of the foreign investments in the sample …... 73
Table III Descriptive statistics for the independent variables …... 75
Table IV Correlation matrix …... 76
List of figures Figure 1 FDI mode choices …... 17
ABBREVIATIONS
CIA Central Intelligence Agency
FDI Foreign direct investment
HDR Human Development Report
JV Joint venture
MNE Multinational enterprise
OECD Organization for Economic Co-operation and Development
R&D Research and development
SIC Standard Industrial Classification
UN United Nations
UNCTAD United Nations Conference on Trade and Development
UNDP United Nations Development Programme
WOS Wholly owned subsidiary
1. INTRODUCTION
The ongoing globalization as well as low growth forecasts for the industrialized nations forces more and more companies to adjust their global strategies (Kaufmann et. al, 2005). During the past years volume and importance of cross-border expansions of multinational enterprises (MNEs) increased, thus stimulating the scholarly debate (Agarwal, 1997; Dikova and Brouthers, 2008; UNCTAD, 2009). While foreign direct investment (FDI) is an inherent part of today´s world economy, the understanding of its determinants is still insufficient (Brouthers, 2002; Buch et al., 2005).
1.1 FDI mode choices – a critical issue
Firms that want to expand their business across the boarders of their home country face at least three strategically important decisions (Dikova and Brouthers, 2008). First, a choice between non-equity (exporting, licensing and franchising) and non-equity (foreign direct investment, FDI) modes of entry has to be made. The second decision deals with the percentage of ownership an MNE desires in the foreign venture (wholly owned or joint venture), which is known as the entry mode choice (Dikova and van Witteloostuijn 2007). The third decision presented in the scholarly literature concerns the establishment mode choice (Brouthers and Brouthers, 2000). MNEs can choose to create a new foreign operation from scratch (invest in a greenfield) or acquire an existing venture (acquisition).
Those choices are crucial for MNEs, because they affect the survival probabilities for foreign subsidiaries, and thus have a significant effect on performance in international markets (Chung and Enderwick, 2001; Nakos and Brouthers, 2002; Root, 1994). Making the wrong choices can be costly and might force an MNE to leave the foreign market again. In order to provide a better understanding of the underlying forces that drive foreign direct investment, research focuses on examining the impact of a big variety of firm-, industry- and country-specific factors on the FDI mode choices.
1.2 Why focus on German FDI?
While most existing entry and establishment mode studies are either based on data for a wide range of nations (e.g. Brouthers, 2002; Dikova and van Witteloostuijn, 2008), or data for the USA and Japan as both host and home country for FDI (e.g. Delios and Beamish, 1999; Hennart and Park, 1993), scholars highlight that future research should focus on other countries in order to expand the applicability of their findings (Brouthers, 2002; Dikova and van Witteloostuijn, 2007).
In response to this suggestion, I will analyse FDI mode choices for a sample of German based MNEs. Cultural background and business practices of German MNEs vary from those of firms headquartered in the US or Japan, therefore it is interesting to analyse to what extent results of studies based on US and Japanese data can be generalized.
1.3 Problem statement
Despite the vast growth of research on FDI, due to immanent flaws in existing studies as well as open issues, further research in this field is necessary (Dikova and Brouthers, 2008).
The majority of researchers focuses either on the entry mode (Delios and Beamish, 1999; Yiu and Makino, 2002; Chan and Makino, 2007; Cui and Jiang, 2009) or the establishment mode choice (Hennart and Park, 1993; Brouthers and Brouthers, 2000) in isolation. However, it would be misleading to assume that those decisions are made separately (Dikova and van Witteloostuijn, 2007). There is a growing recognition that it is necessary to bridge both streams of FDI literature, since both types of FDI decision are influenced by similar factors. Some researchers, such as Dikova and van Witteloostuijn (2007) as well as Kogut and Singh (1988), tried to resolve this weakness by integrating both decisions in one single model, however they conclude that in order to find out to what extent their findings are generalisable to other countries and to MNEs from other parts of the world further research is necessary. Therefore, the following research question is formulated:
In a foreign direct investment (FDI) decision of German based MNEs, which determinants are crucial in explaining the entry and establishment mode choice?
sub-questions are formulated:
1. What potential firm-, industry- and country-level determinants of the entry and
establishment mode choice can be derived from existing literature?
2. Does a significant relationship between the potential firm-, industry- and country-specific
determinants derived from the literature and the entry and establishment mode choice of German based MNEs exist?
Those research questions imply a thesis structure, which consists of two main parts. The first sub-question is answered by performing a literature study. Theoretically exploring the determinants of the FDI mode choices, this part derives hypotheses regarding the potential factors influencing the entry and establishment mode choice. The empirical analysis that follows intends to answer the second sub-question. This empirical part serves as a reflection of the expectations established in the literature study. It explores the applicability of the created conceptual model for a sample of foreign investments made by German based MNEs.
1.4 Contribution
This thesis intends to contribute to the existing FDI literature in several ways. In general, it presents a theoretical and empirical extension to previous research, in analysing the effect of firm-, industry-and country-specific factors on international entry industry-and establishment mode choices.
Whereas most of the existing literature examines the entry and establishment mode choice separately, I aspire to integrate both mode choices in one single analysis focusing on shared elements.
Furthermore, I contribute to the growing debate concerning the explanatory value of traditional transaction cost economics (e.g. Ghoshal and Moran, 1996; Kogut and Zander, 1993; Madhok, 1997) by integrating insight from the institutional theory and the organizational learning perspective in my analysis.
1.5 Thesis structure
This thesis consists of two main parts; an extensive literature review and an empirical analysis on FDI mode choices.
The literature review starts with exploring the entry and establishment mode choice, in general, and wholly owned subsidiaries and joint ventures as well as acquisitions and greenfield investments, in particular (chapter 2). Thereafter, theories that are commonly used to explain the internationalization of MNEs are presented (chapter 3). The focus is on the transaction cost and institutional theory as well as on the organizational learning perspective. In order to determine possible predictors of the FDI mode choices, chapter 4 focuses on shared elements of both streams of research. The hypotheses comprise the firm-, industry- and country-specific determinants.
2. FDI MODE CHOICES
In order to explain the strategic options MNEs have when entering a foreign market, this chapter pays attention to the FDI mode choices in general and to joint ventures, wholly owned subsidiaries, greenfield investments and acquisitions in particular. Both the entry and establishment mode choice are strategically important decisions (Dikova and Brouthers, 2008; Dikova and van Witteloostuijn, 2007; Gomes-Casseres, 1990). Each mode choice can be associated with a number of intrinsic benefits as well as risks.
Research in this field lacks consensus concerning the definition of the mode types that vary with regard to a number of characteristics including: Incentive intensity, ability to manage and control the foreign operation, enforceability of legal rights, and ease of knowledge transfer (Brouthers and Hennart, 2007; Hennart, 1988, 1989; Williamson, 1991). Since previous studies have used more than 16 different categorizations concerning entry and establishment mode choices,1
it is crucial to define the terminology that I will use.
2.1 Entry mode choice
The entry mode choice, which deals with the percentage of ownership an MNE desires to hold in the foreign venture, can be considered a critical decision since it affects both the foreign subsidiary ´s chance of success (Stopford and Wells, 1972) as well as its likelihood of survival (Li, 1995). Root (1994) defines an entry mode as “an institutional arrangement that a firm uses to market its product in a foreign market in the first three to five years, which is generally the length of time it takes a firm to completely enter a foreign market”.
Focusing on the choice between joint venture and wholly owned subsidiary, previous research has applied five major perspectives in order to define subsidiary ownership (Chan and Makino, 2007). The first perspective suggests that each entry mode can be associated with a different level of control in the foreign subsidiary (Anderson and Gatignon, 1986; Caves, 1982; Geringer and Hebert, 1989; Root, 1994). In this context control is defined as influence on operational and strategic decision making (Hill et al., 1990). The second perspective emphasizes that the parent firm inserts a certain amount of resource into its subsidiary (Vernon, 1983; Anderson and Gatignon, 1986). Resource commitment denotes to assets that cannot be redeployed to alternative uses without loss of value (Hill et al., 1990). Due to the fact that each of the presented entry modes involves resource
commitment (regardless of the varying levels), an ex post change of the initially selected mode is difficult assuming that it might cause a considerable loss of time and money (Pedersen, Petersen and Benito, 2002; Root, 1994). The third perspective views ownership as a governance mechanism that serves as a safeguard against dissemination of firm specific knowledge to the local partner (Davidson and McFetridge, 1985; Hill et al., 1990). According to the fourth perspective ownership can be regarded as a resource access channel facilitating the combination of complementary assets and capabilities of the partners of a subsidiary (Blodgett, 1991; Hamel, 1991; Inkpen and Beamish, 1997). The fifth perspective argues that ownership structures can be seen as the result of negotiations between MNE and host government or local partner, viz. the outcome depends on the relative bargaining power of the parties (Lecraw, 1984; Gomes-Casseres, 1990).
2.1.1 Joint venture
An international joint venture can be defined as a partnership or conglomerate that combines resources from at least two legally and economically independent organizations in order to create a new organizational entity (Inkpen and Beamish, 1997). The joint venture partners share decision making as well as risk; and invest, in addition to capital, personnel, material and intangible resources (Himmelmann and Hungerbach, 2005). Since the percentage of equity ownership varies, a distinction between majority-, parity- and minority joint ventures can be made (Delios and Beamish, 1999; Hellmann, 2007). Even though the distribution of ownership shares in a joint venture has an influence on the degree of control and resource commitment, their extent can be positioned between that of wholly owned subsidiaries and licensing agreements (Chan and Hwang, 1992).
2.1.2 Wholly owned subsidiary
A wholly owned subsidiary is a legally independent company entirely owned by the foreign investor (Haas, 2006). While the foreign subsidiary is usually responsible for the day-to-day operations as well as some strategic decisions, the MNE headquarters hold the ultimate control. All costs of opening up and serving the foreign market have to be paid by the MNE parent.
2.1.3 The choice between joint venture and wholly owned subsidiary
1989). Differences between those two internationalization strategies concern the degree of control and resource commitment in the foreign venture as well as the risk involved (Peng, 1995).
First of all, wholly owned subsidiaries offer the benefits of total profits and full control over the foreign subsidiary (Hill et al., 1990). Hence, sole ventures guarantee managerial autonomy and protect unique, firm-specific knowledge (especially technological and research and development (R&D) expertise) from unwanted dissemination (Dikova and van Witteloostuijn, 2007; Hennart, 1982; Madhok, 1997). Due to the fact that control is shared in a joint venture, it can be challenging to conciliate diverging capabilities, interests and goals (Dikova and van Witteloostuijn, 2007; Inkpen and Beamish, 1997). However, it is important to mention that control related uncertainties will decrease as an MNE becomes acquainted with the partner organization, its resources and willingness/ ability to cooperate (Chia and McGuire, 1996).
Secondly, wholly owned modes alleviate the knowledge transfer from headquarters to subsidiary (Dikova and van Witteloostuijn, 2007; Madhok, 1997). Approved internalized routines can be transferred and the utilization of firm-specific endowments can be optimized (Steensma and Corley, 2001). Participating in a joint venture, on the other hand, allows an MNE to gain valuable knowledge about the foreign market and the joint venture partner and thereby reduces decision-making uncertainty (Chi and McGuire, 1996). A better understanding of the market enables a firm to conquer location-specific disadvantages rooted in institutional context differences (Anand and Delios, 1997; Dyer and Singh, 1998; Oliver, 1997).
Furthermore, the costs and risk involved in a wholly owned subsidiary are higher compared to a joint venture (Hill et al., 1990). While an MNE has to cover all expenses in a wholly owned subsidiary itself, in a joint venture it gains access to resources/ technology controlled by its local partner (Dikova and van Witteloostuijn, 2007). Copying an entire operation is often not feasible, either because of high costs or missing access to resources in the host country (Inkpen and Beamish, 1997).
It might be difficult for an MNE to overcome the burden of foreignness without the support of a local partner (Dikova and van Witteloostuijn, 2007). Participating in a joint venture facilitates the adaptation of local norms, values and expectations and thereby enables an MNE to obtain legitimacy in the host country environment (Dacin et al., 2007).
2.2 Establishment mode choice
An MNE can choose to create a new foreign operation from scratch (greenfield) or acquire an existing venture (acquisition) (Dikova and van Witteloostuijn, 2007). Research concerning the establishment mode choice is rather limited compared to the extensive body of literature dealing with the ownership structure of foreign subsidiaries (Barkema and Vermeulen, 1998). Hennart and Park (1993, p. 1055) state, „there is no well-developed theory of the determinants of the choice between … greenfield investment and acquisition as alternative ways of entering foreign markets”. However, scholars suggest that institutional, cultural, transaction cost as well as experience factors have to be considered in order to explain the diversification mode choice (Brouthers and Brouthers, 2000).
2.2.1 Greenfield
A greenfield investment into a foreign market can be defined as the establishment of an entirely new organization abroad by an MNE headquartered outside the country (Hennart and Slagen, 2007). The subsidiary can either be fully (wholly owned subsidiary) or partially owned (joint venture) by the MNE (Barkema and Vermeulen, 1998). The establishment process of start-ups is often supported by expatriates who are assigned the task to select and hire local employees (Hofstede, 1991) and to build up the business step by step (Simmonds, 1990; Teece, 1982).
2.2.2 Acquisition
2.2.3. The choice between greenfield investment and acquisition
An MNE chooses its internationalization strategy, by means of greenfield investment or acquisition, with the objective to sustain and enhance its competitive advantage in the global market (Pradhan and Alakshendra, 2006). In order to better understand the underlying motives of an MNE that makes an establishment mode choice, it is important to highlight the differences between start-ups and acquisitions, which are contingent on characteristics of the investing firm, the target industry and the host country. Taking the intrinsic benefits as well as risks of greenfield investments and acquisitions into consideration, an MNE will choose the mode that serves as the superior competitive tool compared to other available options.
First of all, Caves (1982) argues that both information and management costs and the risk involved in an acquisition are lower compared to a greenfield investment. The fact that an MNE, which is buying an existing firm, gains new technological resources and knowledge of the foreign market, can indeed be considered an advantage (Prahald and Hamel, 1990). However, it is important to mention that returns of acquisitions are generally lower compared to start-ups, since an acquisition necessitates remunerating the capitalized value of rents (Brouthers and Brouthers, 2002; Larimo, 2003).
Furthermore, the establishment process of an acquisition is faster compared to a greenfield investment (Biggadike, 1979; Larimo, 2003). Caves and Mehra (1986) argue that, all else constant, the opportunity costs of a de novo entry increase in line with the speed of growth of the host market. In order to claim its relative market position, especially in oligopolistic industries, an MNE might engage in an acquisition with the objective to speed up the entry process (Davidson, 1982; Graham, 1974). Dikova and Brouthers (2009), on the other hand, allude that a speedy entry might cause post-acquisition problems. Cheng (2006) highlights that greenfield investments give an MNE the chance to adjust to the local business environment at its own pace. Thus, the ideal establishment speed for a foreign venture depends on the objectives of an MNE as well as the external circumstances.
2.3 Combining insight from both streams of research
subsidiary’s assets, the later rests upon the effective combination of the parent´s and local assets (Gatignon and Anderson, 1988; Gomes-Casseres, 1989). Hennart and Park (1993) argue that both choices are independent, because the desired degree of ownership, which represents the MNE´s need for control, can be obtained in a greenfield as well as in an acquisition.
Even though this approach dominates (early) research in this field of study, it would be restrictive to assume that the FDI decisions are isolated from one another (Dikova and van Witteloostuijn, 2007). The case study by Meyer and Tran (2006), for instance, shows that the distinction is not that clear. The growing recognition concerning a possible relatedness of both FDI mode choices is based on the refutation of the argumentation that the entry and establishment mode choice are conceptually different. Kogut (1988), for instance, emphasizes that joint ventures are not purely an issue of equity control, but also constitute a government mechanisms, which is subject to strategic motivations. In addition, Gatignon and Anderson (1988) suggest that the degree of ownership is influenced not only by the need for control, but also by other factors, such as the bargaining power of the parties. Furthermore, the analysis of previous studies discloses that both types of foreign investment decision are influenced by similar factors (e.g. asset specificity, host country institutional advancement, etc.), thus suggesting that a relationship between both FDI mode choices exists (Dikova and van Witteloostuijn, 2007).
Establishment mode
Greenfield
Acquisition
Entry mode
Shared
ownership
1. Greenfield JV2 3. Partial acquisition
Full
ownership
2. Greenfield WOS3
4. Full acquisition
Figure 1. FDI mode choices
Source: Brouthers and Hennart (2007), p. 399
2 JV = joint venture
For the reasons elucidated above, I will adopt the approach presented by Dikova and van Witteloostuijn (2007), who conducted a study on FDI mode choices in transition economies. A conceptual synthesis can be created by solely focusing on elements, which are shared in both streams of research. Thus, creating a framework, which can be used to analyse how the same set of independent variables influences the dual entry-establishment mode choice in case of a single foreign investment.
3. THEORETICAL BACKGROUND
In this chapter, the relevant literature on entry and establishment mode choices will be discussed. Existing studies in this field of research tend to consult traditional rational models (Dikova and Brouthers, 2008). Since most of the prevalent theories focus only on one or two of the three proposed levels, researchers highlight the need to consider multiple perspectives complementing one another at least partly. While the firm-level variables refer to transaction cost considerations, the industry- and country-level variables mirror entry barriers and patterns of oligopolistic behaviour (Kogut and Singh, 1988). Taking this suggestion and data availability constraints into consideration, I will develop and test a model of FDI mode choices that draws from both the transaction cost as well as the institutional theory and additionally includes the organizational learning perspective. The integration of those three perspectives in one framework is well accepted in FDI mode research and therefore will be applied in this thesis (e.g. Brouthers and Brouthers, 2000; Delios and Beamish, 1999).
3.1 Transaction cost theory
Transaction cost theory, one of the core international business theories simultaneously and independently developed by McManus (1972), Brown (1976), Buckley and Casson (1976), Rugman (1981), and Hennart (1977, 1982), is the dominant theoretical perspective applied in FDI mode studies (Hennart, 1991). Researchers consult transaction cost economics (Williamson, 1975, 1985) concerning both establishment as well as entry mode choice issues (Zhao et al., 2004). Amongst others, Kogut and Singh (1988), Hennart and Park (1993), Andersson and Svensson (1994), Brouthers and Brouthers (2000) and Larimo (2003) contribute to this international business theory by analysing the choice between greenfields and acquisitions. Gatignon and Anderson (1988), Kim and Hwang (1992), Delios and Beamish (1999), Meyer (2001) and Brouthers (2002), on the other hand, focus on the choice between wholly owned subsidiaries and joint ventures.
Transaction cost theory argues that the neoclassical assumptions of perfect knowledge and perfect enforcement do not apply in the real world (Teece, 1981; Dunning and Rugman, 1985). Since markets are never fully efficient and natural market imperfections exist, market transaction costs (the sum of information, enforcement and bargaining costs) arise (Hennart, 1991). Hence, MNEs can be characterised as institutions for bypassing these imperfections and optimizing transaction costs (Chiles and McMackin, 1996).
mode choices by considering the cost of operating in a foreign market and the efficiency of alternative organizational structures (Robins, 1987; Madhok, 1997). Williamson (1973) defines opportunism as “an effort to realize individual gains through a lack of candor or honesty in transactions.” It arises either from asymmetrically distributed information or the inability of firms to write complete contracts as well as flaws in the monitoring of those agreements. While opportunism can be regarded as a behavioural uncertainty, bounded rationality can be classified as a contextual uncertainty, which means that decision makers´ cognitive capabilities are limited (Williamson., 1989). Transaction cost research identifies asset specificity as third dimension of transaction (Zhao et.al, 2004). It can be defined as the “degree to which an asset can be redeployed to alternative uses and by alternative users without sacrifice of productive value” (Williamson, 1989).
If incentives for bargaining and dishonesty are high, it might be beneficial to coordinate the interests of seller and buyer (Hennart, 1991). By internalizing markets, MNEs create a hierarchy and turn independent agents into employees (Hennart, 1982, 2000). In a system in which individuals are rewarded for following the MNE´s directives, cheating is no longer profitable. Hence, transaction cost theorists argue that the main drivers behind FDI are the reduction of transaction costs as well as the internalization of non-pecuniary externalities (Hennart, 1991). The aim of MNEs investing in foreign markets is the exploitation of its firm specific advantages at low marginal costs (Zhao et al., 2004). An important precondition for FDI is that hierarchical coordination causes lower costs compared to coordination through prices (Hennart, 1991).
Despite producing encouraging results, researchers criticize that transaction cost theory tends to understate the significance of contextual factors in the choice of a FDI mode (Yiu and Makino, 2002). Thus, combining transaction cost, experience, cultural and institutional context variables in one unifying framework will enhance the understanding of international entry and establishment mode choices (North, 1990; Roberts and Greenwood, 1997).
3.2 Institutional theory
with the goal of attaining the most “efficient” ownership structure, the latter pays attention to the role of “legitimacy” as the central determinant of the FDI mode choices (Delios and Beamish, 1999). Concerning the aspiration to obtain legitimacy, organizations strive to affiliate structures and practices that are “isomorphic” to those of other organizations in the environment (DiMaggio and Powell, 1983; Dacin 1997). Isomorphic pressure exists due to several factors (Yiu and Makino, 2002). According to Scott (1995), three pillars of the institutional environment exist. The regulative pillar alludes to the rules and laws within a society; the normative pillar defines the area of social values, cultures and norms; and the cognitive pillar illustrates the established cognitive structures in society, which are taken for granted. Even though conformist behaviour does not ensure that an MNE is more efficient compared to its more deviant peers, being similar to other organizations will be rewarded (Yiu and Makino, 2002). Meyer and Rowan (1977) argue that MNEs become matched with their environment either because parallelism enables them to successfully manage technical and exchange interdependencies or because it is taken for granted that doing so is the proper way to organize.
3.3 Organizational learning perspective
3.4 Why use an integrating framework?
While early research in this field was mainly based on the transaction cost theory, scholars highlight that combining different theoretical perspectives will improve the understanding of FDI mode choices (Brouthers et al., 2003; Delios and Beamish, 1999; Heide and John, 1988). As elaborated above, more recent studies have started to integrate the level of international experience of the foreign firm as well as the cultural and institutional environment of the host country in their argumentation (Brouthers, 2002; Brouthers and Brouthers, 2000; Brouthers et al., 2003; Delios and Beamish, 1999).
4. HYPOTHESES DEVELOPMENT
Previous studies of FDI mode choices have considered a big variety of determinants. Therefore, it is important to select and construct the hypotheses wisely. In order to provide a broad understanding of MNEs´ entry and establishment mode choice, it is essential to include firm-, industry and country-level predictors in the analysis. Focusing on the hypotheses which are used in both streams of FDI mode research and additionally taking data availability, theoretical consistency and previous empirical results into consideration I develop the subsequently presented hypotheses. Based on the different elements included in this thesis, the following conceptual model is proposed. (see figure 2)
Figure 2. Overview of the relevant determinants of the entry and establishment mode choice
4.1 Firm-specific predictors
The first group of variables discussed here are firm-specific predictors of the entry and establishment mode choice. The focus is on international experience and firm size.
4.1.1 International experience
learning perspective. MNEs obtain important knowledge through FDI activities concerning both host countries in particular as well as international operations in general (Barkema et al., 1996; Delios and Beamish, 1999). MNEs that operate in diverse national markets are forced to deal with new ideas, practices and consumer demands (Barkema and Vermeulen, 1998; Miller and Chen, 1994). Due to this fact those firms are prone to develop new technological and organizational capabilities, which are gradually transformed into routines that serve as basis for future actions (Padmanabhan and Cho, 1999). As an MNE´s level of experience increases, disadvantages rooted in the lack of local knowledge as well as the risks associated with it decrease and the ability to make a commitment to the foreign venture increases (Brouthers and Brouthers, 2000; Chang, 1995; Johanson and Vahlne, 1977).
Scholars argue that internationally inexperienced MNEs tend to prefer low control modes (Delios and Beamish, 1999; Inkpen and Beamish, 1997). Those MNEs might face difficulties in exercising control over a foreign venture due to the lack of know-how (Gatignon and Anderson, 1988). It is predictable that errors will occur when an MNE exercises control without understanding how to use it (Teece, 1976). Hence, it might be beneficial to transfer management tasks to a local joint venture partner which is better qualified (Li, 1995). Makino and Delios (1996) highlight that the relative utility of shared ownership diminishes as MNEs´ experience increases. Being less dependent on the knowledge of a local partner, experienced MNEs tend to favour wholly owned ventures, because this mode of entry avoids potential conflicts as well as costs associated with integration (Agarwal and Ramaswami, 1992; Li, 1995).
Previous empirical research confirms that international experienced MNEs tend to prefer high control modes (e.g. Barkema et al., 1996; Delios and Beamish, 1999; Gomes-Casseres, 1990; Johanson and Vahlne, 1977). Therefore, I hypothesize that:
Hypothesis 1a:
The greater the level of an MNE´s international experience, the higher the possibility for a wholly owned subsidiary.
and Mehra (1986) argue that the likelihood of entry via acquisition decreases as international experience increases. Mature MNEs, on the other hand, which have acquired the necessary knowledge in-house, prefer greenfield investments in order to avoid paying a take over premium (Caves, 1982; Cho and Padmanabhan, 1995; Wilson, 1980). Furthermore, it is easier to transfer firm-specific knowledge to a start-up. In case of an acquisition an MNE inherits the acquired firm´s labour force and company culture. New rules, procedures, conventions and organizational strategies have to be introduced to the employees (Levitt and March, 1988). This task can be extremely challenging and time-consuming, especially when the difference between the nature of parent and subsidiary is large (Barkema and Vermeulen, 1998).
Even though the theoretical explanations are plausible, previous empirical studies have produced mixed results. While some researchers found a positive relationship between the level of international experience and the likelihood of a greenfield investment (e.g. Barkema and Vermeulen, 1998; Brouthers and Brouthers, 2000; Dubin, 1976; Hennart and Park, 1993; Larimo, 1997; Padmanabhan and Cho, 1999; Yamawaki, 1994), other studies indicate the opposite relationship (e.g. Caves and Mehra, 1986; Forsgren, 1984; Harzing, 1998). Due to the fact that previous empirical findings can not clearly proof how the level of international experience influences an MNE´s establishment mode choice, further research is necessary. Since many well-known scholars support the theoretical argumentation presented in this thesis, the following hypothesis is formulated:
Hypothesis 1b:
The greater the level of an MNE´s international experience, the lower the probability for an acquisition.
4.1.2 Firm size
Larger MNEs are more likely to posses enough resources to engage in this form of international expansion (Agarwal and Ramaswami, 1992). Motivated by the benefits of total profits and full control, large MNEs are often willing to accept the higher risk involved in high control modes, since those firms usually have buffers to compensate in case of failure (Hill et al., 1990; Taylor et al., 1998).
Multiple studies have provided empirical evidence confirming the theoretical prediction that a positive relationship between firm size and level of control in the foreign venture exists (e.g. Agarwal and Ramaswami, 1992; Erramilli and Rao, 1993; Gomes-Casseres, 1990; Taylor et al., 1998). Hence, I hypothesize that:
Hypothesis 2a:
The larger an MNE, the higher the possibility for a wholly owned subsidiary.
Focusing on the establishment mode choice, Hennart (1982) argues that larger firms are prone to favour acquisitions over start-ups. Since buying an existing firm is costly, larger MNEs are more likely to possess the resources needed and are willing to implement an acquisition in order to utilize the benefits of this enstablishment mode.
Even though this simple theoretical explanation seems to be plausible, empirical research has produced somewhat mixed results. While many researchers found that larger MNEs prefer acquisitions (e.g. Andersson et al., 1992; Barkema and Vermeulen, 1998; Cho and Padmanabhan, 1995; Larimo, 1993, 2003), Dubin (1976), who analysed the establishment mode choice for a sample of U.S. based firms, found that smaller MNEs prefer buying an existing firm. Based on the theoretical explanations as well as the preponderance of empirical evidence the following hypothesis can be formulated:
Hypothesis 2b:
The larger an MNE, the higher the possibility for an acquisition.
4.2 Industry-specific predictors
agree that industry-level variables are important factors influencing the entry and establishment mode choice. The industry-specific R&D intensity has been included in past studies, however further research examining this predictor is necessary in order to generalize previous findings.
4.2.1 Research and development intensity
MNEs that operate in industries which require a high level of R&D face potential opportunism problems (Delios and Beamish, 1999; Larimo, 2003). Shirking, free-riding or technology dissemination are threats an MNE that depends on a local partner is exposed to (Chang and Rosenzweig, 2001; Hennart, 1991; Hill et al., 1990). Due to the fact that protecting firm-specific knowledge from misappropriation can be expansive, MNEs might utilize higher control governance structures when a transaction requires specialized investments (Gatignon and Anderson, 1988; Taylor et al., 1998). When the level of R&D is low, on the other hand, MNEs are less worried about safeguarding firm-specific knowledge (Hill et al., 1990; Williamson, 1985). In this case, the benefits of a high level of control would not offset the costs necessary to achieve it (Brouthers, 2002). To sum it up, based on the proposition that the risks associated with exchange through markets and the incentives to internalize transactions increase in line with R&D intensity, transaction cost theory argues that the level of ownership is positively related to the degree of R&D intensity.
Early research by Stopford and Wells (1972) supports the hypothesis that as the R&D intensity decreases, MNEs tend to prefer joint ventures over wholly owned subsidiaries. In addition, Gatignon and Anderson (1988), who conducted an empirical test of transaction cost explanations for a sample of MNEs headquartered in the US, provide evidence that industries with a high level of R&D expenses are positively related to entries through high ownership modes. Based on the transaction cost theory and prior research findings the following hypothesis is formulated:
Hypothesis 3a:
The higher the level of R&D intensity of the industry an MNE operates in, the higher the possibility for a wholly owned subsidiary.
hand, is the most efficient way to transfer firm-specific knowledge (Hennart and Park, 1993). Furthermore, managers prefer greenfield investments when R&D intensity is high, because the risk, in terms of organizational control, involved in this mode is smaller compared to acquisitions (Brouthers and Dikova, 2009; Yoshida, 1987).
Empirical research largely supports the arguments presented above (e.g. Andersson and Svensson, 1994; Brouthers and Brouthers, 2000; Chen, 2008; Chen and Zeng, 2004; Cho and Padmanbhan, 1995; Drogendijk and Slagen, 2006; Hennart and Park, 1993; Larimo, 2003; Tsai and Cheng, 2004). However, I have to mention that a few researchers support a competing point of view (Brouthers and Brouthers, 2000). For instance, Caves and Mehra (1986) as well as Kogut and Singh (1988) argue that technologically intensive MNEs might lack specific knowledge concerning the host market and therefore prefer acquisitions. Since this perspective is empirically not supported, I hypothesize that:
Hypothesis 3b:
The higher the level of R&D intensity of the industry a MNE operates in, the lower the possibility for an acquisition.
4.3 Country-level predictors
Finally, I examine the influence of country-level predictors on MNEs´ entry and establishment mode choice. While cultural distance is the most frequently studied country-level determinant, other variables have attracted less attention. However, researchers agree that country risk also plays an important role.
4.3.1 Cultural distance
information about local firms and monitoring the performance are reduced due to the fact that values and customs are partly shared.
Scholars argue that the greater the cultural distance an MNE faces when entering a foreign market, the lower the degree of control in the new venture (Gatignon and Anderson, 1988; Kim and Hwang, 1992; Kogut and Singh, 1988). A lack of comfort and familiarity as well as disagreement with the values and practices of the host country might discourage mangers to assume a high level of involvement (Davidson, 1980; Richman and Copen, 1972). In addition, it can be problematic to transfer approved internalized routines from parent to a dissimilar operating environment (Dikova and van Witteloostuijn, 2007; Steensma and Corley, 2001). MNEs might cooperate with a local joint venture partner in order to avoid high information cost by passing the management on to the affiliate (Gatignon and Anderson, 1988; Root, 1994). Drawing from institutional theory, one might argue that cooperating with a local joint venture partner facilitates the adaptation of local norms, values and expectations, and thereby enables an MNE to obtain legitimacy in the host country environment (Dacin et al., 2007). Overcoming the burden of foreignness without the support of a local partner can be especially challenging when cultural distance is high (Dikova and van Witteloostuijn, 2007).
Empirical evidence supports the argumentation presented above. Several researchers found that MNEs tend to favour joint ventures when cultural distance is high (e.g. Davidson and McFetridge, 1985; Gatignon and Anderson, 1988; Kim and Hwang, 1992; Kogut and Singh, 1988). Based on the transaction cost and the institutional theory as well as on previous empirical findings, the following hypothesis is formulated:
Hypothesis 4a:
The greater the cultural distance between host and home country, the lower lower the possibility for a wholly owned subsidiary.
With regard to the establishment mode choice, Dikova and Brouthers (2009) emphasize that cultural distance between host and home country is the most frequently studied country-level determinant. However, previous research presents two lines of argument concerning the relationship between cultural distance and the choice between greenfield investment and acquisition.
Singh, 1988; Levitt and March, 1988). Hence, MNEs might prefer to launch start-ups and fully transfer managerial practices as well as technological routines developed at the headquarters (Cho and Padmanabhan, 1995; Drogendijk and Slagen, 2006). Many studies, which rest upon this argumentation, found empirical proof that as cultural distance increases, MNEs favour setting up a new venture (Barkema and Vermeulen, 1998; Chang and Rosenzweig, 2001; Cheng, 2006; Drogendijk and Slagen, 2006; Elango, 2005; Harzing, 2002; Herrmann and Datta, 2006; Kogut and Singh, 1988; Slagen and Hennart, 2007; Tsai and Cheng, 2004).
Secondly, other researchers highlight that MNEs´ insufficient knowledge regarding the political, cultural and social standards in culturally distant host countries can be a barrier to subsidiary survival and success (Dikova and Brouthers, 2009). The fact that MNEs, which are buying existing firms, gain essential knowledge of the foreign market, can be considered an advantage of acquisitions (Anand and Delios, 1997; Prahalad and Hamel, 1990; Brouthers and Brouthers, 2000; Harzing, 2002). For this reason, MNEs entering a dissimilar environment tend to favour acquisitions, while MNEs investing in a similar environment are inclined to prefer start-up investments (Brouthers and Brouthers, 2000; Ruiz-Moreno et al., 2007).
Dikova and Brouthers (2009) try to explain this discrepancy arguing that cultural distance is not central to transaction cost economics and therefore leaves space for opposing predictions. It can also be argued that the conflicting evidence in the existing literature is due to the fact that some studies ignored the moderating effect of international experience (Slagen and Hennart, 2008). This thesis takes the first position, since international experience is incorporated in the analysis. The majority of greenfield investments included in the dataset are established by internationally experienced MNEs. Therefore, the second position is less convincing. Hence, I hypothesize that:
Hypothesis 4b:
The greater the cultural distance between host and home country, the lower the possibility for an
acquisition.
4.3.2 Market growth
Scholars argue that MNEs´ FDI mode choices are amongst others influenced by the market potential, which affects both market capacity and opportunity costs (Agarwal and Ramaswami, 1992; Dikova and Brouthers, 2009; Kim and Hwang, 1992).
feature a relatively big receptiveness for productive capacity, thus enabling MNEs to enhance their firm efficiency (Brouthers, 2002). Since over capacity is a potential hazard, MNEs might have an antipathy to make larger investments in stagnant or shrinking markets. Furthermore, it can be assumed that opportunity costs increase in line with the market growth rate, because it influences expansion opportunities and pricing strategies. Due to the possibility to exploit economies of scale and establish a long-term market presence, scholars anticipate that wholly owned subsidiaries are favoured in high growth markets (Agarwal and Ramaswami, 1992; Taylor et al., 1998). In slow growing markets, on the other hand, MNEs tend to prefer joint ventures, because this mode of entry requires a lower level of resource commitment, which MNEs might regard as a benefit considering the relatively low expected returns in slow growing markets (Kim and Hwang, 1992).
Empirical research evaluating how the industry growth rate influences the entry mode choices has produced mixed results. While Hennart (1991) found that MNEs entering high-growth industries favour shared ownership, Makino and Neupert (2000) found that full ownership is preferred in this case. Based on the theoretical background, I hypothesize that:
Hypothesis 5a:
The faster a market is growing, the higher the possibility for a wholly owned subsidiary.
Focusing on the establishment mode choice, scholars highlight that while creating a new venture from scratch increases the number of firms competing in the marketplace, acquiring an existing venture avoids capacity expansion (Caves and Mehra, 1986; Cheng, 2006). Zejan (1990) argues that MNEs tend to favour acquisitions in slow growing markets in order to avoid potential conflicts with competitors. Taking away customers from well-established firms, which experience a decline in sales and profits, might evoke cravings for revenge (Brouthers and Brouthers, 2000; Elango and Sambharya, 2004). It is self-evident that incumbent firms would take actions in order to defend their market share and discourage expansion of new entrants (Zejan, 1990). It can be assumed that it is easier for foreign firms to gain a market share in fast growing markets. Therefore, MNEs tend to favour greenfield investments. Since many new customers are available, MNEs are not directly seen as competition. In addition, it can be an opportunity to implement a niche strategy.
preponderance of encouraging results, some studies found a U-shaped relationship (e.g. Caves and Mehra, 1996; Hennart and Parl, 1993). Nevertheless, I formulate the following hypothesis:
Hypothesis 5b:
The faster a market is growing, the lower the possibility for an acquisition.
4.3.3 Economic development
Meyer and Tran (2006) argue that emerging economies are characterized by a high growth potential, but also feature a lack of economic development and institutional advancement. It is important to mention that the term economic development is not identical to economic growth (Nafziger, 2006). While economic growth refers to a country´s income or production, economic development alludes to a change in output in combination with a transformation of the economic structure and the output distribution.
Scholars highlight that MNEs investing in transition economies usually face high transaction costs for any of the enlisted reasons: “Foreign firms lack information about local partners; they must negotiate with agents inexperienced in business negotiations; and they face unclear regulatory frameworks, inexperienced bureaucracies, underdeveloped court systems and corruption” (Meyer, 2001, p. 358). In addition, transition economies are often characterised by weak property rights protection, which makes contracting and engaging in a joint venture costly (Williamson, 1996). In order to protect firm-specific knowledge and to reduce transaction costs, MNEs prefer wholly owned subsidiaries in transition economies (Delios and Beamish, 1999).
Empirical research evaluating the influence of the level of economic development on MNEs´ entry mode choices is rather limited. However, the study by Dikova and van Witteloostuijn (2007) supports the arguments presented above. Hence, I hypothesize that:
Hypothesis 6a:
The lower the level of economic development, the higher the possibility for a wholly owned subsidiary.
developed economies. Even though, this line of argument is contrary to traditional transaction cost logic, it dominates exiting studies concerning the establishment mode choice (Dikova and van Witteloostuijn, 2007). Scholars justify this approach by arguing that long-term returns in transition economies are more uncertain and that an MNE might only profit from an acquisition in the short-run.
Wilson (1980), Andersson et al. (1992) as well as Cho and Padmanabhan (1995) found empirical support for the argumentation presented above. Hence, I hypothesize that:
Hypothesis 6b:
The lower the level of economic development, the lower the possibility for an acquisition.
4.3.4 Country risk
Host country risk is a wide-ranging concept, which comprises factors concerning the economic, political and legal environment (Brouthers, 2002; Davidson, 1982; Delios and Beamish, 1999). According to Agrawal and Ramaswami (1992, p.6) country-specific investment risk can be defined as “the uncertainty over the continuation of present economic and political conditions and government policies which are critical to the survival and profitability of a firm´s operations in that country”.
Focusing on the entry mode choice, scholars argue that MNEs entering countries characterized by low country risk prefer wholly owned modes with the aim to obtain higher returns (Erramilli and Rao, 1993; Kim and Hwang, 1992). When setting up a subsidiary in a more uncertain and volatile environment, on the other hand, MNEs favour shared ownership (Anderson and Gatignon, 1986). It is advisable to reduce the level of resource commitment in “dangerous” countries, in order to be more flexible and minimize the exposure to country risk (Brouthers, 2002; Chan and Hwang, 1992; Delios and Beamish, 1999). In addition, participating in a joint venture allows a firm to gain valuable knowledge about the foreign market, which can be especially important in an uncertain environment, and thereby reduces long-term costs (Beamish and Banks, 1987; Chi and McGuire, 1996).
Hypothesis 7a:
The higher the investment risk in the host country, the lower the possibility for a wholly owned subsidiary.
Furthermore, Hauser (2005) suggests that the value of a greenfield investment suffers more from an uncertain environment compared to the value of an acquired venture. It can be assumed that local firms are familiar with the uncertainty in the host market. Buying an existing firm usually involves the take over of at least some local managers, whose knowledge can be considered a crucial component for an MNE´s success in the foreign market (Hennart and Park, 1993). It seems to be plausible that these skills are more valuable the higher the uncertainty in the target country. When country risk is high, the benefits associated with buying an existing venture offset the higher costs. Since empirical evidence is limited, further research is necessary. Hence, I formulate the following hypothesis:
Hypothesis 7b:
An overview of the hypothesized influences of all independent variables of the entry and establishment mode choice is shown below. (see table 1)
Entry mode choice Establishment mode choice Hypothesis Independent variable Expected relation
to preference for wholly owned subsidiary Expected relation to preference for acquisition H1a, H1b H2a, H2b Firm-level variables International experience Firm size + + -+ H3a, H3b Industry-level variable R&D intensity + -H4a, H4b H5a, H5b H6a, H6b H7a, H7b Country-level variables Cultural distance Market growth Economic development Country risk -+ -+ +
5. VARIABLE CONSTRUCTION
This section discusses the variables used in the empirical analysis. The operationalisation of the variables is clarified and justified using existing research on FDI mode choices. In section 6.1 the dependent variables entry and establishment mode choice are discussed. Section 6.2, which describes the independent variables, is subdivided into firm-, industry- and country-level determinants.
5.1 Dependent variables
The MNEs´ entry (full vs. shared ownership) and establishment mode choice (greenfield vs. acquisition) into one of the selected countries represent the dependent variables.
First of all, entry mode is defined as a dummy variable, which is coded 1 in the case of full ownership and 0 in the case of shared ownership. It is important to mention that research in this field lacks consensus concerning the distinction between wholly owned subsidiary and joint venture. While some researchers argue that a wholly owned subsidiary is owned by only one parent, others use a 95% (e.g. Anderson and Gatignon, 1986; Comes-Casseres, 1989; Hennart, 1991; Padmanabhan and Cho, 1996) or even a 80% cut off line. In accordance with the definition used by Brouthers (2002), this thesis distinguishes between wholly owned subsidiaries (95% or more ownership) and joint ventures (5% - 94% ownership).
The establishment mode choice is also captured by a dummy variable, which takes the value of 1 in case of an acquisition and 0 when the MNE made a greenfield investment.
5.2 Independent variables
After having discussed the operationalisation of the dependent variables, this section will focus on the independent variables in the research.
5.2.1 Firm-level variables
Taking both the standard operationalisation (e.g. Barkema and Vermeulen, 1998; Caves and Mehra, 1986; Kogut and Singh, 1988; Ruiz et al., 2007; Slagen and Hennart, 2008) as well as Larimo´s (2003) specification into consideration, the independent variable international experience is measured using the natural logarithm of the number of foreign countries in which the MNE has manufacturing investments before the subsidiary in question.
logarithm of firms´ assets (e.g. Barkema et al., 1996; Chang and Rosenzweig, 2001), total sales (e.g. Agarwal and Ramaswami, 1992; Larimo, 2003) or employees (e.g. Elango and Sambharya, 2004; Erramilli and Rao, 1993; Gatignon and Anderson, 1988) in the year of entry. All of those operationalisations are scientifically accepted. Since data availability is best for number of employees, I chose to use this measure.
5.2.2 Industry-specific variables
In accordance with Larimo (2003), the independent variable R&D intensity is operationlised using the OECD (high, medium-high, medium-low and low) technology field of industries classification. Since R&D intensity can be considered a structural feature of industries, it can be assumed that it is a stable measure (Gatignon and Anderson, 1988).
5.2.3 Country-level variables
Cultural distance between Germany and each target country is calculated using Hofstede´s (1980)
cultural dimensions (power distance, uncertainty avoidance, masculinity/ femininity and individualism) and the index of Kogut and Singh (1988), which is defined as:
4
(1) CDj = [ (Iij – IiGER)2 / Vi ] / 4 i = 1
Where Iij is the index of the cultural dimension i in country j, Vij is the variance of the index of the
dimension i, and CDj is the cultural distance between country j and Germany (GER).
The independent variable market growth is measured as the average annual percentage growth rate of the host country´s real GDP during a 3-year period prior to the entry. This operationalisation was commonly used in previous studies on FDI mode choices (e.g. Alvarez, 2003; Brouthers and Brouthers, 2000; Larimo, 1997, 2003).
In accordance with Larimo (2003) the independent variable economic development is captured by a dummy variable which is coded 0 in the case of a non-OECD country and 1 in the case of an OECD member country.
Country risk is a multidimensional phenomenon comprising factors of the political, legal and
based on pre-1970 data, Barkema and Vermeulen (1998) argue that categories should be adjusted for the window of analysis (1978 - 2010). Hence, I modify the classification for countries that seem to be not well represented in the original cluster using more recent data bi-annually published by Euromoney since 1982 (Euromoney, 2008). Rearrangements are necessary, if the difference between the individual country´s score and its cluster´s average exceeds one standard deviation (Barkema and Vermeulen, 1998). In those cases the country is assigned to the next category. Furthermore, the classification system by Goodnow and Hansz (1972) does not categorize all countries included in this study. Missing countries are also assigned to a category using the Euromoney risk index. Following the approach presented by Erramilli and Rao (1993) country risk is captured by a dummy variable, which takes the value of 1 when entry is into high-risk countries4
6. METHODOLOGY
In this chapter, the data collection method is discussed and the composition as well as the selection of the data sample is clarified. In addition, an explanation of the data analysis used in this thesis is given.
6.1 Method of data collection
This thesis is based on secondary data, which is collected from different sources. Data collection takes place in 2010.
Data regarding the dependent variables entry and establishment mode choice as well as the independent variable international experience is gathered from Amadeus, a comprehensive, pan-European database containing financial information on over fourteen million companies (Bureau van Dijk, 2009). In addition, publicly disclosed information such as company websites and annual reports of the sample firms are consulted.
Thomson Reuters´ Datastream, a comprehensive, worldwide databases containing current and historical financial data as well as profile data on public companies provides information regarding the independent variable firm size for the time period 1981 to 2010 (University of Zurich, 2009). Company websites are consulted in order to complete missing data.
Data concerning the independent variable R&D intensity is collected from the OECD science,
technology and industry scoreboard (OECD, 2001b). This publication provides a classification of
manufacturing industries based on technology.
Geert Hofstede´s (2010) personal website provides a dimension data matrix, which exhibits the
cultural scores needed to determine cultural distance between Germany and each target country. The World Resources Institute (WRI, 2010), the Central Intelligence Agency (CIA, 2010) and the
United Nations (UN, 2010) provide data regarding the independent country-level variable market
growth.
Data concerning the independent variable economic development is gathered from the OECD website (OECD, 2010).
6.2 Data sample
The sample comprises FDI activities of 62 German based MNEs in various countries within the time period from 1978 to 2010. Below a detailed description of how the sample was constructed is given. In addition, descriptive features of the sample are presented.
6.2.1 Sample construction
First of all, this thesis focuses exclusively on German based manufacturing FDI with at least 5% ownership. Several researchers argue that manufacturing and service firms respond different to transaction cost stimuli (Anderson and Coughlan, 1987; Brouthers and Brouthers, 2003; Delios and Beamish, 1999; Gatignon and Anderson, 1988; Hennart, 1991; Klein et al., 1990). Erramilli and Rao (1993) conclude that it is necessary to modify traditional transaction cost economics to make it applicable to service firms. In order to avoid that results are biased, the sample is restricted to firms with four-digit Standard Industrial Classification (SIC) codes between 2000 and 3999.
While all countries presented by Deutsche Bundesbank (2010a) are considered as possible targets, the selected firms actually have activities in 51 of those countries. It can be argued that this study includes the most important receiver of German FDI, since those countries obtain more than 95% of German outward FDI.5
The contribution of each manufacturing industry to the overall German outward FDI is chosen as final criterion.6 Due to data availability and time restrictions it is impossible to include all German
based manufacturing MNEs in this study. Therefore, the aim is to create a sample that is representative for the population.
Taking the considerations presented above into account, a set of companies is selected from Amadeus. Using information provided by this database as well as company websites and annual reports, the final sample of foreign investments is created.
6.2.2 Descriptive statistics
While the initial sample consisted of 438 observations, 60 (~ 14%) of those cases had to be excluded due to data availability problems. It was especially difficult to obtain sufficient information concerning the independent variable firm size for the early investments, since Datastream only provides data for the selected companies starting from 1981. Hence, the final sample includes 378 investments made by 62 German based MNEs. While the firms in the sample
