How does general international experience and host-country experience moderate the relationship between cultural distance and institutional distance on equity entry mode choices in knowledge-
based FDIs? A Transaction Cost Economics Perspective
Amelia McCormack
21.08.2015 Email: amelia.mccormack123@gmail.com
University of Groningen: s2681870 Newcastle University: 091917760
Master Thesis International Business Management (Dual Award) Supervisor University of Groningen: A.A. J.H. van Hoorn Supervisor Newcastle University: Dr J. Rodriguez
Abstract
Current research shows that cultural and institutional distance and experience considerations have conflicting predictions regarding entry mode choice. This paper investigates whether general international experience and host-country experience moderate the relationship between institutional distance on entry mode choice and cultural distance on entry mode choice in differing and opposing ways using data of 142 US firms between 2004 and 2013.
Based on a transaction-costs perspective, this research aims to explain how firms choose between
‘hybrid’ (JV) and ‘hierarchical’ (acquisition) categories not only on an estimation of the cultural and institutional differences, but that this choice is moderated by the degree of general international experience or host country experience.
The results show that cultural and institutional distance increases the likelihood of a more hierarchical and controlling entry mode. Host-country experience has a negative moderating effect on the relationship between institutional distance and entry mode choice, whereas general international experience positively moderates the relationship between institutional distance and entry mode choice and positively moderates the relationship between cultural distance and entry mode choice. Host-country experience was found to have a positive but insignificant moderating effect on the relationship between cultural distance and entry mode choice. This suggests that distance and experience have distinct and complex interactions concerning entry mode choices, and that the guidance offered by the current literature needs to be amended with more detailed studies that examine the complex interplay between distance and experience.
Keywords: Entry Mode Choice, Acquisitions, Joint Venture, Host-country experience, General international experience, Cultural Distance, Institutional Distance, Transaction Costs.
Table of Contents
Abstract ... 2
List of Tables ... 5
List of Figures ... 5
Abbreviations ... 5
1. Introduction ... 6
1.1 Problem Statement and Research Objective ... 6
1.2 Research Question ... 11
1.3 Thesis Outline ... 11
2. Literature Review ... 12
2.1 Transaction Cost Economics ... 12
2.1.1 Knowledge-based FDI ... 14
2.2 Entry Mode Choice ... 16
2.2.1 Equity Modes ... 16
2.3 Environmental uncertainty caused by distance ... 18
2.4 International Experience ... 20
2.5 Hypotheses Development ... 21
2.6 Conceptual Framework ... 28
3. Data and Methodology... 30
3.1 Data ... 30
3.2 Variables and Method ... 32
3.3 Empirical Model ... 33
3.4 Dependent Variable ... 33
3.5 Independent Variables ... 34
3.5.1 Institutional Distance ... 34
3.5.2 Cultural Distance ... 35
3.5.3 Moderator Host-country experience ... 36
3.5.4 Moderator general international experience ... 36
3.6 Control Variables ... 37
3.7 Testing for Multicollinearity ... 39
4. Empirical Results ... 40
4.1 Descriptive statistics and correlations ... 40
4.2 Goodness-of-fit Model Test ... 40
4.3 Results ... 43
4.3.1 Cultural Distance ... 45
4.3.2 Institutional Distance ... 45
4.3.3 Host-country experience ... 45
4.3.4 General international experience ... 45
4.3.5 Moderating effect of Host-country experience on the relationship between CD and entry mode choice ... 46
4.3.6 Moderating effect of general international experience on the relationship between CD and entry mode choice... 47
4.3.7 Moderating effect of general international experience on the relationship between ID and entry mode choice ... 48
4.3.8 Moderating effect of Host-country experience on the relationship between ID and entry mode choice ... 49
4.4 Summary and discussion of results ... 50
5. Conclusions ... 55
5.1 Theoretical and research implications ... 55
5.2 Managerial implications ... 56
5.3 Limitations ... 57
5.4 Future Research ... 58
References ... 59
Appendices ... 75
Appendix 1 ... 75
Appendix 2 Zephyr database screenshot ... 76
Appendix 3 Normal Distribution ... 77
Appendix 4 Multicollinearity Check – Number of Iterations ... 78
Appendix 5 Correlations ... 79
Appendix 6 Chi-squared Model ... 80
List of Tables
Table 1 Sample composition………..32
Table 2 Industry Sectors ... 38
Table 3 Results of Generalised Autologistic Models ... 41
List of Figures Figure 1 CD-entry mode Conceptual Model ... 29
Figure 2 ID-entry mode Conceptual Model ... 29
Figure 3 Moderating effect of Host-country experience on CD-entry mode ... 47
Figure 4 Moderating effect of General international experience on CD-entry mode ... 48
Figure 5 Moderating effect of General international experience on ID-entry mode ... 49
Figure 6 Moderating effect of Host-country experience on ID-entry mode ... 50
Abbreviations CD Cultural Distance
GLOBE Global Leadership and Organisational Behaviour Effectiveness Project ID Institutional Distance
FDI Foreign Direct Investment JV Joint Venture
MNE Multinational Enterprise TCE Transaction Cost Economics WOS Wholly-owned subsidiary
1. Introduction
1.1 Problem Statement and Research Objective
Based on transaction cost economics (TCE), this study investigates the relationship between cultural distance (CD) and entry mode choice, and the relationship between institutional distance (ID) and entry mode choice. Cultural and institutional distances are identified in the literature as major causes of knowledge-related complications, which play a central role in entry mode choice (Henisz, 2000). TCE argues that a higher degree of uncertainty (e.g. due to a lack of knowledge) may result in a greater need for control to resolve contracting difficulties (Hennart, 1988: 368).
Thus, a higher distance may result in firms seeking more control over the distant partner.
Gatignon and Anderson (1988: 310) introduced alliance experience as a factor for consideration by arguing that due to agency costs (Jensen and Meckling, 1976), measurement and performance monitoring problems can have “considerable transaction cost ramifications” (Alchian and Demsetz, 1972) for a firm. Therefore more experienced firms may increase their investments and control of a subsidiary because their experience allows them to efficiently monitor and control it.
In contrast, a less experienced firm may reduce their investments and control due to a lack of knowledge (experience), which may increase transaction costs. On the other hand, according to Dekker and Abbeele (2010) and Duysters, Heimeriks, Lokshin, Meijer and Sabidussi (2012), alliance experience may reduce the need for control. In short, there are conflicting prescripts regarding entry mode choice based on the distance and experience considerations.
This paper focuses on four areas regarding foreign entry mode: 1) role of cultural distance (CD);
2) institutional distance (ID); 3) host-country experience and general international experience;
and 4) the correspondence of ID, CD and international experience.
An entry mode is the institutional arrangement that facilitates a firm to enter a foreign market (Root, 1987). Williamson (1991) identified that forms of governance are organised into either
‘hierarchical’ (acquisition), ‘market’ (contractual agreements, exporting) or ‘Hybrids’ (joint- ventures, JVs). Many studies deal with the hierarchical versus market governance mode but hierarchical and hybrid governance modes are often used interchangeably. This is remarkable because Kogut (1988) argued that hybrid and hierarchical governance modes differ in cost and
the power of incentives. A firm being acquired wants to mark up the value of its assets; after the acquisition, these incentives are reduced due to the loss of residual rights on output (Pisano, 1989). Another debate in TCE literature regarding governance mode (foreign entry mode) concerns the amount of control rights and investments a firm pursues with increasing uncertainty caused by ID or CD. Traditional TCE posits that increasing uncertainty increases appropriation hazards, which may increase monitoring costs, resulting in firms choosing the hierarchical mode (Oxley, 1999; Hill et al., 1990). However, post-acquisition costs of hierarchical modes may increase. For example, a larger CD or ID may increase integration costs, so firms may want to reduce investments in uncertain foreign markets (Kogut and Singh, 1988; Li, Boulding and Staelin, 2010). Similarly, Oxley (1999) argued the lowest-cost governance mode depended on the interaction between the monitoring cost and the post-acquisition integration cost; the choice between hybrid and acquisitions is unspecified.
Distance is not a uniform construct, as argued by Ghemawat (2001) and Hutzschenreuter, Kleindienst and Lange (2015), but a multidimensional phenomenon that can be measured in cultural, administrative, geographic and economic dimensions, for example. Based on Scott’s (1995) conceptualisation, this study finds support that institutions and the distance between them is not a uniform construct. This is reinforced by Xu, Pan and Beamish (2004) who argue that the three pillars of institutional differences (regulative, normative, and cultural-cognitive distances) are independent and have different effects. According to Xu and Shenkar (2002), the differences between the cognitive-cultural pillar relate to CD and the difference between the national regulative pillars relates to ID; these can be considered as two different distances (Arslan and Larimo, 2010). According to Xu and Shenkar (2002), CD influences the cooperation and the firm’s global strategy, ID influences the firm’s multi-domestic strategy. This study follows Xu and Shenkar (2002) and Xu and Beamish (2004) and assumes that the institutional (regulative) and cultural pillars are independent and have different effects on MNE, ID and CD. These are tested separately to look at the two independent distance effects on entry mode choice.
International business literature recognises that experience may be an important explanatory variable in explaining the CD and ID paradox (Padmanabhan and Cho, 2005). Sadly literature often considers experience as a uniform construct. Distinguishing between different types of
experience is important because only attainment of the relevant types of experience reduces entry mode monitoring costs (Pisano, 1990).
Distinguishing between different types of experience on organisational strategy has recently attracted academic attention. Most studies suggest general international experience and host- country experience complement each other (Claver and Quer, 2005; Davidson and McFetridge, 1985; Padmanabhan and Cho, 2005). Meyer (2009), however, states these two types of experience have differing effects on entry mode choice, depending on the institutional context.
Meyer argues that host-country experience has a negative effect on the entry mode in emerging countries but a positive effect in developed countries. In contrast, they state general international experience has a positive effect on entry mode in developed countries but a negative effect in emerging countries. Meyer’s findings are based on two different theoretical perspectives; he states that experience of the ‘competence building effect’, which he defines as ‘competence in line of business’, and experience of ‘partner selection’, which he defines as ‘building a relationship with a partner’. This conflict regarding theoretical perspectives makes the foundation of the article disputable.
Meyer’s (2009) findings appear to apply to some developed and emerging markets but not to others. For instance, he found that host-country experience only had a negative effect in China but not in other emerging economies, and general international experience only had a positive effect in Europe but not North America, hence the results are not consistent with the predictions.
Moreover, Meyer’s (2009) results may only be specific to the chosen population of firms in the Taiwanese electronics industry. Although his research is a very interesting and useful contribution to the international business literature, it creates more questions than it answers. His explanation of the deviating result, by arguing that China is a unique case, is less than satisfactory. Furthermore, the imbalance that Meyer (2009) creates in his model by attributing partner selection to country-specific experience but not to general international experience is questionable.
This thesis agrees with some of Meyer’s (2009) findings but posits that the differences between host-country experience and general international experience are based upon and provide
different transaction costs and monitoring costs that may result in differing and opposing effects on the relationship between distance and entry mode choice. The theoretical chapter will explain how different monitoring costs and governing requirements are related to a specific type of experience.
Padmanabhan and Cho (2005) state that decision-specific experience (prior experience with an entry mode) mitigates CD to a greater extent than host-country experience or general international experience. They define host-country experience as similar to general international experience ‘but narrower and more location-specific’ (2005: 313). However, Padmanabhan and Cho’s (2005) conclusions did not consider monitoring costs and regarded host-country experience (prior experience in the host country) and general international experience (prior experience in other foreign countries) to both have the same positive but insignificant effect on entry mode choice, whereby greater experience led to more control.
This study agrees with some of Padmanabhan and Cho’s (2005) findings but argues that host- country experience and general international experience significantly affect entry mode choice and provide different monitoring and governing requirements, which will be explained in more detail in the theoretical chapter. Padmanabhan and Cho (2005) argue that decision-specific experience is based on previous entry mode decisions but this argument does not differentiate between the kind of experience; in other words, decision-specific experience can be considered to overlap with host-country and general international experience. This may explain their inconclusive findings regarding those factors. The present study aims to investigate the specific moderating influence of host-country and general international experience on CD and ID on entry mode choice, thereby differentiating between the different forms of experience.
CD and ID are knowledge-related complications in an alliance between two firms (Johansen and Vahlne, 1977). These knowledge-related complications may increase when the goal of the alliance is knowledge-exchange (knowledge-based FDI). An example of a knowledge-based FDI was the NUMMI JV between General Motors and Toyota, in which General Motors needed the knowledge to build a compact car and Toyota needed a partner to enter the US due to trade
barriers and the knowledge to manufacture in there (Inkpen, 2008).
Knowledge-based FDIs differ from other FDIs because knowledge-based alliances are specifically created for the purpose of exchanging intangible assets (knowledge) (Jones, 2009:
18). The problem with exchanging knowledge is that it is difficult to value and is highly firm- specific (Song and Shin, 2008). This uncertainty and high asset specificity creates complications in both the transaction costs (monitoring costs), which would, according to TCE, result in a hierarchical governance structure (acquisition); post-acquisition integration costs would result in a hybrid or market governance structure.
Interestingly, most knowledge-based FDI research is case-based so there is a lack of specific generalisable empirical literature about the governance choice of knowledge-based FDIs.
Badaracco (1991) is one of the sparse contributions that specifically looked at US knowledge- based FDIs and argued that U.S firms are increasingly involved in such transactions. This thesis posits that by differentiating between different kinds of experience, the entry mode paradox might be solved.
This is the first thesis that addresses the gap of how specific differences between general international experience and host-country experience moderate the influence of ID on entry mode choice and CD on entry mode choice for knowledge-based US ventures through the perspective of transaction cost theory.
1.2 Research Question
“To what extent does host-country experience and general international experience moderate the relationship between cultural distance and entry mode choice and institutional distance and entry mode choice?’’
1.3 Thesis Outline
This thesis is structured as follows. Chapter 1 introduces the topic and contribution of this thesis to the existing literature. Chapter 2 provides a literature overview. Chapter 3 presents the hypothesis and conceptual framework. Chapter 4 presents the methodology, analysis and descriptive statistics. Chapter 5 reports and discusses the main research findings. Finally, Chapter 6 concludes the main results with regards to the literature review and provides limitations of the study.
2. Literature Review
2.1 Transaction Cost Economics
TCE literature focuses on three central concepts: asset specificity, behavioural uncertainty and frequency (Anderson and Gatignon, 1988). The TCE theory defines asset specificity as the dedication of resources to a transaction between two parties that is not easily redeployed in alternative uses (North, 1990). TCE presupposes that appropriation concerns in an alliance relationship are induced by asset specificity, which impacts the choice between high or low- ownership modes (Casciaro, 2003). Behavioural uncertainty arises from a firm’s inability to predict a partner’s possible opportunistic behaviour (Brouthers and Nakos, 2002). TCE posits that when there is greater uncertainty between alliance partner’s cooperation, the effect of asset specificity intensifies (Casciaro, 2003). Frequency represents the rate at which a firm procures the transaction of a good or service (Meyer et al., 2009). Frequency is significant in its interaction with entry mode because TCE posits repeated transactions with the same exchange partner will be more efficiently internally organised using a ‘hierarchical or hybrid mode’ (Williamson, 1985:
60; Krzeminska, 2009).
The five TCE central concepts (Meyer and Nguyen, 2005) are:
1. Opportunism (risk of the partner acting in their own interest);
2. Uncertainty of the environment (makes it difficult to predict transaction outcome);
3. Bounded rationality (of the firms knowledge for decision making);
4. Few partners to choose from;
5. Local know-how and assets held by the local partner.
TCE theory predicts that a firm’s decision to internalise transactions depends on the degree that transaction costs arising from the five central concepts are present. Moreover, TCE is based on assumptions that may not necessarily hold, such as the notion of efficient markets (North, 1990).
In practice, markets are not fully efficient, as governments impose different regulations, tariffs,
and taxes that may restrict ownership or increase difficulty for firms to define and enforce contracts, which may force firms to use sub-optimal entry modes (Erramilli and Rao, 1993).
TCE posits that firms choose a governance mode to minimise transaction costs (e.g. costs of monitoring and enforcement) of dealing with the other firm in the alliance (Houston and Johnson 2000; Rindfleisch and Heide 1997). Johansen and Vahlne (1977) state that CD and ID is a lack of knowledge of foreign markets, which is a significant cause of perceived uncertainty between alliance partners’ originating countries (Erramilli and Rao, 1993). TCE literature provides very little consensus regarding the uncertainty caused by ID or CD on entry mode. Some associate distance with JV (Makino and Delios, 1996) because greater uncertainty, caused by CD or ID, may lead to greater post-acquisition costs (e.g. integration costs) associated with high-control modes (Li, Boulding and Staelin, 2010) leading firms to choose shared-control modes. Other TCE theorists posit greater uncertainty, caused by a lack of knowledge of distant markets (Gaur and Lu, 2007); this may result in firms seeking a greater need for control to resolve contracting difficulties (Hennart 1988: 368) to reduce transaction costs of exchange with a foreign partner.
Moreover, in the presence of asset specificity, increases in uncertainty may lead to market governances, increasing the attractiveness of hierarchical modes (Williamson, 1985: 75) such as acquisition.
To contribute to these conflicting results regarding distance and governance mode, TCE theorists contribute that increased knowledge gained from experience may alter the uncertainty and the magnitude of different sets of costs associated with the governance mode (Li, Boulding, Staelin, 2010).
Partners aimed at foreign expansion often choose to run their alliance across multiple countries, which exposes the alliance to greater uncertainty as a result of unpredictable government policies and different consumer tastes, and so on (Li, Bouldin and Staelin, 2010). For alliances, such as shared-ownership modes, broad (general) international experience may increase information asymmetry between partners and difficulty of quality control (Zeithaml et al., 1985).
Additionally, shared-control modes facing a diverse range of unfamiliar markets must adopt multiple strategies to address multiple environments. Thus transaction costs are higher for
alliances in multiple countries than firms operating in only one country (Oxley, 1995).
Consequently, given the higher uncertainty associated with alliances in multiple markets, TCE theorists suggest firms with general international experience choose hierarchical governance modes rather than shared-control modes.
TCE literature suggests that prior experience with an alliance partner has an opposing effect on governance mode than general international experience. Repeated relationships with the same firm should enable a firm to develop relational ties, thereby reducing uncertainty and information asymmetries that influence transaction costs (Gulati et al., 2000) and governance mode (Heide, 1994), and reducing monitoring costs (Stump and Heide, 1996), resulting in reduced transaction costs for shared-control modes. On the other hand, a positive experience of prior alliances may make firms commit greater investments to foreign ventures to increase any expected pay-off (Li, Boulding and Staelin, 2010) using hierarchical modes.
2.1.1 Knowledge-based FDI
The literature argues that knowledge is a firm-specific asset (Deeds and Decarolis, 1999) that can provide the firm with a competitive advantage because it acts as an isolation mechanism to the competitors. However, due to the specificity of knowledge, which is the result of the interaction between organisational culture, technologies, and techniques (Bhatt, 2001), knowledge becomes context-dependent and thus difficult to trade (Bhatt, 2001; Tsoukas, 1996). In a knowledge exchange or alliance, the specificity of the knowledge resources is difficult to value and monitor for the alliance partner concerning performance and output (Moschandreas, 2000) and may act as a transaction cost increasing mechanism, because of this specificity.
Unlike other forms of FDI, knowledge-based FDIs encounter greater complexities in the international exchange process of intangible assets. The complexity of knowledge-exchange (Song and Shin, 2008) is increased by performing this knowledge exchange between partners with different cultural and institutional frameworks. The complexity is in both the exchange process and the product/service exchanged. As discussed in the introduction, TCE does not specify which governance mechanism it prescribes for knowledge-based FDI. Asset specificity
would result in a hierarchical governance mechanism (Reid et al., 2001) while environmental uncertainty would result in a market or hybrid governance mechanism (Kogut and Singh, 1988).
This makes knowledge-based FDIs an interesting and challenging alliance from a TCE perspective, and the TCE does not specify the most appropriate governance structure in knowledge-based FDI.
Knowledge is difficult to value and knowledge-transfer requires frequent communication and is very time-consuming to share (Kogut and Zander, 1993) because it is difficult for foreign firms to share a specific type of knowledge with another partner that provides them with a unique competitive advantage (Dussauge et al., 2000: 99).
Unlike other FDIs where entry mode follows the efforts of a firm to enhance their tangible resources and scale economies (Levitt, 1983), knowledge-based firms’ motivation for alliances (acquisition or JV) is to secure and augment intangible resources (knowledge) to sustain competitive advantage (Grant, 1991). This is particularly when we consider the competitive advantage (rents) dependent on the specialisation of resources (Amit and Schoemaker, 1993).
This specialisation acts as a double-edged sword; specialisation creates isolating mechanisms in relation to competitors but also to alliance partners, as argued by Amit and Schoemaker (1993), where firms’ resources are a trade-off between specialisation and robustness. Moreover, if this study considers an alliance partner to be selected because of its unique knowledge or resources, subsequently this study argues that both alliance partners could have different frames of
reference, and therefore exchanging knowledge-based recourses may be a challenge. These different frames of reference and the projected economic advantages of the alliance create a high level of interdependency between firms (Reid et al., 2001) that exchange a high level of asset specificity (Kostova and Zaheer, 1999). Moreover, division of labour between alliance partners and the high interdependency of knowledge-exchange processes (Reid et al., 2001) can lead to information asymmetry and thus potential problems of opportunism (Doherty, 1999).
2.2 Entry Mode Choice
2.2.1 Equity Modes
A ‘transaction’ – also known as an entry mode – is an entry into a different geographical market with a specific product/service (Pehrson, 2006). International entry modes are institutional arrangements for carrying out business transactions (Root, 1994). Prior research has typically classified entry modes using universal scales, e.g. low versus high control (Chen and Hu, 2002);
equity versus non-equity; or shared versus full-control (Datta et al., 2002). However, universal scales cannot reveal distinct differences between all entry modes.
Acquisition is the full-ownership of one firm by another (Bruner, 2004:11). JV involves ownership that shares resource commitments, decision-making and risk between two firms to create a new venture (Wolf, 2000:50). JV and acquisition are both equity modes (Parmigiani, 2003) that require direct investment (Jung, 2004).
The differences between ‘hybrids’, ‘hierarchical’ and ‘market’ entry modes can have vastly different outcomes on transaction costs and a firm’s degree of ownership and control (Dikova and Witteloostuijn, 2007) which explain the relevance of researching ‘hybrids’ and ‘hierarchical’
modes rather than ‘market’ modes for this investigation. ‘Hierarchy’ refers to internalising activities (full-ownership modes such as WOS or acquisitions) and ‘market’ refers to outsourcing from the market (contractual agreements or licenses) (Pan and Tse, 2000). Hybrids are transactions between two firms that share some of the characteristics of ‘market’ and
‘hierarchical’ arrangements in the same entry mode (JVs or alliance agreements) (Duarte and Garcia-Canal, 2004).
Previous literature assumes firms choose an equity entry mode to engage in FDI (rather than contractual agreements) to develop firm-specific assets. However, unlike other firm-specific assets, knowledge is an intangible specific resource (North, 1990). Ongoing transactions between knowledge-based alliances can be expected to have high reciprocal interdependency, which requires continuous coordination (Thompson, 1967), and high task complexity which requires a high level of information processing (Casciaro, 2003; Galbraith, 1977). Alliances with a high level of coordination and information processing imply that output and performance
specifications cannot be determined ex-ante. Non-equity (contract) based alliances are less applicable, because contracts can only be enforced when the output can be specified.
In other words, contractual agreements are not efficient for two parties that exchange knowledge- based products that cannot or can only be partially specified in advance (Maekelburger et al., 2012). For knowledge-based firms, contractual agreements become less optimal (Hamel, 1991).
For knowledge-based firms, the transfer of explicit knowledge via ‘markets’ is unlikely to be a source of competitive advantage because it is easily duplicated by other firms (Reid et al., 2001).
Knowledge-transfer between firms requires proximity by establishing a foreign affiliate in another country (Chung and Yeaple, 2008). Hence, knowledge-based alliance partners tend to organise ventures that require ongoing transactions, frequent communication and task interdependence through hierarchies or JVs (hybrids) rather than contractual modes (Pisano et al., 1989). Hybrids and hierarchical modes are thus the best suited governance modes for this investigation.
TCE theory posits the aim of a chosen entry mode is to reduce transaction costs (Makino and Neupert, 2000; Williamson, 1996). However, JVs and acquisitions incur different types of transaction costs, control, learning and resource commitments that provide advantages and disadvantages to entering different markets, making the choice between the two an important one (Brothers and Hennart, 2007; Meyer et al., 2009).
A TCE perspective frequently posits that firms’ choice of entry mode is dependent on the level of asset specificity. TCE theory commonly argues that when assets are highly specific, firms choose acquisition to reduce the transaction costs incurred by risks of opportunism by the local partner (Agrawal and Ramaswami, 1992) and to secure control of profit returns (Anderson and Gatignon, 1988; Kim and Hwang, 1990).
However, this finding only considers single transactions; in practice entry modes are not mutually exclusive. Repeated transactions in the host-country (Krzeminska, 2009) increases experience (Gulati, 1995) and familiarity with host-market idiosyncrasies, weakening the influence of asset specificity and the need for high-control hierarchical modes. Beamish and Chung (2005) found that familiarity between alliance partners and experience can loosen the need for high-ownership
modes. Experience in other countries can help develop efficient organisational practices which may be transferrable to other markets (Barkema and Vermeulen, 1998). Moreover, Yiu and Makino (2002) found JVs to be advantageous for firms making their first entry into culturally or institutionally distant markets, because a foreign partner can provide local know-how of the institutional environment (political and social norms) and of cultural idiosyncrasies. TCE posits that JVs provide an exploratory position in a foreign market without the high level of commitment (Casciaro, 2003) of acquisitions (Hill et al., 1990). However, disadvantages of JVs include possible control and coordination problems and transaction costs incurred by the risk of a local partner acting in self-interest (Jung, 2004).
This thesis addresses the apparent conflicting arguments in the literature regarding the separate complexities of CD and ID on the entry mode decision between acquisition (hierarchies) and JVs (hybrids). Very few studies state that the complexities caused by distance on subsidiary control can be reduced (Beamish, et al., 2008) with experience (Tröhler and Barbu, 2012: 18). This thesis suggests that the inclusion of interaction effects between distance and firm experience would provide better explanatory power of entry mode choice than a stand-alone CD and ID measures.
2.3 Environmental uncertainty caused by distance
There are other forms of distance (e.g. physical) but this thesis focuses on CD and ID. CD has been referred to as a lack of knowledge of another culture (Christiansen, 2012:169) which is a major source of complexity (Hutzschenreuter and Voll, 2008). ID is a lack of knowledge of institutions surrounding transactions which constrains a firm’s access to market information (Khemakhem, 2010). Both ID and CD between the home and host-country are major sources of transaction costs (Brothers et al., 2008) as distance increases the costs of monitoring, communication and managing a dispersed network (Tan and Mahoney, 2006). Another reason CD and ID increase transaction costs is that unfamiliarity and a lack of networks in the host- country environment increase firm-specific costs, which firms try to overcome by adjusting their entry mode (Ramsey et al., 2013).
‘Knowledge’ is specific to the context and host-country in which it was developed. It is embedded in the local people and the cultural and institutional setting (Matarazzo and Resciniti,
2014). Complexity originates from a lack of knowledge of the interaction between multiple distance dimensions, so a focus on one dimension underestimates the complexity of the environment (Hutzschenreuter, 2015).
CD refers to differences in the host country’s normative environment that firms have to conform to (Yiu and Makino, 2002). Some associate CD with acquisitions because for firms that exchange a high level of asset specificity (knowledge-based firms) and enter unfamiliar markets, the costs of monitoring the alliance partner increase. The costs of using a partner therefore offset the benefits (Kostova and Zaheer, 1999), increasing firms’ choice of acquisition. Conversely, some associate CD with JVs (Ramsey et al., 2013) – a less costly method to enter a market for the first time. However, the majority of the evidence that focuses on knowledge-based firms points towards the former. An explanation for the conflicting results can be attributed to its methodology. Prior literature often calculated CD as the sum of the squared Euclidean distance of Hofstede’s (1980) dimensions. However, Hofstede’s (1980) CD index has received mixed statistical support in past studies and has been challenged recently for perceiving culture as a static concept (Jones, 2009: 18) with an outdated measure (Mutch, 2007). Some develop their own items to measure CD (e.g. Luo, 2002); the variety of items to capture CD makes it difficult to compare and draw conclusions from different studies (Hutzschenreuter, 2015). As recommended by Reus and Lamont (2009), this study uses GLOBE project indicators, along with Kogut and Singh’s formula, to improve its methodology.
The core argument for introducing ID is that CD does not entirely capture the complexity of cross-border activities (Hutzschenreuter, 2015). ID encompasses differences in the regulatory and normative pillars (Scott, 1995). The institutional environment provides the ‘formal rules’ which shape the host-market economy (North, 1990) and entry mode choice (Chen, 2003; Harzing, 2004). There is a lack in agreement on the conceptualisation of ID, which makes it difficult to compare studies, as the items used to capture ID differ greatly across many articles. Some use World Governance indicators (Gallego and Casillas, 2014) whereas others use the Regulatory Factor Economic Freedom Index (De Beule et al., 2014).
Based on TCE theory, in distant environments (caused by CD or ID) high-control modes are necessary because distance increases agency costs; however, the relationship may not be linear (Hutzschenreuter, 2015). Institutional and cultural institutions change relatively slowly (Markus and Kitayama, 1991).
2.4 International Experience
Experience is a specific type of knowledge (Luo and Peng, 1999), developed by the process of
‘learning-by-doing’ in organisations (Gao and Pan, 2010). Learning is an exchange of knowledge and ideas (Nummela, 2010: 246); ‘learning-by-doing’ is associated with a repeated activity/transaction (Collins et al., 2009).
Literature on the relationship between the uniform term experience and entry mode choice has received mixed, ambiguous results. A partial explanation for the conflicting arguments may be differences in the settings of the databases used (Padmanabhan and Cho, 2005).
Another explanation may be that experience is often considered a uniform term. For instance, Padmanabham and Cho (2005) suggest that the capabilities gained by host-country experience and general international experience are similar, with the same effect on entry mode choice.
Host-country experience is defined as years of experience within the host country (Padmanabhan and Cho, 2005), and access to networks and a unique type of local knowledge (Zhao et al., 2004).
Host-country experience has been also associated with acquisitions (Sanchez-Peinado et al., 2007), increasing a firm’s access to local know-how and reducing their reliance on a foreign partner. Alternatively, some studies associate host-country experience with JVs (Dow and Larimo, 2009). Based on TCE theory, control is a necessity. Repeated transactions increase familiarity between alliance partners and reduce the need to monitor and control foreign partners (Gulati, 1995), reducing transaction costs (e.g. monitoring costs) and governance control (Dyer and Singh, 1998). Moreover, Wu (2008) and Yu (1990) empirically show that host-country experience reduces the transaction costs of exchange with long-term foreign partners, making it easier for firms to assess foreign partners, strengthening relationships and facilitating knowledge- sharing. Moreover, Kostova and Zaheer (1999) contribute that experience in the host-country
allows firms to adapt their entry mode to the local idiosyncracies of the host-country environment.
‘General international experience’ is attained from frequent operations in multiple markets (Barkema and Vermeulen, 1998). Some associate high general international experience with acquisition (Zhao, Luh and Suh, 2004). Kostova and Zaheer (1999) posit that firms with multinational experience have greater bargaining power but face greater challenges adapting to the local environment of many different host-countries at a corporate-level. They therefore prefer negotiating the host-country environment by transferring their management practices to foreign subsidiaries via acquisition. Moreover, firms with general international experience that are not familiar with a host-country environment may not know a compatible foreign partner to share the risks of a venture with, and so are more likely to safeguard specific assets using acquisition.
Alternatively, some found that firms with experience had difficulty transferring management capabilities to unfamiliar foreign countries as it caused friction, increasing the preference for JVs (Hoang and Rothaermel, 2005). However, most research points towards the former scenario.
Host-country experience reduces monitoring costs, which is the opposite effect to general
international experience where it is difficult to adapt internationally applied processes to multiple environments (Dikova, 2012). This may lead to a uniform governance mechanism that is not always appropriate and a monitoring system (Brouthers and Nakos, 2002), increasing firms preference for acquisition.
2.5 Hypotheses Development Cultural distance
Firms that enter a new culturally distant market are likely to encounter different cognitive and cultural settings with unfamiliar customs and norms (Hutschenreuter and Voll, 2008). Knowledge about the local culture and customs is necessary to do business; high cultural distance may restrict a firm’s ability to do business in new markets (Hutzschenreuter and Voll, 2008) because a distant unfamiliar culture can cause interpersonal barriers (Cho and Li, 2004). Cultural distance has received conflicting evidence regarding the impact on entry mode choice. The conflicting
evidence shows that some theorists associate CD with acquisition (Beamish et al., 2008) whereas other theorists associate cultural distance with JVs (Madhoka, 1997; Chen and Hu, 2002).
Beamish et al. (2008) argue that environmental uncertainty caused by CD leads firms to choose high-control modes of acquisition to reduce transaction costs and information asymmetry. In addition, knowledge-based firms exchange a lot of highly-specific assets (Deeds and Decarolis, 1999). Due to this specificity, firms find it difficult to monitor and trade as it provides a competitive advantage (Tsoukas, 1996). Hutzschenreuter and Voll (2008) argued that environmental uncertainty (CD) makes it increasingly difficult to exchange a high level of asset specificity between a firm and its foreign partner (Gatignon and Anderson, 1988). For firms that are unfamiliar with a host-country, the greater the CD, the harder it may be for firms to adapt their practices to the new setting (Hutzschenreuter and Voll, 2008). Knowledge-based firms may thus accept the higher integration cost of acquisitions over JVs to reduce transaction costs (e.g.
agency costs) (Tihayani et al., 2005).
Other researchers associate increased CD with lesser-control modes such as JV, based on the assumption that firms choose JVs to reduce integration costs associated with high-control modes (e.g. acquisition) and accept the higher transaction costs (Gatignon and Anderson, 1988) to acquire local knowledge from the local partner (Brouthers and Brouthers, 2000; Chen and Hu, 2002).
Other points of view are acknowledged, but this thesis posits that to avoid agency difficulties and opportunism risks of working with a foreign partner in culturally distant markets, firms choose acquisition rather than JV.
H1: Cultural distance increases the likelihood of choosing acquisition rather than JV.
Institutional Distance
As argued by Kostova and Zaheer (1999), ID is a relevant entry mode driver which refers to differences in legal systems, laws and regulations between the host country and the firm (Harzing, 2004), which raises external uncertainties and challenges for the foreign firm. Root (1994) explains that high ID between home and host countries restricts knowledge transfer between parties, because a lack of knowledge of the institutional environment can lead to conflict
between the firm’s organisational practices and the host-country's institutions (Brouthers et al., 2008).
Regarding entry mode, ID has received ambiguous and conflicting results (Hutzschereunter et al., 2015). Some theorists found that in high ID between home and host countries, where the institutional environment is relatively weak or underdeveloped, firms cannot rely on the institutional environment with inefficient labour, capital and product markets, ambiguous property rights and unfamiliar legal contexts to protect specific assets. This may induce a fear a loss of intellectual property or opportunism of a foreign partner (Reid et al., 2001), enhancing a need for greater control of operations; firms therefore prefer acquisitions (Elango et al., 2013;
Contractor et al., 2014). Also, acquisition as opposed to JV, provides greater control and power to the acquirer’s executives in devising necessary organisational changes in the interest of overall success and acquisition integration (Elango et al., 2013). Furthermore, foreign firms may find it too demanding to adapt to external pressures (required by ID markets) and internal pressures (required by JV), resulting in firms choosing acquisition to avoid management’s attention being diverted from their strategic and operating goals (Hitt et al., 1996).
Other studies argue that high ID increases the likelihood of firms choosing JVs (Kostova, 1999) to reduce institutional conflicts between firms (Xu, Pan and Beamish; 2004), reinforce legitimacy among local government in the host country environment (Alon, 2003) and to capitalise on locally embedded resources (Yan and Gray, 1994).
For the foreign firm’s home-country, this thesis looks from a developed economy with an efficient institutional system (US). A high ID between home and host country suggests that the partner firm in the host-country is in a weaker institutional system. In the context of knowledge- based firms, this thesis posits that in the face of enhanced challenges and uncertainty caused by institutional differences, firms are more likely to choose acquisitions over JVs because complete ownership provides greater control over foreign subsidiaries.
H2: Institutional Distance increases the likelihood of choosing acquisition over JV.
Moderating effect of general international experience and host-country experience
The independent role of distance on entry mode choice is well documented. However, only sparse attention is given to the effects of experience, especially the different effects of host- country and general international experience. This thesis argues that different types of experience (general international experience and host-country experience) moderate the influence between the separate effects of CD and ID on entry mode choice.
General international experience may have an opposing effect to host-country experience on entry mode choice in institutionally or culturally distant environments (CD and ID) as it provides different capabilities and transaction costs than host-country experience (Murray, 2009). Host- country experience facilitates knowledge-sharing (Wu, 2008), gained through continuous exchanges with alliance partners within a specific host-country (Casciaro, 2003; Kostova and Zaheer, 1999). Host-country experience provides detailed and specific knowledge of the host- country business environment (cultural and institutional environment) and builds networking relationships (Luo, 1999) that can allow a firm to adapt to local host-country idiosyncrasies and recognise opportunities in a host-country (Johansen and Vahlne, 1977). This unique knowledge is highly asset-specific (Yu, 1990) and difficult to transfer across borders (Meyer and Li, 2008) so can only be obtained through host-country experience (Pattnaik and Lee, 2014). Hence, host- country experience may reduce monitoring costs and allows a firm to overcome difficulties from CD and ID.
‘General international experience’ is gathered from repeated long-term operations in multiple markets (Barkema and Vermeulen, 1998). General international experience provides a more general knowledge that enables firms to transfer organisational routines to other countries (Dikova, 2012). It transfers uniform managerial processes which are accepted in multiple countries but cannot exploit fine-grained differences in specific host-countries or adapt to complex and specific host-country environments (Kostova and Zaheer, 1999). Hence, CD and ID can increase transaction costs for firms with general international experience that are unfamiliar with host-country idiosyncrasies (Rohn, 2009: 46). Thus, general experience may reinforce the agency effect because the focal firm is unable to understand subtle but essential differences
between the host country and other countries’ cultural and institutional systems (Peng et al., 2008). Moreover, general international experience facilitates capabilities to compete in host- markets that are substantially different from the home-market (Delios and Henisz, 2000). Thus, firms learn how to operate in foreign countries, which enables them to take governance forms with greater levels of ownership (Chang and Rosenweig, 2001).
The literature lacks consensus regarding the influence of host-country experience on the relationship between ID and CD and entry mode choice. Traditional TCE theory suggests that experience lowers the costs of operating high-control modes (Anderson and Gatignon, 1986) so firms become more willing to use high-control modes to increase expected returns (Sanchez- Peinado et al., 2007). However, Dikova (2012) found that even for experienced firms, acquisition can create conflicts with host institutions (Stopford and Wells, 1972). Also, for firms that are inexperienced with specific host-country idiosyncrasies (e.g. due to CD or ID), transaction costs may increase because the foreign market’s environmental uncertainty makes it difficult to monitor agents and collaborate with foreign partners (Rohn, 2009: 46).
However, regarding TCE, Chang and Rosenweig (2001) found that a build-up of host-country experience reduces perceived CD and uncertainty of the institutional environment (Blomstermo et al., 2006) because repeated transactions with an exchange partner reduce the incentive for partner opportunism and free-riding (Dyer, 1997) for fear of losing future business. In turn, the transaction costs of sharing risks with a local partner decrease, increasing the attractiveness of JV (Padmanabhan and Cho, 1999). Furthermore many theorists posit that host-country knowledge can overcome the complexities of CD (Yu, 1990) and ID (Dikova, 2012; Goerzen, 2007), lowering the need for control and increasing the appeal of JVs.
Firms from different countries differ in capabilities, such that strategic objectives may not be compatible (Child and Faulkner, 1998). The divergence of partner objectives may lead to coordination problems that can undermine the suitability of a JV. However, a firm with prior experience in a host-country may have already adapted to the specific host-country context. Host- country experience also supports a firm’s ability to identify and select suitable local partners (Harrigan, 1988) and negotiate a contract suitable for the local environment, which may help overcome coordination problems in a JV (Murray, 2009).
Although other points of view are acknowledged, following Beamish et al. (2008) and Newbert (2007), this thesis posits that knowledge gained from host-country experience enables firms to overcome the complexities caused by CD and ID, lowering transaction costs associated with bargaining and monitoring a foreign partner in shared-control modes (Barney and Hansen, 1994) compared to high-control modes (e.g. acquisition). Hence, host-country experience has a negative moderating role on the influence of CD on entry mode choice and a negative moderating role on the influence of ID on entry mode choice, reducing the preference for acquisition.
H3a: Host-country experience negatively moderates the relationship between cultural distance and choice of acquisition.
H3b: Host-country experience negatively moderates the relationship between institutional distance and choice of acquisition.
The conclusion of many TCE studies suggests that ID and CD pose more of a barrier for inexperienced firms than experienced because inexperienced firms that face uncertainty may conflict with the cultural environment and local institution in a foreign country (Stopford and Wells, 1972). The initially high costs of cultural, political and institutional differences decrease with experience (Majoochi et al., 2015; Peng, 2000).
According to Erramilli (1991), firms accumulate international experience as they enter more foreign countries (Wager, 2004). Dikova (2012) states that the networks and knowledge accumulated from general international experience required to compete successfully in one country are not easily transferred and can differ greatly to the capabilities required to compete in another (Delios and Heninz, 2000). Moreover, Estrin et al. (2009) argued that distance ‘erodes applicability’ of knowledge and organisational practices gathered from general international experience developed in one country to entry mode choice in other distant markets (Slangen and van Tulder, 2009). So general international experience may have limited applicability in institutionally or culturally dissimilar countries.
Luo (1999) posits that general international experience provides firms with different capabilities to host-country experience. Such firms would be more likely to choose high-ownership modes
because they are better positioned to recognise business opportunities (Delios and Beamish, 1999), have developed efficient organisational routines (Luo, 1999), can operate independently (Brouthers and Nakos, 2004) and can assess risk (Chang and Rosenweig, 2001), which may reduce the integration costs of high-control modes, such as acquisition (Hennart, 1991).
Some associate experience with JV (Barney and Hansen, 1994). However, MNEs with high levels of general international experience have greater negotiating power in the local environment during potential conflicts with suppliers and partners in uncertain markets (Barkema and Vermuelen, 1998). TCE argues that distance leads to greater information asymmetry and increases firms’ transaction costs (Chan and Cui, 2014). So for MNEs that operate in many distant and uncertain environments, the transaction costs of exchanges with a foreign partner increase greatly (e.g. human coordination and cooperation costs, monitoring costs) (Chwa and Jwa, 2002: 178). Hence, firms that have accumulated sufficient general international experience from many different business environments may choose acquisition (He, 2002) over JV.
Foreign entry mode is also determined by the degree of conformity to internal pressures (Ferreira and Suk, 2009). Internal pressures include existing organisational structures and practices that are developed from previous international experience (Ferreira and Suk, 2009). Tallman and Yip (2003) argue that for MNEs, absolute adaption of operations to a host country increases transaction costs. Hence, general international experience can lead to the application of best practices and routines to all subsidiaries across many countries (Dikova, 2012) to maintain internal synergies and reduce transaction costs. Thus, the drawback of general international experience is that firms that have entered multiple markets succumb to high internal pressures so cannot adapt to different institutional contexts (Meyer and Nguyen, 2005; Dikova, 2012), which may encumber internationalisation efforts.
Environmental uncertainty (CD or ID) severely encumbers knowledge-transfer because knowledge-based firms are unwilling to reveal and share their asset-specific knowledge or intellectual property (Deeds and Decarolis, 1999) with a foreign partner in uncertain environments, especially if they lack protected markets (Kogut, 1998). In addition, the specificity of the knowledge resources may act as a transaction cost-increasing mechanism. So, in uncertain markets for knowledge-based firms with general international experience, collaborative entry modes are not appropriate because JVs pose problems of opportunism and appropriation, so
transaction costs associated with monitoring and sharing assets or intellectual with foreign partners are significantly higher than costs of high-control modes (Klein et al., 1978; Dyer and Singh, 1998), thus knowledge-based firms choose acquisitions over JV.
Acknowledging other points of view, like Dikova (2012) and Pogrebnyakov (2008) this thesis argues that general international experience has an opposing effect to host-country experience on entry mode choice. General international experience is predicted to positively moderate the influence of CD on entry mode choice and positively moderate the effect of ID on entry mode choice, reinforcing the likelihood of acquisition.
H4a: General international experience positively moderates relationship between cultural distance and choice of acquisition.
H4b: General international experience positively moderates the relationship between institutional distance and choice of acquisition.
2.6 Conceptual Framework
Two separate conceptual models illustrate the independent variables CD and ID. Their interactions with moderator variables examined in this study were analysed separately. Figure 1 illustrates the hypothesised relationships between CD and entry mode, which are moderated by general international and host-country experience variables. Figure 2 illustrates the hypothesised relationships between ID and entry mode, which are moderated by general international and host- country experience variables.
It is predicted that in response to cultural distance to reduce internal uncertainty, firms choose acquisition (H1). Furthermore, it is argued that as knowledge, firm-specific capabilities and networks developed from host-country experience increase, firms’ preference for acquisition is reduced in CD markets (H3a). Whereas it is argued that the moderating effect of general international experience will reinforce the positive relationship between CD and choice of acquisition (H3b).
Figure 1 CD-entry mode conceptual model Source: Author’s construction
Figure 2 ID-entry mode conceptual model
It is predicted that in response to institutional distance, firms choose acquisition (H2). As host- country experience increases, firms’ preference for acquisition is reduced in CD markets (H3a).
On the other hand, knowledge gathered from general international experience increases the positive relationship between ID, reinforcing the choice of acquisition (H3b).
3. Data and Methodology
3.1 Data
Firm data, such as entry mode choice and year of initial and observed market entry, were collected from the Zephyr database of cross-border equity-deals made by US companies (home- country). The data consisted of partner firms from 16 host-countries. All international alliances data of US firms with non-US firms were extracted from the Zephyr database between 2004 and 2013. The following variables were extracted from the Zephyr database: name and country of foreign partner company, announcement date and ticker symbols. This study focused on foreign market entry-modes by knowledge-based firms. Only alliances (JVs or acquisitions) that were expressed in the deal comments field as having been initiated for knowledge-based reasons were chosen. An example screenshot of data extraction from the Zephyr database is presented in Appendix 2. Firm-specific (financial) data and host-country data were collected from the announcement date (year)– Brouthers and Hennart (2007) argue that the most effective method for investigating entry mode choice is at the time they occur.
The sample only included public companies that submit annual accounts, thus facilitating complete records for proper data collection. Firm data, such as number of employees and organisational slack, were collected from annual reports reported on the company websites for the year of the announcement of the alliance. GDP per capita and market growth data were collected from the OECD database (OECD.org) from the announcement date of each specific alliance. Only firms from one home-country were chosen to make the research context for all firms’ contracting costs as comparable as possible and avoid the effect of exogenous factors.
Hence, following other studies (Paul and Wooster, 2008; Kuo et al., 2012), a single host-country was chosen (US) to isolate the influence of host-country differences.
The Zephyr database has comprehensive entry mode data that contains acquisition and JV data (Bureau van Dijk, 2015). Only firms that reported all of the necessary data were included in the database. Multi-partner deals were converted into a dyadic 1-to-1 relationship. The sample size was limited to alliances made for knowledge-based reasons and firms that had made at least one
investment between 2004 and 2013. Host countries with missing CD data were excluded. The final dataset consisted of 142 deals converted into 1-to-1 dyadic observations between the US and 16 target-countries.
A cross-sectional analysis of different US firms provides a ‘snapshot’ of the differences between them (Bhagat. et al., 2011). As argued by Hofstede (2001: 3) culture changes slowly with a cycle time of approximately 500 to 5,000 years; institutional structures also suffer from a certain structural inertia (Hannan and Freeman, 1984) and, as argued by Johanson and Vahlne (2006), building international experience takes time. As independent variables change relatively slowly, cross-sectional analysis is an appropriate analysis for this study. A limitation of this study is that firms’ investments in foreign entries before the sample timeframe were not included.
The distribution of the host countries is depicted in Table 1. The sample shows that the majority (69%) of firms chose acquisition and the remaining (31%) JVs. The dataset includes both developed and developing countries; this therefore shows a representative sample of international investments. Approximately half of the dataset consists of developing countries so both the developed and developing countries are equally represented. The difference between developed and developing countries is accounted for by including the market size, market attractiveness and market growth.
Table 1 Sample composition Source: Author
Country Firm Sample
Size
% JV % Acquisition
Australia 3 0 100
Canada 8 0 100
China 23 48 52
Columbia 9 44 56
Finland 3 0 100
France 15 27 73
Germany 15 13 87
Hong Kong 6 17 83
India 11 45 55
Italy 9 33 67
Japan 8 38 62
Netherlands 3 0 100
Spain 7 43 57
South Korea 7 71 29
Singapore 4 25 75
UK 11 18 82
TOTAL 142 31% 69%
3.2 Variables and Method
Due to dyadic level of measurement of the dependent variable (entry mode) this study uses a binary logistic regression analysis which is executed via the generalised linear models application in the SPSS 21 statistic analytics program.Unlike other statistical techniques, logistic regression analysis can be used when the dependent measure is dichotomous (Hair et al., 1998) and it considers the direction of the variable effects (Hosmer and Lemeshow, 2000). In the regression test, the p-value was examined to determine the significance and explanatory power of the variables on the outcome (Lehmann et al., 2002). A default cut-off of 5% was adopted for variables entering the model (Hennart, 1991). χ2 tested the independence of the variables, as it has been specifically advocated measure for entry-mode studies (Ekeledo and Sivakumar, 2004).
Positive regression (β) coefficients of variables indicate a higher chance for acquisition rather than JV. A negative coefficient sign suggest a higher chance for JV rather than acquisition.
3.3 Empirical Model
Model CD-entry mode
Y(entry US)=β0+β1X1(CD)+β3X3(gen.intl.exp)+β4X4(host-country.exp)+β5X5(gen.intl.exp*CD)+β7X7(host-
country.exp*CD)+Controls.j(β9Industry.sector+β10Firm.size+β11Available.Slack+β12Absorbed.Slack+β13Market.size+β14Market.Gr owth+β15Localfirmdependence+β16Market.Attractiveness)+e
Model ID-entry mode
Y(entry US)=β0+β2X2(ID)+β3X3(gen.intl.exp)+β4X4(host-country.exp)+β6X6(gen.intl.exp*ID)+β8X8(host-
country.exp*ID)+Controls.j(β9Industry.sector+β10Firmsize+β11Available.Slack+β12Absorbed.Slack+β13Market.size+β14Market.Gro wth+β15Local.firm.dependence+β16Market.Attractiveness)+e
As there are two separate analyses of the CD and ID variables, two separate regression equations are proposed. The hypotheses will be tested with a logistic regression model, which assumes a common slope parameter for the predictor variable. The dependent variable entry mode choice is coded as a dummy variable 0 for JV or 1 for acquisition represented by Y in both equations. β1, β2, coefficients for the linear combination of the independent variables CD (X1), ID (X2), (X3) represents moderator general international experience; (X4) is moderator host-country experience; (X5) represents moderator general international experience on CD; (X6) is moderator general international experience on ID; (X7) is moderator host-country experience on CD; (X8) is host-country experience effect on ID and (B0) is the intercept term; (e) represents the error term;
and (controls J) represents the control variables. The models express the likelihood that general international experience, host country experience, CD and ID influence entry mode choice.
3.4 Dependent Variable
This study acknowledges control as a crucial entry mode factor. Therefore, to assure the comparability of the results (following Canabel and White, 2008) this research focused on the level of control differences between two entry modes: acquisition (1) and JV (0). To make a clear
distinction between a JV and acquisition, a dichotomous variable was used. Pan and Tse (2000) found that insignificant variables within a differentiated classification scheme are entry-mode predictors with a dichotomous entry-mode variable.
3.5 Independent Variables
3.5.1 Institutional Distance
The hypotheses distinguish two aspects of distance that represent formal and informal institutions of the country (Scott, 1995). Formal ID concerns laws and rules that influence business strategies and operations. Following Estrin et al. (2009), the ‘Regulatory Factor’ of the Heritage Foundation Economic Freedom Index is applied to measure ID between US firms and host-countries.
Specifically, the sub-indexes are: business freedom, trade freedom, fiscal freedom, government size, monetary freedom, investment freedom, financial freedom, property rights, freedom from corruption, and labour freedom. For similar reasons to Estrin et al., (2009) this thesis found the
‘Economic Freedom Index’ to be the most appropriate ID measure, because it covers a broad range of countries and formal institutions that allow the entry firm to freely operate a business in the targeted country. As institutional systems change relatively slowly, this study calculated the institutional index of each country for the reference year 2013, and computed the formal ID as the absolute value of the difference between the measures of the home and host country.
Similar to Kogut and Singh (1988), the Euclidean distance of these dimensions was applied to reflect the institutional differences of country risk power between home and host countries. The Kogut and Singh (1988) distance measure, despite some critical evaluations, remains the most used and accepted. Building on Håkanson and Ambos’ (2010: 207) idea that a five-dimension distance calculation was preferable to a four-dimension calculation, this study included a large number of institutional variables to increase the explanatory power of the distance calculation.
ID is calculated as:
