FOREIGN PARTNER FIRM CHARACTERISTICS,
UNCERTAINTY AND THE OWNERSHIP DISTRIBUTION IN
IJVs
Author: Claudio Russo Student number: S1909487 Email: c.russo@student.rug.nl
Degree courses: MSc. International Business & Management and
MSc. Business Administration – Organizational & Management Control First supervisor: dr. M.M. Wilhelm
Co-assessor: prof. dr. J. van der Meer-Kooistra
January 2016 Word count: 15.057
Abstract
Within IJV research a lot of attention has been paid to IJVs by contrasting them to alternative modes of entry. The study of how IJV parents design their ownership strategy to cope with different sources of uncertainty is a far less studied aspect however. In this research TCE and RO are used complementary to investigate how IJV parents design their ownership strategies in response to several sources of uncertainty. Moreover the boundary conditions under which RO predictions hold are explored by investigating the moderating effects of prior international alliance experience and managerial risk preference on the relationship between uncertainty and ownership share. The results show almost no support for the TCE perspective while some support is found for RO predictions. Moreover, the results pointed to the importance of experience as moderating factor on the relationship between uncertainty and the foreign partner ownership share in IJVs. As a result, this thesis provided several directions for further research to further explore RO predictions and their boundaries, and the circumstances under which they hold.
2 ACKNOWLEDGEMENTS
The writing of this Thesis was the final step I had to take to graduate both my masters: Business Administration: Organizational & Management Control; and International Business &
Management. Thereby an end has come to an amazingly instructive period which formed me to the person I am now. First of all I want to thank my first supervisor Ms. Wilhelm who provided me of a lot of valuable feedback, support and criticism during the writing of this thesis. Next, I also want to express my gratitude to Ms. Van der Meer-Kooistra for her constructive feedback on my work. Their feedback really helped me to improve my work and achieve a better end result. Moreover I want to thank Ms. Bos for granting access to the SDC Platinum database which played a central role in the execution of my research. Last of all, I want to thank all other people who supported me throughout the writing of this Thesis.
3 TABLE OF CONTENTS
Section Title Page
1. Introduction 4 - 7
2. Literature review 8 - 23
2.1 Control in IJVs 8 - 10
2.2 Uncertainty and IJVs 11 - 14
2.3 Foreign partner firm characteristics 15 - 16
2.4 Hypotheses development 17 - 23 2.5 Conceptual model 23 3. Methodology 24 - 28 3.1 Research method 24 3.2 Sample characteristics 24 - 25 3.3 Dependent variable 25 3.4 Independent variables 25 - 27 3.5 Control variables 27 - 28 3.6 Statistical method 28 - 29 4. Results 30 - 34 5. Discussion 35 - 40
6. Limitations and further research 41 - 42
7. Conclusions 43 - 44
References 45 - 51
Appendices 52 – 58
Appendix A – Overview: home countries of foreign partners 52 – 53 in the dataset
Appendix B - Overview: SIC industries frequencies in dataset 54 – 55
Appendix C - Collinearity statistics 56
Appendix D - Assessment of data distribution 57
4 1. INTRODUCTION
As competition becomes more global and emerging market economies are rising, multinational firms increasingly face highly uncertain environments in which they do business (Cuypers & Martin, 2009; Inkpen & Beamish, 1997). While there are a range of entry modes and other risk-handling initiatives, one that has received much attention is the use of international joint ventures (Child & Yan, 2003; Beamish & Lupton, 2009; Lu & Hébert, 2005). It is argued that IJVs in particular are subject to substantial uncertainty (Cuypers & Martin, 2009; Reuer & Tong, 2005). Moreover, this uncertainty is likely to be magnified in an international setting with parents from culturally different backgrounds (Reuer & Tong, 2005). Especially firms that enter emerging and transition economies have to cope with high levels of uncertainty, as such countries often have different political and technological systems as well as different economic and social challenges compared to advanced markets (Brouthers and Dikova, 2010; Brouthers et al., 2008; Cuypers and Martin, 2009). Therefore, IJVs are really suitable to investigate how JVs cope with various and powerful sources of uncertainty.
Although an IJV mode of entry itself is a way to deal with uncertainty, as equity IJVs align the objectives of both IJV partners (Gulati, 1995), it is suggested that uncertainty also determines how much ownership firms seek over their foreign JVs (Aharoni, 1966; Cuypers and Martin, 2009). Selecting the appropriate equity ownership level to enter an IJV is a very important strategic decision that the managers involved in the IJV face (Richards, Yang, 2007). Lu & Hébert (2005) found for example, that the fit between the initial conditions for the IJV and the equity distribution influences the performance of IJVs. Control in IJVs refers to the ability to exercise authority and influence over the IJVs strategic and operational decisions, systems and methods (Anderson & Gatignon, 1986), with equity stakes being regarded as a central element in parent control (Geringer and Hébert, 1989). Control is not always a strict consequence of a parent’s equity position in the IJV and ownership is only one of the many control mechanisms available to firms (Sohn, 1994). Still evidence has indicated that firms do rely extensively on ownership control and relating voting control in IJVs, especially for foreign investors in developing or emerging country IJVs (Luo et al., 2001; Lu & Hébert, 2005). Therefore, this research will focus solely on ownership control.
5
uncertainty with rising transaction costs and potential losses and therefore suggests strategies (e.g. using high equity ownership) to minimize these transaction costs. Although TCE is primarily focused on partner opportunism, asset specificity and uncertainty, it is argued that especially in emerging and developing countries, institutions and other exogenous sources of uncertainty are influencing the equity distribution in IJVs (Cuypers & Martin, 2007; Murrell, 2005). Therefore TCE also included concepts of institutional and cultural uncertainty into their models (Brouthers, 2002; Gatignon & Anderson, 1988; Kogut & Singh, 1988). However, in TCE theory, exogenous uncertainties (e.g. institutional uncertainty) act as conditional factors that magnify information asymmetry and lead to higher transaction costs (Ahsan & Musteen 2011; Cuypers & Martin, 2007). Moreover, the few studies that examined JVs ownership distribution yielded mixed results regarding uncertainty, especially regarding exogenous sources of uncertainty (Delios & Beamish, 1999; Chen, Hu, Hu, 2002; Gatignon & Anderson, 1988).
Therefore, real option (RO) theory has emerged as an important alternative approach to understand strategy while facing uncertainty (Cuypers & Martin, 2007; Kogut, 1991; Li & Li, 2010; Reuer & Tong, 2005). RO theory proposes a different view compared to TCE by suggesting that uncertainty implies possibilities as well as risks to the firm (Li, 2007). More specifically, it suggests that MNEs may want to remain flexible when facing uncertainty, by attaining a low ownership share and keep the options to defer or grow when new information becomes available (Li, 2007). The RO literature generally distinguishes between endogenous sources of uncertainty and exogenous sources of uncertainty of which endogenous uncertainty can be diminished (at least partially) by a firms own actions, as it is dependent upon firms’ learning capabilities (Chi, 2000; Folta, 1998). Exogenous sources of uncertainties’ resolution is unaffected by the firms actions and can resolve only through passive observation of the host country environment, which makes the resolution independent of the investment activity (Ahsan & Musteen, 2011; Cuypers & Martin, 2007). In their attempts to explore and refine the boundaries of RO theory, scholars found that RO predictions with regard to the ownership distribution in IJVs in relation to different sources of uncertainty only hold for exogenous sources of uncertainty (Cuypers & Martin, 2009; Miller & Folta, 2002).
6
2002) while TCE proved to contribute to the analysis of the effect of endogenous uncertainty on the ownership distribution (Gatignon & Anderson, 1988; Richards & Yang, 2007). Therefore, both theories hold great promise as complements to study the influence of both endogenous-, and exogenous sources of uncertainty (Cuypers & Martin, 2007).
This study, which focuses only on two parent firm IJVs, will contribute to the literature in a number of ways. First of all, it will be the first research that uses TCE and RO theory complementary to study the influence of both endogenous- and exogenous sources of uncertainty on the foreign parent ownership share in the IJV. Furthermore, it extends the few RO studies (Cuypers & Martin, 2009; Li & Li; 2010) that empirically investigated the influence of difference sources of uncertainty on the foreign parent ownership share in IJVs in the context of China as a host country. Therefore scholars argued that further empirical research on RO predictions is needed (Ahsan & Musteen, 2011; Cuypers & Martin, 2009; Li & Li, 2010; Reuer & Tong, 2005). Second, and more important, so far RO research has failed to consider how parent firm characteristics may influence the relationship between uncertainty and the initial ownership distribution in IJVs (Ahsan & Musteen, 2011; Li & Li, 2010). However, in the literature it is suggested that foreign parents’ prior alliance experience (Ahsan & Musteen, 2011, Brouthers et al., 2008; Fisch, 2008) and their managerial risk preference (Bowman & Hurry, 1993; Cuypers & Martin, 2007) might have an effect on the ownership distribution in the IJV. Therefore this study will incorporate these foreign parent firm characteristics to examine how they might influence MNEs design of their ownership strategies in their IJVs in response to uncertainty. By adding these possibly moderating forces, this research answers the call of many scholars who ask for further research that explores the boundary conditions under which the RO logic holds (Ahsan & Musteen, 2011; Cuypers & Martin, 2009; Li, Boulding,& Staeling, 2010; Li & Li, 2010; Li, Li,& Rugman, 2013; Richards & Yang, 2007). This study will be the first RO research that empirically investigates if and how foreign partner firm characteristics have an influence on how foreign partner firms design their ownership strategies in their IJVs when facing uncertainty. Therefore, the following research question will be addressed:
7
8 2. LITERATURE REVIEW
2.1 Control in IJVs
An international joint venture is an alliance that combines resources from more than one organization to create a new organizational entity, which is distinct from its parents which are of different nationalities (Inkpen & Beamish, 1997). IJVs are exposed to a great amount of uncertainty, because many IJV partners must monitor operations in settings which are unfamiliar to them (e.g. markets, distribution systems, legal systems); they have to cope with significant geographical separation and time differences; and they have to bridge cultural boundaries (Groot & Merchant, 2000). Furthermore, because IJVs are owned by different independent parent companies, control issues involve additional complexities (Kamminga & Van der Meer-Kooistra, 2007).
9
more hierarchical control mechanisms (e.g. equity control) in IJVs, as trust counteracts the fear of opportunistic behavior of the partner (Gulati, 1995). Although the value of other control mechanisms than equity control should not be underestimated, evidence has indicated that firms do rely extensively on ownership control and the relating voting control in IJVs, especially for foreign investors in developing or emerging country IJVs (Luo et al., 2001; Lu & Hébert, 2005; Wong, Luk & Li, 2005).
10
MNEs facing external uncertainty would be better off with a lower equity stake as it provides flexibility because they commit fewer resources to the IJV. Pan (1996) found similar results regarding the impact of external uncertainty. Therefore, the ownership share distribution in an IJV should correspond to the type(s) of uncertainty that the IJV is facing (Sartor & Beamish, 2014). The pros and cons of the three mentioned options are shown in table 1.
Table 1. Pros and cons of minority, equal, and majority ownership shares in IJVs (Richards & Yang, 2007).
Majority ownership Equal ownership Minority ownership
Pros More control over the activities to Share costs and risks with Cheaper access to
better serve the parent’ global partners. New technologies
strategies. compared to
acquisition.
More profit appropriation. More efficient activity due Less resource to synergy of expertise. commitment.
More likely to prevent partners Equal profit appropriation. Less sensitive to
from stealing technological secrets environmental
during R&D collaborations. uncertainty.
Less sensitive to behavioral uncertainty.
Cons More costs arising from specific resource Inefficient decision making. Less control over
commitment. R&D activities.
More sensitive to environmental Sensitive to both, environmental More likelihood of
uncertainty. and behavior uncertainty. knowledge
appropriation.
More incentives for local partners Less profit
to steal secrets if it is one-time appropriation.
transaction.
11 2.2 Uncertainty and IJVs
This study focuses on how MNEs design their initial ownership strategy in IJVs while facing uncertainty. The term uncertainty, as used in organizational and strategic management theory refers to the unpredictability of environmental or organizational variables that impact a firm’s performance, or the inadequacy of information about these variables (Miller, 1992). More specifically, uncertainty can arise from shocks in exogenous environment of the firm, unforeseeable behavioral choices or actions of the IJV partner or employees, or a combination of both (Lessard, 1989). Miller (1992) distinguished between the following three overarching categories of uncertainty: general environmental uncertainty, industry uncertainties, and firm uncertainties. These overarching categories each encompass a number of uncertainty components. General environmental uncertainty includes political instability, government policy instability, macroeconomic uncertainties, social uncertainties and natural uncertainties (Miller, 1992). Next, industry uncertainty includes input market uncertainties, product market uncertainties and competitive uncertainties (Miller, 1992). Lastly, firm uncertainty includes behavioral uncertainty, operating uncertainty, liability uncertainty, R&D uncertainty, and credit uncertainty (Miller, 1992). As discussed before, organizational responses to control these uncertainties are different, depending on the type of uncertainty, and different theoretical perspectives have opposing views.
In prior research, TCE has been one of the predominant theoretical lenses applied to study subsidiary control in international business research (Brock, Shenkar, Shoham, & Siscovick, 2008). The TCE theory posits that firms involved in an alliance select a governance structure and ownership level that minimizes the transaction costs associated with various aspects of dealing with other partner firms (Li, Boulding & Staelin; 2009). More specifically, TCE explains the use of parent control with asset specificity, opportunism and uncertainty (Williamson, 1985; Child et
al., 2005). As these factors increase transaction costs, firms seek to reduce them by internalizing
12
costs caused by opportunistic behavior of the partner (Ahsan & Musteen, 2011; Richards & Yang, 2007; Williamson, 1985). Transaction costs are mainly influenced by behavioral uncertainty as TCE views environmental sources of uncertainty as magnifying factors that increase information asymmetry and thus increase transactions costs. Therefore, uncertainty figures as an endogenous factor that can be addressed via governance decisions. In the TCE perspective, JVs are perceived on the middle of the continuum from markets to hierarchies, so JVs themselves are already a way to alleviate transaction costs (Crook et al., 2013). Regarding the ownership distribution this entails that the larger the ownership stake in the IJV is possessed by the foreign partner, the more the IJV shifts towards the side of a hierarchy.
13
14
environment disappears (Kogut, 1991). Notice the fundamental difference in proposed strategic response to uncertainty as RO theory suggests lowering the ownership share in response to uncertainty while TCE theory suggests taking larger ownership stakes in order to control the uncertainty.
In the RO literature uncertainties have been placed into two different categories: endogenous uncertainty and exogenous uncertainty (Chi, 2000; Cuypers & Martin, 2009). Endogenous sources of uncertainty can be diminished at least partially by a firms own actions, while exogenous sources of uncertainties resolution is unaffected by the firms actions and can be resolved only through passive observation of the host-country environment (Ahsan & Musteen, 2011; Cuypers & Martin, 2007). Previous research indicated that the RO argument of taking fewer ownership in the joint venture when facing uncertainty, in order to limit the downside risk of the investment, only holds for exogenous sources of uncertainty (Li, Boulding, & Staeling 2010; Li & Li, 2010; Cuypers & Martin, 2007; Cuypers & Martin, 2009). Exogenous uncertainty corresponds to the models of financial options as it is a necessity that the uncertainty is resolved independently of the investors’ behavior (Cuypers & Martin, 2007). However, when uncertainty can be resolved endogenously, MNEs will have an incentive to invest resources and increase control in order to resolve the uncertainty (Cuypers & Martin, 2007). Consequently, the ownership share decision in an IJV should correspond to the type(s) of uncertainty that the IJV is facing. As discussed before, RO theory has proved to be a promising starting point for the analysis of the influence of exogenous uncertainty (Cuypers & Martin, 2009; Li et al., 2010; Li & Li, 2010; Miller & Folta, 2002) while TCE proved to contribute to the analysis of the effect of endogenous uncertainty on the ownership distribution (Gatignon & Anderson, 1988; Richards & Yang, 2007). Therefore, both theories hold great promise as complements to study the influence of both endogenous- and exogenous sources of uncertainty (Cuypers & Martin, 2007). This study will use both theories complementary, as TCE reasoning will be applied to hypothesize how endogenous sources of uncertainty influence the IJVs ownership distribution and RO reasoning will be used to hypothesize how exogenous sources of uncertainty influence IJVs ownership distribution.
2.3 Foreign Partner Firm Characteristics
15
logic holds (Ahsan & Musteen, 2011; Cuypers & Martin, 2009; Li, Boulding,& Staeling, 2010; Li & Li, 2010; Li, Li,& Rugman, 2013; Richards & Yang, 2007). So far, RO research has failed to consider how parent firm characteristics may influence the relationship between uncertainty and the initial ownership distribution in IJVs (Ahsan & Musteen, 2011; Li & Li, 2010).
However, the broader entry mode literature suggests that MNEs prior international experience has an influence on MNEs entry mode choice (Barkema & Drogendijk, 2007; Li, Boulding, Staeling; 2010). It is suggested that prior international experience enhances MNEs capabilities to deal with local suppliers, customers, partners and governments (Herrmann & Datta, 2006) and teaches MNEs to deal more effectively with the uncertainties faced in new markets (Ahsan & Musteen, 2011). Furthermore, various scholars pointed to the influence of prior international alliance experience on particular entry modes (e.g. Li, Boulding, Staeling, 2010). The influence of international alliance experience on the value of the option and the initial ownership distribution in IJVs however remains unclear. As the RO literature failed to investigate the combined effects of prior international alliance experience and uncertainty on the ownership share distribution in the IJV, this paper will attempt to fill this gap.
Another limitation of RO theory is that it assumes that investors and/or managers are rational and objective, while in reality manager’s value their options based on their subjective perceptions of uncertainty (Bowman & Hurry, 1993; Cuypers & Martin, 2007). Consequently, it is important to pay attention to how the uncertainty is perceived by the decision maker. Multiple scholars have argued that managers may not be risk neutral, as they perceive risk differently and have a different risk preference (Brouthers, 2013; Hofstede, 1980). Till now, RO research has failed to investigate how managerial risk preferences influence the relationship between multiple sources of uncertainty and the initial ownership distribution in IJV (Brouthers et al., 2008). Therefore this research aims to address this gap by incorporating managerial risk preferences in the relationship between different sources of uncertainty and the ownership distribution in IJVs.
2.4 Hypotheses Development
16
Furthermore, previous studies argued that very few studies contrasted the influence of two or more sources of uncertainty and their interaction (Cuypers & Martin, 2009; Li, Li, & Rugman, 2013). Two general environmental sources will be included, one industry level source and one firm level source of uncertainty. Of the included sources of uncertainty one is suggested to resolve endogenously while the other three uncertainties are exogenous to the IJV. Next, I will hypothesize how foreign partner firm characteristics, more specifically, prior international alliance experience and managerial risk preference, will influence the relationship between uncertainty and the ownership distribution in the IJV.
Endogenous Source of Uncertainty
As discussed, endogenous sources of uncertainty can be diminished (at least partially) by a firm’s own actions. This research will incorporate behavioral uncertainty as endogenous source of uncertainty that is believed to be present within IJVs.
Behavioral uncertainty is the uncertainty that is related to a transaction partner’s behavior
(Richards & Yang, 2007; Sartor & Beamish, 2014). A rise in behavioral uncertainty will lead to increased transaction cost (Williamson, 1985). According to TCE theory increased transaction cost should lead to more control, and so, in a higher ownership stake in the IJV. Behavioral
uncertainty will mostly arise from the fear for opportunistic behavior of the alliance partner
(Richards, Yang, 2007). Therefore, a factor will be included into the study that increases
behavioral uncertainty on the part of the foreign partner in the IJV. The factor is whether or not
the foreign partner exclusively brings high-tech into the IJV, as joint ventures in which only the foreign partner brings in the high-tech, are likely to be subject to behavioral uncertainty.
17
an equity joint venture itself is a way to control uncertainty as it aligns the objectives of both IJV partners (Gulati, 1995). Because minority ownership in the IJV makes the partner more sensitive to behavioral uncertainty (Richards & Yang, 2007), the following hypothesis is drawn:
Hypotheses 1: In IJVs in which only the foreign partner brings in the high tech activities,
the foreign partner is more likely to take a larger share in the IJV. Exogenous Sources of Uncertainty
As discussed, exogenous uncertainties resolution is unaffected by the firms actions and therefore passive observation of the host-country environment, until the uncertainty is resolved, is the right strategic response (Ahsan & Musteen, 2011; Cuypers & Martin, 2007). This research will incorporate three important exogenous sources of uncertainty that are believed to have a strong presence in the environment of IJVs. The first source of uncertainty is market uncertainty which is an industry variable that refers to volatility of future profitability in industries. Next,
environmental uncertainty will be included, which is a country variable that refers to economic
and political risk in the host country (Anderson & Gatignon, 1986). Third, (the lack off) strength of the host countries intellectual property right protection, which is another environmental source of uncertainty, is included (Oxley, 1999).
18
Hypothesis 2: The foreign partner in the IJV is more likely to take a lower equity share in
the IJV when facing high market uncertainty in the industry.
When MNEs enter new host countries with their IJVs they also have to deal with the uncertainties specific to the external environment in the host country. Environmental uncertainty is defined as the perceived inability to predict the MNEs external environment (Sartor & Beamish, 2014). More specifically, environmental uncertainty can refer to economic and political risk in the host country (Anderson & Gatignon, 1986). In the literature the term country risk has been used to reflect the uncertainty over the continuation of present political and economic conditions (Agarwal & Ramaswami, 1992; Lu & Hébert, 2005). The extent to which country risk threatens the stability of a business operation is generally understood (Davidson, 1982). More specifically, political instability or changes as elections or wars can lead to major policy changes in the country which could have its impact on the MNE (Jung, 2004). Furthermore, macroeconomic instability such as inflation, high interest rates or a recession in the host country also has its impact on the MNE. Cuypers & Martin (2009) argued that for individual firms it is extremely difficult, or even impossible, to influence macroeconomic or political conditions in the host country, especially in larger countries. Because host country risk cannot be alleviated by the MNEs own actions, it is typified as exogenous uncertainty. Therefore I use RO reasoning which leads to the following hypothesis:
Hypothesis 3: The foreign partner in the IJV is more likely to take a lower equity share in
the IJV when facing high country risk in the host country.
19
secured over time (North, 1990). Many countries are signatories to the Paris Convention for the protection of Industrial Property, which requires that foreign nationals are granted the same intellectual property protection as the domestic citizens in a host country (Oxley, 1999). However, the convention does not specify a standard of protection and therefore the actual level of property rights protection varies significantly across the countries (Oxley, 1999). The literature suggests that if the property rights protection in the host country is limited, the MNE risks dissemination of knowledge and resources to local parties and potential rivals (Reuer & Tong, 2005). Because uncertainty stemming from host country’s weak property rights protection cannot be alleviated by MNEs own actions, intellectual property right protection can be typified as an exogenous source of uncertainty. Therefore I follow RO reasoning. The following hypothesis is drawn:
Hypothesis 4: The foreign partner in the IJV is more likely to take a lower ownership
share in the IJV when the intellectual property right protection in the host country is weak.
Foreign Partner Firm Characteristics
Although RO theory provided valuable insights into how MNEs design their ownership strategies in their IJVs in response to uncertainty, scholars argue that there is a high need for further research that explores the boundary conditions under which the RO logic holds (Ahsan & Musteen, 2011; Cuypers & Martin, 2009; Li, Boulding, & Staeling, 2010; Li & Li, 2010; Li, Li, & Rugman, 2013; Richards & Yang, 2007). One area that has received almost no attention within the RO literature is how foreign parent firm characteristics may magnify or alleviate the influence of different sources of uncertainty on how they design their ownership strategies in their IJVs (Ahsan & Musteen, 2011; Li & Li, 2010).
In the broader entry mode literature it is found that a firm’s prior international alliance
experience has an influence on MNEs entry mode choice (Barkema & Drogendijk, 2007;
20
21
Hypothesis 5: The positive relationship between behavioral uncertainty and the foreign
partner ownership share in the IJV as specified in hypothesis 1, is stronger for MNEs with less international alliance experience than for MNEs with more international alliance experience.
As discussed, RO reasoning will be applied on exogenous sources of uncertainty. According to RO theory, international alliance experience should provide MNEs with higher-value growth options. More specifically, RO theory suggests that firms with greater previous international
alliance experience perceive lower ‘risk of loss’, while in contrast MNEs with less international
experience seek downside risk protection (Brouthers et al., 2008). This leads to the following hypothesis:
Hypothesis 6: The negative relationship between market uncertainty and the foreign
partner ownership share in the IJV, as specified in hypothesis 2, is stronger for MNEs with less international alliance experience than for MNEs with more international alliance experience.
Hypothesis 7: The negative relationship between environmental uncertainty and the
foreign partner ownership share in the IJV, as specified in hypothesis 3, is stronger for MNEs with less international alliance experience than for MNEs with more international alliance experience.
Hypothesis 8: The negative relationship between a weak intellectual property right
protection in the host country and the foreign partner ownership share in the IJV, as specified in hypothesis 4, is stronger for MNEs with less international alliance experience than for MNEs with more international alliance experience.
22
decisions and actions will be filtered and distorted by their values and cognitive biases (March & Simon, 1958; Richards & Yang; 2007). Consequently, these values and cognitive biases will influence managers’ risk and uncertainty perception differently, even if they face exactly the same uncertain environment (Richards & Yang, 2007). In the literature it is suggested that national culture is one of the most important factors that determines ones risk perception and risk preference (Hofstede, 1980). More specifically Hofstede’s (1980) uncertainty avoidance dimension is an indication of a managers risk preference, from risk aversion to risk neutrality to risk seeking. Brouthers (2013) argued that future studies should address the issue of different risk preferences among managers and suggested that future studies should adjust their models. However, so far the RO literature has not shed its light on this issue. This research suggests that managers risk preferences moderate the relationship between uncertainty and the ownership distribution in IJVs. As discussed before, RO theory is found only to explain organizational responses to exogenous sources of uncertainty and consequently TCE reasoning will be used to hypothesize the expected moderating effect on the relation between endogenous sources of uncertainty and the foreign partner ownership share. More specifically, the higher the level of uncertainty avoidance in the home country of the foreign partner the less comfortable the managers will feel to take risk, and as a consequence they will take an even higher equity stake in the IJV when facing endogenous uncertainty. Therefore the following hypotheses are drawn:
Hypothesis 9: The positive relationship between behavioral uncertainty and the foreign
partner ownership share in the IJV, as specified in hypothesis 1, is stronger for MNEs with a high level of uncertainty avoidance in the home country than for MNEs with a low level of uncertainty avoidance in the home country.
23
Hypothesis 10: The negative relationship between market uncertainty and the foreign
partner ownership share in the IJV, as specified in hypothesis 2, is stronger for MNEs with a high level of uncertainty avoidance in the home country than for MNEs with a low level of uncertainty avoidance in the home country.
Hypothesis 11: The negative relationship between host country risk and the foreign
partner ownership share in the IJV, as specified in hypothesis 3, is stronger for MNEs with a high level of uncertainty avoidance in the home country than for MNEs with a low level of uncertainty avoidance in the home country.
Hypothesis 12: The negative relationship between a weak intellectual property right
protection in the host country and the foreign partner ownership share in the IJV, as specified in hypothesis 4, is stronger for MNEs with a high level of uncertainty avoidance in the home country than for MNEs with a low level of uncertainty avoidance in the home country.
2.5 Conceptual Model
24 3. METHODOLOGY
3.1 Research method
Although the determinants of the ownership distribution in IJVs are a relatively underexposed research subject within the IJV literature (Cuypers & Martin, 2009; Li & Li, 2010; Richards & Yang, 2007), it allows for a number of testable propositions. Therefore, this study aims to perform a quantitative research to test the proposed hypotheses. The main part of the data will come from the Securities Data Corporation (SDC) strategic alliance and joint ventures database, which is currently one of the most comprehensive sources of information on alliances (Li et al., 2010) and widely used in the literature (Li et al., 2010; Reuer & Tong, 2005; Reuer et al., 2008; Richards & Yang, 2007). In this section, the formation of the dataset, the sample characteristics and the variables will be discussed in more detail.
3.2 Sample characteristics
25
whether or not it is a high-tech IJV, whom of the partners bring in the high-tech, and more. These characteristics make the SDC database really suitable for this study. Furthermore, data on host
country risk and host countries intellectual property right protection are collected from the
Global Competitiveness Reports which are developed by the World Economic Forum.
3.3 Dependent variable Foreign partner ownership share
The dependent variable, foreign partners’ ownership share in the IJV is operationalized as the percentage of ownership of the foreign partner in the IJV. This information can be derived from the SDC database. This operationalization of ownership is general use within the literature (Cuypers & Martin, 2009; Li & Li, 2010; Richards & Yang, 2007). The foreign partner ownership in the sample is 49,06%. The smallest foreign partner ownership share in the sample is 15% and the largest foreign partner share in the sample is 90%. Next, as an alternative measure, foreign partners’ ownership share is also included in the study as a categorical measure. More specifically, minority shares of the foreign partner (<50%) will be assigned the value of 0; 50% shares will be assigned the value of 1; and majority shares of the foreign partner (>50%) will be assigned the value of 2. This categorical measure of foreign partner ownership is included as it may better reflect the amount of control (Chen et al., 2002).
3.4 Independent variables Behavioral uncertainty
26 Market uncertainty
As discussed market uncertainty is an industry wide variable that refers to unexpected changes in the demand for an industry’s output (Miller, 1992). More specifically it constitutes the volatility of future profitability in particular industries as it reflects unstable industry structures, unpredictable consumer taste, and evolving technological standards (Li et al., 2013). Consistent with Li et al., (2010), market uncertainty at time t will be measured as the standard deviation of the monthly return of the value weighted industry i portfolio from time t – 60 months to time t. Time t is the alliance announcement date which can be found in the SDC database. Monthly stock return data can be found for 48 different industries on Kenneth French’s website: (http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/). In appendix B an overview is presented of all 48 industries according to the used SIC codes and an overview of the industry representation within the sample. For every individual IJV in the sample the relevant industry SIC code will be searched in SDC platinum together with the alliance announcement date. After determining the industry and IJV announcement date, a time varying and industry specific measure of market uncertainty will be generated for every individual IJV in the sample, calculated as the standard deviation of the monthly return value weighted industry i portfolio from the monthly stock return data that are found on Kenneth French’s website. This will result in a value for market uncertainty in which a higher value reflects a higher uncertainty.
Host country risk
Environmental uncertainty can refer to economic and political risk in the host country (Anderson & Gatignon, 1986). In the literature the term country risk has been used to reflect the uncertainty over the continuation of present political and economic conditions (Agarwal & Ramaswami, 1992; Lu & Hébert, 2005). Country risk is measured in the literature by country credit risk rating (Pan, 1996). The World Economic forum provides country credit risk ratings with ratings for 144 different countries for all the relevant years. The scores range from 0 (highest risk) to 100 (lowest country risk). Caution should thus be paid to interpreting the statistics as high risk is reflected by a low score on this measure. The highest risk in the sample is 50,1 and the lowest risk in the sample is 87,8.
Intellectual property right protection
27
they measured a score on intellectual property right protection for 144 different countries which is a major advantage. Countries intellectual property protection can be scored between the value of 7 (highest protection) and 1 (lowest protection). The same data are used by Delios and Beamish (1999). Again caution should be paid to interpreting the statistics as high risk is reflected by a low score on this measure. The highest risk (lowest protection) in the sample is 3,1 and the lowest risk (highest protection) in the sample is 5,9.
Foreign partner prior international alliance experience
Consistent with Reuer & Tong (2005) prior international alliance experience will be measured by counting the number of international alliances (involving equity and non-equity) of the foreign partner, from 10 years prior to the current alliance announcement date. The SDC database will be used to find the previous international alliance experiences. Next prior general international
alliance experience will be defined as the log of one plus the number of alliances. The log plus
one is used because the log of 0 (in the case of no alliance experience) is undefined. The firm in the sample with the highest number of prior international alliance from 10 years prior to the current alliance announcement date has had 164 prior international alliance experiences and the lowest score in the sample is 0 prior international alliance experiences.
Foreign partner managerial risk preference
Hofstede (1980) developed a country-level measure of uncertainty avoidance (UA). The score on the UA dimension for the foreign partner as measured by Hofstede (1980) will be included in the research, consistent with (Richards & Yang, 2007). A high score on the UA dimension would indicate that members of this national culture would feel threatened by uncertain situations, and could be typified as risk-averse (Hofstede, 1980). The UA scores in the sample range from 8 to 95.
3.5 Control variables
In order to make the statistical analysis more reliable this study will control for several variables. First, there are several country-level variables that may influence the foreign partners’ ownership share in the IJV. In the literature a lot of attention has been paid to the influence of cultural
distance on entry modes and ownership distribution (Chang & Rosenzweig, 2001; Sartor &
28
included in this study. Second, host market attractiveness is believed to influence the option value of the IJV as fast economic growth in the host country enlarges the upside potential and future gains of the IJV (Reuer & Tong, 2005; Richards & Yang, 2007). This will be measured by the percentage growth rate in GDP of the host country, averaged across the 5 years before the IJV was established, consistent with Reuer & Tong (2005). Next, legal restrictions on foreign direct investment (FDI) have been included as control variable in various studies (Gatignon & Anderson, 1988; Mjoen & Tallman, 1997; Li & Li, 2010). As legal restrictions on FDI may impact the ownership decision the study will control for FDI restrictions in the host country which is measured by the OECD regulatory restrictiveness index in which countries are scored between 0 and 1 of equity restrictions, where 0 means no restrictions. Furthermore, some IJV specific control variables will be included. There will be a control for partner specific alliance
experience as it is found that prior interactions between the IJV partners may lead them to
develop trust, which may lower behavioral uncertainty and impact the ownership distribution in the IJV (Gulati, 1995; Kamminga & Van der Meer-Kooistra, 2007). Consistent with Li et al., (2010) this will be measured as the log of the number of non-equity alliances and joint ventures that both partners previously entered together. Next, firm size of the foreign partner will be controlled for, as this is a measure of a firm’s bargaining power and reputation, and larger companies will be able to exert more influence which could lead to higher ownership shares in the IJV (Brouthers et al., 2008; Li & Li, 2010; Richards & Yang, (2007). Consistent with Richards & Yang (2007) a dummy variable is created which is coded 1 if the foreign partner was a member of the Fortune Global 500 list during the year of the IJV establishment, and 0 otherwise.
3.6 Statistical method
29
30 4 RESULTS
Descriptive statistics and correlations between all the included variables are presented in table 2. Correlations of 0.7 or larger could indicate issues with multicollinearity. As multiple variables have correlations larger than 0.7, a collinearity statistics test was performed to detect multicollinearity. These collinearity statistics test can be found in table 1 of appendix C. Collinearity scores lower than 0.1 indicate collinearity and VIF scores higher than 10 also indicate collinearity. Host market attractiveness which is included as a control variable has the lowest collinearity score indicating that it is the most problematic variable that causes multicollinearity. Therefore the same collinearity statistics test is performed again without the
host market attractiveness variable to see if it solves the issue. As can be seen in table 2 in
appendix C, excluding host market attractiveness from the study solves the multicollinearity issue. As a consequence this control variable is excluded from the study.
31
Table 2. Descriptive statistics and correlations
Mean s.d. 1 2 3 4 5 6 7 8 9 10 11 12 Dependent variables 1. Foreign partner 49.06 9.09 1 equity stake 2. Foreign partner .95 .62 .69** 1 equity stake (categorized) Independent variables 3. Behavioral uncertainty .11 .31 .036 -.021 1 (High-tech brought by foreign p. dummy) 4. Market uncertainty 7.41 2.57 -.036 -.001 -.068 1 (Industry return volatility)
5. Host country risk 68.71 9.99 .006 -.003 .006 -.002 1
6. Intellectual property 3.99 .85 .046 .070 .018 .035 .882** 1 right protection in
the host country
7.Foreign partner general .64 .61 -.024 -.028 .106 -.071 .016 .029 1 international alliance
experience
8.Managerial risk preference 57.40 22.30 -.025 .051 -.067 .040 -.113* -.109* .078 1
Controls
9.Firm size (dummy) .21 .41 -.096 -.033 .008 .002 .094 .119* .577* .142* 1
10.Host market attractiveness 5.34 3.08 -.040 -.101 -.004 -.057 -.870** -.869** -.019 .134* -.086 1
11.FDI Restrictions .17 .11 -.012 -.057 .013 -.047 -.806** -.632** -.004 .129* -.054 .901** 1
12.Prior specific alliance .01 .06 -.038 -.051 -.005 -.004 -.072 -.020 .144** .025 .109* .030 .060 1 experience
32
Table 3. (Overdispersed corrected) Poisson regression estimation results
Model 1 Model 2
(Intercept) -.215(.1590) 3.834(.0418)***
Independent variables
- H1 Behavioral uncertainty (High-tech brought by foreign p.) =0 .112(.1361) .001(.0350) dummy 1=f. partner brings high-tech exclusively =1 0a 0a
- H2 Market uncertainty (Industry return volatility) -.010(.0147) -.005(.0041)
- H3 Host country risk -.037(.0098)*** -.006(.0030)**
- H4 Intellectual property right protection in the host country .319(.0909)*** .067(.0268)**
- Foreign partner general international alliance experience .167(.1752) .094(.0469)**
- Managerial risk preference -.003(.0059) .001(.0015)
Interactions
- H5 Foreign partner general international alliance experience =0 -.200(.1860) -.103(.0501)**
(x high-tech by foreign p.) =1 0a 0a
- H6 Foreign partner general international alliance experience -.007(.0235) .002(.0065) (x market uncertainty)
- H7 Foreign partner general international alliance experience .005(.0130) -.003(.0038) (x host country risk)
- H8 Foreign partner general international alliance experience -.039(.1483) .021(.0431) (x intell. Prop. Right protection)
- H9 Foreign partner managerial risk preference =0 .006(.0061) .000(.0016)
(x high-tech by foreign p.) =1 0a 0a
- H10 Foreign partner managerial risk preference .0001(.0007) .000(.0002) (x market uncertainty)
- H11 Foreign partner managerial risk preference -.00003(.0003) .00004(.0001) (x host country risk)
- H12 Foreign partner managerial risk preference -.001(.0040) -.001(.0012) (x intell. Prop. Right protection)
Controls
- Firm size (dummy) =0 .058(.1145) .067(.0317)**
=1 0a 0a
- FDI Restrictions -1.580*(.6074) -.174(.1706) - Prior specific alliance experience -.723(.6855) -.165(.1725)
Chi-square (likelihood ratio) 8.637 34.153
Pearson chi-square 129.672 491.508
Value/df .417 1.612
Akaike’s Information Criterion (AIC) 763.687 2407.29
1. * p=<0,10 **p=<0,05 ***p=<0,01
2.Dependent variable: model1:Foreign partner ownership share in categories model2: Foreign partner ownership share (actual share)
3.*High scores on intellectual property right protection and host country risk indicate lower uncertainty so positive relations indicate that the lower the uncertainty is, the lower ownership share of the foreign partner will be
4. Regression coefficients B & (Std. error). a. One or both parameter estimates are redundant
33
well. As predicted market uncertainty negatively affects the foreign partner ownership share in both models, but again the effect is statistically insignificant in both models. Therefore H2 which proposed a negative relationship between market uncertainty and the foreign partner ownership
share is rejected.
In both models the negative effect of host country risk should be interpreted as a positive effect between uncertainty and the foreign partner ownership share as a high score on the host country
risk variable indicates lower uncertainty. This effect is statistically significant at the p<0.01 level
in model 1 and at the p<0.05 level in model 2. The statistically significant positive effect of host
country risk on the foreign partner ownership share is contrary to the hypothesized negative
effect, leading to the rejection of H3. As is the the case with host country risk, a high score on
intellectual property right protection also indicates a lower uncertainty. Therefore the positive
coefficients for intellectual property right protection in the host country in both models should be interpreted as a negative effect of uncertainty stemming from a lack of intellectual property right
protection on the foreign partner ownership share. The effect is statistically significant in both
models, providing support for H4. H5 proposes that the positive relationship between behavioral
uncertainty and the foreign partner ownership share is less strong for foreign partners with more international alliance experience. The negative coefficient for H5 in model 2 which is
statistically significant at the p<0.05 level should be interpreted as a positively moderating effect which leads to the rejection of H5. H6, H7 and H8 propose a positive moderating effect of
foreign partner alliance experience on the relationships between the three exogenous sources of
uncertainty (market uncertainty, host country risk and risk stemming from a lack of intellectual
property right protection in the host country) and the foreign partner ownership share. The
expected positive direction is found for H6 and H8 but both results are statistically insignificant leading to rejection of H6 and H8. Contrary to the expectations a negative moderating effect is found for H7. However, this effect is also statistically insignificant leading to the rejection of H7.
H9 hypothesized that the positive relationship between behavioral uncertainty and the foreign partner ownership share would be even stronger if the foreign partner comes from a risk averse country (managerial risk preference). The results provide no empirical support as the results are
statistically insignificant. H10, H11 and H12 hypothesize that the negative effects between the three exogenous sources of uncertainty (market uncertainty, host country risk and risk stemming
34
for foreign partners from risk averse home countries (managerial risk preference). Although this negative effect is found for H12 in both models and for H11 in model 1, these results are statistically insignificant leading to the rejection of H10, H11 and H12. Regarding the control variables, foreign partners that belong to the Global Fortune 500 (Firm size) are found to have a lower ownership share. Next, in host countries with more FDI ownership restrictions the foreign partners will have a lower ownership share. Lastly, as expected, prior specific alliance experience between partner’s leads to a lower ownership share of the foreign partner; however this result is statistically insignificant. An overview of the results is presented in table 4.
Hypotheses Hypothesized direction Found direction Significant
H1 + - No H2 - - No H3 - + Yes H4 - - Yes H5 - + Yes H6 + +- No H7 + +- No H8 + +- No H9 + - No H10 - + No H11 - +- No H12 - + No
35 5 DISCUSSION
The main objective of this study was to empirically address and contrast how MNEs control or react to multiple sources of uncertainty (endogenous and exogenous) with the design of their initial ownership strategy in the IJV, while taking into account the possible moderating effects of the foreign partner international alliance experience and their managerial risk preference. The argumentation on behavioral uncertainty, which is an endogenous uncertainty, is based on TCE reasoning. Furthermore, the main line of reasoning that is applied to market uncertainty, host country risk and risk stemming from a lack of intellectual property right protection in the host country (forms of exogenous uncertainty) is derived from a RO perspective. The research, which used 329 two parent IJVs based in Brazil, India, Indonesia, Japan or the United Kingdom which were announced during the period from 2010 until 2012, partially supports the hypotheses. This study has made several theoretical contributions. First of all, it contributes to IJV research, as numerous scholars have looked at the conditions under which IJVs are a better option than other alternatives like wholly-owned subsidiaries or non-equity alliances (Brouthers et al., 2008; Brouthers 2013; Gulati, 1995; Li et al., 2009; Oxley, 1999), while the determinants of the ownership distribution of IJVs received far less attention within the IJV literature (Cuypers & Martin, 2009; Li & Li, 2010; Richards & Yang, 2007).
Second, it is important to recognize that almost no support is found for the TCE perspective. Contrary to the expectations I found a small negative effect of behavioral uncertainty on the
foreign partner ownership share and the effect was statistically insignificant. The lack of
36
ownership distribution in IJVs. More specifically, a firm first has to decide whether it wants to own the subsidiary outright (integrate) or decide to seek a local partner. If the firm decides to seek a local partner it has multiple nonintegrated options like for example a minority IJV, an equal IJV or a majority IJV. TCE may not be the best theoretical perspective to explain choices between these nonintegrated options. Richards & Yang (2007) however did find a positive effect of behavioral uncertainty on the foreign partner ownership in IJVs. The differences in findings could possibly be attributed to a different proxy and sample used by Richards & Yang (2007). Richards & Yang (2007) constructed a sample containing only R&D IJVs and operationalized
behavioral uncertainty by indicating whether or not the IJVs also engaged in marketing activities
as a proxy for behavioral uncertainty. The current study aimed for a wider scope however, by exploring the effect of uncertainty on the ownership distribution in a wide range of countries, industries and different types of businesses (R&D, manufacturing, service etc.).
Third, this study contributes to and extends the few RO studies that empirically investigate and contrast different sources of uncertainty and their effect on MNE ownership strategies in their IJVs. More specifically, this study extends prior RO studies as they solely focused on China as host country (Cuypers & Martin, 2009; Li & Li, 2010) while the current study has a wider scope of host countries. Prior studies have found that MNEs prefer to take lower ownership shares in their IJVs in response to exogenous sources of uncertainty (Cuypers & Martin, 2009; Li & Li, 2010). The results partially supported the main RO based hypotheses as negative relations are found in both models between market uncertainty and the foreign partner ownership share and
uncertainty stemming from a lack of intellectual property right protection in the host country and foreign partner ownership share. The effect for market uncertainty is statistically insignificant
however. As RO already predicted, higher uncertainty stemming from a lack of intellectual
property right protection in the host country leads to foreign partners that seek to limit their
downside risk by taking lower ownership shares. Therefore it can be concluded that RO predictions regarding exogenous uncertainties’ influence on the ownership distribution in IJVs also applies to other countries than China. The statistically significant positive effect of host
country risk on the foreign partner ownership share is striking however, as it counters RO
37
plays a major role when asset specificity is high, magnifying the need for control that asset specificity creates (Anderson & Gatignon, 1986). More specifically, an interaction effect is suggested between environmental uncertainty and high asset specificity. Unfortunately, I had no access to any data on asset specificity for this particular dataset, so I was unable to test this proposition.
Regarding foreign partner international alliance experience, contrary to the expectation, a significant positive moderating effect was found between behavioral uncertainty and the foreign
partner ownership share. This counters TCE thinking which suggests that experience tends to
reduce transaction costs, which lowers the need for control and thus equity (Li et al., 2010). The positive moderating effect corresponds with RO reasoning which suggests that prior
international alliance experience increases confidence of managers to deal with host-market
uncertainties and this is reflected in higher-control, higher equity entry modes (Ahsan & Musteen, 2011). Moreover, RO theory suggests that firms with more international alliance
experience perceive lower ‘risk of loss’, while in contrast MNEs with less international
experience seek downside risk protection (Brouthers et al., 2008). No interacting effects are found of prior international alliance experience between exogenous sources of uncertainty and the foreign partner ownership share. This is consistent with Li et al., (2010) who found that general alliance experience did not moderate the impact of market uncertainty on the firm’s alliance governance mode. Therefore it can be suggested that a difference exists in the existence of a moderating effect of prior international alliance experience based on the endogenous-, exogenous uncertainty distinction. More specifically, it could be stated that prior international
alliance experience is likely to make MNEs more effective in dealing with uncertainties as is
38
2007). The results indicate that experience has an effect on MNEs ownership strategy in response to uncertainty. The observed effect also complements the stages model of internalization (Delios & Beamish, 1999; Johanson & Vahlne, 1977) which predicts that as experiential learning about foreign markets increases, uncertainty reduces which will reflect in MNEs that incrementally increase their commitment (e.g. taking larger ownership shares) in the foreign markets. However, it should be noted that there is a possibility that the type of experience matters for the effect that the experience has on the relation between uncertainty and the ownership share (Kamminga, 2003; Van der Meer-Kooistra, 2015; Lu & Hébert, 2005). Lu & Hébert (2005) for example argue that experience in the form of local knowledge reduces a foreign partner’s dependency on equity as a mechanism of control. On the other hand, Kamminga (2003) describes a case where the foreign partner learned from a previous IJV experience that a majority ownership is necessary to protect their technological knowledge. So it may be concluded that the type of experience, or learning, could impact the direction of the effect of previous alliance experience.
Furthermore, this study proposed that risk neutral TCE and RO models should be adjusted as managers make their decisions under bounded rationality, that is, that their perceptions, decisions and actions will be filtered and distorted by their values and cognitive biases (March & Simon, 1958; Richards & Yang, 2007). More specifically I proposed that managerial risk preference moderates the relationship between various sources of uncertainty and the ownership distribution in IJVs. No significant moderating effects of managerial risk preference on the relation between uncertainty and the ownership distribution in IJVs are found however. In contrast to Richards & Yang (2007) who found a moderating effect of uncertainty avoidance on the relationship between country risk and the ownership distribution.
39
reporting relationship and more, may be needed to determine real or effective control of the partner (Geringer & Hébert, 1989; Kamminga & Van der Meer-Kooistra, 2007; Schaan, 1983). Next it is important to recognize that ownership represents not only control but also the extent of profit appropriation (Richards & Yang, 2007). Having a larger equity share in the IJV basically means obtaining a larger share of the profits. Especially if MNEs expect the potential payoffs to be large, for example in countries with high GDP growth rates, this may lead to MNEs taking larger equity shares as a way to capture future profits (Li & Li, 2010). Furthermore, as mentioned, Li & Li (2010) found that their results for RO predictions that did hold for low competitive markets disappeared in highly competitive markets, as they argued that competition can considerably reduce the exclusiveness of a firm’s growth option and thus reduce the value of a growth option (Li & Li, 2010; Trigeorgis, 1988). These factors may act as confounding effects that possibly lead to deviations from the foreign partner ownership share as expected by main theories.
40
relationship or in the environment of the IJV relationship (Kamminga & Van der Meer-Kooistra, 2015). Consequently, the organizational context of the parent company can also influence decisions on IJV control, as for example a bad financial situation of the foreign partner may lead to a lower equity share.
41 6 LIMITATIONS AND FURTHER RESEARCH
This research is bounded by several limitations and provides several directions for further research. First of all, there is a limitation of data availability. As discussed, SDC contained a lot of data but unfortunately I had no access to any data to operationalize asset specificity. TCE suggests that environmental unpredictability plays a major role when asset specificity is high; magnifying the need for control that asset specificity creates (Anderson & Gatignon, 1986). This proposed interaction effect could possibly explain the unexpected finding for host country risk. Therefore future research should include asset specificity as an interacting effect as it is possible that this effect leads to deviations from RO predictions. Furthermore, the impact of host market attractiveness and competitiveness on the options within IJVs is discussed (Li & Li, 2010; Richards & Yang, 2007). As there is lack of data on host market characteristics and the used host market attractiveness had to be excluded because it suffered from multicollinearity issues this also provides a direction for further research. More specifically, attractive host countries that achieve large economic growth for multiple years could indicate that MNEs can expect larger profits which lead to a larger static NPV part which MNEs will try to capture by taking larger ownership shares (Cuypers & Martin, 2009; Li & Li 2010). As the majority of the IJVs in the database come from countries that achieved large GDP growth rates in that years this may have led to MNEs taking larger ownership shares as suggested by RO theory. RO research can improve its explanatory power by incorporating host market characteristics that influence the options in IJVs as well as including asset specificity as it may have a moderating effect on exogenous sources of uncertainty.
42
Furthermore, the discussion pointed to major shortcoming of both TCE and RO as both theories ignore the relational aspects of the IJV relation and its impact on IJV control decisions. Further research could enhance IJV control research by incorporating relational aspects and their interplay with TCE and RO factors to their study. Moreover, to really capture these relational aspects and the decisions made regarding the IJV control, the organizational context of the parent firm may also be important to study as it could also have a major impact.
43 7 CONCLUSION
While numerous scholars have looked at the conditions under which IJVs are a better option than other alternatives like wholly-owned subsidiaries or non-equity alliances (Brouthers et al., 2008; Brouthers 2013; Gulati, 1995; Li et al., 2009; Oxley, 1999), the determinants of the ownership distribution of IJVs received far less attention within the IJV literature (Cuypers & Martin, 2009; Li & Li, 2010; Richards & Yang, 2007). Furthermore, RO theory was indicated as a promising alternative theoretical perspective to study international strategy and governance decisions which were mostly based on TCE theory. As scholars found that RO predictions with regard to ownership strategy in IJVs only hold exogenous uncertainty, and TCE contributed to the analysis of the effect of endogenous uncertainty, this study used both theoretical perspectives complementary. Moreover, this study aimed to explore the boundary conditions under which RO logic holds by investigating the moderating effects of prior international alliance experience and managerial risk preference, which are both foreign partner firm characteristics which are believed to moderate the relationship between uncertainty and foreign parent ownership (Ahsan & Musteen, 2011; Bowman & Hurry, 1993; Brouthers et al., 2008; Cuypers & Martin, 2007; Fisch, 2008). By exploring and contrasting several sources of uncertainty and relating them to parent control with the incorporation of the two mentioned firm characteristics, this study attempted to address to following central question:
‘’What is the effect of the foreign partner’s previous international alliance experience and managerial risk preference, on the relationship between (endogenous and exogenous sources of) uncertainty and -the initial foreign partner’s equity share in IJVs?’’
44
