INTRODUCTION
The entry mode of foreign direct investments (FDIs) has been one of the most important research areas of international business. Drawing on various theoretical models and perspectives, numerous studies have explored the factors which influence multinational corporations’ (MNCs’) entry mode choice between joint ventures (JVs) and wholly owned subsidiaries (WOSs) (Anderson & Gatignon, Reference Anderson and Gatignon1986; Gatignon & Anderson, Reference Gatignon and Anderson1988; Madhok, Reference Madhok1998; Erramilli, Reference Erramilli1991; Kim & Hwang, Reference Kim and Hwang1992; Delios & Beamish, Reference Delios and Beamish1999; Delios & Henisz, Reference Delios and Henisz2000; Brouthers & Brouthers, Reference Brouthers and Brouthers2001; Meyer, Reference Meyer2001). However, the majority of studies have focused on firm characteristics, such as proprietary knowledge and assets (Williamson, Reference Williamson1985; Anderson & Gatignon, Reference Anderson and Gatignon1986; Dunning, Reference Dunning1993), capabilities (Madhok, Reference Madhok1998), and experiences of the investing companies (Erramilli, Reference Erramilli1991; Delios & Beamish, Reference Delios and Beamish1999), leaving corporate governance factors mostly untapped, except for a few studies such as Filatotchev, Strange, Piesse, and Lien (Reference Filatotchev, Strange, Piesse and Lien2007).
The absence of such research is unusually surprising, because the study of corporate governance has had a long historical standing. Specifically, controlling shareholders and board members indeed influence corporate strategy and performance, as prior studies indicate, because they make or approve key decisions (see Zahra & Pearce Reference Zahra and Pearce1989, for a review). Their risk preferences shape firms’ strategic moves (La Porta, Lopoz-de-Silanes, Shleifer, & Vishny, Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999; Claessens, Djankov, & Lang, Reference Claessens, Djankov and Lang2000; Thomsen & Pedersen, Reference Thomsen and Pedersen2000; Douma, George, & Kabir, Reference Douma, George and Kabir2006), and they control critical resources that determine firms’ resource endowment (Shrader & Simon, Reference Shrader and Simon1997). Both these factors have strong bearing on corporate internationalization strategies (Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson, & Moyano-Fuentes, Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007).
Among various types of controlling shareholders, family owners may be one of the most important. The family enterprise is the world's oldest and most common form of economic organization and, as noted by La Porta et al. (Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999), family-controlled corporations dominate the global economic landscape. Family ownership is pervasive in US firms across a broad range of industries (Schulze, Lubatkin, & Dino, Reference Schulze, Lubatkin and Dino2003; Gómez-Mejía et al., Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007; Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010), as well as in MNCs from emerging countries (La Porta et al., Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999; Claessens, Djankov, & Lang Reference Claessens, Djankov and Lang2000; Yeh, Lee, & Woidtke, Reference Yeh, Lee and Woidtke2001; Peng & Jiang, Reference Peng and Jiang2010). Despite family control's prevalence and economic significance, few international business scholars so far have addressed how family control influences MNCs’ FDI decisions. This study aims to fill this gap.
Prior studies have found that family-controlled firms are more averse to risks, while at the same time demonstrate higher desire to control, due to their unique concerns regarding the preservation of socioemotional wealth (Gómez-Mejía et al., Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007; Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010). To understand the influence of family control on FDI decision making, one has to take both characteristics – risk aversion and high desire to control – into consideration. Reckoning with these essential characteristics, the researchers use a sample of 1,644 investments from publicly listed companies in Taiwan to investigate how family control influences MNCs’ entry mode choice. Family business owners’ risk aversion and high desire to control, the researchers posit, will make their firms more responsive to environmental uncertainty. That is, when perceiving high environmental uncertainty, firms with higher versus lower family control tend to choose JVs. Nevertheless, when perceiving low environmental uncertainty, firms with higher family control tend to choose WOSs.
This study makes several contributions. First, it finds that family control plays an important role in MNCs’ FDI decision making – it makes firms respond more drastically to environmental uncertainty. When perceiving high environmental uncertainty, all firms generally tend to choose JVs, but firms with higher family control are more likely than firms with lower family control to choose JVs. However, when perceiving low uncertainty, firms tend to choose WOSs in general, but firms with higher versus lower family control are more likely to choose WOSs.
Second, to the research stream of family business, the current study extends from traditional wisdom to provide an alternative point of view on the internationalization of family businesses. Extant studies in family businesses’ internationalization have mostly found that, compared with firms controlled by other types of shareholders, family-controlled firms exhibit a lower degree of internationalization, which reflects in their preference of the choice of lower commitment entry mode (e.g., Kumar & Subramaniam, Reference Kumar and Subramanian1997; Filatotchev et al., Reference Filatotchev, Strange, Piesse and Lien2007; see Kontinen & Ojala (Reference Kontinen and Ojala2010) for a review). Our study suggests that this traditional wisdom may only tell half of the story. Indeed, when family-controlled firms encounter high uncertainty, they are inclined to avoid internationalization, as earlier research predicted. However, our research finds firms controlled by family members behave differently when they perceive low uncertainty – they act more aggressively than other types of firms to assume full equity ownership of their foreign subsidiaries.
Third, examining the influence of family control on entry mode choice, this study finds that the investing firms’ corporate governance matters in their internationalization decision making. Specifically, this study confirms the influence of a specific type of controlling shareholders – family members – on firms’ internationalization. The finding fills the gap created by the scarcity of entry mode studies exploring the influence of corporate governance factors and opens a new window for future research to explore other corporate governance factors that may possibly influence internationalization.
In the following sections, the researchers will first review past literature and develop hypotheses. Then the researchers will describe the sample and our methodology, followed by research findings, discussion, and conclusion.
LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
Choosing an entry mode between WOSs and JVs
When undertaking FDIs to enter a foreign country, MNCs have to choose an entry mode – either JV or WOS. Forming a JV to collaborate with local partners allows MNCs to lower their resource commitment and share risks with their partners, while establishing a WOS enables MNCs to fully control their subsidiary and reap all the returns from their foreign operations. Each choice brings different benefits, but both also suffer from certain drawbacks. Among various theories addressing the pros and cons of each entry mode, the transaction cost economics is one of the most widely adopted (Hennart, Reference Hennart1988, Reference Hennart1993; Dikova & Van Witteloostuijn, Reference Dikova and Van Witteloostuijn2007).
Based on the behavioral assumptions of bounded rationality and opportunism, the transaction cost economics perspective considers the emergence of ‘institutions of capitalism’ (Williamson, Reference Williamson1985) to solve problems of opportunism. To economize on bounded rationality and simultaneously safeguard transactions against opportunism calls for an appropriate governance structure (Williamson, Reference Williamson1985; Coombs, Bierly, & Gallagher, Reference Coombs, Bierly and Gallagher2012). International business scholars apply the transaction cost economics perspective to the choice between JVs and WOSs (Hennart, Reference Hennart1988; Kim & Hwang, Reference Kim and Hwang1992; Brouthers & Brouthers, Reference Brouthers and Brouthers2001). International operation incurs significant administrative costs, they argue. When MNCs enter a foreign country, the costs of monitoring, dispute settling, and reward refining are especially high (Hennart, Reference Hennart1988), making the management of foreign subsidiaries costly and uncertain. Collaborating with local partners via the JV arrangement (typically with a formal contractual agreement) allows foreign MNCs to leverage partners’ local knowledge and resources and thus reduces such costs and related risks (Teece, Reference Teece1981; Hennart, Reference Hennart1988).
Furthermore, sharing the ownership of the foreign affiliate with a local partner to form a JV also allows MNCs to enter that specific country with lower resource commitment and shift certain risks to the local partner (Kim & Hwang, Reference Kim and Hwang1992). With the JV entries, firms can withdraw from the host country when unpredictable environmental changes affect them adversely and redirect the investment elsewhere to seize new market opportunities (Hill, Hwang, & Kim, Reference Hill, Hwang and Kim1990). Therefore, when encountering high uncertainty, JVs appear to be a good choice for firms to avoid risks.
However, the JV choice is not without cost. Foreign partners may behave opportunistically if given the chance (Hennart, Reference Hennart1988). Since human beings are subject to bounded rationality, acquiring sufficient information to foresee partners’ behaviors and safeguard against partners’ potential opportunism is quite unlikely. Thus, the shared equity arrangement of JV bears the ex ante transaction costs of discovering a proper partner, drafting an agreement, and bonding contractual arrangements, as well as the ex post transaction costs of haggling, adaptation, monitoring, enforcement, termination, and the residual loss of cheating and shirking (Williamson, Reference Williamson1985).
In a foreign country, the problem of partners’ opportunism is especially severe. Unfamiliar culture creates difficulties in understanding local partners’ thinking, consequently limiting the accuracy of predicting their behavior (Gatignon & Anderson, Reference Gatignon and Anderson1988; Kogut & Singh, Reference Kogut and Singh1988; Brouthers & Brouthers, Reference Brouthers and Brouthers2001; Soares, Farhangmehr, & Shoham, Reference Soares, Farhangmehr and Shoham2007). When MNCs endeavor to select local partners, the cultural gap blurs the existing criteria of selecting an appropriate partner. Evaluating partners’ potential for opportunism also becomes difficult. Even after a partner is selected, MNCs still need to deal with the challenges of negotiating, monitoring, and enforcing contractual agreements with the partner, and these challenges may incur intolerably high costs for MNCs (Brouthers and Brouthers, Reference Brouthers and Brouthers2001). Under such circumstances, the need to control increases and the WOSs choice appears to be appealing.
Even if the local partners do not behave opportunistically, the JV arrangement still calls for MNCs to share profit with their local partners. However, the WOS choice offers the benefits of full control over foreign operations. Within its own organization, an MNC has full control and makes decisions at its discretion, without worrying about the partners’ potential opportunistic behavior. MNCs can also appropriate all the returns generated from their foreign operations without the need to share with anybody. As a result, the WOS mode of entry is preferred when the costs of arranging, monitoring, and enforcing a collaborative agreement are high (Kim & Hwang, Reference Kim and Hwang1992; Brouthers & Brouthers, Reference Brouthers and Brouthers2001).
In short, JV mode of entry allows MNCs to leverage partners’ local knowledge to reduce risks and share risks with them. On the contrary, the WOS choice provides the opportunity for MNCs to maintain full control of their foreign subsidiaries and reap all the returns from these subsidiaries. To choose between a WOS or a JV depends to a large extent on how high the investing firm perceives the environmental uncertainty (Anderson & Gatignon, Reference Anderson and Gatignon1986; Brouthers, Brouthers, & Werner, Reference Brouthers, Brouthers and Werner2000).
A firm's perceived environmental uncertainty is determined by its internal endowments, as well as external conditions (Milliken, Reference Milliken1987). Past research has found that organization capabilities and experiences highly influence perceived environmental uncertainty (Duncan, Reference Duncan1972; Lorenzi, Sims, & Slocum, Reference Lorenzi, Sims and Slocum1981). Specifically, Milliken (Reference Milliken1987) pointed out that learning facilitates organizations to improve their capability of assessing the environmental changes and better predict future status of the environment. When managers in an organization can foresee the future more accurately, they know how to react and hence perceive lower uncertainty (Duncan, Reference Duncan1972). In FDI studies, scholars discovered similar determinants. Anderson and Gatignon (Reference Anderson and Gatignon1986) and Fisch (Reference Fisch2008) found that international experience help mitigate uncertainty and risks, arguing that MNCs with more international experience perceive lower uncertainty and risks than MNCs with less international experience.
On the other hand, the host country's institutional environment also plays an important role when MNCs assess uncertainty. Inquiries on FDI decisions have indicated that quality of a country's institutional environment signifies uncertainty and thus affects an MNC's willingness to enter that country (Brouthers, Brouthers, & Werner, Reference Brouthers, Brouthers and Werner2000; Yiu & Makino, Reference Yiu and Makino2002; Puck, Holtbrügge, & Mohr, Reference Puck, Holtbrügge and Mohr2008; Slangen & Van Tulder, Reference Slangen and Van Tulder2009; López-Duarte & Vidal-Suárez, Reference López-Duarte and Vidal-Suárez2010). Poor quality of the country's institutions, typically characterized by political instability, capricious policies, ambiguous regulations, corruption, anarchy, or terrorism, brings high risks and discourages MNCs from conducting business there, while outstanding institution quality increases foreign MNCs’ willingness to invest and commit longer (Globerman & Shapiro, Reference Globerman and Shapiro2003; Gani, Reference Gani2007).
In sum, when MNCs posses abundant international experience, or when they are entering a country with good institutions, they perceive lower uncertainty, and thus tend to choose WOSs to enter. However, firms with high family control may act differently when they choose entry mode, because they have distinctive risk preferences and different degrees of desire to control. Compared with firms with low family control or without any influences of family owners, firms with high family control appear to be more risk averse and show a higher desire to control (Gómez-Mejía et al., Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007; Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010). These characteristics may also reflect in their entry mode choice. Below the researchers elaborate details on the two characteristics and how they impact family-controlled firms’ entry mode choice.
Risk aversion
Firms with high family control are dominated by family members who possess the right to make key decisions. Family members usually have their specific objectives – they establish and operate companies not only to make profit but also to maintain socioemotional wealth (e.g., Gomez-Mejia, Makri, & Kintana Reference Gomez-Mejia, Makri and Kintana2010). For family members, their firm provides, in addition to economic benefits, socioemotional functions such as personal attachment, commitment, and identification with the company (De Vries, Reference De Vries1993; Habbershon & Williams, Reference Habbershon and Williams1999; Thomsen & Pedersen, Reference Thomsen and Pedersen2000; Anderson & Reeb, Reference Anderson and Reeb2003).
For the reason of socioemotional wealth, family business owners consider their firm the family's patrimony (Miller & Rice, 1967; Hollander, Reference Hollander1983) or a mechanism to provide employment and financial security for their family (Liebowitz, Reference Liebowitz1986). Hence, they attempt to preserve their firm's ‘familiness’ by recruiting family members and relatives (Bruton, Ahlstrom, & Wan, Reference Bruton, Ahlstrom and Wan2003) or handing over their businesses to their offspring (Beckhard & Dyer, Reference Beckhard and Dyer1983). Those nepotic endeavors exclude non-family members from entering the company or purchasing the firms’ shares. Consequently, managers in firms with high family control are often suffering from lacking diversity, making these firms more vulnerable to environmental changes (Liebowitz, Reference Liebowitz1986).
Previous studies on finance and corporate governance also suggest that family-controlled firms are more averse to risks (Yeh, Lee, & Woidtke, Reference Yeh, Lee and Woidtke2001; Anderson & Reeb, Reference Anderson and Reeb2003, Reference Anderson and Reeb2004). Family equity holdings are usually more concentrated, resulting in a relative lack of financial portfolio diversification and limited liquidity (Anderson and Reeb, Reference Anderson and Reeb200). Consequently, family shareholders may be affected more adversely by the company's specific risks than other types of dominant shareholders (e.g., institutional investors) with diversified portfolios of shares (Maug, Reference Maug2002). Of course, family shareholders may try to mitigate the risks via strategic moves such as product diversification (Chang, Reference Chang2003), but such moves cannot reduce firm-specific financial risks associated with concentrated shareholdings since the company's failure may imply the loss of all returns (Yeh, Lee, & Woidtke, Reference Yeh, Lee and Woidtke2001; Douma, George, & Kabir, Reference Douma, George and Kabir2006). Hence, firms with high family control, compared with firms with low family control or without any family influences, are less tolerable to risks, and ‘their strategic behavior is rather conservative’ (Donckels & Fröhlich, Reference Donckels and Fröhlich1991).
Family-controlled firms’ risk aversion reflects in their prudence in resource commitment. For example, they are less likely to invest in research and development (R&D) and undertake innovation (Jensen & Meckling, Reference Jensen and Meckling1976). The same tendency has also been frequently found in family-controlled firms’ FDI decision. When encountering uncertainty in FDI, a risky investment, family businesses appear to be more conservative and risk averse (Wiseman & Gomez-Mejia, Reference Wiseman and Gomez-Mejia1998; Chatterjee, Lubatkin, & Schulze, Reference Chatterjee, Lubatkin and Schulze1999; Anderson & Reeb, Reference Anderson and Reeb2004). Specifically, Filatotchev et al. (Reference Filatotchev, Strange, Piesse and Lien2007) empirically find that firms with high family control tend to avoid resource commitment and hold lower equity stake in their overseas affiliates when they undertake FDI activities.
The phenomenon that family-controlled firms are more risk averse may also reflect in these firms’ entry mode choices. As elaborated earlier, when perceiving high environmental uncertainty in a foreign country, firms tend to choose JV, the shared ownership arrangement, to reduce risk. Since firms with high family control are more risk averse, they may be even more likely than firms with low family control to choose JV. Hence, the researchers predict that:
Hypothesis 1 : When perceiving high environmental uncertainty, firms with higher versus lower family control are more likely to choose JV as their entry mode.
High desire to control
Family members’ socioemotional motive also reflects in family-controlled firms’ tendency to maintain a high equity stake in their affiliates (Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010). Once outside shareholders buy out a majority of shares of a family-controlled firm, the firm's ‘familiness’ will be attenuated, and the controlling family will no longer be able to dominate the firms’ strategy to meet their family needs. Hence, family shareholders frequently use practices such as ‘pyramidal ownership structures’ (e.g., La Porta et al., Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999; Claessens, Djankov, & Lang, Reference Claessens, Djankov and Lang2000) (in which firm A owns the majority of shares of firm B, firm B owns the majority of shares of firm C, and so on) and ‘cross holding’ (Peng & Jiang, Reference Peng and Jiang2010) (in which firm A owns equity in firm B, and at the same time firm B holds equity in firm A) to exercise their control over a group of firms through a chain of ownership relations (Gomez-Mejia et al., Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007; Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010).
Meanwhile, being the ultimate owners located at the apex of the ownership relations enable family members to allocate and distribute resources available from these successive layers of firms to fulfill their family objective. This rationale partly, if not totally, explains why ‘tunneling’ (the transfer of assets and profits out of firms for the benefits of those who control them) are widespread in family business groups (Bae, Kang, & Kim, Reference Bae, Kang and Kim2002; Bertrand, Mehta, & Mullainathan, Reference Bertrand, Mehta and Mullainathan2002).
For family-controlled firms, layers of firms in the pyramid or firms bonded by cross-holding structure are part of their overall asset portfolio, even these firms are in foreign countries. When undertaking FDI activities, family-controlled firms also regard their foreign subsidiaries as part of their asset portfolio. Compared with firms with low family influence, firms with high family control usually have higher desire to control their foreign subsidiaries, especially via equity ownership. When they invest in a foreign country where environmental uncertainty is considered high, their high desire to control via equity ownership may be suppressed by their risk aversion. Nonetheless, when perceiving low uncertainty, they will be more likely to choose WOSs, the full ownership arrangement. Hence, the researchers predict that:
Hypothesis 2 : When perceiving low environmental uncertainty, firms with higher versus lower family control are more likely to choose WOSs as their entry mode.
Figure 1 summarizes the detailed influence of perceived environmental uncertainty and family control in firms’ entry mode choices between WOSs and JVs.
Figure 1 The influence of perceived environmental uncertainty and family control on the entry mode choice between wholly owned subsidiaries (WOSs) and joint ventures (JVs)
METHODOLOGY
Data and sample
The study analyzes Taiwanese firms’ overseas investments in 13 countries from 1999 to 2008. Financial data and information are taken from the Taiwan Economic Journal database. Taiwan Economic Journal was founded in 1990, and its database contains all overseas investments carried out by publicly listed firms in Taiwan. The regulation of Taiwan requires all publicly listed firms to seek ex ante approval or to carry out ex post filing for all overseas investments, including both JVs and WOSs, so the data is complete and does not place any threshold of investment. The complete dataset contains 12,399 entries. The researchers then set a threshold of investment – at least 5% of the property of the venture located in the host country – to rule out trivial equity holdings. The researchers further exclude investments in China, due to its high cultural similarities with Taiwan, to avoid statistical bias. Investments in tax heavens, such as Cayman Islands and British Virgin Islands, and investments in Hong Kong and Singapore are also excluded due to ambiguousness of investment intention.
After eliminating observations with missing data, the final sample consists of 1,644 observed investments of overseas investments from 505 firms. The data were then triangulated with data from annual reports and the Market Observation Post System of Taiwan Stock Exchange to ensure data reliability. Table 1 lists the overall sample distribution by country. Table 2 reports the industry breakdown of the sample set.
Table 1 Country breakdown of the sample
Table 2 Industry breakdown of the sample
aPercentage is calculated based on the number of investments.
Variables and measurements
Dependent variable
Entry mode
Our dependent variable is entry mode. A WOS is defined as a subsidiary with 95% or higher percentage of shares owned by the parent (investing) company from Taiwan. The distinction between WOS and JV has been inconsistent. In previous studies, some researchers have used an 80% cutoff, following conventional accounting practices that identify the minimum necessary equity level to assume control as 20% (Makino & Beamish, Reference Makino and Beamish1998). However, most researchers (Anderson & Gatignon, Reference Anderson and Gatignon1986; Hennart, Reference Hennart1991; Makino & Neupert, Reference Makino and Neupert2000; Mani, Antia, & Rindfleisch, Reference Mani, Antia and Rindfleisch2007) have adopted a stricter criterion, 95% equity ownership, as the cutoff point to differentiate between a WOS and a JV. On the other hand, 100% ownership for a foreign subsidiary to be defined as wholly owned is not appropriate for Taiwanese firms, since most Taiwanese computer and electronics firms grant certain stock ownership to their employees as part of their employee reward schemes (Han, Reference Han2003; Chen & Huang, Reference Chen and Huang2006). Therefore, in this study, we follow the 95% cutoff point to differentiate between a WOS and a JV. A dummy variable is created, taking the value of 1 to represent WOSs and the value of 0 to represent JVs. We then use binary logistic regression to analyze various factors’ influences on entry mode.
Independent variables
Family control
This variable measures the degree of family control in sample firms. In many institutional contexts, family owners may gain control rights that deviate significantly from cash-flow rights (La Porta et al., Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999). In such circumstances, a family member can exert influence over a firm beyond what its fractional share ownership would indicate. Hence, family members’ fractional share ownership does not fully reflect family control in a firm. Instead, family members holding seats in the board of directors are crucial to the firms’ key decisions, because legally the board is the highest authority in the firm, the ‘fountain of power’ (Demb & Neubauer, Reference Demb and Neubauer1992: 13–16). Therefore, the researchers measure family control with family board presence in the investing firms. This measure has been widely adopted by prior studies to measure family control (e.g., Yeh, Lee, & Woidtke, Reference Yeh, Lee and Woidtke2001; Anderson and Reeb, Reference Anderson and Reeb2003, Reference Anderson and Reeb2004). Family board presence refers to the percentage of board seats held by family members (i.e., the number of board seats held by family members divided by the total number of board seats). This variable is mean centered before running regression.
Perceived environmental uncertainty
The researchers construct the measure of perceived environmental uncertainty with a composite index – Totalindex. This composite index is comprised of two indices – Index_E and Index_L. The former (Index_E) measures the firm's specific aspects of perceived environmental uncertainty, and the latter (Index_L) measures location-related aspects of perceived environmental uncertainty. As elaborated earlier, an MNC's perceived environmental uncertainty is determined by its international experience and the quality of target location's institutional environment. MNCs with rich international experience perceive low uncertainty, and their counterparts with little international experience perceive high uncertainty (e.g., Anderson and Gatignon, Reference Anderson and Gatignon1986; Fisch, Reference Fisch2008). Moreover, the quality of target location's institutional environment also signifies uncertainty – the better the target location's institutions, the lower the uncertainty (Yiu & Makino, Reference Yiu and Makino2002; Puck, Holtbrügge, & Mohr, Reference Puck, Holtbrügge and Mohr2009; Slangen & Van Tulder, Reference Slangen and Van Tulder2009; López-Duarte & Vidal-Suárez, Reference López-Duarte and Vidal-Suárez2010). Therefore, the researchers construct Index_E and Index_L to measure the two different aspects of perceived environmental uncertainty.
Index_E is composed of three factors – the breadth of international experience (i.e., the number of host countries an MNC has already invested in), the depth of international experience (i.e., the number of an MNC's foreign subsidiaries), and the duration of international experience (i.e., the accumulated number of years an MNC has been undertaking its overseas investments). The researchers define co, subs and iexp to measure the breadth of an MNC's international experience, the depth of an MNC's international experience, and the duration (accumulated number of years) of international experience, respectively. Since these measures have different scales, the researchers standardize all of them. After standardization, the three measures share with the common mean of 0 and standard deviation of 1. Finally, the researchers define Index_E as the average of co, subs, and iexp. That is, Index_E = (co + subs + iexp)/3. Larger value of Index_E denotes higher overall international experience and consequently lower perceived environmental uncertainty.
Index_L is constructed based on the ‘Governance Indicators’ developed by the World Bank to measure host countries’ governance quality. Starting in 1996, the World Bank has conducted ongoing surveys on the governance quality of 212 countries worldwide. Using an unobserved components model, the team identified six dimensions of governance quality: voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law, and control of corruption (Kaufmann, Kraay, & Mastruzzi, Reference Kaufmann, Kraay and Mastruzzi2007). Each dimension is assigned a score, with higher values indicating better governance quality of the country measured. In this study, the researchers define six variables, dg1, dg2, dg3, dg4, dg5, and dg6, for the six dimensions of governance quality. The six variables have also been standardized (mean = 0 and standard deviation = 1 after standardization). Finally, the researchers define Index_L as the average of dg1 through dg6. The final Index_L value denotes how good each country's governance quality is, with larger value indicating better governance quality and consequently lower perceived environmental uncertainty.
Finally, the researchers define the composite index – Totalindex– as the average of Index_E and Index_L (i.e., Totalindex = 0.5Index_E + 0.5Index_L) to measure the perceived environmental uncertainty of each investment. Larger value of Totalindex denotes better perceived quality of environment and consequently lower perceived environmental uncertainty.
To differentiate between the impact of high perceived environmental uncertainty and low perceived environmental uncertainty, the researchers further define two dummy variables – High Perceived Uncertainty and Low Perceived Uncertainty. The researchers ranked the perceived environmental uncertainty of all investments based on their scores of Totalindex, and they define the top-ranking quartile as Low Perceived Uncertainty and the lowest-scoring quartile as High Perceived Uncertainty. The researchers set higher thresholds (the upper quartile and the lower quartile) for high and low perceived environmental uncertainty to avoid possible bias of sample firms’ arbitrary criteria in judging the uncertainty of a specific country. In the binary logistic regression, the researchers examine how the two interaction terms – Family control × Low Perceived Uncertainty and Family control × High Perceived Uncertainty– influence entry mode choice.
Control variables
Cultural distance
Prior research has found that cultural distance influences MNC's entry mode choice (Gatignon & Anderson, Reference Gatignon and Anderson1988; Kogut & Singh, Reference Kogut and Singh1988; Brouthers & Brouthers, Reference Brouthers and Brouthers2001; Soares, Farhangmehr, & Shoham, Reference Soares, Farhangmehr and Shoham2007).The researchers therefore include this variable to control its possible effect and follow Kogut and Singh (Reference Kogut and Singh19 to construct this variable as a composite index based on the four cultural dimensions from Hofstede (Reference Hofstede1980, Reference Hofstede2001) (i.e., power distance, uncertainty avoidance, masculinity/femininity, and individualism).
Firm size
Firm size is measured with the firm's total asset, coded by taking the log value. Larger firms usually have abundant resources and better management capabilities and thus have lower needs for resources provided by local partners when entering a foreign country (Gatignon & Anderson, Reference Gatignon and Anderson1988). The researchers therefore control for possible effects of parent firm size on entry mode choice.
Firm age
Firm age is measured in years, coded by taking the log value. It is expected that older, more established firms are more likely to gather information about international operations (Zahra, Reference Zahra2003) and sophisticated management skill (Thornhill & Amit, Reference Thornhill and Amit2003). These intangible resources will affect the firms’ entry decisions.
R&D intensity
R&D intensity is calculated as the ratio of the firm's R&D expenditure to its revenue. Companies with high R&D intensity tend to choose WOSs as their entry mode (Demirbag, Glaister, & Tatoglu, Reference Demirbag, Glaister and Tatoglu2007). Therefore, the researchers include this variable to control for its possible influence on entry mode choice.
Advertising intensity
Past research has found that advertising intensity influence multinational enterprises (MNEs’) entry mode choice (e.g., Demirbag, Glaister, & Tatoglu, Reference Demirbag, Glaister and Tatoglu2007). Advertising intensity is calculated by the ratio of the firm's advertising expenditure to its revenue. MNEs with higher advertising intensity are found to prefer WOSs or high equity entry mode to exploit their competitive advantages in foreign markets (Demirbag, Glaister, & Tatoglu, Reference Demirbag, Glaister and Tatoglu2007). Consequently, the researchers included this variable to control for its possible effects on entry mode choice.
Debt ratio
Debt ratio is calculated by dividing total debt by total assets. A company's debt ratio reflects its degree of financial leverage and signals its current status of riskiness. Li and Meyer (Reference Li and Meyer2009) find that debt ratio is an influential factor to a firm's entry mode. In their study, companies with high debt ratio prefer JV. For this reason, the researchers include debt ratio as a control variable.
Export intensity
Export intensity is defined as the ratio of the firm's export revenue to its total revenue. Prior research has found that export intensity influences MNEs’ FDI decisions – export intensity is positively associated with the choice of WOSs (Chang, Reference Chang1995; Terpstra & Yu, Reference Terpstra and Yu1988; Li & Meyer, Reference Li and Meyer2009). The researchers thus included this variable to control for its possible effects on entry mode choice.
Legal restriction
Legal restrictions on share ownership by foreign firms have frequently been investigated in entry mode research. The results have rather consistently shown that companies are much less likely to establish a WOS in a country with legal restrictions (e.g., Gatignon & Anderson, Reference Gatignon and Anderson1988). Sometimes local governments may not even allow foreign MNEs to do so. In addition, a correlation is likely to exist: some countries with poor governance quality may also have high legal restrictions. Therefore, the researchers included this variable to control for its possible effects on entry mode choice. This variable is a dummy variable, with the value 1 representing legal restrictions imposed on foreign investments by local government and the value 0 representing no legal restrictions.
Industry
The researchers created dummy variables to represent the industry in which a firm participates. Industry categories controlled include electronics, plastics, automobile, and shipping and transportation. (The electronics industry includes semiconductor, computer and peripheral equipment, optoelectronics, communications and internet, electronic parts and components, information services, and other electronics in Table 2.) Electronics firms account for the largest proportion of the sample (67%), and Chiao, Lo, and Yu (Reference Chiao, Lo and Yu2010) find that electronics firms in Taiwan typically put more effort on protecting proprietary assets and thus tend to choose WOSs as their entry mode, so the electronics industry is included. In addition, regulations in the automobile industry, plastic industry, and transportation industry are generally stricter. The researchers thus include these industries as control variables for possible influence on entry mode choice.
Method
Since the dependent variable (entry mode) is dichotomous, we used binary logistic regression to test our hypotheses. This statistical method is commonly used in the research of entry mode (Canabal & White, Reference Canabal and White2008). The regression models are formally expressed as P(yi = 1) = 1/(1 + exp(α + X i β)), where yi is the dependent variable, Xi is the vector of independent variables for the ith observation, α is the intercept parameter, and β is the vector of regression coefficients (Amemiya, Reference Amemiya1994). The dependent variable takes a value of 1 if an observed investment is undertaken via a WOS. Hence, a positive coefficient in the regression models indicates that the specific independent variable increases the probability of the WOS choice.
RESULTS
In the following sections, the researchers first describe the correlations and then report results of the binary logistic regression and family control's effect on entry mode choice.
Descriptive statistics and correlations
The result of descriptive statistics shows that 72% of the observed investments were undertaken via the WOS arrangement. The high proportion of WOS entries is in accordance with Child and Tse (Reference Child and Tse2001), who found that the majority of outward investments carried out by emerging-market multinationals were via WOSs. In addition, in our sample, the average percentage of board seats held by family members is 53%, which indicates commonly high family control among Taiwanese firms. The result is in line with the findings of Claessens, Djankov, and Lang (Reference Claessens, Djankov and Lang2000), La Porta et al. (Reference La Porta, Lopoz-de-Silanes, Shleifer and Vishny1999), and Yeh, Lee, and Woidtke (Reference Yeh, Lee and Woidtke2001).
Table 3 reports the means, standard deviations, and the correlations matrix for the variables. To prevent multicollinearity, the continuous variable used to construct interaction terms, family control, is mean-centered (Aiken, West, & Reno, Reference Aiken, West and Reno1991). We report non-centered means and standard deviations in the correlation matrix to simplify interpretation. The correlations between variables are all low, with the greatest value of 0.34.
Table 3 Means, standard deviations, and correlations
R&D = research and development.
*p < .1; **p < .05; ***p < .01.
The influence of family ownership on entry mode choice
Using binary logistic regression, the researchers further analyze the influence of family control on entry mode. The results are shown in Tables 4–5.
Table 4 The influence of family control on entry mode choice
HPU = high perceived uncertainty; R&D = research and development.
*p < .1; **p < .05; ***p < .01.
Table 5 The influence of family control on entry mode choice
LPU = low perceived uncertainty; R&D = research and development; VIF = variance inflation factor.
**p < .05; ***p < .01.
Models in Table 4 test the entry mode preference when firms with different degrees of family control perceive high uncertainty. Model 1 considers the direct impact from independent variables without including control variables. The LR χ2 value of Model 1 is 89.63 (p < .001), indicating high explanatory power of the model. The coefficient of the term High Perceived Uncertainty in Model 1 is −0.48 (p < .001), showing a negative impact on entry mode. The result indicates that all firms tend to choose JVs when perceiving high uncertainty. The coefficient of the interaction term Family control × High Perceived Uncertainty in Model 1 is −0.01 (p < 0.1), showing a negative impact on entry mode (i.e., more likely to choose JV). The result still holds when including all control variables as shown in Model 2. The results indicate that, when perceiving high uncertainty, firms with higher family control are more likely than firms with lower family control to choose JVs. Hypothesis 1 is thus supported.
Models in Table 5 test the entry mode preference when firms with different degrees of family control perceive low uncertainty. Model 1 considers the direct impact from independent variables without including control variables. The LR χ2 value of Model 1 is 78.45 (p < .001), indicating high explanatory power of the model.
The coefficient of the term Low Perceived Uncertainty in Model 1 shows a positive impact on entry mode, indicating that all firms tend to choose WOSs when perceiving high uncertainty. The coefficient of the interaction term Family control × Low Perceived Uncertainty in Model 1 is 0.02 (p < .01), showing a positive impact on entry mode (i.e., more likely to choose WOS). The result still holds when including all control variables as shown in Model 2, confirming that, when perceiving low uncertainty, firms with higher versus lower family control are more likely to choose WOSs. Hypothesis 2 is thus supported.
DISCUSSION
This study finds that family control indeed influences the entry mode choice – firms with higher family control respond more drastically to environmental uncertainty, in brief. With higher perceived environmental uncertainty, they are more likely than firms with lower family control to choose JVs. However, they have a higher propensity than firms with lower family control to choose WOSs when they perceive low uncertainty. This finding brings a different perspective to the conventional wisdom of the influence of family control on firms’ internationalization.
Extant research has mostly found that family control leads to more conservative or risk-averse decisions in internationalization. For example, family-controlled firms are found to be in favor of countries that are culturally and geographically close (Johanson & Vahlne, Reference Johanson and Vahlne1977), choose the lower commitment entry mode (Kumar & Subramanian, Reference Kumar and Subramanian1997), or hold lower equity ownership stakes of their foreign subsidiaries (Kumar & Subramanian, Reference Kumar and Subramanian1997; Filatotchev et al., Reference Filatotchev, Strange, Piesse and Lien2007). Compared with firms controlled by other types of shareholders, family-controlled firms appear to shy away from internationalization (see Kontinen & Ojala, Reference Kontinen and Ojala2010 for a review).
However, our study suggests that this traditional wisdom may only tell half of the story. Indeed, when uncertainty is high, family control drives the firm to choose the lower commitment entry mode – our empirical results suggest the same as earlier research predicted. However, our research further finds, family control facilitates internationalization when uncertainty is low – firms controlled by family shareholders act more aggressively than their non-family counterparts to assume full equity ownership of their foreign subsidiaries. This finding reveals the unexplored aspect of family control's influence on internationalization decisions.
Gomez-Mejia, Makri, and Kintana (Reference Gomez-Mejia, Makri and Kintana2010) have also found similar patterns in family-controlled firms’ internationalization decision making. They examined diversification decisions of firms controlled by family members, finding that on average, family firms diversify less internationally than non-family firms. However, in contrast, family-controlled firms are more likely than non-family firms to engage in international diversification as business risk increases. Both Gomez-Mejia, Makri, and Kintana (Reference Gomez-Mejia, Makri and Kintana2010) and this study find that family-controlled firms, under certain circumstances, appear to make more aggressive internationalization decisions than firms with other controlling shareholders. This commonality brings significant implications to the research of family businesses, which we will elaborate further later.
Conclusion
This study brings implications and contributions to both family business research and international business studies, but it has limitations as well. The researchers discuss the details below.
Implications for family business research and future research directions
For the research stream of family businesses, the current study takes into consideration both risk preference and desire to control, providing a more comprehensive view on family businesses’ internationalization. Past research on family businesses has focused primarily on the risk factor, finding that family firms are more conservative (i.e., risk averse) (e.g., Filatotchev et al., Reference Filatotchev, Strange, Piesse and Lien2007) when making their internationalization decisions. This study contributes to family business research by finding the unexplored aspect of family firms’ internationalization – family members’ high desire to control, in addition to risk aversion, has a significant effect on internationalization decision making as well. When perceiving low uncertainty, family-controlled firms make more aggressive decisions than their non-family counterparts to assume full equity ownership of their foreign subsidiaries.
The phenomenon deserves further inquiries. For example, it will be of great interest to explore in what contexts family-controlled firms act more aggressively, or, on the contrary, act more conservatively? Will we observe similar behavioral duality brought by family control in other internationalization decisions, besides international diversification and entry mode choices? Whether the behavioral duality can also be observed when family-controlled firms make other strategic decisions? Further research in this direction will benefit our understanding of the influence of family control on internationalization and on other strategic decisions.
Furthermore, whether high responsiveness to environmental uncertainty is beneficial or detrimental to a firm's performance is still unclear. Responsiveness to environmental uncertainty may arise from two different sources – better capabilities of sensing the environmental conditions or over-emphasis on the socioemotional objectives. If the former is the cause, responsiveness to environmental uncertainty may have a positive influence on firm performance. However, if the latter is the reason, then responsiveness to uncertainty may shadow the rationality of decision making and hurt firm performance.
For family-controlled firms and policy makers, responsiveness to environmental uncertainty brought by family control has profound managerial implications. Firms controlled by family shareholders have to trace the root cause of their responsiveness to uncertainty, then they will be able to improve their performance, either by maintaining their sharpness of sensing the environmental conditions or, on the contrary, by avoiding emphasizing socioemotional motives to reduce their side effects. For policy makers, especially government officials who wish to attract family-controlled MNCs to invest in their country, it is especially important to improve the institutional environment to tackle these firms’ concerns on uncertainty.
On the other hand, the current study's contribution is also limited due to insufficient inquiries on family members’ desire to control. Currently, studies on family firms’ desire to control have primarily focused on the socioemotional motives (Gomez-Mejia et al., Reference Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson and Moyano-Fuentes2007; Gomez-Mejia, Makri, & Kintana, Reference Gomez-Mejia, Makri and Kintana2010), arguing that family firms maintain high control over their companies because they intend to preserve socioemotional wealth (De Vries, Reference De Vries1993; Habbershon & William, Reference Habbershon and Williams1999; Thomsen & Pedersen, Reference Thomsen and Pedersen2000; Anderson & Reeb, Reference Anderson and Reeb2003). The researchers expect that motives other than socioemotional wealth, economic motives for instance, can also lead to family shareholders’ high desire to control. The researchers thus call for further inquiries on this aspect to advance our understanding of family control.
Implications for international business studies
This study also contributes to international business studies. Examining family control's influence on MNCs’ entry mode decision, the researchers found that investing company's corporate governance matters. Specifically, the researchers found that investing firms’ controlling shareholders and board members are influential to the firm's internationalization decisions. The majority of past studies on internationalization decisions, including entry mode, location choices, and subsidiary ownership strategy, have focused on the influence of firm characteristics related to resources and capabilities (e.g., firm size, R&D intensity), leaving corporate governance variables mostly untapped. This study fills the gap.
However, the researchers only examine the influence of family control, leaving other aspects of corporate governance unexplored. The influence of other types of controlling shareholders, such as government bodies, institutional investors, etc. has yet been investigated. It will also be intriguing for future research to examine the influence of other corporate governance variables on firms’ internationalization. Factors such as board composition, board structure, and top management team demography may also affect a firm's internationalization decisions and consequently the firm's performance. There is vast room for these factors to be fully probed.
Using a sample of 1,644 publicly listed companies in Taiwan, this study examines how family control influences MNCs’ entry mode choice. The empirical results indicate that firms with higher family control react more drastically to environmental uncertainty when choosing their entry mode. This study makes contributions, but it is just a beginning. Nonetheless, the researchers hope the current study will inspire more related inquiries. After all, we IB scholars know too little of how corporate governance influences MNCs’ internationalization.
Acknowledgment
The authors thank the Editor, Professor Peter Galvin, and two anonymous reviewers for helpful comments on earlier versions of this manuscript. Any mistakes in this manuscript are the sole responsibility of the authors. They are also grateful to the National Science Council of Taiwan for the financial support (NSC 101-2410-H-030-068-).
