Family Heterogeneity and Entry Mode Choice: a bifurcation
bias approach
Jens Nieuwenhuis
(S2960907)
Supervisor: A. Kuiken
Co-assessor: Dr. J. Shin
University of Groningen
June 2020
AbstractFamily firm internationalization has been a substantial area of study in the past years. A point that has been of less interest during this time is the entry mode choice of family firms. This particular choice can be of extreme importance when entering a foreign country, as changing it later can be costly and difficult for family firms. In this study the focus is on family firm heterogeneity and its effect on entry mode decision. Using the bifurcation bias perspective, derived from the transaction costs economics theory, this study looks at the effects of family ownership and family CEO on entry mode choice. Moreover, this relationship will be moderated by the corruption level differences between home and host country. Using a longitudinal logistic regression on 268 entry mode choices of 65 US firms this study finds that family CEOs have a significant influence on entry mode decisions, being inclined to choose wholly owned subsidiaries, whereas ownership percentage does not indicate to have any influence on this choice. This could indicate that the percentage of ownership is not necessarily crucial, but having direct control through a family CEO might.
TABLE OF CONTENTS
1. Introduction
2
2. Literature Review
5
2.1 Transaction Cost Economics
5
2.2 Entry Mode Choice
6
2.3 Family Firms
8
2.4 Family Firm Heterogeneity
9
2.5 Bifurcation Bias
10
3. Development of Hypotheses
13
3.1 Family Ownership13
3.2 Family CEO14
3.3 Moderation of Corruption15
4. Methodology
19
4.1 Sample19
4.2 Dependent Variable19
4.3 Independent Variable20
4.4 Moderator Variable20
4.5 Control Variables20
4.6 Method21
5. Results
21
6. Discussion
25
7. Limitations and Future Research
26
8. Conclusion
27
INTRODUCTION
A couple of the largest and most recognizable firms in the world are Walmart, Volkswagen, Ford and Tata. Besides their success there is something else they have in common, that being that they all are family owned firms. Family owned firms account for two-thirds of all firms on the globe, (De Massis, Frattini, Majocchi, & Piscitello, 2018) and contribute between 70-90% of global GDP (De Massis et al., 2018). They vary in size, from small and medium enterprises to global multinationals. What makes them stand out mostly from other firms is their ownership by a family, instead of multiple individuals. Moreover, this ownership structure influences key strategic decisions of internationalization (Pongelli, Caroli, & Cucculelli, 2016), and compared to nonfamily firms they possess a unique governance mode (Majocchi, D’Angelo, Forlani, & Buck, 2018). The degree of how much the family ownership and governance structure influence the key strategic decisions depends on the characteristics of family firms and how much they differ on this between each other, known as the degree of heterogeneity.
Being widespread and having a certain uniqueness in their governance has led to family firms having been the topic of research for many disciplines. The international business field being one of them, in which the family firm has been the topic of firm internationalization. Family firm internationalization has been widely studied over the years, with mainly research questions ranging from if family firms internationalize to why and how they do this (De Massis et al., 2018).The different strategic decision-making processes involved in internationalization have seen plenty of attention. Most of the studies so far have focused on the distinction between family and non-family firms by using a binary code (Verbeke, Yuan, & Kano, 2019) and either with the intention to see if family firms internationalize less (Gomez-Mejia, Makri, & Kintana, 2010) or more (Hennart, Majocchi, & Forlani, 2019; Miller, Le Breton-Miller, & Scholnick, 2008) than non-family firms, which provided mixed results. A meta-analysis by Arregle, Duran, Hitt, & van Essen in 2017 even found the difference between the levels of internationalization between family firms and non-family firms to be null. And while this dichotomous focus has yielded some interesting results, theories and further avenues of research (De Massis et al., 2018), it also does not outline the full scope of family firm research. In using this dichotomous approach literature has identified unique family firm features and governance (Verbeke et al., 2019). However, this does not mean that every family firm has the same amount and extend of these features, which in turn suggests that this influences the strategic decision making. In order to map out the effect of different magnitudes of family firm features on strategic decisions a variety scale is needed where only family firms are being used (Majocchi et al., 2018). By moving away from the traditional method towards this heterogeneous scale of family firms, literature is able to develop new theories and methods that are solely grounded in family firm research and not as an adaptation of “mainstream” non-family business research.
some disadvantages involved with family ownership, two of the most influential ones being the need to keep control within the family, which limits access to outside capital, and favoring family members for positions in the firm while better options might be available outside the family (Hennart et al., 2019). The asymmetric treatment of family members is called the bifurcation bias (Verbeke & Kano, 2012).
These characteristics of family firms make them unique compared to non-family firms, but can also make them unique compared to other family firms. This is called family firm heterogeneity and can be defined as the variation in family firm behavior and performance, that stems from the combination of governance structures, resources and goals in an idiosyncratic way, which are a result of varying degrees of overlap between business and family systems (Chevy Fang, Kellermanns, & Eddleston, 2018; Chua, Chrisman, Steier, & Rau, 2012). This means that not every family firm characteristic is equally present in every family firm. And therefore, not every family firm looks and acts the same, has different levels of performance and different goals. Family firm heterogeneity has even been suspected to be one of the main reasons why prior family firm research, using a dichotomous approach, is not conclusive (Boellis, Mariotti, Minichilli, & Piscitello, 2016) as the differences between family firms are potentially greater than those between family and nonfamily firms (Chua et al., 2012). For this reason it is important to identify the sources of heterogeneity in family firms (Chua et al., 2012) and how these affect strategic decision making (Dibrell & Memili, 2018).
One such area that is worthwhile to study using a heterogeneity focus is the internationalization of family firms. And more specifically one part of the internationalization process that is perhaps the most detrimental of all: the entry mode choice (Shen, Puig, & Paul, 2017). Using a transaction costs economics (TCE) theory perspective, entry modes are the governance structures used in a foreign country that minimizes costs associated with entering and operating (Brouthers, Brouthers, & Werner, 2003; Canabal & White, 2008). Because different entry modes implicate different levels of control in the foreign market, differences in the resources that need to be committed, and different levels of risks, selecting the ‘right’ entry mode is one of the more important strategic decisions that have to be made when internationalizing, as changing it later can be difficult and costly (Hennart & Slangen, 2015). Selecting the ‘right’ entry mode is perhaps even more important for family firms, as they often have limited resources at their disposal, which could make changing later on a costly endeavor (Maekelburger, Schwens, & Kabst, 2012). Moreover, family firm heterogeneity can influence entry mode decisions in different ways, depending on how many and to what extent heterogeneity factors are present in the firm (Boellis et al., 2016). However, how exactly these heterogeneity factors influence entry mode choice has not yet been a topic of research often.
heterogeneity at the time, such as family ownership, founder ownership, family/external managers (Boellis et al., 2016; Pongelli et al., 2016). Here most findings indicate that family heterogeneity causes the firm to use equity entry modes (Boellis et al., 2016; Pongelli et al., 2016). Most of these family firm heterogeneity studies use socio emotional wealth (SEW) to explain family firm internationalization
At the same time another theory that might be potent in explaining family firm heterogeneity and therefore family firm internationalization is the bifurcation bias, which is can be viewed as an straight family extension of the transaction costs economics (Arregle et al., 2017). Family firms having a high bifurcation bias view family members working in the firm as stewards of the firm, while nonfamily employees are viewed as agent. This means that family members are viewed as very valuable to the firm and its operations, while nonfamily members are viewed as self-serving and opportunistic regardless of their performance or capabilities (Verbeke & Kano, 2012).
Moreover, the institutional environment of the host country is a factor that can play a big part in entry mode decisions (Cuervo-Cazurra, 2006). One aspect of the institutional environment that might be particularly important for family firms is corruption, as any allegation of corruption can have negative impacts on the reputation of the family firm (Rabbiosi & Santangelo, 2019), while reputation is something that is held in high regards in most family firms (Cheng, 2014). Furthermore, corruption differences between home and host country raises barriers to entry (Rabbiosi & Santangelo, 2019) as well as making it more difficult to operate in the host country, as unfamiliarity with the host country increases (Cuervo-Cazurra, 2006) Alongside this corruption increases the operating costs in the host country as uncertainty of operating increases (Cuervo-Cazurra, 2006) This can be impeding for family firms, given that they are often associated with less professionalism (Block, Fisch, Lau, Obschonka, & Presse, 2016) and therefore do not always make the “correct” choices. It is therefore important to expand the international family business with corruption, as there exist different perceptions on the illegality of corruption and it is still a widespread practice in most developing markets (Cuervo-Cazurra, 2016).
In this study the entry mode choice of family firms will be examined using a bifurcation bias perspective. Using the bifurcation bias, the effect of family firm heterogeneity on entry mode choice will be studied, along with the moderation of corruption levels in the host country.
business such as the institutional environment. This is something that has not yet been done, and could potentially provide more complete answers than the transaction costs economics can at the moment.
This research will help to further map out the effect of characteristics of family firms on entry mode decisions from a perspective solely focused on family firms. As of yet the number of studies studying the effects of family firm heterogeneity on entry mode choice is limited. This study will help to expand this literature. Moreover, perhaps even more important, this study will help to decrease the gap between family firm literature and international business literature, which is crucial as noted by Kano & Verbeke (2018). By using family firms as a starting point and integrating this with international business literature ensures that this study is relevant for both areas of literature and thereby helps to bridge the gap between both areas a bit further. The main contribution of this study is that it shows the importance of family CEOs as a form of heterogeneity.
Studying the bifurcation bias and family heterogeneity is not only theoretically important, but can have some managerial implications as well. First of all, being aware of how much bifurcation bias is present in the family firm is the first step in accounting and controlling for it. Moreover, it is useful to know which factors of the family firm heterogeneity influence the effect of the bifurcation bias and what the consequences of this influence can be. One such scenario presented here in the form of entry mode selection. Knowing the potential pitfalls and how to prevent them can limit bad performance.
The remainder of this paper will take the following form. First, it will discuss transaction costs economics and entry mode choices from a TCE perspective, followed by an overview of family firm literature and family firm heterogeneity and the bifurcation bias. After this, hypotheses will be developed on the effect of family ownership and family CEOs on entry mode, moderated by the host country corruption levels. These hypotheses will be tested using a sample of 268 acquisitions and joint ventures by 65 US family firms, using a longitudinal logistic regression. After discussing the results this study will conclude with a discussion of the results, limitations and areas for future research.
LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT
Transaction cost economics
Furthermore, there are three factors that impact transaction costs further: asset specificity, bounded rationality and opportunism (Verbeke & Kano, 2012). First, asset specificity refers to how easily assets used in a transaction can be redeployed in another transaction. Higher asset specificity leads to higher transaction costs as the dependency between the transaction partners increases, which forces both partners to invest more time in searching for the right partner that won’t back out of the transaction and leaves the other with a transaction specific asset that is not usable in other settings (Verbeke & Kano, 2012). The second and third factors (bounded rationality and opportunism) are closely related to asset specificity. Bounded rationality means that it is assumed economic actors behave rational, but have limited capacity to make optimal rational choices (Verbeke & Kano, 2012). This increases transaction costs because there will always be some uncertainty, as you cannot possibly know everything. As a firm you have to account for this uncertainty. An example of this would be when two parties try to formulate a contract. All contracts are unavoidably incomplete (Williamson, 1981), in order to make the contract more complete one needs to invest more time and money in finding out what is missing and how to plug these holes in the contract. This increases transaction costs. Moreover, opportunism, the self-interested behavior of transaction partners, will lead to higher transaction costs as there is a greater need to monitor the partner for undesirable behavior. This assumption however has been criticized for not portraying reality (Verbeke & Greidanus, 2009; Verbeke & Kano, 2012). Therefore, Verbeke & Greidanus (2009) introduce the concept of bounded reliability, which adds to opportunism an explanation for situations where self-interested behavior does not lead to failed transactions.
Entry mode choice
Market entry mode has been a topic of discussion for a long time (Schellenberg, Harker, & Jafari, 2018). The reason for this is that it is widely recognized within international business literature that the strategic success or failure of a firm is determined by the entry mode chosen (Crick & Crick, 2015; Hennart & Slangen, 2015; Schellenberg et al., 2018). Although literature agrees upon what market entry mode is, literature has failed to find a consistent definition and has seen inconsistency in the approaches used for studying entry mode as well as inconsistency in the reasoning for why one entry mode will be selected over another (Schellenberg et al., 2018). A viewpoint that is mostly used by entry mode scholars is of treating entry mode as a selection that is made between distinct alternatives and from there the effect of the choice and implications of the choice are researched (Nisar, Boateng, Wu, & Leung, 2012; Olejnik & Swoboda, 2012). On the other hand, an alternative viewpoint that is adopted by scholars is that of external and internal factors affecting the entry mode, and should therefore be the focus of attention (Hennart & Slangen, 2015; Shaver, 2013). External factors include things such as culture, environmental uncertainty, legal environment, whereas internal factors include asset specificity, international experience and so on (Schellenberg et al., 2018). More and more entry mode scholars call for a shift in focus from what the entry mode choice is, to what the processes of making the entry mode choice is, how the characteristics of key decision makers affect these processes (Canabal & White, 2008; Hennart & Slangen, 2015).
(environment) and with the company itself is important, however what determines this fit can depend on a wide array of factors (Yiu & Makino, 2002).
Over the years literature has defined two categories of entry modes, which are non-equity (exporting, licensing and franchising) and equity modes, which can be wholly-owned subsidiaries (greenfields and acquisitions) or joint ventures (Canabal & White, 2008). Each type of entry mode implies different degrees of ownership and control, with non-equity and joint ventures having shared ownership and control and greenfields and acquisitions having full ownership and control. Moreover, the degree of control also determines the level of (environmental) uncertainty, the exposure to risks and the amount of resources that need to be committed to the entry mode (Herrmann & Datta, 2002).
In the entry mode literature several theoretical frameworks have been brought forth in attempting to explain entry mode choices of firms. The main theoretical frameworks used are the institutional theory, the Uppsala model, the resource-based view (RBV), the OLI model and transaction cost economics (TCE). These different frameworks are each able to contribute something to entry mode research (Schellenberg et al., 2018), however they each also have limitations. For example, the institutional theory is criticized for being not able to be a standalone theory and therefore would need the help of other frameworks (Brouthers, 2013). From early on there have been calls for combining the institutional theory with the TCE, as institutions are the ‘rules of the game’ in which transactions occur (North, 1990; Schellenberg et al., 2018).
In their review on entry mode literature Schellenberg et al. (2018) found three main gaps in the literature. The first gap they found is the focus of entry mode literature on multinational enterprises (MNEs) rather than small and medium sized enterprises (SMEs). An expansion of focus is needed here as SMEs differ significantly from MNEs regarding ownership structures, financing, personnel and management characteristics (Schellenberg et al., 2018). Studies that do decide to investigate SMEs often resort back to the traditional theoretical frameworks and simply swap MNEs for SMEs. This in turn has received criticism, as it does not take into account the unique characteristics of SMEs (Shaver, 2013). Another gap that is prevalent, is that most literature is focused too much on encouragements and discouragements that are present in the host country, but too little on the internal processes in the firm that lead to certain entry mode decisions (Schellenberg et al., 2018). The argument is made that entry mode research has been answering what happens (the entry mode choice), but too little emphasis is put on the why and how of the entry mode choice (Shaver, 2013). What are the internal factors that lead to this decision and how is that decision process structured? Another point made is that more theories need to be integrated, as no one theory can explain entry mode decisions fully (Canabal & White, 2008; Schellenberg et al., 2018).
lowest costs. However, the market is not always perfect, and in most cases opportunism and bounded rationality are prevalent. In a situation with higher transaction costs the firm could be better off internalizing the production into the firm. Most literature finds that non-equity entry modes are preferred when market uncertainty is high, as the risks of entry get shared with a partner firm and therefore uncertainty can be handled better (Bruneel & De Cock, 2016). However, when opportunistic behavior of a potential partner is high, an entry mode with high levels of control is preferred. Such a situation might be present in countries with ineffective property right protection or different cultural backgrounds (Schellenberg et al., 2018). Furthermore, high asset specificity is found to increase the likelihood of an entry mode with high levels of control, as this allows the firm to have more control over how their knowledge is used (Yiu & Makino, 2002).
In this study the focus will be on wholly owned subsidiaries and joint ventures as these require more commitment and strategic decision making compared to non-equity entry modes. Furthermore, as Canabal & White (2008) found in their review on entry mode literature, the most commonly used method for studying entry modes is by using a dichotomous variable classifying wholly owned subsidiaries and joint ventures. By adopting the same method as a majority of other studies, this study can be compared and evaluated with those prior, with results being meaningful. The general consensus on JV or WOS in the literature of TCE, is that higher transaction costs lead to an increase in the chance that a WOS will be selected as the entry mode. The reasoning for this is that as opportunistic behavior increases the costs of monitoring the partner firm for any violations of the contract and opportunistic behavior will also increase. Another determinant for a JV or a WOS is the nature of the relationship and the transferability of the assets (Sestu & Majocchi, 2020). If both firms complement each other than a JV might be preferred, as the chance for opportunistic behavior is low. At the same time, if assets in a firm are hard to transact, it might be preferable to use a JV as without the help of the partner firm the full potential of the asset will not be reached (Sestu & Majocchi, 2020)
Family Firms
Family firms are firms in which, as the name implies, family plays an important role. What role that is depends in each firm, however all family firms are to some extend owned by at least one family. It can vary greatly how much of the firm is owned by the family, how many families own the firm, if the founder still owns the firm or the founder’s descendants do, and so on. Family firms also greatly vary in how involved the family is with the firm. Some family firms are only family owned, while in others the family actively partakes in daily activities in the firm, most commonly it being in a management function (Barontini & Bozzi, 2018). Family firms operate in a wide variety of industries, from high-tech industries, transporting, wholesale, retail to banking, insurance and capital intensive industries (Cheng, 2014).
generations. This also means that family firms in general have less access to outside funds, as most often gaining outside funds means giving up control over the firm. As family firms merely indicates a type of governance structure there are many points on which family firms can differentiate from each other and from non-family firms, while still being able to operate in the same way as non-family firms and in the same industries. This also means that family firm research has long been seen as a variety of ‘normal’ business literature and not as its own discipline (Holt, Pearson, Payne, & Sharma, 2018). Therefore, there is still a large gap between family business research and practice (Holt et al., 2018), meaning that many of the business research so far has to be tailored to a family business discipline.
Most family firm research is focused on the agency problem found in family firms (Cheng, 2014). In this regard literature looks at how the close involvement of family owners on day to day activities affects monitoring in the principal-agency relationship, the clash between long investment horizons of family owners and short investment horizon of non-family members, and reputation building and maintaining (Cheng, 2014).
Other important topics in family business research are: the interaction between family firms and the institutional context (Soleimanof, Rutherford, & Webb, 2018), the role of non-family members in family firms (Tabor, Chrisman, Madison, & Vardaman, 2018) and the social aspects of family firms (Holt et al., 2018). These topics of family research do not only provide interesting new insights and propose new avenues of research, but have also been argued to be able to extend in a family firm setting to a non-family firm setting (Cheng, 2014; Holt et al., 2018; Tabor et al., 2018).
Within the literature on family firms there are different views of degree of family firm, which is mostly caused by the heterogeneity of family firms, which will be discussed in more detail below. Some researchers ascribe huge potential to family firms and their uniqueness, while others see the family aspect of family firms as more of a hinderance or give more weight to the pitfalls of the family aspect. This is evidenced by the views that family firms are sustainable, sociable, trustworthy, traditional and less profit oriented, but are also associated with being inflexible and having a lack of professionalism (Block et al., 2016). A explanation for why the characteristics of family firms are at the same time preferable in a firm while at the same unpreferable, is that next to the focus on economic goals family firm are known to focus on non-economic goals (Kano & Verbeke, 2018). Every firm has economic goals, however family firms have a significant focus on non-economic goals, which don’t necessarily have to be different from each other (Kano & Verbeke, 2018). This provides an interesting interaction within family firms, as family firms need to find a balance between this duality.
Family Firm Heterogeneity
firm heterogeneity has been one of the main reasons why family firm behavior research is inconsistent, as the heterogeneity leads to less predictable behavior (Rau, Schneider-Siebke, & Günther, 2019).
The aspects on which family firms can differ from each other is basically infinite, however it is also obvious that not every small difference is cause for investigation. Some causes for heterogeneity are the goals, governance structures and resources that are present in the family firms (Chua et al., 2012). These result in some major aspects that seem to be influential on family firm performance and internationalization. First, family firms can vary a great deal in the ownership percentage that is in hands of the family. In some cases the firm is completely owned by the family, in other cases there are multiple owners with the family being the controlling shareholder, while in other cases the family owns a minority stake. These different levels of ownership can have great implications for decision making activities.
Furthermore, family firms tend to prefer to hire family members, as this gives a sense of alignment. The function in which this is best reflected However, in order to grow, family firms are often forced to hire non-family members as family members are finite resources (Chrisman, Memili, & Misra, 2014; Tabor et al., 2018). Families are limited in their size and in their capabilities, hiring outside of the family provides more candidates that are better suited for the job than anyone in the family. Hiring non-family members can have benefits but at the same time can create friction between non-family employees and family employees (Verbeke & Kano, 2012).
Bifurcation Bias
One explanation of family firm heterogeneity is the concept of the bifurcation bias. Developed from a TCE view on family firms, Verbeke and Kano (2012) provide new insights on the bifurcation bias and human assets found in family firms. A bifurcation bias derives when family employees are treated as highly valuable, firm specific assets and nonfamily employees are treated as substitutable, short term assets (Verbeke & Kano, 2012) In order to better understand the bifurcation bias in family firms, it is needed to have a look at the theories that lay at the foundation of the bifurcation bias theory. Verbeke and Kano base their theory for extending the TCE on the agency theory and the stewardship theory, two theories that are often found in family firm research. The agency theory deals with conflicting goals between owners (principal) and managers (agents), where the goal is to minimize the agency costs, by aligning the interests of the managers with that of the owners (Fama & Jensen, 1983). The idea early on was that family firms do not face agency issues as ownership and control are almost exclusively in the same hands and monitoring becomes easier as economic and personal relationships are closely linked (Pollak, 1985). However, more recent studies suggest that family firms may indeed not face the same challenges, but instead face unique challenges of their own (Verbeke & Kano, 2012). The stewardship theory is the second theory that the bifurcation bias is grounded in. This theory takes an opposing view from the self-serving view of the agency theory, and argues that managers act in the interest of the firm (Davis, Schoorman, & Donaldson, 1997). Family firm scholars have used this point of view to highlight the unique competitive strengths of family firms. Arguments are made that because of sharing the same name, being related and therefore sharing a family history propels the family managers to act in the interest of the firm instead of seeking self-gain.
and Kano (2012) argue that both theories can be observed simultaneously in family firms, with stewardship thinking being applied to family employees and agency thinking being applied to non-family employees. This is known as the bifurcation bias. In the context of non-family firms this is expressed through the asymmetric treatment of family and non-family managers (Verbeke & Kano, 2012), where family members are viewed as stewards of the firm and nonfamily members as an agent that needs to be monitored closely(Verbeke & Kano, 2012). The degree to which bifurcation bias happens differs, but due to the importance of non-economic goals, bifurcation bias might be inevitable. From a TCE approach bifurcation bias can be explained as an expression of bounded rationality, where the bifurcation bias prevents an objective analysis between family member and non-family member assets.
Viewing family members as stewards of the firm stems, according to Kano and Verbeke (2012), from the fact that family members have undergone a long-term socialization process within the firm, which facilitates the transfer of tacit knowledge and social capital. Furthermore, this makes it more likely that family members are engaged with the economic and noneconomic goals that are often present in family firms (Chua et al., 2012). In this regard family members are seen as family based human assets. Family based human asset specificity refers to the unique employment base that family firms have access to (Verbeke & Kano, 2012), indicated by the long term socialization process. From an early age family members are socialized in the firm while the firm invests heavily in them for a long period of time. This leads to family members possessing idiosyncratic knowledge about the firm, that puts nonfamily members at a disadvantage, simply because they do not possess this knowledge (Verbeke & Kano, 2012).
The bifurcation bias is an expression of transaction costs economics in the following ways: Bounded rationality is observed in bifurcation bias through preventing an objective assessment between family-based and nonfamily-based assets (Verbeke & Kano, 2012). This can have two results on asset specificity. In the case of unproductive or bad family-based human assets this leads to the firm holding on to these assets for too long, which in turn enhances their ‘non-redeployablity’, meaning as time goes on it becomes harder to move these assets out of the firm. At the same time, by viewing valuable nonfamily human assets as inferior, family firms unintentionally enhance the ‘redeployability’ of these assets, as these assets will be driven away from the firm instead of securing their services for the long-run (Verbeke & Kano, 2012). This leads to opportunistic behavior of family members. As they are monitored less than nonfamily members and are more or less guaranteed to stay with the firm, opportunistic behavior can arise in the form of free riding and avoiding responsibilities (Verbeke & Kano, 2012).
misjudging the difficulty of the task agreed upon. This does not mean that others always not trustworthy, but are bound in the reliability they have (Verbeke & Greidanus, 2009)
The bifurcation bias can have lasting effects going both ways. First, by viewing incompetent and unreliable family members as stewards the firm risks bad performance and misbehavior, if these family members would be appointed to crucial function. This could have been avoided if these family members were critically evaluated (Martin, Gómez-Mejía, Berrone, & Makri, 2017). Second, by viewing non-family members as self-serving short-term agents, the firm risks alienating or even losing competent, well meaning, reliable employees that could have achieved great things with the firm. The degree of bifurcation bias will have severe impacts on the performance of the firm (Verbeke & Kano, 2012). Bifurcation bias can be manifested in a couple of ways, some of which are: asymmetric altruism (Chrisman, Chua, & Litz, 2004; Chua, Chrisman, & Bergiel, 2009; Corbetta & Salvato, 2004) and amoral familism (Dyer, 2006). Furthermore, bifurcation bias can manifest in recruitment, performance evaluation, and compensation systems (Verbeke & Kano, 2012).
Family firms need to balance the presence of family-based assets specificity while utilizing their unique human assets and try to avoid the dysfunctional effects of the same human assets. This can be done by training and educating family members in order to recognize and economize on the bounded rationality problems (Verbeke & Kano, 2012)
Bifurcation bias is not equally present in all family firms and the extent to which bifurcation bias can cause problems for family firms depends on a couple of factors such as firm size, technological complexity and organizational structure (Majocchi et al., 2018) Firms size enhances the bifurcation bias effects. Larger firms often rely more on professional non-family managers for expanding their scope and scale, however more bifurcation biased firms are limited in their ability to attract and motivate professional non-family managers, which in turn limits the family firm’s capacity to expand their scale and scope. Furthermore, in technological complex environments family firms are in need of specialized knowledge and high human asset specificity, however due to problems caused by bifurcation bias they experience a hard time recruiting and retaining external talent and instead are stuck with family-based human assets that often do not match the requirements. This problem is not as relevant in low technology environment, as in these cases the skills required are not highly specialized and easily found within the family. Moreover, the organizational structure of the firm can be greatly influenced by the problems caused by bifurcation bias. With an increasing complexity of the organizational structure there is more need for professional managers that create sophisticated routines, that allow the firm to grow and maintain this size and complexity. Due to the bifurcation bias, the family firm is limited in the external professionalism they can acquire.
DEVELOPMENT OF THE HYPOTHESES
Family Ownership
As explained before, the percentage of a family firm that is owned by the family differs with every firm and is one of the larger factors of family firm heterogeneity. This can be of great importance for family firm internationalization as there are strong suspicions that the ownership structures of firms have a significant effect on international strategic choices (Pongelli et al., 2016). In terms of the entry mode selection of family firms, due to the preference to keep control of the firm in the family, family firms would opt for an entry mode with a high level of control, which would mean a wholly owned subsidiary. If the ownership percentage of the family increases, it is more likely that bifurcation bias tendencies also become more pronounced. This can lead to an overestimation of the deployability and transferability of firm specific advantages (FSAs) from the home to the host market (Kano & Verbeke, 2018), which are most often family specific assets. Bifurcation bias causes these assets to be evaluated based on affect and not with rational, which leads to these assets to be overvalued by the owners (Kano & Verbeke, 2018). Furthermore, when choosing between the options to internalize foreign transactions or to use the market, the bifurcation bias will cause a firm to overestimate the need to internalize when it comes down to heritage assets, which causes the firm to push for more control over its activities in the foreign market, leading to high control entry mode options (Kano & Verbeke, 2018). This is further exacerbated by the inability to recognize foreign resources to bundle with heritage assets, preventing a bifurcation biased firm from seeking complementary assets and working together with local partners. This inability is caused by an irrational distrust of outsiders, who are extravagantly seen as opportunistic (Kano & Verbeke, 2018). An example of this is the reputation of family firms. This is often highly valued in family firms and something that has been built over the years in the home country, however is difficult to transfer to foreign market (Majocchi et al., 2018). Bifurcation biased firms underestimate the capabilities of local partners in helping to build a reputation in the host country and instead choose an entry mode that allows them more control over strategic decision making, while reducing opportunistic behavior of nonfamily members.
With higher family ownership percentages it is likely that a bifurcation bias is more present in the family firm, as more decision making power and control lies in the hands of one group with the same values and goals. Higher family ownership percentages allow more room to pursue noneconomic goals. When bifurcation bias increases it distorts the objective, rational observation of entry mode options, therefore further increasing bounded rationality, overvaluing family specific assets and overestimating the opportunistic behavior of outsiders. Therefore, in order to keep control over operations and strategic decision making high a family firm with high family ownership will choose for a wholly owned subsidiary over a joint venture.
Family CEO
A CEO is often seen as a pivotal part of the company, someone that has much influence within the company and is able to have a big impact on strategic decisions and performance, for better or worse (Lin & Hu, 2007). Over the years many studies have investigated the effect of CEO characteristics on firm performance, internationalization and other outcomes (Miller, Minichilli, & Corbetta, 2013; Tabor et al., 2018), in both a family firm and non-family firm setting. This indicates the believe that one person can shape the direction and therefore the future the firm will be going, and shows the importance literature has allocated to CEOs.
Simply said there are two options for the CEO position of a family firm. The first option is appointing a family member for this position. This is an easy and cheap option, as family members fit the firm’s culture and operational needs (Tabor et al., 2018), but at the same time family CEOs are associated with being a convenient option, as they are already present in the family and more or less are obliged by the family to work in the firm. Moreover, family members that fulfill the role of CEO are often viewed as unprofessional and not fit for the position and the general opinion is that there are better options available outside of the family. Unavailability of family members is one important determinant for hiring non-family CEOs. As explained before, family members are finite, and simply said when family firms run out of family members they are forced to hire outside of the family (Tabor et al., 2018). Although this might have a negative connotation in this context, non-family members are found to be instrumental for strategic decision making and expansion into new markets (Tabor et al., 2018). However, literature has found mixed results whether or not non-family CEOs are indeed better for non-family firm performance. Some reasons that have been brought forth as to why family firms are reluctant to appoint non-family CEOs are: a conflict of interest between family owners and non-family employees due to different goals, which may cause the non-family member to be reluctant to perform properly (Chrisman et al., 2014). This divide might be enhanced by a lack of formalization on the family firm’s part, which causes the family firm to ineffectively monitor and improperly (or not at all) reward employee performance (Tabor et al., 2018). Furthermore, due to differences between cultures, training provided to non-family employees by family firms is limited in breadth and depth (Tabor et al., 2018)
Moreover, family members are more prevalent in family firms, as family members are involved in the business at an early age and most often grow up being involved in the firm (Verbeke & Kano, 2012), which leads to them slowly being integrated in the firm. Through years of socialization they fit well within the family firm. At the same time this creates high entry barriers for non-family members to assume leadership roles, as they have not had years of socialization in the firm (Verbeke & Kano, 2012). More entry barriers are present for non-family members in the form of non-economic goals (e.g. family influence, family status) (Gómez-Mejía, Haynes, Núñez-Nickel, Jacobson, & Moyano, 2007). The skills required to achieve these goals can be expected to be present in family members, but are difficult to obtain for non-family members (Verbeke & Kano, 2012). Selecting the right CEO and subsequently monitoring this person gets more complicated by this tug of war between economic and non-economic goals, as a family member might contribute little to the economic goals, but the high human asset specificity this person possesses is valuable for the non-economic goals of the family. Nonfamily CEOs will implement monitoring measurements that reduces the effect of the bifurcation bias and therefore making assessment fairer (Madison, Daspit, Turner, & Kellermanns, 2018).
to undervaluation of potential host country partners and overvaluing the need for family specific assets in the foreign market. The high level of bifurcation bias in family CEOs is due to them having been socialized for a long period of time, which causes them to more easily align with family ownership. This might be exacerbated in the case of the family CEO also being the owner. Nonfamily CEOs do have a bifurcation bias as they have not been socialized from a young age and in general tend to have more professionalism.
Applying all this information to entry mode choices, the prediction made here is that a family CEO is more likely to choose a WOS as an entry mode over a JV, compared to a non-family CEO in a family firm. The reasoning why is because family CEOs first and foremost are aligned with the family firm and are biased towards the capabilities of the family. When entering a foreign country, they focus not only on the economic goals, but also on the non-economic goals that further the prosperity of the family. Their overconfidence in the family results in disregarding local partners, or even straight up mistrusting potential local partners.
Hypothesis 2: Family firms with family CEOs will choose more often for a WOS over a JV, compared to a family firm with a non-family CEO
Moderation of corruption
The choice of entry mode is not only dependent on internal factors of the firm, but is to an extent dependent on the external environment the firm operates in or plans on operating. The external environment can provide motivations for types of entry modes, but at the same time can also impose limitations on the number of entry modes allowed. A well-known example of such limitations is where many emerging countries only allowed entry of foreign firms if this was done by partnering with a local firm in a joint venture.
Of the external environment one of the more prominent factors is the institutional environment (Cuervo-Cazurra, 2006; Rabbiosi & Santangelo, 2019). The institutional environment consists of factors such as corruption, institutional voids, intellectual property rights protection. When the institutional environment provides protection against these factors, the necessity for high control governance structures decreases (Shen et al., 2017), as the need for monitoring decreases and with any misbehavior of partners legal action can be taken. At the same time when the institutional environment is of a low level, the risk of opportunistic behavior of partners increases, in which case a high control governance structure is preferable. In most developed countries opportunistic behavior can be controlled by writing complete contracts. However, because of the low level of institutions in the foreign country the enforcement of contracts is not always guaranteed. In these environments high control entry modes are recommended in order to safeguard intellectual property and assets and knowledge from opportunistic behavior (Maekelburger et al., 2012). Others have proposed to give the subsidiary a great level of autonomy as this will reduce the costs of communication and costs associated with decision making from far away (Rabbiosi & Santangelo, 2019). At the same time this would decrease the reputational and legal costs in the case the subsidiary would be involved in a corruption scandal (Rabbiosi & Santangelo, 2019).
2019). TCE has largely been underutilized by scholars as a concept to study the institutional environment, however TCE provides some interesting explanations for factors such as corruption (Verbeke & Kano, 2013). It can accommodate the variables used in complementary conceptual perspectives and provides a conceptual platform for multinational activities in an institutional setting (Verbeke & Kano, 2013)
Corruption is the abuse of public power for private gain (Cuervo-Cazurra, 2006; Rabbiosi & Santangelo, 2019), and it has been found to decrease economic growth, causes lower numbers of foreign direct investments and lowers the trust firms have in government officials (Boudreaux, Nikolaev, & Holcombe, 2018). These effects can cause firms to use destructive behavior, instead of productive behavior (Boudreaux et al., 2018). Therefore, firms from countries with low levels of corruption often try to avoid entering countries with high levels of corruption or try to limit exposure of equity to corruption (Rabbiosi & Santangelo, 2019). However, avoiding entry in these countries is not always possible because of strategic reasons (Rabbiosi & Santangelo, 2019).
In countries with high levels of corruption, or countries with a bad institutional environment, two points of reasoning can be applied to family firms. On the one hand, literature has suggested that the effect of these institutional differences for family firms is reduced or even neglectable (Hernández, Nieto, & Boellis, 2018). This is mostly attributed to the relational capabilities of family firms, which include informal networks and a long term orientation. (Bertrand & Schoar, 2006). In terms of entry mode choice this could mean that on the one hand a high control entry mode provides the best results when corruption levels become bigger, as the high level of control can substitute the institutional safeguards that are dysfunctional in these countries (Hernández et al., 2018). This would result in a decrease of transaction costs as the costs of internalizing prove to be lower than the costs of negotiating, contracting and monitoring partner firm.
Another point of reasoning uses the bifurcation bias. Although the entry mode outcome stays the same the reasoning behind choosing a WOS or JV mode of entry differs. Corruption increases the effect the bifurcation bias has on bounded rationality, meaning the bounded rationality of transaction will be distorted due to unfamiliarity with local informal customs and social norms, that are clear to local firms, but not to foreign firms (Rabbiosi & Santangelo, 2019). Furthermore, corruption increases the potential for opportunistic behavior, as any form of opportunistic behavior of a partner has the potential to not being properly punished, if it even will be punished or noticed in the first place. This increase in transaction costs will lead to more expression of the bifurcation bias in family firms and the potential reputational and monetary harm that can be done by local partners increases the mistrust of local partners. This in combination with the overvaluation of the family specific assets causes the family firm to use an entry mode with high levels of control in order to reduce opportunistic behavior and have high control over foreign strategic decision making in order to build a reputation and maintaining it. Moreover, it has been argued that “low-trust environments” such as countries with high levels of corruption are causing baseline higher levels of bifurcation bias (Verbeke & Kano, 2012).
not share the non-economic goals. Therefore, family firms entering countries with low levels of corruption enter through a WOS. This provides them with more control over strategic decision making and operating choices, which could result to long term survival.
Therefore, an increase in the difference between both countries’ corruption levels will influence the relationship between ownership percentage and entry mode selection.
Hypothesis 3.1: The relationship between family ownership and entry mode choice is positively influenced by an increase in corruption level differences, such that firms with a higher degree of family ownership become more likely to choose a WOS when corruption differences between home and host country are high
The effects of corruption differences on entry mode decision making by CEOs is most likely somewhat similar to the effect on ownership. Family CEOs are associated with less skills and competence compared to non-family CEOs. As the difference in corruption levels increases, so does the unfamiliarity of the family CEO with the host country’s institutional environment. In this case it is likely that the family CEO will fall back to using a bifurcation bias and chooses to use WOS as entry mode choice, as this entry mode is best suited to the family assets and is in line with the values and visions of the family. At the same time, a non-family CEO is less likely to use a WOS, as they are more competent and therefore are better suited to maneuver through the institutional environment and finding local partners that are a good fit for the firm. Because they are more skilled than family CEOs and only deal with the owning family on a professional level, non-family CEOs are able to make a more objective and independent assessment of the situation and based on this select the entry mode that is most suited. Therefore, an increase in corruption level differences between home and host country increases the effect of the relationship between family CEOs and entry mode choice, meaning that it is even more likely that a family CEO selects a WOS as entry mode.
Hypothesis 3.2: The relationship between family CEO and entry mode is positively influenced by an increase in corruption level differences, such that firms with a family CEO become more likely to choose a WOS when corruption differences between home and host country are high
METHODOLOGY
Sample
Family firms are identified in line with Anderson & Reeb (2003), with the firm needing to have part of its equity owned by the founding family or otherwise by relatives/descendants of the founding family. On top of this another criteria in line with Sestu & Majocchi (2020) is that the ultimate owner of the firm needs to be a family controlled firm. This ensures that family firm values are present in the ultimate owner and no outside non-family values can have an influence on decision making. This is a somewhat straightforward method and while others propose more intricate methods (Hernández et al., 2018), this identification is sufficient for the aim of the research as there is no need to distinguish between family and non-family firms.
The sample selected is based on several criteria. First, the firms selected need to be completely or partially owned by members of the same family. Observations with multiple families in the same firm will be excluded, as it could lead to an overload of information, which in turn would make the study less conclusive. Second, the firms selected need to originated from the same home country. This is necessary to control for any differences in home country factors, such as regulations regarding internationalization. The home country that will be used will be the United States, as relatively a large number firms in the US are family firms (Mazur & Wu, 2016) and one-third of the Standards & Poor’s 500 firms are family owned (Anderson & Reeb, 2003; Iyer & Lulseged, 2013). At the same time by using US firms, the scope of family firm research gets widened, as a large portion of prior studies investigate Italian, Spanish and German firms.
An initial sample of 455 observations was constructed using data available at Zephyr. After a closer inspection of the data provided 186 observations were dropped. This was mostly due to missing data regarding ownership and Zephyr including individual investors within the sample. After removing these observations, the sample consisted of 268 observations from 65 US family firms that have made acquisitions and/or engaged in joint ventures from 1997 to march 2020. Data on acquisitions and joint ventures and ownership percentage is retrieved from the Zephyr database. Ownership percentages and CEO information is retrieved from the Orbis database.
The regression that will be used is a longitudinal binary logistic regression as the dependent variable is a dichotomous variable, which is in line with many other studies (Canabal & White, 2008).
Dependent variable
Independent variables
The independent variables will be family ownership and family CEO. Family ownership is measured by the percentage of shares the family owns in the company engaging in FDI.
Family CEO will be measured as a dummy with a 1 if the CEO of the firm is part of the family and a 0 if the CEO is not part of the family, this is in line with previous studies on family CEOs (André, Ben-Amar, & Saadi, 2014; Barontini & Bozzi, 2018) This information can be found at the Global Family Business Index and Orbis.
Moderator variable
The moderator variable is the difference in corruption levels between the US and host countries. The data on corruption levels is retrieved from the corruption perception index (CPI) created by Transparency International. The CPI scores countries on the level of corruption in the public sector. These scores are comprised of opinions of experts and business executives. In order to find the differences between home and host country, the score of the United States is being subtracted by the host country score.
Control variables
Firm Size. Larger firms have generally more international activity (Majocchi, Bacchiocchi, & Mayrhofer, 2005). Furthermore, larger firms have more funds and more access to funds (Kalantaridis & Vassilev, 2011), and could therefore enter a country with less need for a partner’s funds. Firm size will be measured by the logarithm of the number of employees (Sestu & Majocchi, 2020). A higher number of employees often indicates a firm has more assets and fund, which indicates a larger firm. The data is retrieved from Orbis
Industry. To control for industry factors the USA SIC codes are taken. As international diversification is industry specific, there is a need to control for industry influences (Mitter et al., 2014).
Market size. One of the reasons for firms internationalize is market seeking (Hernández et al., 2018). Controlling for market size is important, as larger markets are more attractive for market seeking (Ma, 2013). In order to measure market size the GDP of the host country is used. In line with Hernández et al. (2018) the logarithm is taken to adjust for right skewness. The data is taken from the World Bank.
Method
A longitudinal design will be used to test hypothesis. The sample consist of 65firms with multiple acquisitions or joint ventures made from 1998 to 2020. The total number of observations in this study is 268. The number of acquisitions and joint ventures made during this time span is not equal for each firm, some firms have been more active than others. Therefore a longitudinal logistic regression will be used, as this research approach makes possible to study firms with multiple acquisitions over a period of time (Ahuja & Katila, 2001). Furthermore, because a binary variable is used as the dependent variable a logistic regression is used. The Hausman test indicated that the regressions need to be run using random effects.
RESULTS
Table 1 provides an overview of the frequency of the host countries. What stands out is the significant large number with which Brazil, Canada, China, Germany and Great Britain provide a destination for acquisitions or joint ventures. These five countries are combined good for around 54% of the observations, compared to the other 37 countries that fill out the other 46%.
Next up, table 2 and 3 provides the summary statistics and correlation of the variables, alongside the variance inflation factors (VIF). For the summary statistics the original values are taken in order to give an understandable overview, except for the GDP, of which the natural logarithm is shown. What we can incur from this, is that 74% of the deals in the sample is an acquisition and 64% of the observations has a family CEO.
Multicollinearity is checked using the variance inflation factor (VIF). The test’s results all have VIF values lower than 10, ranging between 2.50 and 1.00. Furthermore, the tolerance (1/VIF) does not drop below 0.1, therefore the tolerance is met. Given these results there does not seem to be multicollinearity. The initial control variable of GDP per capita had to be dropped, as it had a correlation of 0.91 with the moderator of corruption. Such a high correlation between a moderating variable and a control variable is unsatisfactory, and in this case it is somewhat intuitive, as a low GDP per capita is often associated with developing countries, which are equivalently also associated with high levels of corruption. Dropping GDP per capita also provides better results for the VIF of the corruption moderator.
Table 1: Frequencies of Countries
Target Country Frequency Percent Cumulative
Table 2: Descriptive Statistics
N Mean SD Min Max
WOS 268 0.743 0.438 0 1 Ownership Percentage 268 70.030 23.313 32.390 100 Family CEO 268 0.400 0.474 0 1 Firm Age 268 80.485 51.862 5 167 US SIC Code 268 43.011 21.706 20 89 Number of Employees 268 209584 524491 17 2200000 Corruption Differences 268 5.448 18.197 -18 47 GDP in USD 268 12.236 0.552 8.741 13.134
Table 3: Correlation Matrix
VIF 1. 2. 3. 4. 5. 6. 7. 8. 1. WOS 1.15 1.000 2. Ownership Percentage 1.30 -0.090 1.000 3. Family CEO 1.59 0.242 -0.248 1.000 4. Firm Age 1.97 -0.216 0.227 -0.337 1.000 5. US SIC Code 2.15 0.115 -0.363 0.231 -0.647 1.000 6. Number of Employees 2.13 -0.191 0.069 -0.541 0.557 -0.517 1.000 7. Corruption Differences 1.17 -0.294 0.095 -0.301 0.245 -0.159 0.331 1.000 8. GDP in USD 1.05 0.045 -0.101 -0.087 -0.111 0.160 0.014 -0.043 1.000
Table 4 shows the regression models. Model 1, the baseline model, looks at the control variables in order to provide an insight on how much the variables of interest add. In model 2 the ownership percentage and family CEO variables are added. Model 3 shows the full model, meaning here the interaction terms are added.
Hypothesis 1 argues that an increase in ownership percentage will increase the likelihood of a wholly owned subsidiary being used for entry mode over a joint venture. With a coefficient of -0.001 family ownership shows to have a very small effect, however a negative effect. On top of this the coefficient is not significant. Given these results the first hypothesis is not supported. If we look at Hypothesis 2, a family CEO increases the likelihood of the entry mode being a WOS, we find a positive coefficient which is significant at the 5% level. This gives enough evidence to support Hypothesis 2. Model 3 shows that both the interaction of family ownership with corruption levels and family CEO with corruption levels are significant, however the signs are not as theorized. A positive effect of the corruption levels on the initial relationship was theorized, however it seems the opposite is true. Therefore, hypothesis 3.1 and 3.2 are both not supported.
Of the control variables firm age is significant at the 5% level in the base model and in the model without interactions. On the overall regression it has a negative impact.
Two goodness of fit tests have been used. First, the log likelihood. The general rule here is that the higher the value, the better the fit of the model is. There is however not a reference or critical value that the log likelihood has to assume in order to indicate a good fit. The only thing that can be inferred here is that model 3 has the highest goodness of fit followed by model 2 and model 1 respectively. This indicates that model 3 is the most complete of the three models. Second, the p-value of each model lies below the significance level of 0.05, and even below a significance level of 0.01. This indicates that there is sufficient evidence to assume the results show a good model fit.
Table 4: Regression results
(1) (2) (3)
VARIABLES Model 1 Model 2 Model 3
Ownership Percentage -0.001 -0.001
(0.007) (0.007)
Family CEO 1.101** 0.955**
(0.451) (0.470)
Ownership Percentage × Corruption Differences -0.004*
(0.002)
Family CEO × Corruption Differences -0.899**
(0.457) Ln (Firm Age) -1.660** -1.362** -1.103 (0.695) (0.679) (0.685) US SIC Code -0.012 -0.009 -0.007 (0.010) (0.011) (0.011) Ln (Number of employees) -0.285* -0.077 -0.034 (0.158) (0.185) (0.189) Ln (GDP_USD) 0.168 0.225 0.131 (0.259) (0.265) (0.259) lnsig2u -13.468 -13.372 -12.987 (22.947) (36.160) (23.137) Constant 3.837 1.348 1.747 (3.468) (3.766) (3.673) Observations 268 268 268 Number of Panels 65 65 65 Log Likelihood -144.3 -140.7 -135.7 chi2 14.59 20.23 23.61 P (chi2) 0.00564 0.00252 0.00266
DISCUSSION
In trying to explain entry mode choices of family firms from a transaction costs approach this article has found the following results. It appears that family ownership does not have any effect on entry mode choice. This finding can have multiple explanations. One of them is that prior research has suggested a curvilinear effect on entry mode choice. This has been found in many different contexts such as cultural distance (Malhotra, Sivakumar, & Zhu, 2011) and geographic distance (Malhotra, 2012), institutional distance (Estrin, Baghdasaryan, & Meyer, 2009) between home and host country. Such a relationship suggests that high and low levels of the independent variable mean that a JV is used for entry and for intermediate levels a WOS is used. This study looks at a linear relationship, whereas it might be the case that a curvilinear relationship provides a better representation of the relationship between family ownership and entry mode choice.
Furthermore, this study finds that a Family CEO does seem to have a significant influence on entry mode choice. This could illustrate the level of decision-making power the CEO has within a firm and on internationalization decisions. It seems that as suggested family CEOs are indeed a vassal of expressing bifurcation bias, indicated by the significance and comparative high coefficient. One explanation as to why family ownership represents the bifurcation bias less while a family CEO does, could be because family CEOs have a lot of decision-making power (Chittoor, Aulakh, & Ray, 2019). This power could be enhanced by the fact that often in family firms the owner/founder assumes the role of CEO and therefore has more freedom to operate as they please, without having to consolidate with others. This way the bifurcation bias that might be present within the ownership of the firm will be used in a position where it has more effect on decision-making. It has been found that owner/founder CEOs are characterized by exploratory and entrepreneurial behaviors that lead to riskier investments (Chittoor et al., 2019). On top of this they also are less restricted in their decision making, while they also have a long-term orientation (Chittoor et al., 2019), which is in line with the bifurcation bias.
Moreover, this study indicates that corruption has an influence on the relation between family ownership/family CEO on entry mode choice, however the results show that corruption weakens the effect of the original relation instead of enhancing. This could indicate that in this regard family firms act more like non-family firms and baseline TCE trumps the bifurcation bias theory. This means that family firms could act more in line with non-family firms and choose a JV with a local partner when unfamiliarity with the institutional environment increases, instead of mistrusting outsiders and not using the knowledge of local firms.
This study contributes to the literature first and foremost by following the trend in family firm research by moving away from dichotomous indications of non-family vs family firms to continuous indications of family firms. In doing so this study helps to further uncover the aspects of family firms that still remain in the dark, in the hopes of further paving a path for a proper family firm framework (Sestu & Majocchi, 2020). By focusing solely on family firms this studyhelps to extend the bifurcation bias framework into entry mode selection.
firm will enter a foreign country and how compared to this family ownership does not seem to impact this choice.
This study has some practical implications for family firms, its ownership and management. It supports the view that family firms need to be aware of pitfalls of the bifurcation bias. Although a family first mentality provides some unique assets and strategies it can also turn out to be costly and self-destructive if the level of bifurcation bias within a firm is not properly identified and accessed. Furthermore, this study shows the importance, not of family ownership, but family CEOs on strategic internationalization decisions.
LIMITATIONS AND FUTURE RESEARCH
Although this study provides some insight on family firm entry mode choices from a transaction costs perspective, even more detailed, a bifurcation bias perspective, this study does not manage to answer all questions asked in this study and even adds more questions on top of the existing ones. This means there are some avenues on which to build in future research. This study provides some insight in family firm heterogeneity and entry modes, but manages to only capture a small aspect of this relationship. Future studies should look at more aspects of heterogeneity of family firms and the internationalization/entry modes of firms. As explained in the beginning, most studies on the heterogeneity of family firms look either at family firm versus non family firms, however as this study shows, more relevant contributions can be made by studying the heterogeneity of family firms. In order to develop a complete overview of the relationship between heterogeneity and entry mode choices more aspects of heterogeneity need to be studied. Some of these being: the impact of a founder CEO, how many generations of the family work or have worked in the firm.
Furthermore, given the results of this study, indicating that a family CEO does seem to have a significant impact on entry mode choice, while it does not seem that family ownership has an impact, it might be worthwhile to study how the different aspects of family firm heterogeneity relate to each other. It might be the case that there is a ranking in power between the aspects of heterogeneity, meaning that some aspects have more effect on entry mode choices than others. Exploring this would help to further the development of family heterogeneity theories and models.
Moreover, with the bifurcation bias theory having been developed relatively recent means that still a lot of aspects related to it can be investigated. This is one of the first studies that extends the bifurcation bias across borders. However, this study merely provides a first step between entry mode choices and bifurcation bias in family firms. More studies should therefore extend the bifurcation bias to internationalization and entry mode areas of focus. A point that follows from this is to have more future studies integrate qualitative and quantitative methods, or at the least use qualitative studies to develop measurements for quantitative studies, as also shown in this study, many aspects of family firm heterogeneity and maybe even more so for the bifurcation bias are described theoretically, but are difficult to accurately depict in a quantitative setting. As illustrated by this study, assumptions and proxies have to be used in order to use the bifurcation bias in a quantitative study. By developing quantitative measures of the bifurcation bias results will be more accurate.
