The effects of family ownership and external management on internationalization speed of family owned SMEs, and moderator role of innovation intensity

The effects of family ownership and external management on

internationalization speed of family owned SMEs, and moderator role of

innovation intensity

MSc Thesis International Business and Management University of Groningen

Faculty of Economics & Business

Salvatore Gargiulo Student number: S3887081

s.g.gargiulo@student.rug.nl

Thesis supervisor: Dr. A. Kuiken Co-assessor: J. Shin

The effects of family ownership and external management on internationalization speed of family owned SMEs, and moderating role of innovation intensity.

ABSTRACT

By leveraging key notions from SEW theory, stewardship theory and the relevance of social networks, this study examined the influence of two main sources of heterogeneity among family owned SMEs on internationalization speed. Furthermore, the moderator role of

innovation intensity on the aforementioned relationships has been explored as well. This study made use of a sample of 385 family owned SMEs, from 5 different countries across Europe, with data referring to 2009. Literature on family businesses’ internationalization has not reached a common agreement upon family firms’ distinctive international behaviour. Most of the studies build on SEW theory to demonstrate how family firms are reluctant towards risky strategies such as internationalization. By doing that, the majority limits also the research scope to a comparison between the willingness to internationalize between family and non-family firms. This approach fails to capture the subsequent evolvement of the

internationalization process while treating family firms as homogeneous entities. This study moves further focusing only on family owned SMEs which have already internationalized and differ between them, in terms of degrees of family ownership and presence of external

management. Furthermore, it shows that higher degrees of family ownership favour internationalization speed of family owned SMEs in terms of export intensity. However, it failed to find support for the hypothesized positive relationship between external management and internationalization speed. Finally, the moderator role of innovation intensity has been investigated as well. Surprisingly, a negative moderating effect has been found to influence the relationship between family ownership and dependent variable. On the other hand, no significant moderating effect has been shown to apply on the influence of the presence of external management on internationalization speed.

ACKNOWLEDGMENTS

In this small section I would like to express my gratitude to the people that have made this work possible. First of all, I want to thank my supervisor Dr. A. Kuiken who has supported me during the whole process with her positivity and patience. Then, I would like to thank Dr. J. Shin for her support during the initial phase of this project. Finally, I want to thank Dr. C. Midoes who has served as mediated access to the EFIGE dataset.

TABLE OF CONTENT

1.INTRODUCTION... 2. LITERATURE REVIEW... 2.1 Family ownership and internationalization... 2.2 Restrictive and facilitating approach... 2.3 Socioemotional Wealth and Stewardship theory... 2.4 External management and internationalization... 2.5 Speed of internationalisation………

2.6 Hypothesis development………

2.6.1 Family ownership and internationalization speed of family owned SMEs……… 2.6.2 External management and internationalisation speed of family owned SMEs…… 2.6.3 The moderator role of innovation intensity………

1. INTRODUCTION

Regardless of their size, family firms hold a central role in leading economic growth

everywhere in the world (Zhara, 2003). Indeed, the most prevalent type of firm worldwide is the family firm (La Porta, Lopez-de-Silanes & Shleifer, 1999). Family Firm Institute’s latest statistics state that family businesses (FBs) represent two thirds of all businesses around the globe, generating more or less 70-90 percent of annual global GDP, creating 50-80 percent of jobs in most countries worldwide (Family Firm Institute, 2017).Therefore, the interest in family business research has been growing substantially in recent years (Chrisman et al. 2010). However, a great number of academics and researchers understood the urgency to focus more on family owned small-and medium-sized firms (family owned SMEs onward) due to the fact that family SMEs represent 85% of all the firms in the EU (IFERA, 2003) and an even greater proportion in the developing countries. Therefore it seems more than legit to assume that family SMEs are between the main recipients of new challenges and

opportunities stemming from globalization, technological developments and aggressive national and international competition. Indeed, growth by international expansion is a relevant strategic option for firms of all sizes and ownership types, in order to cope with external environmental changes of such a scale (Lu & Beamish, 2001; Zahra & George, 2002).

has highlighted an inverted-U relationship between family governance and internationalization (e.g. Sciascia et al., 2012; Liang, Wang, & Cui, 2014)

One reason underpinning these mixed results may be because prior studies do not effectively take into account the heterogeneity character of family firms (Westhead &

Howorth, 2007). A growing consensus argues that family firms cannot be merely viewed as a homogeneous entity (Sharma 2002; Chrisman, Chua, & Sharma, 2005). FBs are

heterogeneous. Strategic behaviors such as international expansion may differ, not only between family and non-family businesses, but also between FBs with different properties (Melin & Nordqvist, 2007). Indeed, family firms may differ with regard to the degree

of family ownership and presence of non-family members in the management of the firm (P. Westhead & C. Howorth, 2007). Both distinctive characteristics may affect some facets of the internationalization process, such as speed.

Together with scope and degree, speed represent one of the key dimensions along which the internationalization of a generic firm can differ from others (Zahra & George, 2002), and therefore, be studied. Internationalization speed has gained much more attention in the IB literature since the mid-1990s (Casillas & Acedo, 2013). Specifically, International New Venture, as a research area, stemmed from the need to explain speedy internationalization of SMEs enhanced by globalisation and technological advancements (Oviatt & McDougall, 1994). Subsequently, frameworks such as the one from Oviatt & McDougall (2005) and the revised Uppsala Model (Johanson & Vahlne, 2009), aimed to explain the factors and the forces influencing firm’s internationalization speed. On the other hand, FB literature has heavily overlooked at the concept of speed in the setting of family firms’ internationalization, focusing its efforts mostly on the arguments stemming from the dichotomous comparison of family and non-family firms’internationalization, as stated above. There might be different reasons lying behind this surprising lack of literature efforts. First of all, existing theories explaining internationalization speed do not look at ownership as a relevant factor (Kontinen & Ojala, 2012). This leads to the need for integration between IB literature with FB literature, in order to investigate the effects of sources of heterogeneity among family firms, on

internationalization speed. Second, the concept of speed has generally been operationalized as the time lag between firm’s foundation and it’s first international activity (Rialp, Rialp & Knight, 2005; Zahra and George, 2002). Hence, studying the effects of sources of

such it wouldn’t produce any theoretical advancements in the literature. Moreover, it could lead researchers on a misleading path, studying the speed with which the process begins, rather than the speed with which the process evolves (Morgan-Thomas & Jones 2009). In order to address these issues, this research studies the effects of two main sources of

heterogeneity among family firms, family ownership and external management, on a different measure of internationalization speed. The proposed measure is meant to capture the speed of international growth. It was first theorized by Bonaccorsi (1992) as export intensity, and then re-proposed by Casillas and Acedo (2013) as an alternative measure for internationalization speed of exporting companies. On the basis of the above, the following research question will be addressed:

“How do family ownership and external management affect internationalization speed of SMEs ?”

Despite potential differences in ownership structures and management composition, the predominant position in FB literature assumes that family firms that show greater capability for innovation are more likely to be successful in international expansion (Kontinen and Ojala, 2010).This view is supported by the evidence that nowadays, business growth is mainly supported by the joint effect of innovation and internationalization (Onetti et al., 2010). Indeed, through innovation SMEs are able to open new market opportunities and/or reinforce the existing ones. On the other hand, internationalization gives access to a broader variety of knowledge which in turn, supports innovation (Zucchella & Siano, 2014). However, when taking a deeper look at the propensity to innovate, the current state of knowledge in FB literature is that family firms tend to invest less in innovation, but when they do they are more efficient (Duran, Kammerlander, Van Essen & Zelleweger, 2015). This means that they are able to generate greater innovation output when compared to non-family firms . Hence, on the basis of the above, greater innovation intensity would lead family owned SMEs to greater levels of innovation outputs, which in turn would foster their willingness to internationalize. This would suggest to focus again on the comparison between family and non-family firms, and look at differences in the pre-internationalization stage. Whereas, is not known yet if this effect would apply on the willingness of internationalized FBs to increase speed as well, as a consequence of innovation intensity. Hence, since the focus of this study is on the relationship between two of the main sources of heterogeneity among family firms and

“How does innovation intensity moderate the relationship between family ownership, external management and internationalization speed of family owned SMEs ?”

This research will try to answer these questions by making use of the EFIGE dataset, which has already been used in previous studies about FBs internationalization (i.e. Altomonte, Acquilante, Bekes & Ottaviano, 2013; Hennart et al., 2019). The proposed relationship between family ownership and external management with internationalization speed and the moderating role played by innovation intensity, are tested according to the conceptual model, which will be outlined in the dedicated section. This research aims to bring novel insights to the FB literature, surpassing the classical research approach comparing family and non-family firms by looking more carefully at sources of heterogeneity within the family firms, and their impact on internationalization speed. More specifically threefold learning outcomes are expected. First, deeper insights will be provided about how the internationalization process evolves in FBs, by looking individually at the effects of changes in the degrees of family ownership and the presence of external management on

internationalization speed. Second, it will be clarified how increases in innovation intensity within the FB leads to greater internationalization speed due to higher levels of family ownership or external management. To do so, multiple linear regressions and moderation analysis will be conducted. The analysis will be based on data derived from the EFIGE dataset concerning family owned SMEs from 5 different European countries.

The study starts with an overview of the FB and internationalization literature.

2 LITERATURE REVIEW

This section deals with the literature relevant to the topics of family ownership, external management, and internationalization speed. More specifically it is divided in 5 sub-sections, respectively: “Family ownership and internationalization of SMEs”; “Restrictive and

facilitative approach”; “Socioemotional wealth and Stewardship theory”; “External management and internationalization”; “Speed of internationalization”. Then, hypothesis development will follow.

2.1 Family ownership and internationalization of SMEs

Since the first work of Gallo and Sveen (1991), which highlighted enabling and restraining factors of family firms’ internationalization, FB literature has increasingly explored the role and the effects of family ownership on firms’ internationalization process. Despite the growing literature efforts, scholars have not agreed on a single common definition of family-owned business (Crick, Bradshaw & Chaudhry, 2006). In particular, most of the attempts to express a definition for family firms, focus on the characteristics that distinguish them from non-family businesses (Price, Stoica & Boncella, 2013).This study considers family owned SMEs, in which family members are the main shareholders of the firm (Zahra, 2003; Gomez-Mejia et al.,2010; Arregle, Duran, Hitt & Van Essen, 2016). Since family ownership is especially prominent among small - and medium-sized firms (SMEs) (Hennart et al., 2019), the focus is on family owned SMEs in particular.

Findings about family ownership and different aspects of SMEs’ internationalization, are highly inconsistent (Calabrò, Brogi & Torchia, 2016). Different authors have argued that family ownership can either positively (Carr & Bateman, 2009; Zahra, 2003) or negatively (Fernández & Nieto, 2006; Graves & Thomas, 2006) impact internationalization. Therefore the literature on FBs internationalization has evolved following mainly two approaches to support the different findings.

2.1.1 Restrictive and facilitative approach

FB literature has mainly dealt with questions meant to uncover how family ownership affects the international behavior of firms, especially in terms of willingness and propensity to internationalize. By doing that, scholars have mostly adopted two main conflicting

internationalize, compared to non-family firms. Some of these restraining factors stem from the lack of financial resources, limited access to human resources, scarce managerial

capabilities worsened by the unwillingness to accept non-family expertise, fear of losing control of the enterprise, risk aversion and the need to protect some of the traits that constitute family socio-emotional wealth (SEW) (Fernández & Nieto, 2006; Gomez-Mejıa et al., 2010; Sciascia et al., 2012;). As one would expect that large family-owned firms would have overcome some, if not all of these issues, it is likely that these arguments mostly apply to family owned SMEs (Hennart et al., 2019). Consequently, it would also seem reasonable to expect family owned SMEs to be less prone to internationalize than non-family ones.

On the other hand, the facilitative approach is supported by scholars who argue that family ownership might foster internationalization. Indeed, factors such as family owners and managers’ solid identification with the business induce them to pursue stewardship behavior (Davis, Schoorman, & Donaldson, 1997; Miller, Le Breton-Miller, & Scholnick, 2008) and to exploit international opportunities (Sciascia et al., 2012). Furthermore, informal

communication and long-lasting relationships between family members can promote the sharing of experiences and knowledge. These are the foundation to build trusting

relationships that gives birth to an organizational culture, promoting strategic moves related to long-term gains such as internationalization (Arregle, Hitt, Sirmon, & Very, 2007). On the basis of these assumptions lying behind the facilitating approach, family owned SMEs are expected to internationalize more than non-family ones.

To overcome the inconclusiveness of the findings stemming from both approaches, it becomes essential to surmount the dichotomous approach in comparing family and non-family firms (Hennart et al., 2019) and also to explore what’s next the initial decision whether to internationalize or not. In order to do so, this study poses FBs that have already

2.1.2 Socioemotional wealth and stewardship theory

Family ownership plays very often a major role in defining family firms internationalization and variation in the degree of family ownership is considered a relevant source of

heterogeneity within family businesses (Chrisman et al., 2005; Le Breton-Miller, Miller, & Lester, 2011). Indeed, research on strategy or organization theory sustain that ownership type can influence firm internationalization, because different owners show different interests, values and incentives (Lin, 2012). With regards to family ownership an interesting study from Sciascia et al., (2012) revealed that family ownership experience an inverted U-shape

relationship with the attitude towards international entrepreneurship. This means that high levels of family ownership will negatively affect firm’s willingness to internationalize. Hence, in line with the restrictive approach, undiversified ownership in family firms has been shown to be a primary factor leading firms to avoid strategies such as internationalization or R&D-intensive investments, because of high risk related to these sort of strategies ( Zellweger & Sieger, 2012, De Massis, Frattini, & Lichtenthaler, 2013).

However, once FBs have already internationalized, usually choosing low-commitment entry mode such as exporting (Pinho, 2007), high levels of family ownership might turn things around. Especially with regards to the willingness and ability to increase the speed of internationalization. De Massis (2014) willingness and ability perspective state that different concentration of ownership generates different interests and goals. Hence, the willingness and ability of internationalized SMEs with high level of family ownership to increase speed of internationalization, result to be affected by the extent to which family owners perceive that such a strategy can help to achieve their economics, and non-economic utilities, and their authority to execute the strategy (Fang et al., 2018). Persecution of non-economic utilities is a central topic in Socioemotional Wealth theory (SEW). The latter emphasizes the

noneconomic aspects affecting the decision making process of family firms. More

specifically, family owners will keep socioemotional wealth beside economic considerations, as their fundamental reference point while selecting among different alternatives (Gomez-Mejia, Cruz, Berrone, & De Castro, 2011). Especially when FBs have already

the latter, when family ownership is high in the firm, succession, reputation and family firm identity will represent strong influencing factors in shaping firms goals and ambitions (Evert, Sears, Martin & Payne, 2018). In other words, when family ownership outweighs other shareholders and governance structures, the firm shows a willingness to act in a manner that is supportive of the family objectives, leading to a stewardship attitude towards the fate of the business.

Stewardship theory provides further arguments supporting these assumptions. The theoretical foundation of Stewardship theory is that family owners or managers will place a higher utility on the collective well-being rather than their own, developing a strong sense of duty towards the organization and hence aiming to improve organizational performance ( Miller & Le Breton-Miller, 2006). Indeed, findings suggest that FBs are more likely to show stewardship attitude with respect to the long-term well-being of the firm than non-family firms (Kontinen & Ojala, 2012). Moreover, complementary to preservation of SEW,

stewardship attitude will have its main focus in the survival of the business, with the goal to sustain it over generations (Miller & Le Breton - Miller, 2006). These arguments lead to the assumption that once family firms have internationalized, (e.g. through exports) they might be not really interested in increasing the scope of internationalization process, which is costly, risky and can jeopardise family control of the business. Instead, in line with both SEW and stewardship theory, family SMEs might put more emphasis on strategies that will increase internationalization speed instead that other dimensions such as scale and scope. Indeed, George et al. (2005) found that internal ownership tend to lower the propensity to increase scale and scope of internationalization of SMEs. Subsequently, Lin (2012) expanded these findings in the family business setting, suggesting that high level of family ownership leads instead to rapid pace of internationalization.

2.2 The need for external management for family owned SMEs internationalization

NFMs by definition are not linked to the owning family by blood, marital or adoptive relationships (Klein & Bell, 2007). Once FBs have sourmonted their initial reluctant attitude toward the involvement of external management in their international activities, they can experience several benefits. Banalieva and Eddleston (2011) found that internationalization strategies are often more successful when the FBs rely on the backing of NFMs as well. Sciascia et al. (2012) confirmed this view stating that professional management is required even when the firm is internationally active only with exports activities. In FBs, this is mainly because the influence of NFMs on the decision making process lowers risk and uncertainty (Chang & Shim, 2015). Indeed, the presence of NFMs helps FBs to overcome managerial and market knowledge constraints by adding valuable expertise, competencies and network contacts (Sciascia et al., 2016).

With regards to the latter, IB literature has recognized social networks to be crucial for SMEs international growth (e.g. Lu & Beamish, 2001; Hadley & Wilson, 2003). Bjorkman and Kock (1995) defined social networks as the network of individuals which mostly interact through social exchanges, favoring the information flow and business exchanges as well. Hence, this view put social networks equals to social relationships that might ultimately have an impact on the formation of business ties. In line with this view, when studying the role of social network relationships in the internationalization process, Chen (2003) defined them as the basis from which more formal and tight business networks can be developed in new markets. Furthermore, Liesch et al. (2002) stressed that the creation and maintaining of social network relationships can be considered as part of the internationalization process of the firm. Conversely, for internationalized family owned SMEs is equally important to further develop and maintain social networks relationships. NFMs can be definitely useful in that. Several studies have shown the strategic relevance of social networks for SMEs internationalization, regardless of their ownership structure. Reid (1984) found that social networks helps in raising the awareness of new market opportunities. Ellis ad Pecotich (2001) shown how they can influence firms exporting behavior, and helping to identify new foreign exchange partners (Ellis, 2000). Yeoh (2004) found that personal social networks and contacts, such as the ones of NFMs, positively affect internationalized SMEs’export performance. Indeed, tacit

knowledge about international operations can be derived (Eriksson, Johanson, Majkgard & Sharma, 1997) while helping to broaden further SMEs’ international orientation and

2.3. Speed of internationalization in FBs

The first works of Johanson and Vahlne (1977; 1990) have explained internationalization of firms as a stage model (best known as Uppsala model). According to the Uppsala model, the firm gradually internationalize in terms of speed and scope, as it gains knowledge from foreign markets. More specifically, the Uppsala model states that firms usually tend to establish first a strong domestic market, afterwards they start to export through an agent in geographically and psychically close markets. As they gain more knowledge about these markets, they will increase the amount of resources committed to them (e.g. establishing foreign subsidiaries).

Subsequently, the first work of Oviatt and Mcdougall (1994) shifted the focus towards firms that are international soon or after their inception, giving birth to a new stream of literature best known as International New Venture. Ten years later (2005), the authors proposed a theoretical model of forces influencing internationalization speed. More specifically, in the model, enabling, motivating, and moderating forces affect the

entrepreneurial actor perceptions (mediating force) about the exploitation of international opportunities. This will have a direct impact on internationalization speed. Enabling and motivating forces capture the characteristics of the environment within which the firm operates. They encourage and make internationalization feasible and they are, respectively, the level of technology and competition. Firms, or entrepreneurial actor perceptions represent the mediating force between entrepreneurial opportunities and internationalization speed, subjected to both enabling and motivating forces. Finally, entrepreneurial actor’s perceptions are moderated by the level of knowledge required about foreign markets and the one required to leverage the opportunity. Network relationships also play a moderating role in the model. Hence, moderating forces play a central role in influencing the decision making process of the entrepreneurial actor to internationalize, in order to exploit international opportunities.

On the basis of this definition of internationalization speed as dependent variable, not many studies have tried to expand our knowledge on the contingencies explaining differences in the speed of internationalization in family firms. Gallo & Sveen (1991) argued that family firms tend to internationalize slower compared to non-family firms. This is mainly because they usually experience a knowledge gap as opposed to the latter and to accumulate

knowledge about foreign markets slower. Kontinen et al. (2012) moved the discussion further showing that fragmented family ownership might lead to traditional internationalization pathways (Uppsala Model) while high concentration of family ownership to born-global and born-again global pathways, especially when the ownership was passed in its entirety to the next generation. While Lin (2012) by considering internationalization speed as number of subsidiaries opened per year by family owned multinational corporations, showed that high concentration of ownership leads to rapid pace, narrow scope and irregular rhythm.

However, not all of these findings contribute to a better understanding of how speed of internationalization evolves once the process has already started. They stem from a

conceptualization of internationalization speed which deals more with the pre-stage period rather than the internationalization process per se (Casillas & Acedo, 2013). To overcome the issue, Casillas & Acedo (2013, p. 19) proposed a multidimensional construct of

internationalization speed able to capture three different facets of the concept. These are: “(1) the speed of the growth in a firm’s international commercial intensity; (2) the speed of its increase in commitment of resources abroad; and (3) the speed of the change in breadth of its international markets” The first two tend to capture the intensity with which firms increase international commitment over time, while the latter refers to its degree. Hence, since exports is by far the most common strategy used by SMEs to internationalize (Wolff & Pett, 2000), internationalization speed, in the setting of family SMEs, can be conceptualized as the increase of firm’s export intensity between two specific points in time (Bonaccorsi, 1992).

2.4 Hypothesis development

This section deals with the argumentation leading to the formulation of the hypothesis in order to answer the research questions. It starts with a subsection presenting arguments for the relationship between family ownership and internationalization speed. Following, another subsection will be dedicated to find a link between external management and

2.4.1 Family ownership and internationalization speed

Building on the core concepts of SEW and stewardship theory, I propose two main reasons why family owned SMEs characterised by higher degrees of family ownership will increase internationalization speed. First, once FBs overcome constraints towards the unwillingness to internationalize and enter new markets, preservation of SEW mechanism might start

working all the way around. Gomez-Mejia et al. (2010) argued that high family ownership will likely lead to choices about internationalization which are more reflective of family-centred non economics goals, such as preservation of SEW. Conversely, when facing the decision whether to pursue or not risky strategies such as internationalization, the higher is family ownership the lower is the willingness to internationalize (George et al., 2005; Liu et al., 2011). I argue that once FBs have already internationalized, high levels of family

ownership might lead family principals to get stuck in a continuous loss mode related to their SEW endowment, due to international commitment. This means that they will fear potential SEW losses so much, that they will become more prone to take risky behaviours in order to perform better and avoid them. Indeed, as family ownership increases in FBs, greater stake of family members personal financial and emotional wealth will be linked to the fate of the family business (Ray, Mondal & Ramachandran, 2018). This will make family owners more sensitive to financial and non-financial losses once internationalized, and more prone to behave in a way that allows them to avoid these losses, such as increasing

internationalization speed. In turn, returns are very likely to increase as well, while reducing dependence from the domestic market as main or single source of revenues (Fang et al., 2018).

Second, I argue that high degrees of family ownership will facilitate family members stewardship behaviours to impact speed of internationalization of family SMEs. More specifically, as the degrees of family ownership increases, stewardship attitude towards the long-term wellbeing of the firm will be one of the main drivers of FBs decision making

business, might opt to increase the international competitive trait of the firm, pushing export activities and so increasing internationalization speed.

Taking both arguments together, it can be concluded that once family SMEs have internationalized, high degrees of family ownership will make family owners more sensitive to potential losses of SEW. Hence, increasing speed of internationalization might reduce the risk of losing SEW, linked to potential failure in foreign markets, while increasing the likelihood of SEW gains at the same time. Beside that, the willingness to preserve the business over generations will also be greater in SMEs with higher degrees of family ownership. Hence, increasing speed to keep competitive will be in line with stewardship attitude of family owners. Therefore, as family ownership increases internationalization speed of family owned SMEs increases as well. This leads to Hypothesis 1, which states that: H1: “Higher degrees of family ownership increase the internationalization speed of family owned SMEs”.

2.4.2 External management and internationalization speed

Based on the resource constraints character of SMEs and the importance of social networks, I argue that the presence of professional and qualified NFMs beside family ones, is very likely to increase FBs internationalization speed, for two main reasons.

First of all, as noted by Knight and Liesch (2002, p. 982), “activities such as researching foreign markets, adapting products, finding and contacting buyers, developing foreign channels, moving goods across great distances, and ensuring that products are managed properly on their way to end users, pose considerable challenges to resource constrained, internationalising SMEs”. NFMs can help family owned SMEs to overcome these limitations, since they are more capable to execute professional managerial practices and are usually more focused on financial performance rather than socioemotional returns (Miller, Le Breton-Miller, Minichilli, Corbetta & Pittino, 2014). Conversely, especially in the case of exporting family SMEs, NFMs will direct firm efforts towards the increase of export intensity in order to gain better financial performances. Moreover, they will be more capable than family managers in doing that in most of the cases.

Second, by hiring external and professional managers FBs can gain access to a diversified and larger network of contacts beside acquiring more specific industry knowledge and

depicts markets as systems of social and professional relationships between different agents such as customers, suppliers or competitors (Martineau & Pastoriza, 2016). Findings from Coviello & Munro (1997) and Ellis (2000) states that orientation and initiation to export might depend not only on firm-level factors such as firm size or resource endowment, but also personal and professional networks of managers have a substantial influence. Mechanisms such as reducing information asymmetries, access to opportunities or resources and

experiential learning are enhanced by social networks (Zhou, Wu, & Luo, 2007). In turn, this will not only help in the early stages of internationalization, but it will also play a significant role on the speed with which the process evolves (Coviello & Munro, 1997). A family SMEs which will mainly recruit internally within the boundaries of the family, will be limited in that, compared to family SMEs that recruit NFMs as well. Indeed, NFMs’ social networks represent a potentially effective manner to overcome resource and capabilities constraints, and it will ultimately boost internationalization speed.

Considering the above discussion together, it can be deduced that the presence of NFMs help the firm to overcome serious resource constraints, adding value to the managerial capability of foreign operations. Furthermore, professional and qualified NFMs usually bring valuable social networks ties, which can differ from those of the family owners opening the firm to new opportunities. Therefore, family firms which will hire NFMs, are more likely to increase the speed of internationalization as opposed to the ones that only rely on family management. These assumptions lead to the following hypothesis:

H.2: “Family owned SMEs characterised by the presence of external management increase internationalization speed more those that are not”.

2.5 The moderator role of Innovation Intensity

This section seeks to illustrate how the aforementioned relationships between family ownership, external management and internationalization speed change according to the degrees of innovation intensity within the firm. Hence, the next two subsections will provide some arguments supporting first the relevance of innovation for family owned SMEs

internationalization speed, and then shortly present the current position of FB literature with regards to innovation in family firms. Afterwards, the last subsection will present the

2.5.1 Innovation intensity and FBs

Classen, Carree, Van Gils & Peters (2014) define innovation intensity as a measure that “encompasses all internal and external R&D expenditures, acquisition of advanced

machinery, facilities, software and external knowledge to realize innovation projects, product design, construction, design of services and other preparations for the production/sale and distribution of innovations, internal or external training specifically for innovation projects, and launch of innovations onto the market (marketing campaigns directly linked to product innovations)”. Nowadays, in most industries, innovation has been recognized as one of the most relevant sources of competitive advantage for firms (Cardinal, 2001) especially when competing in international markets. The work from Golovko and Valentini (2011) showed the existence of complementarity between innovation intensity and exporting for international sales growth. More specifically, only SMEs that are both innovative and exporters can experience significant international growth (Love & Roper, 2015).

The predominant view in FB literature states that intensive innovation strategies tend to get little support from family firms due to associated high costs and uncertainty (Matzler, Veider, Hautz & Stadler, 2015). However, when family firms get involved in innovative activities, tend to perform better in terms of innovation outputs quality (number of patents, citations and products) compared to non-family firms (Matzler et al., 2015; Duran et al., 2015). Hence, it becomes interesting to focus only on family firms and investigate how and if innovation intensity moderates the relationship between family ownership, external

management and internationalization speed.

2.5.2 The moderating effect of Innovation intensity

SEW, they might try to push exports towards these markets or even expand in new ones. Indeed, innovative firms have a greater tendency to enter or exploit their presence in new markets to increase sales volume in order to spread the fixed costs of innovative activities (Rogers, 2004). Such a strategy doesn’t come free of risks. However, when facing the potential loss of SEW, FBs characterized by higher degree of family ownership will lower their risk aversion (Gomez-Meija et al., 2007).

In summary, high levels of innovation intensity will make family owned SMEs with higher degrees of family ownership more sensitive to its related costs. Hence, being the firm already internationalized, family owners will increase internationalization speed in order to recoup the amount of resources invested in a shorter period of time, spread costs and preserve SEW. On the basis of these assumptions, the following hypothesis states that:

H.3a: “Innovation intensity positively moderates the relationship between family ownership and internationalization speed, in such a way that, the greater innovation intensity will be, the greater will result the positive effect of family ownership on internationalization speed of family owned SMEs”.

However, the innovation literature is clear on assessing that high-quality skills and networking make difference for innovation and export strategies (Love & Roper, 2015).As it has already stated before, family members are usually likely to miss these specific skills, knowledge or expertise. On the other hand, NFMs will usually be more likely to possess both attributes. Indeed, NFMs depend on more diversified network (Sirmon & Hitt, 2003) and be better able to integrate a set of intangible skills and resources to capitalize from innovative intensive strategies (Chen & Hsu, 2009). Moreover, networks relationships have been shown to be more relevant to FBs than to other firms, and their ability to increase their speed of internationalization will depend on the network relationships they will gather abroad (Graves and Thomas, 2008). NFMs will expose FBs to many if not all of these benefits. Hence, FBs characterised by high level of innovation intensity and presence of NFMs, will have a broader range of network ties to rely on. Consequently, they will have greater chances to capitalize over these investments. In other words, NFMs might be able to exploit better and more intensively their network ties when the FB is highly innovative, increasing the overall export intensity of the firm as subsequent strategy.

networks effect on internationalization speed. This is because high innovation intensity will allow NFMs to exploit better existing network relationships or to develop new ones. This results in the following hypothesis:

H.3b: “Innovation intensity positively moderates the relationship between external management and internationalization speed of family owned SMEs, in such a way that the greater innovation intensity will be, the greater will result the positive effect of external management on internationalization speed of family owned SMEs”.

2.6 Conceptual Model

The conceptual model in Figure 1 provides a visualization of the hypothesis based on the literature review. The model shows the positive relationship between family ownership, external management and internationalization speed. Furthermore, it exhibits the positive strengthening effect of the moderator on the aforementioned relationships.

3 METHODOLOGY

This section provides an overview of the research methodology that has been used to test the formulated hypothesis. First of all, data source and sample derivation are described, followed by an argumentation for the operationalization of the variables in the analysis.

3.1 Data source

All the data that have made possible this work come from the EU-EFIGE/Bruegel-UniCredit dataset (EFIGE ongoing). More specifically data have been derived from almost 15000 surveyed firms, operating in the manufacturing industry, in seven European economies (Austria, France, Germany, Hungary, Italy, Spain and United Kingdom). All the surveyed firms were employing at least 10 workers, in order to exclude micro-enterprises. Data cover the years from 2007 to 2009, and have been derived from responses to a questionnaire which covers different broad areas such as: structure of the firm; workforce; investments; technological innovation; R&D; export and internationationalization processes; market structure and competition; financial structure and bank-firm relationship.

One of the main features of the EFIGE dataset is that data can be fully compared across countries, since they stem from responses to the same questionnaire over the same time period (January to May 2010). Hence, due to its unique features the EFIGE dataset enables to investigate the link between ownership, governance, internationalization and innovation (Altomonte, Acquilante, Bekes & Ottaviano, 2013).

3.2 The sample

were met, ownership structure needed to be defined. First of all, all firms belonging to either national or foreign groups have been excluded, to assure the presence of only independent entities belonging to an individual or a group of individuals. Moreover, only firms that defined themselves as “family-owned entity” were included. Furthermore, since the phenomenon under investigation is FBs export intensity as measure of internationalization speed, firms had to be active in export activities during the period 2007 - 2009 and before. Furthermore, firms had to export directly from the home country and they have had experienced an increase in the value of export activities in 2009 compared to the previous year. Out of the original EFIGE dataset, only 387 family firms fulfil all the given criteria.

3.3 Dependent Variable

Bonaccorsi (1992) stated that especially in the case of exporting companies, speed can be measured as the increase of their export intensity between two moments in time. This will represent the growth of the ratio of firm sales stemming from foreign markets over a specific time period (Casillas & Acedo, 2013). Hence, since family owned SMEs make use mainly of export activities for their internationalization strategies, export intensity will be the operationalization of the dependent variable in the analysis. More specifically respondents to the EFIGE questionnaire were asked if they experienced any reduction or increase in terms of percentage of the value of export activities in 2009, compared to the previous year. All the firms in the sample have experienced this increase, allowing to investigate the determinants of it. However, the variable has been transformed from percentage of increased value of export activities, in its absolute value as share of total sales in thousands euro in 2009. In doing so, the resulting value is meant to capture the absolute value of the increase of the export activities on total sales.

3.3.2 Independent Variables

(D’Angelo et al., 2016; Enriques & Volpin, 2007). Such a measure of external management is meant to capture the presence of professional external managers in the firm. Therefore the sample results to be composed by 139 family SMEs (36% of 387) with no external management at all, while 248 FBs (64% of 387) relied on the expertise of NFMs as well.

3.3.3 Moderating Variable

In this study innovation intensity is included as moderating variable. Following the work of Classen et al. (2014) in the family innovation literature, innovation intensity has been operationalized as the R&D expenditure per employee, rather than an input measure of the innovation process (such as R&D expenditure over total sales) as prior studies have reported this as unreliable, particularly in an SME context (Hoffman, Parejo, Bessant, & Perren, 1998). Respondents from the EFIGE survey were questioned which percentage of the total turnover the firm has had invested on average during the last 3 years before the survey (2007-2009). Consequently the resulting values have been converted into their absolute form referring to 2009 turnover, and divided by firm size. Classen et al. (2014) argue that such a measure of innovation intensity is able to capture firm efforts for the production/sale and distribution of innovations.

3.3.4 Control Variables

have been controlled using the Pavitt’s (1984) taxonomy which relies on technological traits to distinguish among different industrial sectors. Hence the taxonomy defines firms as: “Scale intensive” (21% of 387); “Traditional” (52 % of 387); “High tech” (20% of 387) or “Specialized-suppliers” (23% of 387). Also FDI impact firms’ exports in different ways (Helpman, Melitz & Yeaple, 2004). Hence firms were questioned if they were running at least part of their production activity in foreign countries through FDI. An indicator variable takes a value of 0 (96% of 387) for firms that were not involved in foreign production and value of 1 (4% of 387) for the rest. To end, other studies (e.g. Fernández & Nieto, 2006) also controlled for possible foreign alliances since they can provide resources and mitigate liability of foreignness and uncertainty related to internationalisation expansion of SMEs ( Lu and Beamish, 2001). In this study, an indicator variable has been created to investigate if the firm has foreign affiliates of which it owns at least 10% of shares. It takes value of 1 if firm had such affiliates (13% of 387) and value of 0 if not (87% of 387).

4. METHOD DESCRIPTION

The sample is constituted by cross-sectional data which will be analysed in different ways. First of all, this study makes use of the statistical software Stata 16 in order to conduct the necessary analysis. Furthermore, the Ordinary Least Square (OLS) multiple linear regression has been chosen among other statistical techniques for the fulfilment of the research objectives. Indeed, it is more suitable to test linear relationships between several independent variables and a dependent one. Also, it is reasonable to stress that the dependent variable in this study takes the form of a continuous variable in its natural logarithmic form.

Next section aims to provide proof that all these assumptions are met, before moving on and conduct the required analysis to test the hypothesis. Hence preliminary analysis will be conducted to test the reliability of the sample.

4.1 Preliminary analysis

Following Oscar Torres-Reyna’s (2007) work, preliminary analysis have been conducted to check for the validation of OLS assumptions. However, due to the cross-sectional nature of data in the sample, there has been no need to test for the independence of errors (Best & Wolf, 2014). Hence, preliminary analysis starts with tests designed to check for normality assumption.

First, normality assumption states that residuals of the model behave “normal”. Hence, following a normal distribution. According to the theory a normal distribution should follow the “Bell Curve” , while not be biased either on the left or right side (Best & Wolf, 2014). To test for normality assumption, a histogram for the distribution of residuals of the model has been generated in order to compare it to a normal distribution. Figure 2 provides the graphical examination.

Figure 2 shows that the distribution of residuals results to be slightly biased on the left side, and that it reaches a peak around 0 going a little bit over the Bell curve. However, Zeller & Levine (1974) argue that normality test is primarily used to identify abnormal distributions, while a relatively small misalignment is generally tolerated. To provide further proof, a non-graphical test has been conducted as well. Hence, Table 1 in Appendix 1 shows results from Shapiro-Wilk w test. This particular test has been defined as the most accurate to check for normality (Razali & Wah, 2011). The test poses as null hypothesis that the distribution of the residuals is normal. In this case the p-value is 0.08 and consequently I fail to reject the null hypothesis at a 95% confidence interval. Hence, on the basis of both graphical and non-graphical examination, I can conclude that residuals do not follow any other distribution such as Leptokurtic, Bimodal or Uniform and that normality assumption holds.

Second, homoscedasticity of the residuals had to be assured as well. The homoscedasticity assumption states that the variance in the residuals has to be homoscedastic around the regression line (Oscar Torres-Reyna,2007).This implies that residuals must not vary for lower or higher values of fitted values. Figure 3 gives a graphical visualization of the variance of the residuals in the model. Since residuals seem to decrease slightly as the fitted values increase, a Breusch-Pagan test has been conducted to investigate further. The test is shown in Table 2 in Appendix 2. This particular test poses that residuals are homogenous as null hypothesis. Since results show a p-value well above 0.05 (Prob>chi2 = 0.765), I fail to reject the null hypothesis and conclude that the variance of residuals is homoscedastic.

Third, linearity assumption needed to be checked. For this reason a probability-probability plot (P-Plot) of the standardized residuals in the model is shown in Figure 4. Due to the fact that residuals followed almost perfectly the line, the linearity assumption is proven to hold.

Figure 4: Probability-probability plot.

Moving on, the sample needed to be checked for multicollinearity as well. To pass the test, variables do not have to be highly correlated between each other, otherwise the significance of the OLS model could be negatively affected (Oscar Torres-Reyna, 2007). However, a correlation matrix will be presented in the next section when providing descriptive statistics. The matrix will help to understand if multicollinearity is present between the independent variables of the model.

Finally, to check for the presence of outliers a box and whisker plot has been generated, with particular regards to the dependent variable. Outliers are observations visualized as points outside the plot, on the up-side or down-side of it. Figure 5 shows that none of these fall outside the plot. Hence no outliers are present.

5. RESULTS

This section is meant to provide a meaningful presentation of the results stemming from the analysis. As first, the outcomes from descriptive statistics will be presented followed by a correlation table for independent and control variables. To end, result from hypothesis testing will be provided and relative significance will be discussed.

5.1 Descriptive statistics and Correlation table

Following the example of D’Angelo et al. (2016), descriptive statistics concerning the variables in the model are presented in two different tables to favour clarity and ease of reading. More specifically, Table 3 provides descriptive statistics of indicator variables, while Table 4 summarizes descriptive statistics of continuous variables. With regards to the latter, it is worth mentioning that a log-transformation has been applied to all of them, using Stata 16. Hence, all the continuous variables in the model are log-transformed, in order to correct abnormal distributions. Finally, a correlation matrix is presented in Table 5.

Some of the information displayed below in Table 3 were already discussed in previous section when listing and describing control variables. However information about country distribution of the sample were missing. Hence, Table 3 shows that Spain has the highest concentration of family firms in the sample (36% of 387), hence, following research

Table 3: Descriptive statistics for indicator variables

At a first glance, it can be easily noticed from Table 4, that the variable Innovation intensity has two missing values compared with the others. Hence, I will expect to test the hypothesis on 385 observations instead than 387. Furthermore, having a closer look to the minimum and maximum of all the continuous variables shown in Table 4, the effects of the logarithm transformation can be observed. As stated before, by doing that non-normal distribution among variables are less than an issue.

5.2 Regression and moderator analysis results

The outputs from the statistical analysis are shown in Table 7, in which 6 different models are displayed. The models have the purpose to give numerical values to the impact of the main independent variables, control variables and interactions between some of them, on internationalization speed. It is worth mentioning that the impacts of Family ownership and External management on the dependent variable, have been analysed in two different models. Hence, External management is missing in Model 2, while Family ownership is missing in Model 3. Besides that, control variables are taken into consideration in all 6 models. Finally, the moderator variable has been added from Model 2 moving forward, to detect its individual impact also on internationalization speed.

As Table 7 shows, the first model is meant to measure the influence of the control variables on internationalization speed. Therefore, only variables capturing financial performance, age, specific industry classification and size of the firm were included, beside country from which the firm exports, presence of foreign affiliates and FDI. However, due to the presence of two factor variables, Pavitt classification and Country, one category from both variables has had to be used as reference category in order to avoid the Dummy variable trap. Following the work of D’Angelo at al. (2016), which also controlled for Pavitt classification and a regional variable similar to the one used in this study (Country), I will take the categories which are more relevant for both (respectively: Traditional; Spain), and I will pose them as reference categories for the regression analysis. As stated before Model 2 and 3 embody the effect of the main independent variables on internationalization speed separately, adding Innovation intensity in the analysis as well. The models are meant to test for hypothesis 1 and 2 respectively. Model 4 includes the interaction between Innovation intensity and Family ownership, and it has the purpose to test for hypothesis 3a. Moving on, Model 5, introduces the interaction between Innovation intensity and External management and it tests the validity of hypothesis 3b. Finally, Model 6 includes all the variables plus the interactions in a single model.

Model 3 has been designed to test for hypothesis 2. Therefore, external management has been added to the model excluding family ownership. Although, Model 3 is strongly significant, R-squared lowered down to 0.213 compared to Model 2. This suggests that External management hasn’t had any significant effect on internationalization speed. Indeed its B-value is not statistically significant for each level of significance (p<0.1; p<0.05; p<0.01). Therefore on the basis of these results, Hypothesis 2 assuming a positive linear relationship between external management and internationalization speed, cannot be confirmed.

Model 4 provides the outcome from moderation analysis in order to test for hypothesis 3a. This states that as the value of innovation intensity increases, the positive impact of family ownership on internationalization speed also increases. However, the results are surprising, showing the opposite scenario. More specifically, a B-value equal to -0.225 emerges with moderate significance (p<0.05). This means that a 10% increase in Innovation Intensity leads to a decrease of 2% in internationalization speed in SMEs with increasing levels of Family ownership. Hence Hypothesis 3a is rejected since the relationship between family ownership and internationalization speed turned out to be negatively affected by Innovation Intensity. The determinants and implications of this unexpected finding need to be examined deeply, thus they will be addressed in the next section.

5.3 Robustness check

It’s a common practice in empirical research to conduct robustness check to test for the reliability of findings and research method. More in detail, a robustness check tests if the validity of the findings holds when some conditions are changed in the regression analysis. When the same results can be obtained again, research structure is valid and results are acceptable and robust (Lu & White, 2014). In this study, few robustness checks have been conducted to test the validity of the findings under different conditions.

More specifically, Table 8 shows the results from 6 different models designed to test for Hypothesis 1,2,3a and 3b again. However, this time the analysis are conducted only on a sub-sample (only family firms from Spain). The reason of the choice of this particular robustness check is to observe if any country effects influence the results of the analysis. Hence, Model 1 tests for the effect of control variables on internationalization speed, and is composed of 143 observations; while 142 observations will be under analysis in the other models. Results should confirm what has been stated in the previous section. However, different results emerged for some control variables compared to Table 7: ROA turned out to be not significant (B = -0.049; p>0.1) as opposed to Specialized supplier (B = 1.029; p<0.05) in Model 1; Foreign affiliates proved to be significant in Model 4 (B = 0.489; p<0.1) and Model 6 (B = 0.490; p<0.1) but not in Model 1 (B = 0.445; p>0.1); while Innovation intensity in Model 5 (B = 0,213; p>0.1) lost its significant effect (B = 0.213; p>0.1).Despite these differences, all the other coefficients have confirmed the same conclusion reached in section 5.2. Indeed, the effect of the independent variable Family ownership on internationalization speed remains significantly positive on a 95% confidence interval (B = 0.460) in Model 1. On the other hand, the moderating effect of Innovation intensity on Family ownership turned out to be even stronger in Model 4 (B = -0.394; p<0.01) compared to Table 7. While the coefficients for the impact of external management on the dependent variable stayed insignificant (p>0.1) in both Model 3 and Model 5.

However, in this study External management has been operationalized as an indicator variable. Hence, it could be argued that a better measure of the presence of NFMs within the firm, could give more insightful results. Indeed, the indicator variable used in the models displayed in Table 7, fails to capture at what extent the increase of the number of NFMs affects internationalization speed. The variable takes value of 1 even if the firm has hired only 1 NFM. Hence the effect of the presence of 1 NFM on internationalization speed has been equalized to the effect of the presence of multiple NFMs in the firm. To overcome this bias, it is advisable to operationalize External management making use of a count or a continuous variable. Hence, Table 9 shows the results from 4 different models designed to test only Hypothesis 2 and 3b. This time, using a count variable from the EFIGE dataset capturing the exact number of NFMs in the family firm. The models resulted to be composed by 317 observations, due to missing values.

Finally, using a count variable for measuring the impact of NFMs on

internationalization speed, might cause a simultaneity bias. In other words, if it can be proven to be true that a greater number of NFMs can positively impact

internationalization speed, it might also hold the opposite. This would represent the case in which FBs, that aim to increase internationalization speed, might decide to hire more NFMs to overcome resource constraints and exploit their social networks. Appendix 4 shows results from a test meant to test for it.

6. DISCUSSION

The main goal of this research was to explore the relationship between two of the major sources of heterogeneity among family firms and internationalization speed by investigating how family ownership and external management affect export intensity in family owned SMEs. Furthermore, the moderating role of innovation intensity (H3a-H3b) has been examined as well, in order to observe how the relationship mentioned before vary according to higher degree of innovation intensity within the firm. As outlined in previous sections, full support has been found for Hypothesis 1. On the contrary, Hypothesis 2 could not be confirmed due to only partial acceptance in the model. With regards to the moderator role of Innovation Intensity, surprising findings have been found when testing for Hypothesis 3a. Indeed, a significant opposite effect emerged as opposed to the one originally thought. While no support was found for Hypothesis 3b.

The presence of external managers beside family ones to explain potential increases in internationalization speed, has not been proven to be significant in this research. Hence H2 could not be confirmed. This means that when experiencing an increase in export activities, there’s no difference between firms that have hired NFMs and those that have not. This might be because FBs experience the greatest positive effect of external management mostly in the pre-internationalization stage of the firm, rather than the subsequent evolution of the process. More specifically, external managers are better able to influence the willingness of FBs to start the internationalization process, rather than increase its intensity once the process has already began. Moreover, due to the nature of the firms in the sample (family owned SMEs that have already internationalized) NFMs might be not able to have a significant impact on the decision making process of the firm. Indeed, Miller et al. (2014) found that ownership structure can be critical to positive impact of NFMs. Family owners may be reluctant to delegate strategic decisions to managers that do not share their same attachment to company values and non-economic goals, especially with regards to international operations which are risky in nature and a potential threat for SEW. Hence, the management of the firm might be more centralized in favour of family managers as opposed to NFMs.

reasonable to deduce that when the FB is highly innovative and international, allocating a great amount of resources to innovation strategies as well, international expansion might result to be slower as a result of the few resources left and the willingness to preserve SEW. However, there might be another reason underpinning this finding. Indeed, family SMEs with higher degrees of family ownership might be more willing to exploit their domestic markets to spread innovation costs, rather than increase export intensity towards foreign ones. Indeed, previous studies (e.g. Enthorf & Pholmeier, 1990; Wakelin, 1988) have shown how innovative firms might be more inclined to use innovation to exploit domestic markets first. This might be the case of family owned SMEs with high degrees of family ownership as well. More in detail, SMEs usually reach a good positioning in the domestic market before to internationalize. Hence, family owned SMEs that are both innovative and international, might rather to exploit domestic market first than foreign ones. This might be because they have a better knowledge of the domestic market or a greater market share as opposed to foreign markets. Hence they might be able to lower risk perception about potential loss of SEW, and spread innovation costs better.

With regards to the analysis of the moderator role of innovation intensity on the speed of internationalization of FBs that hired NFMs as well, no significant findings have been produced. This might suggest that NFMs have not the ability and authority to produce significant increase in innovation investments in FBs.

6.1 Theoretical contributions

Some theoretical advancements can be discussed as a result of this research. First of all, studies on internationalization speed of FBs, and more specifically of family owned SMEs, are scarce. The investigation of possible moderators has received no much attention as well. Hence, this research contributes to the literature despite the fact that not all the assumptions made in the study have been proved to hold.

businesses has been surmounted. This study have shown that significant conclusions can be drawn by looking more specifically at the impact of sources of heterogeneity among FBs on internationalization speed. The view that SEW of family firms is mainly a constraint to FBs internationalization has been rebutted. This study proves that high degrees of family ownership turn preservation of SEW into a facilitative factor to increase internationalization speed. Second, this research has shed some light on how the internationalization process of FBs evolves by looking at internationalization speed. Indeed, the vast majority of studies in FB literature have mostly focused on the pre-internationalization stage of the firm. Hence, more emphasis has been put on what is next, and which are the factors behind different behaviours pursued by FBs once internationalized.

6.2 Managerial implications

Findings from this study imply managerial implications as well. Family managers of less or not internationalized firms, can learn that their fear of loss of SEW can turn into a powerful driver for international expansion. Indeed, they can apprehend that firm’s internationalization process can benefit from some distinctive traits of the family firm in terms of speed with which the process evolves after it starts.

6.3 Limitations and future research

Like many other scientific studies, this research does not come free of limitations. First of all, the biggest limitation comes from the cross-sectional nature of the data used in the analysis, since they stem from answers to a questionnaire that has been conducted only once (2010). For this reason, it can be argued that since the focus of this research is on internationalization speed, longitudinal data would have been more appropriate. Indeed, internationalization speed is a process that evolves over several years, thus longitudinal data would have been better able to capture changes to a greater extent, providing better insights. Furthermore, data have been collected during a period of major recession which might have altered strategies of family firms, and not only. Hence, it is advisable for future studies to make use of longitudinal data covering a different time period.

Second, it could be argued that the findings of this study come from analysis of biased sample. Indeed, as Table 2 shows, the sample is composed of 135 Italian firms which represent approximately 35% of the sample, resulting in potentially biased statistical results. Hence, it is recommended for future research to work on more equally distributed samples in terms of firms’ country of origin.

Third, due to dependent and independent variables are derived from the same questionnaire which has generated the EFIGE dataset, it is also arguable that the reliability of the findings might have been affected.

Finally, internationalization speed has been operationalized as increase in export activities between two points in time as suggested by Casillas and Acedo (2013). However such a measure does imply a big limitation. Indeed, increase in export activities might be generated only by a rise from the demand side as well. This would provide less meaningful insights with regards to the real effort pursued by the family firm to increase its internationalization speed.

7. CONCLUSION

This study differs from the research approach of many previous studies in the FB literature that analysed the relationship between family businesses and the internationalization process, for two main reasons. First, it specifically looked at the effects of two main sources of heterogeneity among family firms on internationalization speed. Whereas, great part of previous work has treated family firms as a homogeneous group simply comparing it to non-family family ones. Second, it focused on a specific dimension of the internationalization process: speed. In order to do that, a different measure of internationalization speed has been operationalized as opposed to the one conventionally used. The objective was to shift the attention from the pre-internationalization stage of family owned SMEs, to the speed with which they increase their international commitment over time. By following this distinctive research approach, this study aimed to answer the following research question:

“How do family ownership and external management affect internationalization speed of family owned SMEs ?”