Edinburgh Research Explorer Internationalization pathways among family-owned SMEs

Purpose – To increase understanding of the internationalization of family firms; to investigate how the framework by Bell et al. (2003) on the internationalization patterns of firms could explain the internationalization pathways taken by family-owned SMEs; to identify typical patterns and features in the various pathways taken by family-owned SMEs. Design/methodology/approach – This paper reports findings from an in-depth multiple case study with eight Finnish family-owned SMEs. Findings – The ownership structure had the most important role in defining the internationalization pathways followed by the family-owned SMEs: a fragmented ownership structure led to traditional internationalization pathway whereas a concentrated ownership base led to born global or born-again global pathways. Practical implications – Family entrepreneurs should carefully consider the division of ownership and seek to build new relationships in foreign markets in addition to their primary co-operators. Originality/value – In this study, we extend the integrative model of small firm internationalization by Bell et al. (2003) toward family-owned SMEs. Secondly, this study highlights the most important dimensions in the different internationalization pathways of family SMEs. Thirdly, it integrates the ownership dimension within discussion on differing internationalization pathways. Fourthly, it utilizes a family business specific perspective (the stewardship perspective), in order to understand the specific features of internationalization among family SMEs, and also how these features differ between family SMEs and other firms.


INTRODUCTION
The family is the original economic unit, and from it are derived all other economic organizations (Schulze and Gedajlovic, 2010). As recently as the start of the 20 th century, all businesses were family-owned: the presence of the family in the business was taken for granted, and there was thus no need to label a business as a family business. The world has changed dramatically since those times, not least because of globalization, but family firms are still of great importance to any economy (see e.g. Kraus et al., 2011). Family SMEs form the majority of all firms around the world: about 85% of all the firms in the EU and USA (IFERA, 2003) and an even a greater proportion in the developing countries are family-owned. Furthermore, they account for an enormous percentage of the employment, the revenues, and the GDP of most capitalist countries (IFERA, 2003;Sharma et al., 1996;Shepherd and Zacharacis, 2000).
Despite all this, it was only at the start of the present millennium that the merits of family firms started to be re-evaluated in top-tier management journals (Schulze et al., 2001).
Management researchers in particular have tended to be positive about family governance (Schulze and Gedajlovic, 2010). The unification of ownership and management in family firms enables the CEO to make opportunistic investments and/or rely on intuition (Gedajlovic et al., 2004). Hence, family firms have the potential to adapt to changing environments, launch products, and enter markets that investor-controlled or managerially-led firms are unable to address (Dyer and Whetten, 2006). In adverse economic conditions, family firms have been found to sustain more profitable businesses than firms with other ownership structures (Sirmon and Hitt, 2003). Furthermore, although it was long thought that large multinational corporations had an overwhelming position in international business , it has recently been recognized that substantial numbers of entrepreneurial and family firms are active in the international arena (Casillas and Acedo, 2005). It is this recognition that has led to family business internationalization becoming an important research area Thomas, 2006, 2008;Nieto, 2005, 2006). However, despite the importance of family firms to local economies, scholars have studied the internationalization of family-owned firms to only a very limited extent (Kontinen and Ojala, 2010). Bell et al. (2003) present three typical internationalization pathways of SMEs within a single integrative model. The model includes (i) traditional SMEs, which internationalize gradually and incrementally to foreign markets, (ii) born-again globals, which suddenly internationalize as a result of critical events, such as changes in ownership and management, or a takeover by another company possessing international networks, and (iii) born globals, which internationalize soon after inception (Bell et al., 2001(Bell et al., , 2003. The internationalization of family firms has commonly been seen as following traditional internationalization pathways Claver et al., 2007;Graves and Thomas, 2008). Thus, Bell et al. (2004, 44) view family ownership as linked to a "cautious and reluctant approach to internationalization." However, some findings indicate the existence of family firms that internationalize rapidly to several countries, for instance, after the succession of the firm to the next generation; these, then would follow the born-again global (Graves and Thomas, 2008) or born global pathway. Overall, we do not know much about the factors that could explain the different internationalization pathways taken by family firms. The aim of this study is to investigate how the framework by Bell and his co-authors explains the internationalization pathways of family SMEs, and to find features behind the different internationalization pathways taken by family firms. The research questions addressed here are as follows: 1. What kind of internationalization pathways do family SMEs take? 2. What kinds of features lie behind different internationalization pathways?

THEORETICAL BACKGROUND
Several traditional internationalization theories describe internationalization as an incremental process (Bilkey and Tesar, 1977;Cavusgil, 1980;Johanson and Vahlne, 1977;Johanson and Wiedersheim-Paul, 1975;Luostarinen, 1979) in which firms internationalize their operations from nearby countries to more distant countries. The Uppsala model (Johanson and Vahlne, 1977;Johanson and Wiedersheim-Paul, 1975) is probably one of the most cited traditional theories in the literature of international business. It was developed in the 1970s to explain the incremental internationalization process of multinational firms. In market selection, firms are expected to enter first into nearby markets, where there is a similar language, culture, political system, level of education, level of industrial development, etc. Thereafter, when a firm's knowledge of international operations increases, it gradually starts to develop activities in countries that are psychically more distant. This argument is based on the assumption that the business environments in countries that are psychically close are easier to understand, making business operations easier to implement.
Recently, Johanson and Vahlne have updated their model in parallel with new findings on firm internationalization. In their recent model (Johanson and Vahlne, 2009) they have put more emphasis on networks (as initiated by Johanson and Mattsson, 1988) and opportunity recognition within the internationalization process. According to Johanson and Vahlne (2009), firms are increasingly tending to struggle with the liability of outsidership rather than the liability of foreignness. In other words, they see a firm's problems and opportunities as becoming less a matter of country-specificity and increasingly related to relationship-specificity and networkspecificity.
Because the development of the Uppsala model was based on large multinationals, it has been frequently challenged in the field of international entrepreneurship (see e.g. Autio, 2005;Bell, 1995;Kraus, 2011;McDougall, 1994, 1997). The main critique has been targeted at its inadequacy in explaining the internationalization of SMEs, particularly in hightechnology related industries. INV theory 1 has attracted increasing attention since the seminal work of Oviatt and McDougall (1994). The theory is motivated by the observation that the internationalization of INVs is related to opportunity-seeking behavior, in which an entrepreneur "seeks to derive significant competitive advantage from the use of resources and the sale of outputs in multiple countries" (Oviatt and McDougall 1994, p. 49). It proposes that the international origins of INVs derive from their commitments to valuable resources in more than one country. In INV theory, "international from the inception" means that the founders of an INV seek growth opportunities in foreign markets, having already made some decisions related to the international scope of their activities before the foundation of the venture Oviatt and McDougall, 1994). The theory emphasizes the fact that INVs do not actually have to own their resources, since they are able to use external resources in international markets. Already in Oviatt and McDougall's (1994) theory as originally proposed, network structures were seen as a valuable resource, with cooperation within a network creating new opportunities for INVs. Because these network relationships cross national borders, it is suggested that the founding teams of INVs must already have prior knowledge of international markets (Oviatt and McDougall, 2005). The main difference between the Uppsala model (Johanson and Vahlne, 1977;Johanson and Wiedersheim-Paul, 1975) and INV theory is that INV theory suggests that firms can skip stages, or not have any stages at all in their internationalization process .

The integrative model of small firm internationalization
In their integrative model, Bell et al. (2003) present three different internationalization pathways of SMEs, combining ideas from the Uppsala model and INV theory. In line with previous literature (Kuivalainen and Saarenketo, 2012;Loane et al., 2004;Mathews and Zander, 2007), we use the term "pathways" to refer to a variety of strategies used in firms' internationalization processes. These include several stages, often distinguished by scholars in terms of three dimensions, namely (i) time, referring to the rapidity and pace of internationalization, (ii) scale , viewed in relation to foreign sales, and (iii) scope, with reference to the number of countries in which a firm operates (see Kuivalainen et al., in press).
The first pathway describes traditional firms that internationalize slowly and incrementally to psychically and geographically close markets. The internationalization of traditional firms comes about in an ad hoc manner, and is based on unsolicited orders and enquiries from overseas. The objectives of the internationalization are mainly survival and growth (Bell et al., 2001(Bell et al., , 2003. The product development of the traditional firms focuses first on the domestic market and only thereafter on foreign markets . Generally, the internationalization of traditional firms follows the same pathway as described in the Uppsala model (Johanson and Vahlne, 1977;Johanson and Wiedersheim-Paul, 1975).
Born global firms form the second pathway. These firms internationalize to several foreign markets simultaneously and rapidly, and are less influenced by psychic distance. Thus, they commonly internationalize to the markets where their products sell particularly well. Their internationalization is reactive, and based on striving for first-mover advantage within niche markets (Bell et al., 2003). The products of born globals are developed for the international market rather than purely for domestic customers . This internationalization process is related to INV theory, since in the born global case entrepreneurs seek international opportunities based on industry knowledge and existing networks .
Born global firms are commonly defined as achieving foreign sales in a period of two to five years from their establishment, in addition to having at least 25 percent of their income from foreign sources, and operating in at least five countries (see Kuivalainen et al., in press for a further review).
The third internationalization pathway in Bell et al.'s (2003) model is the born-again global pathway. Born-again global firms have previously tended to focus on the domestic market, but internationalize suddenly as a result of critical events, such as a change in ownership and management, a takeover by another company, or client followership. The change in ownership or management reorients business activities and brings in new decision-makers with an international focus . The takeover helps the firm to acquire more financial resources, managerial capability, international market knowledge, and access to the existing networks of the company taking over. In client followership, a domestic customer internationalizes its operations and the firm follows its customer to foreign markets (Bell et al., 2001). However, the studies of Bell et al. (2001Bell et al. ( , 2003 do not offer any guidance on how long the domestic period should be before a firm starts to internationalize its operations. In an empirical study, Sheppard and McNaughton (2012) used a 28-year domestic period as a criterion for born-again global firms. They found that born-again global are larger in size, and that they spend a smaller proportion of their resources on R&D than born global firms. In addition, it seems that born-again global firms operate in more foreign countries than born global firms (Sheppard and McNaughton, 2012).

Family firms and internationalization
Broadly speaking, the feature that makes a family business different from a non-family business is the involvement of the family in the ownership and management of the firm: a family business is a combination of the reciprocal economic and non-economic values created through a combination of the family and the business systems in place (Habbershon and Williams, 1999). Gersick et al. (1997) described family firms in terms of a diagram containing family, ownership and business dimensions, laid out on three axes (see Figure 1), the idea being that a perturbation in any of the three axes would also influence the other two. In the ownership dimension, one can identify the phases of controlling owner, sibling partnership, and cousin consortium. In the business dimension the phases of start-up, expansion/formalization, and maturity are typical. The phases in the family dimension can be described in the terms of young business family, entering the business, working together, and passing the baton.  Gersick et al. (1997).
The specific features of a family firm have been collectively called familiness. The term refers to the causal relationships between a business-owning family and the resources and capabilities of a business. Familiness is thus defined as "the unique bundle of resources a particular firm has because of the systems interaction between the family, its individual members and the business" (Habbershon and Williams, 1999, p. 11). Familiness may lead to hard-to-duplicate capabilities, and it can allow family firms to survive and grow in an adverse economic environment (Chrisman et al., 2005;Chrisman et al., 2006).
Stewardship theory focuses on the commitment and dedication of managers to the organization. It is based on the idea that the manager, in the role of a steward, feels a strong sense of duty towards the organization and places a higher utility on collective well-being than on individual well-being in aiming to improve organizational performance Miller and Le Breton-Miller, 2006). Miller et al. (2008) argue that family-owned businesses are more likely to exhibit stewardship attitudes towards the long-term well-being of the business (including both employees and customers) than non-family businesses. The family business will exhibit stewardship attitudes if it intends to keep the business vital, with a view to sustaining it over generations (Miller and Le Breton-Miller, 2006). At the same time, however, there is a risk of management entrenchment, leading to the deterioration of the company (Miller and Le Breton-Miller, 2006).
Family involvement in management has been seen as factor tending towards caution in the internationalization processes of family firms Claver et al. 2008;Kontinen and Ojala, 2010). According to George et al. (2005), internal owners generally appear to be riskaverse, with a corresponding decrease in the scale and scope of internationalization; by contrast, the entry on the scene of external owners -in other words institutional owners and venture capitalists -significantly increases the scale and scope of internationalization. Although the internationalization of family firms is commonly characterized as slow and avoiding risk, such firms may sometimes internationalize rapidly, for instance, in the context of a generational change (Graves and Thomas, 2008). The reasons for the slow pace may be, for instance, their limited growth objectives (Donckels and Fröhlich, 1991) and restricted financial capital (Gallo and Pont, 1996). In addition, there could be a connection with limited managerial capabilities (Graves and Thomas, 2006), an unwillingness to accept outside expertise, and a lack of bridging network ties (Graves and Thomas, 2004).
The factors enhancing the internationalization of family firms include a general long-term orientation, and speed in decision-making. In addition, it has been found that the FBs that are likely to be more successful in international expansion are those with a willingness to use information technology, a capability for innovation, and a commitment to internationalization, plus the ability to distribute power and use the resources that are available (Kontinen and Ojala, 2010). Generally speaking, the entry on the scene of new generations has been seen as having a positive influence on internationalization, although generational change has sometimes had no influence, or even a negative influence on internationalization (see e.g. Graves and Thomas, 2008).

METHODOLOGY
The methodology of this study is based on a critical realist case study. Welch et al. (2011) argue that (among scholars of international business) there are four different means of theorizing when conducting case studies: (1) inductive theory-building, (2) natural experiment, (3) interpretive sensemaking, and (4) contextualized explanation. 2 Three of the four means presented above, namely inductive theory-building, natural experiment, and interpretive sensemaking, are wellestablished methods of theorizing from case studies, whereas contextualized explanation, which is applied in the present research, is a more recent addition to the methodological literature (Welch et al., 2011). In recent years, there has been a strong trend towards decontextualization among case study researchers, with theorizing moving towards generalization and away from context (Welch et al., 2011). However, case studies are by nature rich in context, and Welch et al. (2011) see that the method of contextualized explanation has a great deal of potential for future case studies.
As indicated above, a critical realist case study method was applied in this study.
Following Easton (2010, p. 119), case research is here defined as "a research method that involves investigating one or a small number of social entities or situations about which data are collected using multiple sources of data and developing a holistic description through an iterative research process." In a critical realist case study, the research question addresses a research phenomenon of interest, in terms of discernible events, and asks what causes them to happen (Easton, 2010). In this study, the internationalization behavior of family SMEs was studied through an examination of the most important events and views related to their internationalization history. Based on this, and on an investigation of all possible secondary material on the case firms, the most important features related to their internationalization pathways could be identified and discussed. The choice of multiple cases made it possible to identify the subtle similarities and differences within a collection of cases (Brown and Eisenhardt, 1997;Eisenhardt, 1989;Yin, 1994).
The criteria for selecting the case study design as the primary method for the present study were the following: (i) the context: the case study method enables the researcher to study phenomena that cannot be separated from their context (Bonoma, 1985). To understand and explain the internationalization pathways of family SMEs, it was essential to examine the context; (ii) the complexity of the phenomenon under study: in an entrepreneurial process there are several components interacting simultaneously, and the phenomenon is connected to the organizational context. Case study research makes it possible to capture these different dimensions at the same time (e.g. Eisenhardt, 1989); (iii) the limited number of studies on the phenomenon of family SME internationalization.
Since the number of studies concerning family SME internationalization is limited (25 scholarly articles up to 2009, see Kontinen and Ojala (2010)), and since there is evidence that familiness really does make the internationalization of family SMEs different from that of non-family SMEs, it seemed appropriate to conduct a case study. Furthermore, most of the existing studies on FB internationalization have been confirmatory (statistically verifying theory-driven hypotheses), and there have not been many exploratory studies, such as case studies.

Case selection
The research setting was eight Finnish internationally operating family firms 3 . The sampling strategy was purposeful sampling. To be eligible as a case firm, the following criteria had to be fulfilled: (i) the firm had to be Finnish, (ii) the firm had to have fewer than 250 employees in the mid-1990s, hence fulfilling the criteria of the Finnish government and the EU for classification as an SME (OECD, 2003), (iii) the firm had to belong to the manufacturing industry, (iv) it was necessary for the firm to be family-owned, with the family controlling the largest block of shares or votes, having one or more of its members in key management positions, and having members of more than one generation actively involved with the business 4 . Table 1 summarizes the key information on the case firms. The firms were established between 1876 and 1988. The number of personnel varied from 18 to 249 employees, the average being 106 employees. 3 The proportion of family firms in Finland is 80 % (Tourunen, 2009

Data collection
Multiple sources of information were used to gather data from each case firm. The main form of data collection was interviewing, but in addition secondary materials, such as web pages, annual reports, financial records, meeting minutes, and brochures were utilized (see Table 2). The secondary material was used to understand the history and the products of each firm, to form detailed case histories, and to understand the circumstances behind certain events, with particular reference to aspects such as foreign market entries and changes in the operation modes. The secondary material was also utilized to triangulate with the information given by the informants.
The interviewees were selected from those persons who had most in-depth knowledge concerning internationalization, and they included executives (entrepreneurs), managing directors, managers of international affairs, and sales administrators. Following Svendsen (2006), at the beginning of the interview, neutral and non-threatening questions were asked to establish a relationship of mutual trust. The interviewees were first asked to describe their business in general and thereafter their operations related to internationalization. When the main issues of the interview were touched on, short questions such as "Could you describe this? How? Why?" were posed to go deeper into the issue. All these questions were developed according to the guidelines issued by Yin (1994), with the aim of making the questions as non-leading as possible. This encouraged the interviewees to give authentic answers to the interview questions.
Because the interviews focused on the entrepreneurs' past experiences and some of the firms were rather old and started their international operations a long time ago, we followed the guidelines for retrospective studies issued by Miller et al. (1997), and by Huber and Power (1985). Hence, we (i) compared information provided by the informants, (ii) asked about concrete events and facts, (iii) encouraged informants to give precise information rather than past opinions or beliefs, and (iv) utilized the written material of the firm to facilitate the recall of past events.
All the interviews were digitally recorded and transcribed verbatim using a word processor. During the second listening, correspondence between the recorded and the transcribed data was ensured. The complete case reports were sent back to the interviewees, and any inaccuracies they noticed were corrected on the basis of their comments. In addition, e-mail communication was used to collect further information from the interviewees and to clarify inconsistent issues, if necessary.

Method and process of analysis
The method utilized in the data analysis was content analysis. The analysis of the case data consisted of three concurrent flows of activity (Miles and Huberman, 1994): (1) data reduction, (2) data displays, (3) conclusion drawing /verification. In (1) (data reduction), the data were focused and simplified by writing a detailed case history of each firm. This is in line with Pettigrew (1990), who suggests that organizing incoherent aspects in chronological order is an important step in understanding the causal links between events. Thereafter, on the basis of the interviews, the unique patterns of each case were identified and categorized into the patterns observed under the sub-topics derived from the research questions. In addition, checklists and event listings were used to identify critical factors related to the phenomena encountered (Miles and Huberman, 1994). In (2) (data display) the relevant data were collected in matrices, graphs,

charts, networks, and in Tables in Microsoft Excel. In (3) (conclusion drawing and verification)
we concentrated on identifying the aspects that appeared to have significance. At this stage we noted regularities, patterns, explanations, and causalities relating to the phenomena.

Internationalization pathways of family SMEs
Six of the case firms followed a traditional pathway to internationalization, one a born-again global pathway, and one the born global pathway (See Appendix 1). These pathways correspond to the different internationalization patterns discussed by Bell et al. (2003). Concerning the overall procurement of internationalization, it was incremental and gradual among Firms A, B, C, D, G, and H, whereas it was quite rapid in the born global family SME (Firm F), and also in the born-again global family SME (Firm E), following its generational change.
The traditional firms A, B, C, D, G, and H were not regarded as born globals, since none of them (i) had at least 25 percent of their income from foreign sources, (ii) operated in at least five countries, or (iii) employed a direct operation mode during their first five years (see Kuivalainen et al., in press). Furthermore, none of them experienced a sudden change in the scale and scope of their internationalization such as to designate them as a born-again global firm (see Bell et al., 2003 However, in contrast to the findings by Bell et al. (2003), the born global family SME did not conquer many markets at the same time, tending rather to follow a year-by-year progression.
Firm E can be regarded as a born-again global firm, since -after its generational change in the mid-1990s -it internationalized very extensively. This is in line with Bell et al. (2003Bell et al. ( , 2004 who argued that a change in ownership/management may lead to a readjustment of business activities and may cause rapid internationalization. Firm E exported to ten European countries and had a subsidiary in Poland before the son came on the scene. However, the exports were somewhat unsolicited and experimental, and were directed at nearby countries. The establishment of the Polish subsidiary was based on friendship rather than a strategic decision. The second generation radically altered the internationalization strategy of the firm. Initially, the son went to Central Europe and strengthened the firm's international network ties. From this point, exporting became more strategic and regular. Since the generational succession (which occurred in 1995), the internationalization of the firm has been conducted extremely vigorously.
It now has subsidiaries in fourteen countries, and sales in over sixty countries worldwide.
The internationalization patterns of the case firms are summarized in Figure 2 below.

Dimensions behind different internationalization pathways among family SMEs
In this section, the dimensions behind these three specific pathways (traditional, born global, born-again global) will be discussed. The dimensions are based on a careful examination of the interviews and secondary data, as suggested by critical realism approaches, and by the method of theorizing related to contextualized explanation (Easton, 2010;Welch et al., 2011). From our data analysis, we found that the dimensions that best encompass the various internationalization pathways are: (i) ownership structure, (ii) stewardship attitude, (iii) international opportunity recognition, (iv) attitude to psychic distance, (v) the development of networks, and (vi) product. Concerning (i) the ownership base of the case firms (see Table 4), there are differences between the different family firms: when internationalization was launched among the traditional family SMEs, the ownership base was primarily sole ownership by the founder-manager, who embarked on internationalization incrementally. Interestingly, the changes in ownership structure modified the internationalization strategy of the firms to only a very limited extent, since, in the context of possible succession, the ownership was divided among multiple persons: between the siblings and the father (Firms B and C), several siblings (Firm H), several cousins and outside shareholders (Firm A), or siblings and other founder generation members (Firms D and G).
Typically, each owner had about 25% of the shares. For instance, in Firm B, the ownership was divided among two siblings (with 20% each) and the father (60%). The owner-manager (one of the siblings) considered this setting to be extremely demanding, since the siblings had totally different visions. The father, for his part, was trying to strike a balance between the two siblings while wishing that the new generation would decide on matters themselves. This was making it difficult for the owner-manager to develop the firm in any particular direction. In the case of Firm B, the owner-manager was currently planning to buy out the ownership shares of the sibling and the father, with the problem that this would be extremely expensive. As regards Firm C, the situation was similar to that in Firm B, but with the ownership divided between three siblings and the father. In this case the direction of the firm was even harder to define, since the father wanted to have equal ownership with his children and to give all of them the same rights.
The born-again global firm, by contrast, was characterized by concentrated ownership: it was transferred in its entirety from the founder-manager to the son; hence it is at present 100% owned by the current owner-manager, representing the second generation of the firm. The concentration of ownership enabled him to make his own decisions, and being a talented and bold successor, he took advantage of the situation to create a new strategy for the firm. As regards the born global firm, the founder-manager originally owned all the shares of the company, and the owner-manager still owns the majority of the shares (with family members owning over 70% of the shares and about 88% of the voting rights altogether). However, the firm was very founder-centered, meaning that despite the shares belonging to the other family members, he decided on everything related to the firm. In 1999 the firm went public, with 20% of the shares of the firm becoming owned by investors. In spite of this, going public has not really accelerated the internationalization of this firm. Altogether, these findings suggest that ownership structures other than institutional or venture capital ownership (George et al., 2005) can produce successful internationalization. However, as indicated above, fragmented ownership seemed to lead to cautious internationalization strategies. This puts forward the findings of Bell et al. (2004) that in addition to general ownership structures there can be variations in the ownership base that may impact on firm internationalization pathways. As regards (ii) the stewardship characteristics (e.g. Habbershon and Williams, 1999) of the case firms, it is possible to see in this case also a difference between the traditional and the born global and born-again global firms. Among the traditional family SMEs, the desire to guarantee the survival of the firm for future generations came through in all their thinking, whereas in the born global family SME, the owner-manager expressed a preference for having his sons create something new for themselves rather than just having the firm passed on to them. Nor did the owner-manager of the born-again global family SME make any reference to a need for careful management of the resources of the firm: on the contrary, he was prepared to make major decisions and investments at all times. However, both of these entrepreneurs were very proud of being important employers in their home region, and had a strong desire to guarantee that this would be so in poor times also.
In line with Gallo and Pont (1996) and Gallo and Sveen (1991), the traditional family SMEs seemed to be committed to their domestic tradition, and were not interested in a new kind of thinking that would develop the firm in the context of their FME. This being the case, they mainly had agents in the foreign markets. This was an entry mode with a high level of control: they were not ready to act in the foreign market in a different manner (Firm D), nor did they give entrepreneurial freedom to their subsidiary staff (Firm A). It should be noted that this feature is also related to ownership: when several family members have an equal share in the firm, it is hard to make any radical decisions in cases where the family members disagree on future strategies.
When one examines (iii) the context of international opportunity recognition (IOR), one can again see a difference between the traditional, the born global, and the born-again global family SMEs. For the traditional family SMEs, the context was either that of international exhibitions or an unsolicited order, thus they internationalized in a more ad hoc manner, as argued by Bell et al. (2003). This meant they started to internationalize when they were offered an opportunity to export, or when they met a suitable person in the international exhibitions. For the born global family SME, the IOR was based on an existing, strong, network tie and it differed significantly from the other case firms. In the case of the born-again global family SME, the original IOR was based on attending international exhibitions, but thereafter on the ownermanager's own active search for new international opportunities. Hence, in the case of the born global and born-again global family SMEs, there was more existing social capital or more of a self-initiated active search in the background of the IOR.
In terms of (iv) the attitude to psychic distance, there was again an obvious difference between the traditional firms and the born global and born-again global family SMEs. The traditional family SMEs experienced psychic distance very strongly and were willing to cope with it themselves. In contrast, the owner-managers in the born global and born-again global firms found no difficulties in dealing with the foreign customers and could see no important differences between the Finns and other nationalities. Hence, the psychic distance emphasized by Johanson and Vahlne (1977) was very strong in the case of the traditional family SMEs, but did not appear to have any influence on the FME of the born global and born-again global family SMEs. This is also consistent with the findings of Bell et al. (2003), to the effect that born-global firms in particular tend to internationalize to the leading markets for their products, irrespective of psychic distance.
In their most recent article, Johanson and Vahlne (2009) emphasize the liability of outsidership, and this view is strongly supported in the present study, even if the born global and born-again global family SMEs coped much better with this challenge. Hence, with regard to (v) network development, the traditional family SMEs actively developed trust with their cooperators, just as in the case of the born global and born-again global firms; however, the born global and born-again global family SMEs were able to do much more with their networks. In addition to this, the born global and born-again global family SMEs were able to draw back from controlling their network partners too strongly: they were able to give entrepreneurial freedom to their co-operators -an aspect which was regarded as extremely important by the co-operators themselves. The born-again global family SME was able to go still further with its networking: in addition to trust-building and the ability to give entrepreneurial freedom, it concentrated on developing new network ties and on replacing the poor network ties with new ones. This was something that the traditional family SMEs did not do: they concentrated solely on their primary partner in the foreign markets. The born global firm was situated in between these two: based on its strong network ties, it only built new network ties in markets where it experienced some difficulties, whereas in the markets with extremely trustworthy partners, it concentrated on maintaining ties with its existing partners.
Concerning the (vi) product perspective, here too there were differences between the traditional, the born global, and the born-again global firms. The products of the traditional internationalizers are traditional, consisting of high-quality manufactured goods, such as sauna stoves, log houses, fire safety equipment, and wooden toys. The product of the born-again global firm is an innovative manufactured product with strong environmental values and wide diversification for differing customer needs. In the born global firm, there are high technology products (pipettes and analyzers). In the traditional case firms, the product development procedures have not been very intense in comparison with those of the born global and bornagain global case firms. Nor have the traditional firms differentiated their product for the needs of foreign markets to a significant extent. By contrast, the born global and born-again global family SMEs have continuously renewed their products and strategies, and listened to their co-operators. Thus, the born global family SME has been characterized by customer-oriented product development and by new inventions as on ongoing activity. It has also continuously cooperated with universities and people in industry. The born-again global family SME, too, has changed its product strategy in parallel with changes in the market. It has planned a comprehensive product line, just like the born global firm. This means that it sells individuallytailored product entities all over the world. It has a carefully considered strategy according to which it sells, in addition to its packaging materials, tailored production units to produce its packaging material, provided the unit in question is not too close to its existing production and sales units. This is in line with Bell et al. (2003Bell et al. ( , 2004, who suggested that traditional firms tend to design their products for home markets, whereas the product development of born-again globals and especially born globals tends to be targeted at global markets.

CONCLUSIONS
This study makes several contributions in the fields of SME internationalization and family business. In addition, it extends the integrative model of small firm internationalization proposed by Bell et al. (2003), applying it to family-owned SMEs. First of all, we discovered that the ownership structure of the case firms played a central role, determining their internationalization pathways. We found that the founder-manager pattern of ownership generally implied incremental pathways involving traditional internationalization. If within the succession process this kind of ownership was divided between several family members and/or outside shareholders, the internationalization strategy remained the same. However, if the ownership was passed on in its entirety to the next manager, or if it remained concentrated, the internationalization strategy faced radical changes and/or was much more intense, with consequent born-again global or global pathways. It should be noted that the internationalization theories that exist in the background of the Bell et al. (2003) framework (the Uppsala model and INV theory) do not consider ownership to be a determining factor in the internationalization process. Furthermore, in the framework adhered to by Bell et al. (2003), the impact of ownership is described on a very general level. Hence, our findings contribute to internationalization theories in suggesting that ownership structure should be integrated within theories encompassing the internationalization of SMEs. The findings also contribute to the framework used by Bell et al. (2003) by giving more nuanced explanations as to how ownership structure and changes in the ownership structure impact on internationalization pathways. As an additional point, we add to the results obtained by George et al. (2005) by demonstrating that concentrated ownership may enhance internationalization, as may also the presence of external owners. In terms of our contributions to family business theory, we have been able to contribute to an understanding of how different ownership structures and forms of succession influence FB internationalization. As noted by Gersick et al. (1997), a change in any of the dimensions of a family firm affects the other two.
Our own data demonstrate that the division of ownership occurring in the context of generational change clearly affects the international business of family firms. Succession is an important area of research (see e.g. Kraus et al., 2011) in the family business research field, but the detailed influence of succession on the internationalization behavior of firms has not been investigated (see Kontinen and Ojala, 2010).
Secondly, a strong stewardship attitude seemed to lead to a traditional pathway, whereas a weak/moderate attitude was related to born global or born-again global pathways. This indicated that a strong sense of duty (see e.g. Miller and Le Breton-Miller, 2006) towards the family members led to cautious internationalization strategies. However, the born global and born-again global entrepreneurs found it very important to be important employers in their home region, and had a strong desire to guarantee that this would be so in the future. This contributes to family business theory by demonstrating how the influences on family business internationalization are connected with stewardship attitudes.
Thirdly, we found some interesting features in the networking behavior of the case firms.
These added some features to the integrative model of Bell et al. (2003). We observed that the case firms following a traditional pathway to internationalization concentrated solely on their primary partner in the foreign markets, whereas the born-again global case firm actively formed new networks in addition to its primary co-operators, so that it would be able to change operation modes and partners if necessary.
These three characteristics (ownership structure, stewardship attitude, and the development of network ties) seem to explain fairly well the different internationalization pathways among family SMEs, and may also be useful in explaining the internationalization behavior of SMEs in general. The two remaining dimensions (attitude to psychic distance, and product characteristics) are in line with earlier literature (Bell et al., 2003. They provided good explanations for the internationalization pathways of family SMEs, but do not appear to show differences from the internationalization patterns of SMEs in general.
Concerning further research directions, family firm internationalization clearly seems to be a topic of importance, which needs to be studied more in detail by internationalization researchers. Especially quantitative studies comparing the family firms and their non-family counterparts within one sample are needed. It would also be important to investigate how changes in ownership structure affect the internationalization of firms. The possible effects on stewardship attitudes of the age of the firm should also be investigated in the future, since the influence of age remained unclear in the present study. In addition, longitudinal studies would offer more detailed dimensions to the discussion related to the intertwinement of ownership and internationalization. Longitudinal studies would also help to avoid potential memory biases that are always problematic in the studies focusing on past events. For instance, in this study it was impossible to interview entrepreneurs who were involved in the first foreign market entries of Firms A, B, and H. This study also suffers from a small amount of born global and born-again global firms compared to traditionally internationalized firms. Hence, both qualitative and quantitative studies are needed to validate the findings here.
From the managerial perspective, family entrepreneurs need to pay more attention to the ownership structure of the firm, especially in the context of succession planning. For instance the division of ownership in between several siblings might cause problems to the sibling who is in charge of the company if the other siblings just take the company as a portfolio of investment and they do not want to risk any resources for pursuing internationalization. Family entrepreneurs should also be more active in building foreign networks, also outside their primary cooperators.