Analysis of the Inﬂuence of the Moment the Internationalization Process Begins on the Internationalization Intensity of Family and Nonfamily Businesses: An Approach Using a Tobit Model

: The speciﬁc characteristics of family businesses as well as the internationalization path followed can inﬂuence the intensity of the internationalization process. Many studies have analyzed how family character can inﬂuence the internationalization process of family businesses, and the results obtained have not been conclusive. Nevertheless, previous research has not sufﬁciently addressed the inﬂuence that the moment of initiation of the internationalization process has on the levels of internationalization achieved. Based on the behavioral agency model, the unique set of business resources (familiness), and the socioemotional wealth (SEW) perspective, this study examines, the internationalization intensity of family and nonfamily businesses in two deﬁned groups (early internationalization and internationalization from the local market). Likewise, the effect that the entry of the second generation has on the internationalization of these companies is analyzed. To perform this analysis, Tobit regression models are estimated from a data set of panel data from the Spanish Survey on Business Strategies for small and medium-sized Spanish family businesses from 2005 to 2016, ﬁnding that family ownership and management have a negative inﬂuence on the intensity of exports, regardless of the path of internationalization followed, and that the entry of new generations has a positive relationship with the level of internationalization of these businesses. Finally, implications of the ﬁndings for research and management are discussed.


Introduction
The globalization of markets has turned internationalization into an essential strategy for businesses, so the number of such firms expanding their operations into international markets has increased significantly. Internationalization provides businesses with a variety of potential benefits, but it also involves challenges and risks (for example, Barkema and Vermeulen 1998;Hitt et al. 2006). In the literature, there are opposing opinions on whether family businesses internationalize more or less than nonfamily businesses.
Family businesses have a unique set of business resources (familiness) due to the interaction between the family's systems, its individual members, and the business (Habbershon and Williams 1999;Habbershon et al. 2003), and this can play an important role in the internationalization process.
However, the literature on the internationalization of family businesses is contradictory. Some authors have found limitations in the internationalization of this type of business because they are less susceptible to internationalization (Gallo et al. 2004;Okoroafo 1999) and have limited resources and capabilities that constrain their international performance (Fernández and Nieto 2005;Fernández-Olmos and Díez-Vial 2015;Gallo and García-Pont 1996;Li 2007;Olivares-Mesa and Cabrera-Suarez 2006).
However, another group of authors have observed that the set of resources of family businesses can increase the willingness to expand internationally (Zahra 2003). In this regard, owners of family businesses have a long-term orientation (Claver et al. 2008(Claver et al. , 2009), they supplement a lack of resources through network relationships (Wright et al. 2005) are more susceptible to risk (Calabrò and Mussolino 2013;Chen 2011;Pukall and Calabrò 2014;Zahra 2003), and the preservation of socio-emotional wealth (SEW) can lead them to take these risks (Berrone et al. 2012;Hernandez-Perlines 2018).
As in many strategic decisions, internationalization involves possible positive and negative results. Strategic decisions can be considered as mixed bets and in the case of internationalization can explain the differences between family and nonfamily businesses (Alessandri et al. 2018). In this regard, nonfamily businesses are mainly focused on economic goals and are willing to exchange risks for higher returns, while family businesses seek economic and non-economic objectives because of their concern for maintaining SEW (Alessandri et al. 2018;Hernández-Perlines et al. 2021).
The lack of consensus in the studies published may be due to the different moment in which the businesses started their internationalization, since they develop this process based on two approaches: an early or from-the-start internationalization (hereinafter, EI) and an internationalization from consolidation in the local market (hereinafter, LMI).
In the case of small and medium family businesses (hereinafter, FB), its specific characteristics can clarify and justify the existence of different results with small and medium nonfamily businesses (hereinafter, NFB). In this regard, this study explores whether FB have different outcomes in terms of internationalization intensity, depending on the time when the process started.
Furthermore, in the previous literature, different studies have shown that generational change, which is a specific factor in family businesses, and the incorporation of second and successive generations in the management of these companies can have effects on the strategy followed by family businesses. In this sense, this study also explores whether the entry of the second generation represents a strategic change that enhances the internationalization of FB.
This paper aims to contribute to research into the internationalization of family businesses in various ways: firstly, to cover the gap that exists due to the lack of sufficient studies about the relationship between the start time of the process and the internationalization level achieved in family businesses, and secondly, to analyze the influence that the specific characteristics of family businesses have on the levels of internationalization obtained.
This study also aims to contribute by reaching certain practical conclusions to help the owners and managers of FB to correctly choose the time for starting the internationalization process and how quickly they should move forward.
The document is structured as follows. The second section presents the theoretical framework, which serves as the foundation for the hypotheses. The third section describes the data set and statistical approach used. The fourth section shows the results of the empirical analysis and provides an interpretation of these. Finally, the fifth section contains the conclusions from the results obtained in the previous section, offering implications for owners and managers, pointing out the limitations of the study, and proposing lines for future research.

The Behavioral Agency Model and the Perspective of Mutual Commitment in Family and Nonfamily Businesses
The theoretical framework used in this study is based on the behavioral agency model. This model explains how risk preferences change with problem formulation and the importance of benchmarks (Wiseman and Gomez-Mejia 1998). Most strategic decisions can be seen as a mixed bet (Gomez-Mejia et al. 2014;Martin et al. 2013). Decision makers prefer to protect existing wealth, avoiding risky options. However, the potential for future wealth can relax loss aversion (Martin et al. 2013).
Some researchers have applied the behavioral agency theory and mixed bet to explain the differences in the strategies of family and nonfamily businesses (for example, Chrisman and Patel 2012; Gomez-Mejia et al. 2011).
Family businesses use two benchmarks, financial wealth and SEW, and weigh the potential financial gains and losses against profit and loss potential of SEW (Alessandri et al. 2018). However, preservation of SEW is the most important goal to make major strategic decisions (Berrone et al. 2012).
As for nonfamily businesses, their main reference point is to maximize financial wealth (Gomez-Mejia et al. 2014). These businesses are more willing to exchange risks for higher returns and are less likely to limit their internationalization strategies than family businesses (Gomez-Mejia et al. 2010). As a result, the pursuit of future wealth reduces concerns about the protection of the current wealth (Martin et al. 2013).
In summary, the literature applying the behavioral agency theory and mixed bet suggests that, given the different benchmarks of family and nonfamily businesses, family businesses tend to be more cautious in relation to nonfamily businesses; however, family businesses are not homogeneous in their prioritization of protecting SEW when making strategic decisions.

The Specific Characteristics of Family Businesses
Family businesses have a unique configuration of resources and capabilities, and their uniqueness explains the nature of their competitive behavior (Cabrera-Suárez and Olivares-Mesa 2012; Habbershon and Williams 1999;Sirmon and Hitt 2003;Tokarczyk et al. 2007). In the definition of a family business, Astrachan et al. (2002) include a set of features specific to family businesses that serve as defining elements for them (Rau et al. 2018).
This set of specific characteristics of a family business have been collectively called familiness or a unique set of business resources arising from the interaction between the family's systems, its individual members, and the business (Cabrera-Suárez and Olivares-Mesa 2012; Habbershon and Williams 1999;Habbershon et al. 2003;Kontinen and Ojala 2012). These specific characteristics, such as trust, altruism, and social capital (Calabrò and Mussolino 2013; Pukall and Calabrò 2014;Segaro 2012;Zahra 2003), long-term vision (Claver et al. 2008(Claver et al. , 2009, and greater entrepreneurship (Calabrò and Mussolino 2013;Chen 2011;Hernández-Perlines et al. 2020;Zahra 2003), can play an important role in increasing the willingness of the business in question to expand internationally (Zahra 2003).

The Different Internationalization Paths
Businesses do not always follow the same internationalization paths and those chosen can pose important challenges for them. For the purposes of this study, we will consider two types of processes. First, early, or from-the-start internationalization (EI), which corresponds to new international ventures (Oviatt and McDougall 2005) or born-global businesses (Bell et al. 2001(Bell et al. , 2003. They are young businesses which lack the resources necessary for internationalization and have not accumulated the experience necessary (Rennie 1993;Rialp et al. 2005) to deal with the risks and uncertainties associated with their lack of knowledge about their destination markets (Efrat and Shoham 2012;Zaheer 1995).
The second, an internationalization process starting from a local market (LMI) in which firms begin their adventure abroad after a period of consolidation of their local business. This second group corresponds to the traditional approach represented by the Uppsala School Vahlne 1977, 2009;Vahlne and Johanson 2017) and to the born-again global businesses (Bell et al. 2001(Bell et al. , 2003, which begin their internationalization suddenly after their consolidation in the local market. With respect to family businesses, they also follow the paths mentioned above, obtaining different performance depending on the path followed (Varas-Fuente et al. 2022). Some authors suggest that FB internationalize their businesses after consolidating their position in the domestic market, once they have accumulated the resources and skills necessary for their growth there (Segaro 2012). In keeping with the propositions established by the Uppsala internationalization model (Claver et al. 2007;Graves and Thomas 2008), they follow a sequential process into countries that are close from a geographical and/or cultural viewpoint (Child et al. 2002) and favor indirect methods when entering markets (Kontinen and Ojala 2010). However, other businesses follow a faster or accelerated path, rapidly internationalizing into several countries following two models: global born, starting their international expansion from their foundation, and born-again global, which are internationalized, for example, after the second generation takes over management (Fernández and Nieto 2005;Graves and Thomas 2008).
Family businesses have specific characteristics that can motivate strategic behaviors that lead to decisions on the path to take, either internationalization from the start or after consolidation in the local market. The interest of this work is focused on analyzing the relationship between the process followed and the intensity of internationalization achieved by this type of business.

Early Internationalization
As for the businesses that start an IE process, these are start-ups that, in general, will be led by founders.
According to Fernández and Nieto (2005), the founders of FB are often reluctant to make changes in the organizational structures and professional management systems that favor decentralized decision-making. SME family managers tend to be conservative and reluctant to internationalize as they perceive that it can lead to loss of business control, family wealth, family reputation, social status, and greater chances of family conflict (Arregle et al. 2012;Carney et al. 2015;Lahiri et al. 2020).
In addition, even if the founders are entrepreneurs, they may be reluctant to embark on internationalization due to their concentration of personal wealth in the business and their objective of consolidating the business's position in the national market (Menendez-Requejo 2005). In this line, Sciascia et al. (2012) showed that family participation negatively influences international entrepreneurship in family businesses. Zahra (2003) notes that the founders of some US family businesses avoided international expansion since it requires large commitments of resources. Fernández and Nieto (2005) observed that the percentage of Spanish export businesses is lower for first generation family businesses. For its part, Basly and Saunier (2019) found that the more searching of the owning family to maintain control and influence over the businesses, the lower the exports of family SMEs.
In this sense, it seems that the founders of family businesses will give greater importance to the preservation of SEW when making strategic decisions, assuming fewer risks in the internationalization process. Therefore, the following hypothesis is proposed: Hypothesis 1. Family ownership and management have a negative influence on export intensity when businesses follow an EI process.

Internationalization from the Local Market
Regarding businesses that follow an LMI process, these are businesses that start their internationalization processes after having consolidated their local businesses. The founder has already managed to launch his business and begins to worry about maintaining it for their descendants. Therefore, they have the experience obtained in these local markets and begin gradually or suddenly to expand into new foreign markets.
The research suggests that ownership significantly influences the strategic decisions of a business (Zahra 1996;Zahra and Pearce 1989), especially in family businesses where owners have a significant participation.
According to the behavioral agency theory (Wiseman and Gomez-Mejia 1998), family owners are not, a priori, risk adverse or risk seeking. Depending on the situation they will be willing to take big risks or no risks and to avoid the loss of SEW they will be more willing to assume risks than their nonfamily counterparts (Chrisman and Patel 2012; Gomez-Mejia et al. 2007, 2011, which can be accelerated when family business owners perceive that their SEW is at risk (Pukall and Calabrò 2014).
Internationalization can create new employment opportunities, growth of businesses, and new members of the family according to the non-financial goals that coexist in a family business with financial objectives (Gersick et al. 1997;Menendez-Requejo 2005).
The altruistic behaviors of family members (Schulze et al. 2001) imply that, if internationalization is important for the long-term success of the business and for increasing the employment of family members, then the owner-managers can follow this strategy, even when the perceived risks are high (Gallo and García-Pont 1996;Zahra 2003).
All of this leads us to think that once the risk aversion of the first years of the company's life has been overcome, the managers of family businesses will be more willing than the managers of nonfamily businesses to assume greater risks associated with internationalization with the goal of increasing socio-emotional wealth (SEW). For these reasons, the following hypothesis is proposed: Hypothesis 2. Family ownership and management have a positive influence on export intensity when businesses follow an LMI process.

The Entry of the New Generations
FB with greater family involvement in management opt for internationalization strategies through entry modes that involve greater risk and resource commitment (Andreu et al. 2020), and this greater family involvement usually occurs when the second generation appears on the scene. Different generations of owners usually have different interests, management styles, and objectives (Okoroafo 1999;Fernández and Nieto 2005).
Each new generation brings new strategic ideas that are based on the underlying competencies (Ward 1997) and it is possible to expect that subsequent generations will be more qualified (Fernández and Nieto 2005).
The second generation may be more motivated and ready to start projects abroad, while the founders may prefer stability (Menendez-Requejo 2005). Gallo and García-Pont (1996) note that family businesses tend to internationalize when the second and successive generations have joined the business, because they are better trained in international affairs and are looking for new responsibilities in the business.
For Calabrò et al. (2016), the participation of the new generations can be understood as a particular episode that can lead to an era of rapid and dedicated internationalization of the family business, after a time of action focused on the local market. Additionally, the presence of young successors favors internationalization (Bannò and Trento 2016).
Basly and Saunier (2019) found that the more the owning family seeks to renew family bonds through generations, the higher exports are in family SMEs.
In summary, subsequent generations of the founder have more training and information, being more prepared to undertake the process of internationalization.
For all of the above, the following hypothesis is posed: Hypothesis 3. The intensity of internationalization in FB is favored by the presence of the second generation.
The specific research model proposed is shown in Figures 1 and 2 (2019) found that the more the owning family seeks to renew family bonds through generations, the higher exports are in family SMEs.
In summary, subsequent generations of the founder have more training and information, being more prepared to undertake the process of internationalization.
For all of the above, the following hypothesis is posed:

Hypothesis 3. The intensity of internationalization in FB is favored by the presence of the second generation.
The specific research model proposed is shown in Figures 1 and 2.

Sample Selection and Data Collection
To test the theoretical hypotheses proposed, this paper uses the Spanish Survey on Business Strategies (SSBS). This database has been used in several previous studies on This database has an annual panel data structure and is produced by the SEPI Foundation, sponsored by the Ministry of Finance and Public Administrations. The reference population of the SSBS is Spanish businesses belonging to the manufacturing sector. The selection of a business was carried out through random sampling.
SSBS is chosen for its representativeness and for being panel data. While most researchers agree that the internationalization process is dynamic and time-dependent,   (2019) found that the more the owning family seeks to renew family bonds through generations, the higher exports are in family SMEs.
In summary, subsequent generations of the founder have more training and information, being more prepared to undertake the process of internationalization.
For all of the above, the following hypothesis is posed:

Hypothesis 3. The intensity of internationalization in FB is favored by the presence of the second generation.
The specific research model proposed is shown in Figures 1 and 2.

Sample Selection and Data Collection
To test the theoretical hypotheses proposed, this paper uses the Spanish Survey on Business Strategies (SSBS). This database has been used in several previous studies on This database has an annual panel data structure and is produced by the SEPI Foundation, sponsored by the Ministry of Finance and Public Administrations. The reference population of the SSBS is Spanish businesses belonging to the manufacturing sector. The selection of a business was carried out through random sampling.
SSBS is chosen for its representativeness and for being panel data. While most researchers agree that the internationalization process is dynamic and time-dependent,

Sample Selection and Data Collection
To test the theoretical hypotheses proposed, this paper uses the Spanish Survey on Business Strategies (SSBS). This database has been used in several previous studies on inter- This database has an annual panel data structure and is produced by the SEPI Foundation, sponsored by the Ministry of Finance and Public Administrations. The reference population of the SSBS is Spanish businesses belonging to the manufacturing sector. The selection of a business was carried out through random sampling.
SSBS is chosen for its representativeness and for being panel data. While most researchers agree that the internationalization process is dynamic and time-dependent, almost all existing models are of a static nature (Leonidou and Katsikeas 1996) and the use of crosssectional data limits the understanding of this process (Cabrera-Suárez and Olivares-Mesa 2012; Coviello and McAuley 1999). For this reason, this study has opted to use panel data.
From this database, the study has taken as a reference those businesses established since 1980 and with sales abroad, with the aim of obtaining a time series to analyze their evolution over time.
The next step was to apply a filter for the small and medium sized businesses in accordance with the definition established by the European Commission, which on 6 May 2003 adopted Recommendation 2003/361/EC, amending the recommendation from 1996 (businesses which meet the following parameters: fewer than 250 employees, business turnover that is less than or equal to 50 million euros, and a total balance sheet of less than or equal to 43 million euros). This left an initial 268 small and medium businesses (hereinafter, SMB); a total of 30 had mostly foreign social capital and were thus eliminated because the objective of this study is to investigate Spanish SMB. The final study sample was made up of 238 SMB.
For the 238 units finally selected, the data described in the following section have been obtained for the period between 2005 and 2016, both inclusive, which has resulted in an unbalanced data panel with 2.340 observations (2016 is the last year for which complete information is available in the official database used).
The above sample has been divided into two sub-samples (EI and LMI), based on the criteria set by Bell et al. (2001). Bell et al. (2001) define born global businesses as those that begin their intensive internationalization after 2 years from their establishment and born-again global businesses as those that begin their intensive internationalization after 10 years.
Applying the above criteria, the final sample to be studied includes the following number of businesses: 78 SMB (712 observations) that have followed the EI process and 160 SMB (1628 observations) that have followed the LMI process.

Dependent Variable
One dependent variable was used to evaluate the international participation of FB, export intensity, which measures the proportion of foreign sales over total sales. This variable is one well-established measure of internationalization (for example, Bonaccorsi 1992;Calof 1994;Nieto 2005, 2006;Wakelin 1998).

Independent Variables
Family business. The business belongs to a family with one or more members who hold management positions, which allows an identification of the family's capacity for effective control Nieto 2005, 2006;Menendez-Requejo 2005). From a question from the SSBS that looks at whether a family participates actively in the control and/or management of the business, a dummy variable with a value of 1 was added when the family participated actively in the management and control, and 0 in all other cases.
Second generation. Family businesses are classified into two groups according to the family generation in charge of them. Following Fernández and Nieto (2005), a dummy variable with a value of 1 was included to indicate a second generation of the family (business that are more than 30 years old) and 0 when it was the first generation of a family (less than or equal to 30 years).

Control Variables
In keeping with previous research, the following control variables, which may influence the dependent variables, are used: Business size. Following previous studies, this study uses the logarithm of the total assets of the business (for example, Lin 2012; Martínez et al. 2007;Muñoz-Bullón and Sánchez-Bueno 2012).
Leverage. It is known that the amount of leverage affects business performance (Lu and Beamish 2004). In keeping with previous research, this study measures this variable as debt as a percentage of the total assets of the business (for example, Gomez-Mejia et al. 2010;Graves and Shan 2014;Lu and Beamish 2004;Muñoz-Bullón and Sánchez-Bueno 2012).
R&D activities. Through R & D activities, businesses can create intangible assets that can allow them to gain competitive advantage when venturing abroad (Lu and Beamish 2004). This variable has been used in other studies through the concept of research intensity, calculated as R&D spending divided by total sales (for example, Almodóvar and Rugman 2014; Nieto 2005, 2006;Graves and Shan 2014;Lu and Beamish 2001).
Products diversification. When the business has a diversified production, it can count on a competitive advantage in the different markets and influence its internationalization (Muñoz-Bullón and Sánchez-Bueno 2012). From a field included in the SBSS that asks if the business is diversified or not, a dummy variable with a value of 1 was included when the business presented diversification and 0 when it was not diversified.
Sector. Following Nieto (2005, 2006), to capture the characteristics of the sector, the average export intensity of the industry and year is used.

Methods of Analysis
A Tobit model has been estimated to analyze the determinants of export intensity. The methodology is adjusted to process panel data.
A Tobit model is a regression model in which the range of the dependent variable is limited in some direction (limited dependent variables model). In this case, the dependent variable "export intensity" adopts the value 0 for several observations. This characteristic destroys the linearity assumption, which means that least squares methods are not appropriate.
According to Wooldridge (2006), the Tobit model expresses the observed response, y, in terms of an underlying latent variable: Therefore, the observable variable y is equal to y*, when y* ≥ 0, but y = 0, when y* < 0. In this research, several Tobit models are estimated to study the impact of different variables on the intensity of exports. Table 1 shows the distribution between FB and NFB for each of the two sub-samples analyzed.  Tables 2 and 3 show mean, standard deviation, variance inflation factor (VIF), and correlation matrices between the variables selected for each of the two groups analyzed.  The existence of multicollinearity between the explanatory variables has been verified by calculating the variance inflation factors (VIFs) of each of these variables. In all cases, the VIF obtained is lower than the cut-off value of 10 (Kleinbaum et al. 1988). Therefore, a priori, we could conclude the existence of non-multicollinearity. Table 4 shows the results obtained from the application of the Tobit model to the set of variables and observations for the EI group. In Table 4, model 1 includes the control variables. Model 2 includes the independent variable family business. Table 5 shows the results obtained from the application of the Tobit model to the set of variables and observations for the LMI group. As in the EI group, in the first model (3) the control variables are included, and in the second model (4) the independent variables family business and second generation are included. Hypothesis 1 was tested, using model 2. In this model, it is observed that the coefficient of the family business variable has a negative sign and a weak significant relationship (p < 0.05) with export intensity. For these reasons, we obtain a weak confirmation of the negative relationship between family participation and internationalization that is proposed in Hypothesis 1.

Analysis, Results, and Discussion
Model 4 shows that the coefficient is negative for the family business variable and has a significant relationship with export intensity. Therefore, we obtained the opposite result to that proposed in Hypothesis 2. This suggests that there is no difference according to the internationalization path followed, and in both cases, family participation has a negative influence on the degree of internationalization in relation to nonfamily businesses. This result is in line with that obtained, for example, by Alessandri et al. (2018), Arregle et al. (2017, Bannò and Trento (2016), Fernández and Nieto (2005), Graves and Thomas (2008), and Merino et al. (2015), who found a negative relationship between family ownership and level of internationalization of the business.
A possible explanation for this result is that, according to the behavioral agency model, when making important strategic decisions (such as internationalization), family businesses are faced with a mixed bet (Gomez-Mejia et al. 2014;Martin et al. 2013) with two points of reference (financial wealth and socio-emotional wealth-SEW) that must be weighed (Alessandri et al. 2018). Additionally, in this mixed bet, family managers have the preservation of SEW as their main objective, regardless of the internationalization path followed (Berrone et al. 2012), avoiding risky options. In this sense, Pukall and Calabrò (2014) conclude that the lower rate of internationalization observed in family businesses is related to the concern to preserve SEW. On the other hand, Bannò and Trento (2016), also based on socio-emotional wealth, discover that family ownership has a negative effect on internationalization. In addition, Arregle et al. (2017) conclude that, due to family ownership and management, family businesses are particularly affected by restrictive elements, such as SEW protection.
As for Hypothesis 3, in model 4 it is observed that the coefficient of the secondgeneration variable is positive and has a significant relationship with export intensity, which supports this hypothesis. This result is consistent with that obtained, for example, by Bannò and Trento (2016), Basly and Saunier (2019), Calabrò et al. (2016), and Fernández and Nieto (2005).
In terms of the control variables, in the EI group, variables' business size, R&D activities, and sector are significant in the two models included in Table 4, presenting positive coefficients, which is in accordance with previous research. The variables' leverage and diversification are not significant in the two models.
Regarding the control variables in the LMI group, variables' business size and sector are significant and with a positive sign in the two models included in Table 5, coinciding with previous research. However, the leverage variable is significant and has a negative sign, which means that higher debt leads to a reduction in internationalization levels. This result coincides with that obtained in previous research (for example, Chang et al. 2014), although it differs from others, which suggests a positive relationship between leverage and internationalization (Fernández and Nieto 2005). Finally, the variables diversification and R&D activities do not show significance in either of the two models.

Conclusions, Limitations, and Future Lines of Research
This study has proposed an analysis of the internationalization intensity of FB depending on the speed with which the process begins. Previous research has not focused on the differences between the two groups, and studies have addressed the problem as a whole. This paper examines this issue by exploring a set of variables that may influence the level of internationalization obtained by FB, from the consideration that the effect of these variables is different depending on the starting moment of this process.
From a theoretical perspective, this research contributes to the literature on the internationalization of family businesses, showing that the behavioral agency model and the concept of mixed bet can help to evaluate how the objectives of owners and managers of family businesses preserve socio-emotional wealth and influence the internationalization of these businesses. To the best of our knowledge, this research is the first that, based on the conceptual agency model combined with the unique characteristics of family businesses and with the SEW perspective, evaluates the influence that the internationalization path followed by businesses has on the levels of internationalization achieved.
The results obtained lead us to conclude that family ownership and management have a negative influence on the intensity of exports, regardless of the internationalization path followed.
Secondly, we have confirmed the positive relationship that the entry of the second generation has in the level of internationalization of these businesses.
This study extends the empirical literature by contrasting that the moment when the internationalization process starts does not involve different results of the internationalization intensity. In addition, the findings contribute to filling the existing gap in relation to studies on the internationalization process of FB that had not been sufficiently addressed for this type of businesses. Particularly, it looks at the influence of family participation and the presence of second generations of families. It has been found that in all family businesses (regardless of the internationalization path followed), family ownership and management have a negative influence on export intensity, a similar result as that obtained in previous investigations (for example, Alessandri et al. 2018 or Fernández andNieto 2005). It has also been found that there is an enhancing effect on international intensity due to entry of the second generation in the management team.
The results obtained have implications for the managers of the FB who are analyzing their possible internationalization. In the first place, they must consider that the participation of the second generation has a positive effect on the levels of internationalization. Second, they must consider that business size has a positive effect on those levels of international activity, and that investment R&D activities also has a positive influence when an early internationalization is followed.
The results of this study may also be important for authorities when implementing public programs aimed at helping businesses in general, and FB in particular, to begin and extend their internationalization journey. They must encourage the implementation of long-term internationalization strategies and help the owners and managers of FB to acquire international capabilities that allow them to achieve an optimum level of internationalization.
Finally, we should point out the limitations of our study, which open future lines of research. Our research focuses on a sample of small and medium-sized family businesses belonging to a single sector (manufacturing) and to a single country (Spain), which limits the generalization of the findings to other sectors and sizes of businesses (large family businesses and no family businesses), as well as to other countries. Therefore, it would be helpful to extend the study to sectors other than manufacturing to test whether the findings obtained can be extended to FB in other sectors. In addition, the research could be extended to a segment of large family businesses. The hypotheses proposed could also be tested for family businesses from other countries.