Organization type and international firm performance: The moderating effect of board membership and female owner presence
By:
Coen Donders S2013819
c.j.donders@student.rug.nl
MSc International Business and Management Faculty of Economics and Business
University of Groningen
Master Thesis 2014-2015; Semester 2
Supervisor: Dr. M.J. Klasing
Co-assessor: Dr. D.H.M. Akkermans
Organization type and international firm performance: The moderating effect of board membership and female owner presence
Abstract
This study examines the influence of firm type on the international firm performance. In addition, it investigates two moderating effects on this relationship. The first moderating variable investigates the effect of the firm owner presence on the board or Top Management Team (TMT). The second investigates the effect of female owner presence on the board or TMT. This study was conducted with a Beeps dataset among 3,345 firms from Germany, Portugal, Greece, South Korea and Vietnam. The results show that family-owned firms have better international firm performance than individual-owned firms. However, results also show a large influence of the control variables. Furthermore, in family-owned firms where the owner is active as manager or director have better international firm performance than when he or she is not active. Female owner presence on the board or TMT resulted in no significant influence.
Keywords: International firm performance, family-owned firms, individual-owned firms, manager, director, female owner.
Table of contents
1. Introduction 4
2. Literature review and hypotheses development 7
2.1 Firm performance 7
2.2 Ownership 8
2.3 Manager or director: 11
2.4 Presence of female owner on the board 13
3.Methods 16
3.1 Sample 16
3.2 Variables 16
3.2.1. Dependent variables 16
3.2.2. Independent variable 17
3.2.3. Moderating variables 17
3.2.4. Control variables 18
3.3 Data analyses 19
4. Results 21
4.1 Descriptive statistics and correlations 21
4.2 Regression results 23
4.2.1 Total export 23
4.2.2 Export growth 26
5. Discussion 30
6. Conclusion 35
6.1 Answering the research question 35
6.2 Theoretical and practical implications 36
6.3 Study limitations 37
6.4 Further research 38
References: 40
List of figures and tables
Figure 1. Conceptual model of family-owned firms p. 15 Figure 2. Conceptual model of Individual-owned firms p. 16
Table 1. Making Descriptive statistics p. 22
Table 2. Correlations p. 22
Table 3. Linear- and multiple regression analyses total export p. 26 Table 4. Linear and multiple regression analyses export growth p. 29
1. Introduction
With today’s rapid economic and technological developments, companies are increasingly competing on a global scale (Barkema & Shyrkov, 2006). After a small slump during and after the global financial crisis, global FDI returned to growth to US$1,45 trillion, with inflows rising 9 per cent in 2013. For the coming years, the UNCTAD
1projects that the global FDI could rise to a higher level, with a relatively larger increase in developing countries (United Nations, 2014). Hence, there is an increasing need for boards of directors and to Top Management Teams (TMTs) identify expansion opportunities beyond familiar domains into countries or regions that are new to the company, such as South East Asia, Central and Eastern Europe, or Latin America (Barkema & Shyrkov, 2006). Emerging and transitioning economies have enlarged the set of countries that may potentially constitute an international opportunity for the future. Furthermore, things like deregulatory trends and technological changes in the international marketplace have dramatically affected product mobility (Kaczmarek, 2009). Strategy researchers have focused on the role of board of directors and TMT diversity as a driver of strategic innovation (Barkema & Shvyrkov, 2006), and in this way on the firm’s international performance. In relation to the role of boards or TMTs and their diversity, there has been significant interest in examining the role of ownership in corporate governance (Connelly, Hoskisson, Tihanyi & Certo, 2010; Johnson, Schnatterly, Johnson & Chiu, 2010). Is there a possibility this has some influence on the firm’s international performance?
Besides the diversity of boards and TMTs, the type of ownership of a firm plays an important role in the characteristics of a firm. The different degrees of risk aversion and availability of the resources that international strategy requires are characteristics that have an influence on the corporate strategy and performance (Fernandez & Nieto, 2006). The type of firm ownership depends on the largest shareholder and the percentage level of ownership. Any enterprise owned and run by one family, or a small number of families, is defined as a family firm (Stern, 1986). Nevertheless, despite the dominance of the family-owned firms in the organizational landscape, many other types of ownership can be found (Patel & Cooper, 2014). Differences between family-owned firms and non-family-owned firms are manifested in terms of values, objectives and strategic behavior (Fernandez & Nieto, 2006). For example,
1 The United Nations Conference of Trade and Development
family-owned firms tend to focus on the goal of preserving socioemotional wealth, which can be described as the emotional values of the family members. Characteristics of this type of values are trans-generational transfer, loss averse executives with risk adverse behavior (Patel
& Cooper, 2014). Since families have usually invested most of their private wealth, families are a unique type of investors who have exceptional concerns over firm survival and strong incentives to monitor management closely. Besides a high level of knowledge of the firm and market specific technologies, the long-term presence in the firms is also one of the main strengths of family-owned firms (Andres, 2008). One gap in the literature is the fact that it fails to adequately address the question about whether families as largest shareholders are more internationally successful than other non-family controlling shareholders, or vice versa (Andres, 2008).
In contrast with large family investors, large non-family investors such as individual investors have a large enough stake that it pays for them to spend private resources to monitor management. Because of their large stake, these individual investors have the ability to provide a solution to the free-rider problem (Shleifer & Vishny, 1986). In addition, characteristics of individual owners include more supportive behavior for change and growth, a focus on risky strategic initiatives to leverage their firm-specific human capital and increasing their potential rewards in the current firm (Patel & Cooper, 2014). In line with that, they have higher incentives to actively decrease the agency costs, and it is much easier to coordinate their actions and put pressure on managers. Since the voting power of family owners is close to family members, putting pressure on managers is much harder for them. If managers repeatedly act against the wishes of the large investors, managers of an individual- owned firm are more likely to be soon displaced in comparison with family-owned firms.
Therefore, this type of ownership differs from family shareholders in that they do not only have the incentive to decrease agency costs, but also the power and the skills to do so (Shleifer & Vishny, 1997). In the end, greater agency problems ultimately result in worse firm performance (Masulis, Wang & Xie, 2012). Consequently, it also will have influence on the behavior and resources for dealing with the international performance of the firm.
Isakov and Weisskopf (2012) reported that classical agency conflicts between managers and
shareholders may be reduced for firms with a large shareholder (both family- and individual-
owned firms), because these have an incentive to control management. However, when the
owner is pushing for too much extraction of private benefits, this can potentially result in a
value destroying effect on the firm. These two opposite agency problems can have total different consequences in different types of firm ownership. In the literature are the conclusions mixed for so far. Several studies show that family-owned firms create more value and are more profitable compared to individual-owned firms (Anderson & Reeb, 2003). At the same time is concluded by other researchers that family-owned firms do not outperform other types of firms (Chen, Chen, Cheng & Shevlin, 2010). Despite these widespread conclusions about the level of profitability in general, there are also no clear results of international performance since the number of international performing firms is increasing.
In addition to the largest shareholder, and in relation to international firm performance, this study also examined variables that influence this relationship. Since family- and individual- owned firms show completely different characteristics in terms of leadership, network communication etc., the expectation is that moderating variables will influence outcomes of these two ownership types. One moderating variable was if an owner is also a manager or director of the firm. According to Aguilera (2005), are boards by definition the internal governing mechanism that shapes firm governance, given their direct access to the two other parties in the corporate governance triangle: managers and (largest) shareholders (owners).
Licht (2002) defines governance as the rules and structures for wielding power over other people’s interests, including the use and abuse of power. In this way, this can have influence on the international performance of firms. Erhardt, Werbel and Schrader (2003) performed empirical research and they have concluded that gender diversity on boards result in improved financial performance. In contrast, other articles show that homogenous boards, with for instance only males on the board, will a have better performance. For this reason, the presence of a female as largest shareholder on the board or TMT was used as a second moderating variable.
Based on these multiple unanswered questions and the growing amount of international
business, this study will report on the influence of ownership type (family- or individual-
owned firms) on the international firm performance. This relationship will be measured on the
basis of two moderating variables. The first moderating variable is regarding whether the
principal owner is also a manager or director. The second moderating variable is the question
of whether there is a female principal owner on the board. Data on firm level was collected
from five countries (Germany, Greece, Portugal, South Korea and Vietnam), and in this way
the international firm performance was examined.
This article is divided into the following sections. The first section reviews the different elements within international firm performance, namely ownership types, managers, directors and the presence of female owners on a board or TMT. This section will also include the hypotheses and the conceptual models. The second section, the methodology of this study is explained. The third section shows the results, followed by the discussion in the fourth section. In the final section, the conclusion, some limitations and suggestions for further research will be discussed.
2. Literature review and hypotheses development 2.1 Firm performance
Firm performance has historically been measured in many different ways. Samiee and Roth (1992) used the return on investment, return on assets and the sales growth for this measurement. In 1993, Szymanski, Bharadwaj and Varadarajan (1993) used market share. A more encompassing interpretation was used by Albeum and Tse (2001): they defined firm performance related to competitive advantage. Waheeduzzaman and Dube (2003) used the return on sales and sales growth. International research was performed by Lages, Jap and Griffith (2008), who measured the amount of export performance improvement. Schilke, Reimann and Thomas (2009) used all these different views for reporting standardization on international level. According to Cavusgil and Zou (1994), export performance can also be defined as the extent to which a firm’s objectives, both strategic and financial, are achieved via the execution of the firm’s export marketing strategy. In the literature review by Sousa, Martinez-López and Coelho (2008), as many as 40 different determinants of export performance were discovered. This research concerned export marketing strategy, firm and management characteristics, and internal factors.
The focus of this study is on the international performance of firms. Much of the empirical
literature on firm internationalization has already focused on this, but this study will also
focus on export performance as measured by ‘export intensity,’ i.e., the ratio of export sales to
total sales (Denicolai, Zucchella & Strange, 2014). Furthermore, this study will also measure
if the total export has experienced an increase, decrease or no change over the previous 36
months. In this way, this study will use the ratio of total export relative to sales and the
growth rate of the total export relative to sales over the previous 36 months.
2.2 Ownership
The role of the dominant owner, defined as the largest shareholder, is an understudied area in corporate governance (Filatotchev & Wright 2011). The largest shareholders of a firm are the dominant shareholder because they have increased prominence. Useem (1996) notes that the dominant shareholder of a firm is the recipient of special attention from firm management.
Others found that for access to preferred information about the company there are more possibilities for the largest shareholders (Wright & Lockett, 2003; Schnatterly, Shaw &
Jennings, 2008). Consequently, the largest shareholder has a unique position to monitor, discipline and influence the board of directors and TMT (Muller-Kahle, 2015). During this study, a dominant shareholder is defined as someone with the majority of the shares, meaning a minimum of 51% of the total shares. With this majority, the dominant shareholder is the owner of the firm.
Ownership has been identified as a key mechanism in corporate governance. Muller and Kahle (2015) performed research in line with this topic. According to their research, the agency theory suggests ownership provide increased incentive to monitor firm management, which should lead to better firm performance. However, there are many types of shareholder with varying inclination and ability to act as an effective firm monitor. It is argued that the owner of a firm has different incentives to monitor their investment in the firm. Furthermore, Muller and Kahle suggest that the type of ownership plays a large role in whether a manager has the inclination to engage in this specific type of firm monitoring.
Overall, the empirical findings regarding examining the impact of large shareholders are mixed. According to Muller and Kahle (2015), some research reports that large shareholders can have a positive effect on firm performance (McConnell & Servaes, 1990; Dalton, Daily, Certo & Roengpitya, 2003). However, another meta-analysis has found no consistent link between one specific ownership type and firm performance (Dalton et al., 2003). In the case of the largest shareholder of the firm, and as mentioned previously, this paper will research how different types of ownership influence the international firm performance. Two types of ownership will be used, namely family- and individual-owned firms.
These two types of firms have two different types of governance structure. Boubaker et al.
(2012) reported that until recently, most public companies were widely held. In this type of
firms, providing a remedy for the classical agency theory has been seen as one of the main
challenges. In 1932, Berle and Means reported as one of the first that separation between ownership and control can cause problems, and also decrease agency costs. In 1976, Jensen and Meckling did further research and reported the same conclusion. However, in recent studies is reported that a large number of listed firms do not have a widely-dispersed ownership structure. Based on these studies, it can be concluded that all these firms have one or a few large shareholders and can be categorized as individual, families, states and other industrial or financial firm. Claessens, Djankov and Lang (2000) observed that in Asian countries, families own approximately two-thirds of firms. Even in the US, Anderson and Reeb (2003) established that approximately 35 % of companies in the S&P 500 are family- controlled. This view of ownership structure calls for a careful analysis of the effects family ownership has on agency theory and international firm performance. Additional to the results of Berle and Means (1932) about incurring agency costs by separation of ownership and control, they also reported that family shareholders would have better incentives to control management. Since a large part of their wealth is invested directly in the firm, there is a higher inclination for good results influenced by a long-term view (Isakov & Weisskopf, 2014). It seems that family ownership might therefore be a good ownership type for more value creation. The potential benefits associated with the presence of a family as largest shareholder are not new. But the question then is, do these characteristics also result in better international performance? Or do family-owned firms perform better in local or national markets?
Isakov and Weisskopf (2014) wrote much of their research in line with this topic. They considered the identity of the controlling shareholder. Most of the time, this is the owner of the firm. They were looking especially to the distinction between firm types in order to determine which of the effect prevails. Differences could be explained by different types of shareholders having different incentives and motivations that will affect their way of perceiving the company. In addition, the reputation, professionalism and perception of the family within its social environment could be one of the elements that possibly results in a long-term strategy and behaves in a way that is not purely self-centered In contrast, for companies with other types of large shareholders such as individuals, different goals result in different incentives (Isakov & Weisskopf, 2014).
Among other studies, Anderson and Reeb (2003) found that family firms outperform their
individual counterparts in the S&P 500 because only family-owned firms have real incentives
to reduce agency costs. It seems that active management by the family, especially by the founder, is an important incentive for the firm to create value and to be more profitable (Isakov & Weisskopf, 2014).
Isakov and Weisskopf (2014) also outlined that ownership and control level are drivers for better market and accounting performance. After observing a nonlinear relationship between control and performance, the results were interpreted as a sign of increased family opportunism and extraction of private benefits if the family gains too much power. On the other hand, also was shown that the profitability increased when the control by the family was high. This confirmed the research with showing that family management improves firm efficiency.
In other research, performed by Fernandez and Nieto (2006), it is mentioned that a family held firm would be differently governed than a firm with an individual owner. This distinction should be considered when examining the performance of a firm internationally.
Characteristics such as the long-term view, culture as a source of pride, direct control, limited bureaucracy and informal management can especially form strengths for performance and weaknesses for internationalization. The lack of qualified personnel, financial resources and intangible resources are also common in family-owned firms. These factors have an impact on a firm’s resource endowment and its ability to sustain a competitive advantage which is in line with a long-term view. These characteristics of family firms limit their opportunities to acquire resources, particularly intangible knowledge-based assets such as technologies, well- known brands or qualified personnel. Internationalization requires the implementation of more complex structures and formal controls (Fernandez & Nieto, 2006).
According to Westhead et al. (2001), managerial skills specifically play an important role for
the internationalization. Efficient management structures, access to financial resources and
professionalism are common characteristics for individual-owned firms (Fernandez & Nieto,
2015). In addition, contacts provided by the individual owners are more widespread and this
widespread network can result in knowledge about foreign markets. An active participating
role in combination with the knowledge from their network could be an important resource
for opportunities inside the international market. These specific international characteristics
are not mentioned regarding the family-owned firms. Graves and Thomas (2006) mention that
family-owned firms experience obstacles to developing their managerial capabilities as they grow. Based on these characteristics, the first hypothesis was formulated:
Hypothesis 1: An individual-owned firm will have a better international firm performance than a family-owned firm.
2.3 Manager or director:
The first hypothesis was formulated in order to measure whether family-owned firms or individual-owned firms have better international firm performance. These results can be influenced by specific characteristics of the type of ownership. In addition to the type of ownership, this study will also examine if owners play a role as manager or director.
The reason for participation as manager and/or a director is due to the different structures between firms among different countries. Each country has its own structural differences as to how recruitment of board members is organized. Some countries have a clear two-tier system with two separate segments, an executive board and a supervisory board (Van Veen &
Elbertsen, 2008). In this way, decision management and decision control are formally separated (Maassen & Van den bosch, 1997). Second, some firms have a one-tier structure, which distinguishes between executives and non-executives. Decision management and control are less separated to a lesser degree. Lastly, some countries have a system that is between both these extremes. In these countries, most firms have chosen the two-tier board, although there seems to be a slow movement toward more one-tier boards (Van Veen, &
Elbertsen, 2008). For this reason, a manager as director can be seen as a person who is in a leading role in the firm. In all three types of structures, the moral hazard problems are the most debated topic when we speak about power and decision-making. Monitoring is necessary to induce managers to work hard; only concentrated share-ownership induces shareholders to monitor.
According to Huddart (1993), effective monitoring of management by shareholders is a public
good. This public good should be privately incurred. Because of the inability to locate costs
and benefits of the supervision of an individual, this is an underinvested area. When a
shareholder owns more than a threshold amount of stock, he has the incentive to monitor the
managers. Additional, Huddart wrote that this also means that the manager's compensation
scheme is in hands of the largest shareholder. These incentives could result in an endless
increase of the concentration of the shares for this person. For internationalization however, diversified and accurate knowledge is needed and has to be transferred inside the firm. To enable this, it may be important that the largest shareholder has a role as manager or director.
This is the person with knowledge and power to implement decisions he or she can make because of his/her position as firm owner. Not only decisions for themselves, but also for the results of the firm. In line with this, research by Sanders and Carpenter (1998) examined the relationship between internationalization and governance characteristics of a firm. One of the main conclusions was that internationalization leads to an increase in organizational complexity that then creates information asymmetries. This in turn leads to a misalignment of goals between management and owners. Based on this information, this study predicts that an owner as manager or director has a positive influence on the relationship between firm type and international firm performance. In this way it is possible to reduce agency problems, transaction costs and time.
However, the presence of the owner on the board or TMT can have a different effect in
family- and individual-owned firms. According to Singla, Veliyath and George (2014), family
controlled and family managed (FCFM) firms behave differently compared to family-
controlled and non-family-managed firms (FCNFM). This is due to dissimilar goals and
preferences with regard to the shared and/or the private benefits of control. One of their main
arguments was based on
different preferences for the private and the shared benefits of control among family owners. Two arguments were given for a weakened relationship between family presence on the board and international firm performance. The first argument was that since the family’s welfare is closely intertwined with that of a firm, they would like to retain greater control over a firm that is engaging in relatively risky strategic maneuvers (such as internationalization). The second argument was that in order to better manage the increased complexity and information asymmetries resulting from international operations, a greater proportion of outside directors on the board is recommended. However, the controlling family in FCFM firms may not be receptive to having more outside directors monitor management activities, as these outside directors would demand greater accountability and transparency from management. Based on these two reasons: hypothesis 2 was formulated:Hypothesis 2: A family owner as manager or director of a firm will result in weaker international firm performance in comparison with a family-owned firm with no presence of the family owner on the board.
Furthermore, the study by Westhead et al. (2001), confirmed that businesses with principal founders, with a dens information and contact network, with a lot of resources, and having management know-how are significantly more likely to be exporters. With the earlier- mentioned characteristics of individual owners seeking increased risk, a more international view and a more international network, this study expected a positive influence of the presence of the individual owner on the board in the relation to the firm type and international firm performance.
Hypothesis 3: An individual owner as manager or director of a firm will result in better international firm performance in comparison with an individual-owned firm with no presence of the individual owner on the board or TMT.
2.4 Presence of female owner on the board
Particular at the top level of firms, women’s talents are an underutilized topic (Terjesen, Aguilera & Lorenz, 2015). Simpson, Varter and D’Sausa (2010) found that the presence of women on boards is often seen as a good business decision and results in an increase of the value of the firm through their performance as director. Simpson et al. (2010) also mentioned that female directors are not substitutes for male directors of equal ability and qualifications, but that women instead may have unique attributes that increase the performance of the board and ultimately the performance of the firm. Currently, compared to the number of women in the general population, the percentage of women directors on the board or TMTs is relatively low. A female participation ratio of 15,2 % is indicated in the 2009 Catalyst Census of the board seats of Fortune 500 corporations held by women (Catalyst 2009).
In contrast, there are also articles with the conclusion that there is no significant difference in
firm performance with female presence on a board or TMT (Rose, 2007). Erhardt et al. (2003)
mentioned that while diversity within boards of directors could be created in order to
demonstrate an absence of discrimination, it is unclear if diversity within boards of directors
has an impact on organizational performance. Maznevski (1994) examined the literature on
group diversity and challenged the previous research findings that homogeneous decision-
making groups perform better than diverse ones. The keys to improved performance are integration and communication. Hambrick, Cho and Chen (1996) concluded in their research that homogeneous top-management teams actually outperform heterogeneous ones. The heterogeneous teams were slower in their actions and they respond less likely to competitors’
initiatives compared to homogeneous teams.
Women are increasing in number among corporations’ boards of directors, yet their representation is far from uniform across firms (Hillman & Canella jr., 2007). But is the role of women on the board the same in a family-held firm compared to a firm with individual large shareholders? A family is a very dense group, with gender diversity. Issues such as mother- and fatherhood can play a key role in the interaction and communication between family members. Connell and Schau (2012) mentioned that
motherhood plays a prominent role in women’s gender performance. The role of fatherhood has experienced changes over the previous several decades.An example of a change could be: ‘From a father as breadwinner and economic provider to nowadays both responsible financially for his offspring and as equal caregiver and nurturer within the family dynamics.’ According to Campbell and Minguez- Vera (2015), family-owned firms are more likely to have women appointed to the board. They found as a result that there is a positive relation to firm performance when there is a female on the board who has family ties to the owner(s). It is interesting that they do not have an explanation for this finding and mentioned it as future research. This relationship is also not the same as the moderating variable of this study, but it explain the unknown positive relationship that family ties could have. The dense family connections could be an explanation for the lack of communication between men and women. Because of the dense organization culture and less bureaucratic decision making levels in a family-owned firm, this type of firm is expected to have a positive influence on international firm performance because they can perform with gender diversity on the board or TMT. These arguments lead to the following hypothesis:
Hypothesis 4: The presence of a female owner as manager or director will allow a family- owned firm to have a better international performance than a firm without a female owner on the board or TMT.
Furthermore, individual-owned firms will have more direct communication with short-term
individual goals. Internationalization takes into account some important aspects of
personality, such as the need for achievement and risk-taking propensity. As mentioned previously, these are higher in individual-owned firms. In addition, the importance of concrete resources such as knowledge intensity and international entrepreneurial orientation are also expected to be higher in individual-owned firms. These characteristics are coupled to an increased degree in homogeneous teams. The complexity of working on the international level requires good communication skills. According to Adams and Ferreira (2004), diverse boards need more contact moments, which result in more board meetings for the same amount of information. This results leads to following hypothesis:
Hypothesis 5: The absence of a female owner as manager or director in an individual-owned firm means that the firm will have a better international performance compared to a firm with a female owner on the board or TMT.
The conceptual models shown in figures 1 and 2 give a clear summary of the concepts and hypotheses that will be examined.
Figure 1. Conceptual model of family-owned firms
Figure 2. Conceptual model of individual-owned firms
3.Methods 3.1 Sample
As mentioned previously, the objective of this study is to examine the relationship between family- and individual-owned firms and the international firm performance. This relationship will be measured on the basis of two moderating variables. The first moderating variable is regarding whether the principal owner is also a manager or director. The other moderating variable is the regarding whether there is a female principal owner on the board. The data comes from an existing data set from BEEPS (BEEPS, 2004). This existing data are based on a survey done by firms from five different countries, namely Germany, Greece, Portugal, South Korea and Vietnam. For this dataset, BEEPS has collected data based on 84 questions with sub-questions, completed by 3,345 enterprises spread over the earlier-mentioned five countries. The first 11 questions were conducted by telephone interview and the final 74 questions are done by face-to-face interview. For this study, several analyses will be done based on the data collected by BEEPS. The dataset from 2004 is especially useful because it includes useful data about shareholder characteristics and ownership types.
3.2 Variables
3.2.1. Dependent variables
As showed in figures 1 and 2, the dependent variable is international firm performance.
According to Zahra and George (2002), firms with good international performance are “firms
that internationalize their operations in innovative and creative ways (and) stand to achieve significant gains that go beyond superior financial performance.” In earlier research, international firm performance is measured by return on assets, sales growth, market share, return on investment and more. In this study, the international firm performance will be measured at two different times on basis of two dependent variables. First, the dependent variables will be measured in relation to the ratio of level of total export relative to total sales (code: Totalexport). For the level of exports relative to sales for this study, both the percentage of directly exported and the percentage of indirectly exported will be used. Level of exports relative to sales can be found in Q7 (b+c) of the BEEPS dataset.
Second, the dependent variable will be measured on the basis of the changes in export relative to the changes in total sales (code: Exportgrowth). These changes will be measured over the previous 36 months (Q55 from the BEEPS dataset). This means the creation of two separate dependent variables in this study.
3.2.2. Independent variable
The independent variable is the largest shareholder of the firm. In the survey, several types of shareholders are listed: individual, family, general public, domestic company, foreign company, bank, investment fund, managers of the firm, employees of the firm, government or government agency, and others. This study will focus on family- and individual-owned firms (Hypothesis 1). For this reason, the type of largest shareholder will be used when this person has ≥ 51 % of the total stocks. According to Schiehll and Bellavance (2009), firms can be categorized as CEO-owned if the current CEO is the largest shareholder based on the percentage of outstanding common stocks held by the CEO.
The independent variable is family-owned firms. Family-owned firms is a dummy variable coded ‘1’ if the principal owner is a family owner; otherwise it is coded ‘0’ when the principal owner is an individual owner. This dummy variable is based on Q4 from the BEEPS dataset (code: Family owned).
3.2.3. Moderating variables
The relationship between firm type and the international firm performance will be measured
by the influence of the moderating variables. The first moderating variable concerns whether
the principal owner is also the manager or director of the firm, based on Q4b from the BEEPS
dataset. Manager or director (code: Manager/Director) is a dummy variable coded ‘1’ if the
principal owner is a manager or director; otherwise it is coded ‘0’ because he or she is not a manager or director.
The second moderating variable regards those who are a female largest shareholder and if this female owner is on the board or TMT. This variable is based on Q4c from the BEEPS dataset.
Presence of a female owner on the board (code: Female-owner/director) is a dummy variable coded ‘1’ if there is a female owner on the board; otherwise it is coded ‘0’ as there is no female owner present on the board.
3.2.4. Control variables
During this study, several control variables are used. The first control variable is the age of the firm (Firmage). Mishra, Randøy and Jenssen (2001) determined that firm age is an important determinant for business characteristics and company goals. Firm age is defined as the total number of years since incorporation. All firms answered the question in which year the firm started. The data was collected in 2004. This means that 2004 minus the start date of the firm is the age of the firm (S1a of the BEEPS dataset).
Second, the size of the firm will be used. Size can be a key determinant; its relationship with export performance has been one of the most extensively studied in the export marketing literature (Moen, 1999), (Sousa et al., 2008). Size will be measured as the total number of employees (S4b of the BEEPS dataset). For firm size, this study will use three categories:
small firm (two – 49 employees), medium-sized firm (50 - 249), and large-size firm (250 - 9999). Therefore, three separate dummy variables are constructed. Small firms (code:
SMALL) are a dummy variable coded ‘1’ if the firm size is small; otherwise it is coded ‘0’
when the firm has a medium or a large size. For medium- and large-sized firms the dummy variable is constructed in the same way (codes: MEDIUM & LARGE). In case of firm size, the dummy variable MEDIUM will be the omitted group.
The main area of activity will be used as a control variable (S3 of the BEEPS dataset).
Distinction is made between firms operating in the industry sector and those operating in the
services sector. Main area of activity (code: Services) is a dummy variable coded ‘1’ if the
firm is operating in the services sector; it is coded ‘0’ when the firm is operating in the
industry sector. Segments inside the service sector are: wholesale and retail trade, real estate,
hotels and restaurants, and other service activities such as radio and television, news agency
activities, and washing and dry cleaning. Segments inside the industry sector are: mining and quarrying, construction, and manufacturing.
The last control variable is the country where the firm was founded. The firms were founded different countries. Therefore, five separate dummy variables are constructed (codes:
GERMANY, PORTUGAL, GREECE, SKOREA and VIETNAM). GERMANY is a dummy variable coded ‘1’ if the firm comes from Germany; otherwise it is coded ‘0’ if the firm comes from Portugal, Greece, South Korea or Vietnam. For firms from Portugal, Greece, South Korea and Vietnam, the dummy variable is constructed on the same way. In case of the country from where the firms come, the dummy variable VIETNAM will be the omitted group.
3.3 Data analyses
As mentioned previously, the dataset used in this study was collected in 2004 by BEEPS. The main statistical method used in this study in order to test the hypotheses was a multiple regression analysis. By using Stata 13, this study estimates the relationship between firm type as the independent variable and international performance as the dependent variable, and two moderating analyses on this relationship.
The analysis will be done twice because of the two different dependent variables for the international firm performance, namely Totalexport and Exportgrowth. Furthermore, some additional information for reading the regressions equations is stated below.
Dependent variable:
IFP = International Firm Performance (Totalexport and Exportgrowth)
Independent variable:
FoF = Family-owned firms (Family owned)
Moderating variables:
MD = Manager or director (Manager/Director)
Fod = Presence of female owner on board (Female-owner/director)
Control variables:
Fs = Firm size (SMALL, MEDIUM, LARGE) Fa = Firm age
Ma = Main activity (Service)
C = Country (GERMANY, PORTUGAL, GREECE, SKOREA and VIETNAM)
β = Coefficient of predictor variables ε = Error Term
First, it is important to know the direct relationship between the type of firm and the international firm performance. This is the relationship between Totalexport and Exportgrowth influenced by family-owned firms. The full model for the linear regression equation can be written as follows:
1. International Firm Performance on family-owned firms IFP = β0 + β1*FoF + ε
Subsequently, the analysis will be done with the control variables and the dependent variables. For this model, this will be the relationship between Totalexport and Exportgrowth as influenced by all control variables. The full model for the multiple regression equation can be written as follows:
2. International Firm Performance on control variables IFP = β0 + β1*Fs + β2*Fa + β3*Ma + β4*C + ε
In the following analysis the independent and control variables will be combined. This will be the relationship between Totalexport and Exportgrowth as influenced by family-owned firms and all control variables. The multiple regression equation can be written as follows:
3. International Firm Performance on family-owned firms + controlvariables IFP = β0 + β1*FoF + β2*Fs + β3*Fa + β4*Ma + β5*C + ε
Two moderating analyses will be done afterwards. First, is the relationship between Totalexport and Exportgrowth influenced by family-owned firms and all control variables with the presence of the large shareholder as manager or director on the board as moderator.
Second, a regression will be performed using the relationship between Totalexport and
Exportgrowth as influenced by family-owned firms, and all control variables with the
presence of female largest shareholder on the board as moderating variables. The full model
for the multiple regression equation can be written as follows:
4. International Firm Performance on family-owned firms + control variables + manager/director
IFP = β0 + β1*MD + β2*FoF + β3*(MD*FoF) + β4*Fs + β5*Fa + β6*Ma + β7*C + ε
5. International Firms Performance on family-owned firms + control variables + female-owner/director
IFP = β0 + β1*Fod + β2*FoF + β3*(Fod*FoF) + β4*Fs + β5*Fa + β6*Ma + β7*C + ε
4. Results
In the following sections the empirical results are shown. First, the descriptive statistics will be shown, followed by the correlations between the variables. Finally, the different regressions are discussed.
4.1 Descriptive statistics and correlations
Two tables are presented in this section, namely Table 1 and Table 2. The descriptive
statistics provide an overview of a number of observations, the frequency of dummy
variables, the mean, the standard deviation and the minimum and maximum scores. All the
variables have a sample size of 2,327. In Table 1, the statistics for all dependent, independent,
and control variables can be seen.
Table 1. Making Descriptive statistics
Variables Obs* Frequency Mean S.D.** Min Max
Totalexport 2,327 4.60 16.87 0 100
Exportgrowth 2,327 -0.31 24.06 -200 145
Family-owned 2,327 0=1,948 1=379 0.16 0.37 0 1
Manager/Director 2,327 0=345 1=1,982 0.85 0.36 0 1 Female owner
/director
2,327 0=1,792 1=535 0.23 0.42 0 1
SMALL 2,327 0=267 1=2060 0.88 0.32 0 1
MEDIUM 2,327 0=2,164 1=163 0.07 0.26 0 1
LARGE 2,327 0=2,223 1=104 0.04 0.21 0 1
Firmage 2,327 14.52 13.75 3 116
Services 2,327 0=906 1=1,421 0.61 0.49 0 1
GERMANY 2,327 0=1,529 1=820 0.35 0.48 0 1
PORTUGAL 2,327 0=2,024 1=325 0.14 0.35 0 1
GREECE 2,327 0=1,905 1=444 0.19 0.39 0 1
SKOREA 2,327 0=1,937 1=412 0.17 0.38 0 1
VIETNAM 2,327 0=2,001 1=248 0.15 0.36 0 1
* Obs = Number of observations
** S.D. = Standard Deviation
The scores are given in Table 1. Because of missing data from the interviewed firms (3,345), Table 1 shows that all variables have the same sample size of 2,327.
After the summary of the descriptive statistics in Table 1, the results of the correlation of all variables are shown in Table 2. It can be seen that Exportgrowth, Manager/Director and Female owner/director are negatively correlated with Totalexport. For Exportgrowth this will have no consequences, because this variable has no combined regression with Totalexport (bot-dependent variables). It can also be noted that Manager/Director also has a small negative correlation with Exportgrowth. Overall, these effects appear, in general, to be relatively low.
Table 2. Correlations
Variables 1 2 3 4 5
1. Totalexport 1
2. Exportgrowth -0.01 1
3. Family-owned 0.05 0.03 1
4. Manager/Director -0.06 -0.06 0.07 1 5. Female
owner/director
-0.03 0.01 0.11 0.23 1
4.2 Regression results
The statistical analysis is performed in the form of a multiple regression. The results are shown in Tables 3 and 4. The regression equations mentioned above are presented in the two tables. Besides the regressions coefficient, the R- and the F-test scores and the significance of the results are also shown in the tables. The R-squared represents the percentage of the dependent variable that can be explained by the results shown. Not all the results are significant. The different levels of significance used in the tables are: *** p < 0.01, ** p <
0.05, * p < 0.1. Both tables will be explained below. First, the results on the dependent variable total export will be analyzed. Afterwards, the results on the dependent variables export growth will be explained and analyzed.
4.2.1 Total export
4.2.1.1. The effect of ownership type on total export
The scores in row 1 of Table 3 show the results of the regression done between the dependent variable and the independent variable. The significant score of 2.19 (p < 0.05) is given and explains that family-owned firms have a higher total export to sales ratio than firms where the main owner is an individual. Based on this score, it can be concluded that hypothesis 1 can be rejected. Individual-owned firms were expected to have a higher total export to sales ratio.
But the opposite has been found; family-owned firms have a higher ratio, which means better international firm performance. The R-squared score of 0.25 % for this regression can be explained for a small part as the influence of ownership type on the ratio total export to sales.
There are several other elements that will have a much stronger influence on this ratio.
Nevertheless, a significant higher total export to sales ratio is found for family-owned firms.
However, it is important to note that hypothesis 1 is rejected without using control variables in the regression. The regression results from control variables can be found under ‘the effect of ownership type and control variables on total export.’
4.2.1.2. The effect of control variables on total export
In the second row the results of the regression between the dependent variable Totalexport
and the control variables are shown. In examining the firm size, some conclusions can be
made. With a high significant negative score for small firms (-8.41; p < 0.01), and a high
significant positive score for large firms (9.42; p < 0.01) it can be concluded that large firms
have a higher ratio of total export to sales. In other words, when a firm has ≥ 250 employees,
it would have a higher ratio of export to sales than when it has ≤ 249 employees. Furthermore,
the firm age also has a significant positive score (0.05; p < 0.1). This means that older firms have a higher total export to sales ratio. For example, a firm with an age of 116 years would have a higher ratio than a three-year-old firm. However, this score is quite low, which would not result in very large differences between old and young firms. Meanwhile, the main area of activity the firm is operating in has a significant influence on the total export to sales ratio.
Firms working in the industry sector have a significant higher total export to sales ratio than firms working in the services sector (-4.09; p < 0.01). For the country from where the firm comes, there is no significant relationship between a selected country and the ratio total export to sales. This reveals that the separate countries do not show an influence on the ratio of total export to sales. With a significant positive score (5.02; p < 0.01), the dummy variable Vietnam is an exception, which means firms in Vietnam have a higher total export to sales ratio than firms not from Vietnam. With the R-squared score of 11.61 %, these results could be explained to an acceptable degree by the influence on the total export to sales ratio.
However, there are still many other elements with much higher influences on this ratio.
4.2.1.3. The effect of ownership type and control variables on total export
In the third row, the ownership type and the control variables are combined for a regression with the total export to sales ratio. In this row, almost the same significant scores are shown as in the first and second rows. However, when combining the independent variable and the control variables, no significant score remains between ownership type and the ratio total export to sales (0.63; p > 0.1). This reveals that ownership type (both family- and individual- owned firms) does not have an influence on the ratio total export to sales. Based on this score, there is no support for hypothesis 1. In addition, firm size still shows that large firms would have a higher ratio total export to sales (small: -8.33, p < 0.01; and large: 9.43, p < 0.01).
Furthermore, a slightly higher ratio of total export to sales can be found for older firms (0.05, p < 0.1). Firms working in the industry sector also show a higher ratio total export to sales than firms working in the services sector (services: -4.08, p < 0.01). Further, for the countries from which the firms come it is also almost the same result, which means that only firms from Vietnam have a significant higher ratio export to sales than firms not from Vietnam (5.15, p <
0.01). Lastly, with the R-squared score of 11.30 % these results could explain an acceptable
part of the ratio total export to sales, but there are still many other elements with a much
higher influence on this ratio.
4.2.1.4. The effect of a manager or director on the relationship between ownership type and total export, inclusive of control variables
In the fourth row, the results are shown regarding the question of whether a family owner is also a manager or director of the firm and the influence that this has on the total export to sales ratio. In this regression, there is no significant score between ownership type and the ratio total export to sales (-4.03; p > 0.1). This reveals that ownership type (both family- and individual-owned firms) does not have an influence on the ratio total export to sales. In addition, the variable manager/director shows a significant negative score as regression on the ratio of total export to sale (-2.3; p < 0.05). This reveals that when an owner (male or female) of a firm is also active as manager or director, the ratio of total export to sales will be lower than when there is no presence of the owner on the board or TMT. Both family- and individual-owned firms are combined for this regression score and have a lower ratio of total export to sales than when there is an absence of the owner. The moderating effect of family owners and their presence on the board or on TMT of a firm shows an exception to this result, with a significant positive relationship (5.19; p < 0.1). This score reveals that only when in a family-owned firm the (or one of the) owner/s is also active as manager or director is there a higher total export to sales ratio than when the owner is not active as manager/director.
Furthermore, this score also reveals that only when in an individual-owned firm the (or one of the) owner/s is also not active as manager or director is there a higher total export to sales ratio than when the owner is active as manager/director. The control variables in this case show almost the same scores as the previous rows in this table. Again, the score of firm size shows a slightly higher ratio of export to sales for large firms in comparison with medium and small firms (< 250 employees), followed by a slightly higher ratio of total export to sales which is found for old firms (0.05; p < 0.1). The older the firm, the higher the results of the ratio of total export to sales would be. For the main activity area and country, the results are almost the same as previously. The R-squared score is also almost the same, namely 11.88 %, which means that this percentage explains the influence on the export to sales ratio.
4.2.1.5. The effect of the presence of a female owner on the board on the relationship between ownership type and total export inclusive control variables
In the fifth row, the results are shown regarding the question of whether a female family
owner is also a manager or director of the firm and if this has an influence on the total export
to sales ratio. In this regression, there is no significant score between ownership type and the
ratio total export to sales (0.28; p > 0.1). This reveals that ownership type (both family- and individual-owned firms) does not have an influence on the ratio total export to sales.
Furthermore, the variable female owner/director does not show a significant score (0.69; p >
0.1). This reveals that the presence of a female owner on the board or TMT does not influence the ratio of total export to sales. The insignificant result of the interaction term family*female (0.79; p > 0.1) has almost the same score. This reveals that the presence of a female owner of a family firm does not have an influence on the relationship between firm type and the ratio total export to sales. In line with the other rows, the control variables show almost the same values, which would result in the same consequences and conclusions. The R-squared score is 11.69 %, which indicates the influence on the export to sales ratio.
Table 3. Linear and multiple regression analyses of total export No
controls With controls Full set
Full set + interaction
effect
Full set + interaction
effect
Variables (1) (2) (3) (4) (5)
Constant 4.24 12.52 12.82 13.96 12.04
SMALL -8.41*** -8.33*** -8.15*** -8.40***
LARGE 9.42*** 9.43*** 9.57*** 9.40***
Firmage 0.05 * 0.05* 0.05** 0.05*
Services -4.09*** -4.08*** -4.04*** -4.20*
GERMANY -0.19 -0.09 -0.08 0.16
GREECE -0.09 0.02 -0.05 0.23
SKOREA 1.27 1.45 0.97 1.71
VIETNAM 5.02*** 5.15*** 5.17*** 5.34***
Family-owned 2.19** 0.63 -4.03 0.28
Manager/Director -2.20**
Family*Director 5.19*
Female owner/director 0.69
Family*Female 0.79
Additional statistics
R-squared 0.0025 0.1161 0.1130 0.1188 0.1169
F value 5.9 42.19 33.92 28.38 27.86
No. of observations 2,327 2,327 2,327 2,327 2,327
Notes: *** p < 0.01, ** p < 0.05, * p < 0.1.
4.2.2 Export growth
4.2.2.1. The effect of ownership type on export growth
The scores in row 1 of Table 4 show the result of the regression done between the dependent
variable and the independent variable. The score of 1.64 (p > 0.1) for family-owned firms in
relation to export growth is not significant, which means that firm type (both family- and
individual-owned firms) has no influence on the export growth ratio. With an R-squared score
of 0.06 %, this regression does not seem to have a large influence on the export growth relative to sales ratio over the previous 36 months as a dependent variable in this model.
4.2.1.2. The effect of control variables on export growth
In the second row results are shown of the regression between the dependent variable (export sales) and the control variables. First, using the size of the firms as a control variable, it is shown that there is no significant score for large firms (1.64, p > 0.1). Based on this score, it may be said that there is no relationship between large firms in comparison with small- and medium-sized firms and the ratio of export growth relative to sales during the past 36 months.
On the other hand, small-sized firms show a significant score of 5.37 (p < 0.05). This reveals that small firms had a significantly higher export growth over the past 36 months relative to sales ratio than medium- and large-sized firms. In other words, when a firm has ≤ 49 employees, it would have a higher ratio of export to sales than when a firm has ≥ 50 employees. Furthermore, as in Table 3, Table 4 shows a significant positive score for the firm age (0.13; p < 0.01). This means that older firms have a higher export growth to sales ratio than young firms. Again, an example could be given of a firm with an age of 116 years, which would have a higher ratio than a three-year-old firm. However, the coefficient is quite low, which would not result in very large differences between old and young firms. In addition, the control variable regarding the main activity of the firm does not show a significant score. This means that there is no influence on whether firms are working in the industry sector or in the services sector on the ratio of export growth relative to sales. The final control variable is the country from where the firms come. All the dummy variables for the home country of the firms have a significant score. Germany and Greece show a negative score of -4.37 (p < 0.01) and -7.38 (p < 0.01), respectively. And, more extreme, Vietnam shows a negative score of - 23.83 (p < 0.01). These negative scores reveal that firms from these countries do have a lower ratio of export growth over the past 36 months relative to sales than firms from the other countries that are formulated in opposition to the dummy variables. In contrast, South Korea shows a significant positive score (9.81; p < 0.01). This reveals that firms from South Korea show a higher ratio of export growth relative to sales than firms from the other countries.
With the R-squared score of 18.13 %, this percentage may explain an appreciable part of the export growth relative to sales ratio.
4.2.1.3. The effect of ownership type and control variables on export growth
In the third row, the ownership type and control variables are combined to perform a
regression with export growth to sales ratio. In this row almost the same significant scores as
in the first and the second rows are shown, and the same variables show insignificant results.
This again indicates an insignificant score of 0.71 (p > 0.1) for family-owned firms in relation to export growth, which means that firm type (both family- and individual-owned firms) has no influence on the export growth to sales ratio. For the control variables regarding firm size, large firms show an insignificant score (1.64; p > 0.1) and only small firms show a significant score (5.46; p < 0.01). This will result in the same consequences as mentioned in the previous section. The same results are seen for the firm age (0.13; p < 0.01), main area of activity (- 0.30; p > 0.1) and the country from which the firm comes (Germany: -4.62; Greece: -7.25;
South Korea: 10.01; Vietnam -23.66; all: p < 0.01). With the R-squared score of 18.13 %, this percentage could predict a large part of the influences of the export growth to sales ratio.
4.2.1.4. The effect of a manager or director in the relationship between ownership type and export growth inclusive control variables
In the fourth row, the results are shown regarding the question of whether a family owner is also a manager or director of the firm, and if this has an influence on the export growth to sales ratio. In this regression, there is no significant score between ownership type and the ratio of export growth relative to sales (-2.58; p > 0.1). This reveals that ownership type (both family- and individual-owned firms) does not have an influence on the ratio. In addition, the variables manager/director and family*director both also show insignificant scores (-0.48 and 0.36, respectively; both p > 0.1). This reveals that when an owner (male or female) of a firm is also active as manager or director, this will have no influence on the ratio of export growth relative to sales. Furthermore, the moderating effect of family owners and presence on the board or on TMT of a firm will also have no influence on the relationship between firm type and the ratio of export growth relative to sales. With no change is the variables, but with the same significant and insignificant control variables, this model has an R-squared score of 18.17 %. This percentage indicates that this model may explain the influence on the ratio of export growth to sales.
4.2.1.5. The effect of the presence of a female owner on the board on the relationship between ownership type and export growth inclusive control variables
Row five shows the results for the question of whether a female family owner is also a
manager or director of the firm, and if this has an influence on the export growth relative to
sales ratio. In this row there is no significant score between ownership type and the ratio of
export growth relative to sales (1.48; p > 0.1). This will result in the same consequences as
those mentioned in the previous paragraphs. The control variables also again show almost the same results. Almost all control variables score good significant scores, with the most remarkable variable again being Vietnam (-23.61; p < 0.01). Further, in contrast to the other rows, the female owner as a director score in this row reached significance. This means that when the female owner of a firm is also working on the board or TMT, that this has positive influence on the ratio (2.24; p < 0.01). It other words, it could be said that a firm with the presence of a female owner on the board or TMT does have a higher ratio on export growth relative to sales than when there is no presence. The moderating variable family*female shows an insignificant score (-3.15; p > 0.1). This reveals that the presence of the female owner of a family firm does not influence the relationship between firm type and the ratio of export growth to sales. In line with the other rows in Table 4, the control variables show almost the same values, which results in the consequences and conclusions mentioned in the previous paragraphs. The R-squared score is 18.26 %, which means that this percentage explains the influence on the export to sales ratio.
Table 4. Linear and multiple regression analyses of export growth No
controls With
controls Full set
Full set + interaction
effect
Full set + interaction
effect
Variables (1) (2) (3) (4) (5)
Constant -0.58 -1.98 -2.25 -1.82 -2.62
SMALL 5.37** 5.46*** 5.43*** 5.32***
LARGE 1.64 1.64 1.70 1.71
Firmage 0.13*** 0.13*** 0.13*** 0.13***
Services -0.31 -0.30 -0.27 -0.52
GERMANY -4.37*** -4.62*** -4.65*** -4.46***
GREECE -7.38*** -7.25*** -7.29*** -7.13***
SKOREA 9.81*** 10.01*** 9.91*** 10.31***
VIETNAM -23.83*** -23.66*** -23.69*** -23.61***
Family-owned 1.64 0.71 -2.58 1.48
Manager/Director -0.48
Family*Director 3.61
Female owner/director 2.24*
Family*Female -3.15
Additional statistics
R-squared 0.0006 0.1813 0.1814 0.1817 0.1826
F value 1.48 64.16 57.05 46.72 47.02
No. of observations 2,327 2,327 2,327 2,327 2,327
Note: *** p < 0.01, ** p < 0.05, * p < 0.1.
