Family Ownership and Control; Institutional Quality; Firm Value

Department of International Economics &Business Master thesis

The Effect of Family Ownership & Control on Firm Value:

An Institution – Based Approach

Magdalina A. Shopova Student number: S2534908 M.A.Shopova@student.rug.nl

Thesis Supervisor Co – Assessor

Prof. Dr. Hans van Ees Prof. Steven Brakman

2014

2 | P a g e

Abstract

This study explores the relation between family ownership and control and firm value. Looking across 27 European Union countries, the research gathers sufficient data for the analysis of 270 firms which are family-owned and controlled firms. The impact of family ownership and control on firm value is associated not only with the level of shareholder protection, but is drawn on multiple measure of institutional quality. Findings indicate that in countries with higher levels of institutional quality, having a family CEO or family ownership through a subsequent generation works as an effective internal control mechanism while in countries where institutional quality is weaker, family CEOs or family ownership through subsequent family generation may result in higher level of expropriation of minority shareholders. In general, the paper highlights the importance of institutional quality for considering the relationships between firm ownership, controls structures and firm value.

Keywords:

Family Ownership and Control; Institutional Quality; Firm Value

3 | P a g e

Table of Contents

Abstract ... 2

Keywords:... 2

Table of Contents ... 3

Introduction ... 4

Theory Development and Hypotheses ... 8

Family Ownership and Firm Value ... 8

Family Ownership and Institutions ... 10

Family Control and Firm Value ... 13

First, Second-, and Third-Generation Family Firms ... 14

Family CEO ... 15

Methodology ... 20

Data: Sample and Variables ... 20

The Dependent Variable ... 21

Independent Variables ... 22

Control Variables ... 28

Analysis ... 31

Findings ... 33

Compound Institutional Effects ... 33

Individual Institution Effect ... 34

Robustness Check ... 36

Discussion... 36

Limitations and Recommendations for Future Research ... 39

Conclusion ... 41

Appendix ... 42

References ... 45

4 | P a g e

Introduction

A family owned firm is a firm in which ownership and control is referred to one family (through one or several members) serving as a controlling shareholder of a corporation (Peng & Jiang (2010). Family owned firms and their operations are an increasingly researched topic not in the least since most firms in the world are controlled by their founders or their founders` descendants (Burkart et al., 2003). For instance, the vast majority of publicly traded firms in Europe, South and East Asia, the Middle East, Latin America, and Africa are family controlled (La Porta et al., 1999). Despite the extent to which this topic is studied, there are still gaps in the literature regarding whether family ownership and control is beneficial or detrimental to firm value?

Amongst (international) business scholars, how and to what extent family ownership and/or control affects firm value is somewhat controversial. In the literature, three different positions are suggested. Fogel (2006) shows that greater family control over large corporations is associated with worse social economic outcomes, in other words such concentrated ownership is

‘bad’

– implying that a negative relationship exists. However, Anderson and Reeb (2003) suggests that family firms perform better than non-family firms indicating that family ownership is ‘good’

, thus it is an effective organizational structure with reduced agency costs. In addition, Minichilli et al. (2010) indicated that the role of family members in the Top Management Team is positively related to the performance of the family firm and contributes to its financial outcomes. And finally, some scholars argue that family ownership and control are irrelevant

for firm value, in other words family firms do not perform better than other firms (Miller et al, 2007).

As is clear, these arguments are focused largely on agency and resource based theories with little room for institutional based approach. Within their (2010) paper, Peng and Jiang suggest an institution based view on family ownership and control and its effect on firm value. According to the institutional based view in countries with less developed institutions family owned firms and control may afford expropriation to minority shareholders. This paper addresses that the relationship between family owned and/or controlled firm and firm value may be affected by the

1 See Fama & Jensen (1983); Schultze et al. (2001); Morck et al. (2005)

2 See Demsetz & Lehn (1985); Arregle et al. (2007); Minichilli et al. (2010)

3 See Daily & Dalton (1992); Willard et al. (1992)

5 | P a g e

level of institutional quality prevailing within the given context. The focus of the examination is on a region with vast concentration of family owned and controlled firms in large firms within the European Union and examines family firm value before and directly after the peak of the global financial crisis experienced after 2008.

In order to further explore the institution based approach this paper aims to provide an updated and extended investigation of how family ownership and control affect firm value by exploring the crucial role of institutions and how does this may vary across types of corporate governance.

Using Peng and Jiang`s (2010) study as the most recent reviewed research statement, this study steps forward in five significant ways. First, it extends the institution – based view of family corporate governance studies by exploring a cross-country analysis. Second, it extends the study into Europe by examining the institutional quality of various European Union (EU) member countries. This is considered to be important owing to the fact that there are no known studies to date which assess the effect of institutional quality on firm ownership and control and firm value across Europe. Europe is home to several different languages, cultures and varying judicial and sovereign systems. This implies that while the EU may be seen as one single region, each independent state varies considerably in terms of the prevailing institutional quality. Because of this, policy and corporate governance standards which are applied in one area of the EU may not be beneficial for all countries as a whole. To this extent, understanding how the quality of institutions affects family performance is detrimental to improving corporate governance levels in the various regions. Third - following Peng and Jiang (2010) this paper uses institutional quality data compiled by La Porta et al (1998). Their index is widely used and validated in recent cross country studies on shareholder protection and governance (Dyck and Zingales, 2004;

Johnson et al., 2000; Schneper and Guillen, 2004). However, in assessing the data available from

La Porta, data for only 16 EU member countries is available. Additionally, the data offered by La

Porta et al (1998) only contains a perception index for Rule of Law, Efficiency of the Judicial

System and Levels of Corruption. To this extent a full understanding of the institutional contexts

and their effects on EU family firm performance cannot be effectively measured. This means that

we need to explore alternative measures of institutional quality for the region. In light of this, the

study makes use of the World Bank World Governance Indicators (World Bank WGI) compiled

by Kauffman et al (2010). This index is considered to be a better index to use owing to the fact

that it has data for 27 EU member countries and provides 6 varying indicators of governance of

6 | P a g e

which 4 are the focus of this article, namely, Regulatory Quality, Rule of Law, Control of Corruption and Government Effectiveness. Fourth – Peng and Jiang (2010) consider the effect of institutional quality in the case of 8 Asian countries at the time of the Asian Financial crisis (i.e.

1997). The scholars argue that their investigation is important because during this time maintaining a large ownership forces blockholders to be not well diversified, in which case ownership is likely to cause large costs to the larger shareholders. In terms of agency theory, this implies that the larger shareholders will incur the costs of intervening in a company’s affairs alone (Maug, 1998). Shleifer and Vishney (1986, p.462)

shareholders may pay for the improvements of firms themselves since they are the ‘largest consumers of the public good’.

Hence, owing to the concentrated ownership and the large level of costs which they will incur, larger owners have great incentives to increase firm value. From a resource-based view, the benefit of having a large shareholder is that it may be crucial that he/she provides key resources and adopts appropriate strategies to increase firm value. During a financial crisis when external resources are limited, the largest shareholder and his/her involvement in the firm may become a rare resource, which cannot be imitated by other firms (Peng and Jiang, 2010). This implies that measuring these effects at the time of financial crisis is somewhat appropriate and gives a conclusive result. Owing to the fact that this paper considers Europe, the Global Financial Crisis offers an opportunity to assess this relationship more clearly. The family firm value is examined in time of crisis by exploring the relationship before and directly after the financial crisis peak in the EU, 2009, which is directly in line with the methodology of Peng and Jiang (2010) who consider the relationship before and directly after 1997.

The fifth and final differentiating factor is that while Peng and Jiang (2010) make a compelling

argument for looking at family control by observing appointing a family CEO and a Pyramid

Structure, with the exception of Sweden and Italy pyramid structures are rather uncommon

throughout Europe and as such is not considered to be important for the study of the 27 EU

member countries. The paper, however, does acknowledge the importance of family generation

ownership of family firms where the relationship between family ownership and firm value is

expected to vary across different levels of generations. This is further investigated in this paper

where different levels of generations of ownership are considered across the varying institutional

contexts prevailing in Europe.

7 | P a g e

The structure of this paper is organized as follows. The next section presents a literature

overview, exploring Agency Theory and Resource Based Theory and considers hypotheses

development. In section 3 the model and empirical analysis is presented. Section 4 presents the

results and it is followed by a discussion. Finally, Section 5 contains the conclusion reached.

8 | P a g e

Theory Development and Hypotheses Family Ownership and Firm Value

In essence family businesses are different from other businesses and considered as unique, due to complexity of balancing family`s personal goals, the well-being of family participants and satisfying multiple commercial stakeholders objectives – profit and sales growth (Steier et al., 2004)

. In addition, characteristics such as trust and altruism

, which are inherent for family firms, encourage long-term commitment and consequently lead to higher long-term performance (James, 1999).

To explain divergent motivations in organizations, the wishes of shareholders/principals and agents to maximize their own personal wealth and utility, agency theory is employed (Moores &

Craig, 2008). Expressed differently, agency theory tries to account for the inability of owners to control their agents effectively (Fox & Hamilton, 1994). To achieve a personal gain, agents will exploit their access to superior information. La Pora et al. (1997, 1998) argued that there is no a single agency model that adequately depict corporate governance in all national contexts. For instance Peng (2003) showed that the predominant model of corporate governance is a product of developed economies (United Kingdom), where the institutional context lends itself to relatively efficient enforcement of arm’s-length agency contracts. In these economies, the ownership and control are often separated and legal mechanism protects owners’ interests (Young et al., 2008), and the conflict that occurs is principal-agent conflict, which Berle and Means (1932) describe as the classic owner-manager conflict, which arises due to large shareholder`s greater incentive to monitor the manager (which is referred to as Agency Problem I). However, in emerging economies, the institutional context makes the implementation of agency contracts more costly and problematic (North, 1990; Wright et al., 2005), due to prevalence of concentrated firm ownership (Dharwadkar et al., 2000). The absence of effective external governance mechanisms, mixed with concentrated ownership, leads to more frequent conflict between controlling shareholders and minority shareholders (Morck et al., 2005). As a result a new perspective on corporate governance is developed, and the conflict is focused on different sets of principals in the firm (Young et al., 2008). Thus, the second type of conflict refer (Agency Problem II) to the

4 See Steier, Chrisman & Chua (2004); Randoy & Goel (2003); Speckbacher & Wentges (2012).

5 Altruism – actions that supports someone else`s welfare, even at expense to ourselves.

9 | P a g e

large shareholder`s willingness to use their controlling positions in the firm, in order to achieve private benefits at the expense of the small shareholders. But when the firm`s main owner is the founding family, the motivation for monitoring managers is greater than when the ownership is dispersed (Speckbacher & Wentges, 2012). Some scholars

have considered that family involvement reduces agency costs significantly and argued that one of the most successful forms of organizational governance is the one of family firms. Others

suggest that family owners may use their controlling position to obtain private gains at the expense of minority shareholders and argued that the entrenchment effects of family ownership lead to inferior performance by family firms (Agency Problem II; Anderson and Reeb, 2003). Thus the greater incentives for monitoring and expropriation when a family is the large shareholder will be likely to lead Agency Problem II to overshadow Agency Problem I (Villalonga and Amit, 2006).

As is previously mentioned during a financial crisis maintaining a large ownership makes blockholders to be not well diversified, in which case ownership will cause large costs to the larger shareholders. In terms of Agency theory the larger shareholder will suffer the costs of investigating in a company’s affairs alone (Maug, 1998) and may pay for the improvements of firms themselves (Shleifer and Vishney, 1986). Hence, due to the concentrated ownership and the high level of costs, larger owners have greater incentives to increase firm value.

Furthermore, Resource-Based Theory focuses on having majority shareholders involved in decision making and affects availability of key resources, which are limited during a financial crisis and cannot be imitated by other firms. The literature

on family firms suggests that the founding family`s relationships and social networks can serve as a means of transferring tacit knowledge and non-codifiable information to employees and other relevant stakeholders. The superior firm performance from a resource based perspective, the competitive advantage derived from a unique resource and a possible source of lower agency costs from an agency theory perspective, lead us to the premise that family ownership and control can be positive, which indicate that family ownership is beneficial (Speckbacher & Wentges, 2012) and especially in a crisis situation. In a conclusion, the findings according to the Agency theory and Resource based view demonstrate that the large family shareholder ownership can be beneficial to firm value, which shows that family ownership is ‘good’.

6 See Becker (1974); Dalton & Daily (1992); Daily & Dollinger (1992).

7 See Gomez-Mejia et al. (2001); Villalonga & Amit (2006); Anderson et al. (2003); Schulze et al. (2001).

8 See Cromie et al. (1995); Carney (2005); Chrisman et al. (2005); Sirmon & Hitt (2003).

10 | P a g e

Following the benefits of having family ownership and control according to the Agency theory and Resource based view a further examination with other mechanisms affecting family control is needed. Therefore this research examines the effect of institutions on firm performance for family-owned and controlled firms.

Family Ownership and Institutions

Most firms in the world are controlled by their founders or founders` descendants and the separation of ownership and control is not so common around the world (Burkart et al., 2003;

Fama and Jensen, 1983). However, it is unclear as to why the levels and implications of family ownership or control vary across different countries. The leading explanation is institutions.

Institutions are the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction in order to reduce uncertainty in pursuit their goals in social, political, and economic exchange (North, 1990). They also are viewed as regulative, normative and cognitive structures and activities that provide stability and meaning to social behavior (Scott, 1995). Consequently, institutions can be divided into formal and informal ones.

In this vein the institution based view may explain that the relation between family ownership and firm values may vary with institutional context and may define under what contingencies family ownership and control are ‘good’ and ‘bad’ for firm value within European countries.

According to Shleifer and Vishni (1997) the quality

of legal protection of shareholders helps determine ownership concentration. Hence a high ownership concentration can be regarded as a result of poor investor protection in a corporate governance system and visa verse – a good shareholder protection measures are associated with lower concentration of ownership (La Porta et al., 1998). This implies that considering the institution based approach the effect of firm ownership on firm value is likely to vary across different context specific conditions owing to changes across the various institutions prevailing in different countries. Therefore this research aims to explore the relationship between family owned and controlled firms and firm value within European Union during (and directly after) the peak of the crisis period. Thus more control may give family largest shareholders more opportunities to protect their firm value by expropriating minority shareholders (Peng & Jiang, 2010).

9 The extent to which the legal protection is good or poor varies across the countries.

11 | P a g e

By using a sample of large family owned US firms, Anderson and Reeb (2003) argued that family ownership is beneficial to firm performance. However, the authors noting that their findings may only hold in ‘well-regulated and transparent markets’. Considering the fact that there are institutional differences between the United States and Europe, it is interesting to engage in a deeper exploration within European Union.

In line with the aforementioned Peng and Jiang (2010) examine the region with extensive concentration of family ownership and control in large firms and explore the family firm value during the 1997 Asian financial crisis. By using a sample of 634 large family firms in seven Asian countries and ordinary least square regression, the scholars find that the interactions of family control and country`s institutional development have significant impact on firm value, namely in countries with more developed legal and regulatory institutions, the negative effect of family ownership on firm value is weaker. However, the scholars are also noting that their study may not be generalizable to the institutional contexts in other countries and the institutions they have used are not the only and there exists more comprehensive list of institutional quality (Kauffman et al, 2010). Peng and Jiang (2010) theoretically argued that the net balance of the benefits and costs of family control in large firms is systematically linked with the legal and regulatory institutions governing investor protection, but these results may not hold true in Europe. Given the fact that there is a growing consensus that institutions matter and there are relatively few studies about how institutions affect family ownership and control of firms and their value, it is interesting to explore this issue deeper within different sets of countries, and include various institutional measurements according to Kauffman et al. (2010) as discussed next.

Gedajlovic et al. (2004) described the complexity of environment as an issue for family owned and controlled firms, stating that environments with low levels of munificence (resource abundance) and complexity are better situated for family owned firms as opposed to environments with high levels of the aforementioned attributes. They suggest that the presence of better formal legal protection of investor rights encourages founding families to attract minority shareholders and delegate day-to-day professional management on the one hand.

On the other hand in absence of formal governance procedures and legal protection rights, giving

management rights to non-family managers could lead to abuse and/or theft, put differently

12 | P a g e

rampant agency problems. Moreover, prospective minority shareholders will be less willing to invest without sufficient protection, therefore forcing concentrated ownership to become the default mode (Peng and Jiang, 2010). Hence, when the legal protection is weaker, the founder designates his heir to manage and/or take ownership of the firm, implying that firm ownership and/or control remains inside the family (Burkart et al., 2003). This leads to the premise that firms and their ownership and/or control are embedded in institutional context and firm value depends on the efficiency of a bundle of governance mechanisms (Walsh & Seward, 1990; Peng et al., 2008, 2009). Rediker and Seth (1995) examine the substitution effect of that external governance mechanism, where internal is lacking, in other words firms require fewer internal control mechanisms where strong shareholder protection exists (La Porta et al., 2002).

Cumulative effort by scholars

reviews that there is evidence showing more concentrated ownership rights when the formal legal and regulatory institutions protecting shareholders are weaker.

Internal corporate governance mechanism or family ownership is supported by both agency and resource – based theory in environments with less developed legal and regulatory institutions i.e.

family ownership may have positive impact on firm value in countries with lower institutional quality (Heugens et al., 2009). But family ownership as an internal corporate governance mechanism is expected to be less important in countries with more developed quality, i.e. family ownership may be ‘irrelevant’ for firm value. Hence, in countries where the legal and regulatory institutions are less developed, larger shareholder plays an important role in maintaining firm value, and family ownership may be beneficial/good.

Thus, this paper hypothesizes that:

Hypothesis 1: The positive relationship between family ownership and firm value is weaker for

firms in countries with more developed institutions (i.e. higher institutional quality).

10 See La Porta et al. (1998); Young et al. (2008).

13 | P a g e

Family Control and Firm Value

Around 80 years ago Berle and Means (1932) argued that with a firm`s growth the separation of ownership and control will unavoidably replace concentrated family ownership and control. But in contrast to their idea La Porta et al. (1999) suggested that except in economies with very good shareholder protection, firms are typically controlled by families or the State. The greater part of large, publicly traded firms in Asia, Europe, and Latin America are family owned and controlled (Faccio et al., 2001)

. La Porta et al. showed that controlling shareholders typically have power over firms’ success significantly in excess of their cash flow rights, primarily through the use of pyramids and participation in management (i.e. a member of the family is the CEO). Thus the considered mechanisms of monitoring family control are pyramid structure and appointing a family member as CEO. As is discussed below, the two types of control have been considered from both an agency and resource-based perspectives. While opinion regarding the appointment of a family CEO in Europe varies, there is little written regarding pyramid structures. Almeida and Wolfenzon (2003) argue that the pyramid structure strengthens the potential for agency conflicts between minority shareholders and the family who retains control, because of the separation of cash flow from control rights engendered by the pyramid. While, the resource- based theory suggests that ‘informal’ business groups offers access to resources, which may otherwise not have been attainable (Hoskisson et al., 2003). However, in both instances, the literature is deemed to be irrelevant owing to the fact that pyramid structures are not commonly found throughout Europe. To this extent, examining pyramid structures, in line with the paper compiled by Jiang and Peng (2010), may prove to be useless. However, one aspect of the literature on family-owned firms across Europe which is highly pertinent is that of the different effects which generations have on the company control. More specifically, various authors have considered the similarities and differences between first-, subsequent generation ownership and control of family businesses (Sonfield and Lussier, 2004; Handler, 1992). More specifically, they have noted that subsequent-generation family firms could be less-influenced by the original business objectives of the founder. This means that the extent to which family ownership affects firm value might be different across generations.

11 See La Porta et al. (1999); de Miguel et al. (2004).

14 | P a g e

First, Second-, and Third-Generation Family Firms

Sonfield and Lussier (2004) indicated significant differences between first-generation family firms (1GFFs) and the subsequent-generation family firms (Second and/or Third – 2GFFs;

3GFFs) and have defined a 1GFF as a family-owned and managed firm, with more than one family member involved, but only of the first and founding generation of the family. A 2GFF and 3GFF are respectively defined as firms in which the second or third generations of the family are involved in the ownership and management of the company. In this case the founders and other members of the earlier generation are retired from the firm or deceased (Sonfield and Lussier, 2004).

Several studies show different level of firm value that derives from next-generation family members and their relationship with their parents, which could result in poor firm performance (Davis and Tagiuri 1989; Dumas, 1998). For example, Barnes (1988) shows that daughters and younger sons who become CEOs cannot easily shake off their family ties to the bottom levels of the family hierarchy. As CEOs, they then become key figures in incongruent hierarchies. Their positions in the two hierarchies can lead to discomfort, tension, and agony for all members of the family. Outsiders who see this pain are often at a loss to know how to deal with family members as the problems become public and sometimes tragic.

Schulze et al. (2003) demonstrate that family owned and controlled firms which are managed by

their founders (i.e. 1GFF) may display more altruism than the second (2GFF) and third

generation (3GFF). The authors confirm that the subsequent-generations may have more

dysfunctional squabbles. Furthermore, Anderson and Reeb (2003) proved that the second- and

third-generation family CEOs has no effect on market value, and firms with founder CEOs

outperform those with professional CEOs. The reason may lie in the motivations, desires, and

concerns of founders or owners in family firms, which cannot be assumed to be identical to those

of the next-generation (Handler, 1992). For example, Villalonga and Amit (2006) show that

negative effects can occur when family descendant serve as a CEO. They concluded that the

subsequent-generation family firms are not significantly different in value from nonfamily firms

i.e. family management enhances value when the founder serves as the CEO of the family firm

and destroys it when descendants serve as CEO. According to the scholars, founders of a firm

bring valuable skills to their firms and they are inspiring leaders, great visionaries, or

15 | P a g e

exceptionally talented scientists. Serving founders as CEO create value for all of the firm’s shareholders, whereas if family firms are managed by descendant-CEO the minority shareholders are worse and would be open to Agency Problem I - classic owner-manager conflict (Villalonga and Amit, 2006). Hence firms run by descendant-CEO may be “bad” for firm value.

On the contrary McConaughy and Philips (1999), who have adopted a more resource based perspective, showed that descendent-controlled firms (2 or 3GFF) are more professionally run than founder-controlled firms, where the first generation family possesses special techniques or business backgrounds essential for the creation of the business, while the founders` descendants face different tasks as maintaining and enhancing the business, and these tasks are better performed in a more professional manner. Firm founders or first generation management (1GFF) may be resistant to changes, but the statistics show that only 30 percent of the family firms survive the transition to the second generation and only 10 percent make it to the third generation (Beckhard and Dyer, 1983a and 1983b). Furthermore, Sonfield and Lussier (2004) maintain that overall the relationships with customers, suppliers and key stakeholders of firms are usually developed and maintained by first generation owners. If this is the case, the authors indicate that subsequent generations will not be able to maintain the same level of success as the original founders, which essentially implies that handing a firm down to an heir may have a negative effect on firm value in the long run.

Family CEO

European public opinion substantially differs regarding family control of public corporations.

Barontini and Caprio (2006) provided evidence that many European companies have prospered under the founding family for a long time, thereby they connect family firms to long-term success of the firm. But from different point of view the scholars suggest that there are different families’ priorities with respect to those of outside shareholders and are seen as potential for conflict of interest that may hinder value creation and the growth of European companies (Barontini and Caprio, 2006).

As is mentioned earlier appointing a family member as company CEO is a common way to

maintain family control within the organizational structure. Appointing a family member as

CEO, however, may be seen as detrimental according to some streams of agency theory and

resource-based view.

16 | P a g e

Minichilli et al. (2010) regarded CEO as the most important and powerful organizational actor, who has the overall responsibility for the conduct and performance of the entire organization (Finkelstein and Hambrick, 1996). Some of the most important tasks of the CEO, beside planning, organizing, coordinating, commanding and controlling (Fayol, 1949), can be characterized as follow: The CEO is the charismatic representative of the organization (Fanelli and Misangyi, 2006); The CEO is the leader of TMT (Wu et al., 2005), and it provides family CEO with motive to behave ‘altruistically’

. Thus, the family CEO will undertake actions that favor profit and profitability for their family firm i.e. their family will gain a benefit (Schulze et al., 2001, 2002, 2003). In this vein McConaughy (2000) presents the classical arguments from agency theory i.e. the family relationship between top managers and owners may reduce agency costs and increase long-term incentives for top managers, primarily for CEOs. Thus firms with family CEOs may outperform firms with non-family CEOs (Lee et al., 2003).

Opposite to his findings McConnell and Servaes (1990) argued that the agency problems may arise when family member operates as CEO, because the payout to outside shareholders will be reduced, due to taken policies that benefit the CEO`s and their families. The scholar argued that usually the CEO is neither the founder of the firm neither the biological head of the family, because family CEO may be unqualified and incompetent. Furthermore, even the family CEO is qualified, but not strictly disciplined, he/she may deviate from shareholder wealth maximization (Carpenter et al., 2003; Gomez-Mejia et al., 2003).

From resource-based view a lot of studies

regarded family involvement with a source of sustained competitive advantage in gaining access to unique resources such as common interest, goal congruence, trust and reciprocity, resources that are hard-to-imitate. Thus family CEO may have more advantages in accessing resources that otherwise would not be available to the firm (Arregle et al., 2007). In a contrast, other scholars argued that family firms must be managed effectively and resources as family ties are necessarily but insufficient to achieve competitive advantage and that family firms must be managed effectively

. In this vein Schulze et al. (2003) described altruistic behavior as impediment for firm value creation and contributor for agency

12 According to Schulze et al. (2002) ‘altruism’ is a moral value that motivates individuals to undertake actions that benefit others without any expectations of external rewards.

13 See Sirmon & Hitt (2003); Habbershon & Williams (1999); Habbershon et al. (2003); Carney, (2005).

14 See Sirmon & Hitt (2003);Eddleston & Kellermanns (2007); Eddelston et al. (2008

17 | P a g e

problem, i.e. deeply altruistic, family members contribute to a curious mix of rationalities, juxtaposing contradictory economic and altruistic (non-economic) motivation . As a consequence of less effective monitoring and disciplining of family managers, due to emotional relationship, sentiments, and informal linkages between family owners and family CEOs, the agency problem become more difficult to resolve (Schulze et al., 2001). Thus, altruism and the inability of the parents to discipline underperforming adult children serving as CEOs will lead to agency problems (Schulze et al., 2003).

Another view is that the opposite of altruism will lead to more ineffective managing of family CEO. Gomez-Mejia et al. (2001) present some costs that can incur as a result of family management, such as sibling rivalry, generational envy, non-merit based competition, and irrational strategic decision. By entering into power competition with other family members, family CEOs enhance their own power and prestige rather than to create a profit. Furthermore, Gomez-Mejia et al. (2001) describe situation in which the 1GFF passes away, and as a result inter-generational squabbles occur. Thus, the family business is harmed and converted into sibling partnership, where undisputed authority lacks, because usually the CEO is neither the

founder of the firm, nor the biological head of the family (Schulze et al., 2003).

Given that arguments lead us to the assumption that family CEO may be detrimental for firm value.

The Institutional Context of Family Control

The central problem for family controlled firms is that minority shareholders suffer expropriation by controlling shareholders (La Porta et al., 2002) which is referred to as Agency problem II.

Peng and Jiang (2010) argue that not all family controlled firm expropriate their minority shareholders, on the contrary they suggest that a large part of family managed firms treat their minority shareholders fair. They do state however that reputation is a poor substitute for formal institutions such as legal protection of minority shareholder rights. In a time of financial crisis even large family controlled firms may find themselves expropriating their minority shareholders, in order to make up for their losses (Johnson et al., 2000). As a result of the abuse

of their minority shareholders, family controlled firms value suffers.

18 | P a g e

Many scholars

argue that although the individual families differ substantially in their propensity to expropriate minority shareholder, legal and regulatory institutions are rooted in cross-country differences of expropriation. In a recent paper Morck et al. (2005) recognize the benefits of shareholder control, but whether these benefits outweigh the drawbacks remain unclear. This study suggests that in countries with higher level of investor protection i.e.

countries with presence of more developed legal and regulatory institutions may be beneficial for having family control by appointing a family member as CEO, due to better investor protection and higher level of family monitoring. Put differently, a more developed legal and regulatory institution makes expropriation of minority shareholders less efficient (La Porta et al., 2002), hence the resource provision benefits of having a family CEO may outweigh the costs of Agency Problem II (Peng and Jiang, 2010). In this vein Young et al. (2008) argue that expropriation of minority shareholders is more observable, in countries with weak rules and regulation, where the legal system is more willing for corruption. Moreover, in such countries with less developed legal and regulatory institutions, controlling families often have a relatively ‘free hand’ in expropriating minority shareholder (Bertrand et al., 2002).

This essentially implies that in the case of weaker institutional contexts, having a member of the family appointed as the CEO is likely to allow for an improved level of firm value as better institutional quality allows family CEOs incentives and actions for private benefits to be more strictly observable.

Thus, this paper hypothesizes that:

Hypothesis 2: The negative relationship between a family CEO and firm value is weaker for

firms in countries with more developed institutions i.e. higher institutional quality.

Additionally, in countries with less developed legal and regulatory institutions to protect investors, having a descendant-CEO or a family CEO may increase the amount of expropriation of minority shareholders (Peng and Jiang, 2010). Villalonga and Amit (2006) indicate that family ownership will only have a significant effect on firm value when it is combined with some kind of family control or management. As previously mentioned, second and third generation (2GFF

15 See Dyck and Zingales (2004); La Porta et al. (2002); Lee and Oh (2007).

19 | P a g e

and 3GFF’s) are likely to allow non-family members into top management. This implies that second or third generation ownership can destroy firm value, which when coupled with poor institutions is likely to have an even greater effect (Anderson and Reeb, 2003).

Given that, this study argues the benefits and costs of family control firms are rooted in institutional context, the expropriating of minority shareholders through subsequent generation ownership will decrease in environments with more developed institutions.

Thus, this paper further hypothesizes that:

Hypothesis 3: The negative relationship between the subsequent-generation and firm

performance is weaker for firms in countries with more developed institutions i.e. higher institutional quality.

20 | P a g e

Methodology

Data: Sample and Variables

The sample considered for this study includes 270 firms, randomly selected from 27 European Union (EU) membership countries.

Table

and a list of firms considered is included in the appendix. The Datasets contains 10 family- owned firms per European country, which were manually selected which provides information on listed and unlisted firms (Orbis, 2014).

While, as discussed, a number of different definitions are used to distinguish family firms, Villalonga and Amit (2004) suggest that family firm refers to either the founder’s family or to individual or family that becomes the largest non-institutional shareholder in the firm through the acquisition of a block of shares. This definition is in line with the definition presented in Orbis database, which states that family firms refers to individual or family that possess path of min 50.01% control. Therefore this study will be based on it.

Both private and public companies have been considered for the purpose of this study; and firms included span across several industries. Country-specific institutional quality indicators which are used to proxy shareholder protection rights, the total quality of legal background institutions and government effectiveness were obtained from Kaufmann et al. (2002).

An overview of the model considered for this study is given as:

represents the constant term and

the error term which are assumed to have constant variance across the firms within the sample. The remainder of the variables is explained below:

16 Due to a lack of available firm level data Malta has been excluded from the sample set considered for this study.

21 | P a g e

The Dependent Variable

Firm Value has, traditionally, been measured using a variety of performance measures in existing literature – all of which are argued to have both advantages and disadvantages.

In general, authors debate the validity of using market-based vs. accounting-based measures for accuracy in determining firm performance.

In general, market-based measures eliminate problems associated with distorted accounting data and are not influenced by firm-specific financial reporting. However, it is noted that market based measures of firm value are based on the pricing determined by the market. To this extent, market based measures reflect the perception of market participants of the firm value of the company and not necessarily its true underlying value. Additionally, speculation and market trends influence the market based measures of firm performance. Within their limitations, Peng and Jiang (2010) suggest that the use of other measures of firm value may be more appropriate.

Given, that this study is done around the peak crisis period (considered to be in 2009) the use of a market-based measure such as the cumulative stock return (as is used in Peng & Jiang, 2010) is considered less adequate. It is expected that the market based measure of firm value may be distorted by the underlying economic and financial uncertainties plaguing the markets during this time. In light of this, this paper will measure firm value using an accounting-based measure of firm value, namely Return on Assets (ROA). Corstjens et al (2004) indicate that higher levels of ROA may not necessarily be reflected by a higher level of market performance and as such, findings for this study may be less accurate based on the use of a market-based measure.

ROA is defined as the net operating income before extraordinary items have been deducted divided by the firm’s total assets. ROA is considered to be a well understood and common measure used in several similar studies, which implies that it is easily comprehensible.

Moreover, it is suggested that ROA is an appropriate measure for assessing firms across several industries.

Firms considered are those which comply with global Generally Accepted Accounting Practices (GAAP) and as such financial reporting distortions are not expected lead to bias within the sample. While many US based studies make use of Tobin’s Q to measure firm value, studies focusing on Europe have, in general, used ROA (Corstjens et al, 2004). This is

17 For an overview see Murphy, Trailer and Hill (1996).

18 See Cannella and Shen (2001); Carpenter (2002); Finkelstein and D’Aveni (1994); and Henderson et al (2006)

22 | P a g e

important because by using ROA in this study we provide an opportunity for easy comparison between our study and other studies of this nature conducted in Europe.

The peak of the global financial crisis is said to have been during the first quarter of 2009 as this is the period in which stock market indices reached their minimum. Following the work of Jiang and Peng (2010) in which they try to capture the effects during and immediately after the crisis, the dependent variable in this study is measured using the ROA of the selected firms measured at the 2009 financial year end.

In order to determine robustness of this measure, the study, additionally makes use of Return on Equity (ROE), defined as the net income divided by the average capital employed, to proxy for family firm performance. ROE is also used commonly within studies of this nature. ROA and ROE have been calculated making use of data denominated in US Dollar Values.

Independent Variables

All independent variables considered for the study lag the dependent variable (i.e. 2008), and are measured before the peak of the global financial crisis to avoid confounding effects associated with the crisis (Joh, 2003; Peng and Jiang, 2010). This is done to ensure that we are able to see the full effect of the relationship between family ownership and control and firm value of family firms both during and directly after the peak of the financial crisis; and is in line with the paper by Peng and Jiang (2010).

Family ownership given by

is measured by the fraction of ownership of each family for each respective firm. As previously mentioned, family ownership is determined in the case that founder’s family or individual or family becomes the largest non-institutional shareholder in the firm through the acquisition of a block of shares. To permit a deeper analysis of the effect of family ownership on firm value, the paper explicitly identifies firms in which a family member serves as an executive on the management team. This is indicated within the study using a binary dummy variable (

) which takes on the value of 1 in the case that a family members serves as an executive or CEO and zero otherwise.

Measuring subsequent firm ownership (i.e. whether a firm is a 1GFF or 2/3GFF) is a difficult

task based on the lack of data availability. More specifically, when a firm is owned by an

individual family, it is not necessarily clear when one generation moves out of senior

23 | P a g e

management and the next generation takes over i.e. there is often an overlaps. Private firms may also not make this information explicitly available. To overcome this obstacle many authors

have considered the generation ownership of firms from a life cycle perspective. In a model developed by John Ward, he suggests that a family owned institution undergoes three main lifecycle stages namely, the early stage, middle and late stage, where he imbeds in the model parts of the lifecycle of the parents or owning generations and their children. In order to capture the effects of generations, he equates generational development patterns to the life cycle of the firm. More specifically, he suggests that a firm aging between 0 and 5 years is in the early stage where the founders are of age 25-35 years and the business is rapidly growing and demanding time and money, while a firm aging 10 to 20 years old is in the middle stage where the nature of business is maturing and the owners are in the age brackets of 40-50 years old, while the children are approaching 15-25 years. During both these stages, it is suggested that the firm is owned and controlled by a founding family member and as such is a first generation company. The third or late stage characterized by the need for strategic regeneration and reinvestment is then dominated by subsequent generations. In this stage the founders are 55-70 years old, approaching retirement, while their children are in the age group of 30-45. In this period the focus is on new interest where the new generation seeks growth and change. When the subsequent generation reaches ‘regeneration’, the model could repeat itself several times in dynamic families.

Furthermore, Ward explains that there are varying challenges of each stage and the transition is very difficult, and as a result many family firms will not continue to exits through the three stages. This method is very ad hoc simplification and as such is not necessarily a decisively accurate method of determining the generation ownership of businesses. To this extent, this paper has sought to determine the exact ownership of firms in order to provide as accurate an insight into firm ownership by founding and subsequent generations as possible. In cases where data was not available for firms included in this dataset, the paper has made use of Ward’s simplification. The paper has identified the exact generation ownership for 161 of the firms included in the dataset and have estimated the remainder using Ward’s model. Family generation (

) is measured using a dummy variable which takes on the value of 1 in the case that a firm is owned and/or controlled by the founding family member and zero in the case that it is owned by subsequent generations.

19 See Neubauer and Lank (1998); Handler (1994).

24 | P a g e

Institutional Quality is a country specific independent variable used to measure the impact of the overall quality of legal background institutions in the various EU jurisdictions considered for this study. As previously indicated, it is understood that the institutional context of a country affects the economic performance of firms; where evidence suggests that legal rules protecting investors and the quality of their enforcement differ greatly across countries and this essentially affects shareholders against expropriation by insiders. Within their 1998 paper, La Porta et al examine the legal rules covering the protection of corporate shareholders and creditors, in 49 different countries, to the origin of these rules as well as to the quality of their enforcement. Their findings indicate that shareholder protection is rooted in the legal structure of a country, which the authors argue is an external governance mechanism of firms. In institutional settings where there are stronger shareholder protection rights fewer internal governance mechanisms are necessary to formulate and implement successful corporate strategies (Gedajlovic and Shapiro, 1998; Walsh and Seward, 1990). The argument of this paper is made from this institution-based perspective and is that the effect of family ownership on firm value depends to a large extent on the prevailing institutional framework. To this extent, a measure of institutional quality is included in the analysis to serve as a proxy for the legal and regulatory institutional development in a country.

Jiang and Peng (2010) draw on the work of La Porta et al (1998) by including three broad institutional measures crucial for the protection of investors; namely (1) efficiency of the judicial system, (2) rule of law and (3) corruption. Judicial efficiency indicates ‘the efficiency and integrity of the legal environment as it affects business’; Rule of Law is the law and order tradition in the country and Corruption is the extent of corruption in the government particularly the extent to which businesses have to pay bribes. In each case the institutional quality is measured in 1997 (i.e. before the Asian financial crisis) and the average across the three scores in the index for each country is used to proxy for legal and regulatory institutional development.

As previously mentioned the dataset compiled by La Porta et al only has data available for 16 of

the 28 EU member countries and considered for this study. To this extent, this paper seeks to

include data from a more comprehensive database namely the World Bank World Governance

Indicators compiled by Kauffman et al (2010). Amongst scholars there is no agreed definition of

institutional quality and so for the purpose of this paper the definition given by Kauffman et al

25 | P a g e

(2010) is used. It represents institutional quality as ‘the traditions and institutions by which authority in a country is exercised’, which includes:

1. The process by which governments are selected, monitored and replaced;

2. The capacity of government to effectively formulate and implement sound policies; and 3. The respect of citizens and the state for the institutions that govern economic and social

interactions among them’.

All institutional quality indicators used in this study have been drawn from The World Bank WGI database.

Table 1: Countries included in the Dataset Higher Quality Institutions

Austria; Belgium; Cyprus; Denmark; Finland; France; Germany; Ireland; Luxembourg; the Netherlands; Sweden; the United Kingdom

Lower Quality Institutions

Bulgaria; Croatia; Czech Republic; Estonia; Greece; Hungary; Italy; Latvia; Lithuania; Poland;

Portugal; Romania; Slovakia; Slovenia; Spain

Table 2: Institutional Variables

Variables Details

(a) The process by which governments are selected, monitored and replaced:

Voice and Accountability (VOAC)

Perceptions of the extent to which a country`s citizens are able to participate in the selection of the government along with freedom of expression, freedom of associations, and free media.

Political Stability & absence of Violence/ Terrorism (PSVT)

Perception of the likelihood that the government will be destabilized or overthrown through unconstitutional or violent means, including politically- motivated violence and terrorism.

(b) The capacity of the government to effectively formulate and implement sound decisions:

Government Effectiveness (GVEF)

Perception of the quality of public services, the quality of civil service, and the degree of its independence from political pressures, the quality of formulation and implementation, and the credibility of the government`s commitment to such policies.

Regulatory Quality (RGQT) Perception of the ability of the government to formulate and implement sound policies that permits and promotes private sector development.

(c) The respect of citizens and the state for the institutions that govern economic and social interactions among them:

Rule of Law (RULW)

Perception of the extent to which agent have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence.

Control of Corruption (CNTC)

Perception of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as ‘capture’ of the state by elites and private interests.

26 | P a g e

The institutional qualities (

) are provided as an index varying between -2.5 (reflecting the weakest institutional quality level) and 2.5 (the strongest institutional quality level). Given that the aim of this study focuses largely on the institutional quality which essentially matters for shareholder protection rights it is deemed that not all six of the indicators should carry equal importance. Looking at Table 2 and the respective definitions it is shown that

‘Rule of Law’ and ‘Control of Corruption’ are similar measures to those considered by Jiang and Peng (2010). However, in the case of the ‘Efficiency of the Judicial System’ there is no exact counterpart provided in the WGIs. In light of this both measures of Government Effectiveness and Regulatory Quality are considered where regulatory quality is thought to be most prevalent for measuring shareholder protection and government effectiveness has been included to ensure completeness.

Owing to the fact that, firstly, political stability and the absence of violence as well as voice and accountability are not shown to be pertinent aspects of shareholder protection by La Porta et al (1998) in their research; and secondly are not considered within the paper by Jiang and Peng (2010), this paper will focus only on the four above-mentioned key institutional quality indicators.

To ensure that the WGIs serve as a sound substitute for the data provided by La Porta et al, it is

necessary to examine the correlation between the different indicators. Because of the different

scaling across the two different datasets, their correlations have been examined graphically. This

is only possible for countries which are common to both datasets. Graphical analysis (see graphs

1, 2 and 3) indicates that the data for 14 countries across both datasets is relatively highly

correlated and thus likely to serve as a very good substitute for one another.

27 | P a g e

Rule of Law La Porta Rule of Law WGI

Corruption La Porta Corruption WGI

Efficiency of Judicial System La Porta

Government Effectiveness WGI

Regulatory Quality WGI

28 | P a g e

In order to assess the effects of overall institutional quality on the relationship between family ownership, control types and firm value a compound indicator of institutional quality is used.

While this averaging of the various institutional quality measures may be a simplification and could weaken the study overall, it is done to keep in line with that of Jiang and Peng (2010).

As previously mentioned, the four measures of institutional quality which are considered to be pertinent for measuring shareholder protection rights are RGQT, GVEF, RULW and CNTC. The compound indicator is, therefore, calculated using the average of each of these four specific institutional quality indicators, measured in 2008, contained in the WGI dataset. Countries are then assessed according to whether they can be viewed as institutionally ‘stronger’ or ‘weaker’

using this average. The average compound indicator of institutional quality across the 27 EU countries included in the sample set is calculated to be 1.06. Countries whose compound indicator lies above 1.06 are considered to have higher levels of institutional quality while those that score below 1.06 are considered to have lower levels of institutional quality. The distinction of the two country groups can be seen in Table 1 – it is indicated that 12 of the countries in the sample are considered to be institutionally ‘stronger’ countries while the remaining 15 have a lower level of prevailing institutional quality.

In order to capture the interaction effect of institutional quality on family ownership characteristics, interaction variables (

;

and

) are included in the model.

Control Variables

The study includes different control variables which are used to proxy for industry and firm characteristics. These include firm age, firm size, firm leverage, country and industry within which the firm operates. These are discussed in detail below:

Firm Size (

) is measured using the natural log of the number of full time employees, as is done in a majority of previously published papers of this nature.

Firm size is considered to be an important control owing to the fact that larger firms are more likely to have greater access to resources; which in turn affects firm performance.

20 Following various other studies the natural log of the book value of total assets is also considered as a proxy for firm size in the study. Owing to the fact that this is considered to be highly correlated to ROA (the dependent variable) and is not found to be statistically significant, its results have been omitted.

29 | P a g e

Firm Age (

) is additionally considered to have an impact on firm performance and as such this is controlled for within the model. Firm age is measured using the natural log of the number of years that the firm has been in existence.

Firm Leverage (

) is considered to matter since firms with a high leverage ratio may not perform as well during economic difficult times – since highly leveraged firms are less likely to be able to borrow or obtain external financing during crisis periods. Firm leverage is controlled for using the 2008 debt-to-asset ratio of firms in the dataset.

Industry Effects (

) have been accounted for my including industry dummy variables. Differences across industries could lead to companies varying levels of performance.

Firms included in the dataset are spread across 8 industries including: Construction (12);

Financial (27); Information and Communication Technology (8); Industrial (113); Logistics (9);

Manufacturing (58); Real Estate (6); and Services (35).

Country Effects (

) have been accounted for using country dummy variables to account for ‘all non-institutional’ difference across individual countries.

Table 3 reports the descriptive statistics of the sample. With respect to the firm level data it is apparent that the average family equity ownership share in the sample is given at 86.93%. It is also indicated that a majority of the firms in the sample set have a family member in their management team (i.e. Family CEO) and are owned by founding firm members (i.e. are first generation). The average family size considered for the study is represented by 2225 full time employees and the average firm’s age is 16.26 years in existence. The average ROA (measuring firm performance/firm value) in the sample is given at 23.01% while the average leverage ratio is 35.79.

The argument of this paper considers the institutional environment most pertinent for firm

ownership and performance, which are indicated in previous studies to be government

effectiveness, control of corruption, rule of law and regulatory quality. Looking specifically at

these institutional qualities prevailing in countries, it is shown that the lowest level of compound

institutional quality is 0.005 (Romania) while the highest level is 2.065 (Denmark). Individual

measures of institutional quality have been included and it is indicated that in terms of RULW

and CNTC Bulgaria scores the lowest with -0.160 and -0,304 respectively. Croatia scores the