The Dutch interlocking directorate network: A persistent decline or a revival?

(1)

1

The Dutch interlocking directorate network:

A persistent decline or a revival?

A qualitative study on the evolution of the corporate elite in the

Netherlands

Master Thesis A.G.N. Langhout S1460773 October 2010 University of Groningen

Faculty of Mangement and Organization Msc. International Business and Management

(2)

(3)

3

Preface

I have written this thesis to finalize my Master International Business and Management at the faculty of Management and Organization of the University of Groningen. I would like to use this opportunity to show my appreciation to those who have supported me to finalize my thesis properly.

First of all I would like to thank my first supervisor Prof. Dr. D.J.F. Kamann for his continuous support, constructive feedback and for the interesting conversations we had during the course of my thesis. Secondly I would like to thank my second supervisor Prof. Dr. L. Karsten for his feedback in the final stage of this thesis. Furthermore I would like to thank my family and my girlfriend for their everlasting support.

(4)

4

Management Summary

Following Helmers(1975), Heemskerk(2003, 2007) and van der Lee(2008), this research entails a qualitative study on the Dutch corporate interlocking directorate network(IDN). Using both an interfirm as well as an interpersonal perspective it tries to give an insight in the corporate network in the Netherlands. The research question is formulated as follows:

To what extend has the decline of interlocking directorates in the Netherlands persevered over the past decade, and into what kind of network the IDN has transformed into.

(5)

(6)

6

Introduction

The past decade the world has been startled with severe mismanagement and even fraudulent management among several large and respected global firms, such as Ahold, ABN-AMRO, Enron and Parmalat, which inevitably led to the near downfall of those corporations. Leaving behind devastated companies, extirpated jobs and mutilated savings mismanagement left a scar on the face of corporate governance. Subsequently the financial crisis dropped anchor in 2007 resulting in the collapse of large financial institutions such as Northern Rock, Lehman Brothers and Merill Lynch; bailouts of banks by governments in the case of Fannie Mae, Freddie Mac and AIG and downturns in stock markets leaving the stock exchanges to hit rock bottom. Although the prime origin of the financial crisis is attributed to a complex web of mortgage backed securities, collateralized debt obligations, deregulations, increased debt-burdens, incorrect risk pricing and other financial innovations and ramifications (Davis and Mizruchi, 1999; Cheng, Hong, and Scheinkman, 2010), corporate mismanagement is the fundamental substance that ignited the greatest financial crisis since the Great Depression of the 1930s. Both corporate scandals and the financial crisis caused a public debate questioning the role of executive and supervisory boards. The shareholder wealth maximalization model is publically criticized. As both the board of executives as the supervisory board are responsible for corporate governance they failed to do so (Krugman, 2002). Intertwined in the quest for continuing growth of shareholder value, growth targets of firms reached extreme high levels compared to industrial average(Cools, 2005) with their CEOs having eight times as many stock options and higher stock option rewards than their colleagues. Although designed to be an incentive for CEOs to battle for shareholder value it ironically enticed them to commit fraud (Cools, 2005). Furthermore, CEOs of fraudulent firms, enjoying positive reputations have been alleged to receive more media attention than colleagues of non-fraudulent firms which led them to believe in their own superiority.

(8)

8 management deceitfully confine the supervisory board’s power or did the latter step aside readily? The answer lies with the supervisory board and their interlocking directorates. At first sight it is not that obvious, but taking a few steps back catches a glimpse of the vast network of top managers dominating the corporate landscape. Since the early twentieth century there have been tenacious claims reporting such networks (Jeidels,1905; Hilferding,1909; Riesser, 1910; Wibaut,1913; Hagemann, 1931). In the 1960s, J.P.P Mertens was the first to blow the whistle on board interlocks publically. Mertens(1968), a Dutch catholic trade union leader, argued that a group of roughly two hundred men controlled the Dutch economy, inspiring a public debate. Some years later, Helmers et. al. (1975) confirmed that all but two out of the sixty-eight largest Dutch companies formed an elaborate network through 873 interlocks which were carried by a group of only 195 directors. Five people fewer than Mertens had stated three years earlier. The interlocking directorate network was made visible, interlocking directors had nowhere to hide anymore.

An interlock exists between company A and company B if a director of A, whose main business activities coincide with A, sits on the board of company B. If that same person also sits on the board of C, then there is also an interlock between A and C (Koenig et al. 1979; Useem, 1980; Pettigrew, 1992; Mizruchi, 1996; Heinze, 2004; Kentor and Jang, 2004; Heemskerk and Fennema, 2009). An interlock is called a primary interlock as there is a direct link from board A to board C or A to B (Granovetter, 1985). These two interlocks however also induce a so-called secondary interlock, the interlock between B and C, which are not directly connected (Sweezy, 1953). Thereupon, when several interlocking directors hold positions simultaneously on the board of one firm, a network of interlocking firms emerges, which is defined as an interlocking directorate network.

(9)

9 Mizruchi, 1999) or on transnational networks (Sklair, 2001; Caroll and Fennema, 2002; Graz, 2003; Kentor and Jang, 2004). The latter is studied less with empirical studies (Heemskerk, 2002), yet not insignificant. Individual perspective studies show that the allocation of board members depends on informal connections and personality characteristics (Galaskiewics et al, 1985; Yeo et al, 2003). It regards the corporation as a holistic concept in which the company is not an actor as it cannot act on its own, the people managing the corporations – its directors - do.

According to Mizruchi (1996), enacting the corporate perspective, individuals carry the interlocks as they have the expertise and knowledge adding value to the firm. Besides meeting one another in board rooms, they also meet at informal occasions, creating a dense network within the corporate landscape, varying in redundancy, visibility and strength (Granovetter; 1978; Fennema, 1982; Burt, 1992; 2005). Thus, in this perspective, the individual actors are mere pawns in a game of chess, set forth in the quest for check-mate. Granovetter (1975) elaborates on this, stating that a link is stronger if it occurs more often, being its multiplicity. Burt(1992) specifies stating that it is key to get as many non-redundant ties as they form bridges to other clusters; the multiplicity of ties together with the avoidance of these structural holes described by Burt, persuades corporations to enhance their network to a broad variety of industries, legitimizing their network power.

Besides interlock multiplicity and redundancy, the actual purpose of interlocking has to be taken into account. Co-optation (Allen, 1974); Trust (Heemskerk, 2002), Financial benefits(Mizruchi, 1986), Intercorporate control and influence (Mariolis, 1975; Koenig et

al, 1981; Mintz and Schwartz, 1981, Pennings, 1982) Centrality (Dooley, 1969; Burt,

(10)

10 In the past, Dutch research (Helmers, et al, 1975; Heemskerk et al, 2002) concerning these networks argued a decline of interlocking directorates between 1976, 1996 and 2001 and with it, a deterioration of the so-called old-boys network. Kentor and Jang(2004) however, dispute these finding by stating that between 1976 and 1998 the number of interlocks have increased in the US. Nonetheless, our scope concerns the Netherlands, and as early research proved that the level of interlocking directorates differs highly between countries, it enables us to refrain from their point of view, restraining our focus on the Dutch corporate interlocks and the peripheral old-boys network.

Throughout history, this aforementioned old-boys network, consisted of members recruited from the ranks of nobility or patriciate through nepotism and allocating them to influential occupations in business or government, creating a network of highly influential upper-class people dominating a country’s economy (Hezewijk, 2003; Dronkers and Stokman, 1984; Fennema, 2007). Therefore it is interesting to see, roughly a decade after Heemskerk’s findings(2002), how the declining course of interlocking directorates and with it the old-boys network has advanced since the 1970s. Furthermore it is also of interest with regard to this old-boys network, which actors in the network are regarded as the most powerful, influential directors. It is of interest to see whether they show commonalities.

This results in several research questions. The first concerns a descriptive issue. It is important to know what the interlocking directorate network consists of and how these corporate interlocks function. This study takes both an individual and a corporate perspective, assessing to what extend similarities arise between the different directors or firms. Clarifying the position of control and ownership in corporate interlocks concerns the explanation of underlying reasons and purpose of the network as well. By analysing how this network evolved in the years passed this thesis hopes to present a proper comparison and analysis of the different links between firms.

(11)

11 using UCINET will take place. Rendering an analysis on the actual state of the interlocking directorate network.

This can be summarized in one general research question:

To what extend has the decline of interlocking directorates in the Netherlands persevered over the past decade, and into what kind of network has the IDN transformed.

To study these interlocking directorates in the Netherlands I will elaborate on the studies of Mokken and Stokman (Helmers, 1975), Heemskerk(2002) and van der Lee(2008). Creating a primary dataset consisting of all stock listed corporations in the Netherlands and their board of executives and supervisors. Their education; former employers and their other supervisory positions will be drawn upon. By doing so I will study the decline of interlocking directorates. Consequently the ties of these directors will be traced to their origins in order to unveil the real network of interlocking directorates rendering a second and third dataset. Using this new reversed selection procedure the network entails smaller non-listed corporations as well as cultural, charitable and educational organizations.

(12)

12

I. Social Network Theory

Over the past decades there has been a considerable interest in social network theory (Granovetter, 1984; Mokken, Stokman; Mintz and Schwartz; Borgatti ; Everett; Scott, 1992;2000; Wasserman and Faust, 1994; Mizruchi, 1996; ). It entails a distinctive methodology encompassing techniques for the quantitative analysis of complex social structures, specifically with regard to relationships among social entities, and for the patterns and implications of these relationships (Wasserman and Faust, 1994). The focus on the relationships and the additional patterns required new concepts and methods. From this point of view Wasserman and Faust(1994) identified four key principles. First of all actors and their individual action are regarded as interdependent rather than independent. Secondly, relational ties – linkages – between actors are channels for the flow of resources. Furthermore network models with an individual perspective view the network structural environment as an opportunity for or constraint on their actions. Lastly, structures are conceptualized as longlasting patterns of relations among the actors.

Graph Theory

All together the graph theory pioneered by Cartwright and Harary(1956) is a fundamental basis of social network theory. They envisioned groups as a collection of points connected by lines, hence a graph. Graph theory entails a set of mathematical principles and formulas describing the properties of the lines (Scott, 2000). Each point represents a person or a corporation and is dubbed an actor. Since this study focuses on a set of actors of the same type – executive board members in the three indices – it looks at a one-mode network (Wasserman and Faust, 1994). The lines connection these actors are called social

ties and they represent the relationships between them. This can vary from friendships,

(13)

13 are to be measured we look at the group as a whole. The collection of ties of a specific kind among members of this group is defined as a relation (Wasserman and Faust, 1994). This might cause some uncertainty as one might confuse it with the aforementioned relational tie. A relation refers to the collection of ties of a given kind measured on pairs of actors from a specific set of actors. Relational ties on the other hand, only exist between specific pairs. A criticism on the graph theory is its lack of dynamics. As a static theory with a binary approach, it is unable to completely measure all relational ties and their strengths. Yet, it is the best theory we have to analyse the interlocking directorate network in the Netherlands.

Having defined all specific terms within the social network analysis we are able to explain the concept of social network properly. Wasserman and Faust(1994) defined it as a finite set or sets of actors and the relation or relations defined on them. Fortunately, social networks can be visualized using Netdraw (Borgatti, 2002).

A social network can have several artificial structures such as a star, circle or line (Freeman, 1980; Wasserman and Faust, 1994). These three graph forms present a hypothetical clean version of possible relations which in reality, are combined throughout the social network. This depends on the centrality of its actors, which is elaborated upon in section IV on the Methods of this study. When two different actors are connected through a tie with a third actors an interlock arises. Cumulating these interlocks results in an interlocking network which will be elaborated upon in further chapters.

(14)

14

II. The Dutch Corporate Elite

Mertens(1968) was the first in the Netherlands to acknowledge the fact that a dense network of individuals controlled the major corporations. Endorsed by Helmers(1975) it became clear that 195 top executives carried the sixty-six largest companies through 873 interlocks. These top executives are known as the corporate elite (Domhoff, 1974; Useem, 1980).

History of the corporate elite

(15)

15 education. Clearly, these three forms are inextricably connected to one another, making it nearly impossible to join the corporate elite.

After the Dutch aristocratic dominion over the past centuries, during which only the nobility and patriciaat had access to cultural capital, universities became accessible in the mid-twentieth century by the bourgeoisie when wealth - and therefore elite education - was not the sole property of the aristocratic families anymore. Therefore the power of Dutch aristocracy declined (Dronkers and Hillege, 1998). As a result, bourgeoisie wealthy families joined the ranks of the corporate elite. Henceforth, from that point on becoming a member of the corporate elite is to a great extend induced by the combined characteristics of a wealthy family and a prestigious education (Maclean, 2006). Where the former is imperative for the latter.

(16)

16 intellectual requirements. Since the 1950 the government decided to subsidize universities and students in order to give every Dutch person the same opportunities. Elite status nor welfare are nowadays minimal prerequisites to enter an university. This discarded the segregation between elite and non-elite education of the 1950s and enabled the commoner to attain executive and supervisory positions in the future.

Up to the 1960s there was no concern on the amount of power that lay with so few, only until Mertens accused the corporate elite to possess too much and uncontrolled power this changed (Heemskerk, 2003). The corporate elite replied defensively by proclaiming that these elite members were only sharing the responsibilities without abusing them (de Bruyn, 1972). This reaction can be explained as those inside an elite inner circle typically have a different view altogether as sharing common social background enhances the sharing of same norms and values and a common enemy (Useem, 1980). They argue the limits of their influence, and the lack of cohesion amongst them..

The old corporate elite falling apart

The public debate on elites can be divided in two different perspectives. A pluralist perspective in which the intransigent dissimilarities were demonstrated between the elite and in which no single omnipresent elite could exist because of a lack of overlap between the various elite sections (Robert Dahl, 1989). On the other hand an elite perspective stressed that a single power elite network dominated society. A closed network of elite members controlling the big businesses (Domhoff, 1970; 1980 and Mills, 2000).

(17)

17 The rise of a new elite

After the Dutch government’s decision to subsidize education in the 1950s enabling all Dutch to attend a university, the gateways to elite positions opened up. Although aristocracy still enhances opportunities of an elite position (Schijf et al, 2003), nowadays an academic title also increases one’s chances to attain an elite position. Inequality on basis of social origin was replaced through new social inequality on basis of educational performances (Beekenkamp and Dronkers, 2002). This concurs with the supposition that acquired qualities have become increasingly important in modern day society, ruling out the old boys network as sole magnate inheritors of the corporate elite. As a result a new elite has emerged. A clique of the nouveau riche stepped up to join the ranks of the corporate elite. However, they did not conjoin the old corporate elite, instead they opposed the old-boys network – especially politically – because of their conservatism, resulting in two separate corporate elites. The new elite embodies a meritocratic ideology in which success depends on a person’s individual and personal qualities (Matthijssen, 1972, 1982; Heemskerk, 2003), which coincides with the liberalization of the Dutch educational system in the 1950s. In the past decades this new elite with is meritocratic ideology, is to be found very successful in segmented markets not controlled by the old-boys network (Hezewijk, 2003). First real estate, subsequently media entertainment and eventually IT gave rise to the new elites extensive growth of wealth. They co-exist next to another, with the new elite filling up an increasing number of supervisory position, limiting the old elite’s power.

Imparted with executive or supervisory powers, the established corporate elite found itself defied with ‘the brutal realities of the new global economy along with fiercened global competition; hostile takeovers; growing interest of financial markets; changed shareholder perceptions, enlarged foreign direct investments, deteriorated public opinion on their positions and the creation of the new corporate elite (Maclean et al, 2006; Fennema and Heemskerk, 2008). The new elite has alarmed the old as a result.

The current situation

(18)

18 past decade of globalisation of business affected the board composition in the Dutch corporate landscape. All over the world corporate board internationalize introducing foreigners to top management positions (Gillies and Dickinson, 1999; Staples, 2007, 2008) as Mandle(2003) predicted earlier. In his studies on globalisation of the world’s largest TNC’s, Staples(2007) found that between 1993 and 2005 only few TNCs board rooms contained only two nationalities and that only 26 nations of the world were represented on this board igniting a discussion on the possible presence (Sklair, 2001; Heemskerk, 2003;; Staples 2008) or absence (Carroll and Fennema 2002, 2004; Kentor and Jang, 2004) of a Transnational Business Class. A class that represents the international corporate elite. The increased presence of foreigners over the past 15 years on management boards cannot be explained only through the upcoming international job market for top executives. In a study by Fennema and Schijf(2004) they reported that only 6% of foreign executives were recruited externally. On the contrary the majority of their presence can be explained by an increase in Merger and acquisition activity between 1993 and 2005 with a peak at the turn of the century(Staples, 2008). Today, this explains the presence of both Dutch and French executives at KLM and the presence of two Mexican supervisors after Heineken’s acquisition of FEMSA in 2010. Globalisation too, supresses the old boys network.

(19)

19 represent 21 different nationalities. Interesting to see however is the fact that after the Dutch the second most represented nationality is American(9%) whereas British(7%), French(6%) and German(3%) respectively are third, fourth and fifth most represented. This enrichment of corporate boards diminishes the influence of the nation’s old corporate elite and changes the corporate landscape.

(20)

20

III. Corporate Governance and Control

Although the fraudulent scandals and mismanagement of Enron, Parmalat, ABN Amro and Ahold really ignited a public debate on corporate governance and control, the discussion had already lingered for a while on the public agenda.

In order to comprehend the failure of a governance system in the Netherlands in which executives are encouraged to maximize shareholders wealth, we need to discuss the Dutch corporate governance system. Above all, since governance systems differ widely within Europe and the rest of the world in terms of institutionalization, structure and its control mechanisms.

Dutch governance institutionalism

In general, there are two different institutionalizations of board systems. The first is a one-tier board system, with executive and non-executive directors in a single board, common in the UK and the US (Maassen and van den Bosch, 1999; Jungmann, 2006). This system is criticized by its composition as the directors have to control their fellow board members and by its CEO duality (Rechner and Dalton ,1991; Pettigrew, 1992; Boyd, 1995) in which the firms CEO also serves as chairman of the board of directors. The second is a two-tier board system, more common in Germany and the Netherlands. It consists of an executive board (Raad van Bestuur) and a supervisory board (Raad van Commissarissen) in which the former involves as a management function and the latter only an auditing function (Maclean, 2006).

From this, two different structuring models arise. The Rhenish model and the Anglo-Saxon model which coincide with the coordinated market economy (CME) and liberated market economy (LME) respectively.

(21)

21 although not unique in this world, build on durable relations and are stakeholder oriented (Heemskerk, 2004; Heinze, 2004). This is made apparent by the Dutch corporate regime which is designed to give firms high levels of autonomy, shielded by statutory – to reduce the influence of shareholders through priority and preferent shares - and non-statutory - certificates refraining shareholders from their voting rights – defences (Heemskerk, 2004). Criticism on the Rhenish model is its closeness in contrast to the Anglo-Saxon model which is funded through the stock exchange, whereas the Rhenish model is funded through banks and family capital. This closeness allegedly simplifies the power of the corporate elite network.

In the Anglo Saxon structure the executive and supervisory role are embedded in the board of directors. These directors convene regularly. Even so, a distinction can be made between inside and outside directors. Inside directors are employed by the firm for a longer time period and often comprise an executive committee, which resembles the executive board in the two tier system (Maclean, 2006). On the other hand the outside directors cannot be compared to the supervisory board in the Rhenish structure as they share their supervisory power with the inside directors. Furthermore they involve a more extensive role in the organization than the auditing function of the supervisory board described before.

Besides, the Rhenish structure and the Anglo Saxon structure differ on a third perspective as well. Within the Anglo Saxon model it is prohibited to serve on multiple boards as it reasons from an economic point of view. The only form of participation stockholders have is the exit strategy through which they put pressure on the stock itself (Nooteboom, 1999). As a result, executives within the Anglo-Saxon model strive for the shareholder wealth maximization (Fennema and Schijf, 1978; Heemskerk, 2004) and keeping the shareholders down securing their own positions.

(22)

22 changes. The polder model can be explained as a good example of the consensual policy maintained by the Dutch. The interesting part of the Netherlands is that it can be seen has highly internationalized, in terms of trade and foreign investments and in terms of fto a lesser extent foreign executives present in the corporate elite. According to Scott (1997) the Netherlands combine elements of both Rhenish and Anglo-Saxon structures because of its strong ties with both of them, therefore it is an even more interesting country to study interlocking directorates.

Hence, the Dutch corporate governance is classified as a close to CME structure Rhenish capitalism model in which a two-tier structure is prevalent although links within the Anglo-Saxon model are present. Consequently the Dutch corporate governance entails two types of legitimate control: executive control – its actual management – and its supervisory regulatory control. The former is entrusted with the legal responsibility for making all important decisions as well as the distribution of its financial assets (Fennema and Schijf, 2004), the latter is accredited with regulatory responsibility overseeing the executive control. Henceforth, corporate success depends on trustworthiness of its corporate and supervisory executives (Blair, 2002). In the cases of ABN Amro and Ahold some of these executives turned out to be all but trustworthy. Although bonded by legal constraints they saw ways to deceive all regulatory institutions proving the importance of cooperative, sincere, trustworthy executives and supervisors.

(23)

23 The lack of regulation by supervisory board members can be assigned to three reasons. First of all outside board members are described as passive, confrontational avoiding, lacking in expertise (Koenig, Gogel and Sonquist, 1979). Second, the time spend by these managers on their supervisory jobs limits to one day a month creating a distance between them and the corporation they are ought to supervise resulting in a lack of involvement. Third, supervisors often lack in sovereignty as supervisory board members embody the interests of the companies they represent as executives in order to control their investments, strengthen their relations or to enhance their power in the industry (Heinze,2004).

Interlocking Directorates

As argued before the criticized closeness of the Rhenish structure simplifies the formation of a corporate network. Hirschman(1970) already pointed out that within the Rhenish model the coordinating function of interlocking directorates are applied repeatedly. He stressed that control and coordination work through direct concern, through interlocks. Others too recognized the construating element of interlocking directorates and dubbed them as a structural aspect of social capital of the corporate elite: the ‘grease of the economy’ (Bourdieu, 1985; Colemann; 1988; Heemskerk, 2007). They agree with Granovetters(1985) that corporations do not function as independent actors because they are embedded in social networks.

Jeidels(1905) and Hilferding(1909) were the first to acknowledge the presence of interlocking directorates. In his dissertation “Das verhältnis der Deutschen Grossbanken

zur industrie mit besonderer berücksichtigung der Eisenindustrie” Jeidels(1905) argued

(24)

24 Decades of research passed and the definition of interlocks has changed. A primary interlock exists between companies I and II if a director of I, whose main business is with I, sits on the board of II. If the same person also sits on the board of III a primary interlock exists between I and III, creating social closure (Heemskerk, 2007) if not, there is opportunity for brokerage (Burt, 1983; Heemskerk, 2007). These two relations, however, necessarily involve an interlock between II and III, and this we call a secondary interlock (Sweezy, 1953; Koenig et al. 1979; Useem, 1980; Granovetter, 1985l; Pettigrew, 1992; Mizruchi, 1996; Heinze, 2004; Kentor and Jang, 2004; Heemskerk and Fennema, 2009). However, the former is accredited with more weight than the latter (Sweezy, 1953). Directors sitting on more than one board (bridges) can prosper from their brokerage function (Burt, 1983; 1992). Combining all the interlocking directorships and their bridging functions on the boards of different firms shows a network of interlocking firms which is defined as an interlocking directorate network.

Furthermore it is of great importance to differentiate between primary interlocks and primary lines. Primary lines are interlocks created when an officer of one corporation sits on the board of executives of another firm. Secondary lines on the other hand, are interlocks created when an outside director of one company serves as an outside director of a second company, the actor in this case only serves as a supervisor. Subsequently we need to distinguish between thin – single directorship - and thick – multiple directorship – lines (Caroll and Fennema, 2002; Kentor and Jang, 2004).

(25)

25 When a director functions as a bridge, broker or spans a structural hole (Burt, 1992; 2005) centrality can be used to measure his influence. Betweenness of centrality is used when the bridge is the shortest connection between two other directors. A high betweenness centrality has a bridging function in which the personal influence of the actor increases (Wasserman and Faust, 1994). Hence, interlock centrality can be perceived as a indicator of general influence of both corporations and persons (Helmers, 1975; Mariolis, 1975; Mintz and Schwartz, 1985; Burt; 1992; Wasserman and Faust, 1994).

Bearden(1975) followed by mentioning three factors on which centrality of corporation depends. First of all the number of firms with which it interlocks. Secondly, the intensity of the interlocks and third, the centrality of the firms with which it interlocks.

A corporate and individual perspective on interlocks

The rationale for interlocking directorates has been the subject of many researchers over time (Dooley, 1969; Bearden et al. 1975; Helmers; 1975; Koenig et al 1979; Fennema and Schijf; 1978; Burt; 1979;1983;1992; Galaskiewicz; 1979; Pennings, 1980; Mizruchi, 1982, 1996; Useem, 1984; Mintz and Schwartz, 1985; Zajac, 1988; Nooteboom, 1999; Windolf, 2002; Caroll and Fennema, 2002; Graz, 2003; Kentor and Jang, 2004;) There is widespread discussion on interlocking functionality. Proponents of interlocking directorates claim that interdependence and interlocking have an interactive effect on organizational effectiveness (Pennings, 1980) which is disputed by Maclean(2006).

(26)

26 1975; Heemskerk, 2007). Therefore Koenig, Gogel and Sonquist(1978) captured the interlocking functions in three different models denominated the financial control model; the class hegemony model and the management model.

The first model – the finance control model – is derived from early Marxism and states that corporations have given up their independence from financial institutions due to the growing necessity of financial funding. Since the early interlocking studies of Jeidels(1905); Hilferding(1909) and Wibaut(1913), the centrality of banks and bankers, has always been a major issue in interlocking research (Fennema and Schijf, 1978; Fennema, 1982; Heemskerk, 2007). Subsequently several researchers proved that financial firms have been disproportionately active in interlocking (Dooley, 1969; Mariolis, 1975; Pennings, 1980).

Both banks and insurance companies gained power because of their control over portfolios and funding. These banks and insurance companies used this financial power to arrange reciprocity agreements with other corporations. Another argument is the possibility for financial institutions to attract new investments and to control these. Hence, banks reduce risks by gaining financial power and monitoring their investments and corporations gain extensive access to funding in order to cope with the global requirements of capital (Dooley,1969). However, financial interlocking became less apparent in the mid-1990s and by 2001 banking centrality has even come to an end. Only five percent of all non-financial companies had a financial executive on their board despite the growing importance of financial expertise within those boards (Heemskerk,2007). Somehow corporations found another way to gain access to this expertise without corporate bank interlocks.

(27)

27 three boards are called big linkers (Fennema, 1982; Useem, 1984; Windolf, 1996; Heemskerk and Fennema, 2009). Big linkers can be seen as highly influential. Through co-optation they elect directors from their midst sustaining their power and influence. Playing a crucial role in social cohesion the interlocking network (Heemskerk and Fennema, 2009) Burt(1983, 1992; 2005) already dubbed them the bridges between boards. They provide order and coordinate whereas their primary function lies with communications rather than direct control (Barrat Brown, 1973; Heemskerk, 2007). Big linkers earn their position by reputation, known as preferential attachment (Barabasí, 2001). When two big linkers meet opportunities arise for creating consensus within the corporate community. As they do not represent one single companies the possibilities of monitoring and coordination emerge and the big linkers tend to function as the elders of the corporate landscape (Fennema, 1982).

The management model – dubbed as the reciprocity model – sees interlocks as partial alliances between the interlocking corporations assisting coordination, monitoring and trust. This model opposes the functioning of interlocking directorates as it claims that directors themselves are powerless to run corporations (Koenig et al, 1978). It vindicates that decisions are made by the management team and not by the supervisory board.

Other critics have claimed that interlocks have little significance in the world, both in number and meaning (Mace, 1971; Zajac, 1988,1992; Davis, 1992). Zajac protests that individuals join boards for financial compensation, reputation and connects useful in their own advantage rather than in the desire to link corporations (Zajac, 1988). However, as Mizruchi contests, it is not fair to claim that interlocks do not influence strategic choices made by the firm as there is much evidence they do (Fligstein and Brantley, 1992). Hence, researchers do agree there is a reason for interlocking. However it depends on which underlying function they assign to the perspective regarding the interlocks.

(28)

28 perspective are intertwined and inextricable connected (Dronkers and Hillege, 1995). This thesis follows the social hegemony model as it studies the decline of the corporate network in the Netherlands over the past decade (Helmers, 1975; Stokman et al 1985; Caroll and Fennema, 2002; Heemskerk, 2007; van der Lee(2008). Stokman et al(1985) proved that a superiority of directors were selected from a small subset of persons, the old boys network, forming an elite class network (Zeitlin, 1974; Scott, 1997; Windolf, 2001). This subset forms an inner circle (Useem, 1984), in which those individuals that already have positions on multiple boards – the big linkers - are more likely to be selected again. By looking at the current composition of the boards of the Dutch AEX, AMX and AslX corporation this study tries to see whether the decline found in the studies of Helmers(1975); Heemskerk(2003) and van der Lee(2008) has persisted over time.

Dutch Regulations

(29)

(30)

30

IV. Method

Data collection

The population for this study includes all top directors, both executive and supervisory, of the Dutch AEX, AMX and AScX listed companies in January 2010. Information on the AEX and AMX listed companies has proved to be more abundant than the available information on AscX listed companies. The purpose of this study is to research whether the decline of the corporate network has persisted over time. Therefore we compare our results with the works of Helmers(1975); Heemskerk(2003) and van der Lee(2008). Looking at all 488 individuals residing on the boards of the 50 firms listed on either one of the indices, this study identifies a group of 57 big linkers. Corporations that are merely a foreign subsidiary like Logica are not accounted for in this dataset. Furthermore, corporations that have not been listed long enough are excluded as well. Thirdly, corporations that have been taken over or merged are not accounted for nor when they have been included with their mother company.

(31)

31 this reversed selection procedure ensures embodying all possible interlocking firms in the network.

Dataset

To effectively analyze the interlocking directorates network decline over the years this study aims to use different datasets in order to comprise a thorough image of the interlocking network. All dataset consists of information on each directors age; current positions within listed companies, positions on board of foundations are universities; former employers and the highest education they have attended. The directors current position on listed companies boards, foundations or universities is used to create a three layered network within UCINET VI. This program is used to calculate a person’s size of the network, his degree centrality closeness and betweenness and furthermore the density of the network. Looking at the entire network we use three datasets. The primary dataset contains all 488 network actors. The second dataset revolves around all active companies, selecting each single company as an actor in the network. The third and final dataset includes the big linkers in the network who act as bridges (Burt, 1983; 1992) connecting the network. This results in three datasets comprising 1010 different actors filling 1168 available positions; 59 listed companies and their equity value; 57 actors identified as big linkers as they have at least two or more executive or supervisory positions on a listed corporation. These 57 individuals link 45 of the 57 companies.

Table I

Datasets used in this study

# Dataset Individuals Positions Firms Supervisory

Positions Executive Positions 1 Basic Network 488 562 57 355 203 2 Extra activities 1010 1168 123 856 312 3 Big Linkers 57 132 45 117 15

4 All Big Linkers 165 477 140 425 52

5 Corporate Network 165 477 140 - -

Sample size

(32)

32 theirs to their names only, not even presenting their career path, age or nationality. In such cases other sources have been addressed, however, not all necessary information could be retrieved. Other issues also surfaced in the data gathering process. First of all, some of the corporations listed on the three indices that were used in earlier research (Helmers; 1975; Heemskerk, 2003; Van der Lee; 2008) do not exist because of delisting, acquisitions or bankruptcy. Smit International has been taken over by Boskalis Westminster earlier this year, but also ABN Amro that was acquired in 2007 by the Royal Bank of Scotland before it was nationalized by the Dutch Government. Furthermore Tele Atlas was acquired by TOMTOM in 2008 and Super de Boer – a Dutch retailer - was acquired by the Jumbo Holding Group. Schuitema was delisted as it separated its interests from Ahold in the fall of 2009 just as Univar and Hagemeyer earlier in 2008. The van der Moolen group was the only corporation to file for bankruptcy in 2009.

Secondly, there are also corporations that are not accounted for since they have are merely listed for financial or accounting reasons such as AMG and Logica. Third, there are companies that are too new to be included in the corporate interlocking network as they would merely function as outliers. These companies are Binck Bank and Crucell. Inevitably the absence of these companies will have influence on the results. However, they are inherent to the changes in an interlocking network.

(33)

33 the supervisory board of van Lanschot Bankiers includes three actors in the initial dataset making van Lanschot Bankiers of significance in the network.

Centrality measures

In order to study the importance of certain persons and corporations in the network this study will look into the prominence of those actors. Wasserman and Faust(1994) refer to Knoke and Burt(1983), Hubbel(1965) and Friedkin(1991) noting that prominence should be measured by looking not only at direct ties, but also at indirect paths involving intermediaries. Using the reverse selection process of which firms to be included, this prominence should be measured qualitatively. Prominence, can be distinguished in two types: centrality and prestige. Centrality focuses on involvement of the actor and its non-directional relationships. Hence, the most central and therefore most influential actors(firms or persons) have the most linkages to other actors. Centrality can be divided in three different measures, degree centrality, closeness centrality and betweenness centrality, which combined create a complete picture of an actor’s influence. Degree centrality is defined as the degree of the node, d(ni) where xij is the number of links from i

to j. C denotes a particular centrality measure which will be a specific function of ni.

Given in a formula this results in:

( ) ( ) ∑

It represents a measure of actor centrality as follows. Actors have multiple ties to other actors who have an advantage to others. Because of the multitude of their ties they are less dependent on other actors. Because of this, they have access to resources of the network as a whole. In this equation there are also brokers active. They play the part of the dealmaker, connecting actors all together (Freeman, 1979; Bonnacich, 1987; Hanneman and Riddle, 2005; Borgatti and Everett, 2006). Hence, degree centrality represents the actors centrality and its potential powers.

(34)

34 relationships prestige has been left out of the equation. The second measurement of centrality is closeness, which measures the closeness to other actors in the network. The basic idea is that an actor is central when it quickly interacts with all others (Wasserman and Faust, 1994). Hence, this stems from a communication and information perspective. Closeness presented in a formula results in the following:

(

) [∑

]

The third measure of centrality is called betweenness centrality. If an actor lies central between other actors geodesics, it implies he or she has a large betweenness centrality (Freeman, 1979; Hanneman and Riddle, 2005;). Ergo, the actor must be between many of the actors via their geodesics. In a formula this results in:

∑

Where

g

jk is the number of geodesics linking the two actors. Assuming all these

geodesics are equally likely to be chosen for the path, the probability of the communication using any one of them is simply 1/gjk. Further, considering the

probability a more distant actor is involved in the communication between the two. In that case gjk(ni) is the number of geodesics linking the two actors that contain actor i. This

results in

g

jk

(n

i

)/g

jk.. In the essence it means that if actor A wants to access Actor E he

(35)

35

Table II : Interlock Descriptive 1996-2010

Based on van der Lee(2008)

Individuals with Interlocking Positions

2 3 4 5 6 7 Interlocking Individuals Interlocked Firms Interlocking Positions

IP/indiv. Edges Edges/

Positions

Degree Betweenness Distance Compactness Density

(36)

36

V. Results

Interlocking from an individual perspective

(37)

37 due to expansion of the network and the decreased level of interlocking density declined over those years. Overall a decreasing degree; increasing betweenness; growing distance and demised density, together with the obvious tendencies in the number of interlocking directorates, interlocking individuals and interlocking firms led Heemskerk(2003,2007) and van der Lee(2008) to correctly conclude the decline of the interlocking corporate network.

Table III

Descriptive of interlocks 2007-2010

The primary dataset of this study involves all companies from the AEX, AMX and the AscX. These 57 companies and 488 individuals covering 562 positions show some interesting findings. First of all, according to this dataset, the number of interlocks has declined from 198 to 131 representing a decline of 34% over the past three years. So is the number of interlocking individuals and the number of interlocking companies. The former declined 36.7% to 57 interlocking directors and the latter also declined 34%. This decline is largely attributed to the sample size as the 96 firms in the dataset used by van der Lee(2008) is also 40% larger than in this primary dataset. To conclude that the decline has ceased, would be premature. To draw a set of conclusions on the influence of the corporate elite it is needed to investigate the centrality measures of the network.

Table IV

Descriptive of centrality measures

IP/indiv. Edges Edges/Positions Degree Betweenness Distance Compactness Density

(38)

38 Comparing the ratio of interlocking positions and the number of interlocking individuals i.e. the average number of interlocks per interlocking director slightly increased over the past three years. However, the number of edges(relations) declined. The rise in degree centrality reveals the cumulation of power within the corporate network. The decline in betweenness shows that the number of ties used by different actors to apply their influence has diminished. Thus, the average influence per actor has declined. With regard to the entire span of the network – the distance – it increased, following the same pattern of the past fifteen years. Density, as a final characteristic, also declined according to the tendency of the past decade. Nonetheless, this can be rationalized by the limited size of the network, only spanning the AEX, AMX and AscX, leaving other companies out of the consideration. Furthermore, the Tabaksblat code is now in effect for over six years, meaning all the corporations have had time to comply with the rules. All the re-appointments made before the establishment of the code had served their time and some of them had to resign according to the rules. As in 2009 the last supervisory position was vacated according to Tabaksblat code restrictions, 2010 is the first year where the entire code is in effect. This explains the decreased density as the IDN has had to cut off multiple interlocking directorates.

Interlocking using the reversed selection procedure

(39)

(40)

40 revitalizing the power of the corporate elite. In essence, they redirected the corporate landscape around the rules of the Tabaksblat code.

Taking an even closer look, stretching beyond mere corporate interlocks, this study reveals that corporate directors also meet at a more informal level. Gathering at the boards of smaller cultural, philanthropic or educational organizations, the corporate elite found a way to meet at a lower, less visible level in which they coordinate their influence on the entire network. Dataset three comprises dataset two and encompasses the aforementioned organizations such as Oranjefonds, Advisory board of INSEAD or the Koninklijk Concertgebouw in Amsterdam. Unfortunately Heemskerk(2003, 2007) and van der Lee(2008) did not gather such information, making it impossible to compare such data. Nonetheless, inclusion of such non-profit organizations, improves the likelihood of a thorough understanding of the interlocking directorate network.

Incorporating these charitable and advisory organizations in the network consequently alters the congruity of the network. By entailing more organizations the number of interlocks consistently rises by 24%. The center of gravity moves to the right, to the heavy interlocking individuals. Furthermore the interlocking positions increases accordingly. Especially to the limited size of these cultural and charitable boards the amount of interlocking positions in the entire network remains about 40% just as in the second dataset. However, it does mean an extensive growth set against the 25% found by the studies of van der Lee(2008). Again, the percentage of interlocked firms in the total network is biased in this dataset as it is a direct result of the reversed selection procedure.

(41)

41 fifteen years the corporate elite was thought to have lost its power and influence in Dutch business. Faced with the restrictions of the Tabaksblat code, the corporate elite set out to regain the power lost. By diversifying their supervisory activities to both stock exchange listed, local companies and smaller – cultural or charitable – organizations and foundations, the corporate elite found a way to regain their strength. Inevitably they had to share their power with more directors, resulting in a larger corporate elite. Having to deal with more nodes to pass through, limiting their own personal power, they are nonetheless able to exert their power once again.

Big Linkers

Focusing on the big linkers themselves will enable a comprehensive view on the individual perspective of this study. First we will elaborate on the primary dataset in which the network is limited to the stock listed corporations. Subsequently we will comment on the big linkers in the secondary dataset which includes the reversed selection procedure dataset.

Limited to the AEX, AMX and AscX – the basis of the network - there are only 57 Big Linkers, occupying 132 positions for 45 corporations. Of these 57 persons, there are only five(9%) women and only three foreigners(5,3%). The average of these 57 linkers is about 59 years old and almost half of them(49%) has been educated economically. Even more intricating is the fact that 36,8% of these big linkers has followed a second master as well. Even so, 16% has earned an MBA title. Ergo, we can conclude that the characteristics of big linkers differ with the other actors in the entire network. Fewer women, more economically educated and less global as found in the appendix.

(42)

42 number of interlocks. If closeness is regarded as the most influential characteristic, there will be an entirely other distribution of power. In this case, Frohn(0,392) would be the most powerful instead of van der Veer(0,383), ensued by Fröhlich(0,383) and former prime-minister of the Netherlands, Kok(0,369). With regard to this closeness, Cremers(0,364) is the 8th most influential actor in the network. Taking betweenness as a central measure, again there is another sequential line in influential actors. Cremers(0,079) is again the most influential directors. After which van der Veer(0,078) closely follows and lagged by Frohn(0,062) and Kok(0,058). Implying Cremers has a strong brokerage position. Unfortunately it remains intangible which director is the most influential. Taking a closer look at Cremers and van der Veer again we can differentiate between the companies they represent. The former represents Fugro, SBM offshore, Unibail-Rodamco and AMX company Vopak. The latter, van der Veer, is supervisor for Shell, ING, Philips and Unilever. Both men, four interlocking supervisory positions, different size. By comparing size of corporations we use equity as a measure of size. Whereas Cremers embodies 15.5 billion euros, van der Veer represent a much larger equity: 202.9 billion euros. To calculate an actor’s influence, we need to look at both the centrality measures – degree, closeness and betweenness – and the size of the company the actor represents. Combined in a formula this results in:

[ ] ∑

(43)

43

Table V

Directors Influence Dataset 1

It is striking that the most influential Big Linkers are so heavily interlocked. Due to the unavailability of annual reports of smaller companies and the lack of them for other non-profit organizations, it is not possible to include the size of those organizations in such an analysis. However, since their positions have been taken into account in dataset three it is possible to see which actors are most influential when embodying all organizations in the network. In this dataset there are 167 individual big linkers of which 28(17%) are foreigners. Within these 17% there are 10 different nationalities, Americans(35%) are most frequent. Women account for 10% of the big linkers. The average age in the dataset is again 59 years old. Once again, Jeroen van der Veer is touted the most powerful. Using this dataset reveals that seven new actors entry the top 20 influential directors. Remarkable that Elverding, Kist, Kok and Hessels, all belonging to van der Veer’s ego network in the primary dataset, drop out of the top 20.

Actor Degree Distance Betweenness Influence(Millon €)

1 J, van der Veer 0,113 0,383 0,078 684,84

2 Wijers 0,06 0,357 0,021 64,87

3 Storm 0,072 0,36 0,044 41,47

4 Cremers 0,088 0,364 0,079 39,37

5 Routs 0,07 0,369 0,054 38,59

6 Frohn 0,111 0,392 0,062 22,94

7 B, van der Veer 0,064 0,351 0,042 19,68

8 Elverding 0,043 0,307 0,026 13,84 9 Klaver 0,053 0,342 0,018 13,65 10 Kok 0,082 0,369 0,058 13,12 11 Hessels 0,053 0,342 0,026 9,40 12 Harris 0,057 0,358 0,033 9,01 13 Fröhlich 0,109 0,383 0,044 8,85 14 Kist 0,051 0,358 0,023 8,21 15 van Wijk 0,053 0,352 0,02 7,97 16 Levy 0,053 0,364 0,019 7,68 17 van Lede 0,055 0,338 0,018 6,67 18 Burgmans 0,053 0,348 0,011 5,41 19 van Oordt 0,053 0,348 0,024 5,25

(44)

44

Table VI

Big Linkers influence dataset 3

Actor Degree Distance Betweenness Influence(Millon €)

1 J, van der Veer 0,121 0,523 0,015 192,57

2 Storm 0,119 0,531 0,019 43,66

3 Wijers 0,072 0,478 0,004 19,85

4 B, van der Veer 0,090 0,502 0,014 13,20

5 van Oordt 0,123 0,530 0,016 12,38 6 Routs 0,092 0,498 0,007 8,87 7 Cremers 0,103 0,503 0,011 8,87 8 Treschow 0,119 0,522 0,010 7,49 9 Levy 0,086 0,510 0,008 7,35 10 Klaver 0,078 0,487 0,004 6,36 11 Zwartendijk 0,082 0,495 0,016 5,37 12 Dahan 0,113 0,523 0,015 4,82 13 van Oosten 0,138 0,538 0,024 4,36 14 van Lede 0,064 0,465 0,007 4,16 15 Vuursteen 0,092 0,511 0,009 3,72 16 van Wijk 0,070 0,490 0,005 3,67 17 Harris 0,072 0,490 0,007 3,30

18 Peter van Rossum 0,076 0,491 0,007 2,96

19 Burgmans 0,076 0,482 0,003 2,93

20 Gagey 0,094 0,497 0,010 2,53

Corporate Perspective

Reasoning from a corporate perspective gives another look at the corporate network. Arguing co-optation, trust, information mobility and other corporate reasons, researchers have regarded the corporations as distinctive actors in the network. In this case, the actors choose the interlocks. Therefore this analysis is limited to the stock listed companies and local organization, leaving out the non-profit organizations as they do not share any function from the corporate perspective.

(45)

45 consequence. AEGON is followed by TNT, which is also highly represented within the big linkers network.

Table VII Corporate interlocks

Degree Index Closeness Index Betweenness Index Overall Indexed 1 AEGON 0,168 1 AEGON 0,546 1 BAM 0,075 1 AEGON 0,999 2 TNT 0,168 1 TNT 0,525 0,979 AEGON 0,074 0,999 TNT 0,973 3 BAM 0,159 0,991 Unilever 0,521 0,975 TNT 0,069 0,994 BAM 0,958 4 Unilever 0,142 0,974 BAM 0,513 0,967

Wolters-Kluwer

0,05 0,975 Unilever 0,922

5 Wolters-Kluwer

0,133 0,965 Shell 0,505 0,959 Shell 0,049 0,974 Wolters-Kluwer 0,901 6 Shell 0,133 0,965 KPN 0,503 0,957 KPN 0,049 0,974 Shell 0,901 7 KPN 0,124 0,956 Wolters-Kluwer 0,497 0,951 DSM 0,048 0,973 KPN 0,891

8 ING 0,124 0,956 HAL 0,497 0,951 Unilever 0,046 0,971 DSM 0,875

From the corporate interlocking viewpoint the connectivity between the most influential actors and the most powerful organization is clear. The directors are pawns moved by the corporations in their great game of chess, co-opting, trusting and mobilizing information in order to set their opponent check-mate. Figure II and III represents a graphical overview of the corporate network in dataset 5

Figure III clearly shows the interlocking in the corporate network. Figure II represent the graphical overview of the interlocking within the three listed indices with 45 out of 57 companies interlocking representing 57 individuals. Figure III represents an overview of dataset 5, with 140 corporations interlocking through 165 individuals.

Financial Interlocks

(46)

46

Figure II.

(47)

47

Figure III.

Graphical overview of the corporate network in dataset 3

(48)

48 in the overview presented in table VIII, it shows that, although AEGON (0,99) , ING(0,864) and HAL(0,859) are highly represented in the corporate network, other parties are less influential. With a high degree(0,168), closeness(0,546) and betweenness(0,074) it holds a strong position in the interlocking directorate network in the Netherlands.

Table VIII

Financial Interlocking firms overview

# Firm Index Centralization Degree Closeness Betweenness

1 AEGON 0,999 0,168 0,546 0,074 9 ING 0,864 0,124 0,486 0,037 11 HAL Investments 0,859 0,115 0,497 0,029 29 Bank Nederlandse Gemeenten 0,790 0,088 0,459 0,016 44 Rabobank 0,738 0,062 0,423 0,016 50 ABN-AMRO 0,733 0,06 0,42 0,013 53 DNB 0,724 0,05 0,413 0,019 58 SNS REAAL 0,718 0,062 0,406 0,009 75 NPM Capital 0,690 0,044 0,391 0,007 77 NIBC Bank 0,689 0,035 0,403 0,002 87 KAS Bank 0,669 0,035 0,376 0,004

105 van Lanschot Bankiers 0,627 0,02 0,342 0,001

(49)

49 corporations. Hence, only 26 companies do not interlock directly or indirectly within ego networks to these financial institutions.

(50)

50

Figure IV.

(51)

51

VI. Conclusion and Discussion

Since Mertens’ outcry in 1968 the interlocking directorate network has been subject of the public debate. After decades of stability the corporate elite was confronted with a decline of their network. In the years to 1996, the multiplicity of the network declined – resulting in a less cohesive network. The following decade even the number of dyadic ties declined, rendering a decline of the corporate network. This study investigated the persistence of this decline. In such, the Dutch corporate landscape has proven to be an ideal country to study interlocking directorates as is closely connected to the Anglo-Saxon and LME model although it is perceived to advocate the Rhenish model, being a coordinated market economy. Whereas most researchers differentiate between an interfirm and an interpersonal perspective, this study addresses both. Proponents of the financial control (Dooley, 1969; Mariolis, 1975; Pennings, 1980) and the management model (Koenig et al.1978) embody the former as the latter is supported by the social-hegemony model in which the network is to be connected through individual actors instead of companies themselves. Combining both perspectives the studies of Helmers(1976); Heemskersk (2003, 2007) and van der Lee(2008) have been studied intensively as they all stress the decline of the corporate network.

(52)

52 interlocking non-profit organizations, rendering a larger network with more individuals and more interlocking directorates. Instead of merely selection the 250 biggest corporations – thereby overlooking the ever changing corporate landscape – this selection procedure effectively selected all relevant interlocking organizations, revealing both formal and informal meetings in which the corporate network is formed. Using this secondary dataset the findings contradict the primary results. The number of interlocking individuals increased to 164, interlocking companies to 140 and interlocking directorate positions to 471. As the entire population increased, the power has been divided by many, explaining the networks decreased density. However, centrality measures showed that network compactness increased and inter-network distance decreased. Conclusively, the network of interlocking directorates increased, sharing its power with more individuals explaining the decreased centrality with respect to the fifteen years prior. Subsequently, we can conclude that individual power differences have increased as they have more important brokerage positions to fulfil, increasing their influence in the network.

(53)

53 This study also attained a interfirm perspective, following the arguments of Dooley(1969) and Koenig et al(1978). Co-optation, information mobility, monitoring and financial control as main reasons for interlocking, they stressed that the firm was to be regarded the actor choosing the interlocks instead of individuals. Looking at the social network analysis of this study, connecting the companies to their indexed centrality measures – since the equity value of the organizations is somewhat ambiguous in this case – we find another classification of importance. Overall AEGON stands out as the most influential corporation in the network. With 9 out of 14 executives and supervisors regarded as big linkers of which 7 of them belong to the top 20 influential interlocking directors. Apparently interpersonal and interfirm influence are connected. Proving that corporations always use the advantages of their managers, it remains ambiguous why that directors is supervisor for explicitly that corporation, through a choice by the organization or through the social-hegemony model. Fact is, the influence of corporations is increased through their supervisors.

Studying the financial organizations within this interfirm perspective, regarded as the core of the corporate elite (Fennema and Schijf, 1978; Fennema, 1982; Heemskerk, 2007), social network analysis shows that AEGON – the most central - and ING – the ninth position in the classification - both have very strong positions in the interfirm network. Where most would expect ABN-Amro to have a dominant position, their absence is to be explained following the naturalization by the Dutch government after the take-over by Fortis in 2007. The appointment of government officials like Gerrit Zalm and the retreat of interlocking directors as Floris Deckers and Tom de Swaan, weakened ABN’s position in the network tremendously. Subsequently, looking at the difference between the financial organizations we can conclude that even here, the power has been dispersed. Where some have increased their influence, other have lacked to do so.

(54)

54 landscape in which most companies are connected. The corporate elite has integrated more outsiders in order to keep the network in tact at the cost of loss of individual influence.

Limitations

(55)

55

References

Albert, M. 1991., Capitalisme contre capitalisme. París: Éditions du Seuil.

Allen, M.P. 1974. The structure of interorganizational elite cooptation. American

Sociological Review 39:393-406.

Barabási, A.-L. 2001. Linked. Cambridge: Perseus.

Barrat Brown, M. 1973. The controllers of British industry. In: J. Urry & J. Wakeford, Power in Britain . London: Heinemann.

Bearden, J. 1975 The Nature and Extent of Bank Centrality in Corporate Networks. Paper presented at the 1975 meeting of the American Sociological Association, San Francisco.

Beekenkamp, G.G., Dronkers.J, and T. Dronkers-Vlaming 2002. President-Directeuren,

posities en patronen. Amsterdam: Aksant.

Belding, Th. C. 2004. Nobility and Stupidity: modeling the evolution of class endogamy in Blau, J. 1980 "When Weak Ties Are Structured." Unpublished manuscript, Department of Sociology, State University of New York, Albany.

Blair, M. 2002. ‘Post-Enron Reflections on Comparative Corporate Governance.’

Georgetown Law and Economics Research Paper.

Bonacich, P. 1987. Power and Centrality: A family of measures. American Journal of

Sociology, 92:1170-1182.

Borgatti, S.P. 2002. NetDraw: Graph Visualization Software. Harvard: Analytic Technologies.

Borgatti, S.P., Everett M.G. and L.C. Freeman. 2002. UCINET for Windows: Software for network analysis. Harvard Analytical Technologies.

Borgatti, S.P. and M.G. Everett 2006. A Graph-theoretic perspective on centrality

Elsevier, 3:1-19

Bourdieu, Pierre. 1973. "Cultural Reproduction and Social Reproduction." Pp. 71-112 in Knowledge, Education, and Cultural Change: Papers in the Sociology of Education, edited by Richard Brown. London: Tavistock .

Bourdieu, P. 1985. The Forms of Capital. Pp. 241-258 in: J. Richardson (ed.), Handbook

The Dutch interlocking directorate network: A persistent decline or a revival?