Board composition:

(1)

Board composition:

An empirical investigation of the aspects influencing board composition

by pension funds in The Netherlands

Author:

R. Krabbe

Supervisor:

Dr. T.B.J.M. Postma

Co-assessor:

Drs. D.B. Veltrop

Internal supervisor:

Drs. S. Baars

Montae Pensioen

July, 2010

Master thesis, Master of Science Business Administration,

specialization Small Business & Entrepreneurship

(2)

ABSTRACT

(3)

PREFACE

After four interesting years, this thesis is the last challenge for finishing the Master Business Administration, specialization Small Business & Entrepreneurship. I started this research in February 2010 and finished it in July 2010. I do not wish to ignore those who helped me in reaching the end of this Master thesis. Although many people were in some manner responsible for helping me during this journey, I would like to thank some people in particular.

First of all, I would like to thank all my colleagues at Montae for the inspiring and interesting moments during my internship, and especially, Margriet Adema who guided me through the context of pension funds. A special word of thank goes to Sander Baars, for his supervising and monitoring. He was a great motivator for me during my period in Rijswijk.

In addition, I would like to express my thanks to both supervisors at the University of Groningen. Drs. Veltrop, who guided me toward the data collection and methodology. And, special thanks goes to Dr. Postma for putting the text under close scrutiny and commenting critically, but moreover because I really appreciated his dedication to this master thesis.

Furthermore, I would like to thank my friends who helped me to improve the quality of this thesis with many useful comments. And last, but absolutely not least, I want to thank my family to provide me a solid basis for personal development, for their support and trust.

Enschede, July 2010

(4)

1. Introduction

Due to different corporate scandals at multinationals (e.g. WorldCom, Tyco, Enron) at the beginning of this century, extraordinary media attention arose in a public debate about corporate governance. These scandals had one thing in common; the organizations‟ directors failed to do what they were supposed to do, namely; to be informed; to be vigilant and to act in the best interest of their company. Therefore, in order to investigate and counteract these failures extensive empirical literature focused on the role of corporate boards of directors and its composition. However, there is still disagreement between the composition of board members and their influence on organizational performance.

As mentioned by van den Heuvel, van Gils and Voordeckers (2006), the board is classified as an element in the governance framework, which affects the outcomes of organizations. To recover the confidence of the society, investors and other stakeholders, uniform guidelines about „‟good corporate governance‟‟ around the World are developed (Aguilera and Cuervo-Cazurra, 2004). In line with these guidelines, the Dutch Labour Foundation (2005) developed the principles for good pension fund governance1, in which significant attention is paid to the role of board members. This is partly due to increasing complexity in board‟s tasks and responsibilities, but moreover the role of the board changed. It developed from a paternalistic trustee model, with an abundant fixation on their own and less information sharing (Boot, 2005) to a democratic trustee model in which competences, transparency and responsibility of the board play a central role (Frijns, Nijssen and Scholtens, 2010).

In academic corporate governance literature, some authors examined the contribution of the board as a group or a team (Gabrielsson and Winlund, 2000; Borch and Huse, 1993), whereas other authors studied the contribution of individual board members (Gabrielsson and Huse, 2005; Whisler, 1988). Both, individual board members and boards as a group could influence organization‟s strategic choices, which in the end affect organizational performance. These strategic choices are not only affected by structural board characteristics (e.g. size, structure, age) (Daily et al., 2003), but also by board processes (e.g. trust, conflict, cohesion, knowledge) (Forbes and Milliken, 1999).

1.1 Problem statement

In corporate governance literature, there has been a considerable interest in boards of directors. In particular, there is growing attention for the effective functioning of the board, also described as; board performance (e.g. Hermalin and Weisbach, 2001; Dalton et al., 1999). However, academic contributions are far removed from business practice (Huse, 2009). Boards should add value2 for organizations that goes beyond monitoring, gatekeeping, value protection for several shareholders and value distribution to certain stakeholders. Boards should create value throughout the whole value chain- as in relation to innovation, corporate entrepreneurship, resource allocation for example and thereby create corporate value (Huse, 2009) i.e. influence financial performance.

Pension fund boards3 have many tasks and functions, which make pension board functioning highly complex. This complexity is reflected in e.g. duties of care, conflict of interests and accountability. The key objective for pension fund boards is to serve the interests of the employer

1

In reference list: Stichting van de Arbeid (2005)

2

The impact of actual board behavior on corporate value creation, in which board task expectations are compared with actual board task performance (Huse, 2009, p35).

3

(6)

and beneficiaries (Maatman, 2007). Consequently, their governance and regulations are issues of public concern, which makes it essential for pension fund board members to add value to the pension fund (Clark, 2004). As mentioned by Daily et al. (2003) and Forbes and Milliken, (1999) both board characteristics and board processes could influence this value creation process. Many theorist and practitioners emphasize the importance of corporate governance practices in pension funds (Clark, 2004; Maatman, 2007). They mainly concentrate on trustees´ competence and expertise and the way how the board communicates with their stakeholders. However, no single study examined the issue of composition of boards in relation with pension fund performance so far. Therefore, the following research question is formulated:

‘’Is an optimal composition of the board of pension funds possible’’? In order to answer the research question, the following sub-questions are formulated:

1) Which determinants of the composition of the board are discussed in the literature? 2) Which moderating variables influence the relationship between the composition of the

board and the financial performance of pension funds?

3) Is there any relationship between the composition of pension fund boards and the performance of pension funds? If yes, which characteristics show statistical evidence that result in an improved financial performance for pension funds?

By formulating the research questions, the structure of the thesis is taken into account. Literature study is performed to answer the first two questions. The first question is about existing generic literature and determinants of the composition of boards. Question number two is pension fund specific. Question number three, inferred through the literature study is tested empirically in the context of pension funds.

1.2 Contribution

Mainstream board research focuses on the „usual suspects‟, i.e. number of board members, insider/ outsider ratio, age of board members (Finkelstein and Mooney, 2003), which we classify as board characteristics. Besides, several publications focus on actors, processes and decision making within the board, also defined as board behavior (Gabrielsson and Huse, 2004). However, academic contributors who investigated both board behavior and board characteristics are far removed from business practice (Huse, 2009). This thesis aims to contribute to the existing board composition literature by addressing both board characteristics and board processes in the context of Dutch pension funds. Due to increasing complexity and responsibility in this sector the role of pension fund boards becomes more complicated. Therefore there is a call for attention of the composition of board members in the context of pension funds (Maatman, 2007; Labour Foundation, 2005). This thesis addresses pension fund board composition and its relation to pension fund performance. We first discuss the literature and based on this discussion we build a conceptual model and indicate some hypotheses which we test empirically.

1.3 Thesis structure

(7)

generic boards and pension fund boards are discussed. The following chapter indicates several hypotheses.

(8)

2. Context

This thesis is written within the context of the Master Business Administration (BA), specialization Small Business & Entrepreneurship (SBE). Based on the requirements of the SBE-master, whereby a student should write his/ her thesis in the context of a small or medium sized enterprise (SME), this thesis is done for a fast growing SME, called Montae. However, due to several restrictions, the role of SME in this thesis is somewhat limited. Therefore, I explicitly discuss corporate governance issues from a SME perspective in appendix 13.4.

2.1 Montae

Advice and administrative support in the field of collective pensions are the main activities of Montae, which is established in 1997 by Mr. van Engelen. Currently, seven partners are managing the organization with an office in Rijswijk (The Netherlands). With more than 35 specialists in pension law, actuarial, asset management, communications, consulting and administrative support, Montae works for industry- and companywide pension funds, unions, employers and insurers. For every new project, the organization works with project teams. These project teams consist of several advisors of Montae, whom have their own expertise and together they try to reach the objectives of the project. The activities of Montae concentrate on: advising pension funds in legal issues, support pension funds to fulfill efficient and accurate administrative processes, and discuss strategic issues about asset management. However, one of the main-activities, and most interesting in this thesis, of Montae is administrative support, wherein governance is a central issue. Based on their experience and expertise, Montae advices pension funds on the design, installation and operation of the management- and supervision model (i.e. internal control, accountability body), and ensure that this model works in practice (www.montae.nl).

2.2 Instruments

(9)

Table 1: Board Functioning Index, with bold the components used in this thesis

2.3 Role author

(10)

3. Corporate governance

There is a considerable interest in corporate governance during the last decades. Why have collapses, such as Enron, Worldcom, Ahold occurred, and what might be done to prevent such collapses happening again? How can stakeholders‟ confidence be improved? In what way are organizations transparent? The answers to all these questions are linked to corporate governance. As mentioned above, emerging financial scandals will continue to ensure that there is a sharp focus on corporate governance issues, especially relating to transparency and disclosure, accountability and control, and board structure (Malin, 2007). In order to realize why corporate governance has become so important, it is crucial to understand what corporate governance actually is.

3.1 Corporate Governance definition

Corporate governance plays a crucial role in business and society, and therefore received an increasing attention in the last decades (Malin, 2007). The Cadbury Committee (1992, P14) defined corporate governance as: „‟corporate governance is concerned with both the shareholders and the internal aspects of the company, such as internal control, and the external aspects, such as organization‟s relationship with its shareholders‟‟. The Organization for Economic Co-operation and Development (OECD, 2004) provides a broader description and describes corporate governance as: „‟…a set of relationships between a company‟s board, its shareholders, and other stakeholders. It also provides the structure through which the objectives of the company are set, and the means of attaining those objectives, and monitoring performance, are determined‟‟. During this research we use the definition of Van Ees at al. (2009), because it explicitly includes stakeholders. This definition is seen as a mixture of the OECD and Cadbury definitions. Van Ees et al. (2009, P307) describes corporate governance as; „‟Corporate governance addresses the nature of interactions and relationships between the firm and its stakeholders in the process of decision making and control over firm resources‟‟.

3.2 The ideologies of Corporate Governance

Corporate governance can be seen as a struggle between different ideologies. The perspectives underlying the expansion of corporate governance, and the areas it encompasses, are drawn from a variety of disciplines i.e. finance, economics, organizational behavior, law (Malin, 2007). Research of corporate governance has been dominated by the agency theory, but currently it is taking various directions (Huse, 2005). Table 2 gives a summary of corporate governance theories that can be used to explain and describe corporate governance institutions. These theories are described below in more detail, but based on the pension fund context, we discuss only the (1) agency theory (2) stakeholder theory, and the (3) behavioral theory in relation to pension funds.

(11)

Table 2: Summary of theories affecting corporate governance development 3.2.1 Agency theory

(12)

However, the extent to control is not costless, since the controlling party also receives remuneration (Jensen and Meckling, 1976). In sum, agency theory focuses on the way how boards of directors may protect shareholders interests from opportunistic and self-serving managers (Jensen and Meckling, 1976; Fama and Jensen, 1983).

3.2.2 Stakeholder theory

Due to corporate scandals (e.g. Enron, Tyco, Worldcom) of recent years, an alternative trend in corporate governance received increasing attention as a result of the unveiling of the conflicts between stakeholders (Kochan, 2003). This trend is classified as a stakeholder perspective. Stakeholders are defined by Freeman (1984: 25) as; „‟all groups and individuals that can affect, or are affected by, the accomplishment of the business enterprise. This definition of Freeman (1984) is characterized as a broad view of stakeholders. To narrow this view, we use a study of Mitchell, Angle and Wood (1997). Their study concludes that managers should pay attention to stakeholders who have power in relation to the firm (i.e. possess valued resources), are deemed legitimate (i.e. are socially expected and accepted structure or behavior) and can muster urgency (i.e. time-sensitivity or criticality of stakeholder‟s claims).

Many stakeholders, like employees, customers, suppliers, and local societies suffer losses because of managers driven by the possibilities of creating private wealth through extreme increases in the market prices of their shares (Kochan, 2003). Therefore corporations were reminded of their corporate and social responsibility (CSR) and should be managed in the interest of all stakeholders instead of only focusing on the interest of shareholders (Laplume et al., 2008). An interesting development is put forward by Jensen (2002). Jensen mentioned that traditional stakeholder theory argues that the managers of an organization should take into account all stakeholders but, because the theorists refuse to say how the trade-off against the interests of each of these stakeholder groups might be made there are no defined measurable objectives. This leaves managers unaccountable for their actions. Jensen therefore advocates enlightened value maximization, which he says is identical to enlightened stakeholder theory: „Enlightened value maximization utilizes much of the structure of stakeholder theory but accepts maximization of the long-run organizational value as the criterion for making the requisite trade-offs between its stakeholders… and therefore unravel the difficulty that arise from multiple objectives that accompany traditional stakeholder theory‟‟. By balancing the interest of all stakeholders over time in the best interest of the organization, (Freeman, 1984) organizational performance could improve (Clarkson, 1995 and Hillman and Keim, 2001).

3.2.3 Stewardship theory

(13)

stewards is collective, because the steward seeks to achieve the objectives of the company (e.g. market growth, profitability). In turn this behavior will benefit the principal. In sum, it can be stated that even when the interests of the steward and the principal are not aligned, the steward places higher value on cooperation than defection (Davis, Schoorman, Donaldson, 1997).

3.2.4 Resource based view

The resource-based view (RBV) argues that firms possess resources, a subset of which enable them to achieve competitive advantage, and a subset of those that lead to superior long-term performance. If there is a productive use of firm resources which are valuable, rare and appropriable, this results in short term competitive advantage (Barney, 1991; Wernerfelt, 1984). To hold the potential of sustainable competitive advantage a firm‟s resources must have four attributes: (1) it must be valuable, in the sense that it sees opportunities and/ or neutralizes threats in a firm‟s environment, (2) it must be rare among a firm‟s current and potential competition, (3) it must be imperfectly imitable, and (4) there cannot be strategically equivalent substitutes for this resource that are valuable, but neither are rare or imperfectly mobile (Barney, 1991; Barney and Wright, 1998).

The resource based view of organizations yields an economic foundation for examining the role of human resources for organizational competitive advantage. There are three basic types of resources that provide this organizational competitive advantage. The first one is defined as physical capital resources, the second one as organizational capital resources and finally the human capital resources (Barney, 1991). We will focus on the latter. Many scholars have applied the RBV to understanding the human resource role in organizations, because of its recognition of the potential for human assets of organizations to realize competitive advantage (Barney, 1991; Wright, McMahan, and McWilliams, 1994; Lado and Wilson, 1994). Most of the time, there is a focus on the characteristics of a firm‟s human resources, including knowledge, experience, skills, and commitment of the employees and their relationship with each other and with those outside the firm (Lado and Wilson 1994; Barney and Wright, 1998). The role of the board is mainly comprised to notice and implement the correct resources in the organization. A failure to implement the needed resources in a particular situation can lead to not fully realize the possible benefits for organizations (Barney, Wright, Ketchen, 2001).

3.2.5 Resource dependency theory

(14)

3.2.6 Behavior Perspective

A growing number of studies have emphasized the need to more closely study behavioral processes and dynamics in and around the boardroom to understand conditions for effective corporate governance (e.g. Huse, 2007; Forbes and Milliken, 1999; Gabrielsson and Huse, 2004; Van Ees et al., 2009; Finkelstein and Mooney, 2003). The importance of studying board behavior is reflected in the increasing attention of scholars to understand what boards really do. Whereas in previous decades boards could be characterized as essentially formal and passive institutions that rarely came under media attention (Forbes and Milliken, 1999), the actions of boards today are closely monitored (Judge and Reinhardt, 1997). More evidence of interest in board behavior can be seen in the increased level of legal scrutiny to which boards are subjected. Furthermore, there is growing competitiveness of the market for corporate control (Kesner and Johnson, 1990). In the behavioral perspective, the purpose of the board is to enable cooperation (Forbes and Milliken, 1999). However, this takes place not only by solving conflicts among coalitions of stakeholders and exerting control, but also by solving cooperation and coordination problems. Furthermore, information and knowledge gathering and sharing are seen as important objectives for boards. As such, the contribution of the board to firm performance is expected to obtain primarily from its involvement in enabling cooperation, as well as collecting and using relevant and timely knowledge, rather than from reducing agency costs (Van Ees et al., 2009).

3.3 Theories in context

The approach taken in this thesis is to assume a pension fund form. Therefore, the theories discussed above should be viewed in the light of pension funds. We discuss, of course, the agency theory which dominated the corporate governance literature. Furthermore, we discuss the stakeholder theory and the behavioral theory in the context of pension funds. Based on the characteristics of pension funds, we will not discuss the stewardship theory, resource based view and resource dependency theory.

3.3.1 Agency theory and pension funds

(15)

3.3.2 Stakeholder theory and pension funds

Balancing the interest of all stakeholders is a challenging objective for pension fund boards (Clark, 2004). Pension fund stakeholders are defined by the Labour foundation (2005) as: all participants, sleepers, pensioners and financially involved employer(s). As mentioned by Maatman (2007), the interests of these stakeholders do not always run parallel. For example, this is reflected in the urgency claim of Mitchell et al. (1997), whereby the time-sensitivity or criticality of a pensioner‟s claim differs from that of a participant‟s claim. Pensioners are more focusing on the short-term objectives of the pension fund, while participants are more interested in the long-term objectives of the pension fund. According to Laplume et al. (2008), who say that organizations should be managed in the interest of all stakeholders, pension funds try to manage this by representing a variety of constituencies (stakeholders) on the pension fund‟s board (Maatman, 2007).

3.3.3 Behavioral theory and pension funds

The contribution of the board to company performance is expected to originate from its involvement in enabling cooperation, collecting and using knowledge within the board. Today, this is closely monitored. For pension funds, these issues (i.e. board functioning, board processes) are especially reflected in the Pension Act (2007)4 and laid down in the „‟Principles for good pension fund governance‟‟. Board behavior, wherein the board should take action in order to the stakeholders of the fund, is one of the main topics discussed in these principles, represented by issues as; i.e. good board functioning and communication. The board should enable cooperation, through which knowledge and information is shared within the board, but also communicated to external parties (Frijns et al., 2010; Pension Act, 2007; Maatman, 2007). To monitor and control this board behavior pension funds are obliged to fulfill internal control. Also two separate external pension supervisors supervise pension funds. The Authority Financial Markets (AFM) controls the pension fund mainly on their communication and board behavior. Where, the Dutch Central Bank (DNB) focuses on the financial position of the pension fund.

3.3.4 Summary of the theories and pension funds

Based on the remarkable characteristics of pension funds, not all theoretical corporate governance contributions could function as a platform for pension funds. First of all, the characteristics of the agency theory, based on a shareholder perspective, show an enormous gap with the characteristics of pension funds. The stakeholder theory mainly comprises the pension fund situation, whereby balancing the objectives of the stakeholders is one of the main objectives for pension fund boards. Therefore, the stakeholder perspective is aligned with the pension fund situation and could be a good reference point together with a behavioral theory, which comprises cooperation and communication of the pension fund board.

4

(16)

4. Theoretical framework

In corporate governance literature, there has been a considerable interest in boards of directors. In particular, there is growing attention for the effective functioning of the board, also described as; board performance (e.g. Hermalin and Weisbach, 2001; Dalton et al., 1999). However, existing academic contributions are far removed from business practice. It is mentioned by Huse (2009) that boards could create value for organizations that goes beyond monitoring, value protection for several shareholders and value distribution to certain stakeholders. By using the knowledge and skills of the board members, boards contribute to e.g. strategy development, networking, innovation (Forbes and Milliken, 1999). Many scholars implicitly deal with the board of directors as a black box between the input and output of boards. These scholars are criticized because they ignore both a behavioral (Finkelstein and Mooney, 2003; Forbes and Milliken, 1999) and a contingency (Aguilera and Jackson, 2003; Davis and Useem, 2002) perspective. Input-output studies mainly focus on the so called „usual suspects‟, i.e. number of board members, insider/ outsider ratio, experience and age of board members (Finkelstein and Mooney, 2003), which we define as board characteristics. Behavioral studies on the contrary are characterized by actors, processes and decision making. Studies that are characterized by an emphasis on the context and seeing the board as an open system are reflected in the contingency perspective (Gabrielsson and Huse, 2004). Evolutionary studies combine both the contingency and behavioral perspective and include organizational history, time, change and learning (Pettigrew, 1992). In this thesis, we address both board characteristics and board processes in relation to organizational performance. The first one is reflected in the input-output approach, the latter in the behavioral approach.

4.1 Input-output approach

(17)

4.2 Behavioral approach

Gabrielsson and Huse (2004) argue that to thoroughly understand board behavior, multiple theoretical perspectives are required. Stewardship theory assumes that managers in general should be contemplated as good stewards. This theory stimulates board members tasks i.e. sharing information, collaborate and mentoring. Furthermore, the stewardship theory promotes board members to be involved by strategy formation and strategy implementation phases (Hillman and Dalziel, 2003; Shen, 2003). Agency theory proponents, restrict board behavior through a controlling and monitoring role, with the objective to avoid managerial opportunism (Huse, 2005). However, behavioral science provides insight in the relationship between organizational performance and board behavior (Forbes and Milliken, 1999; Gabrielsson and Winlund, 2000). Currently, most research treats the actual work of a board as a black box, assuming that board behavior and conduct of board members can be inferred from the demographic characteristics of the board (Daily, Dalton and Canella, 2003; Gabrielsson and Huse, 2004). Bridging the gap between the theory and actual board performance, requires an understanding of what makes a board function, i.e. open the black box (Huse, 2005).

4.3 Board role

Effective functioning of the board (board performance) is associated with several board roles, as mentioned in behavioral corporate governance (Zahra and Pearce, 1989, Van Ees, Gabrielsson and Huse, 2009). Behavioral processes, shaped by board characteristics, are suggested to clarify board performance (Van Ees et al., 2008). Organizational boards should fulfill two main roles that could, as mentioned in behavioral corporate governance literature, affect the organizational performance. These two roles are defined as: the service role and the monitoring role (Goodstein et al., 1994; Zahra and Pearce, 1989, Van den Heuvel et al., 2006).

The service role is grounded in two dissimilar theoretical frameworks, and we split up this role into two sub-roles; the network role which is grounded in the RDT and the advisory role, including the strategy role, is grounded in the RBV. The RDT argues that board performance is affected by linking the organization with the environment and networking to other firms, also described as board relational capital (Hillman and Dalziel, 2003; Zahra and Filatotchev, 2004). Networking activities of board members might secure a stable delivery of resources (Van Ees et al., 2008) and therefore positively influence board performance. From a RBV perspective it can be argued that the expertise of board members (board‟s human capital) might constitute a source of competitive advantage for organizations (Hillman and Dalziel, 2003; Zahra and Filatotchev, 2004). In this role, board members advice the organization. Moreover, the RBV discuss the board members‟ contributions to the strategy of organizations. This strategy role is defined by Zahra and Pearce (1989: p302) as „‟the board‟s involvement in and contribution to the articulation of the firm‟s mission, the development of the firm‟s strategy, and the setting of guidelines for implementation and effective control of the chosen strategy‟‟. This strategy role can have a significant effect on organizational performance.

(18)

5. Pension

Three pillars form the basis of the Dutch pension system: the state pension (AOW), the supplementary collective pension and the individual pension products that each person can arrange individually. Together these pillars determine the amount of pension a person receives after retirement (OPF, 2009). In this thesis the focus lies on the second pillar, namely the supplementary collective pension.

The pension is a periodic payment which eases the financial consequences of the loss of income associated with the attainment of a certain age, incapacity or death. It is an employment condition that is specified in a pension agreement. This agreement is a tripartite contract between the employer, the employee and the pension fund (see

figure 1). A distinction is made between two types of pension arrangements. The first arrangement contains the promise of a pension related to the employee‟s income and length of service. This arrangement is called „‟defined benefit‟‟. The second arrangement called „‟defined contribution‟‟ guarantees an employer‟s contribution which can be used to achieve a pension.

Figure 1: Tripartite relationship

The financial capital which serves to guarantee the obligations of pensions is collected in a pension fund (Maatman, 2007), which is „‟a legal structure in which assets are collected to provide a group of people a pension‟‟ (Alsma and Boshuizen, 2003). On the basis of the pension agreement a fiduciary legal relationship originates between the pension fund and its stakeholders. Fiduciary relationships arise “when one party (the „fiduciary‟) acts on behalf of another party (the „beneficiary‟) while exercising discretion with respect to critical resources belonging to the beneficiary”. In this relationship, there is separation between the legal entitlement (ownership) and the economic interests. The capital belongs to the pension fund, however the economic risk lies with the stakeholders. These stakeholders entrust the capital to the pension fund, with the belief that the pension fund uses this capital in accordance with the agreement (Clark, 2004; Maatman, 2004). The Labour Foundation (2005) describes participants, sleepers5, pensioners and the financially involved employer(s) (Art. 5.4 PSW) as the stakeholders who are affected by the achievement of the pension fund objectives.

The legal entity of pension funds is a foundation, which is legally and financially independent from another organization. If a company gets into financial difficulties, it will not affect the pension fund. Pension funds are financed from the members‟ contributions scheme paid in earlier periods and from the return on investments of these contributions. In the Netherlands, pension funds can be divided into three different categories (Maatman, 2007);

 Corporate pension funds (OPF), affiliated to a single organization

 Industry-wide pension funds (BPF), for a whole sector, such as rubber industry, metal industry

 Pension funds dedicated to independent professionals such as dentists, lawyers

5

(19)

5.1 Pension fund board role

Pension funds are remarkable institutions, which is reflected in the board‟s role. On the one hand, pension funds have substantial responsibilities for the living of their beneficiaries, on the other hand pension fund board members quite often fulfill their board role part-time (Bovenberg and Maatman, 2007). The core mission of pension funds is to manage the stakeholders‟ financial capital to generate potential income for these stakeholders (Maatman, 2007). Therefore it is essential for the pension fund board to act in the best interest of their stakeholders, in which transparence is key.

As we know, there are various kinds of pension funds across the Netherlands. However, broadly defined, the main responsibilities of pension fund boards are for every pension fund similar. These are first, the objectives of the fund, second the strategy and policy, and third the core activities of the pension fund (Bovenberg and Maatman, 2007, Clark, 2004).

The mandate largely requires the pension fund board to perform three functions (although most functions can be out-sourced to industry service providers), namely administrative services, the information provision, and asset-management (Clark, 2004; Maatman, 2004). The first task includes the collection, tagging and protection of pension contributors. These administrative procedures are dominated by the valuation and distribution of benefits to retirees. Information provision includes i.e. feedback about the performance of the pension fund, the risk of the investment mix and the enforcement of procedures concerning the interests of plan sponsors and beneficiaries. The third pension fund task includes asset-liability matching, asset allocation, the evaluation of investment product providers and advice for participants on their investment opportunities (Clark, 2004; Maatman, 2004).

Many authors classify the role pension fund boards fulfill as complex and complicated (Maatman, 2007; Clark, 2004; Koedijk and Slager, 2007). This is reflected in i.e. increasing complexity by the Pension Act, increasing requirements set out in legislation on financial supervision and a cry for more transparency by society (Dautzenberg and de Jong, 2010) Due to the complexity and complicated role boards perform, boards can decide to outsource certain tasks to some extent. If pension funds outsource some of its tasks to i.e. investment funds, administrative offices, then they are described by Maatman (2007) as „‟separately operating pension fund‟‟. A major disadvantage of outsourcing is that it gives rise to opposing interests, agency effects and agency costs, which can be difficult to control (Jensen and Meckling, 1976; Zahra and Pearce, 1989). In contrast, third parties generally have the experience and expertise to manage the outsource tasks. If the pension fund executes the mandate entrusted to it (and the corresponding tasks) itself, we speak of a self-administering pension fund (Maatman, 2007).

5.2 Board composition

(20)

funds and industry-wide pension funds. However, literature argues that both need to possess at least two financial experts in the board and furthermore countervailing power6 (Maatman, 2007).

Occupational pension fund

The board of an occupational pension fund consists of representatives of employers and employees. Besides, representatives of pensioners could be board members. In this situation, pensioners obtain one seat as a replacement of an employee representative. In the board, representatives of employees and representatives of pensioners, together equals at least the number of employer representatives. Only if it is set by law, it is possible that there are more employee representatives in the board than employer representatives. There are no age-limitations for the representatives. The board can choose a chairman, who is already part of the board, and has voting rights. Or, the board can choose for an independent chairman without voting rights (Pension Act, 2007).

Industry-wide pension fund

The composition of an industry-wide pension fund consists of representatives of employers and employees. Both have equal seats within the board. If industry-wide pension funds set guidelines that other representatives could also be member of the board (for example representatives of pensioners), it is at the expense of an employee seat. There are no age-limitations for the representatives. All board members have equal voting rights. The board can choose a chairman, who is already part of the board, and has voting rights. Or, the board can choose for an independent chairman without voting rights (Pension Act, 2007).

5.3 Pension fund Governance

Pension fund governance in the Netherlands explicitly discuss board composition. As mentioned before, a trust based relationship between stakeholders should be build (Labour foundation, 2005). Referring to van den Brink (2010) and van Rijssen (2009), the role of pension fund boards and their social environment changed in several years from a „‟trust me relationship‟‟ to „‟involvement of the stakeholders‟‟ (see figure

2). In more detail, the role of pension fund boards changed from a relationship wherein the board only informs the stakeholders to a relationship with strong involvement of the stakeholders. In the latter, there is a sharp focus on transparency and disclosure.

Figure 2: Changing social environment of pension funds

6

(21)

Good corporate governance in organizations provides the basis for control of business operations. In order to protect all pension fund stakeholders, an adequate system of corporate governance must be arranged (DNB, 2010).This is reflected in the „‟principles for good pension fund governance‟‟. These principles have been drawn up in 2005 by the Labour Foundation and laid down in the Pension Act. The principles for good pension fund governance focus on: good board functioning, internal monitoring, responsibility, competency, transparency and communication, and involvement of the stakeholders (Labour foundation, 2005). To prove that the board is ‟in control‟ on these issues, both an internal control mechanism (i.e. audit committee, visitation body) and an external control mechanism (DNB and AFM) is developed (Frijns et al., 2010).

In the following paragraph, the governance model of pension funds (see figure 3) gives an overview of the structure of pension funds. This figure gives insight in the composition of the board, but also in the advisory and monitoring bodies.

(22)

5.3.1 Council of members

It depends if it is required to have a council of members7. It depends on the fact if a pension fund is an Industry-wide pension fund (BPF) or a corporate pension fund (OPF). A BPF is required to have a council body with an equal representation of participants, pensioners and employer. Also sleepers could be part of the board. Corporate pension funds could choose if it installs a council of members. Based on the preferences of the board, this could be done. If 5 percent or more of participants, pensioners, sleepers and employer ask for this body, they should install a council of members. The council of members advises the board of the pension fund on several parts. The board should inform the board of the council of members as soon as possible if they follow the advice. If not, they should define the reason why they deviate from the advice.

5.3.2 Accountability body

The accountability body8, which is instituted by law, consists of an equal representation of participants, pensioners and employers. Furthermore, this body could involve representatives of sleepers. The accountability body has various similarities with the tasks of the council of members. If there is also an active council of members, the board could choose to integrate the council of members into the accountability body, supplemented with employer representatives. The accountability body should monitor if the board performed the policy and the strategic choices in favor of all stakeholders. Furthermore, the accountability body has an advisory role.

5.3.3 Internal control mechanism

Since 2008 are pension funds obliged to have internal control. Internal control consist of critical assessment of the performance of a pension fund, the way the fund is governed, its policy and governance procedures and processes. Furthermore internal control assess the approach taken by the board vis-à-vis longer-term risks. There are four possibilities to develop such internal control: visitation body, one-tier board (these supervisors are also board members), audit committee or an executive board. In general, only the visitation body is chosen by pension funds, due to rather limited restrictions.

The visitation body is an independent monitoring body with at least three independent experts, chosen by board members. This commission should review the pension fund at least once in every three years. This review consists of the long-term risks, the functioning of the board, and the procedures and processes (www.goedbestuur.nl).

5.3.4 External Control mechanism

Two separate external pension supervisors supervise pension funds. In the current political, financial, economical and social environment, managing a pension fund is quite complicated. The Pension Act distinguishes between prudential supervision, material supervision and conduct of business supervision. DNB is responsible for prudential and material supervision. Prudential supervision relates to supervision focusing on the financial solidity of pension funds and contributing to the financial stability of the pension funds sector. The material supervision contributes primarily to the content of the administration agreement between the employer and the provider and the content of the pension agreement. AFM is responsible for reviewing conduct

7

(23)

of business supervision and actual board behavior, such as external communication (Frijns et al., 2010; Pension Act, 2007; DNB, 2010).

Dutch Central Bank (DNB)

Monitoring by the Dutch Central Bank (DNB) shifted the last years from ‟rule based‟ to ‟principle based‟. Pension funds should not only comply with the legal act, but moreover they should be familiar with the rules. The DNB monitors pension funds on their financial situation, but besides that on their tripartite collaboration.

Pension funds should meet statutory financial requirements. The financial assessment framework (FTK) is the part of the Pension Act that lays down these requirements. It is build up around the principles of risk-based financial requirements, market valuation and transparency. Risk-based financial requirements increase and decrease in line with the fund‟s exposure to risk. Market valuation is determined by discounting expected future cash flows against the current nominal term structure of interest rates. More transparency should trigger the organization to show a clear and objective view of the fund‟s financial position (DNB, 2007; Labour Foundation, 2005).

By introducing the Pension Act, the government clarified the division of responsibilities in the tripartite relationship between employee, employer and pension administrator. The Pension Act draws up pension scheme rules on the basis of a pension agreement between these parties. The DNB controls whether the pension scheme rules of a pension fund comply with this Pension Act, and other relevant legislation. Normally, every pension fund is assessed once every three years. For large funds, the DNB checks the rules annually. This check mainly consists of a comprehensive assessment of the existing pension scheme rules (DNB, 2007; Labour Foundation, 2005). The Pension Act describes that the DNB should control the pension fund by monitoring if the pension fund act in the interest of all stakeholders, whereby the stakeholders feel that they are represented in an effective way.

Dutch authority for the financial markets (AFM)

Under the new Pension Act, the Dutch authority for the financial markets (AFM) has become the conduct of business supervisor for pension funds implementing pension schemes. The supervision carried out by AFM primarily relates to external communications from the pension provider to employees who are acquiring or have acquired pension rights. Pension funds need to share this information on time, understandable, clear and correct (AFM, 2010). Furthermore, the AFM controls if board behavior serve the legal requirements like sharing timely, understandable, clear and correct information by the pension fund to their stakeholders. Besides, the AFM controls if the pension fund informs their stakeholders about investment risks (Pension Act, 2007; DNB, 2010; www.afm.nl). In sum, pension funds should act in favor of their shareholders. The AFM does not control board processes, they only control the behavior to their stakeholders.

5.4 Differences board role pension fund boards and generic boards

(24)

oversee the board. Therefore these are classified as non-executives. The difference between pension fund board structure and ‟normal‟ boards lies in the appointment of board members. In generic unitary boards, shareholders appoint board members at the organization‟s annual general meeting. In a dual board system, shareholders elect the members of the supervisory board, whilst the supervisory board installs the members of the management board (Malin, 2007). As mentioned before, board members of pension funds are appointed by participants and involved employees (Maatman, 2007; Labour Foundation, 2005).

The stakeholder approach explains the composition of pension funds. The stakeholder approach describes that organizations should be run in the interest of all stakeholders instead of only shareholders. For pension fund boards it is the challenge to balance the interests of all stakeholders and maximize long-term financial profit. Contrary, the traditional approach of corporate governance centered at the shareholder-manager relationship finds its theoretical justification in a simplified representation of the firm according to which the shareholders are the exclusive owners of the organization. These shareholders encourage the board to generate a maximum rent, often on the short term (Charreaux and Desbrièresm, 2001). The situation for pension funds is completely different from this viewpoint, based on the fiduciary relationship between pension funds and their stakeholders. In „generic‟ organizations shareholders could easily sell their shares, however pension fund stakeholders have a long-term relationship with their pension fund and it‟s difficult to break this contract.

(25)

6. Hypotheses

6.1 Board size

There is a vast amount of publications concerning the size of the board and their impact on the corporate performance (Yermack, 1996; Zahra & Pearce, 1989; Eisenberg et al. 1998). Some observers are convinced that the performance of a board may decline as the board size increases above a moderate number, while others predict a positive relationship between performance and board size.

There are three corporate governance theories known in the literature. Agency theory proponents argue that as the consequence of a board size increase, decision making could delay and increase costs (Callen et al., 2003; O‟Regan and Oster, 2005; Yermack, 1996). Goodstein et al. (1994) conclude that this is the fact, because in a highly turbulent environment, large boards have limited effectiveness for strategic change. Furthermore, empirical research suggests that monitoring is more effective in smaller boards than in larger ones (Yermack, 1996 and Eisenberg et al., 1998). Contrary to the viewpoint of the agency theory, the resource dependency theory (RDT) argues that there is a positive association between board size and firm performance. With a mixture of perspectives and intellectual knowledge, boards are more able to link the organization to its external environment and secure critical resources (Pearce and Zahra, 1992, Dalton et al., 1999). The resource-based view (RBV) suggest that if these resources are unique and inimitable and competitors find it difficult to duplicate, sustainable competitive advantage could be achieved (Barney, 1991). This could lead to an improvement in the financial performance. Scholars (Zahra and Pearce, 1989; Eisenberg et al., 1998) suggest that the board-performance nexus is explicitly linked to the ability of the board to tap into significant resources that would flow from a larger rather than a smaller sized board. Furthermore, there is evidence to suggest that a larger board could decrease the domination of a CEO, which affects the financial performance of organizations (Forbes and Milliken, 1999; Goodstein et al., 1994). Based on the agency theory on the one hand and the RDT and RBV on the other hand, there is contradictory evidence between these theories and the relationship between board size and firm performance. It is reasonable to suppose that there is an optimum9 in this relationship. We want to investigate the possible existence of this optimum for pension funds.

The complexity of the board of pension funds is especially reflected in the interest of all stakeholders. According to O‟Regan and Oster (2005), board‟s size should reflect these different interests by representing „the society‟. So far, there is little evidence that board size for pension funds is a factor in determining overall firm performance (Maatman, 2007). The goal of section eight is to investigate this issue further.

We believe that the effectiveness of a pension fund board may decline as board size increase above a moderate number. Therefore, we expect an inverted U-shaped curvilinear relationship between firm performance and board size. Taking this background into account, the following hypothesis arise:

H1: There is an inverse U-shaped curve between board size and the performance of pension funds. That is, on the left of curve, the larger the board size is, the better the performance of the pension fund. On the right of curve, the larger the board size is, the worse the firm performance.

9

(26)

6.2 Board age

The Resource Based View suggest that resources owned by organizations can create a competitive advantage. In other words, resources can be used as a basis and as an aid to implement strategies that can create competitive advantage for organizations (Barney and Wright, 1998). Human capital can be seen as the most critical, scarce resource and the key to value creation, whereas human capital primarily focuses on capabilities, knowledge skills and experience (Nahapiet and Goshal, 1998). As age is still one of the criteria for organizations in selecting and retention board members (Cochran et al., 1984), a simple human capital argument would state that increased age would typically translate to more experience. This results in a higher level of human capital (Fisher and Govindarajan, 1992). Vance (1983) argues that older boards provide more experience in decision making, based on their history of earlier decisions and directions.

But it‟s too simple to say that a ‟grey board‟ achieve the best results. There is also the concept of diversity. In corporate governance, this concept relates to board composition and the varied combination of attributes, characteristics and expertise contributed by individual board members in relation to board process and decision-making (Van der Walt and Ingley, 2003). The various types of diversity that may be represented among directors in the boardroom include age and (life) experience (Milliken and Martin, 1996). Younger board members have different perspectives than older ones. Amongst other things this gives the organization entrance to new products, convenience and technology. Furthermore, diversity in age can lead to other networks (Gilpatrick, 2000).

The case for diversity in the board can be made on a number of grounds, including stakeholders (Keasey et al., 1997). Particularly for pension funds, this is important. Responsive for the welfare of beneficiaries, trustees of pension funds act on behalf of others. As mentioned by Fisher and Govindarajan (1992), increased age will in general lead to more experience, but it is also important to have age diversity. This suggests that the effectiveness of a pension fund board may decline as board age increase above a moderate number. On that account, an inverted U-shaped curvilinear relationship is expected between firm performance and board age. This gives rise to the following hypotheses.

H2: There is an inverse U-shaped curve between board age and the performance of pension funds. That is, on the left of curve, the older the board, the better the performance of the pension fund. On the right of curve, the older the board, the worse the firm performance.

6.3 Conflict

Conflict within boards can be divided in task related issues (cognitive conflict), relational issues (affective conflict) and process issues (Jehn and Mannix, 2001). Cognitive conflict refers to differences in opinions between team members about task related issues such as the objectives, strategy and procedures of the firm (Pelled et al., 1999). Affective conflict refers to an awareness of interpersonal incompatibilities, like annoyance and irritation. The third issue; process conflict is defined as the controversies about the way how tasks should be accomplished. This pertains to issues as duty and resource delegation, like who should do what, and how much responsibility should different people get (Jehn and Mannix, 2001). All these types of conflict have potentially opposing effects on board performance, as mentioned by Forbes and Milliken (1999).

(27)

environments (Eisenhardt et al., 1997). Due to the economic crisis and a changing role of pension fund boards (Bovenberg and Maatman, 2007), complexity and uncertainty are an integrated whole with pension funds.

Forbes and Milliken (1999) argue that if there is a very high ability of the board‟s members to continue working together, it may be detrimental to board functioning and could lead to groupthink. However, task conflict may increase the tendency of team members to scrutinize task issue and to engage in deep and deliberate processing of task-relevant information. Huse (2007) also stresses that cognitive conflict is characterized by creative thinking. Therefore the author suggests that high levels of task related conflict within the board of pension funds might increase the performance of the pension fund.

H3a: The performance of a pension fund will increase if the task related conflict between board members increase.

Contrary to the increasing board role performance by cognitive conflict, affective conflict disturbs the interpersonal attraction of board members and subsequently the board performance negatively (Forbes and Milliken, 1999). Affective conflict has been detrimental to board decisions and leads to a decrease in team performance. In particular, because there is a decrease in communication, cooperation and understanding between team members (Jehn and Mannix, 2001). Pension fund boards are characterized by board members who represent different adherents (Maatman, 2007) and were shared patterns of behavior is hard to obtain. For pension fund boards, it can be thought to develop familiarity, necessary for positive patterns of future interaction if there is a high level of affective conflict (Janis, 1982).

H3b: The performance of a pension fund will decrease if the affective conflict between board members increase.

Process conflicts have been the least examined aspect by scholars. But the logic proposed that when a group argues about who does what, board adherents are frustrated with the uncertainty caused by the process conflict. Then they feel a greater desire to leave the board (Jehn and Mannix, 2001). In another study, Jehn (1997) noted that process conflicts interfere with the quality of task content and could lead to a misdirected focus to not relevant discussions of member ability. Furthermore, the conflict will probably increase the rise of communication-, but also coordination costs among board members (Simons and Peterson, 2000).

H3c: The performance of a pension fund will decrease if the process conflict between board members increase.

6.4 Trust

(28)

of relational risk in the board. Sharing information and knowledge, and to be more active in board decision-making grows if board members feel a low risk perception and therefore more confident (Van Ees et al., 2008). This is in line with the situation for pensions fund boards (Maatman, 2007). Pension fund boards often consist of representatives with different backgrounds, whereby sharing information and knowledge is crucial. Due to an increasing complexity and a collective responsibility of pension fund boards, trust within the board could improve the quality of decision-making processes (Maatman, 2007). Therefore we postulate the following hypotheses:

H4: The performance of a pension fund will increase if trust between board members increase.

6.5 Moderator Cohesion

The interaction between board members in terms of conflict and trust, discussed above combined in one moderating construct is mentioned as cohesion (van Ees et al., 2008).. Forbes and Milliken (1999: 492) define cohesion as; „‟the board‟s ability to continue working together‟‟. This is reflected by the degree to which board members are motivated to stay on board and are attracted to one another. Because boards are charged with complex and interactive tasks, the degree of interpersonal attraction between board members is likely to influence the effectiveness with which those tasks are performed (Williams and O‟Reilly, 1998) Effective board functioning is influenced by two behavioral processes; trust and conflict. Both, the absence of affective conflict and the presence of trust may generate a climate that fosters its performance regarding monitoring, advice and networking roles alike (Van Ees et al., 2008).

In addition, board members must trust each others‟ judgment and expertise. As mentioned by Huse (2007), a trusting board is characterized by commitment, creativity and openness. This fosters board cohesion. As mentioned before, also an absence of cognitive conflict among board members fosters group cohesion (Forbes and Milliken, 1999). However, very high levels of cohesiveness are likely to prove detrimental to the quality of choices made by board. In addition, cohesiveness is the most well-known and frequently noted antecedent of „‟groupthink‟‟ (Mullen et al., 1994). Groupthink is characterized by a reduction in independent critical view and an extreme striving for unanimity between board members (Janis, 1983). Janis also stated that for most boards, „‟optimal functioning in decision making tasks may prove to be at a moderate level of cohesiveness‟‟.

(29)

7. Methodology

7.1 Method of Data collection

The secondary data are collected from different data sources. Data regarding the dependent variable pension fund performance is mainly collected from the pension funds‟ annual reports. These are obtained by using pension fund websites and Company.info, which is a database available at the University of Groningen. This database contains an extensive collection of publicly disclosed company specific information, like annual reports and board information. Data regarding the independent variables age, board size, conflict and trust is mainly collected from publicly disclosed information such as annual reports and Board Functioning Index database10. 7.1.1 Descriptive statistics

The composition of the pension fund board in relation to pension fund performance will be examined. The center of the empirical tests is the Board Functioning Index database. A drawback of this database is that it is not longitudinal and possesses for every pension fund board only information regarding one single year. The database started in the year 2006 and is still in progress. Annual reports, which are currently (April 2010) publicly disclosed until 2008 gives rise to test pension fund performance only for the year 2006-2008. That causes a limitation in the data. It could be possible that boards are tested that completed the BFI in 2009 with an average performance based on the years 2006-2008. After careful screening of the data, 69 turned out to be suitable for basis of this research.

7.2 Measures

All process variables, and their constructs, included in our conceptual model are based on statements in the questionnaire. Pension fund board members are asked for component trust and conflict to indicate the level to which they agree or disagree with the statement in these components. A 5-point Likert-type scale with the following responses (1); largely disagree (2); somewhat disagree (3); not disagree/ not agree (4); somewhat agree (5); largely agree measures these statements. The design of the questionnaire is focused on the effectiveness of pension fund boards.

7.2.1 Board characteristics

Data regarding the absolute number of the independent variable board size is simply measured by counting all board members of one single pension fund by using publicly disclosed information such as annual reports in the year 2008 and compare this with the Board Functioning Index database11. Data regarding the average age of board members, is calculated by combining all board members‟ ages and divided this by the total number of board members, in the year 2008.

10

Board Functioning Index is discussed in more detail in chapter 2.2

11

(30)

7.2.2 Board process

There are four board process variables in this thesis, which are: task related conflict, affective conflict, process conflict and trust. By using SPSS we tested these variables on their Cronbach Alpha (see appendix 13.3, table 9). The variable task related conflict is based on 2 items, Cronbach Alpha is 0.72. Affective conflict consists of 2 items, in which the Cronbach Alpha is 0.65. Furthermore, there are 2 items with respect to process conflict. The Cronbach Alpha for this construct is 0.76. The Cronbach Alpha concerning trust is 0.88, based on 5 items. All Alpha‟s are higher than the lower limits of acceptability (usually considered to be between 0.5 and 0.6)

7.2.3 Moderator variable

The effect of pension fund board cohesion is measured as a combination of trust and an average score of conflict (task related conflict, affective conflict and process conflict). Using median splits, each case was classified high/ low on trust and on conflict. A score of low conflict and high trust was classified a value of 2 for cohesion. A score of 0 was assigned for a combination of high conflict and low trust, and further combinations attained a score of 1. Notwithstanding, different problems with median splits, such as every value above or below the median, for example is considered equal and a loss of power, because it is simply harder to find effects that are really there, we use median splits. Since, it is the most appropriate method to test cohesion.

7.2.4 Performance variable

Existing research on pension fund performance can be based on different instruments (van Westen, 2010; de Haas, 2004). To provide a clear and objective representation of pension fund performance, different alternatives are compared and discussed. The first pension fund performance instrument which is used in different studies and which we will discuss is the share investment ratio. As mentioned by van Westen (2010), these investment results cannot give a comprehensive overview of pension fund performance. This is mainly due to the risk aversive objective. The objective is to pay periodically the pensions, and therefore take a balanced risk. Furthermore, there are more investment opportunities for pension funds (e.g. bonds, real estate) than only shares. Therefore, we think that conducting an empirical analysis based on investment in shares is not sufficient.

The Z-score is a method designed to show pension funds how they are performing in relation to a benchmark annually. The performance of one single pension fund is compared with the performance of the benchmark. The difference between these is divided by a factor, dependent on the investment mix (shares/ bonds ratio) However, there are numerous criticisms on this instrument. First of all, it only refers to a self chosen benchmark, with an emphasis on an absolute output. Secondly, the Z-score cannot be translated in an effective absolute performance, because pension funds in the benchmark could perform another investment policy (De Haas, 2004). Furthermore, the Z-score could stimulate pension funds to manage the fund passively, in which they use index investing and therefore have a relative restricted deviation of the benchmark (NPN.nl). Based on these restrictions, we conclude that the Z-score is not comprehensive enough to use during this research.

(31)

indexing in their annual report, which is publicly disclosed. Therefore, data regarding the dependent variable pension fund performance is mainly collected from the pension funds‟ annual reports. These are obtained by using pension fund websites and Company.info. As mentioned in part 7.1.1, the period of interest for indexing pension funds is 2006-2008. Pension fund performance is calculated by taking the mean of the values for the three years 2006-2008.

7.3 Data analysis

Board composition: