POWER AND PERFORMANCE: HOW DOES POWER INFLUENCE PERFORMANCE AND PERFORMANCE STRATEGY?

Faculty of Economics and Business

Masters International Economics and Business

POWER AND PERFORMANCE:

HOW DOES POWER INFLUENCE

PERFORMANCE AND PERFORMANCE

STRATEGY?

Supervisors: Drs. G. van der Laan Prof. Dr. H. van Ees

Keywords: Power, Firm Performance, Strategy, Outside Directors, CEOs, Shareholders

ABSTRACT

TABLE OF CONTENT

ABSTRACT 2

TABLE OF CONTENT & INDEX OF TABLES AND FIGURES 3

1 INTRODUCTION 4

2 AGENCY THEORY: THE PRINCIPAL-AGENT FRAMEWORK 8

2.1 Origins of Agency Theory 8

2.2 Fundaments and Implications of Agency Theory 10

2.3 Mechanisms to Mitigate Agency Costs 13

3 POWER 15

3.1 Interaction between Stakeholders 15

3.2 Power 23

3.3 Review of Literature on the Association between Power and Performance 26

3.4 Overview and Hypotheses 40

4 METHODS 48 4.1 Research Design 48 4.2 Sample 49 4.3 Measures 50 5 RESULTS 58 5.1 Data Analysis 58 5.2 Results 61

6 DISCUSSION & CONCLUSION 67

REFERENCES 74

APPENDICES 84

INDEX OF TABLES AND FIGURES

Table 1: Overview of Agency Theory 12

Table 2: Overview of Literature on Shareholders’ Power and Performance 28 Table 3: Overview of Literature on CEOs’/Insiders’ Power and Performance 31 Table 4: Overview of Literature on Outside Directors’ Power and Performance 35

Table 5: Overview of Variables 54

Table 6: The Sample Distribution over Industries 58

Table 7: Means, Standard Deviations, and Correlations for All Variables (Except Industry) 60

Table 8: OLS Regressions Estimates 61

POWER AND PERFORMANCE:

HOW DOES POWER INFLUENCE PERFORMANCE

AND PERFORMANCE STRATEGY?

WOUTER KOOL

Rijks Universiteit Groningen

1 INTRODUCTION

The interest in corporate governance increased in the beginning of the nineties (Zajac & Westphal, 1994). Corporate governance is defined as “the relationship among stakeholders in the process of decision making and control over firm resources” (Aguilera and Jackson, 2003: 450). Control refers to the allocation of the steps (initiation, ratification, implementation, & monitoring) of the decision process (Fama & Jensen, 1983). According to Finkelstein (1992) power plays a central role in decision making, particularly in strategic choice. Power in organizations is the capacity (or potential) to influence the behavior of a group or person in some intended fashion (Pfeffer, 1981). In this paper I analyze the relationship between the power of stakeholders and firm performance. I focus on stakeholders who have the foremost ability to influence the strategy of the firm (i.e. control one or more steps in the decision process): the shareholders, the CEO2_{, and the outside directors. The interaction between these stakeholders}

can be conceptualized as a principal-agent relationship. Jensen & Mecking (1976) define this relationship as “a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating some decision making authority to the agent. If both parties to the relationship are utility maximizers there is good reason to believe that the agent will not always act in the best interests of the principal.” (Jensen & Meckling, 1976: 308). This conflict of interest, central in agency theory,

can result in the pursuit of different performance strategies3 (e.g. share value maximization and sales increase), and, hence affect firm performance. Therefore, my particularly focus is on the impact of power on performance strategies. In this context power is the ability of a stakeholder to influence the strategic decision process which can affect the performance strategy and can be used to pursue interests.

Recent developments in business have further increased the attention for control and power in firms in both the academic literature and the popular press. First, corporate scandals (e.g. Enron and Worldcom in the United States (U.S.) and Ahold in the Netherlands) fuelled the concern that the power of the shareholders was too limited, and hence management was too powerful. In the period following these scandals a Dutch newspaper stated that “Americans are dealing with the same problems as the Europeans. These problems were bad functioning ‘outside directors’, too powerful CEO’s and inactive shareholders” (Het Financieel Dagblad, 2003)4.

The years following these scandals resulted in the adoption of new corporate governance codes in a wide range of countries5. The general objective of these codes was to enhance governance (transparency, monitoring, and accountability) of firms by empowering and encouraging shareholders and by emphasizing the role of outside directors. According to Walt & Ingley (2004), this has strengthened the position of the shareholder. In addition, these corporate problems could have been a trigger for increased shareholder activism. Paired with the rise of private equity funds and hedge funds, this helped a new phenomenon to surface: strategic shareholder intervention6_{. This is an intervention where a minority shareholder uses various tools}

to intervene in the strategic decision process of the firm. Thus, compared to traditional

3 _{The pursuit of interests can influence strategy in general. This can translate in various strategic focuses (e.g.} diversification). In this paper, my attention is directed towards performance strategies, i.e. strategies that are related to firm performance focuses.

4 _{Translated from: “Amerikanen worstelen met dezelfde problemen als Europeanen ten aanzien van de checks} and balances: slecht functionerende 'non-executives', oppermachtige ceo's en weinig actieve aandeelhouders.” (Het Financieele Dagblad, October 29, 2003, Speciale Bijlage, page 7).

5 _{E.g. in the U.S. (Sarbanes-Oxley, 2002), the United Kingdom (U.K.) (Higgs, 2003), and the Netherlands} (Tabaksblat, 2003), (http://www.ecgi.org/codes/all_codes.php).

shareholder intervention, smaller shareholders attempt to exert control over strategic issues7. The result of the increased shareholder activity is that the shareholder has become a more active player in determining the strategic focus and, thus, performance. The strengthened position of the shareholder combined with the trend of a more active shareholder has resulted in a second, opposite concern in the popular press and by other stakeholders (e.g. community, unions, and government): too powerful shareholders. An important concern is that when shareholders become too powerful, the focus in corporations will be on (short term) shareholder value alone, and will not take in to account the interests of other stakeholders. M. Tabaksblat, chair of the committee responsible for the Dutch code for good corporate governance, shares this concern. He argued that “For years it was the board of directors and management who were urged to pay more respect to the shareholders. Now, the shareholders themselves are under scrutiny.” (Het Financieele Dagblad, 2007)8.

This discussion concerning control and the balance of power in firms is all but resolved by the academic literature. Nevertheless, corporate governance issues related to control and power have received substantial attention (e.g. the role of power in board-CEO relations, tenure, and remuneration). However, the distribution of power in the upper echelons of organizations and its implications for control and firm performance is less well and often indirectly documented. These studies have predominantly focused on the relative power of single stakeholders. The majority of the attention was directed to insiders and CEOs (e.g. Eisenhardt & Bourgeois, 1988; Haleblian & Finkelstein, 1993; Hambrick, Cho, & Chen, 1996; Murray, 1989; Peterson, Smith, Martorana, & Owens, 1993; Pitcher and Smith, 2001; Smith, Houghton, Hood, and Ryman, 2006). The role of power of the outside directors has also been examined (e.g. Bebchuk, Fried, & Walker, 2002; Clark, 1998; Kiel and Nicholson, 2003; McNulty and Pettigrew, 1996; Ocasio, 1994; Peng 2004; Westphal and Zajac, 1995; Zajac and Westphal, 1996). The power of

7 _{E.g. shareholders can demand firms to split up (e.g. the Carl Icahn versus TimeWarner case), to merge (e.g.} the Kirk Kerkorian versus General Motors case), to focus on geographic areas (e.g. the TCI versus ABN-AMRO case) or to focus on functional areas (e.g. the Centaurus and Paulson versus Stork case).

shareholders, although central in the popular debate, is less well investigated (e.g. Bethel & Liebeskind, 1993; Demsetz & Lehn, 1985; Demsetz & Villalonga, 2005).

Although valuable insights have been provided by these studies, the results presented have often been inconclusive and conflicting. Hence, the role of power in the modern firm remains ambiguous. Finkelstein and Hambrick (1996) and Carpenter, Geletkanycz, & Sanders (2004) underwrite this notion by highlighting that the possible effect of the power distribution needs more attention. A limitation of the previous studies is the focus on the (relative) power of individual stakeholders. Additionally, the focus in the power literature is almost exclusively on performance measured in terms reflecting shareholder value. However, since different stakeholders may maximize their utility through different strategies, the conflict of interests (central in the agent-principal relationships) can result in a desire to focus on different strategies by different stakeholders. Hence, the potential implications of power for other strategies, and consequently an important element of behavioral and social interactions between stakeholders, are almost entirely neglected. Finkelstein (1992) underwrites this and states that the limited attention for the relationship between power and (performance) strategy “is surprising, given the importance of power relationships to strategic choice.” (Finkelstein, 1992: 505). In this paper I address the inconclusiveness of prior research (adding the inclusion of multiple stakeholders in a single model) and attempt to fill the briefly described research gap concerning the association between power and strategy. Hence, I attend to the following problem statement: How does the distribution of power between the shareholders, the CEO, and the outside directors influence performance and performance strategy?

power is irrelevant in agency theory (Fama, 1980; Jensen & Meckling, 1976) I focus on the behavioral processes and interactions in the framework separately. Thus, agency theory explains the (theoretical) structural relationships between the stakeholders. Subsequently, in reality various behavioral aspects affect these relationships and results in interactions which can result in a divergence from theory. Power is one of these aspects and is introduced in section 3. I explain its theorized dimensions and sources. Additionally, I attend to the interactions between the stakeholders. Subsequently, in this context I discuss the association between power and performance. First, I discuss the findings of previous studies on the relationship between power and performance. Next, I explore theoretical implications of power in the social and behavioral processes between stakeholders. Drawing on these inferences, I construct a theoretical framework to test hypotheses on the relationship between power and performance, and the relationship between power and performance strategy. I explain the methodology in section 4, followed by a presentation of the results in section 5. Finally, I conclude in section 6 with a summary and discussion. I propose an explanation for the ambiguous influence of power on performance put forward by both previous empirical studies and my results and present some recommendations for future research.

2 AGENY THEORY: THE PRINCIPAL-AGENT FRAMEWORK 2.1 Origins of Agency Theory

Originating in the work of Berle and Means (1932) and Coase (1937) agency theory forms a part of the research area often referred to as the ‘theory of the firm’. Agency theory has a wide range of applications in situations where a principal delegates a task to an agent (Eisenhardt, 1989). An important application of agency theory lies in the attempt to describe control in firms, and explain the alignment of interests of principal and agent (Mahoney, 2005). Previously, classical models had defined firms in terms of being in- and output entities with the sole goal of profit maximization. Berle and Means (1932), analyzed the changing role of the modern corporation in society after the Wall Street Crash of 19299. Their work clarified that in the

modern firm ownership and control is separated. In 1937, R. Coase laid down the fundaments for the transaction cost theory, which was an attempt to explain the existence of the firm as a result of a minimization of transaction costs through vertical integration of activities. His work and the insight that the firm is a ‘nexus of contracts’ would contribute to agency theory. Although the theory of the firm stems from the 1930s, major development took place in the past 4 decades and it developed in multiple theories10.

Prior to the development of agency theory, in the early sixties attempting to explain the rapid growing firms at that time, managerial and behavioral theories of the firms developed. Important contributions were provided by Penrose (1959), Baumol (1959), Marris (1964) and Williamson (1964). Managerial theorists argued that power had shifted from owners to managers and that instead of pursuing profit maximization managers would attempt to maximize their utility by engaging in sales maximizing strategies. It was this proposition of a deviation of interests that would later also contribute to the development of agency theory. Alongside the development of the managerial theories of the firm a behavioral approach to the theory of the firm developed. Stemming from Simon (1947), March and Simon (1958), and Cyert and March (1963), it attempted to describe how decisions were made in firms. They proposed that the firm cannot be treated as a single entity with a single goal, but merely as a group of individuals with their own, and possibly conflicting interests. Hence, firm strategy is the result of the weighted outcome of

century firms evolved in large, international operating, conglomerates that trade publicly in one or more global exchanges. Although small businesses still make up the larger part of the economy, large firms have crucial implications for changes in productivity, welfare and welfare distribution. The firm in the form of the limited liability joint stock company is arguably the best invention of the past one and a half century. The characteristics of the corporation enabled economic entities to conduct business on a very large scale and to gather vast amounts of financial resources. This enabled them to fully capture the possibilities offered by technological revolutions. At the same time, it was the cooperation that had the scale and scope advantages to fully exploit the productivity and welfare benefits from various technological revolutions, thus magnifying its impact. This and the fact that publicly owned firms extract large sums of money from the public, made the manner in which firms and the public interact with each other extremely important. This development of public ownership had created a separation between ownership and management. The conflict of interest between corporate insiders, such as managers and large controlling shareholders, and outsiders, such as dispersed shareholders, are important for understanding the modern public firm (Berle and Means, 1932; Shleifer and Vishny, 1997). This separation of ownership and control forms the basis of the corporate governance discussion.

these interests, and consequently a trade-off. I will return to these theories in section 3.4 when constructing hypotheses concerning power and performance strategies.

Partially relying on contributions made by these theories, agency theory developed in the 1970s. The early work on agency theory investigated the sharing of risk among groups and individuals in organizations (Arrow, 1971; Wilson, 1968). These studies argued that problems between principals and agents surface when these parties possess different views on risk. Authors such as Spence and Zeckhauser (1971) and Ross (1973) extended the focus to dilemmas of coping with incomplete information in insurance industry contracts. Different attitudes towards risk and information asymmetry are essential in agency theory and form the basis of agency problems. Agency theory matured and was generalized to other issues concerning governance in and around firms through the seminal contributions by Jensen & Meckling (1976), Fama (1980), Fama & Jensen (1983), and Eisenhardt (1985 & 1989). These studies established agency theory as a key theory on ownership and control. Therefore, agency theory is the underlying theory in many studies concerning corporate governance issues11.

2.2 Fundaments and Implications of Agency Theory12

Agency theory describes the problems that arise under conditions of information asymmetry

when a principal hires an agent (i.e. when ownership and control are separated13_{) and}

conceptualizes this relationship as a contract (Jensen & Meckling, 1976). The relationship between the principal and the agent is laid down in both written and unwritten contracts (Fama & Jensen, 1983). Consequently, in the view of agency theory the firm can be seen as a group of actors (i.e. stakeholders) who are tied through contracts. Hence, the firm is a ‘nexus of contracts’ (cf. Coase, 1967; Jensen & Meckling, 1976). These contracts set out the rules of the game, i.e.

11 _{However, alternate but related approaches to these corporate governance issues have also been put forward.} These are the (optimal) contract theory (Mirrlees, 1976; Grossman & Hart, 1983; Ross, 1973), stewardship theory (Donaldson, 1990; Donaldson & Davis, 1991), and resource dependency theory (Pfeffer, 1972; Pfeffer & Salancik, 1978; Zahra & Pearce 1989). Since I rely on an agency theory perspective, I will not discuss these theories in detail. However, in the discussion concluding this paper I will refer to their insights.

12 _{In this section I will discuss agency theory briefly and limit myself to presenting the key points. Although not} the most recent study, I refer to Eisenhardt (1989) for an excellent and extensive assessment of agency theory.

the rules of the principal-agent relationship; tasks, goals, evaluation criteria, and rewards (Fama & Jensen, 1983). Agents are expected to meet the principals’ goals (maximize the residual claim), for which they receive compensation. In the context of modern firms this implies that managers are required to maintain the firm as a going concern and achieve a reasonable flow of dividends and capital gain (i.e. shareholder value).

The main concern of agency theory is to “determine the most efficient contract governing the principal-agent relationship given assumptions about people (e.g. self-interest, bounded rationality, risk aversion), organizations (e.g. goal conflict among members), and information (e.g. information is a commodity that can be purchased).” (Eisenhardt, 1989: 58). Since agency theory’s basis lies in economics, agency theory assumes that stakeholders are (a) rational, (b) risk averse, and (c) utility maximizing (Grabke-Rundell & Gomez-Meija, 2002). These assumptions generate the central agency problem in the principal-agent framework: the conflict of interest between the principal and the agent.

TABLE 1

Agency Theory Overview (cf. Eisenhardt, 1989)

Key idea Principal-agent relationships should reflect efficient organization of information and risk-bearing costs

Unit of analysis Contract between principal and agent Human assumptions Self interest

Bounded rationality Risk aversion

Organizational assumptions Partial goal conflict among participants Efficiency as the effectiveness criterion

Information asymmetry between principal and agent Information assumption Information as a purchasable commodity

Contracting problem Contracting problem Agency (moral hazard and adverse selection)

Risk sharing

Problem domain Relationships in which the principal and agent have partly differing goals and risk preferences (e.g. compensation, regulation, leadership, impression management, whistle blowing, vertical integration, transfer pricing)

Efficient contracts will enable firms to be successful when they minimize agency costs, deliver value for the consumer of the product or service and maximize the claim to shareholders (Fama & Jensen, 1983). Additionally, agency theory addresses the allocation of the steps in the decision process, which is also specified through contracts. An adequate allocation of decision rights is essential to a firm’s survival (Fama & Jensen, 1983). They state that the four steps in the decision process are: “(1) Initiation (generation of proposals for resource utilization and structuring of contracts); (2) Ratification (choice of decision initiatives to be implemented); (3) Implementation (execution of ratified decisions); and (4) Monitoring (measurement of the performance of decision agents and implementation of rewards).” (Fama & Jensen, 1983: 303). The authors’ main hypothesis is that in firms ratification and monitoring are allocated to the principal and initiation and implementation are allocated to the agent. They add that in the publicly traded firm the principal’s decision rights are delegated to the board of directors. When the contracts are efficient the ‘nexus of contracts’ is superior to other organizational forms and markets (Kang & Sorensen, 1999). The authors add that separation of ownership and control is the superior organizational form because the outstanding potential access to capital outweighs the agency costs that arise from the separation of ownership and control. An overview of agency theory is provided in table 1.

2.3 Mechanisms to Mitigate Agency Costs

problems of collective action. The benefits of monitoring accrue to all shareholders. This creates the possibility of free riding. Moreover, since the costs of monitoring for the individual shareholder is likely to be larger than the benefits, dispersed shareholders are generally not likely to engage actively in monitoring. Therefore, large shareholders (i.e. block-holders) are potentially important in enforcing governance and reducing agency costs. Because these enjoy economies of scale in their efforts (or simply put, they have relatively more resources compared to smaller investors) and because they must make sufficient returns on their investments, they will be more active in monitoring. Therefore, some scholars argued that firms with concentrated ownership have lower agency costs, and as a consequence superior performance (Fama & Jensen, 1983; Jensen & Warner, 1988; Jensen 1993).

In addition to contracts, agency theorists have proposed three theoretical mechanisms that can reduce the costs resulting from divergence in interests, adverse selection and moral hazard. These mechanisms mitigate the harmful effects of different attitudes towards risk and information asymmetry. These are (1) structure of compensation, (2) the board of directors, and (3) the market for managerial labor. Firstly, if compensation solely consists of a fixed wage, this increases the possibility of shirking by managers. The reason is that their remuneration will be the same regardless of his quantity and quality of effort and hence performance (Eisenhardt, 1985). Therefore, Alchian and Demstez (1972) and Jensen (1983) argued that fixed compensation should be replaced by compensation based on ‘residual claims’ to firm profits. This would reduce the incentive to shirk because compensation would be related to the agent’s effort and performance. These ideas may have contributed to the rise of option based pay in the U.S. (and later other countries). In a sense this type of pay mitigates the effects of different attitudes towards risk through the convergence the amount of risk agent and principals bear.

board’s directors themselves are agents as well, with the obligation to control the other agent. Therefore similar agency problems can rise in the relationship between the owners and the board. The third mechanism to enforce the agency contract is the market for labor (i.e. management) (Fama, 1980). “A poor-performing agent faces the possibility of being replaced by other executives in the market, dealing with takeover threats, and risking the loss of pay, prestige, and a good reputation.” (Grabke-Rundell & Gomez-Meija, 2002: 6).

An additional mechanism to align interest for shareholders is to require firms to rely on debt instead of equity. Debt requires periodic interest payments and repayments. Therefore managers are obligated to generate cash and pay it out rather than use it to invest in new projects or perquisites dividends are easier to defer). Agency theorists suggest that these changes in ownership can improve the efficiency of firms (Jensen, 1989). The structural principal-agent framework according to agency theory is presented in Figure 1.

3 POWER

3.1 Interactions between Stakeholders

FIGURE 1

The Principal-Agent Relationship and Decision Rights in the Modern Firm

3.1.1 Shareholders – CEO

For a good description of the relationship between shareholders and the CEO/insiders I return to the work of Fama and Jensen (1983) and Jensen & Meckling (1976). The relationship between the shareholders and the executive directors is the standard principal-agent relationship in agency theory that assesses agency logics in large publicly held firms. Consequently, academic literature on this relationship is abundant. The principal is the shareholder, the owner of the firm. The shareholder hires the agent, management (in this paper the CEO), to run the firm in a way that satisfies the principal. In return for this effort the CEO is compensated by the agent. This compensation can be (1) fixed, (2) dependent on the CEO’s performance, or (3) a mixture of both. As presented in section 2.2 agency scholars advocate the use of the latter two (Alchian & Demstez, 1972; Jensen, 1983). The relationship between the shareholder and the CEO can be seen as a contract. Contracts are crafted such that the CEO pursues the goals of the shareholder. The costs that are associated with these contracts, with monitoring the CEO, and the costs that

result when executives do fold on their contracts and do not pursue the interests of the shareholders, are referred to as agency costs.

CEOs, the board of directors (including both inside and outside directors), and shareholders meet in the general meeting. At these meetings CEOs report on their operations and performance. Additionally, at these meetings shareholders vote on major strategic issues and the appointment of directors. Since the CEO mainly influences what is discussed at the general meeting, he can to some extent exert control over shareholders. Besides the standard general meetings, shareholders have the possibility to ask for a meeting. Additional actions that shareholders can take to influence the behavior of the CEO are scrutinizing the operations of the firm (and its management) in the popular press (e.g. TCI – ABN AMRO) or by threatening to sell their shares.

The relationship between the shareholders and the insiders is affected by the in section 2.2 presented three factors affecting principal-agent ties; differing attitudes towards risk, information asymmetry, and uncertainty. Firstly, risk results from the investment that a stakeholder in the firm has. When this stake is higher, the stakeholder’s risk is higher. Traditionally, insiders held little stakes in publicly traded firms, a consequence of the separation of ownership and control. This lower stake in the firm, and thus the lower risk of insiders, resulted in a situation that the interests of the shareholders and managers were not aligned. As a consequence of this, the past two decades experienced a sharp increase in equity pay of managers, thereby increasing their stake in the firm and, hence, aligning the interests. Jensen and Meckling (1976) argue that increased ownership does in fact align the interests of the manager and the shareholder. However, it can still result in differing attitudes towards risk because shareholders tend to be risk neutral (since they can diversify their risk in portfolios) and CEO tend to be risk avers (since their fortunes depends to a to greater extent on their increased stakes in the firm and, thus, its performance).

his profession, the manager often possesses the skills to adequately assess this information. The latter aspect of information asymmetry becomes more influential when ownership is less institutionalized and more dispersed. This is the case in the U.S. (Aguilera & Jackson, 2003). Referring to section 2.2, the result of this information asymmetry is that it (a) becomes difficult to monitor the CEO, and, (b) becomes increasingly difficult to set up contracts that depend on the outcome of the executive’s actions. Thus, information asymmetry increases agency costs and complicates the relationship between the CEO and the shareholder.

The third and final aspect that influences the shareholder-CEO relationship is uncertainty. Eisenhardt (1989) argues that when uncertainty increases, it becomes more difficult to craft outcome-based contracts. This makes it more complicated to align the interests of CEOs with those of shareholders, and consequently increases agency cost. Evidence for this effect was provided by Haleblian and Finkelstein (1993).

According to Aguilera and Jackson (2003), the relationship between the shareholder and the CEOs is affected by its institutional setting. Thus, the institutional setting of my research, the U.S., influences the relationship between these two stakeholders. Since the U.S. has well protected capital markets, these markets are characterized by the presence of many small investors, which contrasts to continental Europe where large institutional investors (e.g. pension funds), play a more prominent role. Since larger shareholder blocks often encompass more professionals than smaller, diffused blocks, information asymmetry plays a larger role in a U.S. setting. As demonstrated in section 2.2 this is the result of the fact that smaller investors have more problems obtaining and correctly assessing information. This effect is, however, mitigated by the fact that shareholders’ interests receive more attention in the U.S. than in Europe.

3.1.2 CEO – Outside Directors

work in the firm. Generally they are officers in other firms and work as an outside director in a number of other firms. For their effort they are often compensated per board meeting or through a lump sum per year. Ideally, to avoid bias in their expertise and decisions, the outside directors should not be inside or outside members of the board in competing firms.

According to Fama and Jensen (1983) outside directors in the board are used as a monitoring device of the CEO’s behavior. Thus, as CEOs are agents of the shareholder with the task to run the firm, outside directors are agents with the task to ensure that this other agent complies with the principal. Thus, actually the firm has one principal and two agents. Besides monitoring the CEO and ensuring that the CEO operates in the best interest of the shareholder, the task of the outside director consists of deciding on new directors and their compensation. A third responsibility is the ratification of management decisions, also to ensure that the shareholders’ interests are served well.

The CEO, the other inside directors, and the outside directors, interact in board meetings. It is legally required that notice is given to all board members and that a minimal percentage of the members is present for the decisions that are made to be legally valid. In most common-law countries, the power of the board is assigned not to individual but to the complete board. At these meetings the outside directors fulfill their two most important duties; monitoring the inside directors and advising them on major strategic issues. The insider directors, i.e. the CEO, have some practical control over the outside board members because they influence the information that outside directors receive and influence the agenda.

significant empirical attention, in which various performance measures have been used. Although some studies find a negative association (e.g. Pi and Timme, 1993; Berg and Smith, 1978; Rechner and Dalton, 1991), other studies argue that there is a positive relationship (e.g. Boyd, 1995; Brickley, Coles, and Jarrel, 1997; Daily and Dalton, 1992; Rechner and Dalton, 1989), contrary to the agency theory predictions.

In the academic field there has been an ongoing debate concerning the loyalty of the outside directors. The central dilemma is the direction of control between insiders and outsiders. Agency logics theorize that outside directors control inside directors and serve the shareholder. Yet, the outside directors do not automatically use their authority to advance shareholder interests. When the interests of individual outside directors are different than those of the shareholders, outside directors may take actions that benefit them at the cost of the shareholders (Borokhovich, Parrino, and Trapani, 1996). Even when the interests of outside directors and shareholders are aligned, the outside directors can still fail to comply when inside directors exert power over their outside counterparts.

Fama & Jensen (1983) argue that outside directors actually are an effective control mechanism, and thus a ‘shareholders’ tool’. The authors continue with the argument that outside directors, who likely are important decision-makers in other firms, have incentives to demonstrate to labor markets that they are experts in decision control by pursuing the goals of shareholders. Outside directors enhance their human capital by reinforcing their reputations as good decisions maker and thus being experts in finance, key fields of interest, and/or in management. Inside directors are, more concerned with maintaining their position and the possibility of extracting rent (Bebchuk et al., 2002). Weisbach (1988) argues that it is less likely that inside directors will challenge the CEO, with whom their careers are attached, than outside directors. This is an indirect argument that the outside directors exert control over management, since they are independent.

dismiss and compensate directors, issues that are, thus in fact controlled by the CEO. Zajac and Westphal (1996) indicate that this is an important factor that can enable management to control board and thus outside directors, and that use this influence to direct succession issues.

Theoretically and legally the appointment and dismissal of directors is voted upon by shareholders in the general meeting. However, sometimes directors have the opportunity to dismiss or appoint a director. In reality it might be very difficult for shareholders to remove a director. Often certain majorities are needed, the director has to be notified, which enables him to prepare a struggle against his dismissal, and finally, it can be very costly to remove directors (due to clauses in their contracts such as ‘golden parachutes’).

The dominance over the selection of new directors by CEO’s is also demonstrated by other authors (e.g. Pfeffer, 1972; Mace, 1986; Lorsch and MacIver, 1989; Zajac and Westphal, 1995 & 1996). Shivdasani and Yermack (1999) find that when CEOs sit in the nominating committee the likelihood of inside directors being selected over outside directors increases. This indirectly suggests control by the CEO over the board. Bebchuk et al. (2002) argues that inside directors possess enough control to influence the design of their compensation schemes. Borokhovich et al. (1996) highlight possible reasons for the reversed direction of control. They argue that “if CEO incentives are not aligned with those of the shareholders, they may nominate outside directors who are more inclined to support their decisions. Second, interlocking directorships may reduce the willingness of outside directors to challenge the CEO. Fearing reprisal, an outside director may decide not to challenge the CEO if the CEO is on the board of a firm at which the director is a senior executive. Finally, outside directors who are appointed because of their expertise in a narrow area may feel uncomfortable challenging the CEO on decisions outside their area of expertise.” (Borokhovich et al., 1996: 339).

there is an information asymmetry between the shareholder and the CEO, there exists a similar asymmetry between the inside directors and the outside directors.

3.1.3 Outside Directors – Shareholders

Evidence and theory on the relationship between outside directors and shareholders, is sparse when compared to the standard principal-agent relationship. Therefore, I am to some extent inclined to refer to logics on the other relationships and extrapolate on these. The relationship between outside directors and shareholders is the second principal- agent relationship surrounding the firm. The first was the relationship between the shareholders and the inside directors. The outside directors and the shareholders convene in the previously discussed general meeting.

As put forward in the previous section, theoretically the outside director serves as a tool of the shareholders to monitor and advise the inside directors and management, and through this, serving the shareholders interests. In practice however, they can become tools of management, used by them to enable managers to pursue their own interests. Various studies on different aspects have highlighted the beneficial effect of outside directors on their interests, suggesting that they actually do pursue the goals of the shareholders.

Firstly, Weisbach (1988), investigating CEO turnover, provides evidence that outside directors are more concerned with the shareholders goals than inside directors. Secondly, research on the stock price reactions on important decisions made by the board suggest that shareholders prefer decisions by boards that are dominated by outsiders to those made by boards dominated by a CEOs or inside directors (Byrd and Hickman, 1992; Brickley, Coles and Terry, 1994). According to Gilson (1990), and Kaplan and Reishaus (1990), the outsider’s status is associated with the value of his human capital. Therefore he will seek to maximize his reputation by establishing himself as a loyal servant of shareholders interests (and enhance the possibility of future board appointments).

(Zajac and Westphal, 1995 & 1996). This indirectly suggests that outside directors mainly contribute to shareholders by advising management on important topics.

3.2 Power

3.2.1 Defining Power

According to Pfeffer (1981) power in organizations is the capacity (or potential) to exert influence to change the behavior of a group or person in some intended fashion. The ability to change behavior of other parties, however, must be consistent in order to distinguish power from chance (March, 1966). Finkelstein (1992) and Daily and Johnson (1997) add that the underlying basis of power is the ability of a stakeholder to deal with uncertainty. This uncertainty can emanate from within the firm (e.g. top management and directors) and from outside of the firm (the firm’s task and institutional environment). Uncertainty plays an important role because it complicates the relationship between the shareholders and management (Haleblian & Finkelstein, 1993). When environments become more uncertain it becomes increasingly difficult to predict the outcome of business. When outcomes become unclear it is both more difficult to construct contracts and to monitor whether these contracts are lived up to.

In this study I concentrate on the role of power in the ability of stakeholders to control the decision process, through which stakeholders can pursue their own interests and goals. This definition is consistent with the work of Pfeffer (1981) and Finkelstein (1992). According to these authors, power plays a role in a wide variety of issues, but in this study I focus on its implications for performance and strategy. It is the decision process (and thus control) that can become the focus of a power struggle (Pfeffer 1981).

3.2.2 Sources and Dimensions of Power

disadvantage is that relying solely on this approach will result in loss of critical information concerning the entire selection of stakeholders (Smith et al., 2005). An example of a study adopting this approach is Eisenhardt and Bourgeois (1988). They use perceptions of team members. They gather these perceptions through interviews and process them into quantitative measures. This type of power results from personal interactions and is, thus, more socially constructed.

Secondly, power results from a structural basis (e.g. expertise and positions). This source of power is often assessed through objective power measurements, relying on secondary data. A disadvantage of this approach is that it neglects the social components referred to previously. Finkelstein and Haleblian (1993) and Daily and Johnson (1997) use objective power indicators (cf. Finkelstein 1992) to construct a power label for top management team members.

Smith et al., (2005) rely on a combination of perceptual and objective power measures. They argue that to truly measure power a combination of the approaches is necessary. However, because it is difficult and time consuming to adequately asses the social basis of power the majority of studies rely on objective measures of power.

Finkelstein (1992) is a seminal work on the sources of power relying on a structural approach. The study, which focuses on top management teams (TMT), identifies four main dimensions of power; (1) structural power, (2) ownership power, (3) expert power, and (4) prestige power. The author tests these dimensions by pairing them to perceptual measures that measures the same sources of power. The results demonstrated that three of these dimensions of power were related to perceptual measures for CEO power. Only expert power deviated form this finding. Multiple other studies rely on this model of power (e.g. Smith et al., 2005; Daily & Johnson, 1997). All these dimensions of power are related to a possible reduction in uncertainty stemming from within and without the firm. Although the study focuses on power within one stakeholder group, its logics are useful and can be extended to other stakeholders as well. A generalized version is depicted in figure 2.

FIGURE 2

Generalized Version of Finkelstein’s (1992) Model of Power

mation asymmetry as suggested by agency theory, I define it as power that is derived from the formal and organizational position that a stakeholder holds compared to other stakeholders and the firm and the degree of access to information.

Ownership power accrues to stakeholders in their capacity as agents acting on behalf of the shareholders (Finkelstein, 1992). Ownership gives rise to power directly by increasing the influence in voting, and indirectly by increasing the legitimate position of a stakeholder as the (partial) owner of the firm. Hence, ownership power increases the control a stakeholder has over the decision process. Ceteris paribus, actors with more ownership will be more powerful than actors with less ownership.

Expert power refers to “the ability of managers to deal with organizational contingencies and contribute to organizational success” (Finkelstein, 1992: 509). Referring to agency theory and information asymmetry, I extend this definition by defining expert power as power of a stakeholder that accrues from the ability to adequately assess available information and hence,

Structural Power

Ownership Power

Expert Power

Prestige Power

add value to firm operations. Expert stakeholders may have profound influence over strategic decisions (Finkelstein, 1992).

Prestige power is power that is derived from the personal status or prestige of a stakeholder. It is likely that when the prestige or status of a stakeholder increases, other stakeholders are less likely to resist, and therefore it will be easier to affect other’s behavior (Finkelstein, 1992). This will enhance the control of a stakeholder over decisions and thus the strategic focus.

3.3 Review of Literature on the Association between Power and Performance

Finkelstein’s (1992) types of power relate to the ability of influencing the decision process in firms. Hence, power reflects the extent in which stakeholders have the ability to control the strategy of a firm. In this paper I focus on the effects of power on performance through this potential influence on control in the decision process. This ability can result in both negative and positive associations between power and performance, depending on the stakeholder. E.g., concerning the impact of managerial power, it may imply a negative association because when managers are more powerful it will be more costly to monitor them (they have more influence to cover their actions since they control more processes) and to align their interests (residual losses can be larger because the manager can make more decisions that deviate from the decisions associated with the principal’s welfare) (Demsetz & Lehn, 1985; Demsetz & Villalonga, 2000). A negative association might also stem from the increased time spent on politics when stakeholders are more equally endowed with power (Eisenhardt & Bourgeois, 1988).

3.3.1 Shareholders’ Power and Performance14

In an extensive 2005 article the scholar L. A. Bebchuk elaborates extensively on the case for increasing shareholder power. In the article the author argues that increased shareholder power will improve corporate governance by mitigating agency issues. The author motivates that the limited control of shareholders over management is not justified by information-asymmetry or centralized management arguments. He concludes by arguing that shareholders should be enabled to make rules-of-the-game decisions. According to the author “These rules-of-the-game decisions should be defined broadly enough to allow shareholders to adopt charter provisions that would give them a specified power to adopt binding resolutions regarding business issues as game-ending or scaling-down decisions. The adoption of such provisions could considerably reduce agency costs and produce substantial benefits in many companies.” (Bebchuk, 2005: 913).

From these insights I conclude that an important impact of shareholder power on performance comes through the reduction of agency costs that detriment performance. But if more powerful shareholders will also positively affect performance by enhancing the operations of firms and the decision making remains unclear. Essential in adequately governing firms is the ability to appropriately assess information. This ability reduces the chance of erroneous decisions on a wide range of issues. Arguably, the ability to correctly assess information depends to degree of professionalism of the shareholders. Thus larger and institutional investors (e.g. banks and pension funds) would be beneficial for firm performance (Mitton, 2002). This logic is supported by agency theory. According to Shleifer and Vishny (1997), agency theory suggests that firm performance is positively associated to the attendance of large shareholders (i.e. blockholders).

Ownership concentration has been a frequently used measure that conceptualizes shareholder power and therefore I review literature on this issue. Already in 1932 Berle and Means advocated the proposition that the ownership distribution affects strategic decisions and efficiency in firms (and thus firm performance) (Berle & Means, 1932). This proposition was not undisputed. Some supported it (e.g. Galbraith, 1967; Pfeffer & Salancik, 1978) whereas others argued that the distribution of ownership is not associated with strategy and efficiency (e.g. Fama 1983; Jensen

& Meckling, 1976). The irrelevance of ownership stems from the idea that contracts and other discipline mechanisms, are sufficient to ensure that managers will seek to deliver value. According to Cubbing and Leech (1983) empirical studies have generated conflicting results as a result of the difficulty associated with the construction of valid measurements of stockownership distribution.

However, in the following years evidence was generated that supported the overall association between ownership distribution and performance, strategy and control (e.g. Bethel & Liebeskind, 1993; Hill & Snell, 1988 & 1989). Hill and Snell (1988 & 1989) argue that ownership structure affects the strategic decisions by motivating that when shareholders are more dominant (i.e. more concentrated), more value enhancing strategies will be adopted. Bethel & Liebeskind (1993) find a significant relationship between blockholder ownership and corporate restructuring. The authors conclude that this is an indirect evidence for the statement that many managers only restructure when they are forced to do so by large shareholders. In this conclusion I see support for the statement that control over management is positively related to the presence of large shareholders.

Despite the findings of these studies, the academic world remains, in addition to the inconclusiveness on the overall relevance of ownership for strategy and efficiency, inconclusive on a possible positive effect on firm performance. A positive effect will rise from reduced shirking by managers (e.g. Claessens & Djankov, 1999; Gorton & Schmid, 2000; Mitton, 2002), whereas others (e.g. Demsetz & Lehn, 1985; Demsetz & Villalonga, 2000 Lskavyan & Spatareanu, 2006) argue that there are also costs (e.g. reduced managerial initiatives and non-contractible investments) of high concentration ratios that harm performance. The presence of both positive and negative effects and translates in ambiguous results.

find a positive association between ownership and the firm profitability and labor productivity.

As previously noted other authors find contradicting results when empirically analyzing the relationship between ownership concentration and performance. Demsetz & Lehn (1985) discover no support for a significant relationship between ownership concentration and firm performance (measured as accounting profit). Additionally, Demsetz & Villalonga (2000) explore the relation between the ownership structure and the performance of firms if ownership is made endogenous. The authors fail to find a significant association between ownership structure and firm performance. Hence, the authors conclude diffused ownership, while it may cause agency costs, also generates advantages that generally compensate them. Finnaly, with a focus on monitoring in both developed and transitions economies Lskavyan & Spatareanu (2006) present evidence of an insignificant association between power (concentration) and performance.

3.3.2 CEO’s Power and Performance15

In this section I review work on the relationship between performance and the power of both the inside directors (or top management team) and the CEO. Although in this paper I focus on the CEO, I discus them simultaneously in this section since these are the inside stakeholders and they are strongly related. The latter is according to various studies (e.g. Harrison, Torres & Kukalis, 1988; Hosmer, 1982; Pearce & DeNisi, 1983; Pearce, & Robinson, 1987) the most powerful person in the firm, and, thus, often the focal point of analysis. In more recent years however, attention has shifted towards analysis of TMTs and powerful coalitions. Various characteristics of TMTs, inside directors, and CEOs and their relationship with various issues (e.g. remuneration, succession and performance) have been conducted and they have generated mixed results. Finkelstein and Hambrick (1996) assess 40 articles investigating various TMT characteristics and generate an extensive and valuable overview of the findings to that date. Particularly, heterogeneity in the background of the members of the TMT was under investigations (e.g. Hambrick et al., 1996; Murray, 1989). However, empirical work specifically on the relationship between TMT/CEO demographics and firm performance is not abundant. Moreover, work on the characteristic power and its association with performance is even less abundant (Haleblian & Finkelstein, 1993). In addition, several TMT research reviews (e.g. Carpenter et al., 2004; Daily & Johnson, 1997; Finkelstein, 1992; Finkelstein & Hambrick, 1996) argued that power deserves more attention in future research, since work on this particular characteristic is sparse and has generated conflicting conclusions. Because TMTs and CEOs both represent the inside component of the strategic stakeholders’ equation, when adopting an ‘inter-stakeholder’ approach and reviewing the literature on insiders, work on both TMT power and CEO power requires assessment.

Direct assessments of the relation have provided both evidence of a beneficial and detrimental influence of power on performance. Firstly, support for a detrimental influence is provided by Eisenhardt and Bourgeois (1988) who state that that politics in upper echelons are more frequent in settings with centralized power. Because the political games are associated with poor performance, they believe an unequal power distribution (i.e. the presence of stakeholder

dominance) is not advantageous for firm performance. Additionally, Haleblian and Finkelstein (1993) present evidence for a significant relationship between CEO dominance and firm performance in high-discretion environments, but not in low-discretion environments. Furthermore, they argue that CEO dominance has a negative effect on firms’ performance in turbulent environments and positive in stable environments.

On the contrary, other authors argue that power has a beneficial impact on firm performance. Daily and Johnson (1997) examined both the influence of power on performance and the reciprocal effect; the effect of performance on power. The main contribution of their study is that it exhibits the dynamic nature of the relationship between power and performance. Past performance as a source of power was also an important contribution of Ocasio (1994), who revealed is that CEO power is subject to circulation and shifts towards the board in cases of poor performance and vice versa. Also Smith et al. (2005) provide evidence on a positive connection between performance and power. They present evidence for the claim that a TMT is more strongly associated with good performance when it is dominated by an executive pair, especially when the pair has different world views. Generally, an unequal distribution of power among TMT members is associated with higher performance.

When performance is measured as performance variability, it depends on how variability is appreciated and hence whether the association between power and performance is perceived as beneficial or detrimental. Variability can both be perceived as positive (i.e. by ‘risk-lovers’) and negative (i.e. by risk avers people). Adams et al., (2005) use performance variability as their performance measure, and investigate its relationship with CEO power. They extend the logic that executives can affect the firm if they have influence over crucial decisions. Group thinking mitigates extreme opinions concerning business issues, i.e. a ‘diversification of opinions’, which results in lower variability in performance. Hence, when CEOs are more powerful, the diversification of opinions effect is mitigated, and variability in performance increases. Their results suggest that variability is positively associated with the presence of powerful CEOs.

indirect assessments also fail to provide consensus on the nature of the relationship between power and performance, partially a result of the fact that their studies were not specifically designed for this purpose. Nevertheless, they provide some interesting insights. For example, Peterson et al. (1993) found evidence that the personality of the CEO affects performance by affecting the dynamics in the TMT and can be, depending on the personality, both advantageous and disadvantageous (i.e. power serves as moderator. The impact of this personality increases as the power of the CEO increases, and thus the authors indirectly assess the impact of CEO power on performance. Finkelstein and D’Aveni (1994) find that CEO duality is less common when firm performance is high but remain unclear on the causality between the two. Pitcher and Smith (2001) find a confounding impact of power on processes in TMTs. Through impact on processes power, affect performance, but it remains unclear if this influence is beneficial or detrimental.

3.3.3 Outside Directors’ Power and Performance16

Work on outside directors in general is sparser than work on inside directors and shareholders. The available literature mainly focuses on how outside directors can mitigate agency costs, leaving issues such as power of outside directors and its effect on performance largely unattended. Especially the assessment of the relationship between performance and specifically the outside directors’ power is very scarce. The majority of my inferences on this association stems from extrapolations of conclusions on other issues (e.g. number of outside directors, insider-outsider ratio) that serve as implicit indicators of outside directors’ power. A suggestion has been put forward, namely that relying on single current theories can explain individual cases where directors influence performance, but this approach will not explain the general pattern of results (Kiel & Nicholson, 2007). In this study they review the hypothesized links between the board and performance that are predicted by three important theories in corporate governance; agency theory, stewardship theory and resource dependence theory. The authors conclude by suggesting that a more process-orientated approach is necessary to understand the link between boards and performance.

Direct assessments of the relationship between power of outside directors and firm performance that find that this relationship is positive are limited. Agency theory suggests that more powerful boards are better in reviewing management’s actions, and thus have a positive effect on firm performance. An important measure of board power (i.e. including both insiders and outsiders) has been board size. However, agency theorists additionally argue that there is an upper limit to the positive effect of the size of the board. From a certain size, negative downsides will dominate the positive effect. These downsides are the complicated group dynamics associated with larger boards (Jensen, 1993). Hence, theory suggests that there is an inverted-U shaped relationship between size and performance. Dalton, Daily, Johnson, and Ellstrand (1999) argue that it is not board size that is decisive in the potential impact of board characteristics, but outsider-insider ratio. Agency logics imply that a greater proportion of outside directors will be able to better monitor behavior by managers and by doing so will minimize the agency costs (Fama, 1980). Pearce and Zahra (1991) review an extensive set of studies on board-CEO

relations and derive a typology of their relative powers. The authors argue conclude that powerful boards are associated with superior firm performance for four reasons: (1) powerful boards provide useful business contacts; (2) powerful boards contribute more actively to the development of the firm’s goals; (3) powerful boards are needed for effective corporate governance (i.e. to reduce agency costs); and (4) powerful boards assist in creating a performance enhancing firm identity. Perry and Shivdasani (2005) argue that firms with a majority of outsiders in the board engage more easily in restructuring and that these firms experience a subsequent increase in their operating performance.

However, agency theory can also be extrapolated in such a manner that it supports the idea of a negative association between power and performance. When focusing on the importance of information, it can be argued that inside directors have better information and are therefore able to deliver more value than outside directors. This suggests that higher ratios of insiders on boards will enhance performance. Especially when integrated with stewardship theory, which argues that inside directors serve shareholders’ interests this view provided firm ground for this argument. Klein (1998) argues that is difficult to find a significant link between performance and the overall composition of the board. Hence, the author investigates the internal mechanisms of the board by looking into the role of board committees. He finds several relationships between the structure of the board and the performance of firms. Most notably is the relationship between the percentage of insiders on financial and investment committees and accounting and market based measures of performance. Thus, this article implicitly states that more powerful outsiders (i.e. a higher percentage of outside directors on committees) can be detrimental to performance.

extremely important for the adequate governance of firms. Kiel and Nicholson (2003) put on display that an implicit measure of outside director power, namely board size, is positively related to corporate performance. Besides their conclusions concerning board size, the authors argue that the ratio of inside directors is also positively related to performance. Since the insider-outsider ratio is a measure of power their work is inconclusive on the relationship between power of outside directors and firm performance. Peng, Buck & Filatotchev (2003) find little support that outside members are, as agency theory suggests, performance enhancing and conclude that future research should address 3 issues concerning the predictions agency theory makes: (1) whether the fundamental theory is suitable; (2) whether there are methodological problems in the studies with similar hypotheses; (3) whether there are institutional factors that need to be accounted for. Peng (2004) presents proof of an influence of outside directors on performance when measured as sales growth. However, their impact becomes negligible when outside directors are linked to financial performance measures.

Taking into account that both important theories (e.g. agency and resource dependency) and important (e.g. Fama & Jensen, 1983; McNulty & Pettigrew, 1998), give outsiders an essential role in mitigating agency costs, it is at least fascinating that so many empirical studies fail to find a relationship at all. This can be the result of the fact that outside directors serve their own goals, which can translate in an attempt to satisfy both shareholders and insiders, and by doing so, they mitigate their potential impact on performance. I will return to this argument in later sections, but I argue here that, considering the hypothesized important role of outside directors, more research should be directed towards identifying their interest in the firm. Especially when considering the potential divergence between the interests of the shareholders and outside directors (Bebchuk et al., 2002). An explanation for the focus on outside directors by theory and authors is provided by DiMaggio & Powell (1983). Relying on institutional theory they state that it “occur as the result of processes that make organizations more similar without necessarily making them more efficient” (DiMaggio and Powell, 1983: 147).

structure and R&D spending and previous operating performance, which are presented as the most important determinants of firm performance. The authors contribute that different firms need different governance practices and therefore contrast the view that standardized board structures should be ordered. Westphal and Zajac (1995) and Zajac and Westphal (1996) explore the relationship between board and CEO power and the selection of new directors in firms and find that new directors will be demographically more similar to that stakeholder. Additionally, the authors find evidence for a positive association between the degree of similarity between the board and the CEO and the CEO’s compensation. The authors also find that outside successors are generally demographically different from their CEO predecessors but demographically similar to the board. McNulty and Pettigrew (1998) elaborate how outsiders convert sources of power into influence and find differences between chairmen and other outside directors, however they do not discuss the link between performance and the power distribution among. Clark (1998) adds that outside directors will have increasing role in the company’s strategic direction but makes no inferences about their impact on performance. In several articles and a book (e.g. Bebchuk et al., 2002; and Bebchuk & Fried 2003 & 2004) the authors Bebchuk, Fried, and Walker assess the agency problems in the contract design stage of contracts between agents and principals. One of their overall conclusions is that outside directors and shareholders do not have sufficient power to adequately control the remuneration process. Concerning the outside directors the authors add that often the interests of shareholders and outside directors are not aligned.

3.3.4 Inter-stakeholder Distribution of Power and Performance

In the previous sections I discussed studies in the academic field that investigated the role of power of single stakeholders, i.e. studies that contain an ‘intra-stakeholder’ approach. Since power is relative to the power of other stakeholders these studies indirectly comment on the relationship between power of multiple stakeholders and performance. However, to my knowledge there is no empirical study that incorporates all strategic in a single model. The only work that addresses the power of all three strategic stakeholders in a single study is Jensen and Warner (1988). However, they rely on 17 studies that adopted a single stakeholder approach. Their important findings are: (a) patterns of stock ownership by insiders and outsiders can influence executive behavior, company performance, and shareholder voting; (b) corporate leverage, inside share ownership by executives, and the control market are interrelated; (c) departures from one-share-one-vote have an effect on firm value and efficiency; (d) takeover resistance through defensive restructurings or poison pills is linked with declines in share price; and (e) top management turnover is inversely correlated to share price performance.

Furthermore, an empirical study that specifically approaches the power-performance dilemma with multiple stakeholders, i.e. a multiple stakeholder approach towards the effect of power on performance, is absent. Therefore, I conclude that there is a gap in the academic literature; work specifically attending to a relationship between the distribution of power among multiple stakeholders and performance.

3.4 Overview and Hypotheses

upper echelons of the firms (i.e. where its impact is the highest) and its relationship with performance needs to be assessed. By incorporating all stakeholders in 1 model, this paper will pay more attention to the relative aspect of power. Through this, although structural sources are used, it will include behavior in the association between power and performance. I believe that this is an uncharted area in corporate governance literature.

In addition to the limited approach concerning the number of stakeholders, I believe that previous literature has focused too much on a single strategy, namely the strategy that maximizes the value of shareholders. Without engaging in the normative discussion on what the performance strategy ought to be, I believe that a step backwards needs to be made. If power has the capability to influence performance and strategy through the decision process, it is important to investigate what stakeholders will do with power. This can also be an explanation for the conflicting results in empirical studies. Since in agency theory interests are not aligned, powerful actors may pursue different strategies. Therefore a relationship between power and the performance measures related to the single financial performance strategy may not be significant. Especially the interest of outside directors (and the related performance strategy) remains a black-box. This is strange considering the important role many scholars give to this strategic stakeholder. Thus, before investigating the link between power and performance measures that represent shareholder value, the potential relationships between power and other performance strategies needs to be explained.

Thus, in my opinion previous research, although valuable, has largely ignored the relative nature of power by focusing on single stakeholders (and through this neglecting the behavioral aspects that play a role in the structural relationships) and used a too limited approach towards the measurement of performance. Hence, besides including multiple stakeholders in a single model, I investigate the relationship between power and different performance strategies.

power plays a role in the vacuum regarding behavioral aspects in the relationships as laid down by agency theory by affecting the agency costs associated with the construction of efficient contracts between the principal and the agent. These costs are (a) monitoring costs; (b) bonding costs; and (c) residual losses (Jensen & Meckling, 1976). In sum, power affects (a) performance by influencing agency costs and (b) performance strategy by affecting the ability of stakeholders to pursue personal interest through the strategic agenda. I discuss the association between the various stakeholders’ power and in section 3.4.1, followed by an assessment of the impact of power on (performance) strategy in section 3.4.2. The former is the confirmative part of this paper and relates to the previous literature that mainly focuses on financial performance. However, contrary to previous research I incorporate all strategic stakeholders in my model. Section 3.4.2 is the explorative and focal part of my study. It studies the potential effect of power on performance strategy, a research area that the literature review suggested to be largely unattended.

3.4.1 Power and Performance

As argued in the previous section and in the literature review power affects financial performance through its influence on agency costs associated with the construction of adequate contracts to control agency problems (Daily & Johnson, 1997; Nicholson & Nicholson, 2007). The detrimental or beneficial effects of agency costs on performance depend on which stakeholder’s power affects them. Power of the principal will reduce the agency costs by increasing monitoring and increased the power to control and enforce contracts. Additionally it will be less costly to align the interest of the agent with that of the principal because the principal will have more control over the decision process. Hence, shareholder power is positively related with firm performance (Bebchuk, 2005; Mitton, 2002; Shleifer and Vishny, 1997).