Faculty of Economics and Business
MSc Accountancy & ControllingThe effect of busy directors on
monitoring and advising performance
The moderating impact of stock ownership and environmental complexity
Author: Jelmer Helder
Student number: 2485680
Adress: Slichtland 3, 9302GH Roden E-mail: j.helder.3@student.rug.nl Phone number: 06-11964794
Supervisor: Dr. R.C. Trapp Co-assessor: Dr. V.A. Porumb
2
The effect of busy directors on monitoring and advising performance
The moderating impact of stock ownership and environmental complexity
Abstract
This paper examines whether director busyness affects the monitoring and the advising performance of the board, using a sample of S&P 500 listed firms from 2009 – 2016. There are increasing market concerns regarding the effectiveness of busy directors. It is questioned if busy directors are able to devote sufficient time to carry out their duties. Director busyness has also attracted significant attention of researchers and the prior literature focusing on the relation between director busyness and firm performance shows mixed results. Findings of this study show that director busyness is positively associated with the advising performance of the board and partial support is found for the expected negative association between busy directors and the monitoring performance of the board. Finally, this study examines two factors that might moderate these relations. Evidence is found that environmental complexity strengthens the relation between director busyness and the advising performance of directors. Director ownership does not negatively affect the relation between director busyness and the monitoring performance of the board. These results put the concerns of regulators and policy makers into perspective, as director busyness might also have a positive effect on the advising performance of the board.
Keywords:
Busyness, Busy directors, Multiple directorships, Monitoring, Advising, Board, Corporate governance3
Table of Contents
1. Introduction ... 4
2. Theoretical background ... 8
2.1 Review of literature on director busyness ... 8
2.2 Hypothesis development of director busyness and the monitoring performance ... 10
2.3 Hypothesis development of director busyness and the advising performance ... 12
2.4 Stock ownership of directors ... 13
2.5 Environmental complexity ... 14
3. Research methodology ... 16
3.1. Sample selection and data collection ... 16
3.2 Variables ... 17 3.2.1 Dependent variables ... 17 3.2.2 Independent variable ... 19 3.2.3 Moderating variables ... 20 3.2.4. Control variables ... 20 3.3 Regression models ... 22 4. Results section ... 24
4.1 Descriptive statistics and correlation analysis ... 24
4.2 Busy boards and monitoring performance ... 25
4.3 Busy boards and advising performance ... 26
4.4 Busy boards and the moderating effect of stock ownership ... 28
4.5 Busy boards and the moderating effect of environmental complexity ... 29
5. Discussion ... 31
6. Limitations and recommendations for future research ... 33
4
1. Introduction
After the financial crisis in 2008 there were resolutions adopted in the United States to limit the number of directorships for directors of publicly traded companies. More recently, the Institutional Shareholder Services (ISS) has stated that according to a survey of the National Association of Corporate Directors (NACD), “the average time commitment for board service has exploded”. As a result, they have revised their recommendation, which has become effective in 2017, to vote against a director who sits on more than five boards instead of more than six boards.1 It makes sense that time and effort of directors are not
unlimited, so therefore it is important to know how multiple directorships affect the monitoring and advising roles of the members of the board of directors.
As shown in the first section, director busyness has recently attracted significant attention from regulators and policy makers. Director busyness is also called multiple directorships and can be defined as ‘‘a person affiliated with one organization sits on the board of directors of another organization’’ (Mizruchi, 1996, p. 271). There are concerns that directors take on too many seats on different boards (Kaczmarek, Kimino, & Pye, 2014). Director busyness for example attracts the attention of regulatory bodies, who are concerned that multiple directorships might compromise the quality of the tasks a director must perform, which is demonstrated by the fact that many countries tighten the regulatory frameworks and revise the national codes of good governance (Méndez, Pathan, & García, 2015). This concern is also recognized in the literature, for example by Lei and Deng, 2014 who question if directors with many directorships are still able to devote enough time and effort to their monitoring and advising duties. Studies focusing on busy directors are not only from the recent past, Lipton and Lorsch (1992) were one of the first who identified that a lack of time might cause problems to members of the board to carry out their monitoring and advising duties. Prior research implies that the monitoring activities of a director might come at the expense of the advising role of a director or the other way around, because both of these roles are competing for the director’s limited time and attention (Faleye, Hoitash, & Hoitash, 2011;Armstrong, Guay, & Weber, 2010). Given the fact that time is a finite resource for directors, devoting enough time to the monitoring tasks as well as to the advising tasks can be quite challenging. If a director holds multiple directorships, this becomes even more challenging, because in that case, a busy director has to divide his or her time between multiple boards. Researchers argue that serving on multiple board seats may lead to
1 Institutional Shareholder Services. (2015, November). Americas Proxy Voting Guidelines Updates, 2016 Benchmark Policy Recommendations
5 overcommitment of directors, which will harm their ability to monitor and advice a firm effectively. (Ferris, Jagannathan, & Pritchard, 2003; Fich & Shivdasani, 2006).
On the other hand, holding multiple directorships is perceived as a positive indication of the reputation of a director (Fama & Jensen, 1983). Researchers state that multiple directorship has a positive effect, because it enhances directors’ experience and helps them building a bigger business network. This expertise directors gain and the wide spread network they build up from serving on multiple boards is adding value to the organization (Booth & Deli, 1996; Carpenter & Westphal, 2001; Loderer & Peyer, 2002). So it appears that director busyness might have advantages as well as disadvantages.
Most prior literature about busy directors focuses on the relation between busyness of directors and overall firm performance. When it comes to the relation between busyness of directors and firm performance, the literature shows conflicting evidence. Core, Holthausen, and Larcker (1999) find that a director who serves on multiple boards is related to excessively high level compensation schemes, which in turn negatively influences firm performance. Besides, the more recent literature report that busy directors have a negative impact on firm performance and argue that multiple directorships negatively impact the oversight tasks and the monitoring effectiveness of a director (Ahn, Jiraporn, & Kim, 2010; Fich & Shivdasani, 2006; Jiraporn, Kim, & Davidson, 2008). However, other studies have found a positive relation between busy directors and firm performance (Field, Lowry & Mkrtchyan, 2013; Loderer & Peyer, 2002). These studies argue that busy directors are beneficial to a firm, as the gained experience and knowledge of busy directors will improve their qualities and skill set, which positively affects firm performance. The majority of prior literature focusing on director busyness try to find a link between multiple directorships and firm performance, but the results of these studies are ambiguous. Therefore it is still open to question in what way and how busyness of directors affect firm performance.
This paper will take a closer look at the different roles of members of the board in order to get a better understanding of how director busyness affects the performance of the board. The board of directors have an overall responsibility to steer and guide an organization and their most important activities can be distinguished into monitoring and advising tasks. The board of directors monitor the performance of the top management of an organization and make sure that the managers act in the best interest of the shareholders. This means that they have to provide oversight over the financial reporting quality, the compensation of managers and the implementation of strategies (Jensen & Meckling, 1976). Besides this monitoring role, the board is also charged with the advising function and the activity to design and
6 implement firm strategies. This is considered to be the other primary role of directors (Hillman, Cannella, & Paetzold, 2000; Adams, Hermalin, & Weisbach, 2010).
As mentioned before, prior literature has focused on studying the effect of busyness of directors on overall firm performance. In contrast, this study focuses on how director busyness affects the advising role of directors and how director busyness affects the monitoring role of directors. By making a distinction between these functions, this study tries to provide a more comprehensive picture of director busyness. To summarize, this study examines if director busyness is detrimental or beneficial to the monitoring role of directors on the one hand and the advising role of directors on the other hand. Therefore the research question of this paper is defined as follows: Does busyness of directors affect the monitoring and advising performance of directors?
Furthermore, there may be moderators affecting the relation between busyness of directors and board effectiveness. These moderators might help to explain how busyness of directors affects the monitoring and advising performance of busy board members. As a consequence, the moderators might help to explain why prior studies have found mixed results when it comes to the relation between director busyness and firm performance. This paper includes two moderating factors that may influence the relation between director busyness and board effectiveness. The first moderating factor focuses on how stock ownership might influence the relation between busyness of directors and their monitoring performance. From an agency perspective, an important responsibility of the board of directors is to monitor the management of an organization and to protect the rights and interests of the shareholders, because they are the formal representatives of the shareholders of an organization (Lorsch, 1989). Structural mechanisms are needed to overcome the agency problem and to help align the interests of directors with the interests of the shareholders (Mallette & Fowler, 1992). One important mechanism is to increase directors’ ownership of stock, because directors that own shares of the organization where they serve on the board are found to be better monitors than directors who do not own shares of the organization (Elson, 1996).
The second factor focuses on how environmental complexity might influence the relation between busyness of directors and their advising performance. This moderator can help to get a better understanding of why some organizations benefit more from busy directors than others, by examining the different advisory needs of organizations. Gaining knowledge about how the organization’s operating environment influences the relation between director busyness and the advising performance might help to appoint the suitable directors, which will eventually result in a more effective board composition. Prior
7 research suggest that it is important for organizations to adapt to the environment and to respond to changes in the environment in order to survive or succeed (Lawrence & Lorsch, 1967; Cannon & John, 2007). The resource dependence theory, which looks at a board as a strategic resource that helps a firm to secure its critical resources and to provide a firm with advice, is used to develop predictions in the theory section (Pfeffer, 1973; Pfeffer & Salancik, 1978). Drawing from the resource dependence theory, Aldrich (1979) described environmental complexity as the number of units with which interaction is required and the extent to which an organization requires specialized or sophisticated knowledge to cope with complexity. When a firm operates in a highly complex industry, it has to deal with a rapidly changing, dynamic and competitive environment (Markarian & Parbonetti, 2007). The expertise and advice of the directors becomes more important when strategic changes have to be made to adapt to the environment and in order to survive in a complex environment. Therefore, busy directors might be more useful to organizations that operate in complex environments as the demand for strategic advising increases.
In order to empirically test the effect of director busyness on the monitoring and advising performance of directors, a sample of S&P 500 listed companies from the US that spans from 2009– 2016 is used. This study provides evidence which partially supports the expected negative association between director busyness and the monitoring performance. However, no evidence was found that stock ownership moderates this relation. The expectation, based on the resource dependence theory, that director busyness is positively associated with the advising performance of directors, is supported by the findings of this study. Finally, the results of this study suggest that environmental complexity moderates the relation between director busyness and the advising performance.
The primary contribution of this paper is twofold. It contributes to the prior literature by providing a more comprehensive understanding of the two primary roles of busy directors and by showing that the extent to which firms benefit from busy directors depends on the environmental complexity. Besides that, the findings of this study might be of interest for regulators and policy makers who issue recommendations to limit the number of directorships.
The remainder of this paper is organized as follows: the next section presents a literature review and the development of the hypothesis. Section 3 explains the sample selection procedure, describes the variables and presents the research methods used. The results of the study are presented in section 4. Section 5 draws conclusion and gives theoretical and practical implications. Finally, section 6 discusses the limitations and provides recommendations for future research.
8
2. Theoretical background
The theoretical background starts with a review of the existing literature on director busyness. As already mentioned in the introduction of this paper, the activities of a member of the board of directors can be distinguished into two primary roles. The next sections will further explain these two roles of directors followed by the hypothesis development of director busyness and the monitoring and advising performance. Finally, two moderating factors that might influence the monitoring and advising performance of busy directors are described. All the hypotheses are based upon the literature review and theoretical background.
2.1 Review of literature on director busyness
‘Director busyness’ and ‘busy boards’ are terms that have attracted the attention of corporate governance research and have become more and more popular in recent decades (Elyasiani & Zhang, 2015). Other early papers in corporate governance literature that started investigating director busyness are Core et al. (1999), Ferris et al. (2003), and Fich and Shivdasani (2006). These papers defined director busyness as a director who is holding multiple directorships, or more specifically, a director serving on three or more boards to investigate the characteristics that could explain the number of directorships and the influence of director busyness on firm performance. Despite the numerous studies researching the effect of director busyness on firm performance, the findings are mixed and therefore not conclusive (Benson, Davidson, Davidson & Wang, 2015). Researchers still have not been able to find out if multiple directorships positively or negatively affect the monitoring and advising performance of directors. The next section describes the studies that show detrimental effects of director busyness followed by a paragraph with an overview of literature that shows the beneficial effects of director busyness.
Directors with multiple directorships have been criticized for serving on too many boards, which negatively influences their monitoring and advising duties. For example the National Association of Corporate Directors, the Council for Institutional Investors, and the Institutional Shareholder Services, were critical about multiple directorships and therefore adjusted their guidelines (Jiraporn, Singh & Lee, 2008). Core et al. (1999) argue that busy board members are less effective at monitoring the compensation schemes, because CEO compensation is higher when the board is busy. Faleye et al. (2011) also find evidence that busy directors are less effective, because they overpay the CEO and are more inclined to involve in earnings management. Fich and Shivdasani (2006) report that firms with busy boards are associated with lower market to book ratios and thus weaker profitability. Besides that, their results suggest that ‘‘boards relying heavily on outside directors that serve on several boards are likely to
9 experience a decline in their quality of corporate governance’’ (2006, p. 722). In addition, Jackling and Johl (2009) find evidence that busy outside directors have a negative effect on firm performance. Because serving on multiple boards causes a lack of time for board members to carry out their monitoring and advising duties. Sharma and Iselin (2012) examined the effect of multiple directorships for audit committee members and reported a positive association between financial misstatements and busy audit committee members. Ahn et al. (2010) find a negative market reaction towards mergers and acquisitions by firms with a busy board. Finally, Falato, Kadyrzhanova, and Lel (2014) observe a negative stock market reaction when the busyness of members of the board increased. Based on this they conclude that the market perceives director busyness to be detrimental to firm performance.
In contrast to the previous section, this section gives an overview of the positive effects of director busyness that have been found in previous literature. Loderer and Peyer (2002) find that multiple directorships is positively related to firm value of listed firms. Fama and Jensen (1983) already noted that members of the board of directors are rewarded with additional board appointments by the labor market if they have built a strong reputation as effective monitors. Brown and Maloney (1999) document that firms with directors serving on multiple boards are rewarded with superior returns from acquisitions. Ferris et al. (2003) document a positive relation between firm performance and the number of board seats subsequently held by a director. Besides that, they find evidence that markets react positive to the appointment of busy directors on the board, which implies that superior monitoring abilities and experience can be beneficial. Harris and Shimizu (2004) found a positive relation between boards with ‘overboarded’ directors and abnormal returns, furthermore they suggest that these boards have the ability to make informed contributions and are more efficient decision makers. In line with these findings, Masulis and Mobbs (2011) document that inside directors who hold outside board seats overstate earnings less often and are more likely to make better decisions about acquisitions in comparison to directors who do not hold outside board seats. Field et al. (2013) find that busy directors contribute to firm value, when the firm is at an early stage of its lifecycle. This is due to the fact that young firms are in relatively high demand for advice. Finally, Elyasiani and Zhang (2015) find a positive relation between the performance of bank holding companies and the number of directors serving on multiple boards. This indicates that director busyness has desirable effects on bank holding companies.
This overview of the literature shows that the vast majority of studies on director busyness have been focusing on the overall performance of the firm. Most of the prior literature uses firm financial performance to assess whether director busyness has a positive or detrimental effect on firm
10 performance. Hereby they have overly focused on the advising performance of busy directors, and less focused on the monitoring role. As mentioned before, the monitoring role and the advising role of the board are both important, therefore, this study will take both roles into account in order to try to explain the conflicting evidence that has been found in the prior literature. Furthermore, moderating factors will be included as they might explain some of the conflicting evidence as well.
2.2 Hypothesis development of director busyness and the monitoring performance
Schwartz-Ziv and Weisbach (2013) argue that the focus of board members’ activities has shifted more and more to an independent and active monitoring role, board members spend a great deal of their time monitoring the management of an organization. Vafeas (2005) shows that if the time that directors devote to oversight duties is increased, this will lead to improved monitoring quality. This is why the board of directors is considered as one of the most important internal governance mechanisms of a firm (Fama, 1980). The supervisory duties of board members include certifying the quality of financial reports, appointing or dismissing managers, approving the remuneration schemes of managers and consulting with the auditor. Besides that, board members provide oversight of the strategic decisions and direction of a firm and they assess the overall firm performance (Withers & Fitza, 2017). These monitoring activities and responsibilities of the board can potentially be affected if members of the board serve on multiple boards, as is already discussed in the literature review of director busyness.
A theory that is used to explain the monitoring role of the board of directors is the agency theory. The agency theory helps to explain the relation between principals and agents in an organization and can be described as an agreement between principals and agents. The agents carry out some duties or services and get decision making authority on the behalf the principles (Jensen & Meckling, 1976). Agency problems arise due to a misalignment of interests between the principals and agents because of the separation of ownership and control (Fama & Jensen, 1983). This misalignment occurs for example when a director is assumed to act in a way that benefit the management and himself at the expense of the shareholders (Mallette & Fowler, 1992).
Multiple directorships might cause directors to be overcommitted. This means directors who serve on multiple boards might take on more responsibilities than they are capable of dealing with. Because these additional directorships imply extra work, since directors have to fulfill extended duties when serving on boards of different firms. The contribution of busy directors to the monitoring performance of the board does not only depend on their knowledge and skills, but also on the time they have available to carry out
11 their duties and the amount of time they devote to prepare board meetings (Fahlenbrach, Low, & Stulz, 2010). This makes it reasonable that directors serving on multiple boards are overcommitted.
A busy director must divide the time available between multiple organizations, which limits the time a busy director can spend on monitoring duties. Following the agency theory, it is the task of the board to look after the interests of the shareholders. But a busy director might not have sufficient time to attend the board meetings and to oversee the management of the organization. Consequently, these commitments with multiple firms could negatively affect the monitoring capabilities of busy directors. Lipton and Lorsch (1992, p. 64) state ‘‘Based on our experience, the most widely shared problem directors have is a lack of time to carry out their duties.’’ Consequently, these busy board members might shirk their responsibilities to monitor the management and to carry out their duties. Jiraporn, Davidson, DaDalt, and Ning (2009) find evidence that busy boards are related to lower board meeting attendance.
The following findings from previous literature support the expected negative effect of the monitoring performances of busy directors. Lower attendance rate at board meetings and committee meetings, higher CEO compensation schemes and CEO overpayment, more involvement in earnings management, lower market to book rations, a decline in quality of corporate governance, higher frequency of financial misstatements are all indications of weak monitoring performance of busy directors (Jiraporn et al., 2009; Core et al., 1999; Faleye et al., 2011; Fich & Shivdasani, 2006; Sharma & Iselin, 2012). Furthermore, a relation between director busyness and both low pay-performance sensitivity and high CEO remuneration might be indications of reduced oversight and supervision (Méndez et al., 2015), although supervising is one of the main responsibilities of the board according to the agency theory. Additionally, mangers who are subject to lax monitoring enforce higher agency costs on an organization, which leads to a reduction in firm performance (Ferris et al., 2003).
All in all, busy directors might be overcommitted because they have to divide their time, to carry out their monitoring duties, between multiple firms. This in turn might impact the monitoring performance of busy directors. Besides that, there are multiple studies providing evidence that director busyness may have a negative influence on the monitoring role of directors. Therefore it is expected that director busyness is negatively related to the monitoring performance of the board.
12
2.3 Hypothesis development of director busyness and the advising performance
The advising activities of the board include supporting the actions and decision of the management. As is stated by Mace ‘‘Management manages the company, and board members serve as sources of advice and counsel to the management” (1972, p. 38). This role of the board members is in line with the resource dependence theory. Pfeffer (1972) argues that directors can create valuable links to the external environment of organizations, can help to get access to important resources through various networks and directors can provide useful insights from different perspectives. They are for example responsible for the appointment of managers, the development and effectiveness of a governance system and they can provide strategic guidance for the organization. Adams et al (2010) and Hermalin and Weisbach (2003) also acknowledge the advising role of the board of directors as they advise and guide the top management in formulating and implementing the organization’s strategy. More experienced members of the board are better able to provide advice, given the skills and expertise they gathered in previous activities and board memberships.
Holding directorships on multiple boards leads to more connections and interaction with other board members, which gives access to valuable resources and information. Hereby, busy board members gain valuable expertise and knowledge from serving on multiple board seats (Fich & Shivdasani, 2006). A theory that might explain the advising role of the board of directors is the resource dependence theory. This theory is commonly used to study the advising role of the board of directors in prior literature (Hillman, Withers, & Collins, 2009), and will also be used in this study.Following the resource dependence theory, an organization is characterized as an open system, which depends on contingencies and resources in the external environment (Pfeffer & Salancik, 1978). Board members may aid in managing possible uncertainties in the external environment by providing critical resources (Hillman and Dalziel, 2003). The resource dependence theory originated by Pfeffer and Salancik (1978), underpins this function of the board. They also state that ‘‘when an organization appoints an individual to a board, it expects the individual will come to support the organization, will concern himself with its problems, will variably present it to others, and will try to aid it" (Pfeffer & Salancik, 1978, p. 163). Members of the board provide various benefits to an organization: advice and counsel, legitimacy, access to resources and channels for communicating information between a firm and the external environment (Pfeffer & Salancik, 1978).
The gained knowledge and expertise of busy directors is likely to help directors to provide advice of higher quality to the management of a firm and to the rest of the board. Numerous studies find that holding
13 multiple directorships allows directors to gain specific industry knowledge that helps to improve their advising performance. Besides that, multiple directorships provide board members with a wider range of accounting expertise (Loderer & Peyer, 2002; Masulis & Mobbs, 2011; Field et al., 2013). This perspective is in line with the resource dependence theory, as busy board members bring expertise and experience to the table, which is a valuable resource to a firm. Holding multiple directorships is perceived as a measure of quality, as it might be an indication of previous success. Therefore, busy directors are associated with enhanced business value caused by their high-quality advising expertise. Other studies also provide evidence of increased advising skills provided by busy directors. Adams (2009) for example provides evidence that multiple directorships makes board members more effective advisors, as their business network and access to information increases.
In sum, based on the findings of prior literature and theoretical arguments it is expected that director busyness is positively related to the advising performance of the board.
H2: There is a positive association between director busyness and advising performance. 2.4 Stock ownership of directors
Given the fact that agency problems might arise due to separation of ownership and control (Fama & Jensen, 1983), mechanisms are needed in order to align the interests of principals and agents (Jensen & Meckling, 1976). The board of directors has an important role in aligning the interests of principals and agents. However, as previously stated, busy board members might be less effective monitors than directors that are less busy. If directors own shares in the company where they sit on the board, this might be an incentive to be more effective monitors. The agency theory assumes that managers are motivated extrinsically, so motivation arises from outside the individual. The stock ownership of directors might encourage directors to be more committed and aligned with the interests of the shareholders and extrinsically motivates them to engage in their tasks. Therefore it can be expected that they will be motivated to achieve the same goals as the other shareholders. As a result, directors might put more effort into their monitoring tasks because they have more incentives to monitor since they are shareholders themselves. Prior literature supports the expectation that monitoring performance and share ownership of directors are associated. For example, Bhagat, Carey & Elson (1999) document that there is a connection between directors’ share ownership and better monitoring performance of directors. Besides that, they found evidence that a CEO of a poorly performing company is replaced faster, which might be an indication of effective monitoring board.
14 Making members of the board stockholders, might create a personally based extrinsic motivation to carry out the monitoring tasks more actively, because it might have a more direct effect on their personal wealth if they shirk their monitoring responsibilities. As a result, directors’ stock ownership might mitigate the negative relationship between director busyness and monitoring performance. Elson (1996) finds that problems resulting from the passivity of the board may become less prevalent and directors might be more willing to become active monitors if they own shares of the company where they serve on the board. In line with the agency theory, findings of Beasley (1996) also suggest that directors’ stock ownership might lead to better monitoring performance of directors. The study finds evidence that increased stock ownership of directors reduces financial statement fraud cases, which can be interpreted as a sign of more effective monitoring. In this regard, results of Shivdasani (1993) provide evidence that increased directorial ownership might lead to greater quality of monitoring.
Thus, it is expected that the higher the level of directors’ ownership of stocks, the higher the monitoring quality. Therefore it is proposed that busy directors who own stocks in the organization where they serve on the board are more effective monitors than busy directors without stock ownership. Consequently, stock ownership might have a moderating impact on the relationship between busyness of directors and monitoring performance.
H3: The negative relation between director busyness and monitoring performance is weaker when
directors own shares of the company in which they serve on the board.
2.5 Environmental complexity
As stated before, the resource dependence theory explains one of the primary roles of the board. The theory is based on the principle that organizations employ inter-organizational linkages with actors in its environment to control and acquire necessary resources. This theory looks at the directors of the board as a resource that provides organizations with advice and that helps to make strategic decisions (Pfeffer & Salancik, 1978). Busy directors might be even more value enhancing to organizations since they have bigger business networks and more experience, which are valuable resources they add to a firm. When a firm operates in a highly complex environment, it has to deal with a rapidly changing, dynamic and competitive environment (Markarian & Parbonetti, 2007). An organization needs specific information about the external environment and resources, such as knowledge and a broad network, in order to survive in a complex environment. The advising role might become more important under complex conditions, since a firm has to cope with more inputs and outputs in the external environment (Dess &
15 Beard, 1984). Besides that, also the demand on the directors might become higher. Therefore, the need for strategic decision-making and advising will increase in order to design and implement a well-structured strategy that deals with a complex environment (Aldrich, 1979). George (2005) also states that it becomes more critical for firms operating in a complex environment to adapt their strategy to the environment. Environmental complexity has a significant influence on the enactment of strategic reactions. Links with the environment and resources are of great value to create sustainable competitive advantage, this is supported by Eisenhardt and Schoonhoven (1996) who provide empirical evidence that directors with external ties play a critical role in strategic decisions making and strategy formulation. In line with these findings, Markarian and Parbonetti (2007) also suggest that valuable resources can be enhanced through multiple directorships, because this provides a source of expertise, valuable external links and builds a competitive advantage.
In sum, it must be considered that not all organizations benefit from the advisory performance of the board to the same extent. Organizations operate in different environments, some would require more strategic advice from the board of directors than others. It is expected that firms operating in a more complex environment have higher advising demands than firms operating in less complex environments. This creates a more crucial role for the advising performance of the board. So a firm that operates in a complex environment might benefit more from the experience, skills and network of a director than a firm that does not operate in a complex environment. Therefore, the advantage of having a board consisting of busy directors with comprehensive expertise and experience might be more valuable if the firm operates in a complex environment. For that reason, it might be the case that a complex environment has a positive moderating impact on the relationship between director busyness and advising performance.
H4: Environmental complexity positively affects the relationship between director busyness and
16
3. Research methodology
3.1. Sample selection and data collection
The sample used in this study consists of companies that are part of the Standard & Poor 500 (S&P 500) index. This index includes 500 of the largest companies that are listed on an American stock market like the NASDAQ or NYSE with their headquarters in the U.S. Ferris et al. (2003) showed that director busyness is a phenomenon that predominantly occurs in large firms. Therefore, the sample of this study consists of S&P 500 firms since it is more likely to capture relationships related to multiple directorships when using a sample of large firms, than it would be if a sample consisting of small or medium size enterprises was used.
Data on these firms is obtained from BoardEx and Compustat and spans a multi-year period ranging from the financial years 2009 until 2016. Data is collected from 2009 to make sure that possible influences from the financial crisis in 2008 are excluded and do not affect the results of this study and 2016 was the most recent year for which data was available. Firms that were listed for one or more years on the S&P 500 between 2009 and 2016 are included in the sample. A few screening criteria are imposed to compile the final sample. First, relevant data of the S&P 500 has to be available in the databases that are used. Therefore, observations with missing data from the BoardEx and Compustat database are excluded. Second, utility firms and firms that operate in financial industries are excluded from the sample. These firms are excluded because regulatory effects might limit the role for the board of directors (Fich and Shivdasani, 2006). The final sample consists of 2564 observations across 8 years. A summary of the data selection and the number of observations that are excluded can be found in Table 1.
Table 1 – Data inclusion
Exclusion reason Loss Observations
Initial firm year observations 3647
Missing data -360 3287
Utility firms (SIC 4900 - 4999) -252 3035
Financial firms (SIC 6000 – 6499; 6600 – 6999) -471 2564
17 As already mentioned, the sample that is used in this study is developed from several sources. In order to obtain information regarding individual specific board members and general board characteristics of firms that appear in the sample, the BoardEx database is used. This database provides information like board size, number of current directorships, the individual role, director’s age and director’s stockownership. This data is merged with data from the Compustat database through Ticker codes and fiscal year data. Compustat provides the necessary accounting and financial data for the sample firms. Information about individual directors is aggregated to a firm level in order to match the director’s individual data with general firm data retrieved from Compustat. Aggregating the data to a company level is necessary to carry out the analysis. The dependent variables, the independent variables, the moderating variables and the control variables are described in the following part of this section. An overview of the used variables and the sources of the data can be found in Appendix A. The exact names of the used variables from BoardEx and the abbreviations of the variables retrieved from Compustat are also mention in Appendix A.
3.2 Variables
3.2.1 Dependent variables Monitoring performance
The discretionary accruals (DA) are the proxy that is used to measure the dependent variable monitoring performance. Discretionary accruals measures director effectiveness in monitoring over financial reporting quality, but is used in this study as a proxy for monitoring activities in general. Discretionary accruals are the accruals component that identify managerial choices. Therefore, the amount of discretionary accruals is a metric that does not reflect the business conditions that naturally create accruals, but it reflects the accruals that are a result of management choices. If managers use aggressive accounting methods to overstate or manage earnings, this is indicated by a larger amount of discretionary accruals. Overstating or managing earnings is a sign of poor financial reporting quality, which consequently means that the monitoring performance of directors is less effective. Therefore, discretionary accruals are used to measure the monitoring performance of a director. A model that is extensively used in the existing literature, is the abnormal accrual model by Jones (1991). Dechow, Sloan & Sweeney (1995) adjusted the Jones model, and stated that the modified jones model is the most effective model to capture discretionary accruals. The modified Jones model, which has been widely used in the literature, is also used in this study (Islam & Ahmad, 2011).
18 TACCt = ∆CAt + ∆Casht + ∆CLt + ∆DCLt – DEPt
Where: TACCt = total accruals in year t
∆CAt = change in current assets in year t
∆Casht = change in cash and cash equivalents in year t
∆CLt = change in current liabilities in year t
DEPt = depreciation and amortization costs in year t
Then, the nondiscretionary accruals are estimated using the following model:
NDACCt At-1
=
α1
1 𝐴t-1+
α2
(∆REVt - ∆RECt) At-1+
α3
PPEt At-1Where: ∆REVt =revenues in year t less the revenues in year t-1
∆RECt = net receivables in year t less net receivables in year t-1
PPEt = gross property plant and equipment in year t
At-1 = total assets in year t-1
α1,α2, α3 = firm-specific parameters
Finally, discretionary accruals are estimated as: DACCt = TACCt – NDACCt
Advising performance
The proxy Tobin’s Q (TOBINSQ) is used to measure the advising performance of the board of directors. Tobin’s Q is used in prior literature to assess the impact of intangibles such as R&D investments and future investment opportunities (Tobin, 1969). Hall (1993) for example used it to analyse how R&D expenditures affect the market value of a firm. In the more recent literature, Tobin’s Q is commonly used, in studies that focus on the composition and structure of the board, as a proxy for firm performance (Fich & Shivdasani, 2006). It is a useful measure due to its relationship with investments. Because it measures the market value to replacement cost, the measure is related to the investment rate. A firm will invest capital if the returns are greater than the cost of providing new capital, assuming that the firm’s objective is to maximize firm value. So if the investments are beneficial to a firm, Tobin’s Q will be high. Conversely, if poor investment decisions are made, due to poor advising performance of the directors, Tobin’s Q will be low. An advantage of Tobin’s Q is that both accounting and market data are combined. Using market data might help to reflect expectations that are potentially unrelated to investments, it might for example also
19 reflect developments, specific trends in the market or regulatory changes. Hereby, it minimizes the distortion of tax laws, systematic risks and accounting conventions. The proxy will be calculated by using the book value of total assets and the market value of equity, minus the book value of equity, divided by the book value of total assets.
3.2.2 Independent variable Director busyness
For each firm of the sample three different measures for multiple directorships are calculated. The different proxies that are used to measure busyness of directors are consistent with the prior literature and were first used by Fich and Shivdasani (2006). Based on Ferris et al. (2003), both executive directors and non-executive directors are included in the three proxies that are used to measure busyness in this study. All the data used to measure the different proxies is retrieved from the BoardEx database. The three different proxies are used as alternative specifications to measure director busyness.
The first proxy, the average number of directorships per director (#DIRSHPS), measures for a given firm of the sample, the average number of directorships held by a director of that firm. This is calculated by taking the number of directorships in listed firms per board member, divided by the total number of board members.
The second proxy measures the share of busy board members (BUSYDIR). To determine when a director is busy, a busy director is defined as a board member who holds three or more directorships in listed companies. This three-directorship benchmark that is used to define director busyness is consistent with previous literature by Core et al. (1999), Ferris et al. (2003) and Fich and Shivdasani (2006). The three-directorship benchmark is somewhat arbitrary, but this cut-off is chosen because it is consistent with previous literature, which makes it possible to compare the results of this study with prior studies. Besides that, the mean number of directorships is the closest to three, which results in an even distribution of non-busy and busy directors. This second proxy is calculated by dividing the number of busy board members by the total number of board members.
The third proxy measures when a board as a whole is busy (BUSYBD). Prior studies reported that director busyness is not always a negative phenomenon. Fich and Shivdasani (2006) stated that before busyness of directors becomes problematic, it must first reach substantial mass of busy directors. Therefore, the method of Fich and Shivdasani (2006) is followed, they argue that a board that consists of 50 percent or more busy directors (three or more directorships) is a busy board. In order to calculate when a board is
20 busy a dummy variable is used. This dummy variable gets a value of zero when less than 50 percent of the board members are busy and a value of one when 50 percent or more of the members of the board are busy.
3.2.3 Moderating variables Stock ownership
Stock ownership of directors (DIROWNSHP) is measured as the value of common stock owned by all directors. The dollar value of the common stock is used instead of the percentage of ownership because the actual value is a more direct measure of a director’s incentives (Bhagat & Bolton, 2013). This measure of ownership is also used in the prior existing literature (Farrer & Ramsay, 1998). The value of the shares at the closing stock price of the annual report date is used to calculate the stock ownership of directors.
Environmental complexity
In order to capture environmental complexity (ENVCOMPLX), a proxy is used. The methodology of Boyd (1990) is followed and a Herfindahl index of industry concentration is employed. The Herfindahl index is used in prior literature to measure complexity (George, 2005). In this study, environmental complexity focuses on the number of units with which interaction is required and how dynamic and competitive a market is. The values of this index are calculated by squaring each firm’s market share within a certain industry and sum these values, which reflects the concentration of resources, competition and heterogeneity in the environment. The total sales are used as a measure of a firm’s market share. The index provides a measure ranging from zero to one, highly concentrated environments have index values closer to one, while less concentrated industries have a value closer to zero. Industries are divided on the basis of SIC codes retrieved from Compustat.
3.2.4. Control variables
Prior studies on director busyness (Ferris et al., 2003; Fich and Shivdasani, 2006) suggest that there are both firm characteristics as well as CEO characteristics and board characteristics influencing the dependent variables monitoring and advising performance. Therefore, control variables are created to control for other characteristics that are known to affect the performance of directors.
Board characteristics and CEO characteristics: Board size (BDSIZE) is included in the tests, because firms reduce board size during financial distress and the prior literature found a relation between board size and company valuation (Gilson, 1990; Yermack, 1996). Besides that, small boards are effective monitors,
21 as they might contribute to lower CEO remuneration (Yermack, 1996). This study also controls for CEO tenure (TENURE) and average age of directors (DIRAGE) as CEO tenure and the directors’ age can influence firm performance. Because directors with a longer tenure have more specialized knowledge and are likely to be more experienced, this more specialized knowledge might lead to more effective monitoring and advising performance. The same applies to directors with a higher age (Field et al., 2013).
Firm characteristics: firm size (FSIZE) is included as a control variable, because it is in accordance with similar research (Fich & Shivdasani). The size of the firm might have influence on the monitoring and advising performance, as larger firms are more likely to implement effective corporate governance practices (Beiner et al., 2006). Another control variable that is included in this study is a firm’s growth opportunity (GRWTH). The growth opportunity of a firm might influence the performance of directors and prior literature also controls for this variable (Fich & Shivdasani, 2006). Demands for advising and monitoring are likely to differ for firms with high growth opportunities in comparison to firms with low growth opportunities (Field et al., 2013). Leverage (LEVERAGE) is included as a control variable, because prior literature provided evidence that leverage is correlated with Tobin’s Q (McConnel & Servaes, 1995). This study also controls for accounting performance by using the return on assets (ROA) measure, this control variable is only included in the model that measures advising performance. Other studies use this variable also as control variable (Benson et al., 2015) or found evidence that busy boards are related to accounting performance (Fich & Shivdasani, 2006).
Finally, fixed year effects (YEAR) and fixed industry effects (INDUSTRY) are included in the regressions. The year dummy variable captures factors that might change over time, for instance, the possible changes in a regulatory framework or changes in accounting standards that are applicable to all the firms of the sample. The industry dummy is included to take expected differences across industries into account. Industries are classified based upon two digit SIC codes. Table 2 shows the descriptive statistics for the year and dummy variables used in the regression models.
Table 2 – Sample description
Variable N Percent %
Year
2009 324 12,6
2010 335 13,1
22 2012 351 13,7 2013 354 13,8 2014 342 13,3 2015 351 13,7 2016 157 6,1
Industry (1-digit SIC code)
Mining & Construction (1) 241 9,4
Manufacturing (2,3) 1302 50,8
Transportation (4) 233 9,1
Wholesale Trade & Retail Trade (5) 351 13,7
Real Estate (6) 42 1,6
Services (7,8) 372 14,5
Public Administration (9) 23 0,9
3.3 Regression models
To test the hypotheses that are developed in the theoretical background section, OLS regression analyses are performed. These analyses measure the relation between director busyness and the monitoring and advising performance of the board. The first hypothesis aims to find a relation between busyness of directors and monitoring performance. The second hypothesis aims to find a relation between busyness of directors and advising performance. This leads to the following empirical models:
1. DA = β0 + β1Director busyness + β2FSIZE + β3GRWTH + β4ROA + β5LEVERAGE + β6BDSIZE +
β7TENURE + β8DIRAGE + β9YEAR + β10INDUSTRY + ε
2. TOBINSQ = β0 + β1Director busyness + β2FSIZE + β3GRWTH + β4LEVERAGE+ β5BDSIZE +
β6TENURE + β7DIRAGE + β8YEAR + β9INDUSTRY + ε
The third hypothesis aims to find if the moderating variable stock ownership has an effect on the relation between busyness of directors and monitoring performance. Finally, the fourth hypothesis focuses on the moderating effect of environmental complexity on the relation between busyness of directors and advising performance. An OLS regression analysis is performed to test the third and fourth hypotheses. This leads to the following empirical models:
23 3. DA = β0 + β1Director busyness + β2DIROWNSHP + β3(Director busyness × DIROWNSHP) +
β4FSIZE + β5GRWTH + β6ROA + β7LEVERAGE + β8BDSIZE + β9TENURE + β10DIRAGE +
β11YEAR + β12INDUSTRY + ε
4. TOBINSQ = β0 + β1Director busyness + β2ENVCOMPLX + β3(Director busyness ×
ENVCOMPLX) + β4FSIZE + β5GRWTH + β6LEVERAGE+ β7BDSIZE + β8TENURE + β9DIRAGE +
β10YEAR + β11INDUSTRY + ε
The ε that is used in the models is the error term. Three different proxies are used to calculate the independent variable director busyness. Thus, the independent variable ‘βDirector busyness’ that is used in the four empirical models consists of the following proxies: the average number of directorships (#DIRSHPS), the percentage of busy directors (BUSYDIR) and the busy board dummy (BUSYBD). These three different proxies of director busyness are run in separate regression analyses.
24
4. Results section
4.1 Descriptive statistics and correlation analysis
Table 3 presents the descriptive statistics for the variables included in the regression models. First the independent variables are described. The different director busyness measures show a mean average number of directorships per director of 2.08 and a mean average of busy board members of 30%. The data shows that the dummy variable BUSYBD has a mean of 0.16, so 16% of the firm observations have a board that is busy (i.e., a board where more than 50% of the board members are busy). Furthermore, the descriptive statistics of the dependent variables and control variables are summarized in table 3.
Table 3 – Descriptive statistics
N Mean Std. Deviation Min Max
Independent Variables #DIRSHPS 2564 2.08 .52 1.00 7.38 BUSYDIR 2564 .30 .17 0.00 .92 BUSYBD 2564 .16 .36 0.00 1.00 Dependent variables RDEXP 1446 .0496 .0605 .000 .6367 TOBINSQ 2564 2.297 1.410 .606 20.923 DA 2564 0.0549 0.0696 .000 0.973 Moderating variables DIROWNSHP 2564 61378.06 427514.33 0.00 8517736.23 ENVCOMPLX 2564 .23 .20 .05 1.00 Control Variables BDSIZE 2564 10.8 2.0 5.0 18.0 TENURE 2564 5.0 5.0 .00 45.9 DIRAGE 2564 61.9 3.5 46.8 75.0 FSIZE 2564 28486.98 58917.39 328.96 781818.00 GRWTH 2564 .0654 .0809 .0013 .7515 ROA 2564 .072 .077 -1.227 .398 LEVERAGE 2564 .71 21.75 -776.59 32.246
A correlation analysis is performed in order to control for any possible multicollinearity problems. The correlation matrix of all the variables used in the regression analysis are presented in table 4. As expected the variables #DIRSHPS, BUSYDIR, BUSYBD show high significant correlation coefficients, since they all measure director busyness. Therefore, this is not a concern regarding multicollinearity. The correlation
25 matrix shows that all the other variables have a correlation coefficient less than 0.4. Thus, the variables identified in the matrix do not reveal any problems regarding multicollinearity. In the next sections, the regressions analysis that are performed to test the hypotheses are summarized in table 5-8. Director busyness is measured with three different proxies, therefore each table shows the results of three regressions models. These three regressions that are displayed in every table test one of the previously developed hypotheses.
Table 4 – Correlations matrix
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) #DIRSHPS 1 BUSYDIR ,779** 1 BUSYBD ,535** ,686** 1 DA -,032 -,030 -,038 1 TOBINSQ ,150** ,125** ,088** ,261** 1 DIROWNSHP -,017 -,018 ,004 ,021 ,032 1 ENVCOMPLX -,021 -,009 ,002 -,070** -,052** ,005 1 BDSIZE ,123** ,139** ,056** -,163** -,231** ,013 -,009 1 TENURE -,067** -,074** -,042* ,003 ,107** ,274** ,051** -,118** 1 DIRAGE ,004 ,035 ,038 -,019 -,058** ,067** -,020 ,065** ,235** 1 FSIZE ,165** ,179** ,113** -,107** -,175** ,264** ,006 ,331** ,144** ,080** 1 GRWTH ,003 -,021 -,042* -,016 -,162** -,027 -,090** -,071** -,016 -,005 ,050* 1 ROA -,100** -,075** -,037 ,085** ,331** ,019 -,004 -,048* ,031 ,021 -,053** -,383** 1 LEVERAGE -,016 -,008 -,026 ,032 -,001 -,005 -,019 ,020 -,016 ,004 -,001 -,023 ,009 1
4.2 Busy boards and monitoring performance
The results of the first three regressions are reported in Table 5. These models test the first hypothesis, namely if director busyness is negatively related to the monitoring performance of board members. All three regressions are related to the first hypotheses, however different measures are used. The first regression (1a) uses #DIRSHPS as a measure of director busyness, while the second regression (1b) uses BUSYDIR and the third regression (1c) uses BUSYBD. The results of the first regression reveal a significant negative relationship between multiple directorships and advising performance (β = -0.001, p < .001). Although the coefficient is very small and the low R2 means that only a small portion of the variation of
26 third regression (1c) reveal no significant relationship between director busyness and monitoring performance, but the coefficients do have a negative sign as well (-0.005 and -0.004). All in all, the results partially support the first hypothesis that director busyness is negatively related to monitoring performance.
Table 5 – Busy directors and monitoring performance
Discretionary accruals (1a) (1b) (1c) Independent Variables #DIRSHPS -0.001*** (0.003) BUSYDIR -0.005 (0.008) BUSYBD -0.004 (0.004) BDSIZE -0.005*** (0.001) -0.005*** (0.001) -0.005*** (0.001) TENURE 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) DIRAGE 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) FSIZE 0.000** (0.000) 0.000** (0.0000) 0.000* (0.000) GRWTH 0,066*** (0.021) 0,067*** (0.021) 0,065*** (0.021) ROA 0.061*** (0.019) 0.062*** (0.019) 0.061*** (0.019) LEVERAGE 0.000* (0.000) 0.000* (0.059) 0.000* (0.000)
YEAR YES YES YES
INDUSTRY YES YES YES
Adjusted R2 0.051 0.051 0.051
F-statistics 7.530 7.516 7.572
Observations 2564 2564 2564
P-values are provided in parentheses.
The *, **, *** indicate level of significance at the 10%, 5%, and 1% levels. 4.3 Busy boards and advising performance
The second hypothesis, namely if director busyness is positively related to the advising performance of board members is tested with the three different regression models (2a, 2b, 2c). The same measures of
27 director busyness (#DIRSHPS, BSYDIR and BUSYBD) have been used as for testing the first hypothesis. The results of these regressions are reported in Table 6. First, the results of regression 2, reveal a positive relationship between multiple directorships and advising performance (β = 0.281, p < .001). The second (2b) and third (2c) regression also reveal a positive relation between director busyness and advising performance and thus are in line with the findings of the first regression. Both model 2b and 2c are significant, with model 2b showing the highest regression coefficient (0.615). To summarize, these three regressions all support the second hypotheses.
Table 6 – Busy directors and advising performance Tobin’s Q (2a) (2b) (2c) Independent Variables #DIRSHPS 0.281*** (0.050) BUSYDIR 0.615*** (0.153) BUSYBD 0.245*** (0.071) BDSIZE -0.121*** (0.014) -0.122*** (0.014) -0.125*** (0.014) TENURE 0.033*** (0.006) 0.033*** (0.006) 0.034*** (0.006) DIRAGE -0.028 (0.008) -0.027 (0.008) -0.028 (0.008) FSIZE 0.000*** (0.000) 0.000*** (0.000) 0.000*** (0.000) GRWTH -1.411*** (0.387) -1.452*** (0.388) -1.514*** (0.389) LEVERAGE 0.000 (0.001) 0.000 (0.001) 0.000 (0.001)
YEAR YES YES YES
INDUSTRY YES YES YES
Adjusted R2 0.172 0.167 0.166
F-statistics 27.630 26.710 26.451
Observations 2564 2564 2564
P-values are provided in parentheses.
28
4.4 Busy boards and the moderating effect of stock ownership
The moderating effect of stock ownership on the relation between busy directors and monitoring performance is analyzed next. The measures for director busyness (#DIRSHPS, BSYDIR, BUSYBD) are used in the moderated regression analysis 3a, 3b and 3c. The results of these regressions are reported in table 7. All three measures of director busyness do not reveal a significant effect of stock ownership on the relation between director busyness and monitoring performance. All the regression coefficients are lower than .001 and not significant at conventional levels. This means that there is no statistical evidence to support the third hypothesis
Table 7 – Busy directors and monitoring performance, modified by stock ownership Discretionary accruals (3a) (3b) (3c) Independent Variables #DIRSHPS -002 (0.003) BUSYDIR -001 (0.008) BUSYBD -004 (0.004) DIROWNSHP * #DIRSHPS 0.000 (0.000) DIROWNSHP * BUSYDIR 0.000 (0.000) DIROWNSHP * BUSYBD 0.000 (0.000) DIROWNSHIP 0.000 (0.000) 0.000** (0.000) 0.000** (0.000) BDSIZE -0.005*** (0.001) -0.005*** (0.001) -0.005*** (0.001) TENURE 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) DIRAGE 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) FSIZE -0.000** (0. 000) -0.000** (0.000) -0.000** (0.000) GRWTH 0.067*** (0.021) 0.067*** (0.021) 0.066*** (0.021) ROA 0.061*** (0.019) 0.061*** (0.019) 0.060*** (0.019) LEVERAGE 0.000* (0.000) 0.000* (0.000) 0.000* (0.000)
29
INDUSTRY Yes Yes Yes
Adjusted R2 0.051 0.051 0.051
F-statistics 7.012 7.037 7.042
Observations 2564 2564 2564
P-values are provided in parentheses.
The *, **, *** indicate level of significance at the 10%, 5%, and 1% levels.
4.5 Busy boards and the moderating effect of environmental complexity
The final three models test the fourth hypothesis, namely if environmental complexity moderates the relation between director busyness and the advising performance of board members. Again, #DIRSHPS (4a), BUSYDIR (4b) and BUSYBD (4c) are the variables used to measure director busyness in the different regression models. The results of the final three moderated regressions analysis are reported in Table 8. The variable BUSYBD reveals a significant moderating effect of environmental complexity on the relation between director busyness and advising performance. The significant (p = 0.072) positive regression coefficient of the interaction variable is 0.612. Using the other variables, #DIRSHPS and BUSYDIR a positive regression coefficient was found, this is in line with hypothesis 4. However, a statistically significant result was not found when using these variables, both p-values were higher than 0.327. Only if the dummy variable busy board is used to measure director busyness a moderating effect of environmental complexity is found, this means that the results partially supports the fourth hypothesis.
Table 8 – Busy directors and advising performance, modified by environmental complexity Tobin’s Q (4a) (4b) (4c) Independent Variables #DIRSHPS 0.268*** (0.051) BUSYDIR 0.589*** (0.153) BUSYBD 0.234*** (0.071) ENVCOMPLX * #DIRSHPS 0.264 (0.269) ENVCOMPLX * BUSYDIR 0.737 (0.756) ENVCOMPLX * BUSYBD 0.612* (0.340) ENVCOMPLX -0.610*** (0.135) -607*** (0.136) -615*** (0.136)
30 BDSIZE -0.123*** (0.014) -0.123*** (0.014) -0.126*** (0.014) TENURE 0.033*** (0.006) 0.034*** (0.006) 0.035*** (0.006) DIRAGE -0.029*** (0.008) -0.029*** (0.008) -0.029*** (0.008) FSIZE -0.000*** (0.000) -0.000*** (0.000) -0.000*** (0.000) GRWTH -1.739*** (0.394) -1.737*** (0.392) -1.765*** (0.392) LEVERAGE 0.000 (0.001) 0.000 (0.001) 0.000 (0.001)
YEAR Yes Yes Yes
INDUSTRY Yes Yes Yes
Adjusted R2 0.178 0.173 0.173
F-statistics 26.271 25.411 25.313,
Observations 2564 2564 2564
P-values are provided in parentheses.
31
5. Discussion
The purpose of this research is to find how director busyness impacts the monitoring and advising performance of the board of directors. The effect of multiple directorships on the performance of directors has generated debate among regulatory bodies and corporate reformers (Méndez et al., 2015; Kaczmarek et al., 2014) They are concerned that director busyness might be detrimental to the quality of the tasks a director has to carry out. Therefore, many regulatory frameworks have been tightened and governance codes are revised (Méndez et al., 2015). Furthermore, prior research that try to find a relation between director busyness and the performance of directors provide ambiguous results. Multiple directorships could negatively affect the performance of directors (Ferris et al., 2003; Fich & Shivdasani, 2006), but busy directors might also positively affect the performance of directors (Booth & Deli, 1996; Carpenter & Westphal, 2001; Loderer & Peyer, 2002).
To provide a more precise account of the relation between director busyness and the performance of directors, this study focuses on the monitoring and advising performance of directors, unlike most prior research that focused on the financial performances of firms. To test the developed hypotheses, regression analyses are performed. Data of S&P listed companies ranging from the year 2009 until 2016, resulting in a sample with 2564 firm-year observations, is used to investigate the theory-based expectations of monitoring and advising performance of busy directors.
Based on the results of the regression analyses, this study finds partial support for the expected association between director busyness and monitoring performance. One measure of director busyness shows a significant negative relation between director busyness and monitoring performance. This might be an indication that busy directors shirk their monitoring responsibilities since they have to divide the time they devote to their monitoring duties between multiple firms. These findings are in line with the expectations, based on the agency theory, that busy directors are not able to contribute sufficient attention and time to perform the monitoring duties. In line with the findings of this study, there is prior literature providing even stronger and more convincing evidence confirming that busy directors may not be effective monitors (Lipton & Lorsch, 1992; Fich & Shivdasani, 2006; Jiraporn et al., 2009). This might be due to the fact that this study used a more recent sample data than prior studies. If the measures of director busyness are compared to other studies, this study shows that directors are less busy. The average number of busy board members on the board is for example 30%, while Fich & Shivdasani (2006), who used panel data of U.S. firms from 1989 to 1995, report that on average 52% of the board members are busy. It might be the case that regulatory changes and revised codes of conduct over the past few
