Are all faultlines equal?

(1)

Are all faultlines equal?

Abstract: The purpose of this study is to examine the influence of faultline dimensions on

board effectiveness. Research regarding faultlines is based on unidimensional conceptualization. However, this might be inadequate. I therefore distinguish three faultline dimensions based on the information/decision-making perspective, categorization perspective and the factional perspective to answer the following research question: Is there a difference between the influences of demographic faultlines, task faultlines and factional faultlines on the effectiveness of the board? I tested the hypotheses through a panel data regression from 18.725 firm year observations, ranging from 2009 until 2015. Results indicate that not all faultlines are equal. Task faultlines have a positive influence on the advisory effectiveness, factional faultlines reduce the advisory effectiveness of the board and demographic faultlines have a positive influence on the monitoring effectiveness of the board. The results contribute to prior literature by dividing faultlines into dimensions based on the nature of the attributes leading to the faultlines. Moreover, the results also suggest that the influence of the faultline dimension depends on the task at hand. Implications, limitations and future research are also discussed.

Key Words: Faultlines, Board, Corporate Governance, US

Name: Meike Burgler

Student number: S2550997

Email: meikeburgler@gmail.com

Date: 25-07-2018

Subject: Master Thesis Accountancy (EBM869B20)

Supervisor RUG: Dr. N.J.B. Mangin

Co-assessor: Dr. R.A. Minnaar

Supervisor PwC: R. Somhorst MSc

(2)

1. INTRODUCTION

Stakeholders’ concerns for diversity in corporate governance keeps growing. This becomes clear when Blackrock announced that diversity in the board is going to be one of its top engagement priorities (Kerber, 2017). Regulation is following this shift in focus. In 2009, the U.S. Securities and Exchange Commission (SEC) decided to make it mandatory for organizations registered under section 12 of the Securities Exchange Act 1934, to clarify their diversity policy regarding the board through a proxy statement. There is a high possibility that these rules regarding diversity in the board will become even stricter, considering the SEC is planning to review their current rules (Bochner, 2016).

Pressure from stakeholders and legislation could be clarified due to the positive influence of diversity on the performance of the board (Adams & Ferreira, 2007; Forbes & Milliken, 1999; Klein, 2002; Tziner & Eden, 1985). Moreover, it also provides diverse opinions and task-related knowledge (van Knippenberg et al., 2004), which positively influence complex decision-making (Bantel & Jackson, 1989).

Despite a rich research history regarding board characteristics in corporate governance (Durisin & Puzone, 2009), prior literature is inconclusive about the influence of diversity on the effectiveness of the board. Diverse boards can also have a negative influence on the performance of the board (Jackson et al., 1991; Jackson, 1996; Jehn et al., 1999; Milliken & Martins, 1996; Williams & O’Reilly). Combining this ambivalence with the increasing pressure from legislation and shareholders raises the question whether and under which circumstances these pressures might be good or bad.

The theory of faultlines developed by Lau & Murnighan (1998) can help reconciling mixed expectations and observations about diversity. Lau and Murnighan (1998) define faultlines as “hypothetical dividing lines that may split a group into subgroups based on one or more attributes” (p. 328). The faultline theory measures the diversity in the group, while taking the alignment between different attributes into account (Bezrukova et al., 2009).

The faultline theory (Lau & Murnighan, 1998) is based on the mechanism that diversity in the boardroom leads to board members with aligned attributes, which could cause faultlines. Faultlines create subgroups, by which attributes of board members are aligned within subgroups, but attributes of board members between subgroups differ. For instance, if the board of your firm consists of four women, with an age of 61, 63 and 64. In this group, no faultlines

(4)

are created, because none of the attributes are aligned. However, if your board consists of four persons, two men with the age of 61 and 62 and two women with the age of 31 and 32, multiple attributes are aligned. The difference between the two is the alignment of the attributes. In the first board, none of the attributes are aligned. However, in the second board, the attributes sex and age are aligned and therefore this board is likely to create faultlines, which leads to dividing the board into subgroups.

These fractures between subgroups could have a negative influence on the performance of a group (Lau and Murnighan, 1998; Lau & Murnighan, 2005; Li & Hambrick, 2005). Faultlines cause stereotyping and prejudice (Jetten et al., 2000), conflicts and less cohesion in the group (Bezrukova et al., 2007; Li & Hambrick, 2005). Faultlines also have a negative influence on social integration (Rico et al., 2007). Furthermore, Gibson and Vermeulen (2003) found that individuals are less open for ideas from other subgroups, when faultlines arise.

So far, research regarding faultlines is mainly negative, however some research also suggests that the possible effects of faultlines may vary (Bezrukova & Uparna, 2009; Gibson & Vermeulen, 2003). Faultlines have a positive influence on creativity (Bezrukova & Uparna, 2009) and group learning (Gibson & Vermeulen, 2003). Moreover, task faultlines bring diverse skills to the group (Jonhson et al., 2013; Van Knippenberg et al., 2004), which could lead to complex decision solving (Bantel & Jackson, 1989).

Prior literature gives little insights into the influence of the nature of the attributes leading to these faultlines. However, it is possible that faultline dimensions based on gender of board members could differ from faultlines based on board tenure. Following the social categorization perspective (van Knippenberg & Schippers, 2007; Williams & O’Reilly, 1998) all faultlines are harmful. This perspective explains that members of a board divide others and themselves into subgroups based on aligned attributes. Moreover, individuals feel attracted to other board members with the same attributes. This behavior leads to prejudice and stereotyping (Jetten et al., 2000).

However, the information/decision-making perspective (Bezrukova, 2009; Dahlin et al., 2005; Webber & Donahue, 2001) suggests that the relativeness to the task of the attributes is important. Faultlines based on gender are not related to the task, but faultlines based on board tenure are related to the task. Following this perspective, these faultlines could have a different influence on the effectiveness of the board. Research regarding the different influences of these faultlines is still very sparse (Bezrukova, 2009).

(5)

The aim of this study is to determine whether the nature of attributes aligning to form a faultline matters, i.e. whether some alignments are beneficial while other alignment are harmful. Furthermore, the aim is also to determine whether the impact of the faultline differs depending on the nature of the task performed by the board member. Prior studies provide little insight on these matters. Providing insights could however help firms in composing their board.

Until now, prior literature primarily assumed that all faultlines in the boardroom are equally harmful (Kaczmarek et al., 2012; Veltrop et al., 2015). However, it is not clear how much and if the negative impact between faultline dimensions based on different nature of the attributes differs. By dividing the faultlines construct into three different dimensions based on the kind of the attributes aligning to form the faultline, it could be possible to acknowledge which faultlines firms should avoid and which attributes firms should align, to achieve a less negative or even positive influence on board effectiveness. Moreover, it could also create a better understanding of board composition. This is important, as “One of the key challenges for organizations today is to maximize a group’s ability to meet challenges and minimize process losses” (Thatcher et al., 2012, p. 970).

In this research, I divide the faultlines into the following three dimensions: demographic faultlines, task faultlines and factional faultlines. Attributes leading to demographic faultlines are attributes that could be seen, like age and gender. Attributes leading to task faultline are attributes that are task-related, like board tenure. Lastly, factional faultlines are based on individuals that come in factions, who represent a social entity, like executive and supervisory directors. This leads to the following research question:

Is there a difference between the influences of demographic faultlines, task faultlines and factional faultlines on the effectiveness of the board?

I expect that demographic faultlines and factional faultlines reduce board effectiveness, based on the social categorization perspective. However, I expect no association between task faultlines and board effectiveness, because prior literature is inconclusive regarding the influence of task faultline on group performance. I test these hypotheses on a sample of 18.725 firm year observations about 3.979 U.S. firms from 2009 until 2015. The findings suggest that not all faultlines are equally harmful. Some faultlines even have a positive influence on the effectiveness of the board. More specific: factional faultlines reduce advisory effectiveness and task faultlines have a positive influence on the advisory effectiveness of the board. Secondly, demographic faultlines have a positive influence on the monitoring effectiveness of the board.

(6)

The findings of this research are relevant for the literature of corporate governance. First, it contributes to the theory of faultlines by showing that not all faultlines are necessarily harmful in all situations. Secondly it suggest that the nature of the attributes aligning as well as the task at hand need to be taken into account.

In the following section, I will give a description of the theoretical background including the hypotheses. In the third section, I will explain the method, which leads to the results in the fourth section. In the last section, I will give a conclusion, where after I discuss the results and will give recommendation for future research.

(7)

2. THEORETICAL BACKGROUND

2.1 Function of the board

The Board can be seen as “the common apex of the decision control systems of the organizations” (Fama & Jensen, 1983, p. 311). The functions of the board can be divided into a monitoring function (Fama & Jensen, 1983; Hillman & Dalziel, 2003) and an advisory function (Herman, 1981; Hillman & Dalziel, 2003; Mace 1986).

The monitoring function of board effectiveness is based on the agency theory (Fama & Jensen, 1983) which originated because of the separation between shareholders and managers. Following the agency problem, monitoring effectiveness is important, to ensure that the managers make decisions in the best interest of the shareholders (Beatty & Zajac, 1994), which includes preventing earnings management.

Healy and Wahlen (1999) suggest that “earnings management occurs when managers use judgment in financial reporting and in structuring transactions to alter financial reports to either mislead some stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers” (p. 368). High information asymmetry creates possibilities for earnings management (Richardson, 2000), because information asymmetry implies that managers possess firm specific information that is not available for shareholders. Monitoring effectiveness is important, because this reduces earnings management (Richardson, 2000).

Secondly, the advisory function of board effectiveness is based on the resource dependency theory (Pfeffer & Salancik, 1978). The advisory role of the board implies that it is the board’s task to advise and support the managers (Lorsch & Maciver, 1989).

Providing resources plays a big role in the advisory function (Pfeffer & Salancik, 1978). Prior literature (Wernerfelt, 1984) argues that the board member can support the firm through providing the following benefits: assistance from external relations, communication lines with external relations, advice and legitimacy. Executing the advisory task has a positive influence on firm performance (Hillman & Dalziel, 2003), because it leads to less uncertainty and less transaction costs (Williamson, 1984; Pfeffer, 1972).

(8)

2.2 Diversity in the board

The board’s ability to fulfill these two functions relies on the diversity of director’s attributes. Diversity can be defined as “the presence of differences among members of a social unit” (Jackson et al., 1995, p. 217).

Diversity regarding executive and supervisory board members leads to board independence. Board independence is known for its monitoring effectiveness, which positively influences board effectiveness (Klein, 2002).

Moreover, board members with diverse networks and diverse backgrounds bring different resources to the group, which leads to creativity and complex decision solving and therefore increases advisory effectiveness (Bantel & Jackson, 1989; Bezrukova & Uparna, 2009; Tziner & Eden, 1985; van Knippenberg et al., 2004). Secondly, diverse boards creates possibility to advice managers, while taking the diverse marketplace into account (Carter et al., 2003). Chatman et al., (1998) argue that demographic diversity leads to increasing productivity, which may influence both the monitoring and advisory effectiveness. Moreover, use of diverse knowledge and expertise can positively influence board effectiveness, because it creates the possibility to make informed decisions and to cope with complex decisions (Adams & Ferreira, 2007; Forbes & Milliken, 1999).

While the board’s ability to fulfill its function is based on the diversity of director’s attributes, prior literature argues that diversity can also have a negative influence of board effectiveness. Diversity leads to communicating problems (Triandis, 1960). Furthermore, Jehn et al., (1999) found that diversity leads to less satisfaction, decreases the intent to remain in the firm and decreases the commitment to the firm based on the argumentation that diversity in the board could cause conflicts between board members. Moreover, diversity that leads to conflict leads to less contentment (Jackson, 1996). Diversity regarding demographic attributes leads to a higher turnover rate (Milliken & Martins, 1996), which could also lead to lower group integration (Jackson et al., 1991). Moreover, diversity creates weakness. A weak group is more vulnerable to be negatively influenced by disturbing influences (Williams & O’Reilly, 1998).

(9)

2.3 Faultline theory

The concept of faultlines can help reconciling mixed expectations and findings about the impact of diversity. The concept of faultlines is an addition to prior research regarding diversity. It does not only consider the presence or frequency of attributes, but also their alignment (Bezrukova et al., 2009). Aligned attributes create ”hypothetical dividing lines” (Lau & Murnighan, 1998, p. 328) between board members, which are also known as faultlines. Faultlines lead to aligned attributes within the subgroups, but the attributes differ between subgroups. So, through faultlines a distinction is made between in-groups and out-groups (Ashforth & Mael, 1989). Moreover, not only diverse attributes and alignment of attributes are taken into consideration, but also the possibility of numerous subgroups and the number of subgroups.

Lau and Murnighan (1998, 2005) created the theory by suggesting that faultlines may be detrimental for a workgroup or a team, based on the social categorization perspective (van Knippenberg & Schippers, 2007; Williams & O’Reilly, 1998). The social categorization perspective primarily exists out of the social identity theory (Tajfel, 1978), the similarity attraction paradigm (Byrne, 1971) and the self-categorization theory (Turner, 1982, 1985). This is a process (Turner, 1982, 1985) by which members of a team create a social identity based on their own subgroup (Tajfel, 1978). This leads to team members who divide themselves and others into subgroups based on attributes (Turner, 1985; Tajfel, 1981), whereby members with the same attributes like each other more than others (Byrne, 1971). This creates behavior like stereotyping and prejudice (Jetten et al., 2000). Moreover, this leads to people avoiding communication with members of other groups, because of different opinions (Rosenbaum, 1986).

Prior literature agrees with the suggestion from Lau and Murnighan (1998) that all faultlines are harmful. Bezrukova et al., (2007) explain that faultlines in a group leads to conflicts, because faultlines causes disagreement about ideas between members of a group. This is based on the explanation that faultlines leads to subgroups supporting the ideas in their own subgroup (Wildschut et al., 2002), but that members of teams are less open for new ideas from members from any other subgroup (Gibson & Vermeulen, 2003). Members from different subgroups do not feel related to each other and therefore prefer ideas from their own subgroup. Where after, Rico et al., (2007) found that faultlines have a negative influence on the social integration of the group, through the mechanism that faultlines leads to more connections within subgroups

(10)

than in the whole group. This leads to individuals identifying themselves with their subgroup instead of the whole group (Lau & Murnighan, 2005.)

2.3.1 Faultline dimensionality

The current unidimensional conceptualization of faultline might be inadequate. What may differentiate between faultlines having a positive or negative influence on the effectiveness of the board could depend on the nature of the aligned attributes leading to the faultlines. I therefore divide faultlines into three dimensions based on the nature of the attributes leading to the faultlines. The following three faultline dimensions are included in this research:

Demographic faultlines

Demographic faultlines are based on attributes that could be seen and are obvious to the eye of the board members. These faultlines are based on the social categorization perspective (van Knippenberg & Schippers, 2007; Williams & O’Reilly, 1998). Therefore, these faultlines divide a group into subgroups based on demographic attributes like age, sex and race (Jehn et al., 1999).

Task faultlines

Task related faultlinescan be defined as “differences in knowledge bases and perspectives that members bring to the group’’ (Jehn et al., 1999, P. 743). Task faultlines associate with the information-/decision-making perspective. This perspective argues that attributes that are task related or informational (Bezrukova et al., 2009) or attributes that are important to carry out tasks (Jackson et al., 2003) could lead to task faultlines. These faultlines are for instance based on attributes like board tenure.

Factional faultlines

While taking the concept of Lau and Murnighan (1995) into account, Li & Hambrick (2005) argue that board members come in faction, but not as independent individuals. Factions can be defined as: “groups in which members are representatives, or delegates, from a small number of (often just two) social entities and are aware of, and find salience in, their delegate status” (Li & Hambrick, 2005, p. 794). These factions come with aligned attributes and create faultlines with other factions or members of the board. For instance, board members of merged

(11)

firms are still likely to feel divided into two sections, which causes a faultline and leads to subgroups in the board.

As mentioned, prior literature regarding the influence of diversity on board effectiveness is of the attributes aligning to form the faultline, it could be possible to acknowledge which faultlines firms should avoid and which attributes firms should align, to achieve a less negative or even positive influence on board effectiveness. The hypotheses regarding the influence of these faultlines dimensions on the board effectiveness will be given in the following section.

(12)

3. HYPOTHESES DEVELOPMENT

Following the reasoning developed in prior literature, I expect demographic faultlines to reduce the effectiveness of the board. Demographic faultlines are based on the social categorization perspective (van Knippenberg & Schippers, 2007; Williams & O’Reilly, 1998). The social categorization perspective argues that faultlines lead to in-group versus out-groups, which causes that members favor their in-group over other out-groups (Byrne, 1971). Furthermore, this leads to prejudice and stereotyping (Jetten et al., 2000). Lau and Murnighan (1998) used demographic attributes in their research and claimed based on these attributes, that faultlines are detrimental because they foster conflicts and communication problems. Moreover, faultlines leads to conflicts, members of a team are less open for idea’s from other outgroups and it has a negative influence on the social integration of the group (Bezrukova et al., 2007; Gibson & Vermeulen, 2003; Rico et al., (2007). Li & Hambrick (2005) support the findings of Lau & Murnighan (1998), as they find that demographic faultlines could cause groups performing inefficiently and therefore reduce the performance of a group (Li and Hambrick, 2005). This leads to the following hypotheses:

H1: Demographic faultlines reduce board effectiveness.

Because of opposite processes at work and mixed results, I expect that there is no association between task faultlines and board effectiveness. In addition to the harmful influence of faultlines, there are also other mechanisms at work, which might counteract the negative impact by a positive one. The information/decision-making perspective considers the task-relativeness of the attributes (Bezrukova, 2009; Dahlin et al., 2005; Webber & Donahue, 2001). Task-related attributes could bring diverse skills and information to the board (Jonhson et al., 2013; Van Knippenberg et al., 2004) which could lead to complex decision solving (Bantel & Jackson, 1989) by challenging each other based on different opinions (Eisenhardt et al., 1997). Moreover, task faultlines have a positive influence on creativity (Bezrukova & Uparna, 2009) and group learning (Gibson & Vermeulen, 2003). However, Webber and Donahue (2001) found no association between task-related faultlines and group performance. Prior literature is inconclusive. I therefore predict the following hypothesis:

H2: Task faultlines are not associated with the board effectiveness.

I expect that factional faultlines reduce board effectiveness. Factional faultlines are primarily based on the concept of Lau and Murnighan (1998) regarding faultlines. Factional faultlines

(13)

also consider that some board members do not come as individuals to the board, but in factions, whereby the members represent a social entity. Representing this social entity is an aligned attribute, which leads to the faultline. The influence of factional faultlines is also based on the social categorization perspective (van Knippenberg & Schippers; Williams & O’Reilly, 1998). I therefore expect that factional faultlines also have a negative influence on board effectiveness, based on the same argumentation as for demographic faultlines. This leads to the following hypothesis:

(14)

4. METHODOLOGY

4.1 Sample and data collection

To test whether faultlines dimensions have a different influence on the board effectiveness, a wide range of attributes regarding board members were needed. This sample is therefore based on data available through BoardEx, which contains data regarding firms from the United Stated. The initial sample size was 105.675 firm year observations. However, 68.167 firm-year observations were eliminated because of missing information regarding the attributes leading to the faultlines (see Table 1). Secondly, 5.629 firm-years from financial institutions and 2.657 firm-years from regulated industries were excluded from the sample (SIC 6000-6500 and 4400-5000). Thereafter financial data needed is obtained from COMPUSTAT. After matching the BoardEX data with the data from COMPUSTAT another 10.496 firm year observations were eliminated, because of missing data that was needed to calculate DACCRUALS and TOBINSQ. The remaining sample consists out of 18.726 firm year observations (See table 1). The data obtained contains a selection from the years 2009 up to and including 2016, to avoid the shock of the financial crisis in 2008. All variables calculated and used in this research are displayed in Appendix 1.

4.2 Dependent variables

Advisory

Following prior literature, I operationalize board effectiveness with Tobins’Q. Prior literature argues that Tobins’Q shows the perceived value of shareholders regarding the effective use of

Table 1 Sample

Criteria Observation

Total firm-year observations 105.675

Less missing information attributes (68.167)

Less financial institutions (5.629)

Less regulated industries (2.657)

Less missing information monitoring/advisory (10.496)

(15)

assets in the organization (Gompers et al., 2003). It therefore captures, to some extent, the perceived effectiveness of the advisory responsibility of the board. Tobin’s Q can be measured through formula 1.

(𝟏) (𝑬𝒒𝒖𝒊𝒕𝒚 𝑴𝒂𝒓𝒌𝒆𝒕 𝒗𝒂𝒍𝒖𝒆 + 𝑳𝒊𝒂𝒃𝒊𝒍𝒊𝒕𝒊𝒆𝒔 𝑴𝒂𝒓𝒌𝒆𝒕 𝒗𝒂𝒍𝒖𝒆)

(𝑬𝒒𝒖𝒊𝒕𝒚 𝒃𝒐𝒐𝒌 𝒗𝒂𝒍𝒖𝒆 + 𝑳𝒊𝒂𝒃𝒊𝒍𝒊𝒕𝒊𝒆𝒔 𝒃𝒐𝒐𝒌 𝒗𝒂𝒍𝒖𝒆) Monitoring

Earnings management creates possibilities to mask choices of managers, which is not in the best interest of the shareholders (Leuz et al., 2003). One of the main aspects of the monitoring effectiveness is to align these decisions of the managers with the interests of the shareholders. High monitoring effectiveness, leads to low earnings management and less discretionary accruals (Peasnell et al., 2005).

I follow prior literature when using discretionary accruals as a proxy for earnings management (Peasnell et al., 2005; Xie et al., 2003). Discretionary accruals are measured by using modified Jones model (Dechow et al., 1995) which is based on the model developed by Jones (1991). A limitation of this model is that it assumes that revenues are nondiscretionary, which leads to a biased measure of earnings management (Dechow et al., 1995), therefore Dechow et al. (1995) modified the model of Jones (1991) which leads to the following measurement:

First, the total accruals are measured (see formula 2).

(𝟐) 𝑻𝑨𝑪𝑪𝒕 = ∆𝑪𝑨𝒕− ∆𝑪𝒂𝒔𝒉 − ∆𝑪𝑳𝒕+ ∆𝑫𝑪𝑳𝒕− 𝑫𝑬𝑷𝒕

TACCt is total accruals in year t, ∆CAt is change in current assets in year t, ∆Cash change in

cash and equivalents in year t, ∆CLt change in current liabilities in year t, ∆DCLt change

short-term debt included in current liabilities in year t, DEPt depreciation and amortization costs in

year t.

After the TACCt are measured, the Modified Jones Model (Dechow, 1995) suggests to estimate

the firm specific parameters (α1, α2 and α3) through an OLS regression (see formula 3):

(𝟑) 𝑻𝑨𝑪𝑪𝒕 𝑨_𝒕−𝟏 = 𝒂𝟏 𝟏 𝑨_𝒕−𝟏+ 𝒂𝟐 (∆𝑹𝑬𝑽_𝒕− ∆𝑹𝑬𝑪_𝒕) 𝑨_𝒕−𝟏 + 𝒂𝟑 𝑷𝑷𝑬_𝒕 𝑨_𝒕−𝟏 + 𝜺𝒕

A(t-1) is total assets in year minus 1, ∆REVt is change in revenues in year t, ∆RECt is change in

net receivables in year t, PPEt is gross property plant and equipment in year t. The modified

(16)

non-discretionary accruals and a non-discretionary accruals component. The non-non-discretionary accruals are measured through formula 3. The accruals that could not be explained by this formula are determined as εt, the error term (see formula 3). The error term represents the discretionary

accruals in year t.

Prior research (Dechow & Dichev, 2002) is followed by calculating the absolute value of discretionary accruals. The absolute value is used because earnings management can be managed upwards and downwards (Arun et al., 2015).

4.3 Independent variable

In this research, the faultlines are divided into three dimensions: demographic faultlines, task faultlines and factional faultlines. The attributes leading to these faultlines and the categorization of the attributes are displayed in Table 2. The definitions of the faultlines dimensions are mentioned in the theory section of this research. The demographic faultlines contain attributes that could be seen by other board members; time to retirement, age, gender & nationality. The task faultlines contain attributes directly related to the task; years in role, years in the board, member of audit committee, member of remuneration committee and member of nomination committee. Lastly, the factional faultlines are based on factions that come to the group. These factions are representatives of social entities. I therefore argue that executive directors and supervisory directors come as factions. Moreover, I argue that directors with aligned network sizes comes as factions to the board, because each categorization form the network size represents a social entity. Lastly, number of qualifications are also perceived as factions. Each category regarding the number of qualifications is familiar with the same standards and social rules and therefore represent a social entity.

(17)

The method of Shaw (2004) is used to measure the faultlines. Shaw (2004) measured through the alignment and deviation of attributes within and between subgroups. The internal alignment of attributes in a subgroup is measured as IA and the alignment of attributes between different subgroups is measured through CGAI (cross-subgroups alignment index). Where after FLS is measured through formula 4. The outcome of the faultlines measure (FLS) is negatively correlated with CGAI and positively correlated with IA. So the higher FLS, the greater the internal alignment of attributes in subgroups and the lower the alignment of attributes across subgroups. FLS is a measure between 0 and 1. This approach is measured using the R package asw cluster (Meyer & Glenz, 2013; Meyer & Glenz, 2018).

(𝟒) 𝑭𝑳𝑺 = 𝑰𝑨 ∗ (𝟏 − 𝑪𝑮𝑨𝑰) Table 2 Faultlines

Faultlines Attributes Categories

Demographic faultlines

Time to retirement, 17 categories. 5 years per category

Age 17 categories: 5 years per category

Gender 2 categories: Male and Female

Nationality 72 categories: 1 country per category

Task faultlines Years in role 11 categories

Years in the board 14 categories

Member of audit committee 2 categories: Member and no member Member of remuneration committee 2 categories: Member and no member Member of nomination committee 2 categories: Member and no member

Factional faultlines Role in the board 2 categories: Executive director and

supervisory director

Director network size 19 categories : 1000 individuals per category

Number of qualifications 15 categories. Categories go up per qualification.

(18)

4.4 Control variables

I introduce several variables in the regressions models to control for the effect on board effectiveness. The control variables are divided into firm specific and board specific variables. Board specific control variables

First, the size of the board (BOARDSIZE) is included, because the size of the board influence the knowledge available in the board and therefore positively influence the effectiveness of the board (Taylor & Greve, 2006). However, the size of the board can also negatively influence the effectiveness of the board because of free-riding behavior and coordination issues (Boone et al., 2007).

Furthermore, the independence of board (BINDEP) is also taken into account. Executive board members can have a negative influence on the board effectiveness, because of conflicting interests (Fama & Jensen, 1983). Independent board members are introduced to prevent any agency problems and reduce any agency costs through their monitoring function (Fama & Jensen, 1983). Empirical research support this theory, where after Klein (2002) found that board independence has a positive effect on the monitoring effectiveness of the board.

At last, I have also included the diversity of the board (DIVERSITY) as a control variable, because I do not want to measure the influence of diversity on board effectiveness in this research. The method of Bezrukova et al., (2010) is followed to calculate the diversity control. The blau index (Blau, 1977) is used for the following categorical attributes: qualifications, role in the board, gender, nationality, member of nomination committee, member of audit committee and member of remuneration committee. The coefficient of variation (Allison, 1978) is used for the continuous attributes: network of director, years in role, and years to retirement, age and average years on other quoted boards. Where after the entropy based index (Jehn et al., 1999) is used to compute the overall diversity score.

Firm specific control variables

First, I control for firm size, because smaller firms tend to report less accurate than bigger firms (Brick & Chidambaran, 2010) and therefore report more discretionary accruals (Xie et al, 2003). I therefore control for book value to total assets (FSIZE) (Xie et al, 2003) and expect that FSIZE has a negative influence on the effectiveness of the board.

(19)

Secondly, I also control for leverage (LEVERAGE), because higher leverage leads to more monitoring. High leverage is related to manager/shareholder conflicts (Jensen & Meckling, 1976; Minnis, 2011). Firms with little equity have incentives to engage in asset substitutions through high-risk projects that could be harmful for shareholders (Jensen & Meckling, 1976). Therefore, high leverage leads to more monitoring.

Return on Assets (ROA) is also a control variable, because firms with high earnings are inclined to ‘smooth’ their earnings (Dechow et al., 1995). Earnings smoothing leads to higher discretionary accruals, which leads to a negative influence on the effectiveness of the board.

4.5 Data analysis

A panel data regression with year and firm fixed effects is used to test the hypotheses. Model 5 is used to test the relation between the faultline dimensions and board effectiveness. As already mentioned, the board effectiveness (BOARDEFF) is measured through ABSDACCRUALS and TOBINSQ. An explanation of other variables used in this model is available in Appendix 1. To support the first hypothesis the coefficient of SDFAULT should have a positive and significant influence on ABSDACCRUALS and negative and significant influence on TOBINSQ. To support the second hypothesis the coefficient of TRFAULT should have insignificant influence on ABSDACCRUALS and TOBINSQ. To support the third hypothesis the coefficient of FACFAULT should have a positive and significant influence on ABSDACCRUALS and negative and significant influence on TOBINSQ.

(𝟓)𝑩𝑶𝑨𝑹𝑫𝑬𝑭𝑭

= 𝜷_𝟎+ 𝜷_𝟏𝑺𝑫𝑭𝑨𝑼𝑳𝑻 + 𝜷_𝟐𝑻𝑹𝑭𝑨𝑼𝑳𝑻 + 𝜷_𝟑 𝑭𝑨𝑪𝑭𝑨𝑼𝑳𝑻

+ 𝜷_𝟒𝑩𝑶𝑨𝑹𝑫𝑺𝑰𝒁𝑬 + 𝜷_𝟓𝑳𝑬𝑽𝑬𝑹𝑨𝑮𝑬 + 𝜷_𝟔𝑹𝑶𝑨 + 𝜷_𝟕𝑭𝑺𝑰𝒁𝑬

(20)

5. RESULTS

5.1 Descriptive statistics

The descriptive statistics of the variables used are displayed in Table 3.The average firm of this sample has ASSETS of €4123,-. The average BOARDSIZE is 8 and therefore in line with prior literature (Bezrukova et al., 2012; Kaczemerak, 2012; Li & Hambrick, 2005; Tuggle et al., 2010). The average TobinsQ of 7 is rather high compared to prior literature regarding faultlines (Kaczemerak, 2012). This could be explained because of the different period used in the research of Kaczemark (2012). Kaczemark (2012) used a sample based on firm years between 1999 until 2008, which includes the shock of the financial crisis. The sample used in this research avoids the shock from the financial crisis.

Table 3 shows that all variables used in this analysis have enough variation. Prior to the analyses of the data, data is transformed to create results that are more reliable, because the regressions used assume normal distributions regarding the variables. As already mentioned in the method section, I calculated the the absolute value of DACCRUALS. The skew of LEVERAGE and ABSDACCRUALS was rather high (see table 3). I have therefore calculated the natural logarithm from LEVERAGE and ABSDACCRUALS, which creates normal distributions. To remove any disturbance of outliers on the results of our analysis, I have winsorized ABSDACCRUALS, TOBINSQ, LEVERAGE, ROA and DIVERSITY with 5%. Lastly, all variables are standardized to facilitate the interpretation of coefficients.

The Pearson correlation matrix (Table 4) is used to test all variables for multicollinearity. The results of the Pearson correlation analysis showed that there is no major threat of multicollinearity as all values are below the absolute value of 0,593.

(21)

Table 3 Descriptive statistics

VARIABLES MEAN MEDIAN SD SKEW MIN MAX

DACCRUALS 0,372 0,246 28,180 100,723 -1.388,40 3.555,995 ABSDACCRUALS 0,720 0,279 28,172 113,931 0 3.555,995 TOBINSQ 7,071 3,148 29,442 48,405 -636,837 2.602,154 FAULTLINES 0,093 0,091 0,033 1,536 0 0,410 TRFAULT 0,141 0,120 0,095 1,565 0 0,698 FACFAULT 0,081 0,075 0,065 4,161 0 0,806 SDFAULT 0,041 0,041 0,038 1,166 0 0,417 BOARDSIZE 7,990 8 2,209 0,455 1 18 LEVERAGE 1,569 0,483 103,839 133,312 0 14.081,5 ROA -0,533 0,029 55,535 -133,139 -7.526 219,423 ASSETS 4.122,50 610,128 19.328,83 21,133 0,001 781.818 BINDEP 0,759 0,8 0,137 -1,332 0 1 DIVERSITY 2,53 -0,714 21,81 25,63 -1,59 1.139,1

(22)

Table 4 Pearson correlation matrix

ABSDACCRUALS TOBINSQ TRFAULT SDFAULT FACFAULT FAULT BOARDSIZE LEVERAGE ROA FSIZE BINDEP DIVERSITY ABSDACCRUALS 1,000 TOBINSQ -0,0105 1,000 TRFAULT -0,024*** 0,022** 1,000 SDFAULT -0,035*** -0,073*** -0,109*** 1,000 FACFAULT 0,009 -0,034*** 0,009 0,257*** 1,000 FAULT -0,045*** -0,029*** 0,561*** 0,547*** 0,514*** 1,000 BOARDSIZE -0,014* -0,146*** -0,157*** 0,521*** 0,186*** 0,332*** 1,000 LEVERAGE 0,036*** -0,593*** -0,064*** 0,141*** 0,045*** 0,047*** 0,228*** 1,000 ROA -0,088*** -0,128*** -0,042*** 0,130*** 0,070*** 0,084*** 0,233*** -0,027*** 1,000 FSIZE -0,031*** -0,071*** 0,017** 0,026*** 0,023*** 0,0510*** 0,058*** -0,244*** 0,348*** 1,000 BINDEP -0,086*** -0,012* -0,218*** 0,228*** 0,042*** -0,029*** 0,241*** 0,031*** 0,056** -0,036*** 1,000 DIVERSITY 0,035*** -0,0118 0,063*** -0,107*** 0,008 0,0015 -0,052 -0,053*** 0,029*** 0,021 -0,127*** 1,000 *** Correlation is significant at 1% level

** Correlation is significant at 5% level * Correlation is significant at 10% level

(23)

5.2 Regression results

First, I executed a Hausman test, to validate the fixed effects in the panel data regression regarding DACCRUALS (χ2 _{= 200,75 , p = 0,000) and TOBINSQ (χ}2 _{= 385,13 , p = 0,000). Table 5 presents the}

results regarding the TOBINSQ regressions and table 6 presents the results regarding the DACCRUALS regressions.

Control variables

Model 1 of table 5 and table 6 presents the regressions with only the control variables. Where after I can observe that LEVERAGE (β = -0,062 , p < 0,01), ROA (β = 0,0855, p < 0,01), SIZE (β = -0,0536 p < 0,01) and BOARDINDEP (β = -0,0257, p = 0,03) have a significant influence on ABSDACCRUALS. However, only LEVERAGE (β = -0,6238, p < 0,01), ROA(β = 0,0477, p < 0,01), SIZE (β = -0,0979 p < 0,01) and DIVERSITY (β = -0,0121, p = 0,06), are significant in regard to TOBINSQ. All significant control variables were significant in the negative or positive direction that was expected, based on my justification in the method section.

Faultlines

Model 2 off both Table 5 and 6 present the results regarding the influence of faultlines which include all attributes on board effectiveness. The coefficient of Table 5 suggests that FAULT has a non-significant influence on the ABSDACCRUALS (β = -0,0141, p = 0,108). Moreover, the coefficient of Table 6 shows an insignificant effect of FAULT on TOBINSQ (β = 0,0018 , p = 0,778). The results suggest that the overall faultlines have no influence on the effectiveness of the board.

Demographic faultlines

Model 3 off both Table 5 and 6 present the results regarding the hypotheses. The first hypothesis states that demographic faultlines reduce board effectiveness. The coefficient of Table 5 suggests that SDFAULT has a significant negative influence on the ABSDACCRUALS (β = -0,0273 , p < 0,01) and therefore H1 is not supported. The coefficient of Table 6 shows an insignificant effect of SDFAULT on

TOBINSQ (β = -0,0003 , p = 0,96) which also not supports H1. Task faultlines

The second hypothesis states that task faultlines has no association with board effectiveness. The coefficient of Table 5 suggests that TRFAULT has an insignificant influence on the ABSDACCRUALS (β = -0,0128 , p = 0,13), which supports H2. However, the coefficient of Table 6 shows a positive

(24)

Factional faultlines

The third hypothesis states that factional faultlines reduce board effectiveness. The coefficient of Table 5 suggests that FACFAULT has a insignificant influence on the ABSDACCRUALS (β = 0,0126 , p = 0,13), which does not support H3. However, the coefficient of Table 6 shows a negative significant effect

of FACFAULT on TOBINSQ (β = -0,0117 , p = 0,05) and therefore supports H3.

Table 5 regression results ABSDACCRUALS ABSDACCRUALS Model 1 ABSDACCRUALS Model 2 ABSDACCRUALS Model 3 FAULT -0,0141 SDFAULT -0,0273*** TRFAULT -0,0128 FACFAULT 0,0126 BOARDSIZE -0,0202 -0,0131 -0,0152 LEVERAGE -0,0359*** -0,0355*** -0,0349*** ROA 0,0852*** 0,0855*** 0,0852*** SIZE -0,0534*** -0,0532*** -0,0527*** BOARDINDEP -0,0260** -0,0279** -0,0263** DIVERSITY -0,0121* -0,0121* -0,0124* Constant -0,0224*** -0,0222*** -0,0221*** R-squared 0,0081 0,0083 0,0090 F-value 6,1391*** 7,1391*** 9,1391***

*** Correlation is significant at 1% level ** Correlation is significant at 5% level * Correlation is significant at 10% level N = 18.725

(25)

5.3 Additional analysis

Overstatement discretionary accruals

The absolute value of discretionary accruals is used as a proxy for monitoring effectiveness in this research. The absolute value of discretionary accruals considers equivalent downward an upward earnings management. It therefore informs about the reliance on accruals to smooth earnings. However, the non-absolute value of the discretionary accruals informs about the tendency to understate or overstate earnings. I have therefore ensured that the results regarding the monitoring aspect of board effectiveness were the same for both the absolute and non-absolute value of discretionary accruals. All significant results were in line with prior results. SDFAULT (β = -0,0214 , p = 0,04) has a negative influence on the non-absolute value of DACCRUALS. This suggest that the initial results are driven by positive discretionary accruals, i.e. potential overstatements.

Table 6 regression results TOBINSQ TOBINSQ Model 1 TOBINSQ Model 2 TOBINSQ Model 3 FAULT 0,0018 SDFAULT -0,0003 TRFAULT 0,0113* FACFAULT -0,0115* BOARDSIZE -0,0127 -0,0136 -0,0088 LEVERAGE -0,6238*** -0,6238*** -0,6241*** ROA 0,0477*** 0,0477*** 0,0478*** SIZE -0,0979*** -0,0979*** -0,0978*** BOARDINDEP 0,0048 0,0051 0,0072 DIVERSITY 0,0012 0,0012 0,0014 Constant -0,0328*** -0,0328*** -0,0329*** R-squared 0,2879 0,2879 0,2883 F-Value 6,1391*** 7,1391*** 9,1391***

*** Correlation is significant at 1% level ** Correlation is significant at 5% level * Correlation is significant at 10% level N = 18.725

(26)

6. CONCLUSION AND DISCUSSION

6.1 Conclusion

The purpose of this research is to determine whether there is a difference between the influence of demographic faultlines, task faultlines and factional faultlines on board effectiveness. Pressure in the boardroom regarding diversity is still rising, because of demanding stakeholders and changing legislation. While stakeholders and society desire diversity, prior literature regarding the influence of diversity on other aspects in the boards is inconclusive. More research is needed to create understanding and to make informed decisions regarding the influence of diversity. The faultline theory (Lau & Murninghan, 1998) provides an explanation able to reconcile prior mixed findings regarding diversity. Moreover, the theory of faultlines brings complementary information regarding the composition and the structure of a group, which is not captured by traditional diversity metrics prior literature focuses on. The faultline theory suggests that alignment of attributes in a group, leads to subgroups. Prior literature argues that all faultlines are harmful, irrespective of the attributes, which actually contribute to their formation (Lau & Murnighan, 1998). Prior literature gives little insight in the influence off the nature of the attributes leading to these faultlines is. By dividing faultlines into different dimensions based on the nature of the attributes leading to the faultlines, it is possible to evaluate the influence of different faultline dimensions on board effectiveness. In this research, I therefore included the nature of the attributes and divided faultlines into three dimensions: demographic faultlines, task faultlines, and factional faultlines.

I expected demographic faultlines and factional faultlines to reduce board effectiveness. I did not expect an association between task faultlines and board effectiveness. I tested these hypotheses with a panel data regression with fixed effects on a sample of 18725 firm year observations from the US in the period 2009 until 2016. Board effectiveness is measured through the advisory and monitoring role of the board. The findings suggest that factional faultlines have a negative influence on the advisory aspect of board effectiveness, which is in line with prior findings that faultlines have a negative influence on the performance of a group (Lau & Murnighan, 1998). Factional faultlines are among others based on the deviation between supervisory directors and non-executive directors. This negative influence from factional faultlines on advisory effectiveness could be explained through the theory of friendly boards (Adams & Ferreira, 2007). This theory explains that the non-executive directors depend on the executive directors for firm specific information. The CEO knows that if they share the information, it will positive influence the advisory effectiveness. However, they also know that if they share the information, the

(27)

supervisory directors will monitor more extensively. The non-executive directors could therefore be reluctant to share this information, which leads to a lower advisory effectiveness.

The findings also suggest that task faultlines have a positive influence on the advisory aspect of board effectiveness. This means that the alignment of attributes like years in role, years in the board, being a member of the audit committee, a member of the remuneration committee and a member of the nomination committee have a positive influence on the Tobins’Q of the firm. This could be explained through the nature of the attributes leading to the task faultlines. Task faultlines are based on attributes that are relevant for board members to complete their task, which is based on the information-/decision-making perspective (Bezrukova, 2009; Dahlin et al, 2005; Webber & Donahue, 2001). Faultlines based on task-related attributes means that the subgroups are based on aligned attributes relating to the task, but that these attributes differ between subgroups. The faultlines bring diverse skills to the group (Johnson et al, 2013). Each subgroup could have a different opinion regarding a complex issue or decision. The subgroups could challenge each other based on these different opinions (Eisenhardt et al, 1997), which could lead to complex decision solving (Bantel & Jackson, 1989). These different opinions could also lead to higher creativity (Bezrukova & Uparna, 2009). Moreover, the subgroups can learn from opinions and idea’s from other groups (Gibson & Vermeulen, 2003). Overall, the complex decision solving, group learning and creativity could positively influence the advisory effectiveness of the board. The findings suggest that demographic faultlines have an unexpected positive impact on the monitoring aspect of board effectiveness. This result is not in line with prior research (Lau & Murnighan, 1998), however it could be explained through the mechanism that socio faultlines include attributes based on the social-categorization perspective (Byrne, 1971; Tajfel, 1978; Turner, 1982; van Knippenberg & Schippers, 2007). This perspective suggest that members of a group create their own in-group versus other out-groups (Jetten et al, 2000). Following this perspective, members feel attracted to their own subgroup, but not to the other groups (Byrne, 1971). This could lead to a lower risk of collusion between the subgroups and therefore could result in greater independence, greater independence resulting in tougher monitoring. This reasoning is supported by Marra et al. (2011), who find that board independence has a positive impact on monitoring effectiveness. Therefore, socio faultlines could have a positive influence on the monitoring effectiveness of the board.

Overall, I can conclude that factional faultlines reduce advisory effectiveness and that task faultlines seem to have a positive influence on advisory effectiveness. Secondly, demographic faultlines have a positive influence on the monitoring function of the board.

(28)

6.2 Discussion

The results of this study contribute to the literature by showing that not all faultlines are harmful. The results suggest that literature can distinguish faultlines based on the attributes underlying their formation, since the faultline dimensions do not have the same impact on board effectiveness.

Task faultlines and demographic faultlines seem to have a positive impact on board effectiveness and factional faultlines reduce board effectiveness. The difference between the three faultlines is the nature of the attributes leading to the faultlines, whereby the task faultlines are related to the tasks but could not be seen. Socio faultlines can be seen but are not related to the task. Lastly, the factional faultlines cannot be seen and are not directly related to the task, but are based on factions, which are representatives from a social entity.

These results bring new insights for shareholders and the board. I suggest paying attention to the alignment of different attributes depending on what is the most pressing issue for the firm. If monitoring effectiveness is an issue for the firm, than I suggest aligning demographic attributes which leads to demographic faultlines. This faultline could reduce collusion and could have a positive influence on the monitoring effectiveness. However, if advisory effectiveness is an issue for the firm, than I suggest aligning task attributes which leads to task faultlines. This faultline could foster divergent thinking and discussions, which leads to a higher advisory effectiveness.

Secondly, I also contribute to prior literature by suggesting that whether specific faultlines have a positive or negative impact depends on the task. Since demographic faultlines influence the monitoring effectiveness and task faultlines and factional faultlines influence the advisory effectiveness. This suggest that one faultline dimension could influence one task, but does not influence another task. This contribution does also have a limitation. This research is based on board in the US. The merger from both the executive and non-executive directors defines the board of organizations in the US. The non-executive directors of the board have no executive responsibility regarding the organization, and have an independent role regarding the organization. So the executive directors can be seen as the insiders, and the non-executive directors can be seen as the outsiders. Some countries make use of a two-tier board system, which divides the executive directors and the non-executive directors into a management and supervisory board. It is possible that my results regarding the different impacts on the advisory and monitoring aspect of board effectiveness cannot be generalized to countries with a two-tier board system. The two-two-tier board system creates a clear line between the responsibilities of executive and non-executive directors (Aste, 1999). The task at hand for the two-tier board, therefore

(29)

differ from a one-tier board system. Moreover, the advisory effectiveness is negatively affected when using a two-tier board system (Millet-Reyes, 2010). Therefore, the results regarding the impact of faultline dimensions on board effectiveness in a two-tier system can be different.

Moreover, the findings suggest that factional faultlines regarding executive and supervisory directors have a negative influence on the advisory effectiveness of the board, while the separation of these tasks and roles are implemented to positively influence the board effectiveness. This raises the question if there should be a clear faultline between the task and the role of directors. For future research, I propose to extend this research by using a sample with both one-tier and two-tier boards. This makes it possible to compare the influence of faultlines on the monitoring and advisory aspect board effectiveness in a two-tier board and a one-tier board.

Another limitation of this research is that these findings can possibly only be applied in the US. The US has much legislation regarding the independence of an audit committee. Any public listed firm in the US has to have an audit committee. I have tried to use the independence of an audit committee as a control variable in my research, but the variance was too low. This indicates that the audit committee independence in the US is high. Audit independence can positively influence the monitoring effectiveness (Klein, 2002). The high audit committee independence in the sample of this research could have a positive effect on the monitoring effectiveness. However, it was not possible to control for this effect, as the variance of the audit committee independence was too low. The audit dependence of the US differs from other countries and this suggest that the variance of the audit committee independence in other counties is higher. Extending this research to other countries, could create possibilities to control for this effect. Introducing the audit committee independence as a control, could lead to other results regarding the influence of faultline dimensions on board effectiveness, because audit committee independence influences monitoring effectiveness. I propose to extent this research to other countries, whereby these countries can be used to control for the audit independence. Where after any unwanted influence of the audit dependence on the results is not an issue anymore.

Furthermore, one of the characteristics of faultlines is the activation of faultlines. Faultlines can stay dormant for many years, without any changes in the group processes (Lau & Murnighan, 1998). Dormant faultlines could be activated through faultline triggers. For instance, diversity problems in the organization may activate faultlines regarding the race of members of the board or changing pension policies in the organization may activate faultlines regarding age. The activation of faultlines is not included in this research, but this could threaten the validity of this research. The research of Lau and Murnighan (1998) who argue that all faultlines are harmful is primarily based on the idea that faultlines

(30)

are activated. The results of this research suggest that task and factional faultlines have no impact on the monitoring effectiveness and that demographic faultlines have no impact on the advisory effectiveness. Faultline triggers create possibility to active these faultline dimensions, which could lead to different outcomes. I therefore propose to extend this research by adding faultline triggers in future research. Activating dormant faultlines with faultline triggers creates possibilities to further investigate the influence of different faultline dimensions on board effectiveness.

(31)

REFERENCES

Adams, R., & Ferreira, D. (2007). A theory of friendly boards. The Journal of Finance, 62(1), 217-250.

Allison, P. D. (1978). Measures of inequality. American Sociological Review, 43(6), 865-880.

Arun, T. G., Almahrog, Y. E., & Aribi, Z. A. (2015). Female directors and earnings management: Evidence from UK companies. International Review of Financial Analysis, 39, 137-146

Ashforth, B., & Mael, F. (1989). Social identity theory and the organization. The Academy of Management Review, 14(1), 20-39.

Aste, L. (1999). Reforming french corporate governance: A return to the two-Tier board? The George Washington Journal of International Law and Economics, 32(1), 1-72.

Bantel, K. A., & S. E., Jackson. (1989). Top management innovations in banking: does the composition of the top team make a difference? Strategic Management Journal, 10(1), 107-124.

Beatty, R. P., & Zajac, E. J., Zajac. (1994). Top management incentives, monitoring, and risk sharing: A study of executive compensation, ownership and board structure in initial public offerings. Administrative Science Quarterly, 39, 313-336.

Bezrukova, K., Jehn, K. A., Zanutto, E., & Thatcher, S. M. B. (2009). Do workgroup faultlines help or hurt? A moderated model of faultlines, team identification, and group performance. Organization Science, 20, 35-50.

Bezrukova, K., Thatcher, S.M.B., & Jehn, K.A. (2007). Group heterogeinty and faultlines: Comparing alignment and dispersion theories of group composition. In K. J. Behfar & L. L. Thompson (Reds.), Conflict in Organization Groups: New Directions in Theory and Practice (pp. 57-92). Northwestern University Press, Evanstion, H.

Bezrukova, K., & Uparna, J. (2009). Group splits and culture shifts: A new map of the creativity terrain. Research on Managing Groups and Teams, 12, 163-193.

(32)

Blau, P. (1977). Inequality and composition: A primitive theory of social structure. New York: Free Press.

Bochner, S. (2016) A conversation with Chair Mary Jo White Keynote Session 43rd Annual Securities Regulation Institute. Retrieved from https://www.sec.gov/news/speech/securities-regulation-institute-keynote-white.html

Boone, A. L., Casares Field, L., Karpoff, J. M., & Raheja, C. (2007). The determinants of corporate board size and composition: An empirical analysis. Journal of Financial Economics, 85(1), 66-101.

Brick, I., & Chidambaran, N. (2010). Board meetings, committee structure, and firm value. Journal of Corporate Finance, 16(4), 533-553.

Byrne, D.E. (1971). The attraction paradigm. New York, NY: Academic Press

Carter, D.A., Simkins, B.J., & Simpson, W.G. (2003). Corporate Governance, Board Diversity and Firm Value. The Financial Review, 38, 33-53.

Chatman, J., Polzer, J., Barsade, S., & Neale, M. (1998). Being different yet feeling similar: The influence of demographic composition and organizational culture on work processes and outcomes. Administrative Science Quarterly, 43, 749-780.

Dahlin, K., Weingart, L., & Hinds, P. (2005). Team diversity and information use. The Academy of Management Journal, 48(6), 1107-1123.

Dechow, P. M., & Dichev, I. D. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The accounting review, 77(1), 35-59.

Dechow, P. M., Sloan, R. & Sweeney, A. (1995). Detecting Earnings Management. Accounting Review, 70(2), 193-225.

(33)

Durisin, B., & Puzone, F. (2009). Maturation of corporate governance research, 1993–2007: An assessment. Corporate Governance: An International Review, 17(3), 266-291.

Eisenhardt, K., & Kahwajy, J. (1997). How management teams can have a good fight. Harvard Business Review, 75(4), 77-84.

Fama‚ E.F.‚ & Jensen‚ M.C. (1983). Separation of ownership and control. Journal of Law and Economic‚ 26(2), 301-325.

Forbes, D.P., & Milliken F J. (1999). Cognition and Corporate Governance: Understanding Boards of Directors as Strategic Decision- Making Groups. Academy of Management Review, 24(3), 489-505.

Gibson, C., & Vermeulen, F. (2003). A healthy divide: Subgroups as a stimulus for team learning behavior. Administrative Science Quarterly, 48(2), 202–239.

Gompers, P., Ishii, J., & Metrick, A. (2003). Corporate governance and equity prices. Quarterly Journal of Economics, 118(1), 107-155.

Healy, P., & Wahlen, J. (1999). A review of the earnings management literature and its implications for standard setting. Accounting Horizons, 13(4), 365-383

Herman, E.S., (1981). Corporate Control, Corporate Power. Cambridge University Press, Cambridge.

Hillman, A. J., & Dalziel, T. (2003). Boards of directors and firm performance: Integrating agency and resource dependence perspectives. Academy of Management Review, 28(3), 383-396.

Jackson, S. E. (1996). The consequences of diversity in multidisciplinary work teams. In M. A. West (Red.), The psychology of groups at work (pp. 53-76). Chichester, UK: John Wiley & Sons.

Jackson, S.E., Brett, J. F., Sessa, V. I., Cooper, D. M., Julin, J. A., & Peyronnin, K. (1991). Some differences make a difference: Individual dissimilarity and group heterogeneity as correlates of recruitment, promotions, and turnover. Journal of Applied Pscychology, 76(5), 675-689.

(34)

Jackson, S. E., May, K. E., & Whitney, K. (1995). Under the dynamics of diversity in decision-making teams. In R. A. Guzzo & E. Salas (Reds.), Team effectiveness and decision making in organizations (pp. 204-261). San Francisco: Jossey-Bass.

Jehn, K. A., Northcraft, G. B., & Neale, M. A. (1999). Why differences make a difference: A field study of diversity, conflict, and performance in workgroups. Administrative Science Quarterly, 44(4), 741-763.

Jensen, M., & Meckling, W. (1976) Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics, 3(4), 305-360.

Jetten, J., Spears, R., Hogg, M., & Manstead, A. (2000). Discrimination constrained and justified: Variable effects of group variability and in-group identification. Journal of Experimental Social Psychology, 36(4), 329-356.

Johnson, S. G., Schnatterly, K., & Hill, A. D. (2013). Board composition beyond independence: Social capital, human capital, and demographics. Journal of Management, 39(1), 232-262.

Jones, J. J. (1991) Earnings Management During Import Relief Investigations. Journal of Accounting Research, 29(2), 193-228.

Joshi, A., & Erhardt, N. (2003). Recent research on team and organizational diversity: SWOT analysis and implications. Journal of Management, 29(6), 801-830.

Kaczmarek, S., Satomi Kimino, S., & Pye, A. (2012). Board Task-related Faultlines and Firm Performance: A Decade of Evidence. Corporate Governance, an international review, 20(4), 337-351.

Kerber, R. (2017) Exclusive: BlackRock vows new pressure on climate, board diversity. Advised on 21-2-2018, from https://www.reuters.com/article/us-blackrock-climate-exclusive/exclusive-blackrock-vows-new-pressure-on-climate-board-diversity-idUSKBN16K0CR.

(35)

Klein, A. (2002). Audit committee, board of director characteristics, and earnings management. Journal of Accounting and Economics, 33(3), 375-400.

Lau, D. C., & Murnighan, J. K. (1998). Demographic diversity and faultlines: The compositional dynamics of organizational groups. Academy of Management Review, 23(2), 325-340.

Lau, D. C., & Murnighan, J. K. (2005). Interactions within Groups and Subgroups: The Effects of Demographic Faultlines. Academy of Management Journal, 48(4), 645-659.

Leuz, C. (2003). IAS versus U.S. GAAP: Information asymmetry-Based evidence from germany's new market. Journal of Accounting Research, 41(3), 445-472.

Li, J., & Hambrick, D. C. (2005). Factional groups: A new vantage on demographic faultlines, conflict, and disintegration in work teams. Academy of Management Journal, 48(5), 794-813.

Lorsch, J.W., & MacIver, E. (1989). Pawns orpofentates: The reality of Americas corporate boards. Boston, MA: Harvard Business School Press.

Mace, M.L. (1986). Directors, Myth and Reality. Boston, MA: Harvard Business School Press.

Marra, A., Mazzola, P., & Prencipe, A. (2011). Board monitoring and earnings management pre- and post-IFRS. International Journal of Accounting, 46(2), 205-230.

Meyer, B., & Glenz, A. (2013). Team Faultline Measures: A Computational Comparison and a New Approach to Multiple Subgroups. Organizational Research Methods 16(3), 393-424.

Meyer, B., & Glenz, A. (2018 February 12). Calculating diversity faultlines with the asw.cluster package in R: A step-by-step guide for beginners. Retrieved from http://www.group-faultlines.org/.

Millet-Reyes, B., & Zhao, R. (2010). A comparison between one-Tier and two-Tier board structures in france. Journal of International Financial Management & Accounting, 21(3), 279-310.

(36)

Milliken, F., & Martins, L. (1996). Searching for common threads: Understanding the multiple effects of diversity in organizational settings. Academy of Management Review, 21(2), 402-433.

Minnis, M. (2011). The value of financial statement verification in debt financing: Evidence from private US firms. Journal of accounting research, 49(2), 457-506.

Peasnell, K. V., Pope, P. F., & Young, S. (2005). Board monitoring and earnings management: do outside directors influence abnormal accruals? Journal of Business Finance and Accounting, 32 (7-8), 1311-1346.

Pfeffer, J. (1972). Size and Composition of Corporate Boards of Directors: The Organization and Its Environment. Administrative Science Quarterly 17(2), 218-228.

Pfeffer, J., & Salancik, G. (1978). The external control of organizations: A resource-dependence perspective. New York: Harper & Row.

Richardson, V. J. (2000). Information asymmetry and earnings management: some evidence. Review of Quantitative Finance and Accounting, 14(4), 325-347

Rico, R., Molleman, E., Sa_{́nchez-Manzanares, M., & Van der Vegt, G. (2007). The effects of diversity} faultlines and team task autonomy on decision quality and social integration. Journal of Management, 33(1), 111-132.

Rosenbaum, M.E. (1986). The repulsion hypothesis: On the nondevelopment of relationships. Journal of Personality and Social Psychology, 51(6), 729-736.

Shaw, J. (2004). The development and analysis of a measure of group faultlines. Organizational Research Methods, 7(1), 66-100

Sinan, A., & Alkan, B. B. (2015). A useful approach to identify the multicollinearity in the presence of outliers. Journal of Applied Statistics, 42(5), 986-993.

Are all faultlines equal?