Looking inside the boardroom: Boards and their reflexivity engagement

(1)

Looking inside the boardroom:

Boards and their reflexivity engagement

Master Thesis, MSc Accountancy and Controlling University of Groningen, Faculty of Economics and Business

June 2017 Boelem Frank Buigel

S2181959 Kruitlaan 21a 9711 TW Groningen Tel: +31 (0)681201710 Email: Frankbuigel@hotmail.com Supervisor: Dr. D. B. Veltrop Co-Supervisor: Drs. I. Mostert

Keywords: Board of directors; Board behavior; Board reflexivity; Board

tenure; Board trust

(2)

Abstract

By using a unique data set, this thesis builds on the currently scarce literature studying actual board behavior. By opening the black box of the elite, it aims to provide new insights that contributes to the inconclusive literature regarding board behavior and board performances. It especially contributes to understanding new determinants that can have a direct or indirect impact on board’s behavior in reflexivity. Although the literature reveals that reflexivity has a positive and important influence on performance, no research to date investigated the variables that influence the extent to which boards engage in reflexivity. In particular, the relation between board tenure and board reflexivity is investigated. Results show that board tenure positively affects a board’s engagement in reflexivity. This relation is partly explained by the positive mediating effect of board trust. This paper therefore finds evidence that limiting board tenure can present potential downsides for board performances, as this would negatively affect board reflexivity. Moreover, the results indicate that boards can differ in their reflexivity engagement. This paper can be used as a starting point for more research to investigate other potential determinants of board reflexivity.

(3)

Introduction

Understanding board1 behavior and studying what makes boards effective are in need of investigation (Forbes & Milliken, 1999; Adams, Hamalin & Weisbach, 2009; Heemskerk, Heemskerk & Wats, 2016). The literature therefore urges the study of board behavior inside the actual boardroom (Pettigrew, 1992; Johnson, Schnatterly & Hill, 2013; Adams et al., 2009). The reasoning behind this is that boards carry out important functions in an organization and have strict responsibilities according to the Corporate Governance Code (e.g. Dutch Corporate Governance Code, 2008) and in literature (Hillman & Dalziel, 2003). A board’s main task is to monitor and advise the executive board (Hillman & Dalziel, 2003). They can be animportant instrument for maintaining internal control and influencing the organizational strategy (Baysinger & Hoskisson, 1990). Studying boards can facilitate a better understanding of an organization’s internal behavior, thereby enabling the development of more policies to control and improve organizational activities (Adams et al., 2009). Literature that studies actual board behavior is scarce, however, because boards tend to behave as an elite team of individuals, which makes it difficult to investigate their behavior inside an actual board room (Pettigrew & McNulty, 1995; McNulty & Pettigrew, 1999). Hence, researchers still do not know much about their actual inner workings (McNulty & Pettigrew, 1999; Pettigrew, 1992; Adams et al., 2009; Migliore & Declouette, 2011). This research contributes to the literature by investigating aspects of board behavior through the use of a unique data set that can facilitate its measurements. By developing and testing a new behavioral model, this paper contributes to a greater understanding of the currently scarce literature on board behavior.

Corporate Governance Codes (CGC) can help boards to perform their previously mentioned roles in the most optimal way. Many Governance Code’s mention that boards should evaluate their own functioning once a year (Dutch CGC, 2008; United Kingdom CGC, 2016; German CGC, 2015), thereby stressing the importance of board evaluations. Board evaluations enhance members understanding of their responsibilities and expectations which can improve board teamwork and decision making. In this way, board evaluations contributes in making boards more effective and can increase board performance (Kiel & Nicholson, 2005; Conger, Finegold & Lawler, 1998). Board evaluations, however, are only useful when boards are actively willing to engage in reflexive behavior, or reflexivity in short (Veltrop, Hermes, Postma & Haan, 2015A). In this paper reflexivity is used as ‘the extent to which group members

(5)

5 overtly reflect upon, and communicate about the group's objectives, strategies and adapt them to current or anticipated circumstances' (West, Garrod & Carletta, 1997, p. 296). In current literature that investigates team behavior, adequate findings show that engaging in reflexivity is important and can enhance team performance (Widmer, Schippers & West, 2009; Schippers, Homam & van Knippenberg, 2009; Tjosvold, Hui & Yu, 2003). It helps teams to understand their actual operations and to develop new ideas to respond to emerging situations (Carter & West, 1998). Reflexive teams tend to focus more on long-term consequences than less reflexive teams (West, 2000). Reflexivity can also lead to more satisfaction and commitment within teams (Schippers, den Hartog, Koopman & Wienk, 2003). Moreover, results suggest that reflexivity leads to more team effectiveness and team innovation (Tjosvold, Tang & West, 2004; Schippers, West & Dawson, 2015; Hoegl, & Parboteeah, 2006). Reflexivity can namely help teams to break through certain strict norms or common team values facilitating them to develop potential new insights or solutions (Raelin, 2001; Gray, 2007). By translating team behavior into board behavior and because board reflexivity facilitates board evaluations, it can be strongly argued that active board reflexivity can be linked to improve board performances (Veltrop et al., 2015A). In correspondence, this paper therefore uses board reflexivity as a determinant for board performance.

Research focusing on explaining team reflexivity shows that teams can differ in their reflexivity engagement (Dayan & Basarir 2009, Tjosvold et al., 2003). By stressing the importance of reflexivity, studying what explains these variations are important (De Dreu, 2007; Schippers, Den Hartog, Koopman & Knippenberg, 2008). While MacCurtain et al., (2010) find that Top Management Teams (TMT’s) differ in their reflexivity engagement, this has not been investigated directly for boards. In correspondence with above literature, this paper assumes that there can arise potential differences between boards and their level of engagement in reflexivity. As far as can be determined however, no literature to date has yet investigated board determinants that influence this level of engagement. Schippers, Homan & Knippenberg (2013) hints that certain group processes can limit or can facilitate the likelihood that a team is willing to reflect upon practices. This paper therefore in particular focusses on studying underlying processes in board behavior as indicators to explain the engagement in board reflexivity. By sufficiently identifying determinants this research sheds light on potential policies that can be set to ensure that boards engage in more active reflexivity.

As starting point, this research investigates the influence of board tenure on board behavior. Board tenure has had considerable attention in literature (Johnson et al., 2013). However literature on studying how board tenure affects board performances is inconclusive,

(6)

6 whereas some scholars found a positive relation, others found a negative relation between the two variables (Johnson et al., 2013; Forbes & Milliken, 1999). Some scholars suggest that higher board tenure increases trust and give boards more time to develop specific knowledge of the organization, which can make boards more effective in their practices (Forbes & Milliken, 1999; Fisher & Pollock, 2004; Johnson et al., 2013; Baysinger & Hoskisson, 1990). Other scholars argue suggest that board tenure can result in cognitive entrenchment, lower levels of task conflict and could make board members less active in challenging the executive board (Vafeas, 2003; Michel & Hambrick, 1992; Veltrop, Mollema, Hooghiemstra & van Ees 2015B). In line with these findings, some regulators limit board tenure (Revisited Dutch CGC, 2016; United Kingdom CGC, 2016; French CGC, 2013). Although scholars reveal contradictory results, inherently they do agree that board tenure is expected to influence board behavior in some way. This paper therefore correspondently proposes that board tenure can lead to certain variation in board behavior. A new behavioral model is developed that simultaneously investigates the contradictory findings between board tenure and board performances and how board tenure, in this way, is expected to (indirectly) affect board reflexivity.

The model argues that board tenure can influence board behavior in two distinct paths which, in turn, lead to different levels of engagement in board reflexivity. Path A proposes that higher board tenure negatively affects board task conflict (Hart, 1991; Michel & Hambrick, 1992; Forbes & Milliken, 1999), whereas more board task conflict is proposed to enhance board reflexivity (Tjosvold et al., 2003; Heemskerk et al., 2016). Path A, therefore, suggests that higher board tenure indirectly decreases their engagement in reflexivity, by the negative mediating effect of board task conflict. Alternatively, Path B proposes that higher board tenure increases board trust (Katz, 1982; Sethi, 2000; Forbes & Milliken, 1999), whereas more board trust is proposed to enhance their engagement in reflexivity (Edmondson, 2004; Sankowska & Söderlund, 2015). Path B suggests that, in this way, board tenure could indirectly enhance board reflexivity, by a positive mediating effect of board trust. If the paper finds support for one or both paths it can provide more knowledge on the contradictory findings in literature on the relation between board tenure and performance. Namely, board performance can then partly be dependent on the influence board tenure has on certain board behavioral variables and how they influence board reflexivity.

The research question is as follows: Is there a relation between board tenure and board reflexivity and how does this relationship operate through board task conflict and board trust?

(7)

7 The contribution of this article is threefold. First, it contributes to a greater understanding of board behavior by using a unique data set. Understanding board behavioral dynamics is often very complex and difficult to study (Forbes and Milliken, 1999). By using a unique survey, this research will be able to map and clarify the behavioral processes of boards. The paper develops and tests a new behavioral model, which can provide more knowledge on board behavior. Second, it contributes to the understanding of what makes boards effective, by identifying board (behavioral) variables that can drive the engagement in board reflexivity. This can facilitate the undertaking of more policies to ensure that boards engage in reflexivity. Third, a contribution to the indecisive literature (Johnson et al., 2013) on the influence of board tenure on board performance can be made, by the argument that board reflexivity enhances board performance (Veltrop et al., 2015A). By studying the relation of board tenure on board reflexivity, it can provide the literature with more knowledge on the so far inconclusive findings. Moreover, as some regulators advocate to (further) limit board tenure, this paper can be used to explain how this influences board behavior (Revisited Dutch CGC, 2016). This paper proceeds as follows; it begins with a background on reflexivity and on literature that studied board tenure. Then the underlying model is explained and on a theoretical basis the hypotheses are shaped. In the method section the measurement of the variables and the hypotheses are explained, leading to the coverage of the results in the results section. The paper is then finalized with a discussion that discusses the practical implications and recommendations for future research.

(8)

8

Theoretical Composite

Background on reflexivity

Reflexivity has mostly been investigated within organizations and in literature studying team or organizational behavior (Schippers et al., 2003; Tjosvold et al., 2004; Dayan & Basarir, 2009). Reflexivity in a group entails that team members collectively review their own performance, communicate with each other about problems to adequately value and discuss potential alternatives or solutions. It involves the process of valuing and understanding other member’s points of view (West et al., 1997; Schippers et al., 2008; Sankowska & Söderlund, 2015). Reflexivity is the opposite of engaging in routines, where tasks are always executed in a similar way (Janis, 1982; Schippers et al., 2003). Reflexivity can be used in improving practices during every stage in the execution of a task: before, during and after completion (Schippers, Hartog & Koopman, 2007). Before the execution of a task, reflexivity facilitates the alignment of different team member goal orientations to a more common and shared team goal by means of enhancing active discussions between the team members (Pieterse, Van Knippenberg, & van Ginkel, 2011). During the execution of a task, reflexivity can ease a more adequate adaption to potential emerging situations. In periods with high environmental uncertainty and in rapid changing situations, reflexivity helps to mitigate the risk associated to this uncertainty (Dayan & Basarir, 2009). After the completion of a task, reflexivity helps a team to evaluate their achievements and can open a discussion whether the execution of the task should be improved or altered in the future (Schippers et al., 2007). Hence, reflexivity is crucial for team learning in improving their practices (Schippers et al., 2003). It is therefore not surprising that much literature stresses the importance and benefits of team reflexivity on team performance (Carter & West, 1998; Tjosvold et al., 2004; Schippers et al., 2015).

Although reflexivity is not directly investigated in terms of boards, it can be strongly suggested that board reflexivity can increase board performance based on the studies of team behavior (Veltrop et al., 2015A). In literature that investigates board functioning, the importance of board evaluations are stressed (Kiel and Nicholson, 2005; Conger et al., 1998). Board evaluations are defined as ‘an evaluation where the board as a whole, or the individual director, is evaluated by itself of someone on behalf of the board’ (Rasmussen, 2015 p. 83). Board evaluations can create value for the organization, by providing clarity for board members about the current organizational situation and about setting common future goals. It can also stimulate communication within the board. In this way, board evaluations can improve board effectiveness (Rasmussen, 2015; Kiel & Nicholson, 2005; Vandebeek, Voordeckers,

(9)

9 Lambrechts & Huybrechts, 2016). The willingness of a board to engage in reflexivity is necessary for board evaluations to be effective. Without reflexivity, board members might not be motivated enough to fully commit in these board evaluations (Veltrop et al., 2015A). Thus, besides that team based literature already stresses the benefits of reflexivity, board literature also directly implicates the importance of board reflexivity as it facilitates board evaluations. Based on the positive findings in team and suggestions in board literature, this paper henceforth argues that board reflexivity enhances board performance.

Board tenure as independent variable

The most important role of a board is to adequately monitor, advice and share its resources with the executive board (Hillman & Dalziel, 2003). This important monitoring role is found in the literature, which describes the agency theory, where the executive directors are mostly interested in their self-interest as opposed to that of the organization. Boards must therefore monitor their executive directors to make sure that they do, in fact, put the organization’s goals first (Jensen & Meckling, 1976; Fama, 1980). Board tenure is an important variable that has received considerable attention in literature (Johnson et al., 2013; Veltrop et al., 2015B, Vafeas, 2003). The reasoning behind this is that board tenure could, namely, have a considerable impact on a board’s performance in their monitoring role. The relation between board tenure and board performance has however been inconclusive so far (Johnson et al., 2013; Forbes & Milliken, 1999). On the one hand, literature argues that longer tenured boards could result in members being more likely to become entrenched. When boards are more entrenched, they are less likely to actively challenge the practices of the executive board (Veltrop et al., 2015B). Longer board tenure could increase social ties and develop friendships with the executive board. This could lead to less independence in relation with the executive board (Veltrop et al., 2015B; Nili, 2015; Vafeas, 2003). Longer tenure can also make boards more strategic persistent and therefore less willing to challenge current practices (Golden & Zajac, 2010). In this way, board tenure is argued to be negatively associated with board’s performance to adequately monitor the executive board. In correspondence with these findings, regulators in different countries limit board tenure (United Kingdom CGC, 2016; French CGC, 2013). For example, the revisited Dutch Corporate Governance Code (2016) advocate that boards should even in principle only be appointed for a maximum of 8 years, as opposed to the 12 years set forth in the Governance Code of 2008.

On the other hand, board tenure could give board members more time to develop complex, organization-specific knowledge and skills which can both improve board’s performance (Forbes & Milliken, 1999; Kosnik, 1990). It can take time namely, for new board

(10)

10 members to completely understand the organization and this could make them less able to adequately monitor the executive board in the beginning stages of their tenure (Bacon & Brown, 1973; Kosnik, 1990; Veltrop et al., 2015B). For example, board members are less involved in tasks of executive directors when they are relatively new in the organization (Veltrop et al., 2015B). Kosnik (1990) additionally argues that, by the increased experiences and expertise, longer termed boards improves their ability and willingness to rise against the executive boards. When boards are together for a longer period of time, it can also increase the cohesiveness and trust within the board (Forbes & Milliken, 1999). This can facilitate boards to be more effective in the execution of their tasks (Forbes & Milliken, 1999). Board tenure can, therefore, also lead to behavior that can improve board performances.

This paper builds further on the inconclusive literature between board tenure and board performances, by adding a new variable, board reflexivity to this relation. It investigates what influence board tenure has on board reflexivity, which as previously argued, can be seen as an important determinant of board performance. It can therefore facilitate new insights on the relation between board tenure and board performance and in this way provide more knowledge about what the potential consequences are of limiting board tenure.

The impact of board tenure: two routes to reflexivity engagement

Board tenure is the starting point of the behavioral model that is tested in this study. The suggested model argues that board tenure influences board reflexivity. The relationship, however, is not a simple one. The model proposes an opposing mechanism that could indirectly lead to the engagement or withdrawal in board reflexivity through the influence of board tenure. It argues that board tenure influences certain board behavior, which in turn affect board reflexivity. The influenced behavioral variables by board tenure therefore act as mediating variables to explain the level of engagement in board reflexivity. Literature that is used in shaping the proposed paths are mostly based on scholars who studied team behavior because less research that directly investigated board behavior is available. When available, relevant board literature is added to strengthen the suggested hypotheses. Team literature suggests that enhanced team tenure can result in that members have less professional discussions, or task conflicts, with each other when performing their tasks. Longer team tenure can, namely, result in that team members become more cohesive and could increase their willingness to act in accordance with the group (Hart, 1991; Michel & Hambrick, 1992; Janis, 1982). High levels of task conflict, when adequately managed, inherently involves that members have many interactions with each other to discuss opposing views (Tjosvold et al., 2003). This can make

(11)

11 them more active in searching for alternatives and can stimulate them to evaluate own and team performances. Task conflict can in this way improve the engagement in collective reflexivity (Tjosvold et al., 2003). In correspondence with these findings, Path A suggests that high board tenure can result in lower board reflexivity through the mediation that a board engages in less task conflict. The second path, Path B, however, suggests a positive relation between board tenure and board reflexivity. Literature argues that members can increase their trust in each other when they are longer together, as it can increase cohesiveness and facilitate them with more time to develop mutual understandings (Katz, 1982; Sethi, 2000). Literature also hints in the direction that more trust can make a team more willing to engage in reflexivity, as members would feel more trusted and motivated to discuss and evaluate potential mistakes (Sankowska & Söderlund, 2015; Edmondson, 2004; De Jong & Elfring, 2010). Thus, path B suggests that longer board tenure can enhance board reflexivity through the mediation of higher board trust.

In summary, this paper tries to identify potential determinants of board reflexivity. In particular, it investigates the (indirect) effect of board tenure on a board’s level of engagement in reflexivity. It studies whether board tenure influences certain board behavior and how they, in turn, influence board’s behavior in engaging in reflexivity. The behavioral variables therefore act as mediating variables. Both paths are independently described, hypothesized and tested. Path A investigates the potential negative relation by the mediation of board task conflict. Path B investigates the potential positive relation by the mediation of board trust. To give an overview, the model is made graphically in figure 1.

(12)

12

Path A, the negative path: The mediating role of board task conflict

The literature mentions two main distinctions on how conflict occurs within teams; relational conflict and task related conflict (De Dreu & Weingart, 2003; Nesterkin & Porterfield, 2016). Relationship conflict involves conflict that results from differences between personal characteristics and values of team members. Task conflict relates to how team members differ in terms of the actions that should take place in order to complete their collective tasks (De Dreu & Weingart, 2003; Nesterkin & Porterfield, 2016). The literature reflects consensus that relational conflict has a negative impact on team performance (Tjosvold et al., 2003; Zhang & Huo, 2015; Nesterkin & Porterfield, 2016). The literature on task related conflict and performance, however, is inconclusive and argues that this relationship can vary between situations (Nesterkin & Porterfield, 2016). This paper therefore continues the study of the influences of task related conflict by how task conflict is affected by tenure and how, in turn, it affects behavior and performance in a boardroom setting.

The cohesiveness of a group can be thought of as the “stickiness togetherness” of a group (Hart, 1991, p 251). Longer tenured teams can make members become more cohesive with each other (Michel & Hambrick, 1992). When teams are more cohesive, members seem more eager to follow the norms of practice and are more willing to remain part of the team (Hart, 1991). Hence, in a more cohesive team, members will be less likely to engage in conflict with each other, as they do not want to be forced to leave the team when conflict results in turnover (Hart, 1991). Thus, when team members are longer together, they potentially have a lower willingness to engage in conflict. A more cohesive team can also lead to a team engaging in more groupthink (Hart, 1991; Janis, 1982). Groupthink can result in that members engage in concurrence seeking behavior. Then, members are not able to break through the standard group practices, as the practices should be in line with that of the group and different opinions should be converged (Janis, 1982; Hart, 1991). Thus, when team members are more subjected to groupthink, task conflict is threatened, because members are less willing to challenge the current ways of thought. Correspondently, Katz (1982) argues that in team projects with high tenure, members communicate less with each other and with members across organizational divisions. Katz (1982) also finds that longer tenure lowers the willingness of members to critically look at their performances and are less open for new ideas (Katz, 1982). In board literature, Golden & Zajac (2010) find a U relation between board tenure and strategic change persistence. They argue that longer tenure can hinder changes in practice as board members are then less actively involved in searching and discussing alternative or new solutions. In longer tenure, namely, boards are more likely that they developed preferences to keep the current way

(13)

13 of practices going that maintain their status quo (Golden & Zajac, 2010).

If teams engage in collective long periods they can develop the same schemata (Michel & Hambrick, 1992). These are ‘cognitive structures used to organize knowledge of past experiences and are invoked when people make sense of new stimuli’ (Michel & Hambrick, 1992, p 17). When team members collectively engage in long tenure, they often develop the same schemata because they have similar past experiences that occurred within the organization (Michel & Hambrick, 1992). By developing the same schemata, it can be argued that longer tenured teams respond in the same way to new situations (Michel & Hambrick, 1992), which lowers the level of different insights between members thereby resulting in less task related conflict. Hence, through collective past experiences and similar schemata, task conflict can diminish because opinions on how to complete the task will be more similar. In a sense this is also mentioned by Forbes & Milliken (1999) who argue that board members that are only in function for a short time use their past non board experiences more than when members that are of long tenure, who instead rely on using the experiences they have collected from their time as being on the board. New ‘fresh’ members can therefore create new knowledge sharing and challenge debate (MacCurtain et al., 2010). Correspondently, Tschang and Ertug (2016) suggests that new human capital in a team can help in the search for explorative innovations. As board tenure can increase board cohesiveness (Forbes & Milliken, 1999) and because board members might develop similar schemata and experiences, this paper proposes a negative relation between board tenure and board task conflict. Leading to the following hypotheses:

H1a: Board tenure has a negative relation with board task conflict.

Although the literature studying the relation of team conflict on team performance is still an ongoing debate, some scholars argue that task conflict can in fact be adequately managed by a team (Tjosvold et al., 2003; Tjosvold, Yu & Wu, 2009) and in result can increase team performance (Tjosvold et al., 2004; Tjosvold et al., 2009; Widmer et al., 2009). Task conflict can result in more motivation and collaboration for open-minded discussions and can facilitate learning from each other, by discussing opposing ideas (Tjosvold et al., 2003; Tjosvold et al., 2004). With certain levels of managed task conflict, members are more active in reflecting on their current performances (Tjosvold et al., 2003), while it also improves their attention for new ideas that could improve their current functioning (De Clercq, Rahrman & Belausteguigoitia, 2015). In this way, task conflict can make teams more active to collectively engage in reflexivity (Tjosvold et al., 2003).

(14)

14 Some scholars argue that especially in high organization level decision-making teams it is important that there are certain levels of task conflict, as it increases their performance in the execution of their tasks (O’Neill, Allen and Hastings, 2013). This is also supported by Eisenhardt, Kahwajy and Bourgeois (1997) who studies TMT’s in highly changing and difficult markets. They find that, when TMT’s are able to manage task conflict, team members are more active in challenge each other at higher levels of task conflict. In these teams, task conflict creates a better understanding of the complex situations and facilitates them to develop a better range of alternatives that can make them more able to adequately deal with their objectives (Eisenhardt et al., 1997). Task conflict is easier to manage when members have a shared perception of their collective tasks as it provides a common framework where members can identify themselves with (Tjosvold et al., 2004; Heemskerk et al., 2016). Heemskerk et al., (2016) argue that especially boards understand their roles to a certain extent, which enables them to manage task conflict more adequately. Correspondently, in a direct study on board literature, Heemskerk et al. (2016) argue that boards that allow task conflict facilitates them to be more responsive, more open for finding alternatives and in this way can lead to more reflexive decisions. Having a certain amount of task conflict should therefore, in fact, not be undervalued, as Heemskerk et al. (2016) specifically suggests that by instead avoiding task conflict would be more harmful for a board then allowing them. Avoiding conflict would make boards underperform because it would suppress the allowance for different thoughts, which could lead to indecisive and unreflective decisions (Heemskerk et al. 2016).

Whereas teams sometimes perceive task conflict as negative (De Dreu & Weingart, 2003), especially in high level decision making teams that face complex situations, it is important to have a certain level of it (Eisenhardt, 1997). As boards fit to this description (Forbes & Milliken, 1999), task conflict can help them to improve their ability to collectively deal with emerging situations, improve the development of a better set of alternatives and can increase awareness of current performances (Heemskerk et al., 2016; Eisenhardt, 1997; Tjosvold et al., 2003), all of which can stimulate the reflexivity process. This research therefore suggests a positive relation between board task conflict and their engagement in board reflexivity. Because board tenure is expected to have a negative relation with task conflict (H1a), this paper consequently proposes that board tenure will negatively affect board reflexivity through the mediating role of task conflict.

(15)

15

Path B, the positive path: The mediating role of board trust

The literature presents many different definitions of trust (Costa, 2003). In concurrence with McKnight, Cumming & Chervany (1998) and McAllister (1995, p3) trust is defined here as: ‘the extent to which a person is confident in, and willing to act on the basis of, the words, actions, and decisions of another’. Trust in teams results in increased task performance, team satisfaction and knowledge sharing, making trust an important factor within teams (Costa, 2003; Erdem & Ozen, 2003; De Jong, Dirks & Gillespie, 2016). Trust includes both affective and cognitive dimensions (Lewis & Weigert, 1985; McAllister, 1995). Cognitive-based trust is knowledge driven, based on trusting the trustee’s competences, whereas affective trust is based on emotions and social connection with the trustee (Dayan & Benedetto, 2010; Lewis & Weigert, 1985). The literature argues that both dimensions of trust are highly correlated and complementary, noting however, that at times they could result in different outcomes (Johnson & Grayson, 2005; Lewis & Weigert, 1985). In accordance with prior research, in this research trust is considered to comprise both components and no explicit distinction between the two dimensions in the development of the theory and shaping of the hypotheses is made (Dayan & Benedetto, 2010; Lewis & Weigart, 1985). Both dimensions, however, are at first distinctively calculated in the data analysis, to provide certainty regarding the possible differences in outcomes (Dayan & Benedetto, 2010; Johnson & Grayson, 2005).

Dayan & Benedetto (2010) argue that, in teams, trust can be enhanced by their being together for a longer period of time. Team members can develop a shared history and this can result in a stronger identification between the members (Sethi, 2000). This is also supported by Katz (1982) who argues that longer tenure makes a team more homogenous, leading to a more common and shared identity within the group. Trust and the connection with tenure are especially important when team tasks are difficult and complex, that is, complex task strengthen this relation (Dayan & Benedetto, 2010). With more complex tasks their completion is namely more dependent on active communication and collaboration between the team members (Akgün, Byrne, Keskin Lynn & Imamoglu, 2005). By having more communication and collaboration, over time members can develop more trust in each other, because they are important variables for developing trust (Dayan & Benedetto, 2010). As boards often have to deal with complex tasks (Forbes & Milliken, 1999), the complexity of these could strengthen the relation between board tenure and board trust.

Harrison et al. (2002) argue that potential demographical differences, such like gender or age, between team members can in the initial stage of a team’s conformation have negative consequences on its social integration and cohesiveness. By spending more time together,

(16)

16 however, team members have more time to collaborate, exchange differences and explain their values to each other, which lowers the impact of the perceived demographic diversity in the team (Harrison et al., 2002; Harrison et al., 1998; Pelled, Eisenhardt & Xin, 1999). In addition, boards in particular also have differences in backgrounds, values and perceptions because they generally have different working experiences (Migliore & Declouette, 2011). After some time, board members can become more familiar with each other and understand these differences. This could enable them to adequately deal with or avoid these differences, so that they are not deleterious to trust (Pelled et al., 1999). In board literature, Forbes and Milliken (1999) suggests that longer tenured boards could lead to more cohesiveness and familiarity between members, to which they suggest that it would improve board trust.

Based on Forbes & Milliken’s (1999) suggestion and from the findings on team behavioral studies, this research proposes a positive relation between board tenure and board trust, leading to the hypotheses:

H2a: Board tenure will have a positive relation with board trust

For the engagement in reflexivity to be effective and constructive, team members must be willing to engage in this process (Tjosvold et al., 2004). To facilitate learning, which can be considered closely linked to reflexivity (De Jong & Elfring, 2010), scholars argue that it is important to create an atmosphere in which a team can talk about imperfections and mistakes (De Jong & Elfring, 2010; Edmondson, 1999). It could therefore be important for reflexivity that team members feel comfortable and trusted so that they can share, reflect and discuss these difficulties or feel that there is room to discuss potential alternatives and new ideas. Individual members should feel trusted that other members would respond constructively when discussing them (McAllister, 1995). Edmondson (2004) states that the way in which team members perceive that difficulties have been handled in prior circumstances influences their perception of whether they feel comfortable talking about similar difficulties. Correspondently, team members must arguably feel trusted in each other that reflecting and potentially adapting to new situations is not something negative, but they must embrace it as something to learn and to benefit from. Trust can therefore be of high importance, to ensure that members are willing to engage in the reflexivity process. In correspondence, MacCurtain et al. (2010) argue that TMT’s with high levels of trust facilitates them to involve in more active and collective open minded discussions about problems and solutions, thereby promoting reflexivity. Schippers et al., (2013) also hints in this direction by stating that the willingness of a team to reflect upon practices can in turn be limited when a team faces certain levels of distrust.

(17)

17 When team members develop more trust, it consequently makes them more motivated to perform their tasks as well as possible (De Jong & Elfring, 2010). This would entail their being more willing to engage in reflexivity, as it can increase their ability to perform their task (Sankowska & Söderlund, 2015; De Jong & Elfring, 2010). However, members must actually feel that this engagement in reflexivity would contribute to higher team performance. Team members should therefore develop trust in each other’s competences and in that, by combining them, they must feel that they can reach new insights or solutions that would increase their collective performance (Tjosvold et al., 2004; Tjosvold et al., 2003). Thus while affective trust is an important facilitator that makes team members feel safe to engage in active reflexivity, cognitive trust is also important. It can make members feel that collectively engaging in reflexivity is fruitful, as high cognitive trust can increase member’s feeling that they are dependent on each other (McAllister, 1995).

This paper suggests that an atmosphere of board trust enables boards to be more willing and more motivated to engage in reflexivity. Board members should have an open mind in order to embrace reflexivity. A positive relation between board trust and a board’s reflexivity engagement is therefore suggested. As previously mentioned, board tenure has been proposed to improve board trust (2a). This paper therefore proposes that board tenure can have a positive effect on board reflexivity, by means of the mediating variable of board trust.

(18)

18

Research Method

Measuring board behavior is difficult, because generally the actual practices and processes of boards are not easy accessible (Pettigrew, 1992; Adams et al., 2010; Leblanc & Schwartz, 2007). By using a unique data set in a Dutch setting, this research breaks through this “black box” and can enrich the understanding of board behavior. It builds further on literature like Heemskerk et al., (2015) who also used a survey to map board behavior processes. In that research, board behavior is mapped by using the chief executive as sole participant for measuring board behavior. This survey however consisted of individual answers of every member of the board, thereby providing a more representative set of data than measuring just one executive as a representative of board behavior (e.g. Veltrop et al., 2015B).

The survey process, reliability and validity

In the survey complete anonymity to respondents is provided. This reduces the problem that board members would falsely respond more positively to the questions as it would reduce the risk of social desirability. The incentive for board members to conduct in the survey is because they received a benchmarking score relatively to other boards. This could imply that members conducted the survey more seriously because the results are very relevant for their own learning. The second reason for boards to participate in the survey is because a board could in this way argue that they are in compliance with the Dutch Corporate Governance Code (2008) in conducting a board evaluation. The survey data were collected between 2013 and 2016. Around 150 Dutch boards participated in the survey and all boards were active in the non-profit sector. The boards were working in different segments, either in the housing cooperation, in educational institutions or charity organizations. The survey was conducted on an online site. This site collected, saved and calculated the data provided by the survey. By the fact that members had to apply for making use of this survey and the survey used a convenience sample, there was a very low non-response rate. The questions from the survey were measured on a seven-point Likert scale (1=completely disagree, 7= completely agree). Boards engage in behavior as a group and not as independent individuals (Forbes & Milliken, 1999). Therefore the individual responses of the board members perceptions were translated into a single combined board average. In answering some of the survey questions, board members had to individually grade every member of the board. In this case, the survey answers were aggregated into a member’s average individual board level perception. Consequently a mean board level score was calculated. Additional analysis were conducted to test the usefulness of the

(19)

19 aggregation from individual to board level. ICC1 calculates the extent to which board members responses are interchangeable. It indicates the extent to which an individual member represent all members in a board (Klein & Kozlowski, 2000) and it indicates whether there is a proportion of variance that can be explained by analyzing on board level instead of on an individual member level (Bliese, 2000). The analysis show that ICC(1) was significant for every of the measured board responses. A value smaller than 0.1 is arguably a minimal effect, a value of .10 to .25 a medium effect, whereas a value greater than .25 is considered huge (Veltrop et al., 2015A; Klein & Kozlowski, 2000). For board reflexivity ICC(1) was 0.13, for board affective and cognitive trust 0.5 and for task conflict 0.35. The results therefore indicate that there is meaningful variation, which facilitates analyzing on a board level (Klein & Kozlowski, 2000). Lastly this paper used different control variables to provide higher internal validity.

The Variables

Board tenure

Board tenure was collected by handpicking the variable from annual accounts. Board tenure is measured as the average length of tenure by the combined board. This is consistent with prior literature that measure teams or board tenure (Vafeas, 2003; Veltrop et al., 2015B). In annual reports board tenure was often relatively easy to collect.

Board Task Conflict

As previously mentioned there are two main aspects of conflict, relational conflict and task related conflict (De Dreu & Weingart, 2003; Nesterkin & Porterfield, 2016). In the survey both dimension of conflict were individually measured. For the research only task conflict is used. The two items used for measuring task conflict, in correspondence with prior literature (Jehn & Mannix, 2001), were: “How often do you have different opinions with [name]” and “how often do you have a different insights with [name]?” The Cronbach’s alpha was 0.824.

(20)

20

Board Trust

Trust can have two dimensions, cognitive trust and affective trust (Lewis & Weigert, 1985). When the first analysis were conducted, the results indicated that board cognitive and affective trust had a very high correlation with each other. Therefore both variables were combined into one mean variable and together defined and calculated as board trust. In line with prior literature (Mc Allister, 1995) cognitive trust is measured with the question: “I can trust that [name] will devote to their task with professional commitment” and “I can trust that [name] will be working carefully”. Affective trust is measured with the question: “I can trust that I can speak freely about ideas, expectations and feelings with [name]” and “I can speak freely with [name] to discuss difficulties that occur while performing my task”. The Cronbach’s alpha for board’s combined trust was 0.870.

Board Reflexivity

The last main variable for the model is the engagement of board reflexivity. The survey developed several items based on prior literature (Carter & West, 1998; Schippers et al., 2007). The items that were covered were “to what extent are the goals and workings of the boards discussed regularly” and in “in how much is the board working together effectively”. Furthermore items covered if the board “changed their goals in the light of changing circumstances” and if the “exchange of information between members is done effectively”. The Cronbach’s alpha was 0.853.

Control Variables

In the analysis additional control variables were used to control for unwanted variation. Forbes & Milliken (1999) argue that size can have an important impact on board functioning and board dynamics. Larger boards could enhance task conflict because it can lead to more different insights, but it can also make boards more difficult to engage in good functioning and could lower board cohesiveness and board trust (Forbes & Milliken, 1999). In this way could board size have a potential impact on the main analysis and is therefore used as a control variable. Board size was handpicked from annual accounts. Board meetings are critical for board performances, because other opportunities for communication between members are often limited (Pugliese, Nicholson & Bezemer, 2014). In this way, for example, more board meetings could simply enhance the time for boards to be able to spend on board reflexivity or on the discussion of task conflicts. To control for variation based on the availability of more time, the

(21)

21 number of board meetings were used as a control variable. Board meeting frequency was handpicked from annual accounts. Maccurtain et al. (2010) find that, in a TMT-setting, diversity can influence trust and (indirectly) the engagement in reflexivity. To control for variation in trust and reflexivity by board diversity, two diversity control variables were used. These were age diversity and gender diversity. For both variables, a standard deviation was calculated as the value of diversity in the board. Lastly, also a financial control variable was used. Firm size was used to control for organizational differences that potentially could have an influence on board behavior. The total assets of a firm were taken as a proxy for firm’s size (Fama & French, 2002).

(22)

22

Results Section

This section provides the results from the conducted analysis. In the confidentiality agreement complete assurance was provided to board members that individual answers would not be made publicly available. The individual answers were therefore standardized before conducting any analysis and, additionally, the names of the board members were excluded to provide complete anonymity. This ensures that the responses in the survey could not be translated to individual board members. The standardization led to the exclusion of a descriptive model, since the average would be a zero and the standard deviation would be one. The standardized data values the scores in relation to all scores given, but hides the score of the actual responses given. The standardized data does, therefore, provide the ability to benchmark and analyze the responses between organizations. The correlations between all variables are described in Table 1. Table 2 examines the regression analysis with mediating path A. Table 3 presents the regression analysis that considers mediating path B. In the analysis board tenure was winsorized before it was standardized, because it had a number of outliers that would else have had too much impact on the analysis. An average for board tenure was calculated, and limited variance to a maximum of -3 or + 3 standard deviations of their average. The analysis includes 137 boards. All data, including that of all control variables, were available from this number of boards. The

Table 1 Correlations of variable dataset

Variables 1 2 3 4 5 6 7 8

1. Board Tenure

2. Board Reflexivity .384**

3. Board Task Conflict .113 -.089

4. Board Trust .188* .337** -.356** 5. Board Size -.094 -.018 -.177* -.019 6. B. Meeting freq. -.053 .014 .299** -.303** .099 7. Age Diversity .009 -.049 -.072 -.121* -.077* .064 8. Gender Diversity -.276** -.133 -.186* -.001 .200** -.172* -.019 9. Firm’s size -.022 -.066 .015 .024 .047 .004 -.165* .086

(23)

23 Correlation Table shows that board tenure correlates positively with board reflexivity, board trust and negatively with gender diversity. Board reflexivity positively correlates with board trust. Board task conflict negatively correlates with board trust, board size and gender diversity but positively with meeting frequency. Board trust, furthermore, negatively correlates with board meeting frequency and age diversity. The Correlation Table also indicates some correlations between the control variables, but they are of less importance for this paper. There was no indication of multicollinearity as additional analysis indicated that the Pearson Correlation was always smaller than an absolute value of .7. Correspondently, these findings were also supported by the later discussed regression analysis, which did not indicate a sign of multicollinearity as VIF was always less than 10, with a highest score of 1.7.

Table 2 and Table 3 indicate, in step 2, that board tenure has a significant impact on a board’s engagement in reflexivity (b=.076, t=4.518, p=.000). Table 2 and Table 3 review this relation by adding the two mediating variables independently. In Table 2, step 1 indicates that board tenure has no significant effect on board task conflict, b=.038, t=1.140,p=.276. It can therefore be concluded that hypothesis 1a is not supported. Step 3 reveals that the regression of the mediator board task conflict with board reflexivity while controlling for board tenure, was supprisingly slightly negatively significant, b=-.078, t=-1.870, p=0.064. The correlation table, however, indicated that there is no significant correlation between the two variables. After conducting additional analysis without using the control variables, no significant regression was found between the two variables. The mediating effect of board task conflict is therefore very questionable, so an additional reliability test was conducted by using the application developed by Preacher & Hayes (2004). The application facilitates the estimation of indirect effects and the confidence intervals of these relations. This way, a greater reliability of mediating effects can be provided. As all the used data is standardized before analyzing, the results show similar intervals for the unstandardized, partially and completely standardized confidence intervals as measurement for the indirect relation. Namely, the (partially) standardized intervals also include the calculation of the standard deviations of the board variables in their analysis, but because the data is completely standardized these values are always one and henceforth there are no differences between the intervals (Preacher & Kelly, 2011). Both expected mediations are collectively tested in one model, with a bootstrap sample of 5000. The results have been collected in Table 4. The results demonstrate that there is no certain mediating effect of board task conflict, namely the confidence interval includes zero, ranging from -.0339 to .0064. When the confidence interval, namely, does interact with the zero point, it is not possible to ascertain

(24)

24 whether there is indeed a mediating effect on the proposed relation (Preacher & Hayes, 2004). These results therefore indicate that hypothesis 1b must be rejected.

Table 3 investigates the potential mediating effect of board trust. Step 1 indicates that board tenure has a positive effect on board trust, b=.078, t=2.023, p=.04, providing evidence for hypothesis 2a. Step 3 reveals that the regression of the mediating effect of board trust, while controlling for board tenure, is significant for board reflexivity, b=.137, t=3.767, p=.000. This step, however, also indicates that there is a significant relation between board tenure and board reflexivity while controlling for the mediator of board trust b=.065, t=4.011, p=.000. This implies that the regression analysis in Table 3 so far indicates that board trust partially mediate the relation of board tenure on board reflexivity. For a reliability check of the mediating effect of board trust the relation was additionaly tested by means of the application developed by Preacher & Hayes (2004). The results are described in Table 4. They illustrate that board trust does significantly mediate the relation between board tenure and board reflexivity on a one tailed level, namely with a confidence interval ranging from .0180 to .0994 and an avarage effect of .0490. This provides evidence that there is indeed a mediating effect of board trust, thereby supporting hypothesis 2b.

Table 2 Path A: The Mediating effect of Board Task Conflict

Dependent Variable Board Task Conflict Board Reflexivity Board Reflexivity

Step 1 Step 2 Step 3

Independent variables

Board Tenure .038 .076** .079**

Board Task Conflict -.078*

Board Size -.096* .004 -.003 Meeting Frequency .193** .010 -.025 Age Diversity .010 -.008 -.007 Gender Diversity -.318 -.043 -.068 Firm size .055 -.047 -.043 R square .152 .157 .179 F value 4.149** 4.603** 4.426** N=137, **P<0.01,*P<0,05 (one tailed)

(25)

25 .

Table 3 Path B: The Mediating effect of Board Trust

Dependent Variable Board Trust Board Reflexivity Board Reflexivity

Step 1 Step 2 Step 3

Independent variables Board Tenure .078* .076** .065** Board Trust .137** Board Size .012 .004 .003 Meeting Frequency -.201** .010 .038 Age Diversity -.027 -.008 -.004 Gender Diversity -.049 -.043 -.037 Firm Size .020 -.047 -.050 R Square .133 .157 .240 F value 5.129** 4.603** 5.832** N=137, **P<0.01, *p<0.05 (one tailed)

Table 4: Regression Results from additional Mediation Analysis

Indirect effect, standardized Value LL 90% CI UL 90% CI SE

Board Task Conflict -.0039 -.0341 .0064 .0117

Board Trust .0490* .0180 .0994 .0239

**P<0.01, *p<0.05 (one tailed), LL= Lower limit, UL: Upper limit, CI: Confidence Interval, N=137, Bootstrap sample: 5000

(26)

26

Discussion

Practical implications

This paper contributes to the existing literature by providing more knowledge on the study of board behavior and board performance. Board reflexivity arguably is an important determinant of board performance by providing clear benefits described in team and board literature (Veltrop et al., 2015A). However, as far as can be determined, no research investigated what leads boards to the engagement in or withdrawal from board reflexivity. By using a unique data set, this paper contributes to the currently scarce literature describing board behavior by investigating potential determinants of board reflexivity. As potential determinant, board tenure was used as starting point of the developed model. Board tenure is especially relevant as, for example, in the Netherlands, regulators advocate to further limit board tenure (revisited Dutch CGC, 2016). However, the literature that investigates the impact of board tenure on board performance has produced inconclusive results on the benefits of the proposed regulation (Johnson et al., 2013). By testing the effect of board tenure on board reflexivity this paper, therefore, aims to contribute to the inconclusive literature and to provide more knowledge for regulators about the consequences on board behavior by limiting board tenure. Although the literature suggested that higher board tenure would lead to less task related discussions, which inherently would reduce their reflexive behavior, this paper could find no support for this (Forbes & Milliken, 1999; Michel & Hambrick, 1992; Tjosvold et al., 2003; Heemskerk et al., 2016). Namely, no negative influence was found in a board’s level of task conflict and its willingness to engage in reflexivity at different levels of board tenure.

In fact, the paper reveals that limiting board tenure could have potential downsides for board performance. The paper, namely, finds that higher board tenure positively affects board trust and board reflexivity. As reflexivity enhances board effectiveness and performance (Veltrop et al., 2015A), it can be argued that limiting board tenure in this way indirectly decreases board performance. Longer tenure can give board members more time to become acquainted with each other. They can overcome differences and dedicate time in developing mutual understandings, which could enhance the social identification and cohesiveness between them (Harrison et al., 2002; Migliore & Declouette, 2011; Forbes and Milliken, 1999). In this way, longer tenure could facilitate the development of more board trust. Results also suggests that for boards to engage in reflexivity, it is important for there to be an atmosphere of trust within the board. Board members must feel trusted with each other in order to reflect on their own performances and discuss potential mistakes or difficulties (Edmondson, 2004; Maccurtain et al., 2010). Additionally, higher board trust also increases member’s trusting in other

(27)

27 member’s abilities and competences (McAllister, 1995). This in turn, increases their motivation and willingness to engage in active reflexivity together, as they feel that it would actually contribute to increase their own and board performance (De Jong & Elfring, 2010, Tjosvold et al., 2004). This paper therefore argues the importance of boards developing high levels of trust, especially when board tenure is limited. Additionally, this paper determines that a board’s level of reflexivity engagement is not an independent and static factor. The results indicate that boards can differ in their engagement in reflexivity and that this can be dependent on certain board characteristics and behavioral determinants. This paper can, therefore, be seen as a starting point of the study on board reflexivity and advocates the importance of conducting more research to explain and identify other potential determinants of board reflexivity. In this way, scholars can contribute in identifying more policies that can be used to improve their engagement.

The paper finds no relation between board tenure and board task conflict. The literature has suggested a negative relation, in which members of teams with high tenure are less active in engaging in task conflict because their decisions are influenced by shared experiences, less active communication and more groupthink (Katz, 1982; Michel & Hambrick, 1992; Janis, 1982). One potential argument explaining why no relation is found, is that especially studies that describe team tenure and team conflict expected this relation. It can arguably take a few years in boards for a board member to completely understand the organization (Bacon & Brown, 1973; Kosnik, 1990; Veltrop et al., 2015B). Board members might therefore not engage directly in conflict in early tenure years because they might feel that they do not understand the organization enough to engage in task conflict from the beginning of their functioning. Board members are in addition often involved in and must respond to newly emerging situations (Forbes & Milliken, 1999). Even when board tenure is high, boards might still face new situations and therefore always have to engage in certain levels of task conflict. As was argued, longer tenured teams could lead to routine-wise behavior (Hart, 1991), but by the ever changing nature of a board’s functioning, it can potentially not be possible for them to evolve in this kind of behavior. Boards also often meet only periodically, resulting in that they are less likely to develop a group cohesiveness that would stimulate groupthink and hinder board task conflict (Forbes & Milliken, 1999; Janis, 1982).

This paper proposed that board task conflict would positively influence board reflexivity, by the reasoning that it would make board members more active in discussing opposing ideas which can inherently make them more reflective on their own performances (Tjosvold et al., 2003; O’Neill et al., 2013; Heemskerk et al., 2016). However, no relation was

(28)

28 found. This can be explained from additional analysis, which indicated that board members perceive board task conflict as negatively affecting their trust. Tjosvold et al., (2003) and Eisenhardt et al., (1997) state that only if teams are able to manage task conflict it could positively affect their engagement in reflexivity. The paper therefore suggests that boards might not be completely capable in managing task conflict, as task related conflict decreases their perception of trust. In this way, board task conflict can potentially even harm board reflexivity as board trust is an identified determinant for board reflexivity. Additional research can investigate how boards can manage board task conflict in the way that it does not harm their trust. Then, when a board succeeds in managing conflict, the relation between board task conflict and board reflexivity can be restudied.

Limitations and future research

Board members had to fill in their own perceptions of how they behave, which could lead to differences due to individuals having diverse perceptions about certain board behavior variables. When evaluating the board, members who had been in the board for a longer time had the same “voice” as a member who only just entered the board, although they might not understand their board behavior as well as longer tenured board members.

Second, the study used only considered Dutch boards, where all boards had a two-tier structure. This is different from, for example, the United Kingdom, where boards have a one-tier board (Weimer & Pape, 1999). In a one-one-tier board, executives and non-executives are together in one board, whereas in a two-tier they are separated. Differences in outcomes can therefore arise. Literature however argues that both have many similarities, in as much as the role of the two-tier board is similar to that of non-executives in the one tier board (Hooghiemstra & Van Manen, 2004; Bezemer, Peij, Maassen, & Halder, 2012). The boards studied were all active in the nonprofit sector. Potentially, there can be differences between boards in profit and nonprofit sectors, which can be an interesting study for future research.

Third, this research could only explain some part of the relation between board tenure and board reflexivity, by the mediating effect of board trust. More research that explains this direct relation is needed and would be an interesting topic to study. More research that explains determinants of board reflexivity are also needed, as so far only board trust and board tenure are identified. A potential determinant that, for example, could indirectly influence the engagement in reflexivity is board diversity (MacCurtain et al., 2010). In this research diversity is only used as a control variable which included a separate diversity measure for gender and age. The analysis did not indicate a direct relation of diversity on board reflexivity but it did

(29)

29 show that board diversity can potentially influence certain other board behavioral variables. It can, therefore, be an interesting study topic whether diversity can indirectly affect board reflexivity through some of these variables, like for example through board trust (e.g. MacCurtain et al., 2010). Also, other forms of diversity, like educational background, are not taken into account. A combined diversity variable could, furthermore, be developed in future research to test their combined (indirect) effect on board reflexivity.

Fourth, the analysis identified many other potential relations between board behavioral variables. Additional research could further investigate some of these relations and provide us with better knowledge of board behavior. This paper also primarily focused on explaining inside board behavior by connecting inside board variables and is not centered on how external variables influence inside board behavior. Only firm’s total assets were used as external variable, but it did not significantly influenced the main board behavior variables. Further research could link other external variables directly with inside board behavior. For example, as this paper advocates the importance of board trust for board reflexivity, external variables that potentially influence board trust can be investigated in further research.

Last, this paper used quantitative data to analyze the expected hypotheses. It did not use any qualitative data. Although the quantitative analyses bear significance, it would provide more information about the proposed relations if it was supported with actual board members thoughts beyond their answers based on a survey’s scale. In addition to these quantitative research, for example, interviews could be held with board members to explain their underlying reasoning and their behavior to a greater extent. Moreover, as the research only measured responses at a specific point in time, nothing can be said about causality. By studying more time periods it would provide a greater reliability that the relationships found are, in fact, explained by the underlying theory suggested.

(30)

30

References

Adams, R. B., Hermalin, B. E., & Weisbach, M. S. (2010). The role of boards of directors in corporate governance: A conceptual framework and survey. Journal of Economic Literature, 48(1), 58- 107.

Akgün, A. E., Byrne, J., Keskin, H., Lynn, G. S., & Imamoglu, S. Z. (2005). Knowledge networks in new product development projects: A transactive memory perspective. Information &

management, 42(8), 1105-1120.

Bacon, J., & Brown, J. F. (1973). Corporate directorship practices: Roles, selection, and legal status of the board. New York: The Conference Board.

Baysinger, B., & Hoskisson, R. E. (1990). The composition of boards of directors and strategic control: Effects on corporate strategy. Academy of Management review, 15(1), 72-87.

Bezemer, P. J., Peij, S. C., Maassen, G. F., & van Halder, H. (2012). The changing role of the supervisory board chairman: the case of the Netherlands (1997–2007). Journal of Management & Governance, 16(1), 37-55.

Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations: 349–381. San Francisco, CA: Jossey-Bass.

Carter, D. A., Simkins, B. J., & Simpson, W. G. (2003). Corporate governance, board diversity, and firm value. Financial review, 38(1), 33-53.

Carter, S. M., & West, M. A. (1998). Reflexivity, effectiveness, and mental health in BBC-TV production teams. Small group research, 29(5), 583-601.

(Dutch) Corporate Governance Code 2008, available at: http://www.mccg.nl/download/?id=609

(31)

31 (French) Corporate Governance Code 2013, available at: http://www.ecgi.org/codes/code.php?

code_id=393

(German) Corporate Governance Code 2017, available at: http://www.dcgk.de//files/dcgk/usercontent/ en/download/code/170214_Code_clean_en.pdf

(United Kingdom) Corporate Governance Code 2016, available at: https://www.frc.org.uk/Our-Work/Publications/Corporate-Governance/UK-Corporate-Governance-Code-April-2016.pdf

Conger, J. A., Finegold, D., & Lawler, E. E. (1998). Appraising boardroom performance. Harvard business review, 76, 136-164.

Costa, A. C. (2003). Work team trust and effectiveness. Personnel review, 32(5), 605-622.

Dayan, M., & Basarir, A. (2009). Antecedents and consequences of team reflexivity in new product development projects. Journal of Business & Industrial Marketing, 25(1), 18–29.

Dayan, M., & Di Benedetto, C. A. (2010). The impact of structural and contextual factors on trust formation in product development teams. Industrial Marketing Management, 39(4), 691-703.

De Clercq, D., Mohammad Rahman, Z., & Belausteguigoitia, I. (2015). Task conflict and employee creativity: The critical roles of learning orientation and goal congruence. Human Resource Management.

De Dreu, C. K. (2007). Cooperative outcome interdependence, task reflexivity, and team effectiveness: a motivated information processing perspective. Journal of Applied Psychology, 92(3), 628.

De Dreu C.K., & Weingart LR. (2003). Task versus relationship conflict, team performance, and team member satisfaction: a meta-analysis. The Journal Of Applied Psychology, 88(4), 741-9.

Erdem, F., & Ozen, J. (2003). Cognitive and affective dimensions of trust in developing team performance. Team Performance Management: An International Journal, 9(5/6), 131-135.

Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative science quarterly, 44(2), 350-383.

Looking inside the boardroom: Boards and their reflexivity engagement