Abstract
This article aims to clarify the effects of conflicts on the task performance of supervisory boards in education. Management studies on conflicts on corporate boards sometimes find a positive effect and sometimes a negative effect on the performance of boards. These mixed findings are considered a result of the triggering of harmful personal conflicts through disagreements about process and content within boards. In order to gain more insight into the complex ways in which conflicts influence the effectiveness of supervisory boards in education, a multidimensional operationalisation of conflict has been designed, in which three different types of conflict – task, process and relationship conflict – are distinguished. Chairs of the executive boards of schools in the Netherlands (N = 300) were questioned via a survey on various conflict dimensions and their supervisory board’s task performance. The results of hierarchical regression analyses underscore the need for a multidimensional approach to conflicts on supervisory boards in education. Task conflict has a positive effect and relationship conflict a negative effect, while process conflict seems to have no significant effect on supervisory board effectiveness. Furthermore, task conflicts have a mitigating, rather than a triggering, effect on relationship conflicts. Supervisory boards in education should therefore not avoid substantive discussions on the differences of opinion among their members to be more effective.
Introduction
In order to understand the current situation with regard to the governance of schools in the Netherlands, it is necessary to explore the roots of the dual system in the Netherlands. That dual system of public and private education, which are both publicly financed, is a consequence of the ‘school struggle’. The school struggle is at the heart of the formation of the Dutch nation state in the late eighteenth and early nineteenth centuries. The struggle arose from a deep divide in terms of religion and power relations between the Protestant majority and the Catholic minority, each pursuing their own religious education, and the liberals who had a single, neutral, non-religious public education ideal (Hooge, 2017; Waslander, 2010). The school struggle came to an end when, after almost a century of political debate, an agreement was reached. The revised 1917 constitution encompassed the freedom of education and the financial equalisation of neutral education and religious education.
These provisions in the constitution have since been much debated but have remained unchanged – even after the depillarisation and secularisation in the 1960s. As a result, about two-thirds of Dutch schools are privately founded and governed by parents and/or religious groups, but are still publicly funded. About one-third of the schools in the Netherlands are public, founded by local government and run by citizens and parents in a private legal entity (Hooge, 2017).
Since the 1980s, the transfer of decision-making power and responsibility from the central government to schools has become an increasingly important element in the reform of educational governance in various countries within the Organisation for Economic Co-operation and Development (OECD). Governments retreated and there was little discussion on this decline of central governance, but the way in which schools dealt with the increasing autonomy was problematic. According to the new administrative vision, more market forces and fewer rules would lead to more quality, but they also led to calculating behaviours and the improper use of competences (Bronneman-Helmers, 2011).
The Netherlands is historically exceptional in the degree to which schools are autonomous and therefore provides a good case study for investigating the consequences of increased school autonomy on school governance (OECD, 2013).
Educational reforms and deregulation in the 1980s led to a further increase in the autonomy and the size of educational institutions in the Netherlands (Bronneman-Helmers, 2011; Karsten, 1999). In the 1990s, it became clear that the traditional one-tier governance model – where a governing board consisting of volunteer parents and local notables delegated executive power to a principal or head teacher – was no longer effective for these large, heterogeneous and autonomous schools. This led to a transition towards a two-tier governance system, wherein a supervisory board with non-executive directors supervised the executive director(s) of the school. The supervisory board in education is typically composed of five or seven members with different expertise (such as financial, legal and educational). These members are volunteers, recruited by the supervisory board themselves through an open application procedure. The executive board of directors often consists of just one person: the chair of the board of executive directors (comparable to the chief executive officer in for-profit organisations). The supervisory board has the legal duty to control and advise the executive board: they decide on the yearly budgets, the overall policy of the school and act as employer for the executive director. Supervisory boards meet on average six times a year. Unfortunately, a series of incidents of financial mismanagement and poor quality of education in large Dutch educational institutions resulted in supervisory boards in education appearing to be absent. It made the question ‘where was the supervisory board?’ a hackneyed phrase in the inevitable reports and studies following these incidents. Although there is never a lack of proposals for improvement through more rules, codes and laws, we simply do not know what the characteristics of good educational governance really are. What makes supervisory boards in education effective?
The inner workings of supervisory boards in education are still underexplored – a classic black box. Previous research in corporate governance suggests effectiveness has less to do with the structure and composition of the board than with the actual behaviour of boards (Huse, 2005; Van Ees et al., 2009). This should be the same for school boards, as Deborah Land argues in her review of school board effectiveness: ‘the ability of board members to work together and reach consensus is essential for boards to exercise their authority’ (Land, 2002: 254).
Although most school boards themselves believe they contribute to the quality of education (Hooge and Honingh, 2014), convincing scientific evidence for the relation between board control and school performance is still lacking (Honingh and van Genugten, 2017). There are quite a lot of causal steps between the performance of a school board as a team and the educational performance of the schools they govern, and even more between the educational performance and the supervisory boards in education. Moreover, when a firm or a school performs well, it does not necessarily indicate that the supervisory board also functions well. Pugliese et al. (2014) even find a negative correlation in for-profits: non-executive directors invest less in the quality of their task when a firm does well.
The obstacles to understanding board effectiveness are both theoretical, given the different and sometimes unknown intervening variables, and practical, given the time between changes in supervisory board behaviour and the publication of educational results. It is thus difficult to judge the relationship between good governance and educational performance (Ehren et al., 2016; Land, 2002). Therefore, in this research, ‘board task performance’ is used to assess supervisory board effectiveness. Board task performance is defined as a board’s ability to effectively perform both its control and its service tasks. Classifications of board tasks vary in description, number and nuance, and how one sees the tasks and roles of a board strongly depends on one’s own theoretical premises. Nevertheless, Huse’s (2005) attempt at a unified classification by approaching board tasks from three different perspectives: the internal, external and strategic focused on the two major board tasks, service and control, seems the most common and useful approach. The control task of the supervisory board refers to its task of overseeing management while the service task refers to the advisory role of the supervisory board (Forbes and Milliken, 1999; Heemskerk et al., 2015).
A multidimensional model of conflict in supervisory boards in education
The concise and researchable model presented by Forbes and Milliken (1999) has been one of the most promising avenues in the behavioural study of boards of directors. They see boards as ‘large, elite and episodic decision-making groups that face complex, multifaceted tasks that involve strategic-issue processing’ (Forbes and Milliken, 1999: 492), and they note that since boards are not involved in implementation, their output is entirely cognitive. They integrate the literature on team performance in corporate governance and formulate three behavioural antecedents for effective board behaviour: task conflict, effort, and the use of knowledge and skills.
The notion that effort and the use of knowledge and skills are important determinants of the effectiveness of boards has now become sufficiently well known. Several studies have shown that the effect of both exceeds that of formal rules or regulations for the composition of boards (Minichilli et al., 2012). However, the effect of conflicts is less clear (Heemskerk et al., 2017).
Studies on school board behaviour and effectiveness have often been limited to the negative aspects of conflict within school boards (Ford and Ihrke, 2016; Grissom, 2009, 2012; Land, 2002). This one-sided and one-dimensional orientation leads to the potentially harmful advice to avoid tension and conflicts between, for instance, the principal or superintendent and the school board (Ford and Ihrke, 2016; Land, 2002). An exception is the study by Heemskerk et al. (2015) on supervisory boards in education in the Netherlands. They focus on the positive, cognitive aspects of conflicts as disagreements between group members about the content of the tasks to be performed because of differences in ideas and opinions. Their advice is therefore the complete opposite; to be effective in advising the executive directors, supervisory boards in education need to thoroughly discuss their varied views and perspectives (Heemskerk et al., 2015).
What should boards in education do with this contradictory advice? Should they avoid conflicts or promote robust discussion? Recent studies on conflict and team effectiveness suggest that it is crucial to recognise the multidimensionality of conflicts in order to clarify the effect they have on teams. Conflict is not a monolithic concept; it has different dimensions and different effects on team effectiveness in different contexts (De Wit et al., 2012).
The history of the classification of conflict dimensions, as outlined by Jehn (2014), shows a tendency to a tripartite distinction between task conflicts, process conflicts and relationship conflicts. Task conflicts are focused on the content of the task at hand, process conflicts are focused on how tasks should be accomplished, and relationship conflicts are driven by interpersonal incompatibilities resulting in tension, animosity and annoyance among team members (Jehn, 1995, 1997).
Intragroup task conflict refers to the disagreement between board members about the content of the tasks to be performed due to differences in viewpoints, ideas and opinions (Forbes and Milliken, 1999; Jehn, 1995). Task conflicts are regarded as crucial for improving the quality of decision-making in boards (Heemskerk et al., 2017). Two meta-analyses of the conflict-performance literature provide a complex and detailed understanding of the relationship between conflict and group performance (De Wit et al., 2012; O’Neill et al., 2013). They identified two conditions in which task conflicts have a more positive effect on team performance. First, task conflicts tend to be more positive in strategic and complex decision-making and more negative in routine tasks (O’Neill et al., 2013). Second, the effects of task conflicts are less negative or even more positive in teams functioning higher in the organisational hierarchy (De Wit et al., 2012). Since supervisory boards in education are typically involved in non-routine and complex decision-making at the apex of the organisational hierarchy, intragroup task conflicts are expected to enhance the performance of these boards.
Intragroup process conflicts, on the other hand, generally have a negative effect on the effectiveness of teams because members of the group are steered away from the performance of their duties. Process conflict within a supervisory board in education has a place as internally focused conflict, that is, conflict about the organisation of the board itself (Who is on which committee? How do we evaluate ourselves? What does the resignation schedule look like?). These process conflicts are related to task conflicts, as conflicts over the content and execution of tasks are often intertwined. Task conflict and process conflict admittedly are interrelated (Greer et al., 2008), but both are conceptually and empirically so different that it is important to include process conflicts as a distinct dimension in the behavioural study of boards (Kerwin et al., 2011). Where task conflicts have a positive effect by bringing different perspectives and possible objections to the table, process conflicts are found to increase anger, animosity and negative attitudes towards the team (Behfar et al., 2011; Jehn, 1997). Process conflicts distract team members from the team’s task goals, leading to procrastination or unproductive reticence among team members (Margarida Passos and Caetano, 2005). The effect of process conflict on team performance was even found to be more negative in the non-routine or complex decision-making that is typical of boards in non-profits (Hamm-Kerwin and Doherty, 2010).
Intragroup relationship conflicts are conflicts within the board that are driven by interpersonal incompatibilities resulting in tension, animosity and annoyance among board members. Relationship conflict results in low cohesion and is likely to hamper the effectiveness of the group (Jehn, 1995). Westphal and Bednar (2005), for instance, have found that low cohesion between non-executive directors contributed to failures in communicating their concerns about their organisations’ strategies.
Since the dimensions of conflict vary in different types of teams (see e.g. Conlon and Jehn, 2009, on conflict in rock bands), and because both the conceptual and the methodological delineation between these three common types of conflict is not yet completely clear (Bendersky et al., 2010, 2014), it is necessary to consider the specific dimensionality of conflict in the context of boards. An important additional distinction typical for conflicts within supervisory boards in education is the difference between intragroup conflicts, that is debate and arguments among the non-executive board members, versus those between non-executive board members and the executive director(s) (Heemskerk et al., 2017). As with intragroup conflicts, these executive board conflicts can be divided into three types: executive board task conflicts, process conflicts and relationship conflicts.
Executive board task conflicts are essential for a supervisory board’s effectiveness, as it is the supervisory board’s task to control and advise the executive directors by challenging and judging their assumptions and strategic choices. In previous research, professional tension between the executives and non-executives was found to be a sign of a well-performing supervisory board (Heemskerk et al., 2017). Cohesive governance teams with a focus on collective decision-making, such as supervisory boards in education, are prone to what Sundaramurthy and Lewis (2003) describe as ‘reinforcing cycles of collaboration’ in which managerial entrenchment leads to low-performing boards. To avoid such detrimental reinforcing cycles, supervisory boards need to balance collaboration and control, and thus at times take a critical stance towards their executive directors. Executive board task conflict is essential in this balancing act as it stimulates critical feedback and counters groupthink (Sundaramurthy and Lewis, 2003).
Executive board process conflicts are mostly about the cooperation and definition of roles between the non-executives and the executives. Due to the increasing autonomy of schools and the accompanying structural and cultural changes in their governance, many supervisory boards in education are looking for their proper role. This leaves room for different views about this among executive directors and supervisory boards. Chairs of the board of executive directors sometimes complain that supervisory boards are too involved with the details of daily task execution, while supervisory boards sometimes do not feel involved and informed by management. Process conflicts between executive directors and supervisory boards in education will hinder effective task performance by these boards.
Executive board relationship conflicts are likely to have a negative effect on supervisory boards’ effectiveness, as they could cause executive directors to be less inclined to seek advice from their supervisory boards (Heemskerk et al., 2017). Westphal (1999), for instance, found that friendship ties between executive and non-executive directors were associated with the frequency with which the executive directors sought strategic advice from their supervisory boards. Relationship conflicts might undermine these social ties and fuel distrust and political battles between executive directors and supervisory boards in education, reinforcing each other in a vicious circle (Sundaramurthy and Lewis, 2003). A model depicting the different dimensions of conflicts in supervisory boards in education is presented in Figure 1.
A multidimensional model of conflict in supervisory boards.
Chair participative leadership
The supervisory board chair plays an important role in steering the behavioural group processes within the supervisory board and particularly in creating an environment in which a supervisory board through participative decision-making can be effective in its performance. As The UK Corporate Governance Code puts it: The chairman is responsible for leadership of the board and ensuring its effectiveness on all aspects of its role…The chairman should also promote a culture of openness and debate by facilitating the effective contribution of non-executive directors in particular. (Financial Reporting Council, 2014: 7)
Triggering effect
Intragroup task conflicts are found to have a more positive effect on group effectiveness when they are weakly correlated with relationship conflicts (De Wit et al., 2012). This is consistent with a frequently suggested explanation for the negative aspects of task conflict in which task conflict functions as a trigger for relationship conflicts (De Dreu and Weingart, 2003; Mooney et al., 2007; Simons and Peterson, 2000). Task conflicts can have a more positive effect on the performance of a team when these conflicts are prevented from escalating into relationship conflicts. Eisenhardt et al. (1997), for instance, found that the management teams of tech companies that were most effective both in speed and quality of decision-making were those that were able to keep constructive conflict about issues from degenerating into dysfunctional interpersonal conflict. Intragroup process conflict likewise is often considered a possible trigger for relationship conflict (Greer et al., 2008; Martínez-Moreno et al., 2012). Discussion about who should do what easily degenerates into personal struggles and animosity.
Board cohesion
Cohesion within a supervisory board in education is a group-level construct (Hogg, 1993) that refers to both social cohesion, which is the attraction of members to the group, and task cohesion, which is the willingness to work with the group to accomplish shared objectives (Carron and Brawley, 2000). Cohesion improves participation and communication within a group and also increases the acceptance of the goals, tasks and roles among group members (Casey-Campbell and Martens, 2009). However, Forbes and Milliken (1999) do not assume a linear positive correlation, rather an inverted U-shaped relationship between cohesion and board effectiveness. That is, while insufficient cohesion is negative to task performance, too much cohesion might cause groupthink within a board. Groupthink develops when members feel so deeply involved with a group that in their wish for unanimity they lose their ability to reasonably consider alternative decisions (Janis, 1982).
The relationship between cohesion and performance is, in reality, even more complex, and although groupthink is a serious threat to supervisory boards, cohesion in itself does not seem to be the sole cause of groupthink (Mullen et al., 1994). The emergence of groupthink also depends on the source of the cohesion within a team. If the cohesiveness of a team is grounded in its members’ joint commitment to their task (as is often the case with supervisory boards) instead of on members’ interpersonal relationships, groupthink is less likely to occur (Bernthal and Insko, 1993).
Part of the complicated relationship between cohesion and performance seems to run through conflicts within a team. The ‘triggering effect’ of task conflict on relationship conflict seems to be mediated by cohesion within a group, as suggested by Ensley et al. (2002). They argue that groups with a high degree of cohesion are better able to prevent task conflicts from escalating into relationship conflicts.
Previous studies point to the crucial role of the supervisory board chair in dealing with conflicts within boards (Heemskerk et al., 2017). The mediated effect of the board chair’s participative leadership on the relation between conflict and performance – as described in hypothesis 2 – is largely due to the decreasing probability that task conflict and process conflict degenerate into personal struggles and feuds within a board. Good chairs who are able to let everyone participate in the deliberations and create a climate for open debate reduce the risk of disagreements on the issues at hand deteriorating into detrimental relationship conflicts (Sundaramurthy and Lewis, 2003).
Method
Gaining access to the black box of the boardroom is seen as the main challenge for behavioural research on good governance. Moreover, social processes are harder to quantify than the formal characteristics of a board, such as the composition or legal structure. In this study, therefore, a survey was designed with several Likert scale items to test the aforementioned hypotheses, providing both access to the inner workings of supervisory boards in education and quantifiable data on the social processes within these supervisory boards.
After pretesting the survey through cognitive interviewing of 10 school governance experts, chairs of boards of executive directors and supervisory board members, the adjusted survey was distributed in 2016 among all educational organisations with membership of the Dutch association of supervisory boards in education (Vereniging van Toezichthouders in Onderwijsinstellingen, VTOI). This sample included almost 500 organisations with differences in size and type of education, from primary school to vocational education and training (VET).
Chairs of boards of executive directors were invited to take the online survey through a personalised email emphasising the anonymous and confidential processing of the data in order to increase the number of responses as well as reduce the potential for common method bias (Podsakoff et al., 2003). Chairs of boards of executive directors, as the chief executive within the organisations, know their organisations well and are also, more so than other school board members, in a position to make statements on the functioning of the board as a whole (Minichilli et al., 2012).
After two rounds of reminders, 300 (60%) out of 498 responded. Of these respondents the vast majority, 62%, managed a medium-sized educational institution (between 2000 and 8000 pupils), 9% were chair of the board of executive directors of a large educational institution (over 8000 pupils), and 29% worked for a small school with fewer than 2000 students. Most respondents worked for an educational institution for primary (45%) or secondary (30%) education. A smaller number associated with an institution for VET (6%) or working in special education (4%). In addition, some of the respondents managed an educational institution where primary education and secondary education (5%) or primary education and special education (7%) were both provided. Five respondents (2%) were chair of the board of executive directors of regional partnerships that often focused on special educational needs.
The dependent variable of board task performance was operationalised through two variables: control task performance and service task performance. Control task performance was measured through six items on a seven-point Likert scale questioning the controlling participation of the supervisory board on sector-relevant issues (e.g. educational quality) and two seven-point Likert items inquiring as to what extent the supervisory board plays an active role as the employer of the chairs of the board of executive directors and to what extent it adequately monitors the performance of the executive directors. Service task performance was measured by questioning the advisory participation on the same six issues and through three items inquiring as to what extent the supervisory board uses its networks to aid the organisation, acts as a sparring partner and is an important adviser to the chair of the board of executive directors.
The independent variables were based on seven-point Likert items inspired by the suggestions of Forbes and Milliken (1999), Wageman et al. (2005) and Sellevoll et al. (2007). Additionally, the items on conflict made use of relevant items from the Intragroup Conflict Scale developed by Jehn (1995). Supervisory board chair participative leadership was assessed using seven-point Likert items based on relevant participative decision-making items from the Empowering Leadership Questionnaire (Arnold et al., 2000).
To test whether the various Likert scale items from which the dependent and independent variables were constructed had been sufficiently consistent, we determined Cronbach’s α for these variables. Given the exploratory nature of our research, the values of α seem to confirm the internal consistency of the items. Based on the reliability analysis and an EFA, three of the five initial items on intraboard and executive board task conflict were excluded from the analysis, as they seemed to measure a concept more related to the shared cognition of the board as a team (e.g. board members think and reason in similar ways) and seemed less related to the occurrence of task conflicts.
Table 1 provides an overview of the reliability statistics and the number of items, as the interpretation of α depends on the number of items from which a variable has been constructed.
Reliability statistics of constructs.
Both dependent and independent variables are measured using the same survey instrument. It is often believed that such a research design leads to inflated correlations through common method variance or bias (Podsakoff et al., 2003). The risk of common method bias, however, seems somewhat exaggerated. Studies that control for common method bias almost never find it to have a confounding effect, and in simulation studies the effect is negligible (Fuller et al., 2016; Spector, 2006). A Harman’s single factor test of the data in this study showed that less than 50% of the variance could be attributed to one factor (Chang et al., 2010). Common method bias thus does not seem to pose a threat to the reliability of the correlations in this study.
Results
Table 2 displays the mean number of cases, standard deviations and the bivariate correlations of all variables. The correlations in Table 2 support the main hypotheses, as both intraboard and executive board task conflict are positively related to both service and control task performance, while intraboard and executive board relationship conflict seem to have a negative effect on a supervisory board’s task performance. Contrary to expectations, there appears to be a positive relationship between intraboard process conflict and a supervisory board’s service task performance. It is also noteworthy that there is a negative correlation between intraboard task conflict and relationship conflict, which seems to contradict hypothesis 3 in which it was expected that task conflicts trigger relationship conflicts.
Correlations.
*p < .05; **p < .01; † p < .1.
To test the first two hypotheses a hierarchical multiple regression analysis was performed on four models. In model 0 we included the control variables: chair of the board of executive directors’ tenure, frequency of supervisory board meetings and executive board chair consultation. Next, in model 1, the three dimensions of intraboard conflict – task conflict, process conflict and relationship conflict – were added. Model 2 added the supervisory board chair’s participative leadership to the analyses. Model 3 further added the dimensions of executive board conflict. Model 4 explored the interaction effect of chair participative leadership and conflict. The findings of the hierarchical multiple regression analysis are presented in Table 3.
Regression analyses of main effects.
*p < .05; **p < .01; † p < .1.
As expected from the correlation analysis, the control model only showed executive tenure and the frequency of executive board chair consultations to have significant effects on the performance of the control and service tasks respectively. The fit of the different models furthermore shows that the model with the control variables has little explanatory value (
The first model, with the three dimensions of intraboard conflict, shows, as expected, a significant positive effect for task conflicts and a negative effect for relationship conflict, but, contrary to the correlation analysis, intraboard process conflict has no effect on control task execution and the impact on service task performance now turns positive. Adding chair participative leadership in the second model leads to a further increase in explanatory power (
The hierarchical regression analysis partially confirms the first hypothesis. Different types of conflict indeed have different effects on task performance by supervisory boards in education. Task conflicts have a positive effect, relationship conflicts have a negative effect and, contrary to the expectation in hypothesis 1, process conflicts have a small positive effect on supervisory boards’ task performance. The dimensionality of executive board conflicts does not seem to be significant. Only executive board process conflict appears to have a small negative impact on the control task alone. This is consistent with the hypothesis that the effect of executive board conflict would be stronger for control than for service task performance in supervisory boards in education.
Chair participative leadership appears to have a solid standalone effect on the performance of supervisory boards in education in the hierarchical regression analysis. However, this effect, unlike the expectations expressed in hypothesis 2, cannot be sufficiently explained by interaction effects with the various dimensions of conflict. A small significant relationship with the controlling task of supervisory boards has only been found for the interaction effect of chair participative leadership and executive board relationship conflict, as presented in model 4 in Table 3.
The hypotheses 3, 4 and 5 with regard to the triggering of relationship conflicts by process and task conflicts and the moderating effect of cohesion and chair participative leadership were tested by a second hierarchical regression analysis, with relationship conflicts as the dependent variable. Three models were assessed in this analysis. There was a control model with three control variables: executive tenure, frequency of supervisory board meetings and executive board chair consultation. The first model looked at both task and process conflict and the second model assessed both of the assumed moderators: cohesion and chair participative leadership.
Hypothesis 3 on the triggering effect was partially supported. Process conflict was indeed found to have a significant positive effect on the occurrence of relationship conflict within supervisory boards in education. Intraboard task conflicts, on the other hand, seem to rather mute or prevent, instead of trigger, relationship conflicts, since their effect was found to be significantly negative.
Adding cohesion and chair participative leadership to the regression model seems to fully absorb this damping effect of task conflicts. Board cohesion has a strong negative effect on the occurrence of relationship conflicts within a supervisory board, while participative leadership has a small positive effect. A supervisory board chair who invites members to voice their opinions seems to run the risk of triggering relationship conflicts. However no interaction effects of cohesion or chair participative leadership were found to be significant, except for the combined effect of participative leadership and process conflict.
To further test hypotheses 4 and 5, subsample regression analyses were performed. An overview of the results is presented in Table 4. First, the initial sample was divided into more cohesive and less cohesive supervisory boards with the mean level of cohesion as the cut-off point. Contrary to the expectations formulated in hypothesis 4, the mitigating effect of task conflict on relationship conflict was found to be significant in less cohesive and insignificant in more cohesive supervisory boards.
Regression analyses of triggering effect on relationship conflicts.
*p < .05; **p < .01; † p < .1.
This could be explained by the fact that – as shown in Table 5 – there is a strong linkage between cohesion and a supervisory board’s level of relationship conflict. In less-cohesive boards the mean level of relationship conflict is 2.15 (SD = .87), while in more-cohesive boards this drops to no more than 1.32 (SD = .48) on average. With so little relationship conflict occurring in more cohesive supervisory boards, it is not surprising that there remains no significant mitigating effect of task conflict, since there is little left to mitigate.
Subsample analysis of the triggering effect on relationship conflicts.
*p < .05; **p < .01; † p < .1.
In a second subsample regression analysis the initial sample was divided into more participative leading supervisory board chairs and less participative leading chairs with the mean level of chair participative leadership as the cut-off point. The regression analysis in Table 4 shows how the mitigating effect of task conflict on relationship conflict found in the previous analysis depends on the level of participative leadership by the chair of the supervisory board. In boards with less participative leading chairs the mitigating effect is insignificant, while in more participative chairs the effect is rather strong. Moreover, more participative chairs also seem to dampen some of the triggering effect of process conflict on relationship conflicts.
Discussion and conclusions
The findings in this research underscore the need for a multidimensional conceptualisation of conflicts in boards. Previous research on conflict in school boards is often solely focused on negative conflict dimensions within the board of education, such as ‘destructive conflict’ (Ford and Ihrke, 2017) or ‘personalised disagreements’ (Ford and Ihrke, 2016), which have a negative effect on the performance of schools and school boards. This creates the impression that conflicts in the governance of education should be avoided. Grissom (2012: 311) thus states that: ‘Conflict reduction and conflict management appear to be important strategies for board success, particularly in complex decision contexts’. However, especially in complex decisions at the executive level, task conflicts are found to contribute to effective decision-making (De Wit et al., 2012; Heemskerk et al., 2015). The findings in this study also indicate that the reality is more complex. Conflict is found to be a multidimensional concept, and where relationship conflicts are indeed detrimental to the effectiveness of governance, substantive debate within the board contributes to its performance.
In the context of Dutch supervisory boards in education, as expected, task conflict seems to have a positive effect and relationship conflict a negative effect, while process conflict seems to have no significant effect on a supervisory board’s effectiveness. Different types of conflicts thus appear to behave differently. The fact that executive board conflicts do not affect the effectiveness of supervisory boards – with the exception of the effect of process conflicts on the controlling task – seems to suggest that this dimension does not add much to the concept of conflicts. At the same time, it also underscores how such executive board conflicts behave differently compared to intraboard conflicts.
That only process conflict between chairs of boards of executive directors and school boards has a significant negative effect on a supervisory boards’ controlling task execution fits the uncertainty within many, especially recently formed, Dutch supervisory boards in education on the role of the supervisory board and the demarcation of the responsibilities and tasks of the chair of the board of executive directors versus the supervisory board. Supervisory boards in education that are too busy discussing the degree to which they are in control and that argue with executives over what is whose task could well fall short of their monitoring tasks by either distancing themselves too much or because their executives become reluctant to adequately inform them in order to avoid detailed discussions about implementation issues (Heemskerk et al., 2017).
Personal conflicts among board members indeed appear to be triggered by process conflicts within a supervisory board, but, contrary to our expectations, task conflicts seem to have a dampening effect rather than a reinforcing effect on relationship conflicts. This gives supervisory boards in education even less reason to avoid substantive confrontations on the issues at hand. This mitigating effect of task conflicts on relationship conflicts is fully absorbed by the much stronger mitigating effect of cohesion within a board. This is in line with previous research on supervisory board behaviour that shows that task conflicts have a significant positive effect on cohesion within supervisory boards in education (Heemskerk et al., 2015). If board members feel comfortable with each other, an open and substantial discussion only further strengthens these mutual relationships.
The leadership style of the supervisory board chair appears to be of great importance for the successful functioning of boards, both in their controlling and in their advising tasks. Contrary to expectations, this positive effect of chair participative leadership appears not to be an indirect effect that runs through task conflict. Apparently, the positive effect of participative leadership goes beyond creating an optimal environment for constructive conflict. The role of the chair and the functioning of participative leadership in supervisory boards in education requires further research to gain insight into how it enhances board performance. It seems obvious to suppose participative leadership also has an impact on the two most important behavioural drivers of the effectiveness of teams: the use of knowledge and skills, and the level of effort (Forbes and Milliken, 1999).
The fact that chair participative leadership seems to have a limited triggering effect on relationship conflicts within a supervisory board is not unexpected. Participative leadership provides space to receive input from several viewpoints around the table, but it can also cause or exacerbate tensions. It seems that this effect does not run so much through the task conflicts stimulated by participative leadership, which are found to have a limited and mainly damping effect. Supervisory board chair participative leadership does seem to trigger relationship conflict through the process conflicts it gives room to.
This research shows that conflict within a supervisory board is not an unambiguous matter and that more research is needed to clarify the dimensionality of conflicts in different governance contexts. The most important practical implication of this research is that relationship conflicts on boards do not appear to arise due to strong substantive discussions. In fact, a good debate seems to make a board an even closer team. Supervisory boards in education should therefore not shun substantive discussions on the differences of opinion among their members.
A limitation of this study is that the chairs of the boards of directors in this study not only report on the dynamics within the supervisory board but also on the dynamics between themselves and the supervisory board. This could possibly lead to self-report bias, especially when it comes to conflicts, for these are ‘undesirable behaviours’ that may result in socially desirable answers (Donaldson and Grant-Vallone, 2002). This might have contributed to the non-significant results for executive board conflicts on the effectiveness of supervisory boards. Future research should seek a response from both parties on the dynamics between executives and supervisors to avoid such self-report bias.
For future studies, it would also be interesting to see whether the conclusions found here for Dutch supervisory board governance in education also occur in other countries with other governance arrangements. In addition, the choice of a cross-sectional research design makes this study a snapshot. Future research could provide more insight into the dynamics of the relationship between conflicts and the effectiveness of supervisory boards in education through a longitudinal research design.
Footnotes
Acknowledgements
The author thanks E.M. Heemskerk for advice and comments on preliminary versions of this article.
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is part of the research programme Governance as Behaviour with project number 023.003.029, which is financed by the Netherlands Organisation for Scientific Research (NWO).
