Are school boards and educational quality related? Results of an international literature review

ABSTRACT

School boards have a central position in educational governance. They have to guarantee quality, monitor results and intervene if needed. In this article we report the results of a systematic literature review meant to identify what school boards do to contribute to educational quality. After five steps of literature selection and refinement, our initial database of 4939 was reduced to 16 empirical articles. This shows that there is actually very little empirical research on the relation between boards and educational quality. Looking more closely at the 16 remaining studies, we notice differences in the reach, quality and applied methods. Moreover, the differences between boards are considerable when looking at board composition and the scope of their autonomy. Altogether, these differences make it difficult to come up with general findings on how boards can contribute to educational quality. Nevertheless, we also found promising large-scale studies that offer a better understanding of the complex nature of the effects that boards may have on educational quality. Furthermore, in almost all of the studies, educational quality is operationalised as students’ test results, indicating a narrow perspective on educational quality.

KEYWORDS:

• School boards
• educational quality
• systematic literature review
• educational governance
• comparative study

Introduction

Since the 1980s, publicly funded schools in the Western world have been granted more autonomy (OECD 2012a). School governance has been decentralised, to enable schools to respond to diverse and local demands (Burns and Köster 2016). Schools have been put in charge of their own management. Instead of direct government control, boards have been established to manage the organisations (Wilks 2007). It was expected that school boards would ensure that the execution of their public task would be done in the most efficient and effective way (e.g. Cornforth 2003; Howard and Seth-Purdie 2005). Boards would also make the organisation directly accountable to the “users” of the services provided by the school and offer them a better choice.

Coinciding with this increased autonomy, school boards’ accountability for educational quality has increased. They are accountable for the performance of their schools and are therefore expected to monitor and enhance the quality of their schools (Hooge and Honingh 2014). Since evidence about school and student performance has become more readily available, parents and other stakeholders are becoming more demanding with regard to educational quality and hold their school boards accountable. Parents and other stakeholders require boards to show and explain how they are strengthening educational quality (e.g. Hess 2002). Consequently, countries like Australia, England, the Netherlands and districts in the USA have developed guidelines, codes of conduct and recruitment profiles for school board members, which are believed to indirectly improve educational quality (Hess 2002, OECD 2012b). To sum up, expectations of what school boards can achieve are high. These expectations raise the question of whether there is solid scientific evidence supporting the claim that boards can contribute to educational quality. Our study aims to answer this question.

Almost a decade ago, several scholars concluded that there was no strong empirical evidence supporting the idea that school boards and educational quality are related (Alsbury 2008; Delagardelle 2008; Johnson 2010; Land 2002). According to Land (2002), literature on school boards is “rife with conclusions and recommendations based on personal experience, observations and opinions, and a heavy reliance on anecdotal evidence rather than on well-designed research studies” (p. 265). Land (2002) concludes that there are “no studies in which governance variables are clearly defined, measured, and their impact on governance effectiveness and, more specifically, students’ academic performance quantitatively assessed” (p. 265). In a relatively small literature study, Johnson (2010) argues that “the educational literature is saturated with many opinion-based articles regarding effective governance” (2010, 28). He focuses on a few qualitative and quantitative studies, aiming to summarise the factors that seem to contribute to a board’s success in improving student achievement. From his review, he derives 12 board practices that appear to contribute to student achievement.

Although the literature studies of Land (2002) and Johnson (2010) are valuable in themselves, they have some limitations. Firstly, neither study provides a systematic description of the methods used for collecting and analysing the literature. Whereas Land initially used the Education Resource Information Centre (ERIC), she decided to broaden her search strategy by including studies about current and future roles of school boards and used more general studies and handbooks on “school governance” (see p. 230). Similarly, Johnson’s study lacks a clear description of the selection of articles also we did not find details of the criteria he used for selecting his 12 essential board practices. The scope of this study also seems to go beyond school boards, as its analysis reveals a strong focus on leadership (e.g. Leithwood and Jantzi 2008; Marzano and Waters 2009; Murphy and Hallinger 1988).

Second, both review studies have a relatively narrow geographical scope, focusing on boards and their schools in the USA. The recent changes in the position of publicly funded school boards in several Western countries indicate a clear need to include literature from other countries. Norway, Sweden, the UK, New Zealand and Australia have adopted acts or charters creating schools that are run by private foundations and operate autonomously from local authorities, but remain publicly funded. In Sweden, for example, grant-aided independent schools (friskolor) have been introduced, and in England academies have been founded (e.g. West 2014). This development shows that there is growth in the number of more autonomous school boards. The recent changes in the role and position of school boards in educational governance fuels the need to examine whether there is indeed scientific proof for the effects that boards can have on educational quality. From a methodological perspective, the inclusion of a larger number of countries challenges the research design, of course, as it becomes more difficult to draw comparisons and to draw conclusions. As such, it is crucial to formulate clear selection criteria beforehand and to remain critical on differences in demographics, governance structures, school board roles and school systems between regions and countries when interpreting the data. In our study, we formulated a set of criteria for our country selection (next section).

The purpose of this systematic literature review was to generate a systematic overview of the literature available on the relation between school boards and educational quality. In particular, we looked at compositional and behavioural factors of school boards. We aimed to identify key publications in the scholarly literature about educational governance and, in so doing, to answer the questions of whether and how school boards and educational quality are related.

The next section explains the selection of countries and types of school boards in this study. The methodology section elaborates on the exclusion and inclusion criteria during the subsequent steps of the systematic literature review. The results section presents descriptive results and reflections to answer our main question.

School boards, countries and board models

To answer the questions of whether and how school boards and educational quality are related, it is crucial to examine the extent to which school boards have decision-making authority. In order to do this, we first had to select those countries in our study where decision-making is decentralised at the level of the school board. This is the case in the Netherlands, England and the Flemish Community of Belgium (OECD 2014).

Because of recent experiments with more autonomous boards, we also had to include new forms of school boards in our study, like the friskolor from Sweden, charter schools in the USA and academies in the UK (Chapman and Salokangas 2012; Waslander and Hooge 2015; West 2014). The particulars of these schools and the policies that manage them may differ from one country to another, but the common idea is that their boards have more decision-making authority because the schools are publicly funded and privately run. As the earlier literature reviews of Land (2002) and Johnson (2010) focused on school boards in the USA, we wanted to examine whether the body of knowledge currently available has grown substantially in recent years. We therefore included studies focusing not only on charter schools in the USA, but on regular school boards as well.

Despite similarities between school boards, we also had to acknowledge particulars in: (a) the board models that are in place and (b) the composition of the school boards. As to board models, it is clear from the literature that there are many different types of boards (e.g. Carver 2006; Van Thiel 2015). Here, it is crucial to distinguish between one-tier and two-tier boards. A one-tier board consists of directors, with both executive and non-executive members, for example the Anglo-Saxon management board. A two-tier board consists of two boards: one board of directors or executives, and one board of overseers. Each board has different roles and competencies. In this article our main focus is on executive board members, as they carry the responsibility for the educational quality in their schools and have to intervene if needed.

As to the composition of boards, it is important to mention that board members can be either elected or appointed. In Belgium, England and the USA, at least some board members are elected, while school boards in the Netherlands and the boards of most academies and friskolor are appointed. Whether or not board members are elected is no mere detail; it has far-reaching consequences for the extent to which boards have to operate in a political context. Compared to the USA and New Zealand, for example, school boards in the Netherlands lack democratic accountability mechanisms and operate at a relative distance from the political dynamics of government (Hooge and Honingh 2014; OECD 2016). In countries where it is common to vote for board members, constraints are imposed to keep the balance between parents, staff governors, local authority governors, foundation governors from the school’s founding body, partnership and sponsor governors from the wider community, and associate (non-voting) members, all of whom are appointed by the governing body (James et al. 2011).

Methodology

Systematic review methods are increasingly being used to comprehensively assess the current state of knowledge, by applying rigorous, objective and transparent steps and criteria to reach conclusions from a body of scientific literature (Petticrew and Roberts 2006). In contrast to a traditional literature review, a systematic literature review aims to avoid any bias, intentional or unintentional, in the selection of the publications by identifying all likely relevant literature in transparent and explicit steps (Petticrew and Roberts 2006). In public administration, such systematic reviews are growing in popularity: their thorough analysis enables identification of areas where substantial progress has been made and where future research could be directed (De Vries, Bekkers, and Tummers 2015; Tummers et al. 2015; Voorberg, Bekkers, and Tummers 2015). In this study, we aimed to examine whether there has been any progress in assessing the relation between boards and educational quality and to distil key publications that may help to answer questions about what works and what does not. The method therefore helps to identify areas of uncertainty or where relevant research is lacking and shows where new studies are needed. Since the previous literature studies point to a lack of empirical evidence on how school boards and educational quality are related, a more systematic literature review seemed the right approach.

The first step in this study was to design review protocols to ensure a transparent and rigorous selection of studies on the relation between school boards and educational quality. The review began with a broad focus on concepts that would fit into the aim of our study, because the scientific discussions of educational quality and diversity of board models, and the literature on school board functioning, are dispersed. Literature was selected using two main inclusion criteria: (a) reviews, editorials, peer-reviewed articles, published and online first, to ensure the quality of the included articles; (b) articles explicitly designed to identify characteristics of school boards to contribute to the school’s educational quality.

We used an elaborate list of search terms of both the independent variable (composition, behaviour - and process characteristics of boards) and the dependent variable (aspects and forms of educational quality). We performed two searches in two databases: the Web of Science (WoS), one of the largest scientific databases for social research; and Education Resource Information Centre (ERIC), the domain-specific database that collects only educational research. The first search was formulated to find literature on the actions of school board members, how they plan and how they analyse information; the second search was for literature on the roles and behaviour (or behavioural aspects) of boards (see Figure 1).

Figure 1. Schematic representation of the systematic literature selection.

The refinements we made in WoS and ERIC helped us to select relevant research domains, countries, publication years, and educational sectors. In WoS we selected: educational studies, public administration, management and governance and the countries we selected for our study. In ERIC we were also able to select primary and secondary education.

Results

The sample included articles in English, published in 1996–2016 (WoS) and 2007–2016 (ERIC). As the main aim of this study was to determine whether, and how, school boards are related to educational quality, there was no need to go further back. These inclusion criteria yielded 4939 unique articles, after elimination of overlap between the ERIC and WoS databases.

The abstracts in both databases were selected for further analysis based on two exclusion criteria. First, we limited our sample to articles with an explicit focus on school boards by examining the key words, titles and abstracts. In this step we removed articles from the dataset that were not related to the study topic, for example, studies on white boards, teacher boards, dartboards and Board versus Brown. Second, we excluded articles that dealt with other school sectors, such as pre-school, higher education and medical schools. We also had to exclude articles reporting on areas and countries that were not part of our geographical selection, such as South Africa, Canada, Japan, Germany and China. These exclusion criteria yielded 277 eligible articles for full textual analysis. To determine the extent to which both coders reached the same conclusions, the intercoder reliability¹ was calculated on the coders’ judgments by a randomly selected subset (n = 897) of abstracts. The coders reached the same judgment in 97% of the abstracts, which is more than acceptable.

In the next sifting step, based on the full texts, it became clear that some articles paid relatively little attention to school boards. They were often mentioned in the text just to give more background and to illustrate the schools’ context. We excluded all articles that did not focus substantially on the board. Using these assessment criteria, progressive focusing yielded 71 articles that were considered eligible for further analysis.

The full texts of these 71 articles were carefully re-read and analysed by both researchers. A data extraction table was designed drawing on research into school boards and public-sector board behaviour (e.gNielsen and Huse 2010; Brown and Guo 2010; Leblanc and Gillies 2003) to systematically collect quantitative and qualitative data from the literature. The table was designed to provide data to answer the research questions and included the following categories: bibliographic information, focus of the study, methodology, nature of the empirics, country, board model, variables measured at the board level, conceptualisation of educational quality, and findings.

While reading the articles, we also found articles that were mentioned but which we did not find through WoS or ERIC. We made a list of these articles and examined them by using our search criteria and by checking whether they provided empirical information on the relationship between boards and educational quality. This method, known as snowballing, yielded 13 extra documents. Thus, in the end our database had 84 articles, which were carefully examined by coding them. The coding of articles using a coding protocol is the fifth and last step in a systematic literature review, as formulated by Petticrew and Roberts (2006). We created a codebook to code board aspects (e.g. board model, number of schools per board), board conduct (e.g. horizontal dialogue and quality management), focus of the study, research findings and aspects of educational quality (e.g. student achievement, reputation, safety, satisfaction and financial data). The codes were retrieved from literature on board dynamics and educational governance (Milliken and Martins 1996; Nielsen and Huse 2010; Brown and Guo 2010; Wellens and Jegers 2014; Ranson 2011).

Besides these codes for the dependent and independent variables, we also coded other characteristics, such as research methods used, countries on which the empirical data were based, and whether the study concerned primary or secondary education. During coding, it became clear that new codes for the selection, composition and background of board members needed to be added to the codebook. The coding of the literature was conducted using software for analysing qualitative data (NVivo, version 11).

Although systematic reviews are designed to be as comprehensive and transparent as possible, some limitations to this approach need to be considered (Petticrew and Roberts 2006). First of all, only peer-reviewed publications were included, to ensure scientific and methodological rigour in the analysed studies. Second, the review included only two scientific databases, WoS and ERIC, to combine the benefits of a general scientific database and a specifically education-oriented one. Finally, the review was limited to material written in English, although more evidence on the contributions that school boards make to educational quality may be available in other languages.

Analysis

As we were attempting to answer whether and how school boards and educational quality are related, it was crucial to include only those articles that clearly define school board level variables and as such provide empirical data on both board level and educational quality. During coding, we noticed these criteria to be of particular interest, as many studies seem to include the board level in their analysis without operationalising variables or presenting data on board level (e.g. Fusarelli, Kowalski, and Petersen 2011; Mintrom 2009; Roberts and Sampson 2011; Ranson 2012). There were also studies that focused on the functioning of the board without analysing its further implications for the school and its educational quality (e.g. Demie 2003; Du Bois et al. 2009; Hess 2002; Namit 2008; Trujillo 2012). The coding process helped us to distil the 16 studies that we hoped would give us an answer to our main question.

In the remainder of this article we will draw on these 16 studies, as they contribute to a better understanding of the composition of these boards, the behaviour of board members and their interventions in schools, and how this is related to educational quality. Twelve of these 16 articles provide data on USA school boards, two present data on UK school boards and two present data on Dutch school boards (see Table 1). The origin of the data matters because the number of schools governed by a board does vary. Small boards in the USA govern only four schools, whereas big boards can govern over 1000 schools. Boards in the USA generally operate at a greater distance from their schools. It is notable that most of the studies cover primary and secondary education, but authors do not explicitly distinguish between the sectors when presenting their results. The Dutch studies are the exception here, as both studies deal with primary education only.

Table 1. Crosstable countries and methods.

	USA	UK	Netherlands	Total
Quantitative (survey)	5	0	2	7
Qualitative (case studies)	3	0	0	3
Mixed methods	4	2	0	6
Total	12	2	2	16

Taking a closer look at Table 1, we notice a variety of methods that have been used in the 16 studies. In seven studies, the only method used was surveys of the board members. In most cases, the board chairs were approached to respond to the questionnaire. Recurrent themes in these surveys were: topics on the boards’ agenda, connections inside and outside the board (e.g. Andrews and Brewer 2014; Saatcioglu et al. 2011; Shelton 2010) and board composition (e.g. Hofman, Hofman, and Guldemond 1999; Meier and O’Toole 2004). Only two out of seven surveys were set up as longitudinal studies (Saatcioglu et al. 2011; Shelton 2010). The study by Van de Grift and Houtveen (2006) also had a longitudinal character but did not draw on survey data. Only two studies conducted multilevel analyses to determine board effects on pupils’ performances (Meier and O’Toole 2004) and on pupils’ math scores and well-being (Hofman, Hofman, and Guldemond 1999).

The presented case studies (Table 1) all turned out to be single case studies and had a reflective nature. Three out of three case studies describe the dynamics of a single board (Johnson 2012; McAdams 2000; Springfield 2008). Johnson (2012) not only reflects on his own experiences, he also uses documents and his own literature review to back up his work. These case studies partly confirm the idea that studies about school boards are based on a combination of personal experience, observations and opinions – which means that research frequently relies on anecdotal evidence. This cannot be said about the often carefully designed mixed-methods studies that present survey data combined with interview data (e.g. James et al. 2011; Plough 2011; Ranson et al. 2005). Occasionally, data in mixed studies have also been collected over a longer period of time, as was the case with Delagardelle (2008), who reports on the Lighthouse study.

Here, we have to draw attention to methodological issues, as most of the presented studies suffer from some methodological weaknesses. A first weakness in most of the quantitative studies was the difficulty researchers experienced to isolate variables. This was mainly due to the complex multi-layered nature and mutual dependency between board members, school leaders, teachers and practices in the classroom. As a consequence, it remains difficult, if not impossible, to speak of causal relations. The qualitative comparative studies also suffer from some weaknesses since the case selection was done based on the dependent variable rather than on the independent variable (Swanborn 2008). This design cannot provide valid causal explanations for student or district achievements. Moreover, almost all authors report difficulties they faced with the availability of data, the need for a longitudinal design and the representativeness of respondents. All of these issues show that we should not be too optimistic about the research evidence so far.

As the relation between boards and educational quality was at the core of this study, we first analyzed the data collected in the 16 studies on board level. Table 2 provides an overview of seven themes the articles deal with. It shows that researchers focused on formal aspects of the boards (Hofman, Hofman, and Guldemond 1999; James et al. 2011; Meier and O’Toole 2004; Ranson et al. 2005; Wong et al. 2007), such as their composition. Some also focused on softer aspects of governance, like behaviour (Delagardelle 2008; Hofman, Hofman, and Guldemond 1999; IASB 2001; James et al. 2011; Ranson et al. 2005) and the extent to which board members reflect on their own functioning (Johnson 2012; Plough 2011, 2014; Springfield 2008).

Table 2. Themes addressed in the studies.

Data on board level	Frequency
Structure and composition of the board	5
Governance style (self-perception)	5
Self-reflection of the board (self-perception)	4
Board relations with the school leader	4
Efforts to foster improvement	3
External orientation (innovation versus consolidation)	2
Focus on performances	1
Total*	24

* Some articles address more than one theme. In the 16 articles we identified 24 themes.

Structure and composition of the board

In most of the articles on formal aspects of school boards, the main variables are structure and composition (Hofman, Hofman, and Guldemond 1999; James et al. 2011; Meier and O’Toole 2004; Ranson et al. 2005; Wong et al. 2007). Some of these articles also examine soft skills (Hofman, Hofman, and Guldemond 1999; James et al. 2011; Ranson et al. 2005).

Ranson et al. (2005) look into the composition of the board (e.g. professional expertise and geographical background) and the working methods applied by the board. They conclude that the board’s working methods are indirectly related to pupils’ achievements on standardised tests. Hofman, Hofman, and Guldemond (1999) focus on the board’s structure, composition and culture. They conclude, using a multilevel approach, that involving people from outside the board in decision-making has a positive indirect effect on pupils’ math scores and well-being. James et al. (2011) conclude that pupils’ achievements are related to the board’s working methods, the pressure the board perceives and the extent to which boards face difficulties getting new members. Meier and O’Toole (2004) find indirect effects of the board composition on pupils’ achievements. These findings concern Latinos in the board and Latino pupils. Wong et al. (2007) focus on the role and effects of mayors appointing school board members. They conclude that the composition of the board alters when this happens, resulting in better achievements. These findings not only indicate the indirect nature of the relation between boards and student achievement. They also clearly show that countries and districts vary with regard to the composition of school boards.

Governance style

The softer aspects of governance, such as style and culture, are taken into consideration by Delagardelle (2008), Hofman, Hofman, and Guldemond (1999), James et al. (2011), IASB (2001) and Ranson et al. (2005). In the IASB (2001) (Lighthouse) study, the focus is on behavioural aspects of the board and school leaders in districts with unusually high- or low-performing pupils. The study shows that boards of districts with high-performing pupils have high expectations of their pupils and a clear vision of their role in improvement processes. Delagardelle (2008) can be seen as a follow-up study that builds on the IASB (2001) findings. Delagardelle also looks into the knowledge, competencies and behaviour of the board and concludes that boards should focus on facilitating an optimal learning context.

Self-reflection of the board

Table 2 shows that four studies report on board members reflecting on their own functioning (Johnson 2012; Plough 2011, 2014; Springfield 2008). Plough (2011), Plough (2014) presents two mixed-methods studies that report self-perception data about the behaviour and attitudes of school board members. In both studies, school districts that are doing well are compared to districts that are performing less well. Springfield (2008) and Johnson (2012) provide a case study in which personal experiences and reflections can be found on how board members can contribute to better-performing pupils.

Board relations with school leaders

A promising research avenue can be found in studies that focus on the dynamics and relation between boards and school leaders. In these studies, it turns out that the relationship between school boards and the superintendent is essential to educational quality (Andrews and Brewer 2014; Johnson 2012; Saatcioglu et al. 2011; Shelton 2010). Shelton (2010), for example, studied whether board members and superintendents differ in their values and attitudes to pupils’ achievements. Here, it is important to emphasise that the superintendent in the USA needs to function as CEO and instructional leader while also being a member of the school board. When compared with the position of school leaders, it becomes clear that the superintendent’s position in the USA differs from the position that school leaders have in other Western countries, such as the UK and the Netherlands. Moreover, the concept of leadership is preserved in the Netherlands for school leaders and not used so much for board members, as is the case in schools and school boards in the USA (e.g. Delagardelle 2008). Such differences need to be kept in mind when interpreting the findings.

Efforts to foster improvement

We found three studies about the efforts that boards make to foster improvement (Delagardelle 2008; McAdams 2000; Springfield 2008). Two present personal reflections of a board member (McAdams 2000; Springfield 2008), with a wide range of board characteristics, styles, models and behaviour. Because of the anecdotal nature of these studies, we cannot draw strong conclusions. This is relevant, as both studies give readers practical, “how-to-do” tools. Delagardelle’s study is very different, as it includes not only board members’ perceptions of their role and their responsibility to strengthen educational quality, but also their knowledge, attitudes and behaviour. This study compares school boards in districts that do well, with districts that do less well, to identify seven differences between boards that are related to bringing about positive change and improvement: (1) connections across the system; (2) knowing what it takes to change achievement; (3) workplace support; (4) professional development; (5) a balance between district wide direction and building-level autonomy; (6) a strong community connection; (7) and distributed leadership.

External orientation

Saatcioglu et al. (2011) and Andrews and Brewer (2014) discuss the board’s external relations. Andrews and Brewer (2014) address whether an innovative, outward-looking strategy or a defensive, inward-looking strategy leads to better performances. They link their findings to which of these strategies has the greatest influence on the relationship between social capital and performance. They conclude that schools have to operate in a complex world and that generally speaking the performance of a public or semi-public sector organisation is not a simple, one-dimensional concept. Andrews and Brewer's (2014) results also show the importance of contingency: the need to take local peculiarities into consideration and to synchronise for the benefit of strategic alignment. Saatcioglu et al. (2011), using a longitudinal design, measured the effects of the board’s social capital: whether internal (bonding) and external (bridging) contributes to the financial position and academic achievement scores at district level. They show that bonding has stronger effects on the schools’ finances and academic achievement than bridging, although they are not able to explain how social capital and academic achievement are related. One of the important findings of this study is that significant effects of the efforts made by the board can only be determined with a two-year delay.

Focus on performances

Van de Grift and Houtveen (2006) report on a longitudinal study analysing data of the Dutch Inspectorate of Education. The authors find differences between underperforming and outperforming schools and the extent to which school board meetings are geared to improving the quality of educational processes. The findings in this study seem to corroborate the findings of many studies into the effectiveness of schools taking process, organisational and situational variables into account.

Educational quality

When looking at the effects presented in the studies to analyse the relation between boards and educational quality, we have to conclude that there is little variety in the way educational quality is operationalised. In all but one of the studies, educational quality is operationalised as the results on exams or (standardised) test scores. Only one study also includes student satisfaction (well-being) as a dependent variable (Hofman, Hofman, and Guldemond 1999). We therefore have to conclude that there seems to be a strong focus on pupils’ achievements as the most accessible and measurable indicator of educational quality. Interestingly, Land (2002) and Johnson (2010) have already acknowledged that educational quality is being reduced to pupils’ achievements. Other authors warn that this might lead to a one-sided story about educational quality and a too-narrow focus on performance, given that education has multiple purposes (e.g. Biesta 2013; Keddie 2015; Moffett 2011). However, despite the fact that many authors problematise the idea of presenting pupils’ achievements as an indicator of educational quality, they argue that there is simply no better alternative.

The effects

One of the main findings of the present work is that there is little evidence to be found on the relation between boards and educational quality. Taking a closer look at the 16 studies in our final analysis, we must conclude that all studies agree that the effects boards may have are indirect. All studies confirm the idea that the school leader and the school leader’s functioning are key for a school, as is keeping the board aligned to all that takes place in the school. As such, school leaders and their relations with their boards can hardly be overestimated. The importance of these connections is affirmed by promising studies that focus on the relation between boards’ internal and external connections and pupils’ achievements. These studies show the relevance of better understanding of boards’ interactions (e.g. Andrews and Brewer 2014; Saatcioglu et al. 2011).

There are, however, some methodological concerns. Although authors point out that there may be a relation between certain behaviour of board members and educational quality, it remains almost impossible to consider this as solid evidence, as it is extremely difficult to isolate variables. Following Land (2002), we doubt whether it is even possible. In the same vein, one can question the possibility of causality. This is due to the complex nature of, and mutual dependence between, board members, school leaders, teachers and classroom practices (Saatcioglu et al. 2011). Several comparative studies can also be criticised for their case selection, where this was based on the dependent variable rather than the independent variable (Swanborn 2008). Case selection builds on student achievement or a district’s effectiveness (e.g. Delagardelle 2008; James et al. 2011; Plough 2011, 2014). Such a design does not provide valid causal explanations for student achievement or district (in)effectiveness.

It is important to note that there are clear differences between boards within countries and between countries. As a consequence, boards’ and schools’ specific contexts cannot be neglected when interpreting the findings. Because of these differences, one cannot easily draw general conclusions or transfer ideas of working methods from one country, or even district, to another.

It is also important to note that a large share of the surveys report board members’ self-perceptions. These surveys may suffer from self-reporting bias and, in some cases, from common source bias (Podsakoff et al. 2003). Moreover, we need to consider whether these studies present data about the actual behaviour of board members. In most studies, the chair of the board was invited to complete the questionnaire. This can affect the reliability of data, for example: about how boards behave, whether board members feel sufficiently confident to speak up, and how they make efforts to foster improvement (e.g. Delagardelle 2008; McAdams 2000; Plough 2014; Ranson et al. 2005; Springfield 2008).

Conclusions

This article draws on a systematic literature review to show that there is a lack of solid empirical evidence on the relation between boards and educational quality. This means that we know less than is reflected in policy assumptions about school boards. The ambitions for school boards and the expectations upon them are not evidence-based. The number of studies covering empirical data on the relation between boards and educational quality is small, especially when compared to the large number of studies on educational quality, school effectiveness and school improvement.

Looking more closely at the 16 remaining studies, we noticed differences in the reach, quality and applied methods, as well as country and regional differences. Altogether, these differences make it difficult to come up with more general findings on how boards can contribute to educational quality. Our study partly reconfirms Land’s (2002) conclusions that studies on school boards are often based on a combination of personal experience, observations and opinions, which means that this research frequently relies on anecdotal evidence. Nevertheless, we also found promising large-scale studies that offer a better understanding of the complex nature of the effects that boards may have on educational quality.

Our study clearly indicates that it is crucial to be aware of the differences that can be found between school boards. Their context may differ, as do the actual board models (one-tier or two-tier), board size, presence of parent members, whether boards are elected or appointed, number of schools that boards govern, school demographics, differences between primary and secondary schools, and position of the school leader. These factors are found both within and across countries. For practitioners this is of great importance: when interpreting findings, they have to take the context of the studied boards and schools into account and not copy practices blindly. When implementing some of the findings, they have to consider whether the findings are valid for their board and school: Is there a reason to assume these findings will also turn out to be effective in a given case?

Despite the fact that educational quality in policy and societal discussions seems to be defined in a broader sense, we have to conclude that this is not the case in research covering the relation between boards and educational quality. The narrow focus on test results is understandable from a methodological point of view. However, this does not reflect the full concept of educational quality. As such, we know little to nothing about the efforts that boards may make in other aspects of educational quality, such as citizenship education, pupils’ well-being and safety. Future research should also take these effects into consideration.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Funding

This work was supported by the Netherlands Initiative for Educational Research under Grant number 405-15-710-043.

Notes

1. We would like to thank Hanneke van den Akker for her contribution to this research project.