School district leadership styles and school improvement : evidence from selected school principals in the Eastern Cape Province

The purpose of this study was to investigate how leadership styles in the Eastern Cape school districts support school improvement. Mixed methods research was employed and data was collected through the use of questionnaires and semistructured interviews with school principals in various districts. The study was guided by the following questions: (1) what are the most common leadership styles among the school district officials in the province; and (2) how do the prevailing leadership style/s appear to support or hinder change and school improvement in the district? The quantitative data was analysed using Statistical Package for the Social Sciences (SPSS) for statistical analysis, while qualitative data analysis followed the iterative approach as suggested by Miles and Huberman (1994). Findings revealed prevalence of more authoritarian top-down leadership styles, which tend to have negative effects on school improvement. The paper ends with recommendations for more empirical work that would uncover district leadership approaches that influence the success of the districts and support school improvement.


Introduction
There is universal agreement on the significance of leadership for improved school performance and successful implementation of large-scale reform initiatives in education (DeVita, Colvin, Darling-Hammond & Haycock, 2007;Leithwood & Jantzi, 2000;Leithwood, Jantzi, Earl, Watson, Levin & Fullan, 2004a).School improvement and effectiveness is important in the context of global pressures that subject education system performances to public scrutiny, through well-known international tests and rankings.According to Spaull (2013) South Africa's performance in these international tests has been consistently poor, when compared to its emerging economy counterparts, while the local tests reveal gross inequalities within the education system.These educational outcomes have dire implications for the fledgling democracy and a struggling economy, exacerbated by a widening socio-economic gap.The role of education thus remains central to abating this situation, hence the country's investment in education and the concomitant role of state and local decision makers in educational resource allocation (Bantwini & Letseka, 2016).Bantwini and Letseka (2016) identify leadership as playing a crucial role in ensuring great returns in the investment.It is against this backdrop that we sought to investigate the role of education district i leadership as an intermediary between government and schools, and particularly the styles of leadership used for school improvement.The significance of district leadership in improving schools and student learning is central to driving educational reforms and achieving greater educational quality in the emerging economies, which makes this paper not only relevant to South Africa, but to other similar contexts.
The South African education management system is decentralised across four levels from national to province to districts to local schools.The Department of Basic Education's Policy on the Organisation, Roles, and Responsibilities of Education Districts (DBE, Republic of South Africa, 2013), clearly states that education districts play a key role in school success and in ensuring that all learners have access to high quality education.Specifically, the policy mandates district offices to: "work collaboratively with principals and educators in schools, with the vital assistance of circuit offices, to improve educational access and retention, give management and professional support, and help schools achieve excellence in learning and teaching" (DBE, Republic of South Africa, 2013:11).This policy thus, re-affirms the role of districts in the delivery of quality education and confirms their accountability for school improvement and broader educational reforms.Existing evidence on the role of district for school improvement shows the significance of school district leadership in driving educational reform initiatives (Bantwini, 2015;DeVita et al., 2007, Leithwood, 2010;Leithwood, Louis, Anderson & Wahlstrom, 2004b); with districts as a conduit between government and schools (Bantwini & Diko, 2011;Christie, Sullivan, Duku & Gallie, 2010); and districts as "institutional actors" in educational reform (Rorrer, Skrla & Scheurich, 2008).This body of literature suggests that effective district leadership is essential for the success of not only districts themselves, but for the whole education system, and particularly for improved student learning in schools.Furthermore, the literature highlights ways in which district conditions can influence student learning (Leithwood et al., 2004b).Some of these conditions include, for example; the district culture and collaboration (Bantwini, 2015;Rorrer et al., 2008), and the provision for professional development opportunities of teachers and provision for leadership succession (Leithwood, 2010).Aligned with school and district policies, district goals and programmes of professional development, and underpinned with a clear picture of the district priorities held by the schools, school districts can affect student learning (Leithwood et al., 2004b).
Several characteristics of effective and supportive districts have been identified in previous studies (Bottoms & Schmidt-Davis, 2010;Duke, 2010;Iatarola & Fruchte, 2004;Leithwood, 2010;Waters & Marzano, 2006).Collectively, these studies established that highly supportive districts firstly promote school leaders' confidence in their ability to succeed, and in their belief that improved school practices are important to their students' future (Bottoms & Schmidt-Davis, 2010).Secondly, they were found to hold district leaders and staff accountable for working collaboratively with principals, their school leadership teams, and staff, to implement a strategic plan and to hold principals accountable for creating excellent leadership teams (Bottoms & Schmidt-Davis, 2010;Leithwood, 2010).This was similar to Murphy and Hallinger's (1988) earlier work, in which they found instructionally effective school district providing a substantial amount of direction to the schools.Although there appeared to be tight control, there was also greater autonomy for schools, which enabled a degree of involvement in decisionmaking.Thirdly, these highly supportive districts were found to be sharing a common vision of high expectations for all groups of students and had a strategic planning framework that enables school leaders and staff to customise a set of strategic goals and actions for their school and providing schools with relevant data (Bottoms & Schmidt-Davis, 2010;Leithwood, 2010).Fourthly, highly performing districts were found to invest in recruitment and retention of talented personnel (Duke, 2010) ensuring continuity and preservation of organisational memory.
In view of these already established characteristics of effective school districts, we wanted to explore the role of school districts in school improvement, with a specific focus on district leadership within the selected province.Specifically, we were interested in examining the prevailing styles or forms of leadership within the district as perceived by school principals, and how these help provide support for school improvement.For purposes of this article, we use district leadership to refer to leadership provided by officials based in district and circuit offices, who deal directly with school principals.Our investigation was therefore guided by the following research questions: (1) what are the common leadership styles in the school districts?; and (2) how do the prevailing leadership style/s appear to support or hinder change and improvement in schools?We believe that by honing in on leadership styles of district officials, which have thus far received minimal research attention, the study holds the possibility of making significant contributions to extant knowledge on the role of district leadership in school improvement in general, and student learning in particular.The general lack of support to schools by districts in South Africa has been noted previously (Bantwini, 2012(Bantwini, , 2015;;Bantwini & Diko, 2011;Christie et al., 2010), but much less research evidence exists on the role of district leadership styles and their effect on school improvement.

Leadership Styles and School Improvement
The concept of leadership styles is one that is often contested in the literature, with different authors using either term (styles, forms, approaches, strategies, models) to categorise leadership practice or theory.In his seminal work theorising educational leadership, Tony Bush has highlighted the overlap in these terms, and sometimes the contrast, in using the same term to "denote different practices" (Bush, 2011:33).Eagly, Johannesen-Schmidt and Van Engen (2003:569) define leadership styles simply, as "patterns of behaviour displayed by leaders", while Leithwood et al.'s (2004b:6) definition is a little more substantial, denoting leadership styles as: "labels that primarily capture different stylistic or methodological approaches to accomplishing […] essential objectives critical to any organisations' effectiveness".According to Leithwood et al. (2004b), these objectives include directions setting and moving towards achievement of the set directions, within a particular organisation.A myriad of leadership styles (including; democratic, authoritarian, instructional, transformational, transactional) have been identified and associated with different leadership practices in the pursuit of organisational tasks and goals.The literature suggests that different leadership styles appear to have worked for different leaders in different situations, leading to assertions that there is no single best leadership style (Hersey & Blanchard, 1988), ii and that most successful leaders are likely to adopt most or all the different styles (Bush, 2007(Bush, , 2011)).Leithwood et al. (2004b) cautioned that most of these leadership styles are defined through mere adjectives that should attract scepticism, rather than acceptance.They argued that some of these adjectives obscure the real meaning of leadership practice that is essential in understanding successful leadership practice.In line with this argument, Johannsen (2014) suggests that leaders must always ask themselves what type of leadership style works best for them and their own organisation.He cautions against pigeonholing, arguing that there is not one appropriate answer to this, as leaders can draw from a wide repertoire of leadership styles depending on the situation with which they are faced.Although Hersey and Blanchard's (1988) situational leadership theory bares resemblance to this, Printy, Marks and Bowers' (2009) conception of the 'integrated leadership model', offers a closer representation of this mosaic of leadership styles.Thus, we loosely use leadership styles as a distinctive term to refer to the different approaches, strategies or forms of leadership exercised by leaders at district level, in order to argue that some leadership styles or some combinations of leadership styles are more prone to lead to school improvement than others.
Evolutionary origins of leadership styles trace back to the seminal work of Kurt Lewin and colleagues that identified three main styles of leadership: authoritarian, iii democratic and laissezfaire.This work showed that there was less cooperation in teams that were led in an autocratic way, when compared to those that were led in democratic and laissez-faire styles (Lewin, Lippit & White, 1939).This work and others (e.g.Gastil, 1994;Vroom & Yetton, 1973) comparing democratic and autocratic leadership styles have suggested that leaders who demonstrate democratic styles of leadership encourage subordinates to take part in decision-making, and those who are autocratic discourage subordinates from taking part in decision-making.Nonetheless, Gastil (1994) concluded that democratic leadership is more effective when it is 'emergent' rather than externally imposed (see also Woods, 2004).
Internationally, more recent work on leadership styles made comparisons between transactional and transformational leadership styles (Abu-Hussain, 2014;Eagly et al., 2003;Judge & Piccolo, 2004), and with distributed leadership approaches (Duke, 2010;Harris, 2004).Focus on transformational leadership styles suggested that leaders who display transformational behaviours motivate, inspire, mentor and empower followers with a shared vision and participative decisionmaking (Leithwood & Jantzi, 2000;McCann, 2011).In contrast, transactional leaders are those who display more traditional behaviours that set clear responsibilities for subordinates and reward them for satisfactory performance and success, correcting mistakes and failures.Although these approaches appear different, and are often dichotomised, Eagly et al. (2003:573) found both transformational and transactional leadership styles to be effective.These authors have, in fact, argued that transformational leadership and the contingent reward aspects of transactional leadership can "provide a particularly congenial context for women's enactment of competent leadership".While this work and others (Eagly & Johnson, 1990) support the centrality of gender to leadership behaviour with women displaying more transformational styles, other studies dispute gender differences (Kent, Blair, Rudd & Schuele, 2010).
However, as a leadership style, transformational leadership has also been found to work together with instructional leadership (Day et al., 2016;Leithwood & Sun, 2012;Printy et al., 2009), leading Day et al. (2016:251) to conclude that "both transformational and instructional strategies are necessary for success" in improved student outcomes.
Currently, distributed leadership is receiving attention in the literature, where it is strongly associated with democratic and participative leadership styles, and places emphasis on less heroic and more shared approach to leadership (Harris, 2013;Spillane, 2012).Parallels and dichotomies have been drawn between democratic, transformational, participative and distributed leadership (e.g.Harris, 2004;Woods, 2004), due to their collaborative nature, involving decision-making that is found to be more inclusive.It is this inclusive nature of these leadership approaches that also attracts connections to the more Africanoriented form of leadership of ubuntu (Bush, 2007;Msila, 2008).Msila (2008) posits that interdependence, interconnectedness and compassion are central to ubuntu, inspiring trust and collectiveness in decision-making.More local studies are needed in this area.It is observed that the literature on leadership styles is largely international and context specific, leaving a gap on effective leadership for local contexts.However, there seems to be overall consensus that effective leadership styles encourage more collaboration.In this sense, the choice of term is not just cosmetic, or semantic, but denotes a more substantial link to improvement and change.

Methodology
This paper draws from mixed methods research conducted with school principals in some selected districts in the Eastern Cape Province (EC).The EC Province is the second largest province in the country, and is known for being clouded by a myriad of education challenges that range from lack of infrastructure, material resources, teaching and learning resources to teacher shortages (Bantwini, 2010(Bantwini, , 2012)).
The study adopted a sequential mixed methods design (Creswell, 2014), where quantitative data was collected through the use of questionnaires, and analysed, before qualitative interview data was then collected.The questionnaires were administered to five of the 23 districts in the Eastern Cape that could be conveniently reached by the researcher.Each district was given 20 questionnaires (100 in total) to distribute amongst the school principals and 19 completed questionnaires (20%) were collected.The limitation of this poor response rate and the possible bias of the convenience sampling are engaged later in the discussion.All the questionnaire respondents were Africans, of which 10 were males, and nine females, from both primary and secondary schools.
The questionnaires were sequentially suplemented with qualitative semi-structured interviews that were conducted with 18 school principals, drawn from the five districts where we administered the questionnaires.The participants were selected school principals, who were willing to participate in the study, and whose school locations could be reached within the time of data collection.Careful consideration was taken to purposively include principals from both primary and secondary schools to match the profile of questionnaire respondents.Nine principals were selected from primary and nine from secondary schools and eight were females and 10 males.
The interviews lasted between 45-60 minutes and all the interviews were audio recorded with the participants' permission and later transcribed for analysis.The qualitative data coding and analysis followed an iterative process, as suggested by Miles and Huberman (1994), which include reading and affixing codes to the transcript notes while noting reflections or other remarks in the margins; sorting and sifting through the materials to identify similar phrases, relationships between variables, patterns and themes, while the quantitative data was imported into SPSS, and analysed using the statistical frequency distribution.The study followed strict ethical conduct, based on permitted access and consent to participation, as well as ensured protection of participants and secured data.

Findings
To establish leadership styles of district officials, and how they support school improvement, school principals in the selected Eastern Cape districts responded to questionnaires and interview questions.This section presents both questionnaire and interview findings using themes developed from research questions as subheadings.

District Leadership Styles
The first set of questions on the administered questionnaire asked participants to select a true or false response to the statements that were establishing perceptions of school principals on district leadership styles.In particular, the first items measured how principals perceived their levels of involvement by their circuit manager, when it comes to making decisions concerning the principals' own schools.These statements would determine the leadership styles of the principals' circuit managers, as illustrated in Table 1 below.It must be noted that the leadership styles column has only been added to the table as part of the analysis.The leadership styles were not part of the administered questionnaire.Questionnaire items were developed from the literature and both tables 1 and 2 were constructed by the authors.The circuit manager closely monitors principals to ensure they are performing well.
Authoritarian 100 0 10 The circuit manager entrusts tasks to other team leaders within the circuit.

Democratic 100 0
Items 1, 4, 7 and 9 denote an authoritarian leadership style, while Item 2, 5, 8 and 10 denote a democratic leadership style, and Items 3 and 6 denote a laissez-faire approach.Our analysis of the questionnaire results suggested that circuit managers use a variety of leadership styles, with high level of involvement and participation suggested by 100 percent true to items 8 and 10.Although there could be a possible contradiction between statements 3 and 4, we note that both of these statements could be interpreted to suggest some level of involvement, through either instruction, or own initiative.Thus, from the quantitative findings, there was evidently an equal distribution between authoritarian and democratic leadership styles depicted.
However, we were more surprised when we asked the school principals, through interviews, to describe their levels of involvement and participation in making decisions that concern their schools.Although there was a mixture of responses showing varying levels of involvement of the principals by the circuit and district, the findings leaned towards mostly no involvement in decisionmaking.Only one participant thought the district's involvement was "participatory in nature", while the majority thought the district approach was the opposite, using phrases such as "autocratic with a top-down approach".The majority of the participants expressed their frustration with the district and circuit offices, citing how they would not involve them in decision-making, even in matters concerning their schools.Below is an extract from a male principal, who was not involved in decisionmaking.The question required them to describe their levels of involvement in decision-making within the district: Well, I would say it is somehow autocratic in a way because as principals we are not involved in decision-making.We are called by the district only when they want to communicate their decisions.So we are not fully involved in the decisions on how things must be run.

And another principal responded to the same question in the following manner:
As principals, most of the time we are just instructed to do as we are told.And they use a policy that says 'you need to comply and complain later'.

Another confirmed:
In terms of decision-making, there are decisions that directly affect our schools that are taken at district level, without our involvement as school principals.That one is visible when you understand that principals' meetings are called once a year, at the beginning of the year, and we are given the marching orders to say this is what is to be done.
The overwhelming majority of the responses suggested lack of or no involvement at all in decision-making by the district.Although the questionnaire responses suggested a combination of styles of leadership, we found the interview resmore telling of the circuit managers' leadership styles.The majority of participants used a "top-down approach" to describe their lack of involvement in decision-making.

District Leadership for Change and School Improvement
In this section, the focus is on establishing the extent to which and ways in which district leadership styles promote or hinder change and improvement.Table 2 below summarises the questionnaire responses to the items that established ways in which district leadership promoted school improvement.The majority of responses (80 percent) were positive (agree and strongly agree), indicating that principals are largely of the opinion that the districts lead in ways that enhance school improvement.However, comparison of some items to the qualitative data reveal some inconsistencies: for example, 89 percent of the respondents were of the opinion that the district provides autonomy to principals to lead their schools.Yet, the qualitative data suggested that the majority of principals do not have autonomy.The majority of participants interviewed did not feel that they had the autonomy to initiate change within their schools as the following extract reveals: And the district curriculum plan does not devolve to schools.It does not give us that autonomy to have our own curriculum plan to be infused into the district curriculum plan, so that we can have a cohesive type of an arrangement, where we say the school curriculum plans and school improvement plans actually talk to the district improvement plan, which will then culminate into [sic] the provincial improvement plan.Hence, you will see these discrepancies that schools are having [sic] their own way of doing things, and districts are having [sic] their own way of doing things.
This extract shows that schools and districts do not work together when it comes to planning.The extract confirms the presence of district curriculum plans, which would also suggest that there are plans and goals set for the learning programme of the schools, by district.The presence of the set goals can be further confirmed by 84 percent of the participants, who agreed that the district has clear strategic goals for instructional programmes.However, the interview findings reveal no collaboration between district and schools in developing these plans, which could potentially end with school improvement plans that are not informed by district development plans.This lack of collaboration was found to be unsupportive of school improvement.
The quantitative data above suggest 94 percent of the were of the opinion that the district takes responsibility for staff development.In the interviews, only a few school principals agreed that staff development happens, as shown below: We will be invited at times to workshops where we will be work-shopped.And we are not only invited as principals […] but teachers as well.We have a very good circuit manager.
This indicates that there are different practices between districts.The majority of principals, however, referred to the lack of professional development for school principals, one participant stated: At least the new principals are lucky, because after they have been appointed, they are inducted by the district office.But in our case, we get that information through circuit management meetings.

Another one retorted:
Schools lack people who have leadership of high quality.Principals are not supported.It is rare for us to be called for a workshop.
Thus, there are different practices between districts and between circuits within districts, and although it was the majority of participants who indicated lack of professional development, there is evidence of principals' professional development in some circuits.Professional development for school principals has implications for capacity building for leading instruction (Rorrer et al., 2008), and for leadership succession planning (Leithwood, 2010), and collective system-wide development (Naicker & Mestry, 2015).The findings in this section suggest that while there is evidence of strategic planning through the presence of district curriculum plans, there is a lack of autonomy, lack of involvement in decision-making by school principals, and a lack of professional development for school principals.It is thus our opinion that the prevailing district leadership styles appear not to be supportive of school improvement.

Discussion and Conclusion
In addressing the research questions directly, we note that firstly, the quantitative findings suggest a stronger combination of district leadership styles, which are both autocratic and democratic, with an element of a laissez-faire attitude, while the qualitative interview responses suggest prevalence of a more autocratic approach to leadership.The latter is seen through the lack of principals' involvement in district-wide decision-making and lack of collaboration that does not create opportunities for engagement from the principals.We find this incongruent with effective district leadership behaviour literature (Duke, 2010;Leithwood et al., 2004b;Waters & Marzano, 2006) that promotes a high level collaboration between district officials and principals for school improvement.We are therefore more inclined to conclude that the prevailing leadership styles in these districts are top-down autocratic approaches with traces of consultation and participation, as seen in 40 percent of the questionnaires.In contrast to the qualitative data, there was little evidence of participative or democratic styles of leadership from the qualitative data.Instead, it was found that there is an absence of collaborative district-wide decision makingwhere school principals lament their lack of involvement in decision-making, even when it comes to issues that concern their schools; and district-wide collaborative planning, where school principals felt that districts and schools 'operate in silos', and that there is a lack of autonomy at the level of the school.While participants are not crying out for full autonomy of their schools, they do decry certain levels of what they regard as undermining their own school plans and processes.The lack of autonomy may suggest lack of confidence shown in them by their school district, but it could arguably be explained in part by the bureaucratic nature of the education context, which makes schools accountable to government (Bush, 2016).According to Bush (2016) autonomy is linked to accountability, and hierarchical authoritarian leadership styles are a feature of bureaucratic models (Bush, 2011).Naicker and Mestry (2013) also found autocratic and hierarchical leadership styles of school principals prevailing and influenced by the bureaucracy in which schools operate.Given this, it is perhaps not surprising that school principals do not have full autonomy to run their schools.However, the non-involvement in decision-making suggests a low level of confidence bestowed on school principals by the district, and does not encourage principals to take ownership and pride of the new changes or plans to be implemented by their schools.This, we argue, is not helpful for school principals to achieve school improvement.While we are not arguing that autocratic leadership is a bad leadership style in its own right, we argue that autocratic district leadership is problematic in the context of this study, where it did not encourage school principals' collaboration in decisionmaking.Further, we contend that leadership approaches that discourage participation do not only undermine the DBE policy that requires districts and schools to work together in collaboration, but also undermines the democratic principles that underpin the South African Constitution.These approaches are therefore neither progressive and nor effective, particularly in an emerging economy that ought to diversify its decision-making capacity.
As a way forward, we suggest that more collaborative leadership styles that encourage participation in decision-making from school principals, may be more helpful for school improvement.This practice may set an example for principals to extend the culture of collaboration to their teachers with ease.Leithwood (2010) suggested that a district-wide focus on student achievement is one of the key characteristics of effective districts.Rorrer et al.'s (2008:323) characterisation of districts as actors foregrounds collaboration, where districts work as a collective with schools to achieve the set agenda.Strong collaboration that drives coherent reforms and learning programmes would require development and empowering of principals as local leaders so that they become influential decision makers.We argue that a certain degree of autonomy, with support and monitoring, is needed, to enable meaningful involvement in influential decision-making.Leithwood et al. (2004b:12) argues that empowering others to make significant decisions enables "greater voice to community stakeholders" and that successful district leadership practices in emerging economies rely on "capacities and motivations" of these local leaders, all of which are essential for driving change, school improvement, and broader educational reform initiatives.
By way of reflection, we acknowledge that existing literature on effective district leadership is mostly from western, developed contexts and is based on large-scale data sets.While useful in providing a benchmarking framework, this literature is applied with scepticism.The current study is neither large scale nor focused enough to enable us to make assertive conclusions about specific cases on how certain leadership styles promote or hinder school improvement.Although we are confident in our assertion that the prevailing leadership styles were autocratic, ineffective and unhelpful for the twenty-first century district leadership, we also acknowledge that the study did not focus on the behaviours of district officials directly, but relied on the school principals' perceptions and interpretations of them through the latter's own experiences.Nonetheless, the study is a useful contribution that has opened up an important research avenue, with possibilities of district officials' leadership influencing school improvement in a developing context.We therefore invite more research of districts as units of analyses so as to examine leadership practices that work from perspectives of both leaders and followers.We believe this holds better research prospects for understanding leadership behaviour that can be correlated so as to enhance school improvement and successful large-scale educational reform initiatives.
As a concluding remark, while the study used a combination of methods, we acknowledge the methodological challenges (of neither reaching too deep nor too wide), and contextual challenges (of administering questionnaire in rural contexts), as limitations.Due to the small convenient sample size that may not be representative, we are conscious that our results are not generalisable to the Eastern Cape Province.However, the findings flagged some serious and important issues that need to be pursued through further research.Our use of sequentially combined methods (Creswell, 2014), was an attempt to mitigate the negative effect of the low response rate and minimise bias.However, this may have raised some further complications in what could be perceived as contradicting findings between questionnaire and interview findings.We do not necessarily view the 'diverging' results as a problem, but attribute the limitation to sampling and low response rate errors.Harris and Brown (2010:9) note that lack of 'confirmatory data' happens due to the length of interviews that gives participants "more time to expose the variabilities and inconsistencies within human thinking"the time that participants do not have when filling in a questionnaire.Thus, although significantly more research ought to be undertaken, the study aim was limited here to producing an overview of district leadership styles and the extent to which they promote school improvement, where more questions were exposed in doing so.However, we want to caution future researchers to take these research complexities (Day, Summons & Gu, 2008:341) into consideration, for more appropriate and suitable methodologies, that will result in "richer, synergistic understandings" of leadership practices.

Table 1
Leadership style

Table 2
District leadership and school improvement