9 District support for curriculum management change in schools

Anne Mc Lennan, Mike Muller, Mark Orkin and Hanlie Robertson

Author and publication details

Anne Mc Lennan, Wits School of Governance,; Mark Orkin, Development Pathways to Health Research Unit, University of the Witwatersrand,; Hanlie Robertson, Research in Education and Labour Centre (REAL), University of the Witwatersrand,

Published in: Christie, P. & Monyokolo, M. (Eds). (2018). Learning about sustainable change in education in South Africa: the Jika iMfundo campaign 2015-2017. Saide: Johannesburg.

Download PDF

Through the conversation, the Circuit Manager is able to understand our challenges. She also directed us to areas where we could improve. This was shared with the School Management Team and educators. The Circuit Manager will be looking at this at the next follow up meeting (Comment by Principal, Programme to Improve Learning Outcomes [PILO], 2016 internal review of secondary school Principals, 2017).

This comment reflects the Theory of Change that drives the Jika iMfundo Campaign, as well as the role of districts in improving learning outcomes through scaled system-wide interventions. The working Theory of Change for Jika iMfundo is that “constructive professional conversations”, based on evidence and leading to action, shift curriculum coverage routines and patterns of support within schools and between schools and districts, to improve learning outcomes. The programme of interventions is systemic, working with all schools, circuits and district officials in the two districts of King Cetshwayo (KC) and Pinetown (Pt) in KwaZulu-Natal.

Jika iMfundo district support is an integral part of the overall change strategy in the two districts. The purpose of the support is to build the capacity of circuit specialists (CES/CMCs) and managers (CMs) to provide institutional curriculum management support. The goal is to develop new, more professional (cooperative rather than bureaucratic compliance) ways (behaviours) of supporting schools. The PILO team used various strategies, including capacity development workshops and coaching, to build the confidence of CMCs (circuit clusters/centres) to take leadership of curriculum coverage practices in circuits and build CM accountability to schools and vice versa (see Figure 9.1). Circuit managers (CMs), the focus of this study, developed tools to guide school visits and were assisted to monitor, supervise and support Principals in solving problems related to the management of curriculum coverage. Each intervention – within the school, as part of district support and in the province – was supported and reinforced upwards and downwards to change practices and establish new curriculum coverage routines.

Jika iMfundo district support

Figure 9.1 Jika iMfundo district support

The significance of the Principal’s statement, quoted above, is that interventions at district level (and in schools) may only be seen in relationships and routine practices several months later. While there is evidence of changes in routines influencing curriculum coverage in schools, it is difficult to correlate these with shifts in district support. However, research suggests that districts, particularly empowered Circuit Managers and Subject Advisers, are important levers for change in complex, unequal and dynamic contexts (Leithwood & Azah, 2017; Moorosi & Bantwini, 2016; Fullan, 2015). However, the ability of districts to monitor, report and respond (MRR) proactively and in support of schools, can be limited by institutional structure, as well as bureaucratic, compliance-driven working cultures. CMs and Subject Advisers (SAs) with different line loyalties tend to work in silos in supporting schools. In addition, there is the ever-present tension between ensuring compliance to curriculum policy and providing support.

Many researchers and policy decision makers view districts as critical supports for learning improvement (see, for example, Fullan, 2015; Corcoran, Fuhrman, & Belcher, 2011; Gustafsson & Taylor, 2016). Similarly, the South African Policy on the Organisation, Roles and Responsibilities of the Education Districts (300 of 2013), which provides guidelines on the size and staffing of districts, as well their administrative and professional support roles, assumes that district planning, support, oversight and public engagement will contribute to learning improvement in schools. However, there is a gap between these policy intentions and realities on the ground, as many districts lack the resources and capability to provide professional curriculum management support (JET, 2014; Prew, 2012; Moorosi & Bantwini, 2016). To examine how and in what ways districts and CMs may enable and support curriculum management change in unequal and resource-scarce conditions, this chapter seeks to explore the effects of the Jika iMfundo district change programme on the curriculum management relationships between schools and circuits. While it is too early to establish strong correlations, it is possible to explore whether assumptions about the role of districts, as a change lever, are plausible with a focus on the ways in which districts and schools connect to improve learning through professional accountability.

Can districts be developmental?

South African policy on education districts accords with a global conviction about the developmental role of districts in helping school leadership to deal with curriculum coverage and management challenges. Districts, as intermediaries between National and Provincial Departments on the one hand and schools on the other, are required to implement curriculum policy by monitoring progress, as well as to provide support to schools struggling with curriculum management. The underlying rationale is that Principals, as instructional leaders, build a conducive teaching and learning environment, and districts set goals, monitor and support where necessary. Fullan (2015) refers to this as “leadership from the middle” – the top being the state and the bottom the school – depending on the country context. What is common is the idea that districts can and should drive and support curriculum management and coverage in schools as part of “a deliberate strategy that increases the capacity and internal coherence of the middle … in pursuit of greater system performance” (Fullan, 2015, p. 24). Elmore (2016) is more sceptical about a system approach that is flexible enough to accommodate different school contexts.

International best practice suggests that districts can be catalysts and supports for development and learning improvement (Bates, 2013; Glewwe & Muralidharan, 2015; Pritchett, 2015). However, much of this literature works from the experiences of the USA and other developed countries (see Moorosi & Bantwini, 2016; Anderson, 2003; Rorrer, Skrla, & Scheurich, 2008). A key difference between South African and USA district systems is that, in South Africa, district leaders are not elected. District officials are educators promoted upwards in the system and are subject to a line accountability that is different from democratic review. Historically, districts and circuits, much like provinces, were either conduits of national policy, or the police of teaching and learning (Mc Lennan, 2003). Despite a strong policy commitment to decentralisation and school autonomy post-1994, districts have been consistently under-resourced and focused on compliance (Chinsamy, 2002). The policy on districts (DBE, 2013a) reinforces the role of districts in the delivery of quality education, without due consideration of district officials’ capability and reality (Bantwini & Diko, 2011; Moorosi & Bantwini, 2016). The kind of professional support envisaged is uncommon in practice. District offices are mandated in Section 20 of the Act 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.

Even in stable and better-capacitated conditions, a tendency for district offices to limit change due to over-bureaucratisation of work processes, internal politics and weak capacity has been noted (Narsee, 2006; Ouchi, Cooper, & Segal, 2003). Globally, traditional line authority has shifted to a focus on performance or compliance reporting to ensure accountability and control outcomes. More and more measures are generated and evidence gathered to show where the improvement is or is not occurring. Districts have become the locus where much of this information is demanded from schools for reporting purposes up the system. However, an excessive focus on compliance can redirect the core work of districts and schools away from support by structuring the organisation of day-to-day routines towards compliance. This means that the greater the emphasis on compliance, the less on professional responsibility (Flinders, 2011; Popkewitz, 1996). Professionals, bogged down in meeting stipulated and monitored requirements, revert to compliance rather than professional judgement. In the process, professional agency is lost as teachers, Principals and officials work to rule.

In the South African context, the general lack of support to schools by districts, in the face of perpetual accountability demands, has been emphasised by researchers (see Bantwini, 2012, 2015; Bantwini & Diko, 2011; Christie, 2010; Christie, Sullivan, Duku, & Gallie, 2010). Nevertheless, the district provides an important social context for school change which influences what Principals know and how they use their knowledge (Mangin, 2007). Evidence confirms that district leadership matters when it comes to driving curriculum reforms, as well as improving schools and student learning (Anderson, 2003; Elmore, 2000; Bottoms & Schmidt-Davis, 2010; Fullan, 2016). This is also true for achieving greater educational quality in the context of emerging economies (Moorosi & Bantwini, 2016). The insight generated internationally that “… many Principals cannot be successful without the best possible district leadership” (Barber, Whelan, & Clark, 2010, p. 3) signals the importance of districts as enablers of school performance (Anderson, 2003; Leithwood, 2010).

Academic research on district support, school change and improving learning outcomes continues to argue, in the main, that strengthening district support capability, alongside school autonomy, has been and still is an effective lever to achieve positive education outcomes through oversight processes which track progress and ensure accountability. For example, Pritchett (2015) suggests that education systems that focus on learning can only be achieved if “greater control [is given] to local officials, parents and teachers.” Changing school routines to focus on curriculum coverage for learning arguably requires different kinds of support skills and abilities. These should be located within districts but also across the education system, linking into schools and districts (Pont, Nusche, & Hopkins, 2008). If districts are the drivers of reform, the capacity and level of influence of the district should be indirectly proportional to the number of effective schools. Anderson (2003, p. 5) concludes that “there is a notable convergence in findings around common strategic principles and policy-linked actions correlated with success.” He also points to efforts to synthesise case research on district roles and effectiveness in the current standards and accountability-driven reform context (see, for example, Marsh, 2002; Hightower, Knapp, Marsh, & McLaughlin, 2002).

Some evidence has been produced in support of a Theory of Change that links effective districts to improved educational performance, suggesting the causal relationship that has been lacking thus far (see for example Gustafsson & Taylor, 2016; Kanjee & Bhola, 2014; Leithwood & Azah, 2017). District leaders, pursuing a well-articulated long-term strategy for improvement, have been found to link to student performance in ethnically and economically diverse settings (Leithwood & Azah, 2017). Numerous studies (see, for example, Anderson, 2003; Murphy & Hallinger, 1988) cite the potential for the district office to support learning if correctly configured. Spillane (2004) points to the powerful influence exerted by school district officials on the degree of coherence (or confusion) in instructional practice and guidance provided to teachers. Effective districts provide direction and support to the schools, including the provision of job-enabled professional learning, facilitating professional learning opportunities, building of a professional community, addressing problems with administration and implementing standard operating procedures (Murphy & Hallinger, 1988; Anderson, 2003; Clarke & Wildy, 2011).

Various studies identify the characteristics of instructionally effective and highly supportive districts. For example, such districts build school leaders’ confidence in their ability to succeed and support the belief that improved curriculum practices are important for their students’ futures (Bottoms & Schmidt-Davis, 2010). The focus is on student achievement across the district (Leithwood, 2010) through an emphasis on instructional support and coherence – curriculum management, as well as trust-based professional relationships (Leithwood & Azah, 2017). This creates a shared commitment to build SMTs to achieve high standards for learning, teaching and leadership (Clarke & Wildy, 2011). Monitoring data are used to enable schools to improve, as well as identify those that need support (Bottoms & Schmidt-Davis, 2010; Leithwood, 2010). Schools that improve against the odds in difficult circumstance are acknowledged and used for learning across the district. They promote a spirit of collaboration between district officials and Principals for school improvement (Duke, 2010; Leithwood, 2010; Waters & Marzano, 2006).

Districts that are developmental and effective in supporting learning improvement tend to have the professional and operational capability to build supportive relationships with Principals (Khosa, 2013). In South Africa, this capability is varied and stretched. Many district institutions have been shaped by the history and context of the province and the locality. Most district teams include corporate functions, specialised education services, governance and management development, and, most importantly, subject and curriculum support (Mthembu, 2014). Districts are responsible for implementing all aspects of education operations, including curriculum, finances and resourcing, but have limited scope beyond planning and focusing goals to district contexts (Khosa, 2013). Although district policy is an explicit attempt to respond to inequality and uneven capacity by providing norms on roles, authority and resourcing, many districts lack the resources and capacity to meet these basic requirements for school support. Inequality between districts and provinces can be entrenched by the uneven distribution of resources and professional capacity to support learning. For example, ratios for district officials to support schools can range from 1 to 30 to 1 to 100 (DBE, 2013).

Districts are caught between their responsibility to ensure policy implementation within a bureaucratic structure that requires them to work and report upwards to the PED, and to provide critical support to schools challenged by poverty and poor teaching and learning practices (Mthembu, 2014). The frustration inherent in this role has shaped how district staff view their ability to contribute. Historically, district officials have a negative association as an inspectorate with a biased process of performance management. This association undermines relations of trust between district officials, Principals and teachers. Focus group research on levers for change (WSG & BRIDGE, 2016) showed different relationships between schools and districts in better resourced areas. Schools in rural areas feel that districts do not provide adequate support. For example, parents from the schools in Limpopo believe that district Circuit Managers do not provide help with the rebuilding of dilapidated schools – “parents and SGBs never get clear answers regarding when services will be delivered.” Principals and teachers note that the district only came once a year to check how far they were with the curriculum because they do not have the means or resources to come to the rural areas. Learning and social mobility also reflect these divisions (Moses, Van der Berg, & Rich, 2017).

Jika iMfundo district capacity building and change management intervention

The district support programme was designed by PILO in response to on-the-ground realities to change the working practices of CMs to support schools through better planning, regular school visits, conversations based on evidence, improved reporting, sharing and reflection. We present this case to relate some of the practical learnings to the insights generated from the international literature. The research makes the conceptual assumption that institutionalised daily routines and interactions shape relationships. System interventions that introduce new routines must be embedded to shift practices and create new support dynamics. Evidence of developing relationships of trust and support between schools and districts will provide a new perspective to the work linking effective districts and improved school performance.

The qualitative data available for the case study were collected by PILO in the form of interviews and routine reviews of interventions across schools and districts, over a period from 2015 to 2017. An analysis of the comments on captured reviews, as well as transcribed interviews, provides insight into the perceptions of Principals and CMs about the programme. This allows an examination of some of the challenges that emerge as the intervention proceeds, a gap that has been identified in the “salient features” literature (cf. Corcoran, Fuhrman, & Belcher, 2011). In addition, during 2017, one of the authors attended a series of CMC meetings in the two districts, as well as workshops with school and district coaches. These, too, provide insight into the challenges.

Among the varieties of quantitative data gathered by PILO were the school-level assessments made by CMs, reflecting their visits in the first and third terms of 2016, to schools within the circuit assigned to them (see Table 9.1). The CMs’ assessments cover curriculum management and general management at the schools under four and seven headings respectively, using the CM Tool which is intended to provide the evidence for constructive conversations about curriculum coverage and management. It contains several questions per heading and is 14 pages long. It is signed off by the school after completion. The statistical analysis focuses on General Management (section B) with scores from the CM Tool aggregated per school under seven headings. The quantitative analysis concentrates, for simplicity, on four of these headings: SMT and Planning (Strategy), Relationships, Finance, plus the total of all seven headings.

Planning & tracking Supervision & teamwork Assessment
Monitoring by subject & phase Supervising and supporting HoDs Assessment plans and programme
Planning tools for HoDs & teachers Staff development plans Internal moderation
SMT & planning Relationships Staffing Attendance Finances LTSM Infra, safety & security
Meetings & agendas SGB activities PPN Learner attendance Expenditure management Records & challenges Maintenance & repairs
Performance & targets Parents Forecast Teacher attendance Records Safety and security

Circuit Manager Tool summary

Table 9.1 Circuit Manager Tool summary

The data from PILO had to be extensively cleaned to combine school data and CM Toolkit data. The school names had been entered manually in each set of data and contained many variations in the way names were entered, as well as typing errors.1 A number of variables that could potentially have been valuable for the quantitative analysis proved to be unusable owing to large proportions of missing information or a lack of variation in the responses. Apart from the first-time inexperience of the CM assessors, the latter problem suggested that the possible responses supplied for the particular items could have been better formulated (Lietz, 2008). Recommendations arising from this examination of the data, as the project is enlarged, are that common numerical codes for schools be used in all data sets, that the questionnaires be piloted to ensure that a sufficient range of responses is obtained, that checks are made for missing information at an early stage and attempts are made to obtain more complete data.

Two serious limitations have to be noted with the data available at this early juncture. First, scoring on items in the CM Tool in the first round of visits tended to be uniformly high. There is more plausible differentiation in the scores from the second round, as CMs gained experience and could be more discerning. Secondly, there was a serious drop-off in schools that were recorded as having been visited for the second round: 36% of the schools had second-round scores (33% for Pt and 39% for KC). Further research would establish possible tendencies in the nature of schools under-visited. For both these reasons, the analyses of change in the schools offered below are made by way of illustration, of what will be possible subsequently. But, they are worth considering on their merits, nevertheless, for their possible implications, given the rapid pace at which the project is being expanded. The data were analysed in terms of repeat visits to schools by CMs, by quintile, for each district and then by CMC to find possible correlations or learnings. Numerous other correlations may be sensibly drawn to discern any salient insights and assess the reliability of various measures for possible future use. In addition, the statistical tests on the CM Tool scores indicate which of the illustrative findings from the current data are statistically significant. These flag potential concerns for future management and prioritisation.

The two districts together have 1170 schools of which 65% are primary. The schools fall into 37 circuits, averaging 26 schools per circuit. King Cetshwayo (KC) has 659 schools in 22 circuits arranged under five CMCs. Pinetown (Pt) has 502 schools in 17 circuits arranged under three CMCs. Physical and staffing resources are limited in many schools and districts. A key resource challenge for KC is staff and travel distance. For example, there are five Subject Advisers for 470 schools in foundation phase, one for 696 schools for Grade 8 and 9 maths and two for 226 FET schools in English FAL. Pt is not much better off with four Subject Advisers for 383 schools, three for 556 and two for 173 schools. CMs are more numerous, but many of the posts are vacant. This leads to the CES acting as a CM, or Principals being seconded into acting CM positions (CMC meeting notes, July 2017).

Table 9.2 indicates that most schools in KC are more than 25 or up to 50 km away from the district office. In contrast, in Pt, most schools fall within a 25 km or less radius. In addition, while schools in KC have water, electricity and security, only 15% have access to any form of sanitation or waste removal (DBE, 2015). Similarly, 69% of district households have no refuse removal and 53% have inadequate sanitation. Pt does not have the same sanitation challenges, but still, 25% of schools do not have any toilet system at all. In addition, less than 25% of district households, on average, lack basic services (DBE, 2015).

Distance from DO Number % Number %
Less than 5 kms 19 3% 35 7%
Between 5 and 10 kms 16 2% 74 14%
Between 10 and 25 kms 160 24% 342 64%
Between 25 and 50 kms 211 31% 85 16%
Greater than 50 kms 266 40% 0 0%
Average distance to DO 4.8 kms 1.7 kms

Distance of schools from the district office

Table 9.2 Distance of schools from the district office

DBE district profiles, 2015

In the two pilot districts, a district coaching team worked closely with the District Director and CMCs to build commitment to change (agency), as well as capacity to do the same work in a different way (using evidence-driven conversations as a basis for action, as well as reporting, sharing and reflecting). PILO attempted to capacitate officials to work in multifunctional teams across silos, to use school data to prioritise and solve problems, and to direct appropriate support and resources to schools. This involved change management, to support the district to absorb and make changes, as well as capacity-building and practices to drive and guide the district to manage the system, through school monitoring and support, in a way that effectively drives improvements in curriculum coverage (and thus learning outcomes).

The process includes:

District Charters

Figure 9.2 District Charters

The CM Tool is a primary resource which was co-developed with CMCs to provide the evidence to support the professional conversations (see Table 9.1). The intention is to reorient schools towards professional accountability by using evidence as the basis of the conversation for school visits and build instructional leadership skills:

Our Circuit Managers have developed the Circuit Managers’ Tool which they call ‘district diagnostic tool’ because, at first, our Principals did not understand that curriculum management is part of their work; they saw themselves as managers for general management not curriculum management (District coach, 24 September 2016).

In addition, the Tool enables CMs to classify schools in the form of a “heat map”, providing a quick way to identify schools most in need of support. The scores of each school are recorded on a matrix and these are colour-coded red, amber and green using defined cut-points for each performance area. For example, the variable Strategy is scored out of sixteen; and scores of thirteen and over are coloured green, scores from eleven to seven are coloured amber and scores of six or lower are coloured red. There have been two rounds of scoring, earlier and later in 2016.

Figure 9.3 demonstrates where and how the PILO team supported the department to improve within the existing system and context. The Theory of Change assumes that agency, a clear focus on curriculum coverage management and the responsibility to drive change is activated via PILO support and leadership workshops and appropriate tools and training. Training is supported by on-site coaching. Sharing is encouraged to show the possibility of reaching goals in real resource constrained working contexts.

PILO change intervention

Figure 9.3 PILO change intervention

At every stage of implementation, a conscious attempt is made to locate responsibility and accountability where it will remain. Circuit Managers are empowered to monitor, report and respond (MRR) to curriculum coverage challenges in schools by working with Principals and SMTs. The goal of using data for monitoring and reporting is to enable the district to support schools on a differentiated basis, responding appropriately from where schools are to identify achievable goals. The districts should be able to segment schools in terms of the strength of their ability to improve curriculum coverage. Steps 3 to 9 in Figure 9.3 provide a schematic of the district development process.

Can the districts leverage curriculum management change?

The qualitative and quantitative analysis of the intervention explores whether Jika iMfundo has influenced how districts and CMs might enable and support curriculum management change in unequal and resource-scarce conditions by building professional, supportive relationships between schools and circuits. The case presented provides a description of the intended goals of the programme, as well as the implementation and rollout process. This section provides an analysis of these interventions, using the key characteristics of supportive districts as a guide. The primary support offered by districts for curriculum coverage includes professional engagement through productive and collaborative working relationships; coherent curriculum management support enabling Principals to share challenges; and the collection, review and analysis of evidence of progress. One of the challenges highlighted in the research process is how PILO tracks whether CMCs and CMs are building productive relationships and helping to change routines in schools.

In this study, evidence of changed routines related to curriculum coverage conversations, reporting, sharing and reflecting practices are: 1) whether meetings, within CMCs and between CMs and Principals, are consistent and provide space for the discussion of curriculum coverage and management meetings; 2) if Principals and others perceive their relationship with CMs to have improved and, more importantly, have been assisted in dealing with curriculum management problems; and 3) what the CM Tool heat maps reveal as a dashboard for identifying curriculum challenges. The socio-economic context and district management and leadership cultures also shape practices. Together, these provide a “picture” of a possible shift away from compliance towards learning improvement. They enable us to explore if Jika iMfundo interventions are gaining traction in the shape of changed relationships and practices and tentatively identify possible lessons for learning improvement and system change.

District contexts

Pilot districts overlaid with homeland departments

Figure 9.4 Pilot districts overlaid with homeland departments

It is common knowledge that the socio-economic, political and historical context can affect change interventions. One challenge for Jika iMfundo is to be flexible enough to manage this. School and district systems are diverse and work well where they work (usually well-resourced public or independent schools) or are fragmented where they do not (most schools) (WSG & BRIDGE, 2016). The KwaZulu-Natal (KZN) system is no different, but has its own contextual challenges as an amalgam of the legacy of provincial, Bantu and homeland education systems. When the provinces were rationalised post-1994, KZN integrated three provincial Own Affairs education departments (for whites, coloureds and Indians), the provincial Bantu Education department and the former homeland of KwaZulu. Figure 9.4 shows that a sizeable chunk of former homeland schools fall into both districts. However, KC, as a largely rural area, is more affected by poverty than Pt. KC is characterised by traditional and urban settlements with some urban areas. Pt is urban with informal settlements and some traditional rural settlements. A large portion of Pt (one-half of the Ethekwini Metro) comprises formerly white and Indian own affairs schools which were better resourced.

Inequality and poverty impacts on learning, in terms of family and community resources to support learning and in terms of school resources for learning (Moses, Van der Berg, & Rich, 2017). The consequences are mediated in the schooling environment in terms of what resources exist, how they are used, what CMs can do and what teachers and learners do in classrooms (Van den Berg et al., 2016). Resource distribution and socio-economic context is reflected in the classification of schools into the quintiles (Q) indicating the level of poverty in surrounding areas – Q1 to Q3 are no-fee schools and Q4 to Q5 are fee-paying.

In this regard, Pt and KC are opposite. Descriptive statistics of schools in relation to quintiles show that 39% of schools in KC are in Q1, compared to only 1% in Pt. Conversely, 9% of schools are fee-paying in KC, whereas, in Pt, the proportion is 51%. Nine out of ten schools in KC are no-fee schools and the corresponding proportion for Pt is less than half at 48% (see Figure 9.5). There is great variation between circuits in Pt in the proportions of Q5 schools, the highest by far being Mafukuzela Ghandi, with Umhlathuzana the lowest, reflecting historical apartheid divisions for white, Indian and homeland areas and administrations. Figure 9.5 illustrates the proportions of schools per quintile, showing, for example, that within the CMCs of KC, there is great variation in the proportions of Q1 schools, with the main concentrations in Nkandla and Umlalazi, the heart of the former Zulu homeland.

Distribution of schools by quintiles: overall, districts, CMCs

Figure 9.5 Distribution of schools by quintiles: overall, districts, CMCs

Research over decades has debated whether supposedly unbiased proficiency tests at school – literacy, numeracy, etc. – are, in fact, biased against children from disadvantaged home backgrounds and/or in disadvantaged schools (see for example Reddy et al., 2015). Quintiles of schools, as defined, are an indirect measure of both at the same time. Do the ANA test results vary in this way, by quintile? On this data, the findings are both revealing and mixed. The following results incorporate testing with analysis of variance (ANOVA) to see whether differences are statistically significant or not.2 Looking at the left panel of Figure 9.6, for primary schools, one sees that ANA3-Maths and ANA6-Maths scores are significantly higher in Q5 schools than in Q1–4 schools. Other research (Reddy et al., 2015) has shown that Q5 schools are indeed markedly better off and have more advantaged catchment areas than the other four quintiles – and evidently in the two districts, the benefits do carry strongly into maths performance.

By contrast, the ANA3-English scores do not differ significantly across schools in the five quintiles. This is also true for ANA6-English scores. But ANA3-Zulu scores have an oddity: the scores do not differ significantly across Q1–4, but scores in Q5 schools are significantly lower. Perhaps fewer children at Q5 schools are home-language Zulu-speakers. When differentiated by district, ANA scores do not vary significantly between Pt and KC at primary level, but are higher in Pt than KC at secondary level. As expected, they inter-correlate well across levels. Looking at the right panel of Figure 9.6, for secondary schools, one sees that ANA9-English scores are nearly flat across the quintile range and, statistically speaking, are indistinguishable. Finally, one needs to look at the graphs more broadly and notice the relative heights of the lines with respect to the vertical axis. Excepting Q5, average scores are approximately 60% for ANA3, 45% for ANA6, 30% for ANA9-English and a disturbing 11% for ANA9-Maths. This trend highlights the urgency of prioritising learning improvement by capacitating CMs and SMTs to support the instructional core.

ANA scores broken down by quintile, for primary and secondary schools

Figure 9.6 ANA scores broken down by quintile, for primary and secondary schools

An additional contextual factor in the take-up change interventions in CMCs and among CMs is the leadership offered by the District Director. KC had the same director for the duration of the interventions enabling the coaching team to work closely with the CMCs to support CMs on their change journey. He was committed to the Jika iMfundo Campaign and actively drove change with his change team, as well as the CMCs. In contrast, Pt was beset with leadership challenges driven by internal politics evident in changing District Directors and numerous CM vacancies. The Pt District Coach notes (Interview transcript, 23 September 2016):

… the situation of Pinetown led to continuous change of district director in the past two years and acting people. We have had about four up to this point. … The role of district director is important within a framework of being in authority and an authority, due to position and also being competent and skilfully driving that process towards change.


A key purpose of the intervention was to enable CMCs and CMs to have professional, supportive conversations about the curriculum with Principals and SMTs, as well as plan and schedule visits, meetings and report backs. There is evidence from the various district management and CMC meetings to suggest that curriculum coverage is a standing item on meeting agendas, if not the primary focus for planning and engagement. For example, at a KC district management meeting in Empangeni on 1 June 2017, CMCs reported that they have regular weekly or biweekly meetings with CMs to discuss curriculum management (among other issues). One suggested that these meetings enabled him to “know what is happening in the schools.” Pt CMCs also noted the weekly meetings at a similar meeting on 2 June 2017. One CMC CES noted, at a meeting on 2 June 2017, that “we are in a terrible space – we are not coping” in reference to having three out of four vacancies.

However, the meetings still seem to focus strongly on the lack of human (and other) resources in circuits and schools, as well as a need for stronger “consequence management” for defaulting schools or poor performance. Working away from authoritarian line instructions to professional conversations based on evidence is a continuous challenge. One KC CMC said, “You sometimes have to be harsh to get them to implement” (21 June 2017). Some of the schools interviewed suggest that districts are more focused on curriculum management but there are still gaps in understanding:

We do not have a common vision and do not make use of the same instruments. They come in with their own instruments, which are different, and if he just does not understand your instrument, he just leaves it as it is. That is what I have noted so far. The district officials are more concerned about numbers, not quality, whereas I am more worried about the skills learners acquire (Interview transcript with school 10 Principal, 19 November 2016).

District CMC % revisit
Pinetown Ave. 32.8% Mafukuzela Ghandi 19.0%
Umhlathuzana 26.2%
Durban Northwest 52.2%
King Cetshwayo Ave. 38.6% Umlalazi 13.6%
Imfolozi 18.3%
Nkandla 35.5%
Mthonjaneni 51.9%
Umhlathuze 58.9%

Re-visit proportion by CMC

Table 9.3 Re-visit proportion by CMC

The CM Tool does seem to have provided a useful basis to start conversations about curriculum coverage and management from some CMs and schools:

This programme has not only brought the Circuit Manager closer to the school but it has made the Circuit Manager understand better the challenges that Principals face. The Circuit Manager has now reason to visit and monitor the school progress based on the MRT findings (Interview transcript for school 13 Principal, 25 November 2016).

However, visiting consistency varies across circuits, perhaps due to a combination of CMC CES leadership, as well as resource availability. In the round of CM visits to schools in early 2016, nearly all schools in the analysis dataset were reached: 947 out of 968, that is, 97.8%. But, in the late-2016 round, only 347 (36%) of the schools were re-visited. The proportions were 32.8% for Pt and 38.6% for KC perhaps reflecting the different leadership cultures. The reasons for this decline in visits may be related to the difficulty of working in district contexts where officials are constantly battling vacancies and crises, or a consequence of leadership and management. This is reflected in the variation between CMCs in Table 9.3 where Durban Northwest in Pt and Mthonjaneni and Umhlathuze have revisit rates over 50%. Timing, as well as other priorities, means that visits may slide, or take place in different forms – such as over the phone. Most CMs are preoccupied with National Senior Certificate (NSC) exams in the fourth quarter.

Perceptions about change

A primary focus of the intervention in districts is to change the working relationships within and between schools and districts to build learner focus, and professional and reciprocal accountability. Most circuits and schools have norms and routines which define the way things work. Many of these are rooted in the struggle against apartheid education. Cultures of mobilisation, entitlement and dependency continue to define relationships in many schools and departmental contexts. The difficult experience in Pt, with shifts in leadership and political positions, demonstrates the effect of deep-seated and underlying tensions based on “historical professional animosity and persistent mistrust” (PILO Workshop Notes, 10 May 2017). The consequence of this is that the intervention had far less traction at district level Pt than KC. This is less evident at school level in relation to SMT training. Collaborative relationships between CMs and CMCs also vary. This is demonstrated in a lack of trust in the accuracy and validity of CM reports on school visits. Also, as shown in both district charters, CMs and SAs work in silos, often duplicating visits, instead of sharing resources and knowledge.

Many of the values which characterise collaboration ‒ trust, reciprocity, tolerance and accountability ‒ were eroded in the struggle over apartheid education. In some cases, this led to a disintegration of an accepted normative code, evident in a fragmented and dysfunctional engagement, low expectations and unprofessional conduct. The district stops and starts in Figure 9.2 reflect these engagements, highlighting the need for more professional conduct in relation to schools. While it is difficult to demonstrate a shift in relationships, many of the interview transcripts and monitoring reviews suggest that there has been a change in the way in which CMs interact with schools:

Their approach to the school is developmental now. They are not coming for fault finding but for conversation based on evidence. They are more supportive (Interview transcription with Principal in primary school 3, 11 October 2016).

The relationship is good because the Circuit Manager was here and he stayed for about 2h30 minutes. They support us a lot and are approachable. If we are having a problem, we just phone to request their support. They sometimes phone us to check whether everything is okay (Interview transcription with Principal in primary school 7, 5 November 2016).

In addition, the CM Tool seems to provide a basis for conversations about curriculum:

It has improved a lot because, whenever the officials come here, they come with tools that have to be improved, unlike in the past when I will not be able to provide what they require from me. At the present moment, I am able to provide evidence of my work as there are tools that I use and everything is recorded in the SMT files. If I do not have specific tool that they need, I request them to help and support me as I have tried to develop it. They are able to provide me with the necessary support (Interview transcription with Principal in secondary school 8, 12 November 2017).

The extent and consistency of changed relationships is unknown at this stage. After three years, CMs are still in the initial stages of changing practices. A district coach notes:

There is traction in routines. There is also contracting to the needs of professional conversations that are based on facts and data which are not personalised but focusing what on what needs to be done (Interview transcript with district coach, 23 September 2016).

A school Principal confirms that:

… personally, I think there is more focus on curriculum coverage, monitoring and on supervision that has made us cope as leaders (Interview transcription with Deputy Principal in primary school 10, 19 November 2016).

One of the benefits of the engagement between the various parts of the system seems to have been the clarification of roles and responsibilities within and between districts and schools. Leadership roles and responsibilities are not always clearly differentiated or understood in policy or practice. Some CMs and Principals are beginning to understand their own role in achieving curriculum coverage by focusing on curriculum management support and leadership respectively. School leaders are in the front line of the struggle to improve learning in schools, but different policy frameworks create a confused, compliance approach. This can be exacerbated by policy enforcing and blame finding engagements:

There are different units within the district structure, so those that are related to curriculum delivery seem to be improving the relationship with us drastically. But, with those distant units to curriculum delivery, there is still a problem because they do not understand how the Jika iMfundo come in (School 10 Principal, 19 November 2016).

Heat maps as dashboards

The scores generated by the CM Tool, as summarised in the “heat maps”, are based on assessments of the multiple items by CMs, on the basis of their engagements at the schools. Even at their best, they are not precise or verifiable data like, say, the schools’ quintiles. However, they are useful in several ways. The first is for Principals to identify curriculum management problems in their own schools and use this as a basis to ask CMs for help and support. They also provide a foundation for discussions on curriculum coverage. In addition, they can be used as aggregated or higher-level “dashboards” to identify challenges within circuits, across CMCs and in the district.

The statistical analysis sought to see how this was working in practice, albeit at a very early stage of the project. The start-up limitations were noted earlier: that, in the first round of visits, the CMs were still getting the idea of assessment, so the recorded assessments were evidently too uniform and too high; and that, although scoring in the second round was better differentiated, only 37% of schools were visited. So, the analysis should be taken as merely illustrative. It focused on the General Management scores for strategy, relationships and finance, plus a total, for all the actual schools in a specified circuit (summarised in the heat map below) and a second round of visits, for the proportion of schools re-visited, with the first round. Additionally, aggregating of the circuit heat-maps for higher-level managers was examined.

As an example, Figure 9.7 reflects the aggregation upwards across the by-school maps in the circuits, to yield a summary at CMC level. In row 6, the Durban Northwest CMC in the Pt District, one sees that (from the first visit to the second) strategy has improved from 12 to 13 out of 16, taking it from amber to green. In KC, rows 1–5, one sees in the two total columns that all five CMCs have improved from amber to green.

Round 1 Round 2
Strat/16 Relat/11 Finan/8 Total/56 Strat2/16 Relat2/11 Finan2/8 Total2/56
King Cetshwayo 1 Imfolozi 6 10 7 45 13 10 7 49
2 Mthonjaneni 14 10 6 46 12 11 7 48
3 Nkandla 12 9 5 39 11 10 6 47
4 Umhlathuza 14 10 6 46 12 10 7 49
5 Umlalazi 15 9 6 43 13 10 7 51
Pinetown 6 Durban Northwest 12 10 7 50 13 11 7 51
7 Mafukuzela Gandhi 13 10 7 47 13 10 7 50
8 Umhlathuzana 13 10 7 48 13 11 7 51
Total 13 10 7 49 13 11 7 51

CMC-level “heat map” showing three areas of general management

Figure 9.7 CMC-level “heat map” showing three areas of general management

While remembering the strictures about the start-up data, an analysis of variance (ANOVA) was again undertaken, now applied to the changes summarised in the General Management heat map to illustrate what will be possible in future, with the CMs having experience behind them for first- and second-round scoring and with a better proportion of second-round visits.

The outcome is shown in Figure 9.8. To interpret the figures, one may begin with the bottom left-hand panel. It displays changes in the finance scores3 from the earlier to the later visit. The lower, red line refers to Pt. The left-hand point shows that there has been a slight improvement of about 0.3 units among primary schools in Pt, while the right-hand point shows that there was a slightly larger improvement of 0.4 in secondary schools in Pt. Similarly, the upper, blue line refers to KC, where the changes were greater in both primary and secondary schools than in Pt. The left-hand point shows that the improvement in finance was 1.0 units for KC primary schools, slightly greater than the 0.8-point improvement shown by the right-hand point. Other panels are interpreted in the same way. The bottom-right panel shows a similar pattern to the bottom-left panel in changes in the total of performance scores4 but the top two panels present a different pattern. For instance, at top-left, dealing with changes in relationships, in primary school changes from first to second visit are much less for Pt than KC, as before; but in secondary schools, changes in Pt are roughly as great as in KC.

Bearing in mind the limitations of the current data, what might this analysis tell us about whether Jika iMfundo is working at district level? Firstly, changes are larger in KC than in Pt (with the one exception), especially at primary level. Anecdotal evidence from the PILO team suggests that this is because the KC district director, with his CMCs, were committed drivers of change. This suggests that there was more follow up on the results. Also, KC had more room to improve given the large number of no-fee schools. Secondly, changes within Pt are larger for secondary schools than primary and vice versa in KC. Thirdly, across the districts, the differences between them at primary level are greater than at secondary. Particular variations, visible on each line of the heat maps at the lower levels, will be of interest to the CMs visiting the schools, or their CMC supervisors. But the gain from the averaging for the high-level picture is that, with the large numbers involved, the statistical tests would be able to indicate which of the myriad differences were large enough to be significant – yielding the “big picture” for overall programme management and prioritisation.

Comparison of changes at district level at primary and secondary levels

Figure 9.8 Comparison of changes at district level at primary and secondary levels


This research set out to explore if it is plausible to assume that districts are an important lever in systemic change, especially in contexts where resources are limited and morale tends to despair or apathy. Does the improving of relationships between districts and schools provide a path towards the building of professional accountability and collaborative commitment to change, despite the context? The answer is yes. There is qualitative evidence and – to the extent that the incomplete start-up data are nevertheless indicative, some quantitative evidence – to suggest that relationships and interactions between some of the schools and their respective circuits are more interactive and less-compliance oriented, perhaps more so in KC than Pt. This is evidenced in better planning, scheduled visits and a shift away from compliance checking to “how can I help you?” A skilled and empowered district official, working with Principals and SMTs can, as Fullan (2015) suggests, “lead from the middle”. The CM Tool provides a useful basis for professional conversations to begin. However, a tendency to focus on “management compliance” rather than “curriculum management” persists, perhaps because finances or human resources are more concrete and seem easier to control. Where the tool is used effectively, both CMs and Principals begin to see themselves as instructional leaders. The challenge is to ensure that the CM Tool does not become a compliance instrument as working to rule is deeply embedded. It is only effective if it is used as a basis for a professional conversation as part of a learning support process.

Tools and using evidence for planning and conversation allow new routines to be introduced and practiced. It is this reinforced practice that enables relationships to shift focus and be institutionalised. It is too early to tell now, but this would confirm the Theory of Change which underpins the intervention. The district and school coaches are a vital part of the process of embedding the routines by encouraging CMCs, CMs and SMTs to try to work with the tools and the evidence. Some consideration needs to be given, in the long term, about where this prodding will be located within the system. The heat maps and limitations noted provide a means for identifying curriculum management challenges, particularly as part of evidence based discussions and plans. Some schools and CMs are focusing on resolving curriculum management problems. A limitation of the current model is that Principals, CMs, CMCs and districts do not have a way to track their own change progress. Revised and moderated heat maps will enable this process.

Elmore’s (2016) scepticism about how well systemic change responds to differentiated school needs is borne out by research. The socio-economic context, combined with leadership cultures in districts and circuits, had had effects on implementation, especially on Pt as a district, but also in some circuits. Many of the socio-economic and resource inequalities between schools, circuits and districts are structural and hard to change. Apartheid, race, class and spatial legacies do structure relationships and learning conditions. Also, leadership and resource challenges make change harder to achieve. The PILO approach of building commitment allows officials to see and understand what can be done despite the prevailing conditions. Flexibility is accommodated by starting the journey where the CM or CMC is at, but the risk is that schools or CMs may opt out or be left behind. There may be conditions which make change unworkable but the utility of the focus on agency and constructive conversations suggests, contrary to many school change interventions, that very depleted dysfunctional schools or districts need not be overlooked.

I like Jika iMfundo because, before 2014, the managers will come once after six or seven months, now maybe twice a term to support, give development, check curriculum coverage, to sit down one-on-one for you to explain your challenges and discuss how you can improve. We are no longer scared of them, because now we just communicate. They now use the CM Tool to ask you questions that will result in support. So, the relationship has really changed a lot (Research transcript of school Principal 1, 3 October 2016).

The authors thank Jon Hodgson for his assistance in providing and harmonising the data used in this chapter.



  1. For example, the designation of a school as primary could be “Primary”, “prim”, “P”, or even “Primary”. 

  2. At the customary level of p < 0.05. 

  3. The finance scale runs from 0 to 8. Differences can be positive or negative, but change – if there is any – will typically be 1 or 2 points from one visit to the next. So, the averaged change scores reflected on the vertical finance scale, for primary or secondary, are not negligible. 

  4. The total score, across the seven areas (of which we are attending to three for this analysis) runs from 0 to 56. Changes, if any, will typically be a few points. So, as in the previous endnote, the averaged change scores are not negligible.