Not a panacea, but vital for improvement? Leadership development programmes in South African schools

The performance of South Africa’s educational system in national and international tests indicates that a large proportion of its public schools are underperforming. Ineffective leadership has been put forward as one reason. This paper analyses a leadership development programme, Partners for Possibility (PfP), which seeks to strengthen leadership in underperforming schools. PfP employs a one-year, cross-sector social partnership that pairs school principals with business leaders. Literature on leadership development attests to the problem of learning transfer. Interviews were conducted with 9 PfP principalbusiness leader dyads and 4 PfP unpaired participants. The research also included observations of PfP meetings and analysis of PfP reports. Although PfP’s partnerships were found to support the school principals’ learning, the cost of the programme limits its scalability. PfP’s group learning component was also found to sometimes divert partnerships from transformational to transactional forms. This study contributes to the understanding of learning transfer within social partnerships. It also highlights a limitation of programmes such as PfP, which aim at school improvement, but are not focused on instructional practice. The article explores ways in which partnership programmes could be strengthened.


Introduction
Effective leadership and management are vital to the success of any school (Bush, Kiggundu & Moorosi, 2011;Leithwood, Louis, Anderson & Wahlstrom, 2004;Ngcobo & Tikly, 2010). Leadership contributes significantly to learner's learning albeit indirectly through the principal's influence on teachers' instructional practice, which directly determines learner achievement (Leithwood et al., 2004;May & Supovitz, 2011;Supovitz, Sirinides & May, 2010). Nevertheless, in most countries, including South Africa, school principals assume office without specific preparation (Bush & Heystek, 2006;Bush & Oduro, 2006). The South African government's decision to allow for greater school self-management has augmented the principal's workload (Bush & Heystek, 2006;Christie, 2010), with, inter alia, fundraising and human resource management (Christie, Butler & Potterton, 2007;Ngcobo & Tikly, 2010;Taylor, 2008). Furthermore, difficulties resulting from apartheid-determined historical disadvantage in the majority of South African schools (Christie et al., 2007;Ngcobo & Tikly, 2010) make school leadership an onerous task. As Zuze and Juan (2018) affirm, this may necessitate differing school leadership and management approaches. International and local research highlights the significance of context in the consideration of school leadership models (Christie et al., 2007;Eacott, 2013;Gunter, 2013;Ngcobo & Tikly, 2010;Spaull, 2013a). That about 75% of South Africa's public schools are underperforming in national and international tests (Spaull, 2013a(Spaull, , 2013bTaylor, 2008) points to what Spaull (2013a) refers to as a dualistic schooling system. He warns that "modeling a single schooling system when there are in fact two school systems can lead to spurious results and misleading policy conclusions" (Spaull, 2013a:436). Leithwood et al. (2004:5) argue that in underperforming schools leadership has a greater impact on learner outcomes, hence the need to address leadership capacity of such schools.
Partners for Possibility (PfP), a leadership development programme that started in South Africa in 2010, aims to improve South Africa's education system by strengthening leadership in underperforming public schools. It attempts to do this by pairing school principals with experienced business leaders (henceforth business leaders) for a period of twelve months, during which, these dyads engage in peer learning while participating in a structured process comprising formal workshops and practice-based activities. PfP's leadership development programme is a trisector, cross-sectoral social partnership (CSSP). Trisector CSSPs involve business, government, and civil society (Googins & Rochlin, 2000), here represented by the business leaders, the principals of public schools and PfP. To date, 977 public school principals, from all South Africa's nine provinces, have participated in the PfP programme (Partners for Possibility, n.d.-a).
There is dissension around the appropriateness, importance, motivation, and results of business involvement in education due to the different values, cultures, and interests of the public and private sectors (Googins & Rochlin, 2000;Mickelson, 1999). These are valid criticisms. Nevertheless, South Africa's weak education system and the lack of leadership training justify examining initiatives aimed at improving underperforming schools. Two features of the PfP model make it stand out.
The first of these is the emphasis on peer learning, as opposed to mentoring, which is found in most leadership development programmes (Bush, 2009;Bush et al., 2011;Fullan, 2009). Mentoring differs from peer learning since "mentoring relationships involve a one-way helping dynamic while peer relationships involve a two-way exchange" (Kram & Isabella, 1985:129). This article explores this two-way exchange in the PfP programme.
The second feature that particularises the PfP methodology is the demand placed on both school principals and business leaders to participate, as dyads, in its training programme. Mentors in school leadership development programmes receive training to empower them in their supportive role (Bush, 2009), which is usually distinct from that which their mentees receive.
This article contributes to understanding how the learning acquired by school principals in a leadership development programme can, through peer learning relationships with business leaders, be transferred to schools to improve performance. Learning or training transfer refers to the application of trained knowledge and skill in the job (Burke & Hutchins, 2007). Developmental interactions such as peer learning and mentoring facilitate the transfer of learning to the practitioners' specific contexts (D'Abate, Eddy & Tannenbaum, 2003). International research on school leadership development interventions indicates that there are challenges to how learning imparted in these programmes is turned into practice, since leadership is a "hands on" activity (Bush, 2009;Fullan, 2009;Leithwood et al., 2004). Leadership development programmes in South Africa face similar challenges, as illustrated in Naicker's and Mestry's (2016) study of a three-year leadership development programme in Gauteng, which established that learning transfer was the least effective feature of the programme. In their evaluation of the national Advanced Certificate in Education: School Management and Leadership (ACE Programme), Bush et al. (2011) found that the effects of mentoring were inhibited due to mentors' tendency to prescribe solutions to school principals.
The article also highlights the importance of content in the design of programmes such as PfP's. Leadership development programmes aiming at school improvement are more likely to succeed if the content of such programmes is focused on curriculum and pedagogy. Improving learning outcomes is dependent on these two instructional tasks (Bush & Heystek, 2006;Taylor, 2008). This means that if PfP aims to improve learning outcomes in underperforming schools via improved principal leadership, its content should also empower principals to be leaders of learning, the absence of which points to a critical limitation in PfP's programme design.
The following section contextualises school leadership in relation to teaching and learning, leadership development and social partnerships in South Africa, as well as outlining PfP's programme design. The research methodology employed is then described, followed by the findings discussed in four related themes demonstrating how learning in PfP's programme improved school principals' leadership, the effects of this on teaching and learning in underperforming schools, and the challenges encountered in PfP's programme design and execution. The discussion considers how learning transfer occurred, the effect of PfP's design on the programme's outcomes, and problems related to content and credibility of this intervention. We conclude by noting the resource challenge faced as well as the importance of aligning the content of programmes to context, which in PfP's case, is improved academic performance of underperforming public schools.

Leadership Influence on Teaching and Learning
School leaders contribute to learners' learning indirectly through their influence on people or features of their organisation (Leithwood et al., 2004). Since learning is directly influenced by the quality of teaching, the latter is of particular concern in South Africa where classroom practice is hampered by weak teacher subject and pedagogic knowledge (Spaull, 2013b;Taylor, 2008), teacher absenteeism, inadequate infrastructure and teaching resources, undisciplined learners contributing to difficult classroom conditions, and a lack of support from parents and local communities (Christie et al., 2007;Taylor, 2008). School leadership is a key lever in school reform, second to teaching (Leithwood et al., 2004;Supovitz et al., 2010), as it affects learners' learning through initiatives that build teacher capacity such as the formation of professional learning communities (Harris, 2010;Little, 2002;Stoll & Louis, 2007), and by providing targeted feedback on instructional practice (May & Supovitz, 2011). Although the centrality of teaching and learning is affirmed in South African studies on well-functioning schools, effective leadership in these studies goes beyond instruction by focusing on the creation of a safe and secure environment (Christie et al., 2007;Ngcobo & Tikly, 2010;Zuze & Juan, 2018). Because school improvement requires that school principals be leaders of learning (Bush & Heystek, 2006), it is reasonable to expect that a leadership development programme targeting school principals and aiming at improving school performance should include strengthening instructional leadership within its objectives.

Leadership Development
Although related to teaching, school leadership is nonetheless distinct, hence the need for leadership development (Bush, 2009(Bush, , 2012. Leadership is here characterised by influence and the provision of direction (Christie, 2010;Leithwood et al., 2004). The concepts "leadership" and "educational leadership" are not without critique (Eacott, 2013;Gunter, 2010Gunter, , 2015. Gunter (2015:32-33) argues that the current approach to school leadership is about managerialism and is little concerned with pedagogic processes. Thus, school principals are perceived, erroneously, as managers rather than educational leaders (Lingard, Hayes, Mills & Christie, 2003).
Increased interest in educational leadership in the 21st century, both locally and internationally, is due to greater awareness that school headship requires specific and systematic preparation (Bush, 2012). According to Bush and Glover (2012), the central purpose of school leadership development is to improve leadership capabilities and abilities for school improvement. Leithwood et al. (2004) argue that effective leadership sets direction, develops staff capacity, and creates an environment that favours teaching and learning. Leadership development activities should thus aim at building capacity in leaders to improve their effectiveness (Bush, 2012). Most reform efforts fail because their theories of change tend to neglect capacity building (Fullan, 2006;Harris, 2010).
A successful leadership development programme should allow for learning that is job-embedded, organisational-embedded and system-embedded (Fullan, 2009). Although some leadership development programmes incorporate job-embedded learning, Fullan (2009:45) maintains that other types of learning are seldom incorporated. Job-embedded learning refers to learning onthe-job as it is grounded in daily work, thus connecting theory to practice (Bush, 2009). Learning in context also allows practitioners to tap into their previous experience and knowledge of the learning process (Leithwood et al., 2004:67-69). Whereas job-embedded learning improves the individual leader, organisational-embedded learning is about building a school's collective capacity to improve the organisation's culture, structure, and processes to generate organisational change across the system (Fullan, 2009). System-embedded learning extends beyond the school and encompasses the entire education system, enabling learning between schools and between the various levels of the system (Fullan, 2009).
According to Bush (2009), the formal content in leadership development programmes has had a limited impact on leadership practice. Various elements characterise these programmes, such as mentoring, action learning, and school visits, customising the learning to the practitioners' specific needs, while providing for group or collective learning (Bush, 2009:379-381). This shift from content to process can be found in leadership development programmes in South Africa, such as the ACE Programme. Bush et al. (2011) highlight the central role of mentoring in the ACE programme but note, nonetheless, that mentoring and networking were both weakly executed, pointing to a lack of jobembedded and collective learning. Similarly, although Naicker and Mestry's (2016) research on the three-year leadership initiative in Gauteng indicates that collective learning was taking place, their findings point to the lack of job-embedded learning, rendering the initiative's impact on educational outcomes moot.
The difficulties encountered in implementing process elements in these studies point to the significance of Fullan's (2009) notion of the need for all three types of embedded learning to be present for successful leadership programmes. The limited impact of leadership training on practice also points to a transfer problem (Baldwin & Ford, 1988). Learning or training transfer refers to the application of trained knowledge and skill in the job (Baldwin & Ford, 1988;Burke & Hutchins, 2007). Factors affecting learning transfer relate to learner or trainee characteristics, intervention design and delivery, and the work environment where supervisory/peer support is considered a critical component of supporting skill maintenance (Baldwin & Ford, 1988). The success of a leadership development programme will thus be influenced by the presence, or otherwise, of these factors.

Partnerships for Social Initiatives
The PfP leadership development programme is a trisector CSSP. CSSPs are said to respond to the complex nature of social issues, such as poverty, healthcare, and education, requiring multi-sector solutions. It is argued that these partnerships have the potential to succeed as it is to everyone's advantage that these problems are resolved (Googins & Rochlin, 2000).
Partnerships can be transformative or transactional, depending on the manner in which the parties in the partnership establish their working relationship (Butcher, Bezzina & Moran, 2011). Transformational partnerships are characterised by partners' shared vision and by the impact of their activities on the organisations' culture (Butcher et al., 2011;Lasker, Weiss & Miller, 2001). This is unlike transactional partnerships, which are established to meet specific goals without effecting fundamental changes to the parties involved (Butcher et al., 2011).
Although partnerships are associated with synergy (Lasker et al., 2001), marginalisation of weaker partners can occur (Googins & Rochlin, 2000;Lasker et al., 2001). Social partnerships also face the problem of legitimacy, exerting pressure on business entities to justify their involvement by displaying the results of their activities (Van Tulder, Seitanidi, Crane & Brammer, 2016). Among the dangers of business involvement in education, Mickelson (1999:504) describes the absence of evaluation as the most serious because "there is very little reliable evidence that these [involvements], in fact, lead to positive school outcomes for students." Nevertheless, Timpane and McNeil (1991) maintain that business can provide support to educators and policy makers. However, they caution that business involvement should be limited as education is a public good and therefore a public responsibility. With this in mind, we look at lessons that can be derived from the PfP leadership development programme.

PfP Programme: Overview and Design
The PfP leadership development programme is a project of Symphonia for South Africa, a non-profit organisation (NPO) and social enterprise. Social enterprises attempt to address social needs through the application of business practices (Malunga, Iwu & Mugobo, 2014). PfP aims to reform South Africa's educational system by improving the leadership in underperforming schools, which is expected in turn to lead to better educational outcomes. Enrolment into the PfP programme is voluntary for business leaders (Collins, 2015:196). School principals may also join the PfP programme voluntarily, or they may be recommended to the programme by district education officials (Collins, 2015:114). Before embarking on this programme, both school principals and business leaders receive preparation directed at effective participation (Collins, 2015:114).
The PfP programme consists of nine training elements executed over twelve months. The design includes three workshops, namely: "Time to Think," a one-day workshop aimed at partnership building held around week three of the programme, "Flawless Consulting," a two-day workshop which takes place around week fifteen, imparting skills on how to contract with stakeholders in the schools, and "Community Building," a two-day workshop occurring around week twenty-seven and aids the PfP dyads to mobilise the community around the school (Collins, 2015:97-104). The PfP training content, therefore, does not include school management and leadership issues. The PfP design does include process elements such as coaching, action learning, community of practice (CoP) meetings, journal keeping, and a portfolio of evidence. Action learning in the PfP design occurs within the partnerships formed between principals of underperforming schools and business leaders. Action learning is an experience-based approach aimed at enhancing skills for professional development purposes through the interaction of a diverse problemsolving team (D'Abate et al., 2003). The PfP partnerships meet between 3 to 5 hours a month to implement the learning derived from the programme. PfP provides expert coaching to the dyads through its learning process facilitators (LPFs). Each LPF is assigned to a CoP consisting of between eight to 10 PfP partnerships where networking and cross-partnership learning occurs. The LPFs facilitate CoP meetings, which occur every six weeks for 2.5 hours.
The PfP design seeks to bring about changes in underperforming schools at four levels: the principal, the school management team (SMT), the educators, and parents/community. i A hierarchy of goals is discernible in the PfP change process since the programme's immediate concern is to capacitate the school principal as leader. This is anticipated to improve the way in which the SMT and educators work, in addition to allowing for greater parental and community involvement in the school.

Method
The study applied a qualitative approach situated within the interpretive research paradigm. The population consisted of 193 partnerships that had been registered with PfP in the Gauteng Province at the time of the study. Through purposive and quota sampling, a total of 43 partnerships were selected as possible participants. Effort was made to secure complete partnerships for triangulation, for coherence, and the validation of data. Verification of data collected was enabled through the various reports made available by PfP, as well as through the observation of a CoP meeting and two celebration events, which marked the end of participation in PfP's year programme for two groups. A total of 22 partners were interviewed between August and November 2016: nine dyads (18 people, each interviewed separately), one school principal, and three business leaders whose PfP partners did not participate in the study. Six of the nine partnerships had been initiated in 2015, two in 2014, and one in 2012. All principals were heads of no-fee schools: six situated in townships, three in poor suburbs, and one in an informal settlement. Of the 10 school principals interviewed, nine headed primary schools and one a secondary school.
Semi-structured in-depth interviews were conducted with study participants. A pilot study pointed to the need to examine the dynamics arising within the leadership programme. A large amount of data was collected from the respondents and this was coded using categories emerging from the data and grouped into topics derived from literature and from the collected information. A potential weakness of the study resonates with Bush's (2009:384) critique of evaluations of leadership development activities as these often rely on selfreported data.
Ethical approval for the research project was granted by the Wits Human Research Ethics Committee. Since the study examined the PfP process, participation was restricted to persons who had completed PfP's one-year programme. An ethical concern deriving from this study was that some participants were curious to learn what their PfP partners had said in the interviews. A careful balance, therefore, had to be maintained to safeguard confidentiality without compromising the research relationship (Legard, Keegan & Ward, 2003).

Findings
Four themes pertinent to how learning transfer occurs in a leadership development programme emerged from the study. The first considers the effects of improved leadership on teaching and learning in underperforming schools. The second looks at PfP as a capacity building programme for school principals, since this initiative focused on capacitating principals with work-relevant skills. The third considers how the learning in PfP is transferred to schools, as this is a recognised challenge in leadership programmes. The fourth theme presents the challenges encountered in the PfP programme design and execution. The problem of ascertaining the impact of the PfP programme on school outcomes indicates the limited focus on teaching and learning inherent in the PfP design. The four themes highlight how leadership development models could be strengthened.

Effects of Improved Leadership on Teaching and Learning
All school principals from the nine PfP partnerships interviewed reported that they had become better leaders since they listened better, were more inclusive, empathetic, and confident. Although these skills cannot be directly linked to teaching and learning, they can positively affect the educational project. One school principal related that the most valued lesson gained from her business partner was to avoid stressing when faced with challenges as this would cloud her thinking. She shared how she had put this lesson into practice. One of her teachers, though competent, was often absent from school. Conscious of the benefits of remaining calm, she narrated how she entered her office and sat back in her chair: The above example shows how change in the principal's leadership perspective enabled her to constructively engage the SMT to ensure that teaching and learning continued, despite the teacher's absence.
The study also indicates that the PfP dyads' activities in schools can have a direct impact on teaching and learning. Of the nine partnerships interviewed, seven were found to have engaged in activities which favoured teaching and learning: enrolling teachers to the "BrainBoosters" ii programme; a discipline programme that was organised for teachers in a particular school by the principal's business partner; iii computer training for teachers in several schools; and lastly, numeracy and literacy programmes in three schools. The principal of a secondary school explained how his business partner introduced him to the principal of a nearby private school and that, thanks to this relationship, his teachers and learners could then access mathematics and science material online. The principal attributed the improved pass rate, which had risen by 7%, to the partnership's initiatives (Partners for Possibility, n.d.-e).

Building the Capacity of School Principals
All participants affirmed that PfP capacitates school principals with leadership skills. The formal content delivered in the three PfP workshops was relevant to the school principals' work, despite the training being on leadership in general. Several school principals reported having improved relationships with parents and educators, better conflict management, better delegation to other SMT members, and enhanced fundraising skills. One principal shared how she managed to involve more male parents in the school by appealing for their assistance in converting an old toilet block into a classroom. She reported that she drew this idea from the third PfP workshop on community building:

If I did not attend that community building workshop, I wouldn't have been brave enough to say: 'men: come' ... You should see the classroom: it's beautiful. So, I just wanted to make men to be involved, because you call parents meeting, in numbers women, fathers just a few. So, I just wanted to make them feel that the school needs them.
The process elements of the PfP programme discussed earlier were found to complement the learning derived from the workshops, especially the action-learning component of the PfP partnership model. This was affirmed by all the school principals in the nine partnerships that participated in this study. A principal shared that although she had previously managed to mobilise resources from companies, the lack of a long-term vision limited the impact of these one-off donations. She stated that the PfP training and her business partner helped her develop a working strategy: Collective learning was confirmed by all the principals interviewed, especially in the CoPs. Principals maintained that they had learnt from one another as well as from the business partners. One principal expressed this as follows:

I don't think if [Alex] was not my partner I would be able individually to come up with
We had CoP meetings. We will exchange the knowledge of how to handle issues. You will find different principals sitting together, and different businesses, and then when we approach issues like governance, you will find out that solutions are there. The other principals, they handle these things this way, the business people from the corporate, they can solve this problem this way.

Learning Transfer through Social Partnerships
The school principals applied the content learnt in the PfP workshops through the support provided by the social partnerships. Nevertheless, the fact that some partnerships accomplished more in their schools than others points to the differing degrees in the transfer of learning occurring. The social partnerships, the central feature in the PfP design, are intended to be transformational, that is, capable of giving rise to substantial and sustainable changes in schools and subsequently leading to improved educational outcomes (Collins, 2015:85).
Learning transfer is optimal in a partnership that is transformational rather than transactional, due to the effects of the transformational partnership on school culture and practice. One of the partnerships interviewed organised a team-building activity in the school where the principal was concerned about staff rivalry. This led to improved staff relations and a change in the school culture as the teachers improved in time-keeping. The school principal felt that this activity had led to a significant change in the way the staff at his school worked and related with each other, since its effects were still being felt a year later: In quite a lot of [the staff], it still sounds like, you know, they went through that team building just yesterday or just last week ... People have really internalised that whole process. It is not just a oneoff event.
Another school principal shared that prior to engaging in the PfP partnership, a concern was that his staff would call in sick on the day that they were due to submit assigned tasks. The principal explained that his business partner had made him aware of the need to communicate better with his staff, and to follow up on their progress as they worked on the assigned tasks. This was because the school principal was not in the habit of discussing the specifics of the task with the staff or of following up their progress for them to meet the deadline. Subsequently, his staff struggled to meet their targets. The school principal expressed this as follows: [Kabelo] spotted that most of the people that I am working with will struggle with details. And I will have left them far [behind] because I have gone [on ahead] ... These days, when I bring a project, I try to close my weakness.
Transactional partnerships, in contrast to transformational ones, are those partnerships that aim to meet specific needs in the targeted schools, usually infrastructure-related, with little or no impact on school culture. Yet, infrastructure-related projects are appealing to PfP partnerships for three reasons. The first is that underperforming schools also tend to be under-resourced. Several of the school principals interviewed admitted that a key motivation for joining the PfP programme was the hope of the school gaining financially. Another reason why a PfP partnership could opt to engage in infrastruc-ture improvement in the school is that the business leaders have little expertise in education-related issues. Thirdly, activities aimed at facility improvement are easier to undertake and yield clearly visible results. A business leader noted this as follows: [Resource-related activities] are possibly, for some business leaders, the easy thing to do … So [they] do a lot of stuff, but it's resource-stuff: it's things. It's not leadership; it's not management; it's stuff. The flip side is whether the principals actually prefer that stuff over [leadership and management related activities] because that is more tangible … The fact that some of the partnerships generated a greater impact in schools than others can be linked to some PfP partnerships being transactional. This transformational-transactional aspect should be monitored in any project employing developmental interactions for professional development.
Business leaders also learned through their participation in the PfP social partnerships. Most of the business leaders admitted in the interviews to having been anxious at the start of the partnership due to a lack of expertise in issues related to education and schools. Given that eight of the 12 business leaders who participated in the study were White, their involvement in the PfP partnerships exposed them to the difficult socio-economic challenges of communities that had been marginalised under apartheid. By working alongside the school principals, the business leaders gained a better understanding of the structural injustices still affecting South Africa's education system. This was articulated by a business leader with an educational background: [On enrolling into the PfP programme] business leaders get surprised at how little they know the educational field and the school environment, and the challenges facing schools and school principals. Whereas, before they would blame teachers for poor performance, they come to understand that it is not the educators. The problem or issues are big.

Challenges Encountered in PfP's Design and Evaluation
Three challenges were noted in relation to the PfP design and evaluation. Firstly, the vision informing the PfP design can promote unrealistic expectations with regard to the kind of transformational changes that the partnerships can bring about in schools in one year. The PfP programme design suggests that after twelve months, the partnerships will have helped the principals become better leaders and that this will have trickled down to the SMT, the educators and the parents for the benefit of the learners. However, the research indicates that the realisable goal after the first year of the programme is restricted to improved leadership skills in school principals. All nine pairs of PfP partners interviewed felt that at the end of the partnership year, they had noted the improvement in the principal's leadership, but that a year was too short to see changes in the SMT, the educators, and parents. Secondly, elements denoting unhealthy competition, such as feelings of comparison and envy, were expressed by some of the partners in relation to the CoP meetings. These feelings were found to generate a sense of insecurity, which hindered the way some partnerships operated. One school principal commented that, although he faced no personal challenges in his PfP partnership, the fact that his school had gained very little financially in comparison with the other partnerships in his CoP made his partnership appear to be non-performing: The CoP meetings tended to be competitive. It was a subtle competition, with partners showing off what they have achieved. If your achievement is however not material, then you felt the need to apologise. Consequently, some partnerships that are doing well become apologetic.
The negative competitive element referred to above encourages the formation of transactional partnerships at the expense of transformational relationships, since the former give rise to visible and immediate results.
Thirdly, the interviews also pointed to the difficulty in determining the impact of the programme on education outcomes in the PfP partnership schools. Although all the school principals maintained that they had become better leaders and that their schools had improved, the measures of improvement were subjective. None of the respondents interviewed monitored learners' performance at their schools. iv It was thus impossible to objectively assess the sustainability of the PfP programme, as noted by one of the business leaders: What is hard to prove is the sustainable improvement in results. School principals say that they are better off [but] there is need to show sustainable results.

Discussion
This article has illustrated how learning transfer occurs in the PfP leadership development programme, and the ensuing challenges. In relation to the fourth theme, we now focus on the content of programmes such as PfP, which aim at school improvement but are not focused on instructional practice. The following discussion comprises three subsections: how learning transfer takes place in cross-sector partnerships; the influence of design on the partnership's effectiveness; and the centrality of instruction in school leadership programmes and the challenge of credibility.

Transfer of Learning: Mixed Results
As previously noted, Fullan (2009) maintains that leadership development programmes should allow for learning that is embedded in the job, organisation, and system. The PfP programme helps to build capacity in school principals through jobembedded learning, actualised through peer learn-ing between the principal and the business leaderone of the achievements of the programme. This shows that peer-learning in social partnership interventions in aid of school improvement can provide for deep learning. Nevertheless, as Bush et al. (2011:36) note, the potential for deep learning in relationships providing individualised support may be restricted by the cost and the limited availability of mentors. Subsequently, massification of PfP partnerships would be hindered by the cost of the programme since participation in PfP is pegged to a fee. v The study also demonstrates that partnership interventions with models similar to that of PfP could have potential for organisation-embedded learning. This potential is however dependent on whether the partnerships are transformational or transactional in nature (Kirori, 2017). Partnerships with predominantly transformational content are likely to improve the school principals' leadership and to encourage greater cohesion and sharing practices among the staff at the school. This improvement indicates growth in collective capacity, which, Fullan (2009) argues, is necessary for organisational change. Organisation-embedded learning is likely to be hampered in transactionalfocused partnerships as the activities of these relationships hardly impact the way in which the school functions. The limited impact of transactional partnerships on school practice and thinking is supported by Butcher et al. (2011) who maintain that transactional relationships tend to meet specific individual objectives, hence their short-term impact. Figure 1 illustrates the types of partnerships formulated in the PfP process and their outcomes.
System-embedded learning appears to be inhibited in the PfP leadership development model by the competition noted in relation to the programme's group learning processes. The feelings of comparison, envy, and inferiority found within and between some partnerships involved in the PfP process point to the presence of inequalities among the participants in this programme, reflecting that neither the school principals nor the business leaders were homogenous groups. This finding is supported by critical action-learning theory which holds that people forming a learning group can begin to measure themselves against one another due to the dynamics arising from emotions, politics, and power among group members (Vince, 2012). The self-evaluation of members of PfP CoPs against others in the group can inhibit learning. The need to show tangible results also may have contributed to the negative competition noted, a limitation associated with social partnerships (Van Tulder et al., 2016). This suggests the need for a system-thinking approach which, as Fullan (2006:10) explains using an example, is when principals interacting across schools become equally concerned about the success of other schools in their network. Applying the foregoing to PfP means that the PfP dyads would not only want to see improvement in their partnership schools, but in all the schools in their CoP, due to the understanding that their school forms part of a larger system. vi Figure 1 Transformational and transactional PfP partnerships and their outcomes (Kirori, 2017:42)

Design Limitations on Programme Effectiveness
Although PfP literature states that it requires three to five years to bring about significant change in schools (Collins, 2015:82), PfP promotes itself as a one-year programme, primarily due to resource limitations. This research shows that at the end of the one-year programme, the PfP dyads, and particularly the business leaders, expect to see positive changes in all the four levels targeted by this process: headship, SMT, educators, and parents. However, the goal that these partnerships achieve successfully on participating for a year in the PfP programme is generally limited to improved headship.
Limited resources is a key challenge of social enterprises since the nature and scope of the social concerns targeted often lead to the establishment of large and complex institutions with many financing and staffing needs (Malunga et al., 2014;Zahra, Gedajlovic, Neubaum & Shulman, 2009). This is the case with PfP, with a staff complement of 50 and which admits funding for organisational sustainability among its major challenges (Partners for Possibility n.d. -b, n.d.-d). Related to the challenge of limited resources is the problem of legitimacy and the subsequent need to show results (Zahra et al., 2009) which, as illustrated, encourages the formation of transactional partnerships.

Centrality of Teaching and Learning and Evaluation for Credibility
Research on the PfP leadership development programme shows potential for building leadership capacity among school principals of underperform-ing schools. Seven of the nine partnerships involved in this study were found to engage in activities that were targeted at improving teaching and learning at their schools. Studies on wellfunctioning South African schools in challenging socio-economic contexts show that the leadership of these schools went beyond instruction by focusing on resource mobilisation through community involvement, in addition to creating a safe and secure environment (Christie et al., 2007;Ngcobo & Tikly, 2010;Zuze & Juan, 2018). However, since school principals' key role is to be leaders of learning (Bush & Heystek, 2006) by focusing on curriculum, pedagogy, and assessment (Lingard et al., 2003;Taylor, 2008), the fact that none of the participants interviewed were found to monitor the outcomes of their schools signals a serious drawback of the PfP programme. PfP's exclusive focus on leadership in general, explicitly excluding school leadership and management content in its programme design, is a major weakness of this model.
The importance of content in school leadership development programmes may be drawn from the evaluation of the ACE Programme (Bush & Glover, 2012). Although the effectiveness of this programme was hindered by the limited focus on leadership and management practice and by poor mentoring and networking practices (Bush et al., 2011), the ACE programme led to improved learner outcomes (Bush & Glover, 2012). The programme led to new leadership learning among the participants who worked on improving the teaching practice in their schools through monitoring the educators and class visits with constructive feedback, among others (Bush & Glover, 2012).
While the present study showed that most of the school principals interviewed engaged in activities with potential to improve teaching and learning, the evidence provided is subjective. The lack of objective evidence about the outcomes and impacts of social partnerships on societal problems affects the credibility of these processes (Van Tulder et al., 2016). Granted that the problem of attribution makes it difficult to ascertain the direct effects of leadership on learners' performance (Bush, 2009;Leithwood et al., 2004), the PfP dyads could be encouraged to monitor school performance via existing (statutory) data, in addition to tracing other intervening variables that impact learning directly, such as school and classroom conditions (Leithwood et al., 2004). The monitoring process ought to extend beyond the partnership year since, as Bush argues, "[i]t is widely recognised that the impact of interventions, such as a leadership programme, takes time" (2009:384).

Conclusion
The PfP programme contributes to our understanding of how learning transfer can occur in a leadership development programme through social partnerships between school principals and business leaders. Through the peer learning element in the PfP partnership model, principals are enabled to apply the learning gained to the specific contexts of their schools. Despite this capacity for transfer exhibited in PfP, monitoring and evaluation of the effectiveness of the programme was problematic. Negative competition due to inequalities in and between the PfP dyads further restricted this learning capacity by encouraging the formation of transactional partnerships. Although beneficial, the provision of personalised support to school principals is costly, limiting the scalability of this model to all underperforming schools in the country. This study also shows a limitation of school leadership development programmes involving non-educationalists, such as business leaders, that are aimed at school improvement and whose content is not centred on improving instructional practice. This indicates that the content of leadership development programmes aiming at school improvement, like PfP, should be scrutinised. The development of school leadership is a means to an end, but the ultimate end is improved classroom practice. A focus on corporate style and leadership values appears to have distracted PfP from this point. Leadership development is not a panacea to the problem of a weak educational system but is part of a wider multipronged approach to educational reform.