Principals’ perceptions of the motivation potential of performance agreements in underperforming schools

A formidable challenge most school leaders in South Africa face is to improve the academic results in state schools. In terms of their contracts, principals are accountable for the academic results as reflected in examination and test results for their schools. The National Department of Education (currently the Department of Basic Education) has made attempts to implement a performance agreement with principals and deputy principals, which would hold them directly and specifically accountable for the examination results. The article explores the proposed performance agreement and its potential influence on principals’ motivation to improve their own, and therefore also the teachers’ and learners’ academic performance. The focus group interviews conducted with principals and deputy principals indicate that principals do not want to be held accountable, because there are too many factors outside their control. They perceive a performance agreement of this kind as potentially demotivating because they do not feel they would be able to achieve the goals it sets.


Introduction, Background and Orientation to Assessment of Teachers' Performance
Education in South Africa, one of the emerging economies in the world, needs to be understood in its historical context. During the apartheid regime, the education system was characterised by huge disparities between the state schools provided by the Department of Education for the different racial groups. After the election of a democratic government in 1994, a new constitution was drawn up, which guaranteed all South Africans the right to basic education. A Plan of Action was drafted in 2003 by the then Department of Education (DoE) (since 2009 the Department of Basic Education (DBE)) to provide access to free and quality basic education for all (DoE, 2003). However, although South Africa currently spends about 5.5% of its Gross Domestic Product (GDP) on education, and has attempted to implement wide-ranging educational reform, the educational landscape continues to be marred by huge disparities. According to Kanjee and Sayed (2013) and Taylor (2009), most state schools in South Africa are underperforming. In this context, underperformance reflects the level attained on the Grade 12 (school leaving) examination results, as well as Grade Three and Six literacy and numeracy (litnum) results. According to district officials in the Western Cape as well as North West Province, there are no official documents which provide criteria for underperforming schools. A general guideline is a pass rate of less than 60% or 70% in the North West province. The term underperforming is therefore a vague but powerful term, because it labels schools and principals, although the criteria for its application to them are not clear (Heystek, pers. comm., 9 December 2014).
The DBE is committed to improving school performance. As has happened in many other emerging economies (Bush, 2012), it has identified leadership as a potential contributor to achieving the goal of quality education for all learners. The emphasis on the importance of leadership is reflected in the financial support the DBE provides, so that principals, deputy principals and heads of departments can register for the Advanced Diploma in Education Leadership (ACE). This ACE is offered at various universities in South Africa. DBE via the provincial departments of education pays their academic fees and travel costs and accommodation costs. The official vision for the programme was the development of a corps of education leaders, who apply critical understanding, values, knowledge and skills to school leadership and management in line with the vision of democratic transformation (DoE, Republic of South Africa, 2008). The pronounced expectation was that this program must be a 'quick fix' to the problems in education more broadly, for example, the underperforming schools as expressed by the National departmental representative in 2006 at the launch of the programme at the University of Stellenbosch, where I was present. The students for the ACE are selected from principals, deputy principals and departmental heads, since the aim of the programme is to develop principals and aspiring principals (throughout the rest of the article referred to as principals, because the heads of departments as well as deputies are included in the ACE).
In general, the DBE feels it can hold school principals accountable for the performance at their schools. They all sign contracts that explicitly refer to their accountability for performing the tasks and functions expected of principals in the South African Schools Act of 1996, and the Personnel Administration Measures (PAM) (DoE, 1999). In 2003, the Department of Education introduced the Integrated Quality Management System (IQMS) process to measure what teachers are achieving. All staff members at schools (including principals) are required to complete the IQMS forms annually as part of this process. However, this system does not hold teachers accountable for the quality or success of their performance (as measured by whether their learners pass or fail the subject they teach) (Mosoge & Pilane, 2014). No disciplinary or even specific developmental action is taken in cases where performance does not improve. The implementation of the IQMS has therefore not proved to be an effective means of motivating teachers to improve their performance (Mahlaela, 2012;Mbulawa, 2012). To address the issue of the accountability of all teachers, including principals, the DBE drafted a document, the Performance Agreement (PA -used here in reference to the proposed performance agreement document) in 2011, for principals and deputy principals. This PA would hold them accountable for the performance of the teachers, and therefore also the learners' examination and test results. As part of the process, necessary attempts were made by the employer (DBE) to negotiate the PA with the teacher unions, as representatives of the principals and deputies in the Education Labour Relations Council (ELRC). This initial document, drafted in June 2011 was not accepted by the teachers unions. Since then, the DBE has developed a more comprehensive document, the Quality Measurement System (QMS) (South African Council for Educators (SACE), 2013), which includes all teachers, such that it is not limited to principals and deputy principals. Nielsen (2014) points out that this trend towards stricter performance management and goal-orientated outcomes by governments is international, and not unique to South Africa. Although the role of the unions in schools makes the implementation of PAs a more complex process here, it still falls within the global trend to more visible, performance-driven assessment. The South African Democratic Teachers Union (SADTU), by far the biggest teachers' union and thus the most powerful roleplayer in negotiations on teachers' benefits, remains strongly opposed to the principles of accountability and potential disciplinary action in PAs. In its view, this is not in the interests of its members.
The initial proposed PA was never implemented, although the national Minister of Education (Gernetzky, 2012) publically announced the proposed PA. However, principals in the Western Cape Province did feel the potential threat of the proposed PA. As early as 2012, the Western Cape Province, one of the nine provinces of South Africa, wanted principals to sign the PA, even though the DBE and the unions had not yet reached agreement. Had principals signed this, it would have meant that disciplinary steps could be taken against them if the examination results at their schools did not improve. Later in this article, the focus group interviews and the analysis will be used to explore the principals' responses to the possible implementation of the PA.
The project described in this article stems from a desire to explore principals' responses to what they had heard about the PA. This article explores their responses in order to raise issues about such agreements, and their possible effect on school leaders' level of motivation in South Africa. The question framing this research is: to what extent do principals perceive the PAs as motivational in the process of improving quality education? The question is premised on the notion that motivation is what moves one to participate in an activity, and is what affects one's desire to continue with the activity (Enhanced Motivation, 2004). The research draws on motivational theories to explore whether a PA could be a means of motivating principals to improve the quality of education in most underperforming schools. It should be noted that the performance management policy has not been implemented, and that the participants in the research did not have any firsthand knowledge of it. The research arises from my personal experience with principals during discussions in different forums. The anecdotal evidence I gleaned suggested that these school leaders had reacted negatively when they heard that a performance management process might be introduced. This will be discussed in more detail later in this article.
It cannot be said that the DBE assumed that if the PA were accepted, improved performance would necessarily result. It is more accurate to speak of the DBE's expectation (Fredericks, 2012;Gernetzky, 2012). This point will be explored in what follows.

Performance Agreement as Potential Threat
Motivated people are likely to perform better over a longer period, than people whose actions result from an agreement which is potentially threatening to them (Latham, 2007;Owens & Valesky, 2011). The proposed performance agreement (QMS) is an example of such a potentially threatening agreement. The rationale for this assertion is that accountability requires measureable goals to be met; and therefore, these goal-driven agreements or control processes have the potential to motivate people to improve their performance level (Ryan, 2012). However, for this to be true, those being assessed or appraised must feel they have a realistic possibility of achieving these goals (Reeve, 2009). Although there is not really an official crisis of trust in South African education system, Saunders, Dietz and Thornhill (2014) have argued that many individual managers are struggling to manage positions of trust and distrust. The notion that a level of trust is important, resonates with McGregor's X and Y approaches to people and management (Kressler, 2003); if leaders do not trust those they lead, they tend to implement stricter control criteria and actions. This may explain why the national and provincial ministers of education and the directors generally feel the need for mechanisms to ascertain whether their plans to improve the quality of education are achieving the desired goals and standards.
However, as Sun and Van Ryzin (2014) indicate, goals must be measurable and realistic. The potential problem with the goals of the PA is that although the examination results are measurable, the input of teachers and principals are difficult to measure. This compromises the reliability of the PA. A performance agreement of this kind thus focuses on the task, rather than the person (Crawford, 2009;English, 2008). Chen, Eberly, Chiang, Farh and Cheng (2014) refer to a paternalistic leadership (or rather governance) style, which may be said to typify the DBE's over protective or "I know what is best for you" attitude towards principals and schools. The performance goals are predetermined and externally developed; in this case, the examination pass rates or litnum levels have to be achieved, regardless of the context and circumstances of the contracted principal. This type of 'paternalistic' PA may thus be described as a managerial and task-orientated approach to school improvement.
DeNisi (2011) points out that the problem does not lie with the criteria or the content or the scale, but with the implied threat. When humans feel threatened, they revert to a defensive survival mode, and may lose the drive to develop the new ideas which are necessary to lead an underperforming school to sustainable improvement (Hallowell, 2011;Kressler, 2003). Mosoge and Pilane (2014) indicate that performance management without provision for development is unlikely to motivate teachers to enhance their performance. Underlying the research on principals' attitudes to the draft PA for principals, was the need to explore whether this would have a motivational or demotivational effect. In order for a PA to be motivating, the employees concerned need to feel that they can achieve the goals set out for them (Ryan, 2012). A determining factor is whether or not they feel that they have sufficient control over the contextual factors, intrinsic and extrinsic. For this reason, it is argued here that PAs or performance contracts should not have deterrent or other negative effects.
These insights have direct implications for the way the performance agreements are implemented or applied, as well as for the importance of making the purpose clear. A performance agreement may become a positivist, bureaucratic and inhuman process if employees are assessed against examination result targets, without regard to the context of the school and community. Performance agreements and appraisal are associated with goal path and expectation motivational theories (Pintrich & Schunk, 2002;Sheppard, Canning, Tuchinsky & Campbell, 2006). It is not clear from the draft of the PA whether or not the intended purpose was to motivate principals to in-crease their performance, but the Minister's statement indicated it was rather a managerialistic approach to keeping people accountable, rather than to motivating them in their performance.

Performance Agreements and Motivation
DeNisi (2011) emphasises that for assessment to be a performance-changing instrument, the process and criteria and purpose of the appraisal must be fair, valid and reliable. Otherwise, the appraisee will change his or her behaviour only to a limited extent, if at all. The performance agreement, including the document used, for example the IQMS or proposed PA, has to be seen as legitimate and acceptable by the appraisee. A performance agreement is not inherently problematic if the agreement is properly negotiated, and accepted by all involved. Nielsen (2014) emphasises the positive potential influence of PAs, when he states that when performance feedback shows that performance falls short of aspirations, this provides a signal to the organisation that some sort of change is required. Chen et al. (2014) in China, as well as Moreland (2009) in England and Wales, discuss the motivational factor of performance appraisal, and indicate that the level of trust between the people involved in the PAs is an important factor in making performance appraisal a positive process. What is most important is not the control, assessment, appraisal, inspection, or whatever the actual process is labelled, but the motive behind the control functions it enacts. A school-related PA must be more than an accountability tool; it must also make provision for the development and support of the principal (Cardno, 2012). McGregor's X and Y approaches to people and management (Kressler, 2003;Maslow, 1998) offers a way of understanding the philosophic base that underlies PAs or appraisals. Although, like most social theories, McGregor's theory has been subjected to a great deal of criticism, it remains useful to the current argument, where his basic assumption is that there are two contrasting views that one may apply to the process: one is that people are inherently not driven to work hard (theory X); and the other is that people are highly motivated and dedicated (theory Y). Theory X implies a lack of trust, and hence, a "must know and control" attitude, which centralises PAs as vital. However, theory Y leads to a very different attitude. Principals, for instance, would be seen as professionals who do not need to be controlled and managed by means of PAs in the same way as workers involved in forms of industry. Subscribers to this view would see intrinsic motivation as preferable, because it is more likely to provide a longer-term solution to the need to provide quality education than a threatening control process (James, 2005;Latham, 2007). Moreland (2009) concurs that in England and Wales, a positive attitude to the way teachers are treated makes the performance process a positive experience. This is very different from the South African context, where a theory X approach is taken towards principals.
The job description in the contracts principals sign at state schools explicitly states that they are accountable for the quality of education in their schools. Knapp and Feldman (2011) refer to this as internal accountability or internal motivation. Externally driven accountability, however, is more of a matter of compulsion, where the person involved meets requirements due to a sense of legal obligation. Expressed in simple terms, people may feel that I do not really believe in performance process, but because legislation requires that I must do a thing, I will therefore (or, I must) do it.
Another problem with a performance agreement is that it is applied according to a one-sizefits-all policy. According to Sun and Van Ryzin (2014), there is sufficient evidence to show that an individual's personality, attitude and values, as well as contextual factors like teachers and the socioeconomic environment, can make a difference to the approach taken to implementation, and the consequent success or failure of any performance system. Moreland (2009) stresses that when performance management is experienced, as at least partly focused on individual as well as organisational management and goals, the participants feel positive about it. Sun and Van Ryzin (2014) indicate that in the United States of America (USA), performance management is correctly implemented, which includes a positive relationship between the leader (or employer) and employee. There is, therefore, the possibility that principals in South Africa could be motivated by a PA that is appropriately managed and implemented. However, as Sun and Van Ryzin (2014) point out, performancerelated data should not be an end in itself: the process should allow principals to evaluate how much progress their leadership has achieved over time, help them to make better decisions, and adapt to new priorities or needs, so as to bring about the desired improvement.
Martin and Dowson (2009) define motivation as a set of interrelated beliefs and emotions that influence and direct behaviour. For this reason, PAs may be expected to fulfil a few basic functions: first, to ensure that the work being done meets the minimum requirement of quality or predetermined goals. Secondly, such an agreement may serve as a motivational factor to workers, because they may earn more money, or some remuneration or, preferably, feel inherently motivated if they achieve the agreed standards or even surpass the required standards. Martin and Dowson (2009) argue that relationships affect achievement motivation by directly influencing the constituent beliefs and emotions of motivation. Interpersonal and intrapersonal relationships will negatively influence motivational levels if principals are made to sign agree-ments. However, a positive self-relationship, but also positive relationships with superiors and followers, may be significant motivational factors that minimise the potential demotivational factors in these agreements. Thirdly, PAs may also serve as a smoke screen for employees to eradicate unproductive workers, as one of the participants in this research indicates.
A further problem is that there are too many factors, both external (environment or fellow human beings) and internal (personality, values), that could influence the motivational levels of a person; or constrain the ability of a person (such as principal) to motivate all the teachers in a school simultaneously and to the same level, to get them to improve their performance significantly (Kressler, 2003;Sheppard et al., 2006). Each individual, as a cognitive, emotional and values-driven entity, thus needs a specific kind of motivation to make him or her 'move' to achieve or do what is expected. External factors, such as the circumstances, also determine the kind and level of motivation needed (Pintrich & Schunk, 2002). These include the school's socio-economic environment, or the infrastructure of schools, which often lie outside a principal's sphere of influence. These factors may significantly affect the academic achievements of learners in the school (Reeve, 2009;Ryan, 2012). If the principals feel that there are external factors that are beyond their control, which will prevent them from achieving certain performance objectives, they may not see a PA as an acceptable accountability instrument.

Research Design
The research employed a constructivist and interpretative perspective. In this qualitative approach, the participants construct their own reality, perceptions and interpretations of what is happening, what they are experiencing, and what they know. This construction is influenced by the participants' own personal reality, context and personality (Cohen, Manion & Morrison, 2000). They constructed their understanding of the PA from what they had heard from different sources, exempli gratia (e.g.) newspapers, friends, unions and departments of education. In this research, little of the information had been gained from viewing the proposed document itself, but is rather based on second-hand information or discussion. This does not make the perceptions linked to their lived experiences of the participants less valid. Lived experience refers to how a person experiences a situation, which directly or indirectly, consciously or unconsciously, influences their perceptions about the specific issue, which in this case is the proposed PA (Savin-Baden & Major, 2013). The participant's own constructed reality is how they "create forecast events and rehearse situations before the actual event occurs" (Cohen et al., 2000:337), which again, in this case, is the potential influence on their motivation and performance level. Although the performance appraisal was not implemented, the participants experienced what may happen if the PA may be implemented. The participants were confronted by the possibility to sign a PA which might have a potentially negative influence on their work as they heard the Minister of Education said.
Most, if not all of the participants came from underperforming or low-performing schools, as identified by the Western Cape Department of Education. Although most of the principals were working in the Western Cape, a province which generally performs very well (DBE, Republic of South Africa, 2013); the schools are located in rural and poor socio economic and previously oppressed communities, which historically tend to form the backdrop for most of the underperforming schools in South Africa (DBE, Republic of South Africa, 2011). A convenient sampling strategy was applied to select the participants. The sample included the school leaders, who would have been directly influenced by the PA if it had been implemented, namely the principals, deputy principals and heads of departments. The purposive and convenient sampling was not conducted to determine if these groups of participants had differing perspectives on the PA, but rather to ascertain a perspective from the leadership group; hence they were included in the groups as equals, and not as representatives of the different levels of leadership in schools.
A convenient sampling method was applied, because the participants were conveniently available. It was also purposive, because the participants in the Western Cape already had information that the performance would be implemented in their province, even while the national agreement was not yet signed. Group D was purposively selected, because they are from another province and also had postgraduate qualifications in education leadership. The purpose was to determine if they had similar perceptions about the proposed PA. They were also conveniently available, since they were the part time lecturers for the ACE programme. The participants in groups A, B and C (see Table 1) were selected from principals and deputy principals in the process of gaining a leadership qualification at a university. The members of Group D were part-time lecturers presenting courses in a leadership qualification, but they were all full time principals and deputy principals. They were selected because of their position as school leaders, and not as lecturers. The fact that they already had either a master's or doctoral degree made them a unique group of principals and deputies, who could offer a different perspective.
It could be argued that groups A, B and C constitute a captive audience. However, these participants, who voluntarily registered for the academic programme, can also be described as "knowledgeable". Their decision to register for the programme is an indication of their own need for development and their potential interest in managerial issues with regard to their own and their school's performance.
The participants explained how they saw, felt, and experienced the announcement of the proposed PA, and what they expected the implications to be. From one perspective, their views may be seen as biased. However, this is not an issue here, since the research is concerned to reflect their reality at that time (De Vos, Strydom, Fouché & Delport, 2011). Eight students (six principals and two deputy principals) from a large town or its adjacent rural areas, 200 kilometres from Cape Town and about 200 kilometres from where Group A live. Group C Five students (three principals and two deputy principals), who live in a small, rural town and in its rural environs, more or less midway between the two other towns, and about 180 kilometres from Cape Town. Group D Three part-time lecturers (principals) who live to the north of the country, about 1,800 kilometres from Cape Town. All had the minimum qualification of a master's degree. Group E Nine principals from the Executive of South African Principals Association (SAPA) Western Cape.
The participants from Group A, B, C and E were from the Western Cape, which is regarded as one of the two best academically performing provinces (DBE, Republic of South Africa, 2013). The members of Group D were selected on the basis of convenience. I had the opportunity to participate in the presentation of the leadership programme at another venue in the country. These principals and deputy principals were part-time lecturers, who helped to present the programme. They were different from the other groups in that they had higher academic qualifications; all of them had at least a master's degree, and some had a PhD degree.
The participants of group A and B were selected from the 200 students attending the academic programme; those in Group C were selected from amongst 25 students attending the programme at this venue; and those in Group D were selected from amongst 24 part-time lecturers. They all volunteered to participate after they had been informed as to the nature of the research project. The interviews were conducted after the classes ended at a convenient venue for the participants. Since a large number of students had volunteered to participate, I decided to work with the participants who arrived at the agreed venue the next day.
The ethical requirement to explain the procedures in detail was met. In all the cases (except in the case of the SAPA principals), I was either their lecturer in the programme, or their former lecturer, at different levels in Education Management and Leadership programmes. I therefore acknowledge that they may have been influenced by my presence, or by the ideas I had presented to them during the programme. I attempted to ensure sufficient validity of the information they provided, by emphasising that I wanted them to give their own opinions. I also ensured that I did not ask any leading questions, and that I even challenged some of their statements when it seemed that the principals might be pandering.
Focus group interviews were the means of gathering the data, with semi-structured interviews as the data gathering tool. This gave the participants the freedom to discuss the issues, but all of the groups were given a basic predetermined set of questions as a discursive guideline. Stewart, Shamdasani and Rook (2007) indicate that focus group interviews are preferable to individual interviews, because they give the participants the confidence to speak their minds openly. They are also able to draw on each other's ideas, making the discussion a building process of gathering data. De Vos et al. (2011) indicate that participants can also support each other, and thus, can emphasise the importance of a particular point; but they could also disagree with each other, which is an indication that the issue was as contentious as the project assumed it to be. The focus group is also an ideal forum in which to construct reality, since differing opinions were sometimes expressed, allowing a deeper understanding of the theme to be reached. During the focus group interviews in this research, I ensured that a single person did not come to dominate the discussions, since that could have prevented all the participants from voicing their views, and thus have influenced the data collected. Certain participants were specifically prompted to give their opinions, so as to allow all voices to be heard. The discussion during the focus group interviews was thus a developmental process, capturing a communal construction of reality with regard to the proposed PA document (Stewart et al., 2007).
The interpretative data analysis was used on that data which captured the direct voices of the participants, reflecting their construction of reality. This reality was then interpreted against the theory and existing knowledge related to the issues addressed in the project. The quotations in this article were chosen because they represent the con-struction of the group, rather than a single person's ideas.
In the following sections, I draw on the data to discuss the issues addressed by the participants.

Principal's Knowledge and Perceptions of the Proposed Performance Appraisal
The official negotiations on the proposed PA took place in the ELRC between the DBE and the unions. From the comments of the principals, it seems that neither the DBE nor the unions communicated with the principals about the PA and its possible implications. Although the majority of the principals had heard something about the PA, few of them had seen the draft PA, or been given any official information about it. In the interviews, the participants expressed their particular concern that the Western Cape wanted to implement the PA before there had been public or interest group meetings to clarify all the issues it raised. The participants in the SAPA (Group E) had had the opportunity to discuss the document in detail, because they had been asked to give official feedback to the Western Cape Department of Education (WCED), as well as to the national DBE. The SAPA participants said they did not have a principled objection to such a document or to the process involved, but they felt that there were many problems with the current form of the document. They too expressed their concern that the WCED wanted to pilot the PA, with the blessing of the DBE, although the official agreement had not been signed in the ELRC. In fact, agreement had been reached in the ELRC in December 2011 that the document had to be revised by the DBE before the unions, especially SADTU, would discuss the issue again.
One of the principals in Group B, who is actively involved in SADTU, indicated that he had been given information about the PA by the union. A few other principals intimated, after some hesitation and a short discussion among the members of the group, that the PA had been mentioned in passing at information sessions organised by the local districts. In their view, neither the DBE nor the WCED seemed to have a structured plan to inform principals about, or to promote the implementation of the PA. Most principals described their knowledge about the PA as information that they obtained 'via the grapevine'. A principal (an active union member) in Group D mentioned that at a union meeting "they just touched on the document" about the PA, but there had not been a specific meeting to inform them or discuss the possible implications. Another principal in this group (not actively involved with any union but with a Philosophiae Doctor (PhD) in Education management and policy), first read about it in a newspaper, and then heard about it in her district.
A Group D principal (active union member) summed up how the group had felt when they heard that the PA might be implemented: The fear and uncertainty expressed above are discouraging to principals, where the PA can be seen to potentially undermine the basic human needs of safety, as indicated by Maslow, because they may lose their work (Aycan, Kanugo & Mendonça, 2014). Most of the participants felt that there was limited motivational value and power in the PA. One of the principals in Group D explained their feelings about the PA as follows: It [the PA] is a threat, because we think it will be the same as with the directors -if you do not perform, you will be recalled (25:10). [In the South African context 'recall' means that you are removed from a post and either employed at another school or in a district or circuit office].
Group B and C commented that the PA was not motivational, because teachers might be empowered to "sabotage" the principal. In this kind of situation, teachers might purposefully undermine the principal's leadership and efforts to improve the school, which could ultimately even lead to the dismissal or replacement of the principal. That is another indication that principals felt factors outside their control that could have a negative influence on their future careers if they signed the PA.

The Performance Agreement as a Potential Motivational and Accountability Tool
Most participants did not perceive the PA as potentially motivating. They mentioned external factors outside their control, which made it impossible for them to feel they could take responsibility for the personal performance of each teacher, and by extension, for the examination results of the school. The participants mentioned all the usual hindrances to performance, such as lack of facilities, teachers' (possibly limited) qualifications, the learners' level of ability, and (possibly detrimental) socio -economic conditions. Significantly, most were adamant that external motivational factors -for example, an increase in their salary -were not necessarily important motivational factors. In their view, for money to be a motivational factor; all the teachers at the school would have to share in any financial benefit. That would have to go along with the teachers' willingness to share in the consequences should the school not perform. The principals were under the impression that the national department of education had not made the additional funds necessary to reward all.
The participants also mentioned other external motivational factors that would have a motivational effect on them, and would empower them to encourage teachers to embrace development, such as better facilities at the school, and better support from the departmental officials at the local level. This would be a more effective means of getting them to work harder than a PA, aimed at forcing them to work harder. They also mentioned the important role parental support could play. External motivational factors can therefore not be discarded as potential shorter-term motivational factors.
Most of the participants did not believe that the PA would be a powerful means of inspiring the principal, and most of the staff members, to improve their own performance. On the contrary, the PA was perceived as an action which might have very different consequences from those envisaged by the different departments of education initiating this process. The principal's internal motivation (Knapp & Feldman, 2011) were not addressed, because they said that the PA would not provide them with any power to use, if and when teachers did not meet the agreed levels of performance. The Group D principals also said that their knowledge did not give them total power, but noted that it did help them.
One of the most important factors which militated against the principals' feeling that they should be held accountable for the quality of the performance of teachers and the academic achievement of the learners, was the teachers' unions. The Group D principals put it this way: We as principals are exposed (by the PA), because you know, the teachers are protected by the unions; we may be recalled, but because of the union's power, an underperforming teacher will [stick] with you forever. [Author: underperforming teachers will rarely or never be disciplined or have formal action taken against them]. iii

Concluding Comments
As an emerging economy, South Africa needs to raise the quality and level of the education provided to its future citizens. In order to manage this process, the DBE seems to use a neoliberal and managerial approach, adopted from countries in the developed world (Apple, 1999), rather than a progressive management process that would em-power local school leaders. In particular, it has not taken sufficient account of the contextual constraints in the school environment.
In general, the participants' initial reaction to what they had heard about the PA was that it represented a threat, rather than a motivation. Their perceptions were based on the view that the locus of control was not in their hands. They saw themselves as unable to achieve the required goals, which would threaten their work and position (Latham, 2007;Ryan, 2012). The potential motivational influence of goals as motivation (Ryan, 2012) was thus negated by the principals' sense that they would be unable to achieve these goals, due to constraints over which they had no control.
Negative emotions have a strong influence on the motivational level (Martin & Dowson, 2009). In this case, the proposed PA offered few, if any, incentives for principals. According to the 2011 draft performance agreement document, there was no offer of a salary increase, or other reward, providing guidelines for performance (SACE, 2011). Therefore, they viewed it as a control mechanism, that was likely to have a negative effect on them, and which was not likely to result in a sustainable improvement of education.
It is important to note that the responses of most participants was based on informal or unofficial information, such as newspaper articles, information, or comments given on the radio, or the informal discussions of colleagues (grapevine). This speaks directly to the expectations of motivational theory. The expectations from the DBE that principals must under all circumstances improve the quality of education and if they cannot do this, ought to be accountable for the low performance was too high. When a goal or expectation is too high and it does not seem achievable for the participants, it loses its motivational value (Pintrich & Schunk, 2002). Outcomes expectations (Reeve, 2009) such as the proposed PA, can be valuable as a motivational factor if the person concerned experiences the expectation as achievable, and if the effort involved to attempt to meet this will result in valuable outcomes aligned with the level of effort. The perceptions the participants cited in this research were based on the Minister's comments, as well as their experience of the redeployment of underperforming district officials. When the Minster made the comment, their perception was that the actions of the Minister were an abuse of power (Guthrie & Schuermann, 2010) or the use of her power to manipulate principals (Owens & Valesky, 2011), because she seemed to have chosen to threaten the principals. They saw this as demotivational. Any potential motivational gains from the performance appraisal system were therefore negated.
What seems plain is that the possibility of identifying realistic goals for individual schools based on their local context indicated in the 2011 draft have been ignored, in favour of the "one-sizefits-all" QMS, which is currently in the development phase. Ryan (2012) refers to this influence on the principals as a blow to their ego. The principals felt they had to exercise self-control to work in the difficult circumstances and to motivate themselves and the teachers they oversee. Once they felt that their self-control was not effective, their ego was depleted and this effect is against any potential motivational value of PAs which further erodes the control principals have on the situation. Although the 2013 document might offer a more subjective assessment process, and be burdensome to administer, it might be a better means of determining both personal and school goals, and thus be more motivational.
Not all principals viewed the agreement negatively. This is strong indication that the PA need not be interpreted as inherently negative, or bad. Reeve (2009) refers to this as the contextual influence associated with the perceived ability to control the situation. Control, or in turn, a lack of thereof, can ultimately have a direct effect on performance. Most of the principals expressing this were from Group E, in which the majority of principals were at better performing schools. There were also a few principals in Group B, who expressed the view that although they were not performing well enough, they did have control over the circumstances in their schools. These principals did not see a PA as negative per se. They possess the personal and professional fortitude to face a potentially threating situation (Reeve, 2009;Ryan, 2012). However, the attitude of most of the principals to the proposed policy seems to be an indication of a certain corrosive energy, which may ultimately erode the positive growth in the organization (Bruch & Vogel, 2011). It seems that the key issue for this group is not the PA itself, but the many factors outside their control that make them feel they cannot be held accountable for examination results. The significant influence of the unions on the principal's experience of low control is explained by Sheldon (2010). The personal and emotional depletion over a sustained period of time is caused by being in a situation in which the problem of external control (in this case by the unions), is unlikely to be resolved in the near future. The influence of the unions makes it very difficult for principals to effect an improvement in the teachers' performance, or in the learners' academic achievements. This does not reflect well on the motives and actions of the unions in a democratic and open society.
Although this research is limited in scope, it could prove valuable when PAs with principals are being considered. The DBE should take account of the union's role as a powerful external influence on the principal's ability to perform, and on account-ability when drafting a policy such as performance criteria for principals. From this research, it seems that PAs are unlikely to motivate principals to improve their own, as well as the performance of teachers and learners at their schools, if full account is not taken of these external contextual constraints. The powerful influence of SADTU on the potential influence of principals to improve the quality of education in their schools must first be negotiated at national level so as to alleviate the pressure from the union on the local principal.

Notes
i.
Verbatim quotation was edited for the publication. ii.
Verbatim quotation was edited for the publication. iii.
Verbatim quotation was edited for the publication.