South African school peace programmes make modest gains but stall on punitive discipline and teacher training

Introduction

School violence in South Africa tends to get talked about as a discipline problem or a crime problem, and a lot less often as something education itself is supposed to deal with. Peace education takes the second view. It treats conflict skills, respect and non-violent ways of settling a dispute as things a school teaches, not just things it punishes people for lacking. Given the history, the idea has obvious appeal here. Baker et al. (2021) describe how child safeguarding policies introduced in Catholic schools from 2012 uncovered deep unhappiness and showed that corporal punishment, abolished in 1996, was still being used. Du Plessis and Mestry (2024) start from the view that violence, crime and abuse against learners have become a national concern, and the rural principals they interviewed describe a situation that matches it.

This paper argues that peace education in South African schools has produced modest and uneven gains (Baker et al., 2021), and that those gains depend on conditions most schools simply do not have (Ndwandwe, 2024). The programmes studied so far, which include learner-run peace clubs, a locally designed action research strategy and a restorative justice programme, all report slow improvement. But they run next to a punitive disciplinary default and a teaching workforce that appears thinly prepared and, in some cases, hurt by the violence it is asked to reduce (Ndwandwe, 2024; Mosito and Sitoyi, 2024). My position is that peace education should be judged as part of how a school disciplines learners and trains its teachers, not as an add-on club. The paper looks first at what the evidence says about violence and bullying, then at the programmes and what they achieved, then at discipline and restorative practice, and then at teachers, before drawing the threads together.

What the evidence on violence and bullying shows

The best national picture comes from a source that was not designed as a violence study at all. Juan et al. (2025) used self-reported data from 11 891 Grade 5 learners who took part in the 2019 Trends in International Mathematics and Science Study, and looked at how much bullying there is and who is most at risk of being a victim. Their results point to clear age and gender differences, with boys and older learners more likely to be victims. They also note that little had been known about which South African learners are exposed, which suggests that school safety work has often had little data behind it. A nationally representative sample is rare in this field, so the finding that boys are more exposed deserves to be taken seriously, even though self-reports in a Grade 5 classroom have their own limits (Juan et al., 2025).

A very different kind of study complicates that picture in a useful way. Mayeza and Bhana (2021) held focus group discussions with young boys in a township primary school and asked them what being a bully actually meant. What came back was messier than the usual victim and bully split. The boys described using violence to hold power over girls, and they also described being victims of violence which they said girls carried out. The authors read this through critical masculinity studies and tie it to the wider social and economic conditions that produce gender violence. Put next to the survey result, it suggests that a headline like ‘boys are more often victims’ says little about the power relations underneath, and that programmes aimed at an imagined bully may miss the point (Mayeza and Bhana, 2021; Juan et al., 2025).

The adults in the building matter too. Du Plessis and Mestry (2024) interviewed principals and school governing bodies at rural schools and observed those schools. They report that unprofessional teacher conduct, teacher absenteeism, a weak criminal justice system, dysfunctional families, missing school leadership and a precarious school environment all feed violence, and that poorly managed schools are more likely to experience more of it. Schools do write codes of conduct, but management at most of them seems unable to apply the codes effectively. Mosito and Sitoyi (2024) then show the other side of the same coin by following two teachers through narratives of school violence, and reporting what they call narratives of disillusionment. The sample is tiny, yet it fits the wider point in their article that many teachers are reportedly leaving the profession because they fear violence or have been attacked. So teachers appear in this literature as part of the problem, as victims, and as the people expected to fix it, all at once (Du Plessis and Mestry, 2024; Mosito and Sitoyi, 2024).

The quality of this evidence is uneven, and that needs saying plainly. One source is a national survey of nearly twelve thousand learners (Juan et al., 2025). The others are interview and focus group studies of a few schools or a few people (Mosito and Sitoyi, 2024; Du Plessis and Mestry, 2024). They say a lot about how violence is experienced and explained, and almost nothing about how it changes over time. Any claim that peace education ‘works’ has to be read against that background, because the baseline itself is shaky.

What the programmes have achieved

Ndwandwe (2024) offers the closest look at a school peace programme operating in practice. The study evaluated the Quaker Peace Centre programme in three secondary schools in the Western Cape, drawing on interviews with the programme manager and principals and on focus groups with teachers and learners. The programme was carried out with modest success outside the formal curriculum, mainly through a multimedia approach built around learner-run peace clubs and anti-bullying campaigns. Then come the barriers, and they are quite a list. Financial constraints, too little time given to peace education, poor parental involvement, overburdened teachers and negative attitudes and societal influences were all named. The author recommends teacher training in peace education, a place for it in the mainstream curriculum, wider stakeholder participation and proper funding (Ndwandwe, 2024).

Ngidi and Kaye (2022) took a different route, which is probably why their study is worth reading beside Ndwandwe’s. They ran action research at one school in Umlazi, KwaZulu-Natal, working in four stages from initial data collection to evaluation. An action team of teachers, community members, parents and learners built a strategy called ‘We Care’. The first round of data had already painted the school as unsafe, with violence connected to substance abuse, theft, vandalism, physical and sexual violence, cyberbullying, gender-based violence, religious discrimination and gambling. The strategy sorted troubling behaviours into categories and supported at-risk learners, victims and perpetrators alike. A preliminary evaluation a year on reported more capacity to deal with these problems and less violent behaviour at the school (Ngidi and Kaye, 2022). Having the community inside the design looks like a real strength, particularly when set against a programme brought in from outside (Ndwandwe, 2024).

The third programme was built by a network of schools, not by one. Baker et al. (2021) describe the Building Peaceful Schools programme that the Catholic Institute of Education developed in response to teachers insisting that corporal punishment was the only discipline that worked. It began in 2013 and moved through three workshops on peacebuilding, conflict management and restorative justice, with a school climate survey of pupils added in 2016 so that change could be measured. Their reading of the results is that improvement was slow but steady. That phrase matters. It is a more believable claim than anything dramatic, and it echoes the modest success reported by Ndwandwe (2024) and the preliminary gains reported by Ngidi and Kaye (2022).

Taken together, the three accounts (Baker et al., 2021; Ndwandwe, 2024; Ngidi and Kaye, 2022) agree on direction and disagree about almost everything else. The programmes differ in who designs them, how long they run and what counts as success. The designs, which are a three-school evaluation, a one-school action research project and a case study with a climate survey, are good at showing what happens when a programme meets a school. They are less good at showing that the programme caused the change, since none of them is built around comparison schools. So I would describe the evidence as encouraging and thin at once. A sensible reading is that peace education is plausible and worth continuing, while the size of its effect on violence is still unknown (Baker et al., 2021; Ndwandwe, 2024).

Discipline, restorative practice and the punitive default

Peace programmes land in schools that already have a way of dealing with misbehaviour, and in many of them it is punishment. Baker et al. (2021) found that teachers were adamant that corporal punishment was the only means of discipline that worked, and that most school discipline processes were punitive and broke down relationships. Obadire and Sinthumule (2021) interviewed experienced educators from disciplinary committees, learner representative councils and school principals at four schools, asking how learners are being disciplined now that corporal punishment is gone. They found that the causes of disciplinary problems varied from school to school, and that effective school management sat at the heart of both discipline and the general academic performance of the school. That finding is a useful brake on any idea that a single peace curriculum could fit every school (Obadire and Sinthumule, 2021).

The sharpest illustration of what a punitive default can do is the case discussed by Ally et al. (2021). In S v L M & Others, four learners who tested positive for drug use at school were channelled into the criminal justice system and detained for months under compulsory residence orders at child and youth care facilities. The authors point out that the law explicitly prohibits referring children to the criminal justice system after a school-administered drug test. So what the case exposes is a failure by school officials, prosecutors and magistrates to follow legislation that already existed. Ally et al. (2021) use it to raise red flags about exclusionary disciplinary mechanisms in general, even lawful ones, and to argue for restorative and preventative approaches. It is a long way from a peace club to a compulsory residence order, but that distance is exactly the point: the same school can offer both.

There is a gap between what is written and what is done. Du Plessis and Mestry (2024) report that schools do draft learner codes of conduct, yet management at most of them cannot apply the codes effectively. They suggest that school management teams should lead on curbing learner violence by enforcing the codes strictly and by training teachers to manage discipline. That recommendation pulls in a different direction from restorative practice, and I think the tension is real rather than cosmetic. A strict code enforced by a stretched management team can easily become the exclusionary approach Ally et al. (2021) warn about, while a restorative approach needs time, trained adults and trust, which Ndwandwe (2024) found in short supply.

The claims for restorative practice run ahead of the evidence. Its logic is persuasive, since discipline that repairs relationships fits the stated aim of peace education much better than suspension or the cane does. The one programme here that was built around it, Building Peaceful Schools, reports slow but steady improvement and little more (Baker et al., 2021). What I would take from this is a point about design. A few workshops on restorative justice, attached to a school whose discipline system still runs on punishment, is bound to struggle. If the approach is to count, it has to become the school’s discipline system and not an item on its calendar (Ally et al., 2021; Baker et al., 2021).

Teacher preparedness and wellbeing

Almost every study above puts teachers at the centre, and the evidence on their preparation is not encouraging. Mawila et al. (2023) worked with 25 second-year Bachelor of Education Foundation Phase student teachers at a university in Johannesburg. In groups of five, the students had to run an anti-bullying campaign for Grade R learners, and each then wrote a reflective essay on how prepared they felt. The study concluded that the students were not prepared to raise awareness about bullying, pointing to inappropriate activities, poor classroom management, poor planning and discord within the groups. The authors want tertiary institutions to strengthen this part of teacher training. The sample is small and from one university, so it proves nothing about the whole country, but it does match Ndwandwe’s (2024) recommendation that teachers be trained in peace education (Mawila et al., 2023).

Preparedness is only half of it. Ndwandwe (2024) lists overburdened teachers among the barriers to peace education, and Mosito and Sitoyi (2024) describe teachers worn down by the violence around them. A teacher who is frightened, tired or disillusioned is not well placed to run a peace club after hours, let alone to model the calm conflict handling that the programmes hope to teach. That is a hard thing to say in a country that needs more teachers. Still, it suggests that support for teachers’ own safety and wellbeing is part of the peace education agenda, not a separate matter handled by someone else (Mosito and Sitoyi, 2024; Ndwandwe, 2024).

There is also the uncomfortable fact that some teachers are part of the violence. Du Plessis and Mestry (2024) list unprofessional conduct and absenteeism among the causes principals and governing bodies identify, and Baker et al. (2021) show how deeply the belief in corporal punishment ran among teachers even after its abolition. Training that treats teachers only as deliverers of a peace curriculum will miss this. A more honest version would ask teachers to examine their own discipline habits, which is what the Building Peaceful Schools workshops seem to have tried to do (Baker et al., 2021; Du Plessis and Mestry, 2024).

Social cohesion and what a credible approach needs

Peace education in South Africa is often justified by the hope that schools can help build a more cohesive society after apartheid. That hope is visible in the way Baker et al. (2021) open their account, with oppression and dehumanisation as the backdrop to the work. Yet the studies read for this paper mostly measure something narrower: violent behaviour, school climate, bullying prevalence, or participants’ views about barriers. None of them directly tests whether learners from different racial, language or national backgrounds relate better to each other after a programme. The gender findings of Mayeza and Bhana (2021) show how cohesion inside a classroom can break along lines of power that a general peace message would not reach. So when programmes claim to build social cohesion, the claim is mostly an inference, not a result (Baker et al., 2021; Mayeza and Bhana, 2021).

Read carefully, the evidence points to a more credible approach with four parts. The first is to treat discipline as the main vehicle, so that restorative practice replaces exclusion in the daily routines of the school instead of sitting beside them (Ally et al., 2021). The second is to put peace and anti-bullying training into initial teacher education, where Mawila et al. (2023) found a gap, and to back it with time and funding in schools, as Ndwandwe (2024) recommends. The third is to build programmes with the school community, as Ngidi and Kaye (2022) did, and to aim them at the patterns that surveys like Juan et al. (2025) identify. The fourth is to measure outcomes in ways that allow comparison between schools, because nobody can say how much of the improvement reported so far belongs to the programmes.

Conclusion

Peace education in South African schools has a sensible logic and a thin evidence base. The programmes examined here report modest or slow gains, whether through learner-run peace clubs, a community-built strategy or a restorative justice programme, and the national data show that the problem they address is real and unevenly spread (Ndwandwe, 2024; Ngidi and Kaye, 2022; Juan et al., 2025). What holds the gains back is not mainly a lack of good ideas. It is a punitive disciplinary habit, schools that cannot apply their own codes, and teachers who are underprepared, overburdened and at times harmed by violence themselves (Ally et al., 2021; Du Plessis and Mestry, 2024; Mawila et al., 2023).

The practical conclusion is that peace education should stop being treated as an extra. It belongs inside discipline (Ally et al., 2021) and inside teacher education (Mawila et al., 2023), and it needs to be evaluated well enough that schools and funders can tell what is working. Until that happens, the better programmes will keep producing slow, local improvements that are hard to sustain and hard to prove (Baker et al., 2021; Obadire and Sinthumule, 2021; Mosito and Sitoyi, 2024).

Reference list

Ally, N., Beere, R. and Moult, K. (2021) Red flags: disciplinary practices and ‘school-to-prison’ pathways in South Africa. South African Crime Quarterly, (70), pp. 23-33. doi:https://doi.org/10.17159/2413-3108/2021/i70a11092

Baker, A., MacKenzie, C. and McCormick, J. (2021) Building peace one school at a time: a case study for Catholic schools in South Africa. International Studies in Catholic Education, 13(2), pp. 217-227. doi:https://doi.org/10.1080/19422539.2021.2010461

Du Plessis, P. and Mestry, R. (2024) Violence in rural schools in South Africa: perceptions of school principals and school governing bodies. Perspectives in Education, 42(1), pp. 198-217. doi:https://doi.org/10.38140/pie.v42i1.7173

Juan, A., Hannan, S. and Harvey, J. (2025) At risk: age, gender, and socio-economic status in bullying victimisation among South African primary school students. Perspectives in Education, 43(1), pp. 126-140. doi:https://doi.org/10.38140/pie.v43i1.7662

Mawila, D., Munongi, L. and Mabaso, N.M.P. (2023) Student teachers’ preparedness to raise awareness about bullying among Grade R learners. South African Journal of Childhood Education, 13(1), article a1230. doi:https://doi.org/10.4102/sajce.v13i1.1230

Mayeza, E. and Bhana, D. (2021) Boys and bullying in primary school: young masculinities and the negotiation of power. South African Journal of Education, 41(1), article 1858, pp. 1-8. doi:https://doi.org/10.15700/saje.v41n1a1858

Mosito, C. and Sitoyi, Z.M. (2024) School violence and its impact on teachers’ well-being. Perspectives in Education, 42(1), pp. 131-147. doi:https://doi.org/10.38140/pie.v42i1.7215

Ndwandwe, N.D. (2024) Barriers to implementing peace education in secondary schools in South Africa. Journal of Peace Education, 21(2), pp. 164-184. doi:https://doi.org/10.1080/17400201.2024.2367966

Ngidi, L.Z. and Kaye, S.B. (2022) Reducing school violence: a peace education project in KwaZulu-Natal, South Africa. South African Journal of Education, 42(2), article 1989, pp. 1-14. doi:https://doi.org/10.15700/saje.v42n2a1989

Obadire, O.T. and Sinthumule, D.A. (2021) Learner discipline in the post-corporal punishment era: what an experience! South African Journal of Education, 41(2), article 1862, pp. 1-8. doi:https://doi.org/10.15700/saje.v41n2a1862