Violence during childhood and adolescence can have lifelong adverse health, social, educational and economic consequences ^{1– 3} . In particular it can increase vulnerability to HIV acquisition ⁴ due to increased likelihood to engage in risky sexual behaviours, cause poor academic performance ^{2, 5, 6} , and can lead to higher levels of depression and suicide ideation ⁷ . Substance use also is a public health concern, as a predictor of violence ⁴ . While children who are not exposed to violence can still be violent, children who are constantly exposed to violence display violent behaviours more often than those who are not ⁸ . Other characteristics can lead to some children being more likely to be bullied than others.

Schools should be a conducive teaching and enabling environment for both learners and educators; however, violence in South African (SA) schools continues to be a problem, despite progressive school safety policies and implementation frameworks by the National Department of Basic Education (DBE) ^{9– 11} . Schools have become unsafe places, where learners are at risk of experiencing violence and in some cases death during school hours, in after-school programs, and on their way to and from school ^{2, 12, 13} . Perpetration of violence and victimization in schools has two dimensions: 1) learner-learner violence where learners may bully, harass, rob or assault each other of their valuables; and 2) educator-learner or learner-educator violence ¹⁴ .

Multiple studies and media coverage over the last decade have highlighted the severity of school violence ^{2, 9, 15– 18} . According to these studies, bullying stands out as a common act of violence in SA, where in many cases learners are perpetrators of violence, with both, learners and educators becoming victims ¹⁹ . In the 2019 Organisation for Economic Co-operation and Development (OECD) Teaching and Learning International Survey, 34% of SA principals report that acts of intimidation or bullying among their learners occur at least weekly in their school, more than double the OECD average ²⁰ . Linked to bullying, data from the 2016 SA school violence against children (VAC) survey with 15 – 17 year olds, showed that 82.0% reported experiencing some form of victimisation whether criminal victimisation or exposure to family or community violence ²¹ . Another study indicated that 22% of learners aged between 12 –18 years across SA having experienced some form of violence while at school in the past year ² . When it comes to sexual harassment, the same VAC study showed that over a third of young people had experienced some form of sexual abuse at some point in their lives and almost 10% had been made to do sexual acts against their will by a peer ²¹ . A qualitative study with 13 – 17-year-old SA learners found that both boys and girls are perpetrators and victims of sexual violence in schools ²² . Research reveals that there is poor management of violence in SA schools, with many educators and learners fearing retribution from the perpetrator ²² . This suggests that reporting the incidence of violence to educators is generally ineffective ²² , resulting in an unending cycle of violence. Regarding drug and alcohol use, principals report weekly incidents related to the use or possession of drugs and/or alcohol at school (South Africa 27%; OECD average 1%) as well as vandalism and theft (South Africa 21%; OECD average 3%), which are comparatively very infrequent in other countries ²⁰ .

With regards to violence perpetrated by educators against learners, Burton ¹⁵ revealed that school principal management reports show that 50% of educators have verbally abused learners, and 25% of educators have physically abused learners. Despite being banned through the National Education Policy Act of 1996, corporal punishment, defined as any kind of violent action inflicted on children by educators or school administrators as punishment for disciplinary purposes, is still used as a common form of discipline in many schools ^{22, 23} . Provincial rates of corporal punishment vary between 22.4% in Gauteng to 73.7% in Kwa Zulu Natal ² .

Learner and educator safety is the mandate of the DBE, led by the School Safety Directorate. In response to an accumulation of safety concerns for both learners and educators, the DBE, the Centre for Justice and Crime Prevention and UNICEF South Africa introduced and launched the National School Safety Framework (NSSF) in 2015, as an update to the Hlayiseka Early Warning System ²⁴ . The NSSF is located within a range of international and national laws and policies that recognise the safety of learners and educators as a prerequisite for quality learning and teaching at school ²⁵ . The NSSF provides an important instrument through which minimum standards for safety at school can be established, implemented and monitored, and for which schools, districts and provinces can be held accountable ²⁵ . The four strategies of the NSSF are as follows:

Late in 2015, the Wits RHI received funding from the Bill and Melinda Gates Foundation to conduct the Girls Achieve Power (GAP Year) trial. As a cluster randomised control trial across 26 schools in Gauteng and Western Cape, GAP Year sought to test the effectiveness of a four-pronged ecological intervention: a sports-based after-school asset-building component, a parent engagement component, linkage to care component and a school safety component ²⁶ . The school safety component included supporting the DBE to implement the NSSF. GAP Year schools were assessed, in partnership with DBE, using the NSSF implementation framework to determine the state of violence prevention, management, and reporting, seeking to create an enabling environment for adolescents by transforming schools into hubs of safety and support. The primary outcomes of GAP Year were to reduce school dropout of adolescent girls between grades 8–10 and increase reporting of gender-based violence (GBV). Complimenting these outcomes, the four-pronged intervention sought to improve adolescent girls’ agency and safety while shifting gender attitudes and encouraging positive behavioural change among adolescent boys.

This paper provides the results of the baseline learner survey, one component of NSSF implementation. It also contributes to the ongoing routine monitoring of the NSSF and the little, yet growing, evidence base of violence and safety at SA public schools: the last school safety survey was in 2016 ²¹ by Artz et al., and before that, 2012 ² , by Burton et al., on behalf of the DBE. This manuscript also seeks to inform the development and design of effective interventions and policies to address the prevailing concerns of safety and violence in schools and provides recommendations for the use of and analysis of the NSSF learner survey tool.

A total of one thousand and thirty-four (1034) Grade 9 to 11 learners completed the NSSF learner survey ( Table 1 and Table 2). Overall, 52.9% (n=547) were female and the majority were aged between 16-18 years (58.2%, n=602). The mean age was 16 years (SD=1.36), and all learners were African. A description of learner demographics by sex are found in Table 1. Most participants were from Soweto (47.3%) and were enrolled from Grade 9 (37.1%). Over half of the participants (52.9%, n=547) were female and most females majority were aged between 13-15 years (61%). Of those aged >=19 years, 74% were males. The full, deidentified survey responses are available in Underlying data ²⁷ .

Overall, we found high rates of violence among 1034 learners in the 15 government-funded South African high schools who participated in this research. Almost 30% experienced peer provocation and relational aggression, 14% experienced peer perpetrated sexual harassment and almost 50% reported they felt unsafe to and from school and almost 90% were exposed to illicit drugs on the school property. Despite this, 76.8% indicated that discipline was enforced in their schools and 58.3% felt able to comfortably report any form of violence to educators.

We found sex differences in experiences of safety at public high schools with males experiencing more peer provocation and relational aggression at school while female learners were more concerned with safety threats in the community, on their way to and from school. The role of the School Management team in ensuring safety and enforcement of discipline was viewed differently between male and female learners, with females feeling more comfortable to report any form of violence to educators. Our overall findings regarding learner experiences of violence are consistent with other research findings and reports, reinforcing that school safety remains a concern in SA schools ^{2, 20, 38, 39} .

Our findings showed four key sex differences in experiences of violence, which will be outlined in turn.

Our findings indicate that male learners are more susceptible to peer provocation and relational aggression compared to females. This is confirmed by existing literature recognising men’s vulnerability to violence ⁴⁰ especially perpetrated by male strangers or acquaintances ^{41, 42} . Constructs of masculinity and femininity that position men as dominant and highly sexually active and women as subordinate and acquiescent have been found to contribute towards gender inequality and in turn, to violence ⁴³ . Interventions such as the SASA! intervention could be implemented, through community mobilization, tackling the social and cultural norms to address the primary prevention of GBV and HIV ⁴⁴ . The SASA! intervention was associated with lower past-year of experience of physical and sexual intimate partner violence among women, highlighting that addressing social and gender norms can result in reduction of violence ⁴⁴ .

Community safety was a concern for female learners. This is confirmed by literature that indicates that community crime and violence is a concern in SA ⁴⁵ , highlighting the need for a more ecological approach to safety, including community safety with an inter-governmental approach. Various evidence-based interventions could be deployed such as the Walking Bus Project ⁴⁶ , self-defence workshops ^{47, 48} and the development of community safety plans through Community Safety forums ⁴⁹ , with the South African Police Service (SAPS) playing an active role in mapping and ensuring safety.

The poor enforcement of discipline relating to any violence at school experienced by male learners, correlates with and is confirmed by the high rates of peer provocation and relational aggression experienced at a school level. We suggest that learners are involved in the enforcement of discipline as well as strengthening the code of conduct to allow for a more ‘whole school approach’ to addressing this issue, rather than top down. Strategies for restorative discipline should be implemented ⁵⁰ , encouraging cycles of peace, rather than punitive discipline, which is currently the norm ⁵¹ .

We found that females felt more comfortable reporting violence to educators, than their male counterparts. More research could be done to better understand why males are not comfortable reporting and also to explore if this finding translated to an increase in reporting of violence at a school level by females. However, with the recent case of a Limpopo learner committing suicide after being bullied ⁵² , schools need to re-think and strengthen the anti-bullying policy intervention strategies at all levels ⁵³ to identify more subtle types of violence, like bullying, correctly and swiftly. The introduction of anonymous reporting systems could also help encourage reporting of all types of violence. In addition, interventions, like the GAP Year intervention ²⁶ , should be done at a school level to educate and empower boys on the reporting pathway in a school environment, increasing their ability and confidence to report whilst reducing the stigma. As with the other interventions noted, their success will depend on the extent to which all stakeholders, including learners, educators and other school staff, as well as parents, and the wider community are committed to reducing bullying ⁵⁴ .

While not statistically significant, our findings on sexual harassment and exposure to illicit drugs are concerning, however they do align with other studies. Our results confirm findings from Kutywayo et al. ⁵⁵ and Ward et al. ¹⁶ indicating that both female and male learners are vulnerable to sexual harassment and violence. The Optimus Study school survey ¹⁶ , and other South African studies ^{2, 56} also found similar results: one in 10 of young people had experienced unwanted sexual touching by a known or unknown adult in their lifetime ²¹ . Campaigns and interventions, supported by the government ⁵⁷ , should be promoted to normalise the reporting of all types of violence and ensure the adequate first-line support to providers to respond to disclosures of violence. First-line support using the LIVES framework ² , provides practical care and responds to a survivors’ emotional, physical, safety, and support needs ⁵⁸ .

Given that in 2001, all South African schools were declared drug-free zones and no person may possess illegal drugs on school premises ⁵⁹ , it is concerning that almost 90% of all learners had exposure to illicit drugs on the school property. These concerns are confirmed by other studies ⁶⁰ . As with the other interventions noted in this manuscript, an ecological approach is critical to address the root cause of this drug problem and break the cycle. The SAPS, the South African National Council on Alcoholism and Drug Dependence (SANCA) ³ and the Substance Abuse Prevention through Academic Excellence program ⁶¹ , provide potential interventions.

Several of these findings suggest there is an urgent need for the DBE to strengthen the functionality of the School Safety Committees (SSC), a substructure of the School Governing Body. The participation of the SAPS in the SSC’s is also critical as they conduct random checks and ensure safety from gang related violence, among other safety prevention methods. The recent digitisation of the NSSF and the Protocol on the Management of Sexual Abuse and Harassment training ⁶² , outlining the roles and responsibilities of the SSC, is one step in creating a more functional SSC. All the interventions mentioned in this manuscript should be led by the SSC, in collaboration with the other stakeholders. One opportunity for future research would be to explore the variables that create an exceptional school.

Findings confirm that school safety and violence continue to be a concern in SA schools. Whilst these behaviours are displayed at school, violence is determined by a range of often inseparable dynamics located at individual, relationship, community, and societal levels ⁶⁴ . This strongly indicates that interventions must work across the various ecological levels to effectively see an impact on violence prevention and reduction. There is also a need to strengthen NSSF implementation to address the concerns that were reported by learners to ensure more effective and efficient learner and educator safety within the school environment. School safety can be strengthened through development and implementation of sustainable action plans, including effective incident reporting system for learners and staff, ensuring that referrals and psychosocial support is provided. We further recommend the adaptation of the NSSF learner survey to increase its utility and implementation thereby allowing it to generate more accurate and timely data for the identification of appropriate safety interventions.