A mixed-methods impact evaluation of an intervention to address school-related gender-based violence in Zambia

• In Event: Digital tools supporting youth and gender justice in educational settings

Proposal

School-related gender-based violence (SRGBV) is prevalent and has significant consequences on adolescents’ development and well-being. SRGBV is defined as “acts or threats of sexual, physical, or psychological violence occurring in and around schools, perpetrated as a result of gender norms and stereotypes and enforced by unequal power dynamics” (UNGEI, 2019). School-based interventions have shown mixed promise in addressing SRGBV in low-resource settings and evidence remains limited, particularly on how interventions might affect boys and girls differently. This paper presents findings from a mixed-methods impact evaluation of the Empowerment Pilot (EP), a school-based program aimed at addressing SRGBV in secondary schools in Zambia. The intervention targeted both teachers and students and aimed to raise awareness about SRGBV; build critical assets including knowledge, resources, and skills to prevent and respond to SRGBV; and change attitudes and norms to improve student safety. While the program was designed as a whole-school approach and the results are reported at the school level, in practice it was primarily delivered through safe-space clubs. Preliminary results suggest no overall effect on SRGBV, but disaggregated results show a 2-percentage point reduction in sexual violence, representing a 33% decrease relative to the control group. One potential mechanism for this change is an improved belief among students that peers would intervene if they witnessed violence. Additionally, we observe a 50% increase in discussions about violence among students during implementation. However, there was no corresponding shift in student attitudes towards SRGBV, suggesting that the reduction in sexual violence may be attributed to a heightened perceived cost of being seen committing violence.

Program: The EP includes a Teachers Curriculum; a Learners Curriculum delivered through separate boys and girls safe space clubs with up to 25 participants each; and a Parent-Teacher Committee Engagement Module. Two teacher-mentors per school were trained to deliver the program. The intervention took place from May-October 2023 in three districts in Zambia with students in grades 8-11.

Data: The impact evaluation included a total of 90 schools and was designed as a cluster randomized controlled trial with the intervention randomized at the school level. Data collection included a school survey and the listing of all students in the beginning of the study, two rounds of qualitative data collection during implementation, followed by one round of quantitative surveys four months after the intervention concluded. Quantitative data were collected from adolescents and teachers in study schools. Twenty-five boys and 25 girls were randomly selected per school to form the all students sample. Qualitative data were collected from eight intervention schools and included interviews, focus group discussions, and observations with boys and girls in and out of clubs, and mentor and non-mentor teachers.

Analysis: Quantitative data were analyzed using ordinary least squares regressions to estimate the impact of the Empowerment Pilot and qualitative data were systematically coded using a combination of inductive and deductive codes.

Preliminary Results: While designed as a whole-school approach, there was little evidence of consistent implementation of activities beyond the clubs. Club members were enthusiastic and found the curriculum relevant. However, some boys believed the content on gender roles conflicted with community norms and a subset of male teachers worried that messages on learners’ rights and positive discipline undermined their classroom authority. Non-club adolescents expressed interest in participating when they were aware of clubs, but often knew little about the program.

The EP was implemented in a context of high SRGBV prevalence—61% of boys and 57% of girls in control schools experienced any SRGBV in the last term. Qualitative data indicate that violence was normalized as part of school culture and victim blaming was common. There were weak beliefs that perpetrators would be held accountable, and fears that reporting would lead to backlash and shame. This fostered an environment where a culture of violence persisted in schools.

In this context, quantitative data show the intervention had no impact on the two aggregate measures of SRGBV during or four-months after implementation among the all student sample. However, it reduced the incidence of sexual violence in the past two weeks by two percentage points, from a control mean of 6%. There was no change in physical or emotional violence.

The intervention did not alter the three aggregate measures of attitudes, descriptive or injunctive norms related to gender roles and SRGBV. However, boys in intervention schools developed slightly more equitable gender attitudes, while girls’ attitudes remained unchanged. The intervention also showed some positive effects on injunctive norms regarding gender roles, but it did not change attitudes or norms related to SRGBV.

While the intervention increased discussions about violence during implementation, it did not enhance adolescents’ knowledge of SRGBV, as their knowledge was already high. Qualitative data indicate that adolescents readily recognized SRGBV, but faced norms that reinforced the acceptability of violent behavior. While reporting of SRGBV was low, adolescents from intervention schools were more likely to report instances of violence they witnessed as bystanders, and demonstrated a greater belief that their peers would act if they witnessed hypothetical situations of bullying.

The intervention did not improve perceived school climate, socio-emotional wellbeing, agency, or school enrollment four months after implementation. Heterogeneity analysis suggests that the intervention may have operated through different pathways in primary compared to secondary schools.

Discussion: It is possible that increased discussion of violence, bystander reporting, and shifting beliefs that a bystander might intervene if they witnessed violence, may have increased awareness of violence and “informal” accountability, which could have led to the observed reduction in sexual violence. However, overall limited impacts of the EP may be due to a number of factors including that the program appears to have had limited “spillover” to those outside of the clubs, and that the EP may have been too ambitious in its aim to address all types of violence without a clear understanding of SRGBV prevalence and drivers of different forms of violence within schools. This points to important program implications, including the potential need design targeted approaches to address different forms of SRGBV (by type and perpetrators).

Authors

• Sophia Friedson-Ridenour, The World Bank

• Carla Zenobia Glave Barrantes, The University of Wisconsin-Madison

• Tara Bedi, Trinity College Dublin

• Wei Chang

• Yenny Chavarria, The University of Wisconsin-Madison

• Menaal Ebrahim

• Anu Jose, Central Bank of Ireland

• Michael King, Trinity College Dublin