The Influence of Schooling on the Stability and Mutability of Gender Attitudes: Findings From a Longitudinal Study of Adolescent Girls in Zambia

"Increasing calls to shift gender attitudes to be more equitable - both as a matter of justice and as a means to improve other outcomes, including health - inevitably lead to the question of when and how to best intervene. Are there times in the life course when gender attitudes are more flexible and open to change?"

Multiple factors at different levels - individual, relational, social, institutional, and structural - interact with one another to reproduce gender norms and behaviours in a reinforcing ecological system. Inequitable gender attitudes emerge early in childhood, and are either reinforced or challenged in places such as schools - with evidence of the pervasiveness of inequitable gender norms across diverse settings. This study investigates the role of education in shaping adolescent girls' gender attitudes using longitudinal data collected as part of the Population Council's multiarmed cluster randomised controlled trial (RCT) of the Adolescent Girls Empowerment Program (AGEP) in Zambia.

As detailed at Related Summaries, below, AGEP was implemented in 10 sites in 4 of Zambia's 10 provinces in an effort to increase the likelihood that adolescent girls remain in school, avoid early marriage, delay sexual activity, and prevent unintended pregnancy, HIV, and other sexually transmitted infections. The study compared 3 different combinations of interventions: safe spaces only; safe spaces plus health vouchers; and safe spaces, health vouchers, and savings accounts.

Unmarried girls aged between 10 and 19 years who had the highest levels of vulnerability (i.e., behind in school for their age) in each AGEP site were eligible to participate in the study. This article focuses on data collected in rounds 1 (2013) to 4 (2017). The response rate at round 4 was high (83%), and the final sample consisted of 4,853 girls. Descriptive analyses of changes in gender attitude scores were restricted to 4,088 girls.

The researchers conducted random effects multinomial logistic regression to determine whether schooling-related factors were associated with shifts in adolescent girls' gender attitudes and explored whether these relationships varied by age. The analysis used the endorsement of violence against women scale as a proxy measure of gender attitudes in the study population. In each survey round, respondents were asked whether violence against women was justified in five scenarios: whether the woman (i) goes out without telling her husband; (ii) neglects the children; (iii) argues with him; (iv) refuses sex with him; or (v) burns the food.

Mean gender attitude scores were significantly higher among urban girls than rural girls across all four survey rounds. Among all girls, there were increases in gender attitude scores over this period. This shift, however, was only statistically significant among urban girls. In both rural and urban areas, older girls were more likely to have stable inequitable attitudes or attitudes that shifted to less equitable over time than relatively stable-equitable attitudes. Age was not associated with attitudes becoming more equitable versus staying relatively stable-equitable.

"Educational attainment was strongly associated with more equitable gender attitudes" - at least among urban girls. Urban girls with higher educational attainment were more likely to have relatively stable equitable attitudes than stable, inequitable attitudes, or attitudes that shifted to inequitable or more equitable. Rural and urban girls currently attending school were more likely to have relatively stable equitable attitudes than stable, inequitable attitudes, or attitudes that shifted to inequitable.

Reflecting on the findings, the researchers note that, overall, "urban girls were more likely to endorse equitable views and become more equitable if they changed their attitudes, and rural girls were more likely to endorse inequitable views and become less equitable if they changed their attitudes. This different pattern in rural versus urban settings highlights the importance of contextual factors in shaping views on gender norms." However, both rural and urban girls attending school were more likely to hold stable equitable attitudes (vs. inequitable or shifting to inequitable). Thus, "schooling appears to be an institutional lever that holds promise for shifting gender attitudes toward greater equality."

The researchers stress that the multiple patterns they observed coexist and vary depending on context, rather than suggesting a predominant explanatory theory. "The interactions documented in this study highlight the importance of taking an ecological approach that considers dynamics across multiple layers of the system to shed light on how gender attitudes are shaped over time, including by school, and how these factors may operate differently in different settings."

In terms of implications of the findings for programming, this analysis reinforces calls for designing more targeted interventions among adolescents. For instance, interventions that "target multiple levels of girls' socioecological environments (e.g., community, schools, etc.) may be more effective in changing views on gender norms. School-based efforts, for example, can target gender attitudes through programs such as sexuality education curricula that address gender and power or civics or social studies classes that critically examine discrimination and inequality. Furthermore, whole school efforts to make schools harassment- and violence-free also reinforce patterns toward more equitable gender views."

In proposing future directions for study, the researchers note: "Deepening understanding of the complexity of gender attitudes' stability and mutability during adolescence highlights the importance of looking longitudinally at the effects of social context and the larger gender system. More quantitative, longitudinal research exploring shifts at the aggregate and the individual level in LMIC [low- and middle-income country] settings is needed. Also needed are studies that explore the effects of different levels of the ecological system and how and why these factors may be associated with more equitable views in one context and more inequitable views in another."

In conclusion: "Ultimately, it is evident that gender attitudes are mutable and that although there are multiple factors across the ecological system that reinforce inequitable norms, elements of the system can be leveraged for transformative change."