Empowering Married Adolescents: An Impact Evaluation in Northern Nigeria

Policy Context

As many as 12 million girls are married annually before the age of 18 (UNICEF, 2022). In northern Nigeria, the problem is both severe and widespread, with estimates suggesting 78% of girls are married before the age of 18 and 48% before the age of 15 (Save the Children, 2021). There is substantial causal and correlational evidence that child marriage worsens outcomes across girls’ lives in educational, economic, and household outcomes (Parsons et al., 2015) and worsens the health of children born from these unions (Chari et al., 2017; Raj, 2010). Relative to women over twenty, girls under fifteen are five times as likely to die in childbirth and girls fifteen to nineteen are twice as likely (United Nations, 2001); younger girls also suffer significantly higher rates of serious pregnancy complications, including obstetric fistula (EngenderHealth, 2003).

Many interventions are focused on delaying child marriage (e.g., Bandiera et al., 2020; Buchmann et al., 2023; Cohen et al., 2023) yet far less research has examined whether post-marriage interventions can address some of these mechanisms. Safe Spaces for Married Adolescents (MAS), an initiative by the northern Nigeria-based Centre for Girls Education (CGE), focuses on this under-studied response. MAS provides safe spaces and life skills to married adolescent girls aged 14 to 19, and connects girls to communities, clinics, and contraception.

Study Design

Starting in 2022, Professor Isabelle Cohen and team implemented a cluster randomized control trial of the program in 40 communities in Kaduna state. Communities were paired for randomization post-baseline, using an optimally greedy matching algorithm (King et al., 2007). A one-year follow-up survey in August 2023 found a 40% decrease in pregnancy rates (.145 percentage points) and a .6 standard deviation decrease in an index measuring intimate partner violence. These results suggest both that post-marriage interventions can mitigate the effects of child marriage, and that it may be possible to decrease fertility without incurring intra-household trade-offs. However, the communities where MAS is being tested are remote, and surveying in these areas is a challenge; the results noted above are still self-reported. Reporting biases and enumerator demand effects may reduce response reliability.

As such, Professor Cohen and team will use CEGA funding to pilot several approaches to gathering data from health centers located in the areas where the study is taking place. The approach involves two strategies: first, surveys of healthcare staff in local clinics to assess their perceptions of service utilization from specific communities; second, attempts to access  deidentified patient records either at individual or community levels. These records would provide the opportunity to match real-world data to the self-reported effects collected in 2022.

Results and Policy Lessons

Forthcoming.