You have full access to this article via your institution.

Source research: Cohen, I., Abubakar, M. & Perlman, D. A big-push community intervention reduced rates of child marriage by 80%. Nature https://doi.org/10.1038/s41586-026-10206-2 (2026).

Messages for policy

• Multipronged programmes can have high upfront costs yet still have a net positive benefit.

• With the right programme and implementation, the age at which girls marry can be delayed.

• Creating opportunities for girls to attend school in cases in which that runs counter to the behaviour of most people in their communities might work only if cost and social barriers can be overcome.

• Educating girls has many benefits, not just for the girls themselves, but also for their families and communities.

The policy problem

Globally, around 650 million girls and women alive today were first married before 18 years of age¹. A 2021 report² estimated that nearly 80% of girls in northern Nigeria had married before they turned 18. Early marriage — particularly when it is arranged by parents, guardians or other parties without the involvement of the girl — has negative effects on many aspects of the futures of adolescent girls. It affects their agency, health and education adversely, as well as their level of income, and increases their risk of experiencing violence³. It has been estimated⁴ that ending early-marriage practices would reduce the share of girls who have a child before they are 18 by 75% and could also increase women’s earnings and productivity, providing large benefits to countries’ economies. For instance, in 2015, an estimated US$7.6 billion in extra earnings could have been generated in Nigeria⁴.

In many contexts, girls marry at a young age because feasible alternatives are lacking; systemic constraints can mean that marriage is the best choice available to them⁵. The existing resource constraints, attitudes and norms that enable child marriage are multifaceted and intertwined, yet many interventions that seek to address child marriage have a narrow focus. Multifaceted interventions could take strategic advantage of various programmes’ complementary components to succeed where others have failed.

The findings

We found that the Pathways to Choice programme reduced by 80% the likelihood that girls in northern Nigeria would be married two years after receiving an intervention that encouraged them to attend school or other training. Girls in the control group, who did not have access to the programme, had a marriage rate of 86%; those in the intervention group, a rate of 21% (Fig. 1). Our intervention also increased the girls’ school attendance by 70 percentage points, and improved their access to social support, as well as their self-perception and ability to advocate for themselves. Moreover, it increased the likelihood that a participant’s younger siblings would be enrolled at school, with an increase of 87% for sisters and 41% for brothers. We find that the programme has net returns of $1,627 per $1,000 invested, and our estimates of its lifetime advantages for the participants show that the intervention has a benefit–cost ratio of 2.41.

These results cover only a short period; understanding the programme’s full effects will require long-term data on participants’ lives. Context matters: we expect that settings in which education is a socially acceptable alternative to early marriage and schools are not currently meeting girls’ needs are more likely to benefit from our intervention than are settings in which high-quality education is already available or in which education is not viewed as a suitable alternative.

The study

The Pathways to Choice programme is a multipronged, community-focused intervention by Nigeria’s Centre for Girls Education in Abuja. It simultaneously tackles several constraints on girls’ education over two years, through community engagement, remedial education and social and in-kind support, encouraging participants to attend school or vocational training. To test its effects, we performed a randomized controlled trial between 2018 and 2020 in 18 communities in northern Nigeria’s Kaduna, Kano and Borno states. We focused on 1,181 unmarried girls who were aged between 12 and 17 years old and not at school at the start of the study. After a baseline survey, the communities were divided into nine pairs, with one of each pair receiving our intervention. Two years later, we conducted a follow-up survey with the same participants. Our results compare the outcomes of girls in communities that received our intervention with those in communities that did not.

References

Competing Interests

I.C. declares no competing interests. M.A. is an employee of the Centre for Girls Education. D.P. has consulted for the Centre for Girls Education and served as its first director until 2016.