Highlights:

• We evaluated the impact of a gender-sensitive poultry marketing intervention on women's empowerment.

• Despite reaching and providing some benefits to women, there was no program impact on women's or men's empowerment.

• The intervention may not have targeted the areas of most significant disempowerment.

• Productive asset transfers may be essential program components for resource poor populations.

Abstract

Understanding the types of food systems interventions that foster women's empowerment and the types of women that are able to benefit from different interventions is important for development policy. SELEVER was a gender- and nutrition-sensitive poultry production intervention implemented in western Burkina Faso from 2017 to 2020 that aimed to empower women. We evaluated SELEVER using a mixed-methods cluster-randomized controlled trial, which included survey data from 1763 households at baseline and endline and a sub-sample for two interim lean season surveys. We used the multidimensional project-level Women's Empowerment in Agriculture Index (pro-WEAI), which consists of 12 binary indicators, underlying count versions of 10 of these, an aggregate empowerment score (continuous) and a binary aggregate empowerment indicator, all for women and men. Women's and men's scores were compared to assess gender parity. We also assessed impacts on health and nutrition agency using the pro-WEAI health and nutrition module. We estimated program impact using analysis of covariance (ANCOVA) models and examined whether there were differential impacts by flock size or among those who participated in program activities (treatment on the treated). Program impacts on empowerment and gender parity were null, despite the program's multipronged and gender-sensitive approach. Meanwhile, results of the in-depth gender-focused qualitative work conducted near the project mid-point found there was greater awareness in the community of women's time burden and their economic contributions, but it did not seem that awareness led to increased empowerment of women. We reflect on possible explanations for the null findings. One notable explanation may be the lack of a productive asset transfer, which have previously been shown to be essential, but not sufficient, for the empowerment of women in agricultural development programs. We consider these findings in light of current debates on asset transfers. Unfortunately, null impacts on women's empowerment are not uncommon, and it is important to learn from such findings to strengthen future program design and delivery.