Integrated Graduation Program and Its Effect on Women and Household Economic-Wellbeing: Findings from a Randomized Controlled Trial in Burkina Faso

Background and Purpose: To address the main limitation of poverty-reduction programs not reaching the bottom poor, a growing number of agencies are adopting a “graduation approach” to moving out of extreme poverty. The graduation approach falls within the framework of “productive safety net programs” that simultaneously provide social protection along with the aim of creating productive and resilient livelihoods. Our study aims to investigate (i) whether participation in the graduation program changes women’s investment in and income and profit from various activities; (ii) whether improvements in income generation translate into greater assets and savings for the women—within the context of the program’s effect on general household assets and financial security; and (iii) whether benefits derived from participation in the program increase household expenditure on children below age 16. The study differs from evaluations of other graduation programs examining the household-level assets and participant’s monthly aggregate income or monthly cash earnings, without explicitly focusing on women’s assets distinct from the household assets.

Methods: The study, implemented in Nord Region of Burkina Faso (ClinicalTrial.gov ID: NCT02415933), uses a cluster-randomized design with two treatment arms and one control arm. Randomization was conducted at the village level to avoid cross-arm contamination. Twelve villages located in the Nord Region of Burkina Faso were randomly assigned to one of the three study arms. Repeated-measures data was collected at baseline, at 12 months, and at 24 months from female caregivers (n=360). We used multilevel mixed-effects models that allow estimating subject-specific effects while taking into account clustering of subjects within villages. We modeled the between-village variability as a random effect. We also accounted for a possible within-individual correlation (due to the repeated-measures nature of the data) by modelling the within-individual variability as random effect.

Results: We found that treatment group women reported significant and large increases in investment from cash crop cultivation, livestock rearing and other income generating activities. With greater investments, came greater income and profit from the respective activities. We also found that increased return from market activities led, in turn, to greater assets (i.e. durable assets and livestock assets) owned by the women. Intervention also had significant effect on improving economic well-being of the household in general, with increases in durable and productive assets; and increase in expenditure on children, although not at the same rate as the increases in income and profits.

Conclusions and Implications:  By providing women with financial tools, cash grants, and access to local markets, the intervention addresses economic security barriers faced by women in the region. Provided at per household cost of USD 208-228, the program generates greater income for women after two years. The success of this intervention provides strong support for the expansion of the graduation approach program to help the poor in different settings. Our findings also point to the importance of taking into account existing social relationships within households and communities. The added value of addressing these intra-household dynamics through appropriate program strategies is shown through increased likelihood of project success demonstrated in our findings.