SHOUHARDO III performance and impact evaluation : endline evaluation of Bureau for Humanitarian Assistance resilience food security activity in Bangladesh, implemented by CARE International : volume I : report

This evaluation employs three methodologies: qualitative inquiry, pre-post comparison, and impact evaluation. The impact evaluation matches communities treated by SHOUHARDO III with untreated communities ex-post, using baseline stunting rates from the 2014 DHS dataset. The evaluation finds that the SHOUHARDO III project engaged more than 40% of households surveyed within target villages and successfully targeted poor and female-headed households. The analysis of baseline and endline statuses (pre-post analysis) of households in the SHOUHARDO IIItargeted areas demonstrates that households from these areas improved across several indicators, including poverty levels, the nutritional status of women and children, women’s empowerment, and gender equity. From a qualitative standpoint, participants from areas where SHOUHARDO III appeared well-implemented offers insights into the potential of the interventions. The qualitative evaluation found mechanisms of change in several areas that can be built upon and enhanced. Qualitative findings show that the program succeeded in promoting multi-sectoral change at household and community levels. They also show that SHOUHARDO III effectively targeted services to the most food-insecure, Poor and Extremely Poor members of communities, and its multigenerational and gender-inclusive approach to its interventions facilitated community acceptance. From the impact evaluation, it is likely that we can credit SHOUHARDO III with improvements in women’s dietary diversity, women and children’s minimum acceptable diet, antenatal care access, and the increase in participation across several sectors. In addition, households in SHOUHARDO III villages experienced statistically significant differences in one resilience indicator, and households in program villages that experienced major shocks were better able to maintain their food consumption than similar households in comparison villages. However, the impact evaluation does not find meaningful differences between households in targeted communities and households in non-targeted communities in terms of women’s mobility and decision-making, children’s nutritional status (including child stunting and underweight status), children’s diarrhea, exclusive breastfeeding, household hunger, and improved use of health and nutrition services overall. Improvements in mostmeasured conditions in the SHOUHARDO III program areas appear to have been matched by similar improvements in non-program areas, suggesting broader forces may account for them. Ultimately understanding differences between program areas and non-program areas can help inform decisions about future chapters of the SHOUHARDO III program and other development food security programs to ensure the most effective programs for vulnerable populations. Understanding the dynamics and mechanisms of change and responses of participants to interventions can also inform future work. Salient findings are also important to highlight for action. The research team concludes this report with recommendations.