Survey data supporting the publication: Governance Depth as the Missing Link: Why Ethiopia’s Nutrition-Sensitive Agriculture Strategy Failed to Improve Dietary Outcomes

This manuscript presents a mixed-methods case study evaluation of Ethiopia's National Nutrition-Sensitive Agriculture (NSA) Strategy (2017–2021), examining why high-level policy commitment failed to translate into sustained, equitable implementation at scale. The study employed a qualitative-dominant, sequential explanatory design, combining semi-structured key informant interviews (n=51) with federal and regional policymakers, program coordinators, development agents, and donor partners across four purposively selected regions, Amhara, Oromia, Sidama, and Somali, chosen to capture variation in agro-ecological zones (highland to pastoral), administrative capacity, and nutrition outcomes. Interviews were conducted between January and June 2022, audio-recorded with informed consent, transcribed verbatim, and analyzed using thematic analysis with NVivo 12, guided by an a priori framework of five governance domains (leadership, financing, accountability, coordination, institutional capacity) while allowing emergent themes. To triangulate and supplement qualitative findings, the study descriptively analyzed secondary quantitative data from the Ethiopia Demographic and Health Survey (EDHS 2016 and 2019) for child dietary diversity scores (children aged 6–23 months), the Agricultural Sample Survey (2017–2021) for production trends of nutrient-dense crops, and annual government performance reports (2018–2021). Documentary analysis included the NSA Strategy document (FDRE, 2016b), the Food and Nutrition Strategy (FDRE, 2021), and the Seqota Declaration implementation plan. Ethical approval was obtained from Jimma University Institutional Review Board . All participants provided written informed consent, and data were anonymized prior to analysis. The study identifies financing and accountability as<em>binding constraints</em>using a causal cascade analytical technique, introducing the concept of<em>governance depth, </em>operationalized as the joint presence of dedicated predictable financing, outcome-oriented accountability systems, and institutionalized sub-national capacity. To ensure reproducibility, the interview guide, codebook, and detailed case selection criteria are provided in supplementary materials. Limitations include retrospective recall bias (mitigated through documentary triangulation), limited generalizability to conflict-affected areas (e.g., Tigray, parts of Amhara), and absence of primary household-level dietary data, which future research should address through integrated household and institutional surveys.