A scoping review of scaling agricultural innovations in sub-Saharan Africa: from diffusion to systems transformation

Abstract

Background Scaling development innovations is critical for advancing agricultural transformation and food system resilience in Sub-Saharan Africa (SSA), yet significant gaps remain in how scaling is conceptualized and empirically assessed. Objectives This scoping review examines how scaling is defined, operationalized, and evaluated in empirical studies, focusing on dominant approaches, methodological trends, and key enablers and constraints. Guided by PRISMA scoping review guidelines, 54 eligible studies published between 2016 and 2026 were included from the initial 2070 studies that were considered for this review. Data were extracted using a structured template capturing study characteristics stipulated in the objective. Results Findings show a strong temporal concentration, with 86% of studies published between 2016 and 2021. Geographically, evidence is uneven, dominated by Kenya (28%), followed by Ethiopia (12%) and multi-country SSA studies (16%) including Malawi, Zambia, Namibia, Tanzania and Mozambique. The literature is heavily quantitative, with 47% using randomized controlled trials, 26% comparative surveys, and 16% panel data designs, policy and evidence synthesis indicating a strong preference for causal inference approaches. 65% of studies focus on agricultural input and productivity-enhancing technologies, while 9% examine value chain or market-oriented innovations. Scaling is predominantly framed as horizontal diffusion, with 70% emphasizing “scaling out” strategies. Key enablers include access to information, extension services, and social networks (60% of studies), as well as financial access (30%). Major barriers include financial constraints (23%), limited access (21%), and knowledge gaps (16%). Notably, fewer than 15% of studies explicitly consider systemic dimensions of scaling, revealing a disconnect between empirical approaches and systems-based frameworks. Conclusion The review shows the need to move beyond diffusion-centric models toward integrated, systems-oriented approaches.