Agricultural Extension for Climate-resilient Farming: A Critical Review of the Emerging Roles of Advisory Services in Adaptation and Mitigation

Abstract

Agricultural extension and rural advisory services occupy an increasingly prominent position in national and international strategies for climate-resilient farming, yet the evidence supporting that position is uneven and frequently over-interpreted. This critical narrative review examines what is known, contested and unknown about the roles that advisory services perform in supporting adaptation and mitigation in farming systems, with particular attention to smallholder contexts in sub-Saharan Africa and South Asia alongside evidence from higher-income settings. Literature was identified through structured searching of multidisciplinary and regional scholarly sources, supplemented by citation tracking and examination of institutional publications, and was appraised for design adequacy, attribution logic, contextual specificity and consistency. Four claims emerge with reasonable confidence: advisory contact is associated with higher uptake of adaptive practices across many observational studies; experimental evidence confirms that information delivery can raise adoption, though average effects are modest and heterogeneous; digital channels reduce delivery costs substantially but do not reliably reproduce the persuasive and diagnostic functions of interpersonal advice; and mitigation remains a marginal, weakly evidenced component of advisory practice in low-income agriculture. Several widely repeated propositions are less secure than their circulation suggests, including the assumption that improved climate information translates predictably into changed farm decisions, that co-production reliably resolves usability problems, and that gender-responsive design follows automatically from targeting women. Persistent weaknesses include reliance on cross-sectional designs vulnerable to selection bias, conflation of knowledge gain with practice change, scarcity of resilience outcomes measured over climatically variable periods, and near-absence of costed comparisons between advisory modalities. Priorities for future work include multi-season trials with pre-specified resilience endpoints, evaluation of hybrid human–digital models, measurement systems capable of attributing emission changes to advisory intervention, and analysis of the political and financial conditions under which climate-responsive advisory reform survives beyond project cycles.