Climate-smart agriculture and multidimensional poverty among smallholder maize farmers in Kenya

Climate-smart agriculture (CSA) has emerged as a central strategy for strengthening agricultural resilience under climate change. However, limited evidence exists on whether adopting multiple CSA practices is associated with improved multidimensional poverty among smallholder farmers. This study examined the association between CSA adoption intensity and multidimensional poverty among smallholder maize-farming households in Kenya. CSA intensity was measured as the number of CSA practices adopted by each household and categorised into high- and low-intensity adoption groups. An endogenous switching regression model was employed to account for potential selection associated with households’ adoption decisions, while propensity score matching was used as a robustness check based on observable characteristics. The findings reveal that CSA adoption remains relatively low, with over half of the sampled households not adopting any of the selected CSA practices and less than one percent adopting three practices simultaneously. After accounting for observed household characteristics and potential selection, households with higher CSA intensity exhibited lower scores on the multidimensional poverty index (MPI) than households with lower intensity. Education, farmer group membership and household experience were also consistently associated with lower MPI levels, while substantial spatial heterogeneity was observed across counties. The findings suggest that promoting the integrated adoption of complementary CSA practices, alongside investments in extension services, farmer organisations and improved access to finance, could help reduce household MPI levels. By linking climate adaptation with multidimensional poverty, this study provides evidence to inform the desi