Heterogeneous states of regenerative agriculture adoption among smallholders in Kenya

Abstract Regenerative agriculture (RA) is widely promoted as a pathway to restore agroecosystems and strengthen resilience. However, empirical studies often reduce these pathways to a binary distinction between adopters and non-adopters, obscuring the differentiated and institutionally mediated pathways through which smallholders engage with principle-based agricultural transitions. This study conceptualizes RA adoption as a continuum of engagement among smallholder farmers in Kenya. We used a sequential mixed-methods design in which an initial qualitative phase informed survey development, the contextualization of RA indicators, and engagement states. This included 33 individual farmer interviews, 14 focus group discussions, 9 expert interviews, and one stakeholder workshop. Building on these insights, we surveyed 949 farmers and developed a seven-state engagement typology based on awareness, adoption timing, practice history, continuity of use, behavioral change, and adoption lag. We used a multinomial logit model to examine the factors associated with farmers’ positions across these states. Our findings show that binary classification identifies 49% of farmers as RA implementers and 51% as non-implementers. However, findings also mask substantial heterogeneity within both groups. The seven-state classification unmasks this and reveals knowledge gaps, conversion barriers, delayed uptake, stagnation, and discontinuation. Credit access was the most consistent factor associated with RA engagement states beyond awareness. Gender, age, marital status, land tenure, and the county context showed state-specific associations. RA engagement among smallholders is markedly heterogeneous; the observed states are consistent with, but do not demonstrate, staged and reversible adoption pathways. Our typology offers a potentially transferable framework for diagnosing where farmers are located along the RA engagement continuum and for designing state-contingent interventions that address both entry barriers and progression constraints.