While land tenure security is recognised as a cornerstone of sustainable land use, understanding of its influence on the adoption of agroecological practices is still limited, particularly in tenure-insecure contexts. This study examines how land tenure security, institutional support, and behavioural factors shape smallholders’ willingness to adopt agroecological innovations. Institutional and behavioural perspectives are integrated within a discrete choice modelling framework to analyse data from a choice experiment conducted with 400 smallholder farmers in Huambo Province, Angola. Willingness-to-accept (WTA) estimates were derived for key attributes, including land tenure security, technical assistance, and financial compensation. A conditional logit model was first employed to establish baseline preferences, followed by a mixed logit specification that relaxes the independence of irrelevant alternatives assumption and captures unobserved preference heterogeneity. The results revealed trade-offs among incentives, with required compensation increasing from 27,250 kwanzas for biopesticide use to 81,749 kwanzas for combined agroecological practices. Secure land tenure emerged as the most influential determinant, reducing WTA by 344%, while behavioural factors, including pro-environmental attitudes and livelihood security perception, moderate farmers’ responses to policy incentives. Overall, the findings indicate that institutional support, particularly tenure security combined with targeted technical assistance, increases adoption and reduces reliance on monetary compensation. These results provide empirical evidence that land governance reforms embedded in supportive institutional frameworks can enhance the cost-effectiveness and scalability of agri-environmental and payments for ecosystem services (PES) programmes. The study offers policy-relevant insights for land-use and agricultural policy design in Sub-Saharan Africa and other regions facing comparable institutional constraints.
Rosário et al. (Fri,) studied this question.