Sustainable Agricultural Practice Adoption and Technical Efficiency among Smallholder Maize Farmers in Contrasting Agroecological Environments of Zambia

Abstract

This study examines the association between Sustainable Agricultural Practice (SAP) adoption and technical efficiency among smallholder maize farmers in two contrasting agroecological environments of Zambia: Mungwi District in higher-rainfall Agroecological Region III and Senanga District in drought-prone Agroecological Region I. Using cross-sectional household survey data from 582 maize-producing households, the study estimated technical efficiency with a stochastic frontier production model. The final complete-case stochastic frontier sample comprised 482 households. A pooled Cobb-Douglas stochastic frontier with a normal/half-normal error structure was used as the main model because its inefficiency component was statistically identified. In contrast, the Translog frontier and district-specific frontiers were retained as diagnostic robustness checks. Secondary fractional logit, interaction, and propensity-score matching analyses were used to examine associations between SAP adoption and predicted technical efficiency. Mean pooled-frontier technical efficiency was 0.401, with higher average efficiency in Mungwi (0.547) than Senanga (0.258). High SAP adoption and SAP count were negatively associated with technical efficiency in pooled models. However, district-level interaction results showed that this association was concentrated in Senanga and not statistically evident in Mungwi. The findings suggest that SAP adoption counts should not be interpreted as automatic indicators of short-run resource-use efficiency. SAP promotion needs to be linked to implementation quality, agroecological suitability, complementary input and water-management support, and location-specific extension services.