Dynamic Agricultural Technology Adoption Models: Theory and Evidence from Nigerian Smallholder Farming

This study develops and empirically tests a dynamic technology adoption model for agricultural innovations in smallholder farming contexts, contributing an original theoretical framework that addresses key limitations of existing static and threshold adoption models. Static expected utility adoption models and threshold adoption frameworks fail to capture the sequential, information-driven, and reversible nature of technology adoption decisions observed in Nigerian farming communities. This study proposes a dynamic discrete choice model of technology adoption that incorporates information updating through learning-by-doing and social learning, option value of waiting under uncertainty, and switching costs between technologies. The model generates predictions about adoption timing, trial behaviour, technology accumulation, and abandonment that are empirically distinguishable from static model predictions. Empirical testing uses a novel six-year agricultural panel survey of 320 smallholder farmers in Benue, Kaduna, and Ogun States, specifically designed to capture adoption dynamics including trial, scaling, and abandonment episodes. Structural estimation of the dynamic model uses nested fixed point algorithms. Findings strongly support the dynamic adoption framework, with social learning and option value explaining a significant share of adoption timing variation that static models cannot account for. Extension demonstration plots generate the largest social learning effects. The study contributes an original dynamic adoption theory and estimation methodology for Nigerian agricultural contexts and recommends information-intensive extension approaches that accelerate social learning.

Keywords: dynamic adoption models, technology adoption, smallholder farming, Nigeria, social learning.