Patterns, Intensity, and Determinants of Modern Rice-farming Technology Adoption in Rajshahi, Bangladesh

Abstract

Rice farming in drought-prone north-western Bangladesh is experiencing an uneven transition from mature mechanisation in land preparation and post-harvest operations to selective adoption of advanced labour- and water-saving technologies. This study examines the patterns, intensity, and determinants of modern rice-farming technology adoption in Rajshahi District using primary household survey data from 250 rice-farming households across nine upazilas, collected between January and April 2024. Adoption intensity was measured as a count of six advanced technologies: combine harvesting, mechanical rice transplanting, direct seeding/seed sowing, laser land levelling, power weeding, and alternate wetting and drying (AWD). Descriptive statistics, cluster-robust Poisson regression, a binary logit robustness model, and technology-specific logit models were applied. Mature mechanisation was widespread, whereas advanced adoption remained limited: the mean advanced-technology count was 0.66 out of six, and 45.6% of households used at least one advanced technology. Farm size, labour scarcity, water scarcity, machinery training, extension contact, machinery knowledge, custom-hiring availability, perceived usefulness, and technology compatibility were positively associated with adoption intensity, while greater distance to machinery services and perceived cost barriers were negatively associated. Combine use was most strongly associated with farm size, labour scarcity, and service proximity, whereas AWD use among Boro growers was associated with water scarcity, irrigation cost, and extension contact. The findings suggest distinct service-access and knowledge-and-water-management pathways for technology adoption in the Barind environment.