Mr. Adoption of Climate-Smart Agriculture Among Rice Farmers in Bangladesh

Abstract

Rice-based smallholder systems in southern Bangladesh face reinforcing pressures from climate variability and soil degradation driven by decades of input-intensive cultivation. This study examines Climate-Smart Agriculture (CSA) adoption among modern rice farmers in this region, with particular attention to joint adoption behavior and variation in adoption intensity. Rather than treating CSA as a discrete technological choice, this study conceptualizes it as a portfolio of interrelated practices such as crop diversification, agroforestry, integrated soil fertility management, small-scale irrigation, and integrated pest management, and estimates adoption decisions jointly using a multivariate probit (MVP) model. An adoption index complements the MVP estimates by capturing depth of engagement beyond binary indicators. The study found that CSA adoption is widespread but uneven. Most households adopt multiple practices.  However, full portfolio adoption is rare. This is consistent with a gradual and selective adjustment process rather than systemic transition. MVP estimates confirm that decisions across practices are interdependent but not perfectly correlated. Determinants of CSA adoption differ significantly across practices. Human capital, farm size, access to credit and extension services, and market information impact adoption. But their effects are not uniform. Exposure to soil salinity and drought increases adoption of specific practices while leaving others largely unaffected. This indicates that agroecological context mediates the incentive structure rather than operating merely as background condition. Climate awareness alone does not sustain adaptation. Institutional support determines whether awareness translates into behavior change. These findings suggest that policy frameworks predicated on technology dissemination alone are unlikely to expand the adoption frontier. Effective intervention requires explicit attention to institutional capacity, localized environmental constraints, and heterogeneity in adoption intensity across households.