Can ICT-based extension partially substitute for weak institutional service delivery? Evidence from smallholder dairy farms in Ethiopia.

Smallholder dairy farmers in Ethiopia face low productivity and food insecurity, in part due to limited use of artificial insemination (AI) and overstretched extension systems. Artificial insemination (AI) requires accurate estrus detection, timely communication, and quick technician response, which are difficult to achieve in rural areas with limited service. This study analysed how ICT-based extension is associated with partial substitution for weak institutional service delivery to improve AI adoption and use, milk productivity, and food security in central Ethiopia. We use survey data from 571 households and apply a double-hurdle model to analyse adoption and use intensity, an endogenous switching regression to analyse impacts on milk productivity and food security, and a structural equation model to examine mediation effects. ICT-based extension is associated with higher AI adoption and use, increased yields, and reduced food insecurity, with greater benefits for those with more education, cooperative membership, market access, and mobile coverage. ICT-based extension is associated with reduced coordination failures, but deeper structural constraints persist, highlighting the need for investment in education, connectivity, and targeted interventions.