Effects of adopting improved bread wheat varieties on smallholder farmers’ yields: a case study of Duna District, Central Ethiopia

Wheat ( Triticum aestivum L.) is a key food and cash crop for smallholder farmers in Ethiopia, yet national yields remain low. Adoption of improved bread wheat varieties is widely promoted to boost productivity, but institutional, socioeconomic, and demographic constraints limit uptake. This study examines the determinants of adoption of improved bread wheat varieties and their association with smallholder yields in Duna District, Hadiya Zone. A multistage sampling procedure selected 264 farm households, including 119 adopters (45.1% of the sample) and 145 non-adopters. Data were analyzed using a binary logit model to identify factors influencing adoption, propensity score matching (PSM) to estimate yield effects, and descriptive statistics. The logit results show that sex of the household head, landholding size, livestock ownership, extension contact, education level, training participation, and cooperative membership were positively and significantly associated with adoption (all effects reported in percentage points of predicted probability, not percent). Access to credit had a significant negative association. Using kernel (Epanechnikov) matching with a bandwidth of 0.01, PSM findings indicate that the average treatment effect on the treated (ATT) corresponds to a 15.5% relative difference in wheat yield between matched adopters and non-adopters (5.59 quintals ha⁻¹, equivalent to about 18.4% relative to the matched control group’s yield); this difference is conditional on the observable covariates included in the matching model, and its robustness to potential unobserved confounding is examined through a Rosenbaum bounds sensitivity analysis. The study recommends policies that empower female-headed households, strengthen education and literacy, expand extension services, and promote training and cooperative participation. Improving livestock productivity, land management, and further investigation of the credit-adoption relationship can inform future interventions. These findings are specific to Duna District and should be interpreted with appropriate caution when generalizing to other agroecological or institutional settings; broader claims about food security and rural livelihoods would require additional evidence beyond wheat yield alone.