Heterogeneous effects of horticulture commercialization and gender-based decision-making on smallholder farmers’ income: Evidence from a quasi-experimental study in Jimma zone, Ethiopia

Horticulture commercialization and gender-based decision-making are valuable contributors to smallholder farmers' income. Many existing studies measure horticulture commercialization and gender-based decision-making using average treatment effects, while only a few examine heterogeneous effects. Investigating distributional effects can offer policy implications for targeted interventions to improve smallholder farmers' income. This study uses quantile regression to assess the heterogeneous effects of horticulture commercialization and gender-based decision-making on income in the Jimma zone using a quasi-experimental design sampling method with location-specific data. This study applies propensity score matching at two levels before quantile regression estimation. First, nearest-neighbor matching is conducted based on key agricultural characteristics to identify control groups at the village (kebele) level. Second, kernel matching is employed at the household level to address selection bias. Then, a total of 610 farmers were randomly sampled; among these, the treated group included farmers who were participants in the Smallholder Horticulture Empowerment and Promotion (SHEP) project between December 2022 and February 2023. The interaction effects between horticulture commercialization and SHEP intervention demonstrate positive heterogeneous effects on horticulture income, notably more pronounced at equal to or lower than 50^th quantiles. Results revealed joint decision-making significantly improves farmers' income at the 10^th and 25^th quantiles; however, the intervention did not provide an added effect to joint decision-making on income, indicating the effects of the SHEP intervention on joint decision-making could take longer to observe, as the gender training primarily focuses on behavior and mindset changes, which require a longer time to be fully internalized. Family labor, livestock holding, and proximity to agricultural cooperatives significantly correlate with income and support those in the lower quantile of income from horticulture production. Thus, for the lower-quantile farmers to benefit from the project, an effective market-driven extension approach such as SHEP is crucial to improve smallholder farmers’ income through horticulture production. More attention is needed to promote and strengthen agricultural cooperatives regarding facilities and capacity building to enhance the effective access of essential inputs to farmers especially for low quantile farmers. Moreover, this methodological innovation could be a valuable reference for future quasi-experimental and impact evaluation studies employing similar approaches in different contexts.