Heterogeneous treatment effect estimation of participation in collective actions and adoption of climate-smart farming technologies in South–West Nigeria

• Published in: GeoJournal 2020-10, Vol.85 (5), p.1309-1323
• Main Authors: Olawuyi, Seyi Olalekan, Mushunje, Abbyssinia
• Format: Article
• Language: English
• Subjects: Agricultural conservation; Agricultural extension; Agricultural practices; Agricultural production; Agricultural technology; Agriculture; Bias; Climate; Climate-smart agriculture; Collective action; Conservation; Cooperation; Data processing; Degradation; Digital agriculture; Environmental Management; Farmers; Farming; Food; Food production; Food security; Geography; Hazards; Human Geography; Land degradation; Negative selection; Participation; Process engineering; Productivity; Propensity; Sensitivity analysis; Small farms; Social Sciences; Variability; Workers
• ISSN: 0343-2521; 1572-9893
• DOI: 10.1007/s10708-019-10024-2

Enhancing sustainable food security requires agricultural production systems to change in the direction of higher productivity and to mitigate lower output variability in the face of climate extreme related hazards such as land degradation. Adoption of resilient food production system capable of withstanding disruptive events is therefore needed to stabilize farmers’ productivity. Consequently, participation in collective actions has been touted as an effective approach to enhance cooperation among individuals within a social system and to advance adoption of climate-smart farming techniques (conservation agriculture). This study investigated this perspective using heterogeneous treatment effects estimation to analyze the data collected from 350 smallholder farmers selected randomly from the South–West Nigeria. The average treatment effects on the treated estimate revealed that participation in collective actions had adoption-increasing effect for each unit variation in propensity score rank, although, negative selection effect was suspected. Similarly, information acquisition, access to extension service and frequency of visit by extension workers are significant features that predict adoption in the study area. However, rosenbaum sensitivity analysis test revealed that the increasing effect of participation in collective actions on conservation agriculture adoption is insensitive to unobserved bias that may double or triple the odds of exposure to treatment. Hence, the average treatment effect on the treated estimate is a pure impact of the participation in collective actions. The study concluded that farmers with high propensity to participate in collective action have high likelihood to adopt climate-resilient farming practices compared to the counterparts with a lower propensity of participation in collective action.