Agriculture Survey on Zero-Tillage Technology in Bihar

Posted In:    Impact Assessment    Recent Projects    Bihar   

cimmyt-01.JPGZero-tillage (ZT) in wheat is one of CSISA's core supported technologies in the Bihar and Eastern Uttar Pradesh (EUP) hubs. In 2013, two independent but interlinked socioeconomic studies were conducted, one focusing on ZT adoption and its performance at the farm household level (Study 1) and one focusing on ZT service provision as a business opportunity (Study 2).

These two studies represented a snapshot at a time when ZT and related custom hire services were still quite new in Bihar. While the number of ZT SPs in Bihar increases since then the development of individual service provision businesses over time and the social inclusiveness of farmers' access to such services under a more competitive business environment are still unclear, as is the continuity of farmers' use of ZT over time and the welfare impact of the technology. Phase III of CSISA presents a valuable opportunity to shed light on such questions, which are of particular importance in view of the long-term sustainability and impacts of related CSISA interventions. Hence the current Study-3 aimed at revisiting the households who participated in Studies I and 2 conducted in 2013, in order to expand the cross-sectional datasets into panel data to elucidate the adoption dynamics of ZT wheat and the growth dynamics of related custom-hire services in Bihar.

Data was collected using Computer Assisted Personal Interviews (CAPI) among those 1000 wheat growing farm households in 40 randomly selected villages of 6 districts in Bihar (Begusarai, Bhojpur (Ara), Buxar, Lakhisarai, Samastipur, Vaishali), that participated in Studies 1 and 2 in 2013. ZT service provider survey was done with 250 ZT service providers in 185 villages of these 6 districts. The study adopted difference-in-difference methods for technology impact assessments, which allows control of not only of observable characteristics (as propensity score matching based on cross-sectional data), but also for unobservable characteristics, and therefore produce more reliable impact estimates.