First stage ordered selection model.

Minimum tillage (MT) is a sustainable farming practice that limit soil disturbance only to planting stations while leaving the rest of the soil undisturbed. It is an important component of conservation agriculture, which aims to raise agricultural productivity, improve the livelihoods of farmers and build resilient farming systems. Despite the growing empirical literature on its adoption and benefits, there is a paucity of empirical evidence on the heterogeneous effect of length of MT adoption on household welfare. This study uses plot-level and household data combined with geo-referenced historical weather data to provide microeconomic evidence of the impact of MT on maize yields, food and nutrition security, and farm labor demand in Ghana. We account for potential selection bias and omitted variable problems by using an ordered probit selection model to estimate two transition-specific treatment effects: from conventional tillage systems to short-term MT adoption and from short-term to long-term MT adoption. The empirical results show that longer cropping seasons of MT adoption significantly increases maize yields and dietary diversity by about 4.33% and 14.22%, respectively, and decreases household food insecurity and labor demand by 42.31% and 11.09%, respectively. These findings highlight the necessity of developing and implementing programs that promote and help smallholder farmers to sustain its adoption for longer cropping season.