Soil and water conservation technologies: a buffer against production risk in the face of climate change? Insights from the Nile basin in Ethiopia

"This study investigates the impact of different soil and water conservation technologies on the variance of crop production in Ethiopia to determine the risk implications of the different technologies in different regions and rainfall zones. Given the production risks posed by climate change, such information can be used by decision makers to identify appropriate agricultural practices that act as a buffer against climate change. Using a household- and plot-level data set, we apply the Just and Pope framework using a Cobb-Douglas production function to investigate the impact of various soil and water conservation technologies on average crop yields and the variance of crop yields, while controlling for several household- and plot-level factors. Results show that soil and water conservation investments perform differently in different rainfall areas and regions of Ethiopia, which underscores the importance of careful geographical targeting when promoting and scaling up soil and water conservation technologies. We find that although soil bunds, stone bunds, grass strips, waterways, and contours all have very significant positive impacts on average crop yields in low-rainfall areas, only soil bunds have significant risk-reducing effects in these areas with low agricultural potential. We also find that irrigation and use of improved seeds have insignificant risk-reducing effects in low-rainfall areas, suggesting that—as currently implemented—these interventions may not be appropriate adaptation strategies for these environments. Regionally, in the low-rainfall areas we find significant spatial heterogeneity, with soil bunds being risk reducing in Oromiya and Amhara, and stone bunds, grass strips, and waterways being risk reducing in the Southern Nations, Nationalities, and Peoples Region. Irrigation was only risk reducing in the high-rainfall areas of Benishangul-Gumuz. These results remain robust even after controlling for the major crops grown on the plot. Res(This abstract was borrowed from another version of this item.)