Currently, only about 4 percent of Sub-Saharan Africa’s arable land is irrigated; the rest is rain-fed, meaning it is susceptible to droughts and floods. Yet, irrigated land can have yields that are up to five times those of rain-fed areas. It must be the case that the costs of irrigation—capital, recurrent, administrative, political—are sufficiently high to outweigh these benefits. But if you take into account the possibility of more frequent floods and droughts, which would make irrigated land relatively more attractive, does the benefit-cost calculation change?
The short answer is yes. In a calculation for the Zambezi basin, Aziz Bouzaher and I estimate that the costs of tripling the irrigated area are about equal to the benefits—if you ignore the effects of climate change. It is not surprising therefore that there has not been much investment in irrigation. But when you include as benefits of irrigation the avoided damage from increasingly frequent droughts (using fairly conservative assumptions), the overall benefits are double the costs. Recognizing that the effects of climate change will increasingly affect rain-fed agriculture may tip the scales in favor of more irrigation in Africa, and lead to higher yields for African farmers.
More information on the costs and benefits of irrigation in the Zambezi River Basin (PDF)