AGRICULTURAL ECONOMICS Agricultural Economics 32 (2005) 141–150 Yield and income risk-efficiency analysis of alternative systems for rice production in the Guinea Savannah of Northern Ghana Augustine S. Langyintuo a , Emmanuel K. Yiridoe b, ∗ , Wilson Dogbe c , James Lowenberg-Deboer d a CIMMYT-Zimbabwe, P.O. Box MP 163, Mount Pleasant, Harare, Zimbabwe b Department of Business and Social Sciences, Nova Scotia Agricultural College, P.O. Box 550, Truro, Nova Scotia, Canada c Savanna Agricultural Research Institute, Tamale, Ghana d Department of Agricultural Economics, Purdue University, 1145 Krannert Building, West Lafayette, IN 47907-1145, USA Received 9 April 2002; received in revised form 12 December 2002; accepted 19 May 2003 Abstract Risk efficiency of rice grain yield and returns to farm operators’ household resources generated from an improved short- duration cover crop fallow system were compared with (traditional) natural bush fallow, and continuous rice-cropping systems. The improved fallow system involved maintaining Calopogonium mucunoides, seeded into a natural bush fallow for 2 years before planting to rice. With no chemical fertilizer application, which reflects farmers’ practice in the area, average grain yield for continuous rice (1,185 kg/ha) and the cropping sequence incorporating a natural bush fallow (1,175 kg/ha) did not differ, but were higher for the improved fallow system (1,304 kg/ha). This suggests that nutrient contribution from the leguminous cover crop made up for critical crop N requirements in the improved fallow. Stochastic dominance of grain yield distributions from the improved fallow system, relative to the other two cropping systems, was more dramatic with no N fertilizer application compared to treatments with 30 kg/ha N. Average returns were highest for the improved fallow system, followed by the natural bush fallow-cropping system, and then continuous rice, under the no N fertilizer treatment regime. With 30 kg/ha N fertilizer, income risk efficiency was less clear (compared to treatments with no N fertilizer), especially between continuous rice and the improved fallow treatment, because of faster N mineralization effects on continuous rice. In contrast, the improved cover crop fallow system completely dominated the natural bush fallow treatment under both fertilizer regimes. Rice production systems that incorporated the leguminous cover crop fallow were superior to the natural bush fallow system, based on both grain yield and average farm income risk-efficiency considerations. JEL classification: D81; Q12 Keywords: Rice; Cover-crop fallow; Stochastic dominance; Ghana 1. Introduction This study compares the risk efficiency of re- turns to farm household resources generated using a ∗ Corresponding author. Tel.: 902-893-6699; fax: 902-897-0038. E-mail address: eyiridoe@nsac.ns.ca (E. K. Yiridoe). short-duration improved leguminous cover crop fallow system, with two alternative rice production systems involving traditional natural bush fallowing, and a continuous rice-cropping system. Farmers in Ghana mine soil nutrients from farm fields to main- tain low levels of crop production (Rhodes, 1995). The gap between current levels of chemical fertilizer