Abstract/Summary

This study analysed exposure to improved rice varieties, and the effect of adoption on output and net rice income, whilst separating production technology gap from technical inefficiency in rice cultivation of 576 Ghanaian households using 2012/2013 production data. This was complemented with qualitative interviews to assess rice varietal diffusion, access and adoption, farmers’ perceptions of varietal traits and constraints to rice cultivation. Exposure to improved rice varieties was estimated to account for non-exposure bias followed by determinants of adoption using treatment effect. A stochastic metafrontier was estimated to separate productivity differences due to technology gaps from technical inefficiency after correcting selection bias. Adoption under incomplete exposure under-estimated the adoption rate as 55.9%, producing a non-exposure bias of 11.3%. The exposure rate and adoption rate of improved rice varieties were 82.5% and 67.2%. Community participation in rice projects, colleague farmers, agricultural extension agents and input dealers were sources of knowledge about improved rice varieties. Adoption was positively influenced by rice projects, participation in model and block farming, agricultural extension, higher rice yield motive, and cultivating irrigated rice. Traditional varieties were cultivated because of localized market demand and perceived resistance to bird infestation due to longer maturity period. Training rice processors on correct parboiling of jasmine 85 can increase its consumption in the local market. Meanwhile, seed, farm size, fertilizer, labour and herbicides application increased rice output of adopters whereas farm size and fertilizer had positive effect on the output of non-adopters. The mean difference in metafrontier technical efficiency of adopters (42.7%) and non-adopters (44.5%) were statistically not significant, although adopters had a higher metatechnology ratio of 0.91 compared with 0.79 for non-adopters. Thus, non-adopters were behind in applying the best available technology represented by the stochastic iii metafrontier. Adoption increased net rice income per hectare by GH¢374.6, whereas the potential gain if the non-adopters had adopted would have been GH¢867.5. Agricultural extension, controlling plot water levels and weeding twice using herbicides increased the technical efficiency of adopters. Applying ammonia fertilizer and weeding increased the technical efficiency of non-adopters. The original contribution of this study is using nationally representative plot level data to establish that exposure to improved rice varieties and subsequent adoption increased the net rice income of smallholder Ghanaian farmers through increased yield, whilst weed and bird infestation, labour constraints, intermittent flash flooding and drought hampered rice cultivation.