Technical Efficiency of Smallholder Wheat Farmers in Ethiopia: A Panel Data Approach

This paper aims to estimate the technical efficiency measures of wheat-producing farmers in Ethiopia using the stochastic frontier panel model. Data from 3482 farm households collected in two rounds of panels (2011 and 2014) was used to estimate the Translog stochastic frontier production function and factors influencing technical efficiency with a one-step maximum likelihood estimator. The production frontier function involves land, seed, inorganic nitrogen, pesticide, oxen power, and labor. The model showed that more than 95% of the total variation in output was a result of factors within the control of the farmer. The result also indicated that land, seed, nitrogen, and pesticide had significant positive effects on wheat output. Most sustainable agricultural practices and plot characteristics included in the production frontier had positive effects on wheat production. The covariates such as gender and education of the household head, credit access, and livestock holding were important in reducing the inefficiency of the wheat producers. However, land size was found to increase the inefficiency of wheat producers. The mean technical efficiency of 2011 and 2014 was 65.3 and 65.4 which was not significant implying limited technological and institutional progress in the wheat sector between the study time. Results revealed that on average wheat output can be increased by 35 percent without additional inputs. Improved access to direct inputs and identified environmental and socioeconomic factors are important in attaining a higher frontier in wheat production in Ethiopia.