Modelling technical efficiency and technology gap in smallholder maize sector in Ethiopia : accounting for farm heterogeneity

This paper examines the effect of farm heterogeneity on the technical efficiency and technology gap of maize production in Ethiopia. We utilize comprehensive household-level data in 2010, 2013 and 2015 for three major maize producing regions in Ethiopia. We estimate technical efficiency using a bias-corrected stochastic meta-frontier model that accounts for technological differences and the consistent 'true' fixed-effects stochastic frontier model, which disentangles unobservable farm heterogeneity from inefficiency. We find that maize farmers produce below their potential output with the technology available and the prevailing environmental conditions in their regions. The results show that Oromia has the highest technical efficiency, followed by SNNP, and Benishangul-Gumuz is the least efficient in maize production. These findings generate important policy implications for achieving increased technical efficiency and reduce the technology gap in maize production. Moreover, the level of education, age of household head, family size, household wealth, and off-farm employment and extension services significantly affect technical efficiency in maize production in Ethiopia.