Decomposing Partial Factor Productivity in the Presence of Input-Specific Technical Inefficiency: A Self-Dual Stochastic Production Frontier Approac

The present paper provides a theoretical framework for the decomposition of partial factor productivity in the presence of input-specific technical inefficiency. Based on Kuroda’s dual approach and using the theoretical foundations developed by Kopp, we decompose the growth rate of partial factor productivity into five sources, namely, changes in input-specific technical efficiency, substitution effect, technical change, the effect of scale economies and a homotheticity and input biased technological effect. The empirical model is based on a generalized self-dual Cobb-Douglas stochastic production frontier and on the methodological approach for measuring orthogonal input-specific technical efficiency suggested by Reinhard, Lovell and Thijssen. The model is applied to a panel data set of 723 cereal farms in Greece observed during the 1994-2003 cropping period obtained from FADN. The empirical results suggest that the labor productivity of cereal farms has been increased by 2.89 per cent annually. Technical change was found to be the main source of labor productivity (70.4%), while changes in technical efficiency also contributed significantly over the period analyzed (34.7%). On the other hand, substitution effect was found to affect negatively the rate of labor productivity (-14.2%).