Deepening the Measurement of Technical Inefficiency in Private Farming in Georgia: Locally Parametric Regression

This study deepens the measurement of technical inefficiency in private maize farming in Georgia, applying iocelly parametric (LP) r eg ression method, whicb builds on the stochastic frantier production function appraach. Detailed survey data for 221 mixed farms for 1997 are used in the estimations. Findings suggest: (i) maize production can be further increased by breaking up large farms into smaller parcels; and (ii) increzsed schooling and farm experience of new private farm operators would further reduce the inefficiency. Furthermore, global and locai estimations of the inefficiency suggest different policy directions as to the future of maize farming. The global estimations, revealing non-disciriminant overestimation of the inefficiency, prejudice braad-based farm reforms. On the contrary, the local estimations, pcinting out relatively large farms as t,he kcy source of the inefficiency, favor the design of specific polities for the effective operation of lage farms.