Bias free measurement of technical efficiency

Technical efficiency, which is a measure of production performance of a firm, has been estimated generally using a primal production frontier. Since the estimation is carried out for a given level of inputs, the efficiency measure includes the effect of 'input-mix' or 'input-allocation' and consequently, the technical efficiency estimate is biased. The objectives of this paper are to gauge the magnitude of 'input-mix' bias in technical efficiency estimate and to suggest a method to measure technical efficiency eliminating the bias. The workability of the suggested method is demonstrated through an empirical analysis using agricultural data from India covering the period 1970-1993.