The Joint Measurement Of Technical And Allocative Efficiency Of The Industrial Enterprise In Guangdong Province

Cost efficiency can be decomposed in to technical efficiency and allocative efficiency.Technical efficiency concerns the success of minimization of inputs given production andinputs proportion. Allocative efficiency is the effectiveness with optimal allocation of inputsaccording to market price. This paper aims to estimate the technical and allocative efficiencyfor industrial enterprise of GuangDong province.First, the paper reviews the methodology about the decomposition of cost efficiency,including the DEA, the parametric estimation and the most popular SFA. Given the featuresof each method and the empirical data collected, the paper choose the SFA as the basis ofestimation. There is a long standing econometric difficulty known as Greene Problem in thestudy of the cost efficiency decomposition. And lots of solutions were given for this problem.Bayesian statistics has an advantage over the complicated Greene problem, so the paper’sempirical research takes Bayesian approach. Second, the paper compares those models aboutcost decomposition through pre-estimation, which are the Schmidt specification, theKumbhaka specification, and the Brissimis, Delis, Tsionas simplified models. And find thatonly the Schmidt specification is appropriate for the empirical data. Thirdly, follows the lineof Schmidt, a system of equations in the form of translog cost function and cost shareequations is set up with some improvement. Finally, with Bayesian approach, the model isapplied to a large panel of4454industrial enterprises with5years’ data. The main results arethat up to90.383%of those4454enterprises have allocative efficiency higher than90%. Andthe main reason of cost inefficiency is due to technical inefficiency.The main contribution of this paper lies in follows: First, this paper compares those threemodel specification through empirical estimation, and make clear their preconditions. Second,the paper improves the Schmidt specification and gives the estimated values of allocative andtechnical efficiency for industrial enterprise of GuangDong province. Thirdly, the paperexplores the implantation of Bayesian analysis for simultaneous equations of stochasticfrontier models.