Evaluation Of Efficiency Of China's Industry Based On The Improved Malmquist DEA Model

Industry is the pillar of economic development,and the proportion of industrial added value to GDP has been maintained at more than 30%.However,in the new normal economy and under supply-side structural reform progress,economic development of China has shifted from high-speed growth to medium-and high-speed growth recently.Therefore,it is important to evaluate the industrial development scientifically and effectively and adjust the industrial structure.Data Envelopment Analysis(DEA),as a data-driven efficiency evaluation tool,can effectively deal with the situation of multiple inputs and outputs of decision-making units(DMUs)in complex systems.It is widely used in the relative efficiency evaluation.At present,many scholars have applied DEA method to the performance evaluation of industry.In this paper,Malmquist DEA is optimized in two aspects: core function and factorization.Firstly,the core function of Malmquist DEA method is optimized based on the common weight.The core function of traditional Malmquist DEA usually uses CCR envelope model.However,CCR model tends to allocate the most favorable input-output weight for each DMU when evaluating the efficiency of DMUs.This self-evaluation system will not only cause the difference of comparison basis between DMUs and the problem that the efficiency values cannot be completely ranked,but also lead to DMUs avoiding self defects and the weight distortion.In view of the above problems,the common weight DEA makes up for the shortcomings of the traditional CCR model.Based on this,this paper optimizes the traditional common weight DEA model and constructs a new common weight DEA model based on "the priority of choosing weight".Taking the new common weight model as the core function,a biennial Malmquist DEA model based on common weight is constructed.The rationality of the new model is illustrated by an example,and the new method is applied to evaluate the efficiency of China's industry during the 12 th Five Year Plan period.Secondly,the effect of learning by doing is introduced to analyze the factors of Malmquist DEA model.The existing Malmquist DEA methods tend to regard technical change as a whole when measuring productivity changes.In fact,technical progress factors may be subject to different internal driving forces.This paper considers that technical change can be further decomposed into independent innovation and learning by doing effect.The concept of independent innovation and learning by doing effect are introduced into the Malmquist DEA model,which are decomposed and quantified to construct a new Malmquist DEA model based on the learning by doing effect.Based on the new model,this paper studies the changes of industrial productivity and learning by doing effect between provinces and regions in 2011-2015,and puts forward some effective suggestions for regional industrial development according to the evaluation results.The empirical study shows that during the 12 th Five Year Plan period,China's industrial productivity shows a first advanced and then backward trend,and the regional industrial productivity gap is obvious,among which the productivity of the northeast and East regions is significantly higher than the average level.It is worth noting that the driving force of technology in the process of industrial development is increasing year by year.As the key to technological change,the "learning by doing" effect shows a V-shaped trend of first decreasing and then increasing in 2011-2015.However,different regions are affected by the effects of research and development and "learning by doing".The northeast region is more sensitive to the effect of "learning by doing".Based on the above conclusions,this paper puts forward the following four suggestions around technological progress.First,optimizing the internal structure of the industry,phasing out low-tech industries and vigorously developing high-tech industries.Second,attaching importance to the role of technological change under innovation driven development strategy,including the important role of independent innovation and "learning by doing".Third,strengthening regional industrial integration,improving and optimizing industrial geographical pattern,and reducing geographical isolation of technological innovation.