A Technical Efficiency Reserch On Chinese State-owned Industrial Enterprises Under Conditional Of Scale Heterogeneity

China is in the process of rapid economic growth, the industry has always been the main driving force of China’s economic growth. In the process of industrialization in China, the transformation and upgrading of industrial structure is an important part of the transformation of economic development mode, the improvement of the quality of industrial development, the progress of the technological level has important practical significance to realize the sustained and stable growth of the industrial sector and the overall economy.Therefore, how to measure and analyze the technological progress of industrial enterprises precisely in China has become a hot issue in the academic field. The most widely used method in the field of efficiency research is the stochastic frontier model, which can be used to measure the technical efficiency and the production decomposition in the same time. Compared with other ones, the hypothesis of stochastic frontier analysis is more realistic. However, the stochastic frontier model has a premise assumption, that is, the sample companies must have a common technical frontier. The scale of industrial enterprises in China spread a large range, and the technical level is very likely to be heterogeneous, which may exist in more than one frontier. If we use stochastic frontier method to measure the efficiency of industrial enterprises directly in China, the efficiency result is likely to have a deviation,even completely wrong.Therefore, to examine the technological heterogenicity between the enterprises in the progress of China’s industrial is important to understand the operating conditions, the productivity structure, the policy formulation, and to realize the steady growth of China’s economy.In this paper, we start from the micro level, based on the Chinese industrial enterprise database, get and processing the enterprise data from 2000 to 2009, using the stochastic frontier model, taking the heterogenicity between the enterprise into accout by using a threshold representing enterprise scale We detect the heterogeneity of sample data, and then to measure the technical efficiency and decomposition productivity of the industry with significant heterogeneity, finally give the corresponding policy recommendations.The main conclusions are as follows:Firstly, there is a significant difference in the change of the industry, the average level and the variation trend of the different industries are different. And the same time efficiency level of the large and medium Technology Group always performs better than that of the small group. Second, the difference in the productivity structure is still evident in the large and medium size groups.The large and medium groups’frontier technical change contribute more to productivity, while small ones benefit more from the technical proficiency. Finally, the results indicates show that technological progress has become the most important driving force for productivity improvement...