| Since the proposal of “Industry 4.0”,a new round of industrial transformation have been promoted around the world.In this context,the impact of efficiency evaluation is becoming more and more significant.In the existing industrial efficiency evaluation,the related studies based on Data Envelopment Analysis(DEA)occupy the majority and have achieved certain results.However,there are still two main problems in these studies.On the one hand,the evaluation perspective is single,which focuses on the “region” or “industry” and selects provinces,regions,or some subdivided industries as the research objects,so the research results are difficult to effectively link the development of various regions and subdivided industries.On the other hand,most of these studies regard the industrial production system as a traditional “black box”.Although some studies explore the internal composition of the system by the network structure,the research on the hierarchical structure of the industrial production system is blank.In this case,this paper proposes a dual evaluation perspective by combining “region” and“hierarchical subdivided industry”,taking the provincial industrial system as the decision-making units(DMUs)and constructing the hierarchical evaluation structure of the provincial industrial system according to the Industrial Classification for National Economic Activities(GB/T 4754-2017).For such DMUs with hierarchical characteristics,this paper puts forward three improvements to the corresponding Hierarchical DEA model.Firstly,this paper introduces the shared inputs and realizes their reasonable allocation among subunits at all levels.Secondly,this paper combines with the global Malmquist index to evaluate the productivity of the hierarchical DMUs and their internal subunits.Thirdly,the hierarchical DMUs are grouped by K-Means clustering algorithm to eliminate the environmental heterogeneity as much as possible,and then the efficiency evaluation of each cluster is carried out.Finally,the improved Hierarchical DEA models are used to measure and evaluate China’s 29 provincial industrial systems above designated size from 2012 to 2016.The empirical study shows that,during the 12 th Five Year Plan period,the efficiency improvement of industrial systems in China is not strong,and there are regional differences in the efficiency scores,which is inversely related to the proportion structure of R&D expenditure allocation.In terms of productivity development,the differences and consistency of regional industrial productivity coexist,and the improvement of productivity in three industrial sectors all rely on leading industries.It is worth noting that the production technology change index is the key driving force of industrial productivity,and the improvement of production technology is affected by the same direction of absolute independent innovation and learning by doing effect,of which the latter contributes more.In addition,the clustering provides a more reasonable evaluation environment for each province,which promotes the parallel development between the quantity of provincial industrial achievements and the quality of industrial development.Based on the above conclusions,this paper puts forward five suggestions: first,take the manufacturing industry as the starting point,to drive the overall efficiency of the industrial system.Second,all provinces should improve efficiency and productivity in combination with the development of the three industrial sectors and promote regional coordinated development.Third,in the production process,we should give overall consideration to the improvement of production technology and technical efficiency.Fourth,all industrial sectors should pay full attention to the leading industries to drive the overall productivity.Fifth,all provinces should take the clustering efficiency evaluation as the benchmark to accurately set the learning benchmark,and comprehensively improve the efficiency of industrial development and the quantity of provincial industrial achievements. |
PDF Full Text Request