Research On Improved Dea Method Of Efficiency Evaluation For Technological Innovation In High-Tech Industry

With the development of high-tech industry and the increasing inputs and outputs in the process of technological innovation, technological innovation becomes the key factors to determine the long-term sustained development of high-tech industry and influence regional economic growth and regional competitiveness. As the scale of innovation resources is often restricted to the objective conditions such as level of regional economic development, base of regional knowledge and so on, therefore, an important approach to increase innovation outputs is to improve the input-output efficiency of technological innovation in the condition of available input scale of innovative resources. First of all, we should to solve the problem that how to evaluate the technological innovation efficiency in the high-tech industry in the process of improving technological innovation efficiency.Generally speaking, traditional production process meets the law of decreasing marginal returns while many scholars have different views on whether or not it still works in the knowledge-based economy. Scholars have discussed the relationship between knowledge and return increasing and mentioned the phenomenon of increasing marginal returns in the knowledge-based economy from a qualitative point. The technological innovation in high-tech industry has obvious characteristics of knowledge-based economy; therefore, it's necessary to reconsider if the law of decreasing marginal returns works universally in the technological innovation in high-tech industry innovation of. The DEA model is an important instrument to evaluate the efficiency, in which the convexity of feasible production sets implies a precondition of decreasing or unchanging marginal returns. If the phenomenon of increasing marginal returns exists for the research objects, the feasible production set doesn't meet the convexity assumption so that the frontier decided by the DEA model will not accord the facts, which will lead to the unjust results and even result in a serious evaluation of the irrational. Therefore, it's necessary to do some research on how to identify the marginal return characteristics from a quantitative point and explore a new method of efficiency evaluation in the condition of decreasing marginal returns. In this paper, we start the research according to above problems as following:After clearly defining the relative key conceptions, we decompose the high-tech innovation process into technological development process and technological transformation process from the input-output perspective. On this basis, technological innovation efficiency can be decomposed into technological development efficiency and technological transformation efficiency. We also analyze the input factors and out factors and design the input-output index systems in each stage respectively.According to the limitation in DEA application, we present the interrelated conceptions of feasible innovation sets. Base on these conceptions, we present the methods to identify the marginal return characteristics and to evaluate the technological innovation efficiency in high-tech industry. The feasible innovation sets has no restriction of convexity so that it describe the input-output activities in conditions of various characteristics of marginal returns. The innovation frontier reflects the actual frontier of innovation activities and the outputs of the projections can respect the optimum outputs the units can achieve so that the new evaluation methods can apply not only to condition of decreasing or unchanging marginal returns but also to condition of increasing marginal returns. In addition, through the expansion of two-stage sequence DEA model, the new evaluation method can not only consider the marginal return characteristics in each stage but also consider the impact of the internal structure in technological innovation in the whole process. In order to eliminate the influence of regional gifts on the technological innovation in the high-tech industry, we develop a new new model to evaluate the efficiency based on the three-steps model given by Ruggiero.Finally, we analyze the characteristic of marginal returns in each stage and also acquire technological development efficiency, technological transformation efficiency and overall efficiency for each region in each year in the empirical research.The research shows that marginal returns are constant in smaller scale and marginal returns are incremental in larger scale at technological development stage and there no characteristics of increasing marginal returns at technological transformation stage. The new evaluation method conquer the shortage of traditional DEA model evaluating technological innovation efficiency so that the results is more rationale and scientific. Therefore, the research we have done is meaningful in theory and practice, which provide a new idea and method about the issues of efficiency evaluation of technological innovation in high-tech industry and provide some theoretical and practical references to improve the technological innovation efficiency and increase the competitiveness in high-tech industry.