空间相关性、创新生态环境与高技术产业创新生态系统创新效率——基于中国23个省份的实证研究

构建两阶段共享投入关联DEA模型，测算中国内地23个省市样本期整体及各子阶段创新效率，并从时间、空间维度进行对比分析，构建空间Tobit模型进一步分析创新生态环境对其创新效率的空间影响机理。研究表明，样本期整体与创新研发阶段效率均处于中等水平且呈M型变化趋势，而成果转化阶段效率较低，呈W型变化趋势；创新生态环境与创新效率具有强正向空间相关性，说明地理邻近空间存在知识溢出效应；在整体创新阶段，劳动力素质、政府支持均与创新效率显著负相关，而市场结构与创新效率正相关；在研发创新阶段，人均GDP对创新效率具有显著负向影响，创新基础设施、对外开放水平具有显著正向影响；在成果转化阶段，除对外开放程度与创新效率负相关外，其余影响因素均与创新效率正相关。

Spatial Correlation,Innovation Ecological Environment and Efficiency of Industry Innovation Ecosystem：an Empirical Study based on High-tech Industrial of 23 Provinces in China

The high-tech industry is the embodiment of a country's international core competitiveness, and it is in the leading position of the overall economic development strategy. At present, under the guidance of the concept of innovation and development, China's high-tech industry and technological innovation have achieved leap-forward development. However, it is also an indisputable fact that the original innovation ability is insufficient and the key core technologies have not yet made a breakthrough. Especially, the current situation of globally spread of the COVID-19 epidemic, the rise of trade protectionism, and the security risks in the industrial chain supply chain pose great challenges to China's high-tech industry. The Fourth Plenary Session of the 19th Communist Party of China Central Committee proposed the statement that "enterprises and all subjects should integrate innovation ... to form a good innovation ecosystem with the proper place, mutual cooperation, and mutual support". Therefore, it is urgent to carry out in-depth research on the innovation ecosystem of high-tech industries under the new development pattern.#br#The previous studies only focus on the whole innovation ecosystem, ignoring the correlation of the internal elements of the system, especially the lack of research on high-tech industries. Based on the new era background of building the new development paradigm, this paper focuses on the interior of the innovation ecosystem of high-tech industries and examines the relationship between innovation environment and innovation efficiency within the system from the perspective of open innovation. A comprehensive and scientific evaluation system of innovation efficiency index based on the perspective of innovation ecological environment is constructed. At the same time, the geospatial perspective is introduced to comprehensively investigate the correlation in between. Studying the influence mechanism of innovation ecological environment in high-tech industrial innovation system on innovation efficiency can provide theoretical support for the study of system operation mechanism, and at the same time provide solutions for better playing the overall role of the system, improving the development quality of high-tech industries, coping with the complex environment and risk challenges at home and abroad and realizing high-quality economic development. #br#Based on the input-output index data of 23 major cities in China from 2003 to 2018, this paper constructs a two-stage shared additional input correlation DEA model to quantitatively measure the innovation efficiency of China's high-tech industry innovation ecosystem and makes a comparative analysis from the perspectives of time, region and value chain. Based on the measurement results of innovation efficiency, this paper constructs an index system of influencing factors of innovation efficiency based on innovation ecological environment, establishes panel Tobit model and spatial Tobit model respectively to quantitatively analyze the influencing factors of overall and two-stage innovation efficiency, analyzes their spatial correlation through Moran'I index of each variable, comprehensively analyzes and compares the above two empirical methods, and sums up the empirical analysis results. Finally, based on the above conclusions and combined with the new era background, the paper puts forward some policy implications to improve the innovation efficiency of China's high-tech industry innovation ecosystem from the following aspects: laying out high-level innovation infrastructure, creating a good environment for innovation and entrepreneurship, strengthening the optimization and integration of regional high-tech industries, establishing a high-level collaborative innovation system, insisting on the combination of innovation and independent innovation, realizing a higher level of opening to the outside world, giving full play to the advantages of the new national system, realizing the synergy between government guidance and market leadership, paying attention to original innovation and application transformation of achievements, and promoting the coordinated development of industrial chain innovation chain.#br#The findings are as follows. (1) The measurement results of innovation efficiency show that the overall innovation efficiency and innovation R&D efficiency of China's high-tech industry innovation ecosystem are at a medium level from 2004 to 2018, while the efficiency of achievement transformation is low. Among them, the change of overall innovation efficiency and achievement transformation efficiency is consistent-both show an M-shaped change trend, and the innovation R&D efficiency shows a W-shaped change. It can be seen that the efficiency of achievement transformation dominates the overall innovation efficiency. (2) The provinces with relatively high innovation efficiency in China's high-tech industry innovation ecosystem as a whole and in two stages are mainly concentrated in first-tier cities such as Beijing, Shanghai, and Guangdong, and coastal provinces such as Jiangsu, Zhejiang, and Shandong. In the four major regions of China, the overall stage efficiency shows a decreasing distribution trend from east to west. (3) By constructing the two-dimensional distribution map of innovation R&D efficiency and achievement transformation efficiency, it is found that the provinces that fall into areas with low innovation efficiency in a certain stage can improve their overall and two-stage efficiency through a breakthrough way of " shore up our weak spots"; however, for the areas with low efficiency in both stages, we can adopt theway of "making up for each other's weaknesses" to promote the path gradually. (4) The regression results of comprehensive panel Tobit and spatial Tobit show that spatial correlation significantly inhibits the improvement of innovation R&D efficiency of the high-tech industry innovation ecosystem, but it has a strong positive spatial spillover effect on achievement transformation efficiency, and the overall innovation efficiency is unbalanced in regions; Labor quality, market structure, and government support are the strong influencing factors to explain the overall efficiency. Per capita GDP, innovation infrastructure, and the level of opening to the outside world are the strong influencing factors to explain the efficiency of innovation R&D; Per capita GDP, innovative infrastructure, entrepreneurial environment, opening-up level, and market structure are the strong influencing factors in the stage of achievement transformation.#br#