Abstract: The efficient operation of science and technology finance has become one of the key fulcrums of China's innovation-driven development. In order to timely adjust and optimize the development modes of science and technology finance in various regions, super efficiency SBM model and Dagum Gini coefficient decomposition method are used to calculate and analyze the regional differences of the efficiency of science and technology finance in China. Based on the consideration of spatial factors, spatial Kernel density estimation is used to investigate the distributed dynamic evolution of the efficiency of science and technology finance in China. The results are as follows:The efficiency of science and technology finance in China fluctuates significantly and its spatial distribution is unbalanced, the main source of spatial difference is inter-regional difference; the efficiency distribution of each region has the multi-polarization trend of different degrees; in the eastern, central and western regions, the spatial spillover exists in the dynamic evolution of the distribution of science and technology finance efficiency, while the spatial spillover is not obvious in the northeast. Based on this, this paper puts forward policy suggestions on how to reduce the regional differences in the efficiency of science and technology finance and promote the coordinated development of science and technology finance in China.

Key words: science and technology financial efficiency, spatial difference, distribution dynamic, Kernel density estimation