Abstract: “Innovation leads development” has become a broad consensus. Chairman Xi pointed out that technological innovation is not only conducive to the Chinese breakthrough of major key core technologies in the field of strategic competition, but also conducive to opening up new industrial development directions and key areas, and cultivating new economic growth points. As the main body of basic research, universities are the main source of major original innovation achievements. They shoulder the four basic functions of talent cultivation, scientific research, social service and cultural inheritance, which are of vital importance in the process of national innovation system building. For the innovation activities of universities, economic value and social value are the value of university innovation activities. With the expression of university innovation activities in the economic development and the enhancement of regional innovation ability, the core importance of university innovation——social value is becoming the key to the construction of an innovative country. However, the existing research mainly focuses on the economic value efficiency of university innovation and lacks deep concern about the efficiency of social value.
Under this background, this paper firstly defines the social value of university innovation as the indirect, altruistic, non-market and common value benefits of the university through the research and innovation activities to meet the needs of enterprises, government and society. Then, based on the synthetical perspective of economic value and social value, the DF-DEA model is used to measure the efficiency of economic value and social value of universities innovation, and the dynamic evolution process of efficiency and the trend of efficiency change are analyzed. Finally, the panel Tobit model is employed to analyze the influencing factors of university innovation efficiency.
The research shows that:(1) From 2011 to 2015, the efficiency of economic value of provincial university innovation in China is relatively high but fluctuating much, and the efficiency of social value is relatively low but stable. Specifically, the regional division of the two types of efficiency is very different. Very few regions are divided into high-efficiency regions with economic value efficiency, and there are more regions with high social value efficiency. From the perspective of different value types, the provinces with medium efficiency are relatively large and widely distributed, covering different parts of China’s eastern, central and western regions. The perspectives of low-efficiency regions vary greatly from the perspective of different value types, and most of the economically inefficient regions are concentrated in the west, while the areas with low social value are mainly concentrated in the eastern region.
(2)Further analysis with kernel density estimation found that the distribution curve of the nuclear density estimation of the innovation efficiency of economic value of provincial universities in 2011-2015 moved to the right as a whole, which more intuitively reflects the overall improvement trend of the economic value of scientific research in China. Specifically, 2014 is a watershed in the change of the nuclear density distribution curve of innovative technology efficiency in universities. Before 2014, the overall nuclear density distribution curve of university innovation efficiency moves to the right and the peak height increases, indicating that the overall value of efficiency of economic value of provincial university innovation in this period increases but the provincial difference increases either; after 2014, the nuclear density distribution curve moves to the left and the peak height decreases, indicating that the efficiency of economic value of provincial university innovation declines in the whole period but the provincial differences decreases. From 2011 to 2015, the nuclear density distribution curve of efficiency of social value of provincial university innovation shows the obvious “one-peak” trend, indicating that social value efficiency is concentrated at a higher level. In addition, the nuclear density distribution curve of efficiency of social value of provincial university innovation moves to the left as a whole and the peak height rises, which intuitively reflects the overall downward trend of the efficiency of social value of provincial university innovation in China.
(3)Using the panel Tobit model to analyze the innovation environment, it is found that economic advantages, location advantages, government support, and infrastructure conditions have a significant impact on the efficiency of economic value of provincial university innovation. In particular, in terms of economic advantages, per capita GDP has a significant negative correlation with the efficiency of economic value of provincial university innovation, which indicates that the level of regional macroeconomic development and the scientific research innovation of universities are inhibited. The added value of the tertiary industry as a percentage of GDP has a positive contribution to the value of the efficiency of economic value of provincial university innovation, indicating that the higher the demand for innovation in high-tech industrial developed areas, the more beneficial it is to the development of economic value efficiency. In terms of location advantages, the efficiency of university innovation in provinces close to the eastern region have a rising trend both in terms of economic value and social value. This may be mainly related to factors such as the scientific research base in the eastern region and the importance attached to technological innovation. In terms of government policy support, the per capita public education funds have a great positive impact on the efficiency of economic value and social value of provincial university innovation, and the government’s influence on social value is greater than economic value. The status of infrastructure is positively correlated with the innovation efficiency of universities. On the one hand, good infrastructure conditions enhance the capacity of cultivating researchers, which is conducive to the development of research and development activities in universities. On the other hand, the optimization of infrastructure level further reduces costs in the process of the transfer of innovation output. Through the two efforts above, the efficiency of economic value and social value of university innovation is improved.
And finally, the paper proposes countermeasures to improve and optimize the efficiency of economic value and social value of university innovation.

Key words: university innovation efficiency, economic value, social value, dynamic evolution, innovation environment