Research On Performance Evaluation Of Government R&D Input

The government R&D input undertaking the dual roles, shall bear the direct investor of R&D activity at the same time, and also plays a role of the indicator and compass of other R&D input. So that the research with the government R&D input as the starting point is particularly important.First of all, using the method of descriptive statistics, exponential function and moving average, based on the data of R&D expenses from 2004 to 2014, we forecast R&D expenditures from 2015 to 2020. Studies have found that R&D expenditure will continue to increase the amount, and the growth rate of testing and development of research funding is the fastest, followed by applied research funding growth, lowest growth of basic research. Compared with previous studies, this study has longer time span and used data updates. Even more important, the study suggests that long-term, basic research expenditures will be v-shaped curve, spending on R&D will face a test of ceiling effects. Introduced latent growth model on development trend and the effect factors of effective patent number, new products development number and science and technology papers number do in-depth exploration, we found three variable characterization of science and technology output basically keeping steady growth, and the growth rate of provinces having more high of effective patent number in initial become lower than those provinces having not so high of effective patent number in initial. The number of new products development in the year has some volatility, but on the whole, also showed the trend of steady growth. The difference is that the number of new products development and science and technology papers is in fixed initial level, the number of effective patents is increased over time, and its growth rate will gradually become faster, and the number of scientific papers and new product development is increased over time, but its growth rate will gradually slow down.Secondly, using principal component analysis method on evaluation of the government R&D input-output performance from 2010 to 2015, we found that Beijing, Guangdong, Jiangsu, Shanghai, Shandong have been ranked the top since 2012, and Gansu, Xinjiang, Guizhou, Chongqing, Inner Mongolia, Ningxia, Hainan, Qinghai, Tibet, this 9 provinces are relatively stable rankings behind. Compared with previous studies, we choosing the index are more comprehensive, the most important is that we evaluate the R&D input performance over the past 5 years, and more fully verify the pattern of strong in the East and weak in the West of R&D funds input-output performance in China, and the objective reality of insufficient investment in the West of China.Thirdly, using least square regression method, we research the occupancy ratio of government R&D appropriation expenditure, the ratio between government and enterprise R&D appropriation expenditure, the effect relationship between imports trade and science and technology input-output efficiency. The results showed that there are the positive effects on the occupancy ratio of government R&D appropriation expenditure, the ratio between government and enterprise R&D appropriation expenditure, and the effect relationship between import trade and science and technology input-output efficiency. There is the negative effect on the export and import structures (the specific value between export trade and import trade) for science and technology input-output efficiency. There is non-significant effect of the occupancy ratio of government R&D appropriation expenditure, the ratio between government and enterprise R&D appropriation expenditure for science and technology direct outputs. Compared with previous studies, this paper makes up for the problem of PCA science and technology input efficiency evaluation results cannot be look back upon. In the case of endophytism of the control variable, we reanalyze the effect indicator of science and technology input efficiency, with an emphasis on direct R&D input-related indicators, including share of government funds in total, and the proportion of government funding and enterprise funds. At the same time, we verified the different effect and discrepancy of import and export trade for science and technology input efficiency, and solved the problem that most of the research focused on import trade, less concerned about the impact of trade structure.Lastly, using poisson regression model and OLS regression model, through stepwise regression method researching the effect relationship in government R&D input, enterprise R&D input and the innovation performance, the result shows that there is the significant positive effect for government R&D funding input on innovation performance; there is the significant positive correlation between government R&D funding input and enterprise R&D funding input; enterprise R&D funding input plays the intermediary role in relationship of government R&D funding input and innovation performance completely. Compared with previous studies, this paper explains the effect mechanism of government R&D input for innovation performance, that is, government R&D input will promote enterprise R&D investment, and ultimately promote the innovation performance. Meanwhile, this paper supports the "inductive effect" of government R&D input for enterprise R&D investment. In existing studies, research and analysis level is "organizations". There have some research deviation due to sample selection issues. We verified the existence of "inductive effect" based on the more macro-level. That means, the government R&D input higher lead to enterprise R&D input higher, rather than lead "lazy" for enterprise. At a more detailed industry classification or other studies, most of the findings of the study and also remain the same with us. In other words, the government R&D input "crowding out" enterprise R&D investment, is not because the subsidy itself, but because of deviations in the process of implementation.