Research on Regional R&D Innovation Efficiency in Chinese Provinces with the Perspective of Innovation Driven: Based on Projection Pursuit and Stochastic Frontier Analysis Model

As there are problems in measuring innovation efficiency with the traditional stochastic frontier analysis model(SFA), combined with the characters of projection pursuit model which can reduce the dimensionality of multidimensional data, the paper improves to SFA in order to effectively measure innovation efficiency. On this basis, regional RD innovation efficiency in Chinese provinces was analyzed empirically by adopting the data from30 provinces of China in 2003-2011, and the result shows: the average of regional RD innovation efficiency is below 0.5 percent, which is generally in a lower level and has the existence of the phenomenon of inefficiency,so there has large spaces to be doing better. The paper analyzes impact factors of innovation efficiency found that the level of economy, foreign opening degree, government funding, emphasis put on educating, property rights awareness, informationization level are favorable factors for RD innovation efficiency.