This paper discusses the decision mechanism for the regional innovation differences from the perspective of innovation efficiency. Specifically, endogenous growth model is adopted for innovation and two-level theoretical framework of regional innovation mechanism is established in this paper. Then innovational production efficiency and innovational economic efficiency as well as their impacts are estimated from the two levels of regional innovation efficiency. Regional differences of innovational production efficiency and innovational economic efficiency in China are analyzed respectively. In the final section, this paper proposes a system of innovation development patterns that are suitable for different regions in China on the basis of rich empirical researches.In the theoretical research section, this paper establishes an endogenous growth model with knowledge-driven R&D, and adds the impact of technology imitation to the innovational production as well as the interaction between intermediate product department and final product department, under the open conditions, which makes the model closer to the reality. The result of the model indicates that innovation's effect on regional economics can be divided into two levels:one is the regional innovational production level itself; the other is the realization level of economic efficiency for regional innovation achievement. It provides theoretical evidence for the whole paper and is also the basis of the empirical research later on.On the first level empirical research of regional innovation efficiency in China, this paper selects 30 provincial administrative regions in China as the decision-making unit of innovational production and data ranging from year 1999 to 2008. Innovational production is estimated based on knowledge production function via stochastic frontier analysis (SFA). Battese and Coelli's model (1995) is employed to estimate the elements that effect innovational production efficiency. In order to find out the different innovational production efficiency of various innovational subjects, this paper divides this level of innovational production efficiency into three categories of enterprises, universities and scientific research institutions, covering the comprehensive innovational production efficiency of 30 regions and of 3 innovation subjects. Finally, clustering method and coefficient of variation are used to conduct summarization, comparison and analysis of variance on these four kinds of innovational production efficiency.On the second level empirical research of regional innovation efficiency in China, from the perspective of economic activities, this paper considers innovation as an important production factor and estimate its production efficiency, so as to reflect the economic conversion capability of innovation achievements, that is, the innovational economic efficiency. In this section, this paper again adopts stochastic frontier analysis and a model based on estimating production efficiency. The innovational economic efficiency of 29 provincial administrative regions is estimated. Again, clustering method and coefficient of variation are used to conduct classification and analysis of variance on the innovational economic efficiency of various Chinese regions, with comparison to the classification and variation of regional innovational production efficiency. Finally, four kinds of convergence inspection are conducted on the innovational production efficiency and innovational economic efficiency of various Chinese regions with several conclusions made on the regional differences trends of innovational production efficiency and innovational economic efficiency.Regional innovation development pattern is an important choice for regional economic development and is a practical way of regional innovation. Differentiated regional innovation development patterns should be established according to various features of innovation development in different regions. This paper proposes a regional innovation development pattern system suitable for different regions and different development stages in China, so as to establish regional innovation development strategies accordingly for different development stages of our country and innovation-endowed regions, which ultimately offers policy foundations for the development of regional economics and the transform of development patterns.The key point in this paper is that it creatively divides innovation efficiency into innovational production efficiency and innovational economic efficiency, which clears the thoughts in innovation effect research and offers new theoretical framework for the empirical research on innovation efficiency. Based on this idea, this paper attains fruitful conclusions, among which include the core finding that the innovation efficiency in all the Chinese regions is significantly lower than 1 on average, no matter whether it is innovational production efficiency, or innovational economic efficiency, or the three innovation subjects of the innovational production efficiency. Technology inefficiency obviously exists. Especially, the innovation production capabilities of most Chinese regions well exceed the demand for innovation products from economic activities. The economic growth patterns in most Chinese regions are still extensive, low-level and resource-consuming. Technology and knowledge has not yet become an effective production factor, which only has rather weak effect on the economic development. The differences of innovation efficiency between regions are large, with regional differences of innovational economic efficiency increasing and that of innovation production efficiency converging. |