ABSTRACT
At present, there exist both regional-scale and regional performance differences in China’s high-tech industry development. Performance differences will inevitably affect resource allocation in regional high-tech industry and in the upgrading of industrial structure. Thus, this article conducts a systematic study of convergence issues in China’s high-tech industry development performance by adopting the spatial panel econometric method and related approaches.
First, to measure the comprehensive performance of China’s high-tech industry development, the study applies the Malmquist index and efficacy coefficient method from the perspective of R&D efficiency and economic efficiency; then, by adopting a spatial autoregressive (SAR) panel data model, it studies the absolute convergence and convergence mechanism of China’s provincial high-tech industry development performance from 2001 to 2013. The results indicate the existence of distinct absolute convergence in China’s high-tech industry development performance since the 21st century, which is mainly due to technology diffusion. The convergence rate of central China is notably faster than that of other regions.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 In China, patent not only refers to invention patent, but also includes utility model and appearance design.
2 The production function would be a more reasonable choice if it includes new product value added, because it contains the intermediate inputs, which may affect the efficiency value. However, in the absence of such accounting in the statistical yearbook, this article chooses the new product output value index to represent new product value added, based on the relevant references.
3 Tibet is not included due to a lack of data, while Hong Kong and Macao are not considered due to their different social systems.
4 In fact, this article chooses the fixed-effect model because the sample regression in this study is confined to certain individuals and there is no need to deduce the overall nature from a specific individual nature. Therefore, the fixed-effect model is a better option.
5 The MATLAB code of the spatial econometric model calculation programme can be downloaded from the website: www.regroningen.nl/elhorst