摘要

The development of high-tech industry has been widely recognized for its importance to economic growth, China has made substantial investment in high-tech industry, aiming to improve its productivity. This study proposes an approach for dynamic Data Envelopment Analysis to measure intertemporal innovation efficiency performance of China's high-tech industry during 2001-2015. The significant difference from existing studies is that the accumulated patent stock measured by Perpetual Inventory Method is regarded as a carry-over variable to link regional innovation activities over multi-periods. Furthermore, this model can deal with the overall efficiency score from a long-term and systematic perspective, and thus the overall efficiency score is decomposed into period scores with endogenous weights. The empirical results indicate that neglecting the carry-over variable may overestimate the efficiency of innovation system in China's high-tech industry; R&D efficiency of high-tech industry has almost improved monotonically throughout the study period; The eastern area has the highest average efficiency score, followed by the western area, and then by the central area; There is no statistical significant relationship between R&D personnel and R&D efficiency, and government grants have negative effects on improving efficiency. Finally, some important recommendations are achieved for policy makers and managers.