科技发展战略研究所知识库

Although a large amount of past research has studied the measurement of institutional research and development
(R&D) efficiency, there has been limited dynamic empirical investigation of this topic. This study differs from
previous studies in its consideration of R&D capital stock, which is defined as a carry-over activity between two
consecutive periods. To further test high-tech institutes’ R&D performance in the pre-reform and post-reform
periods, we evaluate a panel of 29 high-tech institutes of the Chinese Academy of Sciences (CAS) for
2011–2017 with a dynamic slacks-based measure. (a) The empirical value of the average overall efficiency is
0.5559, and the clustering analysis shows that institutes operating above the average overall efficiency level
follow an increasing and then decreasing trend, while those operating below the average overall efficiency level
follow a decreasing and then increasing trend. (b) The efficiency of low-performing institutes has increased from
0.2259 in the pre-reform period to 0.2727 in the post-reform period, which does not reflect the efficiency improvements
of high-performing institutes. (c) High-tech institutes selected as pilot institutes have an average
efficiency level of 0.6524, which is higher than that of non-pilot institutes (0.4765), indicating the benchmarking
effect of pilots. (d) The types of pilot institutes can be ordered as follows: feature institutes > centres of excellence
> innovative academies. (e) The total-factor input efficiency level has decreased in the post-reform period.
Only the total-factor efficiency of high-quality papers has increased in the post-reform period (by 0.0094).