Dynamic network DEA efficiency of Chinese higher education institutions: insights from the teaching, research, and grant application subprocesses

In the education community, there is widespread concern over the technical efficiency of higher education institutions (HEIs). To consider the internal structure of the efficiency measurement, network data envelopment analysis (NDEA) models have gradually been used in HEIs since 2010. In practice, there are input and output conversion processes in which, for example, shared inputs and dynamically changing shared inputs are converted into carry-over items in the education production process. However, existing NDEA studies rarely account for shared inputs and carry-overs for internally linked activities and the dynamic period efficiency evaluation of HEI. This study adopts a multistage dynamic network DEA (DNDEA) in the multiplier form, analyzing teaching, research, and grant application subprocesses for 64 “985 Project” and “non-985 Project” universities in China from 2018 to 2020. Unlike predetermined weights of subprocesses, our proposed DNDEA model utilizes endogenous weight determination and carry-overs embedded across periods, allowing HEIs to determine the relative contribution of each subprocess. On average, the evaluated 64 HEIs achieved overall efficiency of 0.6985, with research and teaching performing relatively well at 0.8373 and 0.9618, respectively, while the grant application stage lags significantly at 0.0218. Teaching performs consistently well across all HEIs. Most HEIs show improvements in teaching and research, while nearly half exhibit a declining trend in grant application efficiency. The results also indicate disparities in the efficiency scores for the research and grant application subprocesses. The “985 Project” universities outperform “non-985 Project” universities especially in research, whereas more “non-985 Project” universities prioritize teaching. Additionally, different weights are endogenously assigned to the three observed subprocesses, reflecting each HEI’s unique strengths and varying priorities for improving efficiency. Research emerges as the dominant stage driving overall efficiency, while grant application represents a critical area for improvement for all HEIs. To enhance overall technical efficiency, expanding collaborative networks, strengthening research records, and targeted training are required, particularly in the grant application.