Evaluating the scientific and technological innovation efficiency of universities in China: evidence from the global Malmquist-Luenberger index model

The evaluation of the scientific and technological (S&T) innovation efficiency of universities is important for the 'double first-class' construction in China. However, related research often only considers desirable outputs while ignoring undesirable ones, which tends to produce biased evaluation results. For this reason, this study takes the ratio of unemployed graduates as undesirable output, and uses the global Malmquist - Luenberger index method to evaluate the S&T innovation efficiency of 31 relevant universities in China during 2013-2017, then develops a panel Tobit model to analyse its influencing factors. The results reveal that: first, during the sample period, the S&T innovation efficiency of those universities improved overall, and technical progress was the main driver. Second, excluding undesirable output will result in an overestimate of the S&T innovation efficiency of universities. Third, the S&T innovation efficiency of universities features regional distribution, with high efficiency in the west and low efficiency in the east. Moreover, different regions have different drivers for the improvement of S&T innovation efficiency of universities, with technical progress for the east and technical efficiency changes for the central and west. Finally, the 'double first-class' construction proposal in China has significantly promoted the improvement of S&T innovation efficiency in universities.