Abstract:  "Emphasizing nature science over liberal arts" is an academic malady accumulated in China′s long-term development process. On the one hand, after the division of liberal arts and sciences, the phenomenon of "emphasizing science over liberal arts " has been triggered; on the other hand, with the nature science disciplines (e.g., science, engineering, agriculture and medicine) playing an obvious and important role in the socialist modernization of new China, the degree of "emphasizing nature science over liberal arts" has been increased. Obviously, the emphasis on nature science rather than literature arts has directly led to the focus on scientific research and engineering. However, it is the basic requirement for education to emphasize both nature science and liberal arts. Therefore, in terms of scientific research efficiency, whether or not nature science is emphasized over liberal arts? This study will discuss this topic.Several studies use the DEA method with fixed production frontier to evaluate the scientific research performance at the university or regional level, but it has not been analyzed by different subjects. Based on the panel data of R&D inputs and output of nature science and liberal arts in mainland China from 2009 to 2014, this paper first uses the SFA method with variable production frontier to measure and compare the scientific research efficiency of different disciplines, so as to investigate whether there is a phenomenon of emphasizing nature science and ignoring literature arts from the perspective of scientific research efficiency. We draw the following conclusion and policy implications:It is found that the scientific research efficiency of natural science is significantly higher than that of liberal arts, which means that "emphasizing nature science rather than literature arts" is still established at the level of scientific research efficiency. Moreover, in the six years of the sample range, the gap is expanding. In 2009, the scientific research efficiency of natural science was 0.8130, slightly higher than that of social sciences, 0.7441. In 2014, the scientific research efficiency of natural science rose to 0.8885, while that of social sciences was still only 0.7474, which shows that the gap is gradually expanding. The scientific research efficiency of natural science has experienced a stable upward trend, with an average annual increase of 1.79% in six years; while the scientific research efficiency of social sciences has experienced a roughly N-shaped fluctuation, with an average annual increase of only 0.09% in six years.The scientific research efficiency in different disciplines shows obvious regional differences. The growth rate of scientific research efficiency of natural science in universities located in the eastern region is relatively slow, while that in the western region is relatively high. Although the scientific research efficiency of nature science in the eastern region is the highest in the initial stage, it has been overtaken by universities in the western and central regions at the end of our sample. In terms of the scientific research efficiency of social sciences, the eastern region with higher initial level was also surpassed by the central and western regions in the later period of the sample interval. The scientific research efficiency of social sciences in the central region showed a more rapid improvement. There are great differences in the scientific research efficiency between different provinces and cities. Among them, for the scientific research efficiency of nature science, Henan Province (i.e., the highest level) is 0.53 times of that in Tianjin, which is the lowest level. Also, the difference in the scientific research efficiency of social sciences among different provinces is even greater. Henan Province (the highest level) is 1.48 times of Shanxi Province, which demonstrates the lowest level.

Key words: R&D technical efficiency, Chinese universities, natural science, social science