TY - UNPB

T1 - Measuring the research performance of Chinese higher education institutions using data envelopment analysis

AU - Johnes, J

AU - Yu, L

PY - 2006

Y1 - 2006

N2 - This study uses data envelopment analysis (DEA) to examine the relative efficiency of over 100 selected Chinese regular universities. Various models are developed to measure the research efficiency of these higher education institutions (HEIs) using data for 2003 and 2004. The findings show that the level of efficiency depends on whether or not a subjective measure of research output (based on experts’ opinions of the HEIs) is included as an output in the model. Mean efficiency is higher when the reputation variable is included (around 90%) than when it is not (mean efficiency is around 55% in this case). However, the rankings of the universities are remarkably insensitive to whether or not this variable is included. Bootstrapping procedures are used to find the 95% confidence intervals for the efficiencies, and indicate that the best and worst performing institutions are significantly different from each other; only the middle-performing 30% of HEIs cannot be distinguished from each other in terms of their performance. Further investigation suggests that regional location, source of funding and whether the university is comprehensive or specialist may all contribute to the observed differences in performance. The regional differences are consistent but not significant at conventional levels of significance; the efficiencies differ significantly by administrative type when the subjective measure of research output is excluded from the analysis; comprehensive universities consistently and significantly outperform specialist institutions. The possibility of regional differences in performance is particularly worrying since the already economically disadvantaged Western region may suffer a continued lag in development if its HEIs are less efficient than those in the better developed Central and coastal regions.

AB - This study uses data envelopment analysis (DEA) to examine the relative efficiency of over 100 selected Chinese regular universities. Various models are developed to measure the research efficiency of these higher education institutions (HEIs) using data for 2003 and 2004. The findings show that the level of efficiency depends on whether or not a subjective measure of research output (based on experts’ opinions of the HEIs) is included as an output in the model. Mean efficiency is higher when the reputation variable is included (around 90%) than when it is not (mean efficiency is around 55% in this case). However, the rankings of the universities are remarkably insensitive to whether or not this variable is included. Bootstrapping procedures are used to find the 95% confidence intervals for the efficiencies, and indicate that the best and worst performing institutions are significantly different from each other; only the middle-performing 30% of HEIs cannot be distinguished from each other in terms of their performance. Further investigation suggests that regional location, source of funding and whether the university is comprehensive or specialist may all contribute to the observed differences in performance. The regional differences are consistent but not significant at conventional levels of significance; the efficiencies differ significantly by administrative type when the subjective measure of research output is excluded from the analysis; comprehensive universities consistently and significantly outperform specialist institutions. The possibility of regional differences in performance is particularly worrying since the already economically disadvantaged Western region may suffer a continued lag in development if its HEIs are less efficient than those in the better developed Central and coastal regions.

KW - data envelopment analysis

KW - efficiency measurement

KW - Chinese higher education

M3 - Working paper

T3 - Economics Working Paper Series

BT - Measuring the research performance of Chinese higher education institutions using data envelopment analysis

PB - The Department of Economics

CY - Lancaster University

ER -