Modelling energy efficiency in China: a fixed-effects panel stochastic frontier approach

This paper combines energy demand modelling with stochastic frontier analysis to investigate the changing trends, variations and determinants of energy efficiency for 27 Chinese provinces over the period 1995 to 2014. An aggregate ` frontier' energy demand function and an efficiency function are estimated simultaneously. We obtained several findings. First, the energy intensity is not a particularly good indicator of energy efficiency. Second, the energy efficiency levels for all the provinces improved during the sample period, but the current efficiency levels are still low, implying great potential for energy saving. In addition, the energy efficiency gap among the provinces seems to have widened over the past 20 years, as the variance has increased by almost three times. Finally, technological progress driven by new investment and the development of market mechanisms are two important drivers of energy efficiency improvement.