Modelling Sectoral Fuel Demand in China in China

Recently Fisher-Vanden et al (2004) have used firm level data to estimate an energy intensity function for a large number of enterprises in China. This is the first study to use the regression approach to examining the relationships between fuel prices and fuel consumption in China. One of the major findings of the study is that firms are very responsive to relative fuel price changes. However, as the authors acknowledged, the data set is comprised of only sampled enterprises and the time frame is too short to examine firms’ dynamic behaviour. The present paper aims to model fuel demand behaviours in China using longer data series at the sectoral level. Unlike the FisherVanden et al study, fuel price data for the present study are first estimated using China’s input-output tables and industrial ex-factory price indices. Then, using annual data on total fuel consumption by sector for the period 1985-2000, the study investigates how sectors respond to changes in the relative prices of various fuels. Four sectors in the Chinese economy are selected for the study, namely, the chemical industry, the metal industry, the non-metal materials industry and the residential sector, which are top energy as well as top coal consumers. Five fuels are considered, namely, coal, crude oil, electricity, natural gas and petroleum products, which accounted for nearly all of the total energy consumption in each of the four sectors. A translog demand system is estimated for each sector using the seemingly unrelated regression method. The results suggest that significant substitutions away from coal to alternative fuels take place in the residential as well as the metal sectors.