• 论文 •

### 基于STIRPAT模型的中国能源压力分析——基于空间计量经济学模型的视角

1. 华东师范大学地理信息科学教育部重点实验室, 上海 200062
• 收稿日期:2010-10-11 修回日期:2011-04-28 出版日期:1997-09-20 发布日期:1997-09-20
• 通讯作者: 季民河,教授。E-mail: mhji@geo.ecnu.edu.cn E-mail:mhji@geo.ecnu.edu.cn
• 基金资助:
国家自然科学基金项目(40671074)资助。

### China's Energy Stress Based on the STIRPAT Model: A Spatial Econometric Perspective

JIANG Lei, JI Min-He

1. The Key Lab of Geographic Information Science, Chinese Ministry of Education, East China Normal University, Shanghai, 200062, China
• Received:2010-10-11 Revised:2011-04-28 Online:1997-09-20 Published:1997-09-20

Abstract: China is now claimed to be the largest energy consumer among all countries in the world, as it more than ever needs energy to sustain its consecutive two-digit annual GDP growth. The huge demand for energy has had a stronger impact on energy production and supply in the country. The distribution of energy consumption among provinces and cities may conceal significant spatial effects that each location has exerted onto its neighbors. When taken it into account in the analysis, these spatial effects may be used to rectify the estimation bias inherent in the traditional energy consumption model. This paper employed the total energy consumption as an index of environmental impacts to evaluate the spatial effect of energy consumption in China based on the STIRPAT model. The energy consumption data of all regions in China was first examined via the Moran's I index for the existence of spatial dependence. For calculating the index, a contiguity rule was employed to establish the spatial weight matrix for the regions. The exploratory analysis results in a value of 0.194 for Moran's I at the significant level of 0.05, which indicates a tendency of spatial clustering of similar consumption values. This warranted further analyses on a confirmatory nature. Two spatial econometric regression models based on spatial lag (SLM) and spatial error (SEM) respectively were then established to analyze the impact of several relevant factors on energy consumption. Results from these two models were compared on the basis of several statistical tests, and the SEM was selected to fit the data. The goodness of fit of the SEM reached 0.898, a 3% improvement over 0.871 of the adjusted R-squared resulting from the traditional OLS model. The results indicated that the average energy consumption between the years of 2006 and 2008 did present spatial interdependence to some degree among Chinese provinces, and the energy consumption behavior was collectively influenced by the internal factors of the province under investigation and its neighbors. There was a significant positive correlation between energy consumption and population, social affluence, and secondary industry. That is, the elastic coefficient of energy consumption increased gradually as these influential factors increased. A set of governmental countermeasures are necessary: moderate efforts should be made to revise birth control targets, civic investment should be increased to advocate a low-carbon lifestyle in China, and new and energy-saving technologies should be rapidly adopted in the Chinese industrial system. Above all, the strategic planning and policy making for the long-term reduction of energy consumption should consider the spatial interaction mechanism of energy consumption among different jurisdictions in the country.

• F426