Spatial Econometric Analysis of Carbon Consumption of Chinese Provinces

Learning from the experience of the C-D function form´s double logarithm, and using five explanatory variables, the authors tentatively build models by themselves and introduce spatial effect into the introduction of the research about Chinese provincial low-carbon economy. First, according to Moran´s index of coal consumption behavior we determine the existence of spatial autocorrelation, secondly, we determine the best choice about the spatial lag models and spatial error models through the five spatial weight matrix, and then analysis conclusions are made (achieved) through the most superior model: spatial effect exists among the provincial coal consumption behavior. The estimated value of spillover effect parameter is up to 0.1286, and gradually becomes smaller with the distance; There exist endogenous economic relations among the dependent variable of coal consumption quantity and the independent variables of industrial structure, population growth, resources talent, income level and industry economy growth, among which the industrial structure and the population growth are the two most important factors affecting the coal energy consumption behavior, while the higher industrial economic growth is not necessarily accompanied by higher coal consumption. It implies the enhancement of coal efficiency. Finally we make conclusion and put forward related policy recommendations.