基于半参数空间模型的空气污染与经济增长关系再检验  

Air Pollution and Economic Growth: Empirical Evidence From a Semi-parametric Spatial Model

作　　者：吴雪萍 高明[2] 曾岚婷 Wu Xueping, Gao Ming ,Zeng Lanting

机构地区：[1]福州大学经济与管理学院 [2]福州大学

出　　处：《统计研究》2018年第8期82-93,共12页Statistical Research

基　　金：福建省高校新型特色智库项目"福建绿色发展研究院"(201705)的阶段性成果

摘　　要：本文选取2001-2014年我国30个省份数据作为样本,考虑空气污染的空间自相关性,采用空间杜宾滞后模型(SDM)和半参数空间滞后模型实证检验经济增长与空气污染的非线性关系。结果表明:①我国空气污染存在显著的空间正相关性,高空气污染水平集聚地随时间推移呈现出"由西向东"的转移特征;②空气污染与经济增长存在一种震荡曲线形式的关系,并不完全吻合传统的EKC倒U型曲线形状,但震荡关系也符合EKC所描述的环境污染与经济增长的关系将长期存在的特征,说明了经济增长并不能自发解决空气污染问题;③半参数空间滞后模型的拟合优度高于普通参数模型,其刻画的空气污染与经济增长的非线性特征验证了前人对二者震荡关系的猜想,结果更为稳健、准确与有效。Based on a Chinese province-level panel dataset from 2001 to 2014,this research employs the spatial Durbin model（ SDM） and semi-parametric spatial model to verify the nonlinear relationship of economic development and air pollution with considering the spatial autocorrelation of air pollution. The results show that the air pollution in China has significant and positive spatial correlation and spatial spillover effect. The concentration regions with high air pollution level have transferred from western to eastern. Moreover,there is a wave relationship between air pollution and economic growth,which does not exactly coincide with the inverted U-shape（ EKC） but conform to the long-standing relationship between environmental pollution and economic development described by EKC. It implies that air pollution can＇t be solved automatically by economic growth.In addition,it finds that semi-parametric spatial model works better than the parameter model in goodness of fit,as it can elaborate the nonlinear relationship effectively. The findings testify the former scholar＇s conjecture.

关 键 词：空气污染 经济增长 空间计量 半参数空间滞后模型 

分 类 号：C812[社会学—统计学]