Abstract

To improve the air pollution of China fundamentally, effective measures should be proposed based on the thorough understanding of the characteristics of air pollution. Based on spatial econometrics, this paper investigates the characteristics and analyzes the determinants of the spatial concentration of PM2.5 pollution in China. Results show that: (1) PM2.5 pollution is highly concentrated in Central and Eastern China, covering 17 regions which accounts for 75% of the total population and GDP (gross domestic product). (2) The PM2.5 values in China show a significant spatial correlation. Provinces such as Shandong, Henan, Anhui, and Hubei are high in PM2.5 concentration. Meanwhile, these provinces are high in population density, GDP, and coal consumptions and have a large amount of civilian cars. (3) PM2.5 pollution shows spatial spillover effects. A 1% increase in the PM2.5 values of neighboring provinces will lead to a 0.78% increase in that of one province. (4) An upward U-shaped relationship is observed between the density of per capita GDP and PM2.5, and the PM2.5 value is far from the turning point of growth. With the further growth of the density of per capita GDP, the PM2.5 value is expected to increase rapidly and continuously. (5) Based on the characteristics of spatial concentration and spatial spillover, this paper proposes several prevention-control measures for haze pollution, such as stressing on the treatment of air pollution in severely polluted provinces, avoiding moving pollution industries to neighboring areas, performing joint prevention and control nationwide. Air pollution may only be rooted by transforming the pattern of economic growth.