基于静态与动态空间面板模型分析城镇化对雾霾的影响

该研究以中国30省区为研究对象,首先考察了雾霾污染的空间效应。全局Moran’s I指数为0.3875,中国雾霾污染存在着空间集聚。Moran’s I指数散点图显示中国雾霾污染存在着正的空间自相关,绝大部分省区位于高-高集聚和低-低集聚。然后,基于雾霾污染的空间效应,建立静态与动态空间面板计量经济学模型,实证考察了城镇化、能源强度、交通压力等对雾霾污染的影响。城镇化与雾霾污染之间存在环境库兹涅茨曲线;交通压力每上升一个百分点,将使雾霾污染上升0.2075个百分点。从效应的分解来看,该地区以及全局城镇化与雾霾污染之间存在环境库兹涅茨曲线。人均GDP的间接效应、总效应显著为负,该地区人均收入的上升可以使相邻地区的雾霾污染下降,并且会减少全局雾霾污染。能源强度下降会减轻本地区雾霾污染程度,但会导致相邻地区雾霾污染上升。能源消费结构直接效应为正,间接效应与总效应显著为正,煤炭消费比例的下降不但会减少本地区雾霾污染,也会显著抑制相邻地区的雾霾污染,进而减少全局雾霾污染。交通压力的直接效应显著为正,但间接效应显著为负,交通压力的上升会明显加重本地区雾霾污染程度,然而相邻地区交通压力上升会抑制本地区雾霾污染。动态空间杜宾面板计量模型中被解释变量滞后一期系数高达0.6114,且通过了1%水平下的显著性检验,中国雾霾污染存在时空依赖。动态空间面板计量模型比静态更为合适,估计结果更为准确,遗漏因素对雾霾污染的影响也很重要。

Analyzing effect of urbanization on haze pollution based on static and dynamic spatial panel model

At first, this article studied the spatial effect of haze pollution, using Chinese 30 provinces panel data from 2003 to 2014. The global Moran's I index was 0.3875, and there was spatial agglomeration of haze pollution in China. The Moran′s I index scatter plot showed that there was a positive spatial autocorrelation of haze pollution in China, and most of the provinces were located in high-high agglomeration and low-low agglomeration. This research empirically investigated the influence of urbanization, energy intensity and traffic pressure on hazy pollution with static and dynamic spatial panel econometric model. The spatial autoregressive coefficient was 0.1449, and the spatial spillover effect of haze pollution was significant. There existed an Environment Kuznets Curve between urbanization and haze pollution. The increase of traffic pressure by one percentage will increase haze pollution by 0.2075 percentages. From the decomposition of the effects, the EKC curve existed between urbanization and haze pollution in the regional and global. The indirect effect and total effect of per capita GDP were significantly negative, and the increase in per capita income in the region can reduce haze pollution of adjacent region and whole region. The decrease in energy intensity will reduce haze pollution in this region, but will increase haze pollution in adjacent regions. The direct effect, indirect effect and total effect of energy consumption structure were all significantly positive. The decline in the proportion of coal consumption not only can reduce the haze pollution in this region, but also can significantly inhibit the adjacent regions of haze pollution, thereby reducing the haze pollution of the whole region. The direct effect of traffic pressure was significantly positive, but the indirect effect was significantly negative. The increase of traffic pressure will obviously increase the degree of haze pollution in this region. However, the increase of traffic pressure in adjacent areas will inhibit the haze pollution in the area. The coefficients of PM10 (-1) in the Durbin spatial panel model was up to 0.6114 which indicated the time and spatial dependence of haze pollution in China. The dynamic spatial panel econometric model was more appropriate and accurate than the static model which revealed the influence of omission factors such as the mode of resident behavior and construction on haze pollution were also very important. These results for the understanding the impact of the energy intensity and traffic pressure on haze pollution in the process of urbanization and promoting win-win of urbanization and ecological environment have important policy implications. In order to control China′s haze pollution, regional cooperation is required and the sustainability of haze management is ensured. During the new type of urbanization process in China, we should take the green as the banner, expand the urban green area, make the urban energy conservation infrastructure construction, and green town planning. In the new normal economy, we should increase per capita income of residents; We should strengthen the energy efficiency; and reduce the proportion of coal consumption, and increase the proportion of non-fossil energy; We should strengthen traffic management and develop the public transport.