Public Participation in Pollution Control in China—Empirical Research based on Spatial Dynamic Panel Data Model

Previous research focused more on emissions reduction effects of formal environmental regulations, but less on emissions reduction effects of public participation in such informal environmental regulations. Based on the provincial panel data from 2000 to 2017, this paper empirically examined the direct effects of public participation on environmental pollution under three different spatial weight matrices by using dynamic spatial panel model, and then tested whether public participation can affect environmental pollution through influencing government's formal environmental regulations. The results show that:(1)Generally speaking, the improvement of public participation can significantly inhibit pollution emissions, but there are regional differences. Public participation only promotes the reduction of environmental pollution in the eastern region, but has not yet shown a significant impact on the central and western regions. (2)Public participation can influence pollution emissions through influencing the government's formal environmental regulation policies, but the effect of influencing the government's environmental legislation and environmental law enforcement is not obvious, and the effect of inhibiting pollution emissions can be achieved through influencing the government's environmental economic regulation and environmental pollution control investment. (3)Environmental pollution has an obvious "cumulative effect" in the time dimension and an obvious "spillover effect" in the spatial dimension, and the dynamic continuous "additive effect" of environmental pollution is greater than the spatial "overflow effect". In terms of different regions, the dynamic cumulative effect of environmental pollution in the western region is more obvious, and the spatial spillover effect of environmental pollution in the central region is more obvious, indicating that the situation of pollution control in the central and western regions is more severe. Even if different spatial weight matrices are replaced, the above regression results are still robust.