Does lower regional density result in less CO2 emission per capita?

Regional density is a useful tool for analyzing regional spatial structure as well as a good starting point for analyzing regional CO2 emissions per capita. This paper empirically analyzes the relationship between regional density and per capita CO2 emissions in China’s prefecture-level administrative regions. We improve the CO2 emission measurement method for prefecture-level administrative regions and estimate the per capita CO2 emissions of 252 prefectural-level cities in China from 2003 to 2013. Using panel fixed effect model regression, and taking the terrain roughness index as an instrumental variable to solve endogeneity, we find that the relationship between regional density and per capita CO2 emissions presents in an inverted U-shape, per capita CO2 emissions first increase with the increase of regional density, and after reaching the turning point, it decreases with regional density. In a mechanism test, analyzing the interaction terms between regional density and industrial structure, and regional density and urbanization level respectively. We found that industrial structure and urbanization are important mechanisms for regional density to affect CO2 emissions. In order to reduce per capita CO2 emissions, we propose corresponding policy implications for the regions in different positions of the “U” curve.