This paper analyzes the spatial distribution characteristics of carbon dioxide (CO₂) emissions in 30 Chinese provinces from 2002 to 2013, and systematically examines the relationship between government and CO₂ emissions from the level of regional legal regulation, the level of public expenditure, and corruption. Research results show that the global Moran’s I values of carbon emissions from 2002 to 2013 falls between 0.2 and 0.3, and spatial lag coefficient ρ stands at 0.2340, indicating that CO₂ emissions in the region increase by 0.234% for each additional 1% of CO₂ emissions in the adjacent areas, which means carbon emissions exist in a remarkable spatial autocorrelation and spatial clustering phenomenon. Then the empirical results find that the level of legal regulation and the level of regional corruption are significantly negatively correlated with regional CO₂ emissions, while the level of public expenditure is significantly positively correlated with regional CO₂ emissions. The results also demonstrate the inverted “U”-shaped relationship between economic income and CO₂ emissions. Based on empirical results, this paper provides beneficial policy recommendations for reducing carbon emissions.