Spatial Patterns of Carbon Emissions in the U.S.: A Geographically Weighted Regression Approach
Hamilton College - Economics Department
March 9, 2012
This paper uses U.S. county-level data to examine the extent of geographical variability in the process liking total emissions of carbon dioxide to measures of population, affluence, and technology. Results from geographically-weighted regression models show that there is strong evidence of geographical heterogeneity and that the magnitude, and in some cases, the direction, of the effects vary within and across the 48 contiguous states in the U.S. These results suggest that we ought to be cautious of policy recommendations based on global models that ignore or account imperfectly for spatial dependence.
Keywords: Carbon emissions, Spatial dependence, STIRPAT model, EKC hypothesis, Geographically-weighted regression
JEL Classification: Q38, Q40
Date posted: March 11, 2012