An Environmental Kuznets Curve Analysis of U.S. State-Level Carbon Dioxide Emissions

Most environmental Kuznets curve (EKC) theories do not apply to carbon dioxide (CO 2 )—an unregulated, invisible, odorless gas with no direct human health effects. This analysis addresses the hypothesis that the income-CO 2 relationship reflects changes in the composition of an economy as it develops and the associated role of trade in an emissions-intensive good (e.g., electricity). To test this hypothesis, I use a novel data set of 1960 to 1999 state-level CO 2 emissions to estimate pretrade (production-based) CO 2 EKCs and posttrade (consumption-based) CO 2 EKCs. Based on the first EKC analysis of CO 2 emissions in the United States, I find that consumption-based EKCs peak at significantly higher incomes than production-based EKCs, suggesting that emissions-intensive trade drives, at least in part, the income-emissions relationship. I have also investigated the robustness of the estimated income-CO 2 relationship through a variety of specifications. Estimated EKCs appear to vary by state, and the estimated income-emissions relationships could be spurious for some states with nonstationary income and emissions data. Finally, I find that cold winters, warm summers, and historic coal endowments are positively associated with states’ CO 2 emissions.