LAUSR

This paper tests for a carbon Kuznets curve (CKC) by examining the carbon emissions per capita–GDP per capita relationship individually, for 21 OECD countries over 1870–2010 using a reduced-form, linear model that allows for multiple endogenously determined breaks. This approach addresses several important econometric and modeling issues, e.g., (1) it is highly flexible and can approximate complicated nonlinear relationships without presuming a priori any particular relationship; (2) it avoids the nonlinear transformations of potentially nonstationary income. For 10 of 14 countries that were ultimately estimated, the uncovered emission–income relationship was either (1) decoupling—where income no longer affected emissions in a statistically significant way, (2) saturation—where the emissions elasticity of income is declining, less than proportional, but still positive, or (3) no transition—where the emissions elasticity of income is (or very near) unity. For only four countries did the emissions–income relationship become negative—i.e., a CKC. In concert with previous work, we conclude that the finding of a CKC is country-specific and that the shared timing among countries is important in income-environment transitions.