The environmental Kuznets curve (EKC) hypothesis proposes that there is an inverted U-shape relation between environmental degradation and income per capita. Explanations for this pattern include proximate variables such as input-output structure that mediate the effects of underlying causes such as changing comparative advantage or increased environmental awareness with higher incomes. A number of studies have tested individual factors that might explain the EKC. But a systematic allocation of changes in environmental impacts or degradation has not been attempted. Stern and Common (1999) show that there are omitted variables in a global sulfur EKC that may result in biased estimates of the EKC in non-random samples of countries. In this study we develop a method to systematically "attribute" changes in emissions to the major proximate causes of the EKC relationship: change in input mix, change in output mix, scale, and technical change. We apply the attribution method to the panel data set used by Stern and Common. The results show that though input and output mix are statistically significant they make only a small contribution to changes in global emissions. Increasing scale and countervailing technical change explain most of the observed change. However, policy may play a role in accelerating technical change and input substitution on a regional scale, though our analysis cannot show if this results in some "sulfur leakage" to other regions. For a comparison, we estimate an EKC for the same period, which yields a monotonic emissions-income relation at the global level. This model does not explain as much variance as the nonlinear model and has poorer statistical properties.
