Spatial Dependence in the Environmental Kuznets Curve

Abstract: This essay aims to detect spatial autocorrelation present in the EKC relationship, as well as to find empirical evidence of said relationship. A panel data of 176 countries for the period 1990-2018 is used for the empirical part. Several spatial models are considered in this essay, namely the SDM, SAR model, SAC model and SEM. The main conclusions are: (i) there exists significant spatial autocorrelation in the CO2 emissions, (ii) significant evidence for a direct EKC relationship is found, (iii) the renewable energy consumption and the CO2 emissions are significantly and negatively correlated, (iv) among all models considered in this essay the SAR model with time and spatial fixed effects is the one with less information loss.