TTIP and the environmental Kuznets curve

This paper investigates the empirical validity of the environmental Kuznets curve (EKC) in the potential Transatlantic Trade and Investment Partnership (TTIP) members. We use data from EDGAR (2015), CAIT (2015), UNFCCC (2015), NEC/NFR (2015), EEA (2015), EUROSTAT (2015), and Stern (2006) for emissions per capita of 10 air pollutants, such as CO2, CH4, GHGs, HFCs/PFCs/SF6, NOx, NO2, NH3 , SF6, SO2, and SOx and per capita municipal waste for all TTIP members over 1989-2013. For our independent variables, we employ data sources from IMF (2015), PENN World Tables 8.0, CIA World Factbook (2015), and Kraufman et al. (2011). We apply several econometric techniques to empirically evaluate the existence of the EKC, and then investigate the effects of the implementation of TTIP on the EKC. Throughout this paper, we employ the usual random and fixed effects approaches. Furthermore, in addition to the usual heteroskedastic robust standard errors, we employ specifications that are robust to contemporaneous cross-sectional dependence and serial correlation effects, respectively. In addition, we run several robustness checks in order to avoid any endogeneity problems related to the possible dual causality issue between each pollutant measure and each of our per capita income variables, or between the former and each of our trade measures. We report generally robust and statistically significant evidence suggesting an inverted U-shaped relationship between economic growth and per capita emissions of CO2, CH4, and HFCs/PFCs/SF6. This implies that at first economic growth increases per capita emissions of the above three air pollutants, but eventually growth reduces their per capita emissions. In other words, initially the scale effect dominates the technique effect, but then eventually for high levels of income, their role is inverted. We report the omitting bias problem, especially in the cases of GHGs, SF6, and NO2. We show practically a monotonically increasing relationship between per capita income and per capita emissions of each of the latter three air pollutants, respectively. Mostly opposite to previous studies, we report statistically significant evidence, suggesting a U-shaped relationship between per capita emissions of SO2 and SOx and per capita income. Further, we show that the values of the EKC's turning point for each pollutant are sensitive to the econometric approach and/or the control variables employed. Finally, we report statistically significant evidence suggesting a U-shaped relationship between per capita emissions of SO2 or SOx, and per capita income.