Economic Growth and Environmental Quality: A Non-Parametric Kernel Estimation of the Environmental Kuznets Curve

The proposed inverted U-type relationship between environmental degradation and per capita income under EKC hypothesis has been re-examined in this paper. A nonparametric kernel estimation technique has been employed to obtain varying point estimates of partial derivatives of sulfur dioxide emissions with respect to per capita income. This technique does not impose any a priori functional restrictions in the empirical testing of EKC hypothesis, and avoids the modelling criticism raised in the literature. More importantly, we present a test of EKC hypothesis by decomposing data into residential, industrial and commercial areas of a city (which are further divided into centre-of-the-city and suburban areas). Our results suggest a qualified support for the validity of EKC hypothesis: it is statistically supported for observations corresponding to centre-of-the-city (either residential or industrial or commercial). The hypothesis does not seem to hold for the suburban areas. Finally, a distinction between coastal and off-coastal cities turned out to be important as indicated by the climatic knowledge