This paper examines the nonlinear relationship between environmental pollution and regional Incomes in Korea during the period 1988-2014. We allow for spatial dependence in emissions across 16 regions to consider the possibility that a region’s emissions are affected by those in neighboring regions. We use the panel spatial Durbin model with both individual and time-period fixed effects, and use three different per capita emission samples: carbon monoxide(CO), oxides of nitrogen(NO2), and sulfur dioxide(SO2). We find that air pollution is significantly influenced by the emissions of neighboring regions. From the total effect, including feedback and spillover effects, there is strong evidence for the N-shaped relationship between per capita emissions and regional incomes. From stochastic kernel analysis, the per capita CO and NO2 gap indicate a tendency toward divergence in regions over the entire period.

[Key Words] Spatial Dependence, Air Pollution, Panel Spatial Durbin Model, Environmental Kuznets Curve Hypothesis