The exact specification and motivation for an environmental Kuznets curve (EKC)
is the subject of a vast literature in environmental economics. A remarkably
diverse set of econometric approaches and candidate regressors have been
proposed, which highlights the degree of model uncertainty surrounding the
relationship between environmental quality and pollution. We introduce Bayesian
model averaging (BMA) to the EKC analysis to examine: (a) whether a sulphur
dioxide EKC exists, and if so (b) which income/pollution specification is
supported by the data. BMA addresses model uncertainty as part of the empirical
strategy by incorporating the uncertainty about the validity of competing
theories into the posterior distribution. We find only weak support for an EKC,
which disappears altogether when we address issues relating to the extreme
oversampling of two industrialized countries in the sample. In contrast, our
results highlight the relative importance of political economy and site-specific
variables (specifically executive constraints and precipitation variation) in
explaining pollution outcomes. Trade is shown to play an important indirect
role, as it moderates the influence of the composition effect on pollution.