Erik Hille

• Given the ambiguous empirical results of previous research, this paper tests whether support for a climate policy-induced pollution haven effect and the pollution haven hypothesis can be found. Unlike the majority of previous studies, the analysis is based on international panel data and includes several methodological novelties: By arguing that trade flows of dirty goods to less dirty sectors may also be influenced by changes in policy stringency, trade information on primary, secondary, and tertiary sectors are included. In order to clearly differentiate between dirty sectors and sectors with high pollution abatement costs, separate measures for pollution intensity and policy stringency are implemented. For the former, two intensities, namely the sectors’ carbon dioxide emission intensity and the emission relevant energy intensity, are used to identify dirty sectors. For the latter, an internationally comparable, sector-specific measure of climate policy stringency is derived by applying a shadow price approach. Potential endogeneity between climate policy stringency, trade openness and the trade balance is controlled for by employing a dynamic panel generalized method of moments estimator. The results provide evidence for a pollution haven effect that is also present for non-dirty sectors, i.e., a sector’s net imports rise in general if the sector faces an increase in climate policy stringency. Moreover, a stronger pollution haven effect regarding carbon dioxide intensive and emission relevant energy-intensive sectors is revealed. However, no support for the stronger pollution haven hypothesis can be found. Keywords: International trade, pollution havens, carbon leakage, global pollution, eironmental policy stringency, shadow prices