A bayesian model averaging approach to the transfer of subjective well-being values of air quality

Abstract

This paper utilizes the subjective well being approach to benefit transfer of the welfare effects of reducing air pollution across a sample of citizens from several European countries. In addition, we compare the results of subjective well being utilizing the ordinal probit model approach with an alternative Bayesian Model Averaging (BMA) approach that selects the best model from the space of potential models. The results shows that the BMA approach significantly reduces transfer errors for air pollution externalities and provides a more robust strategy to modeling SWB data on pollution effects.