Non-market valuation of coastal ecosystem services

Abstract: Marine and coastal ecosystems provide benefits to people through the provision of seafood, maintenance of water quality and several other valuable ecosystem services. The need to understand these benefits in economic terms has never been more pressing. Contingent valuation and choice experiments are survey-based non-market valuation methods that can be applied to estimate the values of these benefits. This thesis aims to address some of the challenges facing non-market valuation of coastal ecosystem services, comprising a summary and four appended papers that are based on two separate studies designed to value coastal ecosystem services in two archipelago areas in Sweden. Paper I addresses the challenge of creating and describing a contingent valuation scenario so as to achieve content validity. A holistic valuation scenario based on ecological measures that are well connected to policy was developed and is described and evaluated. Sometimes when decisions have to be made and new valuation studies are impossible due to time and funding restrictions, estimates from existing studies can be transferred to a new setting, an approach referred to as benefit transfer. Including socio-economic information in choice experiment models for benefit transfer is addressed in Paper II. The findings suggest that a more ‘general’ model based on socio-economic information that can be easily obtained from public databases performs quite well as a model that would require the collection of more detailed information. Paper III addresses distributional effects of environmental policies affecting coastal ecosystem services. The study presents information regarding distributional effects in both monetary terms and in terms of environmental quality for different socio-economic groups. Paper IV addresses the important topic of model selection uncertainty for choice experiments in environmental valuation. To be able to ensure robustness of results, this study proposes an information-theoretic approach for model selection. The findings also indicate that greater attention should be paid to potential policy implications that may follow from different model specifications.