Towards Empirical Measurement of Externalities
This chapter details the approaches used to estimate the value of non-market goods and services attributable to the scenarios. The natural science team of the research project carried out detailed experiments at the power plant, in order to assess how a changed flow would affect fish ecology. In fact, the perturbations or scenarios under study were implemented in a test run and the ecological consequences thus monitored live on site. A contingent valuation experiment was undertaken using a web-based survey instrument. The contingent valuation study targeted respondents in the municipality affected by the considered perturbations in water flows. It was preceded by a focus groups analysis and the on-site ecological assessment described above. These two studies were basic ingredients in the scenario development. A key issue in any survey is the elicitation architecture, i.e. the way of eliciting information from the respondent. We argue that self-selected intervals provides a rich picture of response uncertainty, potentially increase response-rates and maintains a link to recent ideas on coherent arbitrariness. We implemented this approach in the present study by allowing the respondents to choose between a point and a self-selected interval. Thus, if they could not answer the valuation question with a point estimate (as our theory does suggest), they were then asked to submit the answer as an interval. In other words, we assume that true WTP is either stated as a point or contained in the reported interval. The econometrics of this approach is discussed and different estimators are introduced and their properties are compared. Finally, the chapter addresses some other non-priced items, such as the cost of emissions caused by replacement power (assumed to be provided by coal-fired plants).