Methodological aspects of multi‐criteria decision analysis for policy support: A case study on pharmaceutical removal from hospital wastewater

Decision making in public and political contexts can be complex. Multi‐attribute value/utility theory (MAVT/MAUT) can support such decision processes by providing a transparent framework that helps focusing on objectives and corresponding degrees of achievement by different alternatives. Eliciting preferences with MAVT/MAUT can be time consuming and cognitively challenging. Therefore, it might not be feasible to elicit full preference functions with standard methods. To deal with this problem, we suggest a simplified elicitation procedure that combines (a) the elicitation of values instead of utilities at lower‐levels of the objectives hierarchy and conversion to utilities to consider risk attitudes at appropriate higher levels, (b) the use of linear value functions for sub‐objectives with minor effects on the overall value, and (c) sensitivity analyses to check the robustness of results regarding these assumptions and the elicitation process in general. Furthermore, we developed a modified Swing procedure (‘Reversed Swing’) to elicit weights for cases in which the hypothetical alternatives of the conventional Swing technique are unrealistic. We applied this procedure to a case study on pharmaceutical removal from wastewater of a typical Swiss hospital involving 13 stakeholders. Aim of the interdisciplinary research group was to assess a large bundle of combinations of novel point source measures. The ultimate policy objective was to develop consensus solutions which are acceptable to all important stakeholder groups. We hope that the suggested simplified procedure stimulates the application of transparent and conceptually satisfying decision support methods in environmental management, which is needed to justify policy decisions to the public.