Decision problems can be difficult if the decision makers hope to achieve several conflicting objectives. For example, one objective could be ‘low costs’ while another objective is to achieve ‘good environmental performance’. However, the most environmentally-friendly technical option might also be the most expensive one. In environmental decisions, typically multiple stakeholders with differing interests and values are involved. Uncertainty, for instance about the longer-term consequences of decisions, further increases complexity.
Decision analysis supports systematic thinking to structure the decision problem and to better understand the preferences and values of those involved. It aims at integrating all available information: scientific data and ‘hard facts’ together with the subjective preferences of stakeholders. It improves the process of decision making by increasing its transparency. It aims at making an optimal decision that is well accepted by all parties.
Good decision processes based on Multi-Criteria Decision Analysis (MCDA)
Multi-Criteria Decision Analysis (MCDA) allows to analyze the ‘hard facts’, i.e. the predictions about the consequences of choosing a specific decision option. As an example, if engineers have to choose between different wastewater treatment technologies, it is important to know how well nutrients or micropollutants are removed, and how much each option costs. To make sound predictions about the outcome of decision options, modeling techniques or expert assessments can be used.
Predictions should include uncertainty, such as a lack of scientific knowledge. For instance, one might not exactly know how an ecosystem will react to a management intervention. Predictions are also uncertain due to simplifications in the used models, or because expert judgments are uncertain. Uncertainty increases over long time ranges because the future is not known. In such cases, it can be helpful to integrate scenario planning with MCDA, which allows to consider different possible futures.
MCDA also analyzes ‘soft data’, namely the preferences of decision makers and stakeholders. A good decision process ensures that the important stakeholders and the people affected by the decision are involved. It can be useful to select participants with stakeholder analysis, which belongs to the large family of Problem Structuring Methods (PSM). Such methods can also help to ensure that there is a proper understanding of the decision situation before using MCDA and that the main objectives of all people involved are considered, possibly including presumed interests of future generations.
People have different value systems and therefore different preferences in a decision. This especially affects how people perceive the importance of objectives and how they decide under uncertainty. If not all objectives can be achieved, trade-offs between these conflicting goals have to be made. Preference elicitation techniques help to determine, for instance, how important the achievement of an objective is, and how strongly it can be traded-off with another objective. For example, improving the ecological value of a river to foster an endangered bird species may mean that people can no longer use the river bank for picknicks. Such trade-offs are based on the best and worst possible outcomes of each objective for the specific decision problem (e.g. the costs of the cheapest and most expensive option, or the most user-unfriendly/ -friendly technology). Preference elicition processes (interviews or group workshops) are demanding, because people tend to run into systematic biases when answering our questions. People may also be uncertain about their answers, and they may be risk-averse if the outcomes of the decision (i.e. the predictions) are uncertain.
MCDA modeling and results
Preference elicitation techniques capture the decision maker’s preferences as numbers. These preference parameters enter the decision analysis model together with the ‘hard facts’, the predictions. The result of the MCDA model is a ranking of the decision options from best to worst. If several stakeholders were interviewed, a different ranking may result for each stakeholder. Also, some decision options may be more uncertain than others. The MCDA process helps to select good compromise options that perform reasonably well for all stakeholders despite uncertainty. MCDA should be seen as an iterative process, and it is often useful to refine or construct new decision options based on the insights of the process. It is also very beneficial to discuss the results with stakeholders.