1 Department of Transport, Technical University of Denmark2 Transport policy and behaviour, Department of Transport, Technical University of Denmark3 Department of Management Engineering, Technical University of Denmark
Assessing Robustness in Decision making
The evaluation of transport projects has traditionally been based on quantitative descriptions of selected impacts by the use of cost-benefit analysis (CBA). However, environmental impacts are now taking into consideration when decisions have to be made, regarding which large transport projects to implement. This is handled by using environmental impact assessment (EIA). In most real-world policy situations there are many possible alternatives, many uncertainties, many stakeholders, and many consequences of interest. In spite of this, public decision-makers (DMs) have a responsibility to develop and implement policies that have the best chances of contributing to economic development, health, safety, and well-being of their constituencies. DMs are confronted with the difficult task of evaluating potential outcome and choosing policies to achieve the desired outcome in the presence of increasing complexity. Optimisation has been replaced with satisficing, where satisficing is defined as finding an acceptable or satisfactory solution to a problem instead of a socio-economical optimal solution. This calls for new approaches of appraising transport projects that goes beyond what the classical CBA, seems suitable for. As a result of the above mentioned issues regarding conflict of interests, multiple impacts, participation of stakeholders, and sustainability, additional approaches for appraisal have been suggested. Multi-criteria decision analysis (MCDA) has been introduced in the transport planning system, where the appraisal method is an MCDA priority model, with a CBA nested within. However, some concern needs to be addressed; In CBA, the weights (that is, the monetary valuation) are determined on the basis of the best available evidence. In MCDA, weights can reflect evidence, expert opinion, or various preferences. However, defining weights on the basis of preferences introduces subjectivity in the analysis and in the extreme defies its purpose. Therefore it is essential to be able to cope with this introduced subjectivity in order to provide justifiable decision support. This can be dealt with by the use of sensitivity analysis and robustness measures. This also calls for an appraisal methodology which is transparent and able to include the different preferences among experts, the public, and stakeholders. This PhD thesis has three main focuses concerning appraisal of transport projects: • How to reconcile socio-economic and public acceptance to obtain decisions that are both rational and legitimate • How to involve stakeholders in the appraisal • How to apply robustness as mean to improve the appraisal The examination of these main focuses has been two-folded: For each question first its relevance has been argued and afterwards been treated more specifically in regards to methodology and suggestions. The main concern in the PhD study has been to apply robustness as a mean to improve the appraisal of transport projects. In order to accomplish this, sensitivity analysis methods and robustness measures have been elaborated, which examine the subjective part of the MCDA (in form of criteria weights) and its role in decision support making. For this purpose both deterministic and stochastic sensitivity analyses have been developed. In addition, the focus has been formulating a framework which provides a transparent analysis that can be communicated to the DMs. The framework involves CBA for assessing the socio-economic part of the decision problem, EIA to include environmental issues, MCDA to embrace various and often conflicting criteria, and sensitivity analysis for taking into account the interests and preferences of different stakeholders. These various interests and preferences have been revealed by the use of decision conferencing, which engage the stakeholders and provide a common platform for understanding the decision problem. Leading up to this framework, this thesis has also examined the issues of how to provide decision support which both support decisions that are rational in a socio-economic manner as well as being able to legitimatise decisions for the public. Applications of the developed methods are demonstrated by three case studies which are described in four papers written during the PhD study. The potential for transport appraisal demonstrated in this PhD study can be listed in the following main findings. • To reconcile socio-economic analysis and public acceptance it can be recommended to widen the appraisal methodology. • When widening the appraisal methodology, e.g. replacing CBA with EIA+CBA+MCDA+DC, there is a need to balance between the use of sophisticated modelling and emphasising transparency for all. • The conduct of a DC with the participants representing different stakeholders is seen as a way of combining technical, analytical solutions with social group processes aiming for engaging the different stakeholders with consideration of their variability of their preferences. • The different sensitivity analysis and robustness measures developed in the technical part of this PhD study are found to be relevant in supporting the DC group process, as robustness of a recommended solution is major concern in the final steps of decision making. As mentioned, the appraisal of transport projects is a complex issues involving conflict of various interests and this calls for new approaches to the practice of appraisal. The presented appraisal framework is a tool for analysing conflicting preferences, and it is believed that the efficient use of available information from CBA and EIA, combined with the MCDA approach and the robustness measures, establishes a strong decision aid tool for the practical evaluation of larger transport projects.