Decision analysis under uncertainity for sustainable development

• Autores: Fatine Ezbakhe
• Directores de la Tesis: Agustí Pérez Foguet (dir. tes.)
• Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2019
• Idioma: español
• Tribunal Calificador de la Tesis: Alejandro Jimenez Fernández de Palencia (presid.) , Jan Graffelman (secret.) , Germà Coenders Gallart (voc.)
• Texto completo no disponible (Saber más ...)
• Resumen

  • Policy-making for sustainable development becomes more efficient when it is reliably backed by evidence-based decision analysis. Concretely, this is crucial in the planning of public services delivery. By translating "raw" data into information, decision analysis illuminates our judgment, and ultimately the policies we adopt. In the context of public services provision, decision analysis can support the prioritization of policy options and the monitoring of progress. However, most models are deterministic - that is, they do not consider the uncertainty in their evidence. These "incomplete" models, through their impact in policy decisions, can ultimately lead to an inefficient use of resources. The main barriers to a wider incorporation of uncertainty are: (i) the complexity of the approaches currently available, and (ii) the need to develop methods tailored to the specific decision problems faced in public services delivery.

    To overcome these limitations, this thesis intends to facilitate the incorporation of uncertainty in the evidence into decision analysis for sustainable development. We propose two methods. First, a non-compensatory multi-criteria prioritization under uncertainty model. Given multiple criteria and uncertain evidence, the model identifies the best policy option to improve service provision for sustainable development. The non-compensatory nature of our model makes it an attractive alternative to the widely used composite index approach. Second, a compositional trend analysis under uncertainty model to monitor service coverage. By considering the non-negativity and constant-sum constraints of the data, our model provides better estimates for measuring progress than standard statistical approaches.

    These two methods are validated in real case studies in the energy, water and health sectors. We apply our prioritization model to the context of strategic renewable energy planning, and the targeting of water, sanitation and hygiene services. Furthermore, we use our trend analysis model to the global monitoring of water and sanitation and child mortality.

    Our results emphasize the importance of considering and incorporating uncertainty in the evidence into decision analysis, particularly into prioritization and monitoring processes, both central to sustainable development practice.