This thesis examines the issue of participation of affected populations in disaster relief, which is receiving increasing attention from researchers, planners and practitioners. This concern comes out of the widely documented experience in development studies that beneficiary participation is essential for programmes to succeed. Similar arguments are being applied to disaster relief. However, despite much rhetoric, examples of genuine grassroots participation both in relief and development continue to be rare. I review the concept of participation in Chapter One and, in Chapter Two, the many possible reasons as to why participation of beneficiaries continues to be a problematic issue. In Chapter Three, I review the concept of humanitarianism and the implications of changes in humanitarian assistance on participation. In Chapters Four, Five and Six, I present three case studies, different by geographical, socio-political context and type of disaster. All the three studies contain material collected through fieldwork involving a qualitative methodology. I have indicated, in each study, the range of data collection tools used. In Chapter Seven, I compare and evaluate the findings of the three case studies. I present overall conclusions of the thesis in Chapter Eight. The main conclusions of the thesis are that beneficiary participation continues to be a problematic issue because groups that have power derived from ownership of economic resources or politics seem unwilling to share that power with the people they seek to assist. Their unwillingness to do so has, in turn, many causes including, lack of trust by aid organisations of local power structures and organisations, poor bureaucratic orientation, a self-given superiority of moral virtue and technical expertise, and sometimes limitations imposed by operational, structural and accounting procedures. I argue that some of these limitations could be addressed through financially supporting and enhancing the capabilities of member-based grassroots structures. I also argue that more effort needs to be devoted to research on how willingness to adhere to the ideals of humanitarian assistance can be generated on the part of aid agencies and donors.

Les désastres naturels ont des conséquences particulièrement dévastatrices dans les pays en développement où les individus sont hautement vulnérables et les institutions inefficaces. Néanmoins, leurs impacts sur le bien être des ménages et le rôle des autorités publiques restent encore mal compris. En outre, alors que la plupart des études se focalisent sur le risque climatique, les désastres géologiques, et les éruptions volcaniques en particulier, restent peu étudiés. Pourtant, même si elles représentent une fraction marginale des désastres naturels au niveau mondial, les éruptions volcaniques sont une menace majeure dans certains pays tels que l’Indonésie ou l’Equateur. La présente thèse tente donc de contribuer à la littérature à travers trois essais empiriques traitant de l’effet à long terme du risque volcanique sur l’accumulation de capital des ménages, de l’impact d’une éruption sur le capital social ainsi que du rôle des autorités publiques dans les décisions de migration. / Natural disasters have particularly devastating consequences in developing countries where people are highly vulnerable and institutions remain inefficient. Nevertheless, their impacts on households’ well-being and the role of public authorities are, yet, not fully understood. In addition, while most studies focus on climatic risk, geological disasters, and volcanic eruptions in particular, are clearly understudied. However, despite representing a marginal share of natural disasters at the global level, volcanic eruptions are a major threat in some countries, such as Indonesia or Ecuador. The present dissertation tries to contribute to the literature by investigating the long-term effect of volcanic hazard on farmers’ capital accumulation, the impact of an eruption on social capital, as well as the potential mitigating role of public authorities on migration decisions.