Caught in the Maelstrom, Catching the Tide : Micro-level Responses to Climate-Related Hazards and Conflict

Abstract: This dissertation contributes to the literature on the nexus between climate-related disaster and conflict. One of the main sources of uncertainty regarding the effects of climate change on conflict stems from limited understanding of the driving mechanisms and the political contexts under which these effects materialize. People respond to climate variability through a variety of coping mechanisms that, in turn, determine social outcomes. Yet, these remain little understood. Addressing this overarching issue, the first part of the dissertation (Paper I and II) studies the conditions and possible pathways through which climate events shape conflict-related outcomes. Paper I studies the role of local state institutions in mitigating adverse effects of flooding on communal conflict risk in Sub-Saharan Africa. Statistical evidence shows that areas experiencing flood disaster are more likely to experience communal violence in a context where citizens distrust the local state institutions. Paper II examines the mediating effect of migration as a transitional pathway, connecting natural hazards and social unrest in urban areas of Bangladesh. It finds that although natural hazards shape migratory decisions, this climate-related mobility does not affect the likelihood of protests in host communities. The last two essays of the dissertation study responses to climate-related disasters taking place in contexts of ongoing armed conflicts, and focus on household-level compound effects of these two calamities. Using interpretative machine learning methods, Paper III explores distress-related mobility as an outcome of interest. We show that compound effects can manifest differently in the same communities, since a significant amount of variation exists in the mobility patterns of people affected by the same event. Paper IV focuses on how state response to natural-hazard-related disasters and conflicts jointly affect levels of political trust in Pakistan. The evidence suggests that state responses to disasters can mitigate negative political consequences of conflict exposure in the context of non-state conflict, but not when communities have been exposed to violence with the involvement of the state itself. In sum, this dissertation makes both theoretical and empirical contributions to the research on climate and conflict research, as well as to the scholarship on disaster risk reduction more broadly.