Civil Conflict and Rainfall in Sub-Saharan Africa

The scholarship on civil conflicts has actively debated the link between rainfall and conflict onset over the last 15 years, since the 2004 publication of a seminal paper by Miguel, Satyanath and Sergenti. Despite the large volume of research, no consensus was found in the following two areas: (1) whether rainfall and the likelihood of conflict have a positive, negative, or non-linear relationship, and (2) the causal mechanism of how rainfall can affect conflict proneness or conflict intensity other than through economic growth. We make progress in both of these areas. First, we use flexible functional form models and generalized additive models to explore the relationship between the rainfall and conflict onset and conflict intensity, using various social conflicts beyond civil wars. Second, based on the detailed description of each conflict episode in the Armed Conflict Location & Event Data Project (ACLED) data, we coded various motivations of conflicts by different types. In particular, we coded whether a conflict was motivated by (a) citizens’ frustration due to water scarcity, (b) flood, (c) price instabilities, or (d) its severe effects on the main consumers of water for living, such as farmers and pastoralists. We examined through which of these channels rainfall affects conflict onset and intensity in Sub-Saharan Africa