Abstract
With the steady decrease in freshwater availability and the growth of world population, freshwater distribution is likely to become an increasingly prominent source of conflict. Given that civil wars by definition occur within national boundaries and therefore involve factors that vary within the country in which they occur, past studies examining the determinants of civil conflict using the state as the unit of observation are inherently flawed. Improvements in the accuracy of the delineations of such conflict should increase the ability to determine the local determinants of civil war. This analysis focuses on Africa between 1980 and 2001, utilizing a refined version of the International Peace Research Institute, Oslo (PRIO) Armed Conflict dataset at the 0.25-degree resolution, and juxtaposing this data on rainfall data using GIS analysis. Using socioeconomic factors as control variables, we perform a logistic regression analysis, comparing the correlation coefficients from this analysis to those of a previous study using conflict centroids with radii as the units of observation. Thus, this study aims to explore the effect a change in the spatial level of resolution of conflict data has on our ability to analyze geographic determinants of war outbreak. We find that, despite the altered delineation of conflict data, the statistical analysis produces similar results that confirm the causal relationship between drought and civil war outbreak. This calls to attention the need for further international cooperation in assisting developing nations to develop infrastructure for the sustainable management of water resources.
This work is licensed under a Creative Commons Attribution 4.0 International License.