Disaster Dips? The Link between Natural Disasters and Violence in Indonesian Provinces, 1990-2003

Abstract

Climate change is likely to cause an increase in natural disasters such as floods, droughts, and cyclones – factors that have been linked to hunger, diseases, and violent conflict. We test two competing accounts on the effects of disasters on violent conflict. Whereas environmental security holds that increased resource pressure under certain circumstances increases conflict risk, disaster sociology’s main finding is that social cohesion is strengthened in the aftermath of disasters. In order to assess potential future consequences of climate change on violence we use a recently collected dataset on low-level violence in Indonesia and explore if natural disasters affect the likelihood of violence at the province-month level for the period 1990-2003. This represents an innovation to a field which is dominated by national level studies on large-scale violence such as civil wars. Of the two forms of violence we investigate, we find that disasters decrease the risk of inter-group violence. The effect is quite small, however. We find the same conflict reducing tendency, although not significant, for less organized violence. Thereafter we investigate whether disasters have a greater impact on violence when poverty is high, land is scarce, there is high inter-ethnic inequality, or a province is dominated demographically by two groups. Contrary to our expectations, we do not find much support for conditional effects. Following environmental security theory, disasters should increase the risk of violence. Our findings do not support this. The results resonate better with disaster sociology, which emphasize increased social cohesion in the wake of disasters.