Abstract:
We exploit 0.5ê×0.5ê raster data for mainland Southeast Asia from 2010 to 2020 to document a non-linear relationship between extreme temperature days and conflict. We show that the occurrence of conflict events increases with extreme maximum temperature days, whereas days with extreme minimum temperature decrease the occurrence of conflict. Because climate change makes both maximum and minimum temperature extremes more likely, these effects partially offset each other on aggregate. However, our results further suggest that the impact of extreme maximum and minimum temperature days differs for the type of conflict, actors involved and population affected, indicating complex distributional consequences.