Temperature and conflict in latin America

Abstract

Using a unique dataset in Latin America, I exploit the nonlinear relation between temperature and conflict onset in the region. I find that an increase in temperature in a specific month increases the number of conflicts started in that same month with more than 30 percent compared to the average number of conflict onsets. Diving more into the different types of civil conflict, I find that temperature almost exclusively impacts non-state-based conflicts. These results shows the importance of investigating this relationship in Latin America, a region not often used in this field of research. Studying this effect could help policymakers understand where the high levels of conflict are coming from and shift their focus more towards fighting climate change.