The Epidemiology of War, 1816-1980

Abstract

Quantitative research into the causes of war is influenced by the choice that has to be made between time-series designs on the one hand and cross-sectional designs on the other. The assumption of independence is important for both of them. The present article investigates whether warfare is clustered in time and in space. Using a method of analysis proposed by epidemiologists, significant evidence is found for time-space interaction not only in participation in war but also in outbreaks of war.