More than one-quarter of the world’s population lives in water-scarce areas, while most countries share at least one transboundary river. If water scarcity is this prevalent, should we expect riparian countries to fight over the water allocation of shared rivers? To answer this question, I develop a modified one-shot three-stage river-sharing game where countries can resort to force to solve their water allocation problem. Using backward induction, I solve for the probability of the downstream country initiating conflict against the upstream country and the likelihood of the latter responding with force to the former’s hostile actions. I test the model empirically using a set of all upstream–downstream riparian dyads with available data from AQUASTAT and the Correlates of War Project for the years 1960–2010. The main contribution of this paper is that it demonstrates how upstream and downstream riparian countries differ in their decision to use force against the other country when experiencing water scarcity. I find that water scarcity increases the likelihood of the downstream country initiating the conflict, but it has no effect on the upstream country’s likelihood of responding with force. If history is a predictor of the future, then the results imply that as more riparian countries become water-scarce, militarized conflicts between upstream and downstream countries are likely to increase, especially if there is heterogeneity in water availability between the riparian dyad.
Backward induction in games without perfect recall
