Although island disputes have returned to the geopolitical theatre on a small scale, to date, virtually no previous research on territorial interstate island disputes and violent escalation exists. This paper argues that when an island is positioned in a strategic location, because of its unique attacking, defending, and trade capabilities this may induce a willingness towards- and eventual use of violence in the attempt to conquer or defend the territory. This paper attempts to answer: under what circumstances do island disputes escalate? by modelling the influence of strategic locations on violent island disputes. It draws observations from Altman (2020c) and a novel data frame (1920 – 2020) with additional cases and an alternative operationalization of strategic locations along important lines of communication. It finds that island disputes are more likely to occur without BRD than non-island disputes. Meanwhile, island disputes escalate violently more often than they do not. An island’s strategic location, notwithstanding a broad or narrow operationalization, does not have a statistically significant effect on a violent outcome of a dispute. Instead, the presence of military garrisons, ceteris paribus, resulted in the most statistically significant effect. Consequentially, the causal mechanisms were adapted to include military garrisons.

In an attempt to bridge the gap between theories of violent escalation and those of genocide, this paper theorizes genocide to be a strategic choice by leaders in response to a situation which they perceive to lack alternatives. This situation is expected to evolve out of a violent escalation, more precisely civil war. The empirical test consists of a structured focused comparison of one positive and one negative case; namely the conflict in Bosnia-Herzegovina and the conflict in Georgia over the region Abkhazia. The finding gives some evidence to the theory, however a more adequate theory needs to also involve a theorization of the ability to perpetrate genocide and not only of a lack of other alternatives. The study builds on previous research on the relationship between violent escalation and genocide, and findings are in line with existing research.