Conceptualising non-state security

Thesis

Why do states choose to engage with non-state security apparatuses when faced with domestic threats? This research tackles two distinct challenges. First, I seek to understand how non-state security actors are organized and how they operate. Second, I assess the circumstances under which developing states choose to outsource their coercive capabilities to non-state security actors as opposed to their own militaries or other state militaries. I answer these puzzles using a novel dataset on non-state security actors, their characteristics, and their operations. I identify four latent clusters of non-state security actor using unsupervised machine learning techniques. Using this typology, I develop a theory of why states might choose to engage with different forms of non-state security, testing these theories with a number of prediction models. I find suggestive evidence that states with low levels of military diversity are more likely to outsource their coercive powers to non-state security actors who provide multiple services across multiple domains, essentially serving as a surrogate army for a lower cost. Implications are studied and discussed using a comparative case study of four non-state security actors.

Authors: Brailey, TJ

• DOI: 10.5287/ora-pzkreyeqj
• Type of award: MPhil
• Level of award: Masters
• Awarding institution: University of Oxford
• Language: English
• Keywords: mixed methods; defence contractors; unsupervised machine learning; private military companies; private security companies; mercenaries
• Subjects: Civil war; Non-state actors (International relations); Politics and government