Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Arts and Sciences


Political Science

First Advisor

Dr. Clayton L. Thyne


Military mutinies are occurring more frequently in the last two decades than ever before. Mutinies impact every region of the world. Given that they are occurring more frequently, impact every region, and often have disastrous implications, scholars must answer the foundational question: why do mutinies occur? What are the proximate domestic conditions that give rise to military mutinies? This project makes three contributions. First, I set out to formally define mutinies and collect a new dataset that will allow scholars to examine mutinies empirically. Second, I present a theoretical framework that explains when and why mutinies will occur. Finally, I present three novel empirical tests of the theory.

The first portion of this dissertation defines mutinies and describes the data collection process. I present the Military Mutinies and Defections Databases (MMDD). Using news articles from various sources, I code 460 mutiny events from 1945 – present day. I code a number of other variables that give users details about the event, such as: whether or not violence was used, whether or not civilians were killed, and whether or not soldiers defected from the military apparatus.

Next, I utilize a nested principal agent model to describe when mutinies are likely to occur. Agent models describe hierarchical relationships of delegation. A nested structure allows for multiple agents and multiple principals in a given model. I apply this nested structure to the military to generate three various nests. The first examines foot soldiers as an agent of the military leadership. In this nest, policy failures (e.g., bad strategy) secured by the military leadership will drive foot soldier mutinies. The second nest explores foot soldiers as agents of the executive, a civilian principal. In this nest, I expect that situations that place soldiers in conflict with the executive will generate shirking. The final nest considers foot soldiers and military leadership as collective agents of the executive. I theorize that risk aversion and divergent preferences will drive shirking, or mutinies, in this nested structure. The final nest presents an interesting trade-off for a coup-worried leader. I argue that while executives can utilize regime securing strategies, such strategies might actually agitate the military and drive low level military rebellions. Coup proofing, a common practice among executives that are worried they will be ousted by the military, effectively wards of coups but can generate unintended consequences. Specifically, I expect that counterbalancing measures and other coup proofing tactics should spur mutinies because the intended purpose of these measures is to create coordination challenges which likely spur military splintering.

The first empirical chapter sets out to explore the relationship between civil conflict and the likelihood of mutinies. I expect that when civil wars are extremely bloody or long lasting, mutinies will be more likely as war-weary soldiers no longer want to invest in the war effort. I find evidence that indeed civil war intensity and duration contribute to the probability of a state experiencing a mutiny. The second empirical chapter explores scenarios that pit foot soldiers preferences against the executive’s. I expect that scenarios that impose steep costs on foot soldiers, yet provide some benefit to the executive are likely to spur mutinies. I find evidence that protest events and divisionary conflict spur mutinies. The final empirical chapter explores the military apparatus as a whole. I find that coup proofing measures increase the likelihood of mutinies. Additionally, I find that scenarios that are likely to spur widespread dissent among military actors will increase the likelihood of a mutiny in the context of steep coordination challenges that stifle coup activity.

The final chapter concludes by providing policy recommendations. I offer recommendations for leader states (e.g., major powers and democratic leaders in the international system) and for states experiencing mutinies. I conclude by discussing the many possible extensions for this project. This section seeks to emphasis the fact that this is a young, novel research program with many promising avenues for future research.

Digital Object Identifier (DOI)