Political Science ETDs

Publication Date

Spring 4-30-2020


Why do states intervene in civil conflicts? I argue that a third-party actor's decision to intervene relies in part on the behavior of other potential interveners. Because other interveners have their own interests in the conflict outcome and resources to contribute, intervention decision-making is multi-actor game theoretic problem: external actors contemplate their roles in civil conflicts by weighing the projected impact of their own prospective contributions against their preferences for the internal combatants. However, the impact that a given third party can expect to have on the conflict outcome depends on the contributions of fighting effort that others make when they decide to intervene as well. The case of U.S. foreign policy toward Nicaragua during the revolutionary conflict provides some evidence of complex decision-making and highlights the roles of other actors in shaping the decision to ultimately intervene. For one, the prospect of other international actors getting involved shapes the decision to intervene, as was the case with the U.S. response to Soviet and Cuban efforts in Nicaragua. For another, policymakers within the third-party state can vie for political influence and ultimately also contribute to the intervention decision. Over time, as conflict dynamics shift, new actors emerge, and more information is revealed, would-be interveners update their foreign policy strategies and their affinities for the target state combatants, leading to changes in side-choice and intervention decision-making. In addition to the case study, I also test the robustness of several conflict-level and dyadic-level factors that contribute to the likelihood of intervention using large-N statistical analysis. Ultimately, power, geography, and conflict intensity matter to intervention likelihood, but so do the number of other actors involved and the affinity that third parties have for the target state actors. Finally, I explore the prospect of predicting intervention using both the model predictions from the statistical analyses and an original simulation. I compare the predicted probabilities of each politically relevant actor intervening in Nicaragua to the empirical reality of the case, finding that dyadic-level statistical models do poorly to predict intervention. However, I also provide a demonstration of an iterated game theoretic model using empirical data on states' capabilities and other inputs generated from the qualitative case study to predict side choice across an incomplete network. Overall, there is much evidence to suggest that third-party states make and update their decisions to intervene based in part on the behavior of other external actors with interests in the conflict outcome.

Degree Name

Political Science

Level of Degree


Department Name

Political Science

First Committee Member (Chair)

Christopher K. Butler

Second Committee Member

William Stanley

Third Committee Member

Sergio Ascencio Bonfil

Fourth Committee Member

Cassy Dorff




Intervention, Game theory, Third parties, Civil War, Networks

Document Type