Cumulative Dynamics and Strategic Assessment: U.S. Military Decision Making in Iraq, Vietnam, and the American Indian Wars 

Advisors: Stephen M. Walt, Robert H. Bates, Monica Duffy Toft, Richard J. Zeckhauser

My dissertation examines why policymakers often struggle to assess their progress and why they often fail to realize their strategic mistakes. This subject is relevant to many areas of politics and economics, and it is especially salient in the study of armed conflict. For example, the United States has been fighting the Taliban for over a decade and yet officials, experts, and the general public are still not close to consensus in estimating how much longer it might take or how much more it might cost to contain the insurgency. Similarly, in the literatures on U.S. decision making in Vietnam, German decision making in World War I, Japanese decision making in World War II, and many other cases, scholars frequently ask why decision makers find it so difficult to evaluate their policies, or why they often stick to unsuccessful strategies for so long. This behavior is especially puzzling in light of prominent theoretical frameworks (especially the literature on the bargaining model of war) which predict that the longer decision makers go without achieving their objectives, the more pessimistic they should become about their ability to do so, and the more likely they should be to change course. In my dissertation, I challenge those ideas and I explain why we should often expect the very opposite.

The theoretical crux of my dissertation is that standard models of learning and adaptation (as well as many people’s basic intuition) revolve around the assumption that decision makers are observing repeated processes that are analogous to slot machines or roulette wheels – but in war and other contexts, decision makers often deal with cumulative processes which have very different dynamics, and a very different logic when it comes to forming and revising expectations. I formally explain how cumulative processes have counterintuitive dynamics and how they present genuine analytic challenges that scholars do not generally take into account. Then I apply this theoretical framework to shed new light on salient policy debates and historical experience. The case that I examine most extensively is U.S. military strategy during the American Indian Wars, an experience that is unusually well-suited for a study of strategic decision making, and for which I exploit new, event-level data spanning roughly 3,000 frontier engagements.

Together, the theoretical and empirical arguments in my dissertation provide a new perspective on why decision makers in many fields are often so reluctant to change their policies, even when those policies do not seem to be succeeding. This is one of the most common themes in scholarship on international relations and security studies, and I hope that my dissertation leads scholars to rethink conventional wisdom on this subject.

Here is the full version of my dissertation. Here, you will find a current working paper that lays out some of the main arguments from my dissertation in more detail. Please feel free to email me with any questions about the project.