Open Access

This dissertation delves into the complex dynamics of U.S. nuclear weapon deployments. Scholarly work on nuclear weapons has predominantly focused on strategic forces. However, many escalation scenarios foresee the use of tactical nuclear weapons as a precursor, which could eventually lead to an all-out nuclear war. These are exactly the weapons that the United States has deployed across the world, and we know little about their strategic causes and consequences. My study addresses this gap in the literature. The first part of this dissertation investigates the causes of U.S. nuclear weapon deployments. It emphasizes the importance of assurance, which has commonly been neglected or subsumed under non-proliferation. Through a comprehensive conceptualization, I distinguish assurance from other strategies of interstate interaction, explaining how the U.S. uses nuclear weapon deployments to assure allies and maintain its sphere of influence. Utilizing an original dataset, I find evidence that largely confirms my assurance hypotheses. With regard to alternative explanations, the power projection hypotheses are supported, but the extended deterrence hypotheses are not. These results contribute to the underdeveloped literature on assurance and provide valuable insights into the strategic considerations underlying U.S. nuclear weapon deployments. The second part, which is co-authored with Julian Wucherpfennig, examines the consequences of U.S. military deployments, both conventional and nuclear. The theory of free-riding in military alliances postulates that member states piggyback on security commitments by guardian states. We argue that free-riding critically implies that foreign military deployments positively affect subjective feelings of security – in other words: that they assure. In contrast, if citizens hold skeptical views of the guardian or do not subscribe to the (extended) deterrence logic, they might experience an increase in threat perception. We investigate how foreign military deployments impact attitudes toward defense policies in host states through an observational data analysis and a survey experiment. While nuclear and conventional troop deployments decrease citizens’ subjective need for defense, they do not necessarily increase their sense of protection, challenging the logic of free-riding. Taken together, my findings suggest a mismatch between objectives and outcomes. While I find evidence that nuclear weapons are deployed to assure allies, this aim does not appear to be effective at the micro-level.

In a recent contribution to this journal, McAlexander (2020, hereafter McAlexander) seeks to re-analyze Wucherpfennig, Hunziker and Cederman (2016, hereafter WHC) who instrument the effect of ethnic group inclusion on civil conflict by exploiting differences in colonial governance between the French and the British empire. McAlexander proposes a research design that replaces the between-colony dimension of WHC’s comparison with a continuous measure of indirect rule that varies be-tween British colonies. We show that McAlexander’s study hardly poses a viable reanalysis because his approach is compromised by problems of ecological inference and post-treatment confounding, lead-ing to biased inference by design. We propose a more informative reanalysis, indicating that British colonialism built on suitable customary institutions, unlike the French empire. Although tentative, our analytical extension lends support to WHC’s original findings, most importantly that inclusion reduces the risk of post-colonial civil conflict.