Research

This project examines the impact of nuclear proliferation on regime durability and leader survival, proposing two mechanisms for this interaction. First, a bomb program can generate internal legitimacy by rallying support among domestic audiences. Such initiatives are particularly valuable following contentious successions, providing new leaders with a means of consolidating their rule. Second, successful proliferation attempts can boost external security by preventing regime collapse, either deterring foreign intervention or entrapping great powers into providing costly support to tentative allies. I rely on a mixed-methods research design to test these arguments. Survival analysis, Bayesian modeling, and a battery of robustness tests establish a baseline domestic hazard rate relative to nuclear activity. Next, two in-depth case studies trace the processes linking proliferation and political survival in Israel and South Africa. The analysis concludes with fifteen vignettes that highlight the generalizability and limitations of my theoretical framework.

Can an implementation-driven analysis of project success be used as a more granular measure for assessing World Bank project Effectiveness? By focusing on how projects perform, we attempt to capture variation hitherto unexplored in the aid effectiveness literature. This offers greater precision for diagnosing implementation challenges throughout the project cycle, producing a cross-cutting instrument that reaches across the country-, time-, and sector- based approaches. Using newly-coded data from the Global Delivery Initiative’s “Delivery Challenges in Operations for Development Effectiveness” (DeCODE) database and indicators from over 5000 lending projects (1995-2015), we examine project performance and the achievement of development objectives across 42 specific delivery challenges. Bayesian model averaging is used to holistically assess the relative impact of each challenge alongside a battery of structural and contextual covariates. We find that flaws in project design, ineffective monitoring, and weak organizational capacity have systematically hindered the Bank’s performance and the achievement of indicators. Conversely, while financial instability and weaknesses in stakeholder engagement can hinder success, their identification and treatment ultimately improves project performance.

The post-tenure fate of leaders influences conflict escalation, duration, and termination. Golden parachute offers to embattled leaders can help mitigate the costs of a prolonged intrastate conflict by providing the incumbent a safe exit. However, the number of verified cases of a leader receiving and accepting a golden parachute offer remains small. Using a formal model of conflict termination, this paper seeks to identify the conditions under which embattled leaders accept such offers. We find that golden parachute can create a moral hazard, giving incumbents a fallback option and incentivizing them to fight harder or stall for time. The model also indicates that leaders are unlikely to accept offers unless the terms are very generous; however, the high costs of conflict that might justify generous offers also make it difficult for patrons to extend immunity to leaders who have violated human rights. We present these findings alongside a case study on Manuel Noriega's ouster from Panama in the late 1980s.

Recent empirical brush-clearing exercises have suggested that few if any of the established determinants of proliferation have a substantive impact on the pursuit and acquisition of nuclear weapons. However, such exercises risk under-emphasizing important effects by failing to account for missing data, model sensitivity, and variable selection issues that can lead to Type II errors. Using a panel dataset of 163 variables aggregated into 13 conceptual categories (1945-2015), this paper conducts an exhaustive analysis of the established determinants of nuclear proliferation. After accounting for missingness and collinearity issues, an iterated Bayesian logistic regression approach used to examine a model space of over 74.5 billion unique specifications across over 20 meta-specifications. We triangulate the impact of different explanations for proliferation across differently coded measures, showing that a handful of supply- and demand-side indicators are consistently informative as correlates of proliferation. Specifically, factors such as urbanization, industrial capacity, regional conflict, and enduring rivalry have a substantial impact on the pursuit and acquisition of nuclear weapons. Our findings address a larger debate in the literature on proliferation, while our methodological approach contributes to discussions of quantitative research transparency and replicability.