Revisiting the puzzle of endogenous nuclear proliferation

Abstract

Nuclear proliferation literature typically differentiates supply-side and demand-side factors influencing the spread of nuclear weapons. These distinct approaches to the proliferation puzzle raise the following empirical questions: Does nuclear supply stimulate states’ demand for nuclear weapons? Conversely, does the demand for nuclear weapons really facilitate the acquisition of nuclear supply? If such endogeneity exists between the demand-side and supply-side determinants, how would it cause empirical bias in the estimation of their effects on nuclear proliferation? This article aims to unpack endogenous mechanisms of nuclear demand and nuclear supply over the course of nuclear proliferation. In particular, it examines two potential sources of endogeneity: (1) simultaneous interactions between states’ nuclear development decisions and nuclear technological capability and (2) selection bias in nuclear development. To address each source of endogeneity, simultaneous equation models and the duration models with selection are estimated, respectively. Contrary to what recent supply-side literature suggests, the empirical analyses reveal that states’ nuclear demand is primarily driven by external security threats instead of their existing nuclear technology, and that their successful acquisition of nuclear technology mainly follows as the result of nuclear development efforts but does not necessarily depend on individual supply-side factors. This article addresses the typical inference issues in nuclear proliferation research and contributes to our synthetic understanding of proliferation mechanisms.