Nuclear “Pork” Revisited: Organizational Imperatives and Nuclear Weapon Program Abandonment

Since 1941 at least 30 countries have conducted nuclear weapon activities. Over two-thirds of those countries have abandoned their nuclear weapon activities as a matter of policy, making nuclear abandonment the dominant historical tendency. Understanding why countries abandon their nuclear weapon activities holds the promise of allowing policy makers to improve efforts to control the spread of nuclear arms around the world. This dissertation offers a novel explanation for nuclear abandonment. It draws from the literature about bureaucracy and foreign policy to argue that the narrow imperatives of the organizations that conduct nuclear weapon activities can actually make nuclear weapon program abandonment more – rather than less – likely. This stands in stark contrast to the conventional wisdom about bureaucratic imperatives and nuclear weapons, which is that bureaucratic organizations, such as nuclear science agencies and militaries, push governments to acquire nuclear arms. In contrast to that prevailing wisdom, this thesis explains why some of the organizations that governments have charged with executing national nuclear weapon policies face weak incentives to act as nuclear bomb lobbyists. Thus, bureaucratic or organizational imperatives do not only work in favour of nuclear proliferation. They can also work in favour of nuclear abandonment by contributing to a policy environment that makes nuclear abandonment more likely. While there are many explanations for why countries abandon their nuclear weapon activities, most explanations are only substantiated by small, non-random data samples. This dissertation offers a generalizable and probabilistic explanation for nuclear abandonment, and substantiates its theoretical claims with large-N multivariate regressions that use an original data set that covers approximately 670 country-year observations and that exploits natural variation in the types of organizations that have conducted nuclear weapon activities. The large-N empirics are supplemented with two case studies of nuclear weapon policy in Switzerland (1945-1988) and South Africa (1969-1993).