The Half-Lives of Others: The Democratic Advantage in Nuclear Intelligence Assessment

Abstract

An accurate nuclear proliferation assessment requires an understanding of a country's nuclear capabilities and intentions. Capabilities in the form of fissile materials and technologies may be well-hidden secrets that are difficult to detect from abroad. Nuclear intentions are mysteries. All else being equal, the intelligence agencies that can best estimate these mysteries will perform the most accurate proliferation assessments.

To explain the conditions under which intelligence assessments can most accurately capture the likelihood of another country's acquisition of nuclear weapons, this dissertation develops a theory of bounded perception in intelligence assessment. Regime type is the key to this theory: Democratic leaders can politicize assessments, but authoritarian regimes must politicize their assessors. The former episodically corrupts intelligence work while the latter systematically undermines the creation of accurate intelligence assessments. This theory is tested against competing explanations of states as fearful overestimators or rational information processors by comparing the ability of multiple intelligence agencies to assess foreign states' nuclear weapon proliferation intentions and capabilities.

A comparative research design of this dissertation makes it possible to keep several, possibly confounding, variables constant in relation to the intelligence collection capability and nature of the assessed country's nuclear program. In this approach, the strengths of multiple inferential methodologies are leveraged. Each empirical chapter initially conducts tests on a self-coded global panel dataset of nuclear assessments. This step is followed by comparative congruence verifications for assessments of India, Argentina, and Pakistan that were performed by the UK, USA, Sweden, Switzerland, and both Germanys. The theories' causal mechanisms are subsequently tested by tracing the process that led to variations in the accuracy of East German, British, and American assessments of the West German nuclear program.