Individual Submission Summary
Share...

Direct link:

Download

Discovery of Covert Nuclear Facilities: How Secret is Secret?

Fri, August 30, 4:00 to 5:30pm, Hilton, Columbia 8

Abstract

States that develop latent nuclear capacity and nuclear weapons often initiate programs in secret. Covert development may last until a state tests a nuclear weapon or covert installations are discovered by the international community (e.g. Natanz in Iran, Nahal Soreq in Israel). Slightly more than half (132/253) of enrichment and reprocessing (ENR) facilities were initially covert; yet, we know very little about what affects the likelihood of discovering a covert nuclear facility. We examine the duration of secrecy—the time between initial covert facility construction and discovery by the international community (not state intelligence apparatus). We argue that the discovery of a covert ENR facility is affected by a state’s relationship with the US, existing nuclear infrastructure, geographic topography, and commercial satellite availability. We test our hypotheses using data on ENR facilities from 1939-2012 to identify the duration a facility remains secret. First, we find that states with an adversarial position towards the US are less able to hide ENR facilities. We argue the US leverages economic and political mechanisms as well as the nuclear scientific community to promote non-proliferation objectives. However, the US does so in an inconsistent manner as duration of secrecy between friend and adversary discovery is not equal. Second, we find that states with large nuclear infrastructures (e.g. existing military nuclear complexes, civilian nuclear fuel cycle infrastructure) are less able to hide covert ENR facilities. Counterintuitively, we argue that nuclear industry is so regulated and regimented that facilities built outside civilian infrastructure noticeably divert resources (material and personnel), increasing the likelihood of discovery. Third, we find mountainous terrain decreases the likelihood of discovery. A state’s geographic size has no effect on duration of secrecy. Fourth, we find, despite high profile instances of satellite technology assisting in facility discovery, the empirical relationship between satellite technology introduction and decreased duration of security exists but is sensitive to model specification. Similarly, IAEA Additional Protocols are associated with decreased duration of secrecy but are sensitive to model specification. The result is potentially an artefact of the data and limited cases. The results as a whole have broad implications for nuclear policy and international security. In a world where traditional elites are increasingly challenged by leaders with non-traditional backgrounds, pursuit of nuclear latency justified through populist nationalist pride or security concerns challenge the existing inspection regime.

Author