The Stimson Center’s research team followed a three-part methodology: the codebook compilation and review, the data collection and evaluation, and scoring.

Project Codebook

Stimson created a project “codebook” of important security indicators extracted from six International Atomic Energy Agency (IAEA) guidance documents:

1. Code of Conduct on the Safety and Security of Radioactive Sources (IAEA, 2004)
2. Guidance on the Management of Disused Radioactive Sources (IAEA, 2018)
3. Security of Radioactive Material in Use and Storage and of Associated Facilities Implementing Guide (IAEA, 2019)
4. Security of Radioactive Material in Transport Guide (IAEA, 2020)
5. Nuclear Security Recommendations on Radioactive Material and Associated Facilities (IAEA, 2011)
6. UN Recommendations on the Transport of Dangerous Goods, Model Regulations, Rev. 22, Annex I, Part I.

The use of these security indicators builds on an examination of the Code of Conduct (under a previous project funded by the Government of Finland) to identify crucial variables needed for a robust national legal regime to secure radiological sources. For this project, the team sought to integrate additional variables from the five more recent documents to allow for a more complete and current analysis of national legal frameworks for radiological security.  The task of integration proved feasible, but conceptually challenging due to the considerable overlap across the documents. While the overlapping elements buttress the individual guidance and overall security themes found in each, the often-detailed guidance in each document did not always easily mesh. Consequently, some indicators focus on the general theme taken from several more specific guidance elements.

In addition, to help illustrate the enforcement aspects of these legal measures, several of the variables in the dataset identify the kinds of penalties levied, to include imprisonment, fines, and licensing, and the level of authority invested in inspectors to levy penalties, issue warnings, or require immediate action, and other types of penalty options, most typically forfeiture or seizure of materials. This focus stems from one of the purposes of the original study, which was to provide concrete examples of penalty provisions for use in international training activities for INTERPOL and the UN Office of Drugs and Crime (UNODC).

The dataset also includes variables on whether the State has identified a Point of Contact and notified the IAEA of its commitment to the Code, hypothesizing that measures in States that have made such commitments have a greater likelihood of reflecting elements of the Code.

The Stimson team then invited external legal specialists and experts in the security of radioactive material to review the codebook and addressed feedback to the maximum extent possible within the scope and schedule constraints of the project, before starting on the analysis of national legal measures. The resulting codebook comprises 46 security-related variables. 

Together, these security indicators comprise the criteria against which a state’s legal framework is assessed in a Review. For the initial release of the data, the research team grouped these criteria into six thematic/topical categories: Regulatory Body; Source Security; Authorization and Licensing; Training, Awareness, and Information Sharing; Offenses, Violations, and Penalties; and International Commitments.

Project Dataset

In compiling the baseline project dataset of legal measures, the Stimson team relied on publicly available information published by the United Nations Committee Established Pursuant to United Nations Security Council Resolution 1540 (2004), i.e., the 1540 Committee.  UN Security Council Resolution (UNSCR) 1540 requires states not to aid non-state actors in obtaining nuclear, chemical, or biological weapons and their means of delivery, to prohibit a range of associated activities, and to implement controls to prevent the illicit proliferation of materials, technologies, and equipment related to these activities. The 1540 Committee’s Group of Experts periodically compiles information from each UN member state on implementation progress into a country “matrix” that the Committee then approves and publishes on its website.

The Stimson team began its data collection efforts by extracting the names of legal measures potentially relevant to securing radioactive material from the data fields on securing nuclear weapons-related materials found in the approved 2020 matrices for all 193 UN member states. Although the matrices are designed to capture information on nuclear materials, UNSCR 1540 does specifically mention the IAEA Code of Conduct on the Safety and Security of Radioactive Sources in its preamble and further commitment to international treaties within the IAEA framework, which now includes instruments that cover radiological materials, such as the International Convention for the Suppression of Acts of Nuclear Terrorism (ICSANT). In practice, States typically provide information on their management of both radioactive and special nuclear-related materials when they report to the Committee.  These reports establish the foundation for the data collected by the Committee in generating its matrices for each country.

This project supplemented its research with several other sources of information, especially the websites of national regulatory bodies and national gazettes where States officially promulgate their legal measures, where openly available. The team also used information from a variety of IAEA documents, such as regional meetings on nuclear security, and basic internet search tools to point to legal texts and some civil society resources, notably the Nuclear Security Legislation Database compiled by the Verification Research, Training and Information Centre (VERTIC).

After compiling a list of the names of the laws, regulations, decrees, or “hard law” equivalents potentially relevant to securing radioactive material, the research team located the text of these measures and reviewed the text of each legal measure for actual relevance, i.e., a legal requirement to secure or physically protect radiological material.  In most instances, a keyword search for variants on the words radioactive or nuclear could be conducted (including within definitions of hazardous materials or other less obvious provisions). If the measure was found to apply to nuclear and/or radioactive materials, further review of the context set aside the measure if it pertained solely to radiological safety, with no requirement for physical security or protection articulated, or if it specifically excluded radioactive materials to cover only special nuclear materials.

This search and review involved several technical challenges:

• Availability: Although most states freely publish their legal measures online, typically referenced in websites of key government stakeholders, some measures are not available.
• Language: Most countries make their legal measures available online in English translation as well as the original language, but in some instances accurate translations were not available for review. In cases where the research team could not find an English translation and did not read the original language well enough to analyze it for coding, the team utilized Google, DeepL, or Microsoft Word applications to translate the measure.
• Semantics: In many languages, the word for safety and security is the same. As noted above, distinguishing between these two meanings required a closer examination of such texts to interpret if security was a purpose, where further keyword searches for variants of words and phrases like physical protection, guards, malicious intent, and so forth proved useful.

Measures that passed these tests of relevance were then reviewed against the 46 variables in the project codebook and coded for Yes (1), No (0), or Unknown (“.”). These codes were aggregated to produce scores for the state’s radiological security legal framework in six categories and overall.

Planned outreach to national authorities and regulators will confirm completeness (through December 9, 2020) of the legal measure collection underpinning the analysis. National Authority Outreach Status is indicated on each country page as Not Started, In Progress, Complete, or No Response.

Review Scoring and Weighting

Many states have one legal measure which satisfies multiple criteria. Likewise, some criteria are satisfied by multiple legal measures. For the purposes of scoring, if a state had one or more legal measures which satisfied a criteria, then that criteria is noted as a Yes. If the research team was not able to identify relevant legal measures that satisfied a given criteria, then that criteria is noted as a No. In a few rare cases, a criteria is noted Unknown, largely in situations where legal language was unclear, ambiguous or the researcher was otherwise uncertain about how to code. (To be clear,  no bonus is awarded for redundancy or efficiency: If a given state has 3 legal measures that satisfy one criteria, the state receives a single Yes for the criteria. If one legal measure satisfies 3 criteria, then each criteria is a Yes.)

A state’s score in a review is, then, simply the aggregate number of criteria satisfied. While the criteria are organized into categories, they are not weighted; equal weight is assigned to each criteria and category.