Using seizures data to describe the scope and scale of transnational organized crime

A few years ago I wrote a brief for the Global Initiative against Transnational Organized Crime (GI-TOC) for their ENACT (Enhancing Africa’s capacity to respond more effectively to transnational organised crime) programme. The brief was about using seizures data to understand the scope and scale of transnational organized crime (TOC). ENACT were interested in whether they could use reported seizures to contribute to a multi-dimensional set of metrics they wanted to develop to measure the scope and scale of organised crime in Africa.

The Global Initiative decided, quite rightly in my view, not to use seizures data. Separately, we discussed the use of expert elicitation as a way to obtain useful knowledge about organized crime. Since then, they have not only developed a suite of indicators for measuring transnational crime in Africa but also a Global Organized Crime Index underpinned by a multi-layered process of expert assessments and evaluations to obtain this information.

The second iteration of the Global Organized Crime Index will be launched later this year. The Global Initiative also decided that they wanted to publish a revised and up-to-date set of briefing papers that informed their decisions in creating  a robust evidence-based metric. 

In my brief I describe different ways in which seizure records are obtained. These  databases of seizure records provide useful information on the nature, provenance and trade routes of seized items for law enforcement agencies. But, they do not necessarily reflect trends in illicit markets.  I show how changes in the number of reported seizures in the database between countries or over time conflates changes in illicit activity with changes in enforcement and reporting.

I discuss an approach, applied to the illegal ivory trade,  that can disentangle differences in enforcement and reporting from trends in illicit activity.  The outcome is a relative index of how trade varies over time and between countries.

These indices are calculated from fitting a statistical model that estimates the relative rates at which seizures are made (seizure rate) and reported (reporting rate) to the database for each country over time. To do this proxy variables that capture reasons why countries differ in their ability to make and report seizures must be developed.

In the brief I reflect on the qualities of the seizures database needed for such an approach to work. Key points are that (1) the database should not be based only on media reports because it is difficult to capture the whims and caprices of media interest and reporting  to estimate the reporting rate and  (2) for any commodity should cover all parts of the trade chain  including all countries and all modes of transport.  More generally, it is important to understand the whole trade chain when deciding what to collect, measure and model.

I hope the paper will provide useful information for those thinking about using seizures data to learn about transnational organized crime and I look forward to reading the rest of the briefing papers which will be coming out in the next few weeks and months.