Panel Paper: Structure and Dynamics In Dark Networks: Applying Formal Network Analysis to Criminal and Illicit Activity

Friday, November 9, 2012 : 8:40 AM
Schaefer (Sheraton Baltimore City Center Hotel)

*Names in bold indicate Presenter

David Dornisch, US Government Accountability Office

Since the early 1990s a burgeoning literature on the structure and functioning of networks of relationships in “legitimate” domains of social, political, and economic life has developed. Network effects and outcomes have been viewed as increasingly relevant in areas such as business organization, management, public administration, social movements, international trade and organizations, community politics, civil society, and many others. In these areas, large research literatures have developed in which formal quantitative network analysis has been applied extensively.

At the same time, there has been comparatively little research applying formal network analysis to what have been called “dark networks,” networks that enable, result from, or include elements of criminal, illicit, or fraudulent activity. While many researchers have recognized that dark networks and illicit activity are highly relevant in the areas noted above, the emphasis in applying network analysis has been on normative questions regarding how networks influence or are associated with efficiency, competitiveness, collaboration, trust, management, innovation, and other issues. Moreover, the ability to do analysis of dark networks has been limited by difficulties in collecting data on criminal/illicit/fraudulent activity. Moreover, beyond typical difficulties inherent to all network/relational data there are additional difficulties concerning potentially missing or misrepresented dark network data.

This paper reviews the existing literature using formal network analysis methods to analyze, detect, or develop interventions into criminal and illicit activity. This body of research has primarily emerged in the last decade and still remains somewhat limited in its scope and sophistication, including the breadth of topics considered, the types of methodologies and analyses applied, and the sophistication and depth of the research questions asked. Nevertheless, research of this type, which has primarily emerged in the fields of criminology, sociology, and computational science, is growing and holds significant potential for further development.

The paper proceeds in two main sections. First, it reviews existing network analysis applications to dark networks in general. To date, the existing research can be categorized into four types:

1) Interpretive approaches using visualization or core network measurement techniques to explain criminal or illicit activity;

2) Extensions of interpretive approaches to detection of and/or intervention in illicit/criminal networks (esp. in criminology);

3) Statistical analyses of associations and causal relationships between network variables and key behavioral outcomes;

4) Data-Mining Approaches to Analyzing and Detecting Criminal/Illicit Activity .

For each of these types, the review will cover the following issues:

1) Data sources and data collection methods;

2) Research topics and questions, including the relationships and associations that have been examined between network measures and methods and key criminal/illicit behaviors and outcomes;

3) Key Results;

4) Limitations of the research.

Second, based on the analysis in section 1 above the paper assesses potential extensions of the existing research and avenues for new research. Particular attention will be placed on developing ideas for new data sources/collection efforts, research questions/topics, and network analytical methods. Ultimately, this analysis will help move the research literature forward in developing new insights into the underlying structures and dynamics characterizing dark networks.