Review of “The Efficiency/Security Trade-off in Criminal Networks”
This is a review of a recent (2006) research paper in criminology[1] that compares and contrasts a criminal-type covert network with Krebs’ terrorist covert network.
For comparison, the authors of this paper selected a representative criminal network, in the form of the Caviar drug trafficking enterprise operating in Canada in the mid-90’s. Their data was gleaned from court proceedings from trails of known members of the organization (in contrast to news articles, which formed Krebs’ source). In addition, their link types were different. Where Krebs used familial and same-school student relationships, Morselli et al. used a call-graph (e.g. a link between two person nodes is added to the graph if they communicated by phone).
Similar to Krebs, the authors also made a distinction in their analysis between actual perpetrators of drug smuggling (active), and those that contributed financing and other resources to the endeavor (passive). The most striking difference between these two networks was the fact that, when the passive actors were added to the network consisting of active participants, terrorist networks improved efficiency (where efficiency is measured by network centrality metrics) with reduced security. Conversely, criminal networks showed the exact opposite.
The authors posit that criminal networks grow organically from a central core, an aspect of formation that terrorist networks do not share. The implication is that covert networks are not all the same when it comes to leadership, as the centralized “kingpin” model of core leadership in criminal enterprise contrasts sharply with more decentralized nature of terrorist cells. This reminds me of the dragon versus the hydra, where the dragon has a single head (but is able to breathe fire) and the hydra has many heads that regenerate when severed from its body.
In criminal networks, the traditional form of network disruption comes from making deals with lower-level criminals in the hierarchy to try to sever the head of the organization. In terrorist networks, this approach fails, because terrorist networks simply re-grow the head of their cell though back-channel communications. The difficulty of analyzing covert networks is always centered around the fact that their secretive nature hides much of the network from view, and one cannot be certain (even after discovery) that one is able to observe the actions of all participants. A challenge might therefore to find networks that emulate similar structure. With that in mind, decentralized networks, like those of peer-to-peer or blogosphere networks might serve as interesting proxies that exhibit similar structure to terrorist networks.
[1] Morselli, Carlo, Petit, Katia and Giguère, Cynthia, 2006. “The Efficiency/Security Trade-off in Criminal Networks“.