Sex Trafficking Networks in Boston

(Jessica Tsang) 

Background: Human Trafficking is an alarming global phenomenon. In Boston, law enforcement officials are stepping up efforts to find human trafficking rings, and to protect victims[i]. With the overwhelming nature of the crime, tools such as SNA can be used to help law enforcement make sense of the data, and outline the networks, so that information can be used efficiently, and interventions can be planned intelligently, and safely.

Research Question: Are organized crime groups in Boston using the same networks for drug trafficking and sex trafficking, or are the networks specialized?

Hypothesis: There is an overlap in networks for selling drugs and human beings. Established criminal networks could utilize their connections for trafficking in similar ways, regardless of which commodity is being moved.

Alternative hypothesis:Networks could be more specialized for certain cover businesses, such as massage parlors. Some massage parlors in Chinatown have proven to be fronts for sex trafficking[ii]. This type of trafficking is probably more specialized, to massage parlor owners, and victims coming from Asia. There are probably fewer overlaps to drug activity or gang activity in this type of network.

Data: It would be difficult to get all of the necessary data to properly analyze the networks, which is a common problem when studying sex trafficking. An ideal place to start would be with arrest records for sex trafficking charges in Boston, as well as arrest records for drug offenses in Boston. It would be interesting to collect data both on buyers and sellers, to see if customers overlap as well (customers could be a phase II project if necessary). In addition to getting the data, a challenge would be the sheer size of the network. Looking at sex trafficking alone (and eliminating other forms of human trafficking) would help to limit the scope a bit. Ideally, this data would be provided through a partnership with local law enforcement officials, who are already engaged in fighting human trafficking, and looking for better ways to target their resources.

It would also be useful to have cell phone and email data, in order to establish connections between perpetrators, and to determine information flows and directionality. This would probably need to be done once the network is established, and possibly limited to certain important nodes (due to the sheer volume of information and the difficulty in gaining access to it). It may be possible instead to establish connections through attribute data, such as address, phone number, or affiliation with certain gangs or criminal groups.  

Attribute data that would be desired for everyone in the network would include, but not be limited to:

·         Gender

·         Age

·         Nationality

·         Legal status in the U.S.

·         Previous legal offenses

·         Any known affiliations to organized crime or gangs

·         Address

·         Phone number

·         Location/Method used for the sale/purchase (street-based, hotel/house-based, online, massage parlor)

Analysis: To analyze the network, all data would be loaded into UCINET and NetDraw. The first task would simply be to establish the network. Next, it would be important to see which nodes have arrest records for both drug-related offenses and sex trafficking offenses. The percentage of nodes with overlapping charges would be a good indicator of how much overlap exists in the network between drug trafficking and sex trafficking. It would be particularly interesting to isolate the group with overlapping charges and to examine that network more closely based on whether each node was charged with selling (drugs or sex) or facilitating trafficking, or buying (drugs or sex). This would provide some information about the flow and directionality of the criminal activities. Across the whole network, it would be helpful to look at various pieces of attribute data to determine the make up of the network. Gender would be a good starting point – to determine if the network is male dominated, or whether there is more female participation than expected. Examining nationality and age would provide key demographic information, and may establish patterns. Looking at the location/method of the offense (such as massage parlors) would be important in relation to any overlaps with drug charges. It would be crucial to look for patterns, based on the type of location/method, between location/method and overlaps or lack of overlaps. It would also be crucial to examine any data on affiliations with gangs or criminal groups, to look for patterns. On the network level, centrality measures should be calculated to determine the density of the network.

            At the node level, centrality measure should be calculated, and examined more closely. They should be used in conjunction with analysis of key ego networks, factions, and cliques. Faction and clique analysis will be useful to provide specific groups that could be targeted by law enforcement. One hypothesis is that factions and cliques would form based on affiliation with a criminal group – it is likely that a particular law enforcement unit is already aware of that group, and has specialized knowledge which could be used to determine the best strategy moving forward in terms of a raid, arrests, etc. Ego networks of key players may be able to show links between different parts of the network, which might explain movement along the supply chain. To determine key individuals, centrality measures should be assessed. Important measures include: betweenness, Eigenvector, and closeness. Someone with high betweenness could be seen as a “gatekeeper”[iii] or a broker, and could be a very significant node in the network. Someone with high Eigenvector would be well connected to influential nodes in the network, and therefore might either be a leader in the network, or might provide access to leaders in the network. A node with a low closeness provides many connections to other nodes in the network – so interviewing that one individual could provide a wealth of information.

Conclusion: This SNA project would provide a better understanding of the existing networks in Boston used for sex trafficking. Gaining insight into this network can have many benefits, particularly for law enforcement agencies trying to fight this crime. Understanding the links between drug trafficking and sex trafficking, or criminal organizations and trafficking, can help police target their interventions and raids appropriately. It also allows for information sharing across specialized police units, such as those working on human trafficking, and DEA agents working on drug crime. It may illuminate patterns with regards to certain gangs or criminal organizations that are already under surveillance for other crimes. It could also provide surprising results, which might highlight holes in the current strategy for dealing with this growing problem. Some of the conclusions reached in the analysis may make it possible to create tailored prevention programs that could interrupt the trade and perhaps protect some victims.

---

[i] Fox News Boston, Human Trafficking Rings too Close to Home, February 12, 2015. Accessed October 22^, 2015. Available at: http://www.myfoxboston.com/story/28090179/human-trafficking-rings-too-close-to-home

[ii] Kathy Curran, 5 Investigates undercover finds prostitution, ‘human trafficking’: Massage parlors raided by Boston Police, November 24, 2014. Accessed October 22^, 2015. Available at: http://www.wcvb.com/news/5-investigates-undercover-finds-prostitution-human-trafficking/29888320

[iii] Cockbain and Brayley, Internal child sex trafficking: Exploring cooffending and co-victimisation through social network analysis, November 3, 2011. Accessed October 22, 2015. Available at: http://www.ucl.ac.uk/jdi/events/int-CIA-conf/ICIAC11_Slides/ICIAC11_1B_ECockbain