Human Trafficking Index: Using Data to Fight Modern Slavery
The Problem: A Lack of Direction and Data
Human trafficking is one of the most challenging crimes to identify. It usually requires survivors and victims to come forward and report their abuse – a daunting task that may not even be possible for those who are controlled and threatened by their traffickers. Because of these precarious circumstances, the necessary intelligence for understanding how different trafficking operations function and how to stop them remains out of reach for law enforcement.
The state of Missouri confronted this issue strategically by utilizing a federal grant, which provided them with a fresh stream of resources for identifying, interrupting, and combating trafficking crimes. A unique challenge facing Missouri (and other state governments when it comes to informing anti-human trafficking strategies) was where and how to deploy these new resources effectively and equitably.
The Response: The Human Trafficking Risk Index
The MO state government responded to its resource deployment problem by collaborating with IDEA Analytics (supported by HTCBC and with the department of social services for MO) to create a human trafficking risk index that would serve to analyze and showcase trafficking statistics and risk factors at the county level. By isolating and examining smaller geographic sections of the state using the same variables, IDEA could provide Missouri with statistical indicators of where survivors could likely be located. This would help the MO state government decide where its resources would best be allocated.
The human trafficking risk index was created with two goals in mind:
- Help the state government maximize resources and direct them to places that need them the most
- Function as a recreatable tool that other departments can utilize to combat trafficking.
These two objectives come together with the goal to ultimately uncover and provide relief to trafficking victims everywhere.
How the Index Works:
The human trafficking risk index itself considers known indicators of human trafficking to predict the likelihood of victims being located in an area. It then quantifies each county in Missouri with a score between 1 and 7. Counties that classify lower on the scale (closer to 1) most likely don’t have as many human trafficking victims in their jurisdiction as counties that rank higher (closer to 7).
While the index projects where high quantities of victims are likely to be located, counties with low scores may still house survivors as well. The tool is simply meant to guide victim services and law enforcement toward areas that statistically may have a higher likelihood of trafficking victims so resources are utilized more effectively.
Identifying the Variables
One initial and important step was identifying open-source variables that could be combined to help IDEA team members Neil and Jason develop an equation for identifying risk factors for human trafficking in Missouri. Every variable had to be grounded in research and relevant to the conversation before it was included in the project.
Some variables defined by Neil and Jason included:
- UCR data (Uniform crime reporting) from the FBI’s website
- Illicit massage parlors along the interstate
- Data from Polaris
- Polaris is a self-reporting human trafficking victim portal. Survivors, victims, and those who believe they may have witnessed human trafficking can report their information to the organization. Neil and Jason contacted Polaris directly to obtain their trafficking data, which Polaris then provided.
- Truck stops (a known hub for trafficking activity)
- Connection to kidnapping (data pulled from NIBR reporting from FBI)
- Density of foreign-born citizens in a county (data sourced from US census)
- Statistically, foreign-born citizens incur a higher risk of becoming victims of trafficking than individuals born in the US.
Some variables counted more toward overall score than others. For example, UCR data was weighted more heavily than illicit massage parlors along the interstate. IDEA combined the variables to render each county’s score and created the numeric risk index.
Jason and Neil then used the calculated indices per county to develop a choropleth map, which uses color to signify different numbers, to help viewers easily locate areas with high concentrations of trafficking risk indicators. As a county’s score increased, its shade of blue on the map would darken. For example, a county that’s almost white would have earned a score of 0-2, while a county with a score of 6-7 would be dark blue.
The Two Indexes
The resulting index was intricate and used technical data science analytical techniques. It established a research-based and calculated tool for quantifying trafficking risk indicators, but its highly technical data manipulation and language made it difficult for non-data persons to understand.
Because a major goal of the project was to create a tool that many law enforcement agencies throughout the United States could use to understand and reduce human trafficking, IDEA decided to develop a second index. This version was modeled off indices generated by the US Census Bureau that are more digestible, repeatable, and recreateable.
Once the two indexes had been formed, it was time to create a visual representation of human trafficking risk indicators in Missouri.
Mapping the Variables
Using the choropleth map, Neil and Jason created an interactive map that allowed viewers to gain a deeper understanding of where resources could be focused to support victims of trafficking. The tool allows researchers and viewers to zoom in and out and focus on specific counties to examine them in greater detail.
The maps also featured horizontal sliders that allow viewers to easily switch from one map to another. This makes it simple to compare features, such as UCR data and Polaris data, against each other quickly.
After creating the interactive maps, Jason and Neil embedded them into the IDEA Analytics website so that other law enforcement officials and data analysts could recreate the project using different variables that would be relevant to their location.
Human trafficking is a particularly pervasive problem because of data availability. It’s difficult to collect data that will reveal the victim’s location and identity without it being directly reported to official sources. UCR and official police data unfortunately have their own set of limitations because there’s often not a good quality of reporting for a number of reasons.
Although human trafficking isn’t a new phenomenon, government agencies have begun placing more importance on it in recent years. Although this constitutes progress, departments typically don’t have proper training on how to identify human trafficking – neither its indicators nor its victims.
For example, the county of Boone, Missouri, scored very high on the human trafficking risk indicator index. This could be for a myriad of reasons. It could be because Boone truly does have higher human trafficking victim numbers than other counties, but it also may be because Boone may have low to average rates but a police department that’s phenomenal at detecting and reporting trafficking offenses.
This poses an obstacle for Missouri (and states that want to recreate the index) because they may consider sending a significant amount of resources to Boone under the impression that they have high indicators for a trafficking victim population only to discover that the real case is that they have lower numbers and extremely efficient reporting.
Although this challenges researchers, the indexes are founded on concrete data that provides astute predictions. As data on human trafficking becomes increasingly available, the indexes will continue to become more and more accurate and helpful to law enforcement.
Neil and Jason created a 30-page report to showcase their research and findings. It’s currently in the final stages of development. They also presented the interactive storymap, powered by Esri, to the agency, which was enthusiastic about the direction and strategic approach that the tool provided them. Now, the agency has the resources available to help them reach out to victim service providers in counties identified to have high risk indicator scores and inform law enforcement on better training practices.
The indexes are a promising start as an open-source, repeatable solution for decreasing trafficking crimes across the United States. The responsibility falls to other agencies to take this initial step as well so they can determine areas in their jurisdiction that need their attention. This type of data-driven methodology has the power to push back on one of the most elusive crimes in our modern society. For assistance in understanding how data analysis actually works and how to implement it in your organization, schedule a call with the IDEA team here.