7

Key Takeaways and Additional Areas of Focus

David Banks (chair, Planning Committee) began the wrap-up discussion by noting that the workshop participants are a collection of disparate experts who came together with a common purpose in mind. He encouraged participants to consider other venues and scholarly journals where they could continue discussing the themes that emerged at the workshop.

One of the major themes, he noted, was the value of sharing data. He added that reproducibility studies can prove useful when confidentiality constraints prevent the sharing of data. Anonymization can allow data to be shared more broadly, but Banks said the amount of covariate information that is lost in the process can create major data deficiencies. He said that sharing software instead of sharing data—letting the owners of the data run the researcher’s program and then turn over the results to the researcher—may relieve some of the trepidation around releasing sensitive data. Davina Durgana (member, Planning Committee) noted that the U.S. National Science Foundation has taken a similar approach in some of its data analysis. Bernard Silverman (member, Planning Committee) asked the group to consider how taking new and creative approaches to anonymization might improve data sharing.

Sheldon Zhang (member, Planning Committee) reemphasized the importance of having a common understanding of what is at the core of human trafficking measurement, particularly focusing on the varying definitions and questions about thresholds. Banks said he thinks that aligning the human trafficking definition used for research with the legal definition in a given region or country is a start toward achieving consistency. Zhang said although there are legal frameworks already in place—namely, the

Trafficking Victims Protection Act and the guidance from the International Labour Organization (ILO)—it is often difficult to operationalize these definitions for use in different measurement contexts. For example, it is difficult to explicitly and consistently define “coercion” or “fraud” in the context of human trafficking. He noted the tools available to mental health professionals (DSM-5) and law enforcement (Uniform Crime Report) to help them identify men and women who meet certain criteria; he said he would like to see something similar in place to help identify what constitutes labor trafficking and sex trafficking.

Davina Durgana (member, Planning Committee) added that there is also ambiguity in the ways researchers approach trafficking; even experts sometimes have difficulty differentiating between labor and sex exploitation and trafficking. She said her research team at the Walk Free Foundation is working on ad hoc models that look at the most extreme forms of modern slavery and is working backward from there to better understand what actually constitutes trafficking.

Silverman said the U.K’s Home Office produced a typology of modern slavery that provided classification for different possible sectors and forms of exploitation and trafficking.

Sara Crowe (U.S. National Human Trafficking Hotline) said that she has sometimes had trouble classifying cases reported through the hotline. Even with a very extensive data dictionary that also gives guidance on how to evaluate narratives reported to the hotline to see if they meet the legal definition, Crowe said it can still be challenging to apply the guidance to complex situations.

Jordyn White (Committee on National Statistics) asked participants, What data collection would look like if they were able to standardize the information being collected? Where would the data ultimately land? Who would be responsible to process, analyze, and disseminate the data?

Silverman added to the questions: Who would be the right people to construct a road map of a gap analysis that laid out what is still not known about human trafficking? While he is hopeful, he believes it will be a while until researchers develop the overarching mechanisms that will enable that type of research feature.

Banks said two of the key questions that researchers should ask before they begin work on human trafficking are why they are working to obtain an estimate and whether the estimate will drive legislative decision making. He brought up the point made by Annick Febrey (Human Trafficking Institute) about researchers working on other public policy issues not being required to estimate prevalence before they approach policy makers to request funding, but said he can understand how early quantification can help make sure the right resources are in place to do the work. He added

that a toolkit sharing best practices with small agencies and organizations could help bridge the action-funding gap.

Banks also noted how the presentations by both Hanni Stoklosa (HEAL Trafficking) and Michael Shively and Ryan Kling (Abt Associates) noted that hospital emergency departments can be a very critical point of contact and data collection. He reiterated the need to educate health care professionals on how to identify potential victims and effectively gather data in these settings.

Stoklosa said it is important to build up the health sector’s capacity to respond to trafficking.

Meredith Dank (John Jay College of Criminal Justice) added that service-provider training and buy-in is a critical component of estimating human trafficking prevalence.

Katrina Stone (Surveys & Behavior Analytics LLC) commented on the importance of collaborating with other sectors and other federal agencies, such as the U.S. Centers for Disease Control and Prevention and the Bureau of Labor Statistics, to bolster support for counter-trafficking work. She added that experts from other disciplines can also bring a fresh perspective to measurement challenges.

Roy Ahn (member, Planning Committee) emphasized the importance of incorporating data sharing into the research strategy, and he noted that the health care system has developed models that address data-sharing challenges.

Lauren Damme (U.S. Department of Labor) said that hearing from actual survivors throughout the workshop enhanced the discussions and reminded the group of the importance of taking a victim-centered approach when researching such vulnerable populations.

Dank echoed Damme’s reminder and suggested working to include survivors in the data collection and research instead of merely engaging with them in a participant role. She reminded the group about the point made by Manisha Shaw (University of California, Los Angeles) that individuals are more likely to speak freely with their peers than with someone with whom they do not relate.

Banks asked the group to comment on whether looking at trafficking statistics from other countries and then estimating the prevalence in the United States by extrapolating the findings based on the percentage of that nationality in the entire U.S. population could prove helpful.

Zhang noted that while the plan could work, the breadth of industries in the labor market and the gray area around the definition of labor trafficking could make it difficult to quantify the findings. He also discussed the general difference in visibility, understanding, and legislation, between sex trafficking and labor trafficking, and he encouraged participants to be

intentional when considering their research approaches to both forms of trafficking.

Durgana said she disagreed and believes the way in which the United States focuses on the distinction between different types of victims may be detrimental to the research: getting a tailored estimate for each specific labor industry can be very time and resource intensive. She said that in the global estimates generated by the Walk Free Foundation, sex is considered an industry of labor. She asked the group to think about the best cohesive research strategy going forward.

Crowe agreed and said that responses to the National Human Trafficking Hotline have shown her that methods of control and other trafficking conditions are relatively similar across labor sectors.

Erin Klett (Verité) said she has observed variations in the way labor trafficking manifests itself in different sectors and wonders how measuring the sectors together would affect both the richness of the data and the effectiveness of addressing the issue from a policy standpoint. She explained how Verité’s approach starts with a core set of indicators based on ILO governance and then defines them based on context. In that way, although the data collected may look different from sector to sector, they can also be mapped at the indicator level.

In closing, a workshop staff person informed participants that representatives from the Centers for Disease Control and Prevention and from the human trafficking division of the FBI’s Uniform Crime Reporting Program had commented through the webcast that they had been listening to the discussions throughout the workshop and had provided their contact information for follow-up on the topics relevant to their organizations.