An official website of the United States government

Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

You are viewing ARCHIVED CONTENT released online from January 20, 2017 to January 20, 2021.

Content in this archive site is NOT UPDATED, and links may not function.

For current information, go to

The Role of Research in Human Trafficking Intervention:  Observations from the Federal Trafficking in Persons Office

There is a huge challenge before us. This challenge runs through all of our efforts to combat human trafficking. We do not know how to measure the impact of our work – and because we do not know how to measure impact, we cannot tell which of our well-intentioned interventions are working and which ones are missing the mark.

This is why I am so excited to join you here for the Human Trafficking Research and Data Advisory Roundtable. I am grateful to Point Loma University for hosting this event and for all of you who are sacrificing your Saturday to be here this morning. This gathering is important because I believe your emphasis on research and data is essential if we are going to address our challenge of figuring out how to measure impact. We must know how we measure success.

I have had the privilege of working in this movement in a variety of roles. Each has allowed me to view our common work from different perspectives. I have twice worked for NGOs and twice worked for the government. I have worked domestically and internationally, on casework and policy, and with survivors and traffickers.

Now, I serve as the United States Ambassador-at-Large to Monitor and Combat Trafficking in Persons. This position confuses some of my friends. When most people think of “ambassadors,” they think of “Ambassadors in Residence” – ambassadors to a specific place on a wide variety of issues. “Ambassadors-at-Large” are ambassadors not to a specific place, but to all places on one specific issue. I am encouraged that Congress elevated the issue of human trafficking to this status.

Our office has three streams of work. The first is to lead the United States’ global engagement to combat human trafficking. We do this through accurate reporting, targeted foreign assistance, and engagement with governments through both bilateral diplomacy and multilateral institutions. Our second mandate is to support the coordination of anti-trafficking efforts across the U.S. Government as we work to combat trafficking domestically and internationally. Finally, we engage the public, business, academia, faith communities, and the anti-trafficking field about current trends, promising practices, new tools, and innovative ideas. We work to convene, connect, and collaborate around key policy priorities that drive our diplomatic efforts in this fight for freedom.

My work has allowed me to meet thousands of victims, and to better understand their stories of trauma, abuse, and fear and hear their recommendations for improving our work. I have also spent a lot of time with traffickers. In long interviews as they agreed to testify against other traffickers, they explained in detail how they recruited, groomed, coerced, and monetized people.

During these conversations, it became clear to me that traffickers do not share our challenge. Traffickers are not confused about how they measure impact. They have great clarity about how they define success. Traffickers seek to maximize their illegal profits by minimizing their victims’ freedom. They trade others’ freedom for their personal gain. And their criminal schemes are working. Their annual profits exceed $150 billion. This means that they are making more than Apple, Microsoft, Samsung, Exxon, and BP combined. Human trafficking is big business.

It is in this context that governments and people of good will must have clarity about our goal – freedom for all. We must also learn to assess the utility of various strategies and programs so we can determine what should be scaled and where we need to make adjustments.

2020 is an auspicious time to pause and reflect upon our challenge. This year marks the 20th anniversary of the Trafficking Victims Protection Act (TVPA) and the United Nations’ Protocol to Prevent, Suppress and Punish Trafficking in Persons (known as the Palermo Protocol). Since 2000, the TVPA has served as the cornerstone of comprehensive federal anti-trafficking efforts, and it has exemplified bi-partisan and legislative-executive cooperation to combat trafficking in persons.

Anniversaries provide a good time to pause and reflect about what we have experienced and then pivot to consider what is ahead. A few years ago, my wife and I celebrated our 25th anniversary. We trekked to Machu Picchu along the famed Inca trail. It was the trip of a lifetime.

We thought about our journey of gaining degrees and credentials, building careers, starting a family, and parenting kids. We remembered the good times and the many challenging times. Then we thought about what would make the next season successful. We realized that things are about to change for us. We are shifting from raising kids to launching adults, from building careers to training others. The next couple of decades will call upon us to love our parents well as they age.

The TVPA’s 20th anniversary gives us a similar opportunity. As we look back at our anti-trafficking efforts, we can draw out several themes.  Many of you have reflected on these themes and are deep in a variety of important research topics. The theme I would like to focus on this morning is “measurement.”

Much of this movement has been fueled by emotion and anecdote as we have tried to get a better handle on the victim experience and how traffickers operate. To be clear, I find the stories and case studies powerfully moving.  However, we need to continue to add metrics, scholarship, and data to the emotion and anecdotes.

Most of the measurement the field has done centers around measuring activities. We measure how many hours of training occurred, how many beds are in a shelter, how many people were exposed to an awareness campaign, or the number of calls a hotline received.

This is not a criticism of the early years of this work. Instead, it is upon those early advocates’ shoulders that we stand today. In this “crawl, walk, run” progression, we had to first convince people the issue mattered, try quick interventions to prove that we could protect survivors, and now we need to standardize and professionalize our work.

Moving beyond measuring what we have done, to measuring the effect of our work, will fuel the future of this movement. The bottom line is that what we measure matters. We need to take a serious look at how to improve if we are going to defeat the traffickers and ensure freedom for all.

A few years ago, my wife and I packed the car in Washington, DC with our three amazing kids and set off for an epic cross-country drive. Our destination was Yellowstone National Park in Wyoming, just over 2,000 miles away. Waiting for us there was a cavalry of extended family, festive matching reunion t-shirts, and memories to be made.

Not long into the drive, the predictable question came from the backseat, “How much farther?” It is hard to give a satisfying answer. How much farther to the ultimate destination or to the campsite we would enjoy that night? The kids didn’t seem interested in my driving performance: the turns, merges, and effective steering. There were lots of decisions and attractions along the way. They wanted a measurement of progress toward the ultimate goal, not my current activities.

As we combat the crime of human trafficking, “How much farther?” is a difficult question. When driving the car, I could measure our progress in miles traveled. I had the benefit of a GPS and an odometer.

The challenge before us of measuring impact is that we lack a consistent ruler to measure our progress and we do not enjoy maps and road signs. Our journey is more like that of an explorer charting new territory than a family following a well-developed route.  We start our journey with a huge goal but uncertain about the most effective path. Along the way, we must develop the tools we need, including the tools to measure impact.

If we do not tackle this challenge, our well-intentioned efforts to combat human trafficking may be of limited utility. We have to ask if the programs and policies we have worked so diligently to implement have had a direct impact on stopping traffickers or supporting victims.  If we do not solve this challenge, we will be left merely talking about what we have done – which some may call, “busyness masquerading as progress.”

Measurement has long been a top priority for the United States. At the White House Summit on Human Trafficking on January 31, the President signed an Executive Order that commits federal agencies to improve the methodologies used to measure prevalence, utilize them, and make them public. This is an exciting development and part of putting “freedom first.”

The key quantity of interest is the “the prevalence” of human trafficking. We want to know the scope of the problem. If we know the size of the problem we can measure to see if we are making progress toward our ultimate goal.

Yet, we know it is not easy to measure prevalence. Traffickers and victims are hidden. Victims rarely self-identify due to shame, fear of authorities, or ongoing effects of their trafficker’s coercive scheme.  (I, myself, have provided those explanatory talking points over the years.) Yet, whether the number of victims or traffickers have decreased, is one of the truest measures of a program or policy success. It would be the equivalent of me sharing how many miles are left to arrive at Yellowstone.

To date, there have been some efforts to measure prevalence. They center around measuring global or national prevalence. While they have sparked interest in combating trafficking, they have limitations when measuring impact.

On the global level, one of the most widely cited prevalence measures is from the International Labor Organization (ILO).  Knowing that no one method could accurately identify all victims of trafficking, they fielded 54 nationally representative surveys and conducted a systematic review of databases on human trafficking maintained by the International Organization for Migration and other validated sources. This enabled them to estimate that, at any given time in 2016, traffickers exploited 24.9 million people.

That is more victims than ever before in human history. Global estimates are powerful for raising awareness and making the case that the problem is enormous.

There has also been interest in national prevalence estimates. Some national estimates have been done based on existing administrative data on victims of trafficking, which requires a country to have a high level of data infrastructure already in place. They need to have at least two lists of victims of trafficking, hopefully more, and at least 75 identified victims of trafficking. It is a tall order for countries to maintain this high level of data infrastructure, including those that are well-developed.

The challenge for both global and national estimates is that they rely on a single survey instrument to identify a wide variety of different types of trafficking. For instance, we can expect that the survey instrument and sampling methodology used to identify victims of sex trafficking in an urban environment will be unlikely to successfully capture victims of labor trafficking in a factory. Likewise, the survey instrument designed to measure traffickers who compel adults to work in agricultural fields will not be effective to measure traffickers who compel boys to engage in commercial sex. These traffickers recruit, groom, and coerce their victims in very different ways and the victims inhabit different, often tightly woven social networks. These illegal market sectors do not necessarily overlap.

This makes sense. We would never expect one cancer screening test to identify all forms of cancer. The test to measure the presence of blood borne lymphoma is not best to identify skin cancer, liver cancer, or cancer in the appendix. Just because all these are “cancers” does not mean that researchers should employ a single screening instrument.

Imagine a call for a national estimate of economic crime. The survey instrument designed to identify con-artists fleecing the elderly would need to be very different from one to identify Wall Street securities fraud.

Traffickers also operate in significantly distinct ways.  Any single survey instrument seeking to capture the vast variety of sex trafficking or forced labor operations would be deeply flawed. Global and national prevalence estimates may benefit general awareness, but they are insufficient to meet our challenge of measuring impact.

That is why we believe that the future of prevalence lies in focused estimates. They must be both industry-specific and geographically-restricted.  Through sub-national “focused prevalence estimates” we could develop an estimate of adults traffickers compel to work in restaurants in a specific metropolitan area, or an estimate of the nurses coerced to work in eldercare facilities in a specific region. This would be a much more useful figure from which to construct a program or policy to address the crime.

These focused estimates could employ a variety of methodologies: multiple-systems estimation, respondent driven surveys, network scale up models, or something else. What works in one region may not be as effective in others. We can tailor the methodology to the specific type of trafficking we are targeting.

A focused estimate yields many benefits. It allows us to operate programs, conduct awareness campaigns, train service providers, expand services, and improve the criminal justice delivery system around a type of trafficking, and then measure again using the same methodology. We could know if there was a reduction in prevalence and if we introduce a treatment and control, we can tie these results or show a relationship to our program.

There will still be critics. They will argue displacement or disagree with the cocktail of possible causes for the reduction. (Honestly, if we advance the conversation to a “causation debate” we will have accomplished something significant. I hope that at the 30th anniversary of the TVPA, folks are gathered here at this conference trying to figure out what interventions should be credited with a massive drop in the prevalence of human trafficking. That would be a huge win!)

To this end, I have some good news. A few years ago, Congress created the Program to End Modern Slavery (PEMS) to drive transformative efforts to achieve measurable and a substantial reduction of the prevalence of modern slavery. For the first time in its funding history, the State Department’s Trafficking in Persons Office has a program that makes prevalence reduction a central component.

A few weeks ago, we announced the first ever Prevalence Reduction Innovation Forum. This puts human and financial resources towards ideas that we have long thought important. The goal of the Forum is to document the effectiveness of various methodological approaches in prevalence estimation and build a global community of researcher-learners in the field. The State Department is making a $5 million investment in this research, and I am hopeful that number will significantly increase as other federal partners join the Forum.

Let me highlight six distinctives about the Forum.

1. Focused Prevalence Studies

The first distinctive element is that all the studies will be focused. They will be industry-specific and geographically-restricted so that measuring impact will be possible.

2. Coordination

The second is coordination. With the leadership of the African Programming and Research Initiative to End Modern Slavery at the University of Georgia, the Forum will fund at least eight research teams to implement a minimum of two different prevalence estimates on the same target population. The studies will be harmonized through the implementation of a meta-analysis – providing us consistent and comparable research results. As far as we know, research of this scope and style, with this level of coordination, has never been completed before in the human trafficking field.

3. Common Definition

All the studies will align with the United Nations definition of trafficking in persons in the Palermo Protocol. This Protocol is one of the most widely accepted international legal instruments and generally aligns with the United States’ TVPA. It follows the three elements of definition of act, means, and purpose.

The Forum will develop and test a statistical definition of human trafficking based on the Protocol. This will be key. The statistical definition will provide guidance that ensure all research teams are measuring the same crime.

This definition avoids alternate constructions that are inconsistent with international law. Some of these insert additional requirements like “the movement of victims” or they delete essential elements like the “act” or “means,” allowing an overly expansive measurement of generalized exploitation.

By remaining faithful to Palermo’s construction, the Forum represents the first time that we will be able to field more than 16 prevalence studies using different methodologies but the exact same statistical definition.

4. Beyond Vulnerability

All the studies will focus on measuring the incidents of human trafficking offenses and not on merely measuring vulnerabilities to trafficking. We all know that factors like poverty, immigration status, educational level, discrimination, prior trauma, addiction, family and community bonds affect one’s vulnerability to criminals.

Yet, we also know that the vulnerabilities themselves are not the root cause of human trafficking. The root cause of human trafficking is traffickers. Trafficking is always caused by the women and men who elect to exploit people in forced labor or sex trafficking. Traffickers tell us that they often target vulnerable people because they are easier to control. By measuring the crime and not just vulnerabilities, the Forum brings a new level of rigor to the field.

5. Meta-Analysis

The Forum will allow us to compare methodologies within and across specific geographies and sectors. Questions remain on the best type of method to use in any given circumstance. Even though we may never be able to get to a perfect number – that is every victim identified – there are some steps we can take to get us to tighter confidence intervals and lower margins of error.

The Forum will also facilitate a meta-analysis. It is the first time that we are going to get to compare, using a meta-analysis, estimates of human trafficking prevalence across and within multiple international research studies to see what really seems to work globally. Programs to combat human trafficking are generally small, so it is imperative that the prevalence estimates generated are as precise as possible. Only through that precision will we be able to detect a decrease.

An effective comparison of studies can answer a number of questions:

  • About whether face-to-face household surveys result in underreporting because respondents might be unlikely to report sensitive issues to an unknown researcher.
  • About overreporting when respondent-driven sampling is used because some research suggests that referrals are not random and are based on possible bias.
  • About whether reliance on administrative data results in underreporting because programs often leave out segments of the community.
  • About when tech-based methods like Facebook surveys, which are usually cheaper and faster, return estimates that are close to those from face-to-face survey administration.

We hope the Forum’s meta-analysis will resonate beyond the human trafficking field to anyone who uses these methods.

6. Publication

A key component of the Forum is that we are going to require researchers to publish their results in academic journals. Through a unified publishing strategy, we will ensure that all results, including null, will be made public. Furthermore, we hope academic papers will provoke interest and debate within the academic community and encourage researchers from other related fields to consider working on human trafficking. Publication will further transparency and will ensure that the studies generated by the Forum will serve as a toolbox for program design and a starting place for future research.

Our efforts are ambitious, but at the same time, we know we have to be realistic. Here I want to acknowledge what the Forum is not going to do.  The studies will not produce exact numbers of cases. The Forum’s findings should not be generalized into a global or national estimate.  The results should also not be used to suggest the prevalence across regions and industries not studied.

While we are very excited about the Forum and its contribution to our understanding of how prevalence can be used to measure impact, we also know it does not alone solve our challenge. There are many important and essential interventions to combat human trafficking that may not have a significant impact on prevalence. For instance, providing effective trauma-informed care to survivors who are no longer being trafficked, may have little effect on the total number of current trafficking victims if the trafficker has moved on to new victims. We need additional non-prevalence focused innovations to effectively measure the impact of counseling, job skilling programs, shelters, and other protection services. Although the “prevalence ruler” isn’t the only needed metric, developing it will be a significant contribution to the field.

The bottom line is that the U.S. government and the anti-trafficking movement have committed themselves to moving beyond merely measuring activities to measuring actual impact. A focused estimate and good data will strengthen our argument to right-size the world’s approach to this crime. We could point to impact and what is working. We could argue based on metrics to scale up effective interventions. We could also dismiss, or rethink, well-intentioned interventions that do not seem to be effective.

This is the time to engage. We are not powerless to do something about human trafficking. It is not a naturally occurring phenomenon like earthquakes, tsunamis, or weather patterns. We cannot stop those. Our best efforts are found in mitigating their consequences. But we do not have to settle for just mitigating the consequences of human trafficking, we can stop it at its source by stopping the traffickers who fuel this crime!

But the question will come: “How much farther?” Funders and policy makers will want to know if we are making a difference. They want to know if we are on the right track. Very reasonable questions – and your research efforts can help us provide the answers.

Research can also generate hope. I think people of good will want to draw near to people in pain if they think there is something they can do about it.  But if they believe their engagement doesn’t really make a difference they pull back. Letting them know “how much farther” and showing progress toward our big goal of freedom generates the tangible hope that inspires engagement. This is hard work, and there are no easy fixes. It is time for us to prioritize freedom. The world does not need people to explain why things are hard; instead, the world is desperate for people to do hard things. Securing freedom for all is a hard thing, but it is worth doing.

Please stay engaged with our Office. Follow us to get updates and hear more about our efforts to put “freedom first” during this 20th anniversary year.

It has been said that, “We value what we measure and we measure what we value.” If we value freedom, we must measure our progress. Through research and action, we can meet this huge challenge before us: we can measure for impact.  I am glad we are in this together.

Thank you.

U.S. Department of State

The Lessons of 1989: Freedom and Our Future