Global Demographics of Human Trafficking Detection in Data

A Data Visualization Project by Leila Ouhri, an alum of FCI’s SHINE project.

Background

Florida Community Innovation (FCI) is a civic-technology non-profit based in Florida. FCI empowers young innovators to produce scalable social-services technology for Floridian communities. As the organization expanded in 2022 and 2023, an FCI team began to research the anti-human-trafficking space.

After speaking with Florida anti-trafficking community organizations, task forces, and academics, FCI developed SHINE (Stopping Human Trafficking through Industry Networking and Education).

SHINE builds upon Florida Statute § 509.096, which mandates anti-trafficking training for hospitality employees. SHINE will be an evaluation tool and mark-of-excellence provider for anti-trafficking hospitality trainings.

FCI operates in the state of Florida. Initial research on state-specific human-trafficking revealed the large, seemingly intractable problem of data collection.

The Federal Bureau of Investigation ceased to make state-specific trafficking case numbers after 2015, for example.

In any case, crime or case data alone does not capture the full scale of human trafficking, much (perhaps most) of which slips through detection measures and transgresses county, state, country, and even continental boundaries.

For this data project, Leila used a global scope. This project aims to provide an imperfect but best-efforts picture of human trafficking worldwide.

Data Sources

Human trafficking is an underground enterprise. Crime data in general faces serious problems, e.g., under-reporting.

Data related to human trafficking encounters these same problems and still others—survivors, for example, might not ever self-identify as “victims” or “survivors,”¹ much less approach case- or record-keeping organizations, and survivors’ relationship with their trafficker may be profound, personal, and difficult to characterize in any form of data visualization.

Human-trafficking-related data is characterized by gaps and questions. What is reported? What slips by? What is over-counted, and what is under-counted? What is correctly and incorrectly identified?

Polaris, a heavy-weight in anti-trafficking data collection, tracks both hotline calls and criminal cases. This produces, it is thought, a more-balanced number than police records alone, but this “balancing” may double- or triple-count certain survivors and mischaracterize things that are not trafficking as trafficking.

In light of these challenges, this project develops data from two publicly available sources. The first source below offers a more holistic view, integrating Polaris and hotline-derived data, whereas the second source listed below shows the “hard numbers” of what crimes countries detect and report. Both have serious shortcomings.

These shortcomings, which manifest differently depending on the data category, are outlined in the “additional information” section below each figure. The two sources are dealt with separately; readers can find data visualizations from either source in separate sections. These sources are the:

1. Counter-Trafficking Data Collective (CTDC), which derives its data from the United Nations International Office on Migration, Re-Collectiv+. Polaris, A21, and the Observatory on Trafficking in Human-Beings. 
2. 2022 Global Report on Trafficking in Persons (GLOTIP) from the United Nations Office on Drugs and Crime, which relays each United Nations member-country’s self-reported national crime data on human trafficking.

Counter-Trafficking Data Collective Visualizations

FIGURE 1. Reported Victim Age-of-Majority Status at Time of Exploitation (2002 – 2023)

ADDITIONAL INFORMATION ON FIGURE 1.

The CTDC data includes a crucial category—victim age-of-majority status, that is, whether a survivor is legally an adult or a child (minor). The CTDC issues this status based on age “at [the time of] exploitation.” This seemingly simple category is actually complex.

To what time, exactly, does time of exploitation refer? Trafficking spans many years for some victims. Therefore, conceivably, someone may be first trafficked at age 16 but continue to be trafficked until escaping and marked in casework at age 30.

In the CTDC data, is such a person counted as a minor (based on their time of induction into trafficking) or as an adult (because they operated as a legal adult for some years)? This is unclear. The CTDC does not define boundaries that describe what “age at time of exploitation” means.

However, the data is still useful as it indicates an overwhelming slant toward minors. There are 2,246 victims whose status at the time of exploitation is marked as “adult,” but there are 7,044 minor victims. This opens up several possible interpretations.

For instance, in the previous example of a 30-year-old who was first trafficked at age 16, the more likely outcome would be counting her as an “adult,” not a child, a categorization which would obscure her minor status at the start.

That the data skews so heavily toward minors suggests that minors may indeed be overrepresented among trafficking victims. 

Of course, data issues arise here, too, because it is possible that cases involving adults did not always include age, whereas cases involving minors may have understandably flagged or described age more often in reports since such cases are especially sensitive.

Moreover, the CTDC category “age-of-majority status at time of exploitation” is separate from CTDC age-bracket data, which is explored below in FIGURE 2. CTDC age-of-majority status data uses 9,290 cases, but the CTDC age-bracket data has much more information, covering 36,439 total cases.

It can therefore be safely assumed that not all of the cases where age information was available were coded into age-of-majority status categories. (“Present” in the graph title refers to the CTDC’s data collection up until August 2023, when this visualization was developed.)

FIGURE 2. Age of Detected Trafficking Victims (2002 – 2023)

ADDITIONAL INFORMATION ON FIGURE 2.

Although both deal with age, this graphic must be differentiated from FIGURE 1: Whereas FIGURE 1 uses an ill-defined, nebulous term, “time of exploitation,” FIGURE 2 is explicitly concerned with victims’ ages at the time of detection.

“Detection,” here, may have many meanings. It may, for example, indicate the time at which a victim is recovered by police or the time at which a hotel employee calls a hotline about a suspected sighting of a victim. Readers should keep the flexibility of the term “detection” in mind when considering FIGURE 2.

Still, FIGURE 2 appears to show a compelling pattern—adolescents aged 9-17 years old are the most prominent group, followed closely by 30-38 year olds.

Closer examination reveals that this pattern is not necessarily what it seems, however.

Consider the age brackets reported by the CTDC. 9-17 year olds and 30-38 year olds are each grouped in a category that spans eight years. Yet 18-20 year olds are grouped together, a category that spans just two years.

The CTDC does not report the individual ages associated with reports or cases because of the sensitive, theoretically identifiable nature of this data. The data can only be accessed in the form of the counts for each age bracket. 

Let us now re-consider the numbers as they might appear in eight-year (or close-to-eight-year) brackets.

Ages 9-17 years of age is an eight-year bracket with 8,645 marked cases; ages 18-20, 21-23, and 24-26 combined is an eight-year bracket with 12,125 marked cases; ages 27-29 and 30-38 combined is an eleven-year bracket with 9,732 marked cases; and 39-47 is an eight-year bracket with 2,821 marked cases.

There appear to be noticeable fewer cases in which victims at the time-of-detection are over age 39. Older survivors are detected less often. Overwhelmingly, the survivors who are detected appear young. 

FIGURE 3. Proportion of Trafficking Victims By Age Range (2002 – Present)

ADDITIONAL INFORMATION ON FIGURE 3.

FIGURE 3 shares the limitations of FIGURE 2: uneven age bracketing. Uneven age bracketing refers to the fact that one category, survivors aged 0-8 years old, contains eight ages, whereas another category, survivors aged 18-20, covers just two years. However, FIGURE 3 is still instructive.

Trafficking victims older than 17 and younger than 30 account for 40.53% of detected human-trafficking victims, according to the CTDC. 23.72% of detected human-trafficking victims are between the ages of 9 and 17 years old. An additional 4.27% are 8 years old or younger. 27.99%, then, of detected human-trafficking victims are younger than 18 years old. 

FIGURE 4. Detected Victims by Gender (Years 2002 – 2023)

ADDITIONAL INFORMATION ON FIGURE 4.

Sex trafficking dominates the public imagination in human-trafficking-related discussions. Sex trafficking is assumed to—and in fact does— primarily affect women.

Males, however, can also be trafficked. Further, sex trafficking is not the only kind of trafficking: Forced labor and debt bondage are other mainline forms of trafficking in which males, as well as females, are exploited. FIGURE 4 shows the number of detected human trafficking victims by sex.

Women account for 35,534 of detected victims and men for 13,267, with detections of women outnumbering detections of men by almost a 3:1 ratio. Men, then, are 1 of every 4 detected trafficking victims whose sex the CTDC recorded. (Here, as in FIGURE 1,  the titular word “present” in “2002 – Present” refers to CTDC data as it existed until August 2023.)

FIGURE 5. Relationship Between Trafficking Recruiters and Victims

ADDITIONAL INFORMATION ON FIGURE 5.

FIGURE 5 highlights the “black box” nature of trafficking data. “Unknown” relations account for 79.54% of the data available (39,207 out of the total 49,287). Importantly, “unknown” does not exclusively refer to situations in which the trafficked individual was a stranger to or did not know their trafficking recruiter.

The CTDC’s “unknown” relation category also catches gaps in the data. Some cases marked as “unknown” in the CTDC sets refer to cases in which it is simply not known at all what the trafficker’s relationship was to the victim; for example, the relationship may not have been explicitly stated during a survivor’s interview or during their casework.

For at least some victims, though, the trafficker was likely indeed a total stranger to them. Of the 10,080 victims whose relationship with a recruiter was “known” to the CTDC, 16.12% (1,625) of victims were recruited by an intimate partner; 20.79% (2,096) by a friend; and 18.64% (1,879) by a member of their family. 

FIGURE 6. Major Countries of Exploitation

ADDITIONAL INFORMATION ON FIGURE 6.

FIGURE 6 emphasizes countries of exploitation whose numbers made up 2% or more of the total. Countries of exploitation whose total counts were less than 2% are grouped in the yellow “other” category. “Countries of exploitation” refers to countries in which trafficked individuals operated; it does not necessarily refer to the countries from which victims came or in which victims were born.

Notably, based on the CTDC data, the United States is a leading country of exploitation, followed by Ukraine, Moldova, Russia, and the Philippines. FIGURE 6 does not necessarily mean that these are in fact the countries in which the most exploitation occurs.

Countries with greater numbers might instead have more mechanisms to catch trafficking or better data reporting mechanisms. Thus, countries who do “better” at stopping human trafficking may appear “worse” than countries with more human trafficking but less reporting.

Of course, some of these countries may indeed have especially significant trafficking operations. Complete data does not exist. Still, Moldova and Ukraine’s large presence in FIGURE 6 matches trends identified by INTERPOL and other agencies, who flag the countries as high-activity “source” countries.²

FIGURE 7. Victim Citizenship Proportions of Total

ADDITIONAL INFORMATION ON FIGURE 7.  

FIGURE 7 addresses the percentage of victims with citizenship in each nation. Victim citizenship is a good stand-in for a victim’s origin, although the country where a victim has citizenship may not necessarily be the country in which they are first trafficked.

Still, victim citizenship information is valuable because it may point to which demographics (i.e., which nationalities) are most vulnerable. This information may also facilitate better survivor recovery efforts, as victim citizenship may reveal language needs.

FIGURE 7 suggests that a plurality of trafficking victims are citizens of the Philippines (23.42%), with citizenship from the Ukraine (15.99%) and Moldova (12.16%) also prominent. Evidently, two regions account for a large part of the total: Southeast Asia and Eastern Europe. Victims with citizenship in the Eastern European countries of Belarus (3.02%), Moldova (12.16%), and the Ukraine (15.99%) are 31.17% of the total. Southeast Asian countries similarly compose 34.14% of the total, combining Indonesia (4.06%), Cambodia (4.08%), Myanmar (2.58%), and the Philippines (23.42%).

If less than 2% of total victims had a certain citizenship, that citizenship was aggregated into the “Other” category, which was only 8.94% of total cases, although cases where victim citizenship was “unknown,” a separate category altogether, accounted for 18.26%. 

FIGURE 8. Count of Victims in Each Category of Labor Exploitation (2002 – 2023)

ADDITIONAL INFORMATION ON FIGURE 8.  

FIGURE 8 covers different forms of labor in which survivors were engaged. Domestic work—which includes work like cleaning, cooking, childcare, elderly care, and so on—is far-and-away the most salient labor category.

We may hypothesize that the lopsidedness of the categories of labor exploitation (that is, the dominance of domestic work) may explain the lopsidedness of the overall trafficked male-female count. Women may be used more so for domestic work. The next most prominent specified categories are construction, manufacturing, agriculture, and begging. 

FIGURE 9. Victims Per Type of Exploitation

ADDITIONAL INFORMATION ON FIGURE 9.  

FIGURE 9 differs from FIGURE 8 because FIGURE 8 targets labor trafficking categories. FIGURE 9, on the other hand, takes a wider scope: the number of victims per category of trafficking as a whole. According to the CTDC data, then, there are more sex-trafficking victims (16,067 individuals recorded) than labor-trafficking victims (9,772).

FIGURE 9 suggests that sexual exploitation significantly outpaces forced labor exploitation, but labor exploitation is known to be “less frequently detected and reported than trafficking for sexual exploitation.”³ Sex exploitation is more conspicuous due to heightened media coverage and its tendentious nature, whereas signs and cases of labor exploitation can seem banal or easy-to-miss. The term “other exploitation” is mysterious here, not defined by the CTDC.

In practice, other forms of human trafficking may involve forced begging or peddling on the streets (which may or may not be grouped in with labor trafficking), forced involvement in crime, domestic servitude, or organ removal.⁴ Just 168 forced-marriage victims were recorded: This number, of course, is very low and not representative of the full scope of the category.

FIGURE 10. Means of Control Used On Victims

ADDITIONAL INFORMATION ON FIGURE 10.

FIGURE 10 counts the means of control used upon victims. It is important to note that each case—that is, each victim—could be coded for multiple means of control in the CTDC dataset. Most victims experienced multiple means of control (e.g., their trafficker used both debt bondage and psychoactive substances upon them at the same time), not just one.

The most common means of control, per FIGURE 10, was psychological abuse. The next-most frequent control mechanisms were restriction of movement, threats, physical abuse, taking earnings, false promises, and excessive working hours.

Obtaining this data in the first place means getting qualitative information from survivors themselves, who may not conceive of themselves as “victims” or their traffickers, with whom they may share tight emotional, cultural, or social ties, as “criminals.”

The categories themselves—e.g., “false promises” and “threats”—are broad, nebulous, and not precisely defined. They may therefore obscure nuances about the trafficking experience, alternately wrongly including and excluding cases according to the categorizer’s subjective opinion about whether a case fits into such-and-such category. 

GLOTIP Visualizations

FIGURE 11. Global Cumulative Count of Persons Convicted of Trafficking Offenses By Year

ADDITIONAL INFORMATION ON FIGURE 11.  

FIGURE 11 represents the citizenship of detected trafficking victims. It is therefore the GLOTIP equivalent of FIGURE 7, which shows victim citizenship according to CTDC data. Despite the fact that they purport to measure the same thing, the figures draw drastically different conclusions.

FIGURE 7 suggests that only around 7% of detected victims have U.S. citizenship, whereas FIGURE 11 puts that number at 33.6%. FIGURE 11 also emphasizes the number of detected victims with Mexican citizenship (25%), as well as those with citizenship from other countries in the Americas like Honduras and Guatemala.

On the other hand, the regions that dominated the CTDC chart (i.e., Eastern Europe and Southeast Asia) are relatively minor in this GLOTIP-created figure. This mismatch reflects the difficulty of finding reliable answers in the anti-trafficking world.

FIGURE 12. Total Reported Offenses of Trafficking in Persons By Region During 2010 – 2021

ADDITIONAL INFORMATION ON FIGURE 12.  

FIGURE 12 suggests that Asia has the largest number of trafficking-in-persons offenses (roughly 80,000), with Europe (roughly 60,000), the Americas (roughly 30,000), and Africa (roughly 15,000) trailing behind. However, these numbers are totals, not adjusted for the population size of the regions.

Asia, the world’s largest continent with around 4.7 billion people, is bound to have the most of any measurement because of its sheer population. Further, because human trafficking is transnational, survivors or traffickers may begin in one place and migrate to a totally different part of the world to avoid being recognized or to avoid staying in one place long enough to arouse suspicion.

Finally, this data is the sum for each continent, based on the individual cases reported by constituent countries. For an offense of human trafficking to be reported and included in the GLOTIP report, therefore, a country would have to have a law that such an offense violates, an established legal system capable of recognizing and recording that violation, and the ability to transmit that information reliably to the United Nations Office on Drugs and Crime.

Instability and conflict in countries like Sudan, whose civil war has driven it to the brink of state collapse, makes the collection and transmission of that kind of information to global intergovernmental organizations challenging. 

FIGURE 13. Recorded Offenses of Trafficking in Persons by Subregion (2015-2020)

ADDITIONAL INFORMATION ON FIGURE 13.  

This graph reveals general trends—and of course suspicious outlier data points from the GLOTIP data over a five-year span between 2015 and 2020. North America, designated by a dark purple line, is generally the region in which the most trafficking-in-persons offenses are reported. Eastern Europe, a pale purple line, is close but slightly lower than the North American line.

The almost completely flat green lines of Melanesia and Micronesia can be attributed to the regions’ small populations. North Africa, indicated by a light blue color, begins low and rises.

Again, such rises may not indicate an increase in human trafficking in that region in itself but an increase in the state’s capacity to identify and record these reported offenses. The bright red line, South Asia, dominates the figure with an enormous, seemingly inexplicable surge in 2017, followed by a drastic fall. 

FIGURE 14. Global Count of Trafficking in Persons from 2010 – 2021

ADDITIONAL INFORMATION ON FIGURE 14.  

FIGURE 14 demonstrates that same spike of reported offense in 2017. What happened in 2017 to human-trafficking statistics? It is unclear. Based on its surge and dramatic fall, the 2017 situation appears to be an outlier. FIGURE 14 suggests that this effect in the GLOTIP data was a result of something happening in South Asia.

FIGURE 15. Offenses of Trafficking in Persons by Country (2010 – 2021)

ADDITIONAL INFORMATION ON FIGURE 15.  

FIGURE 15 reveals the cumulative totals of each country from 2010 to 2021. Here, again, it is worth stressing that population size here may distort the findings. India and China have, far and away, the highest number of reported cases, no doubt contributing to Asia’s dominance in FIGURE 11.

However, this does not necessarily indicate a uniquely concentrated human trafficking system: China and India’s large populations make them likely to dominate many simple measures. Following China and India are Pakistan, Argentina, Romania, Italy, Nigeria, Russia, Belgium, and Uzbekistan. 

Conclusion

Human trafficking is still a “black box” for researchers. Data collection is enormously difficult for any crime. Data collection for global human trafficking, with its propensity to cross national boundaries and hide survivors, is all the more so.

Even the finest data collectors and agencies, like the UN, with the best chances at access to sensitive documents face serious limitations.

Collecting data primarily from survivors who are no longer experiencing trafficking means that there is not a random sample for trafficking victims; trafficked individuals who are still experiencing trafficking—and their demographic details, like age—are left out. Methods like respondent-driven sampling and the network scale-up approach, which uses respondents’ social relations and surroundings to dig deeper into the respondent’s life, can help shine light on difficult-to-reach populations. 

The UN’s GLOTIP report and the CTDC’s dataset represent enormous strides in the field of human trafficking research, yet at times, as this data visualization paper showed, the results of the data they produce look quite different.

For example, the citizenship of victims differed dramatically, with one set emphasizing Central American nationalities and the other Eastern European and Southeast Asian nationalities. Taking either of those results as a premise for policy or research would have drastically different outcomes. All of these limitations should be kept in mind when viewing the earlier data visualizations.

The visualizations are merely a reflection of the data that has been collected. FIGURE 9, for instance, suggests that sex trafficking is more salient or widespread than labor trafficking, when experts say that exactly the opposite is true: Labor trafficking, while less-often identified as “human trafficking,” is understood to be much more common. Still, the visualizations in this report give an overall impression of trafficking on the world stage, answering, with the limited data available, several crucial questions.

Which nationalities are human trafficking victims likely to be? Where does the most human trafficking get detected? What forms of labor trafficking are the most widespread? Which places are the primary countries of exploitation? Most importantly, though, the visualizations capture one of the anti-trafficking field’s key obstacles: officially tracking and recording reliable information about underground, inaccessible, and vulnerable populations. 

1. Rebecca Surtees, “Trafficking Victim Identification: A Practitioner Guide,” NEXUS Institute, 2021. https://www.iom.int/sites/g/files/tmzbdl486/files/iscm/materials/traffickingvictimidentification.pdf ↩︎
2. INTERPOL. (2022, October). “Combating organized crime, terrorism and firearms trafficking in Moldova,” Operational Assistance in Moldova Initiative. https://www.interpol.int/en/Crimes/Organized-crime/Operational-Assistance-in-Moldova-Initiative ↩︎
3. “UNODC report on human trafficking exposes modern form of slavery,” United Nations Office on Drugs and Crime, accessed 12 November 2023, https://www.unodc.org/unodc/en/human-trafficking/global-report-on-trafficking-in-persons.html ↩︎
4. The International Criminal Police Organization, “Trafficking of Human Beings for the purpose of Organ Removal in North and West Africa.” Published July 2021. https://www.interpol.int/content/download/16690/file/2021%2009%2027%20THBOR%20ENGLISH%20Public%20Version%20FINAL.pdf ↩︎

About the Author

About the Editor