Gewei Cao 曹格玮 - How is human trafficking data accessible for the public and researchers?

How is human trafficking data accessible to the public and researchers?

---- the "data dilemma" problem

Author: Gewei Cao

2026

Keywords: human trafficking, open data, open science, modern slavery

1. Introduction

Human trafficking is a modern form of the slave trade. For researchers who study this issue using quantitative methods, access to reliable data is essential. In practice, however, countries differ widely in how they collect, report, and release data on human trafficking. These differences make it difficult to assemble data and to compare patterns across countries. For this reason, national-level empirical studies face serious limitations, a concern raised by Weitzer (2015). There was one wave of economic studies about human trafficking and sex-work legalization in the 2010s - read Rusch et al. (2026) for more information - especially around 2012, but basically all of them relied on country-year-level data.

Therefore, for quantitative data-driven research, leveraging micro-level (subnational) data is crucial (Weitzer, 2015). In recent years, more and more studies on human trafficking or modern slavery use sub-national level data, just like my paper: Modern Slavery and Mistrust (Cao & Rusch, 2025). However, empirical researchers still face challenges with data quality and availability.

A central challenge in data-driven research on human trafficking is what I refer to as the "data dilemma". Countries with high-quality and transparent data typically have strong institutions and effective law enforcement, and therefore face relatively lower levels of trafficking. By contrast, countries with severe trafficking problems often lack the institutional capacity to detect, record, and report these crimes. This creates a mismatch between data availability and research needs: the countries that most require scholarly attention are precisely those for which reliable data are scarce, limiting researchers’ ability to study and publish on these cases.

Moreover, much of the available data is aggregated at the national level and offers limited variation for empirical analysis. Even in countries with relatively good data access, such as Germany, the small number of states (N = 16) constrains statistical power. For these reasons, sharing practical experience in working with human trafficking data is both necessary and valuable. This blog shares my experience collecting micro-level human trafficking data during one of my PhD projects, Modern Slavery and Mistrust (Cao & Rusch, 2025), and offers practical guidance for researchers facing similar challenges.

In the following sections, I will first share the data sources that I find with the highest quality and accessibility. Then I will discuss the "data dilemma" problem, which we must face seriously for development studies.

2. Where to find detailed human trafficking data

2.1 Global

Currently, to my knowledge, the Counter Trafficking Data Collaborative provides the most detailed and comprehensive victim-level data. The data are anonymized through a specific algorithm to protect the privacy of victims. The most unique information they are providing is the means of controls on victims, and their victim-perpetrator data.

However, this dataset has a strong focus on the USA, where around 58% of the synthetic records were detected. Furthermore, in my personal experience, the geographical variation in this dataset (national level) does not deliver enough statistical power and cannot support sophisticated econometrics analysis. Nevertheless, the citizenship and the country of exploitation of victims might provide insights regarding the trafficking flow.

2.2 Romania

Romania is regarded as the major source country of human trafficking victims in Europe. For more details you can read Bogdan (2025). According to Walkfree (2023), there were 7.5 victims of modern slavery out of 1000 inhabitants in Romania, and the average modern slavery prevalence in Europe and Central Asia is 6.9. Fortunately, as an EU member country, Romania is actively releasing very detailed data on human trafficking victims. The victim-level data can be accessed via the EU Open Data platform. Until January 2026, data from 2015 to 2024 can be downloaded. Here are the links:

>Data from 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024. (in Romanian: victimele traficului de persoane) <

These data provide individual-level information on human trafficking trafficking victims, including age, gender, citizenship, trafficking destination, type of exploitation, educational background, and—most importantly—the county from which the victim originated. In Romania, counties correspond to the EU’s NUTS-3 level, allowing for consistent subnational analysis (N = 42).

Human trafficking prevalence from 2015 to 2017, according to the Romanian open data. From my work: Modern Slavery and Mistrust

2.3 India

India also suffers from the human trafficking problem. According to Walkfree (2023), there were 8 victims of modern slavery out of 1000 inhabitants in India, and the average modern slavery prevalence in Asia and the Pacific is 6.8.

The crime statistics yearbooks Crime In India report the state-level aggregated human trafficking data from the year of 2014. However, researchers should be careful when analyzing the time trends of human trafficking with this data. The variance of the number of victims identified is impressively high, which reflects the time volatility of the data collection process in India. One of the Trafficking in Persons Reports from Office to Monitor and Combat Trafficking in Persons (USA) also mentioned this issue.

2.4 USA

The human trafficking problem has received great public and research attention in the USA, though, according to Walkfree (2023), it is estimated to have only 3.3 victims of modern slavery out of 1000 inhabitants. As a developed country, the USA provides good data accessibility, which helps it avoid the problem of the "data dilemma" to some degree.

The FBI provides state-level aggregated human trafficking victim data from the year of 2014 in the Crime in the US Uniform Crime Reporting Program reports. This series covers the statistics in 2014, 2015, 2016, 2017, 2018, and 2019. You can also find finer-grided data and data after 2019 via the NCRP and NIBRS system.

In addition, the National Human Trafficking Hotline, an NGO targeting human trafficking, releases its hotline data and the state-level aggregated victim data. From NHTH, you can find the yearly aggregated 2007-2014 data, then the yearly data from 2015-2024. An NGO report, Trust and Liberation (2021), used this data.

2.5 Brazil

Brazil is also famous for its modern slavery problem. The renowned book "Disposable People" (Bales, 2012) features a chapter that describes modern slavery practices in Brazil. Brazil is estimated to have 5 victims of modern slavery out of 1000 inhabitants by Walkfree (2023), and the average value is 5 in the Americas. A recent news report on the "modern slavery" practiced in Brazil is about the labor exploitation in the BYD construction site, though the victims are Chinese workers (BBC, 2025).

The yearbooks of crime in Brazil do not release detailed human trafficking data. Passos et al. (2022) collected the human trafficking data from Informatics Department of the Unified Health System in Brazil. From this source, people can access the aggregated state/regional-level victims data. Their online platform supports a conditional filter, which gives researchers more flexibility. However, I doubt that the victim data collected via the Public Health system might be threatened by selection/reporting bias.

Another potential source is from Rangel et al. (2023). They claimed that they collected human trafficking victims from various sources. I did not check if all these sources are openly accessible, but for anyone who is interested in this topic, it is worth trying.

Recently, development economists discovered other sources of the modern slavery measurements. The Ministry of Labor and Employment (MTE) supplied these data through Brazil’s Freedom of Information Law. Araujo et al. (2024) and Cepaluni et al. (2026) use this data source. Most importantly, from Cepaluni & Civitarese's GitHub replication package, we can find the municipality-level (N = 5490) data.

2.6 China

China also faces serious challenges related to human trafficking. Unlike in most developed countries, prominent forms of trafficking include forced marriage and illegal adoption, which have attracted a lot of public attention (e.g., the film Blind Mountain, 2007; 丰县铁链女, 2022). According to my experiences and knowledge, these unique human trafficking patterns can be attributed to the One Child Policy during the 1990s to 2010s, which caused severe gender imbalance in some regions. Official statistics on human trafficking victims are limited and lack a detailed public release. As a result, existing studies have relied on alternative data sources. For example, Liu et al. (2024) used data from the missing-children platform Baobei Huijia (宝贝回家) to construct a city-level dataset on illegal adoption.

For more backgrounds, according to Walkfree (2023), the prevalence of modern slavery is 4 among 1000 inhabitants, and the average in the Asia and the Pacific region is 6.8. According to one of the most famous criminal defense lawyers and criminal law professors in China, Xiang Luo, China is turning from the biggest "exporter" to the biggest "importer" of human trafficking victims. In addition, according to the book 《敬愛的領袖》(Dear leader, by Jang, 2014), refugees from North Korea are also highly prone to human trafficking and sexual exploitation in China.

2.7 Mexico

According to Walkfree (2023), Mexico is estimated to have approximately 6.6 victims of modern slavery per 1,000 inhabitants, slightly above the regional average of 5. The most valuable and publicly accessible source is the open crime incidence data portal of the Executive Secretariat of the National Public Security System (SESNSP, downloading data from: https://www.datos.gob.mx/dataset/incidencia_delictiva). It contains both state-level (N = 32) and municipality-level (N = 2,469) monthly crime incidence data from 2015, as well as the state-level aggregated victim data. After downloading the data, go to column "Tipo de delito", and filter to "Trata de personas".

2.8 Others

I also tried to search data from other countries, countries like Bulgaria and Moldova also provide some detailed information about human trafficking, but they cannot fulfill my demand. It's worth exploring their data for other researchers.

3. Discussion

As shown in the table below, I classified the countries into two dimensions. The first category is how severe the problem of human trafficking is. The second is how the criminal justice and other data are available in this country. If we place countries in this 2-by-2 matrix, we will find that basically the Global North has high data availability and a low human trafficking problem. On the contrary, the Global South usually has low data availability (also quality) and a high human trafficking problem. Romania is a special case: it has a severe human trafficking problem, but it is also an EU member country. This means Romania can release higher-quality data. My matrix illustrates the "data dilemma" clearly.

Countries that support rigorous empirical research are usually also good at controlling or dealing with these social issues. On the contrary, countries that are facing these serious social issues usually lack high-quality data for further empirical research. This "research demand" and "research logistics" mismatch undermines the value of empirical social sciences. Under the current academic production system, most quantitative researchers seek to study the "data-rich" regions or scenarios, and spend taxpayers' money on only marginal improvements. On the contrary, what's really urgent or important topics are handled by the qualitative researchers, who are often underfunded contemporarily. This data problem does not only exist in the human trafficking or modern slave alike "hard-core" crime studies, but also in general empirical studies.

For the "data-rich" developed countries, part of their social unrest nowadays is actually caused by the developing countries. Such as refugees, undocumented migrants, bad human and labor rights protections, human trafficking, and organized crime in and from these countries. I do NOT mean here to stop all immigration, because this will only push people to use illegal market services, and make the organized crime problem more severe. The real solution is to help the victims' and perpetrators' source countries solve their social issues. To achieve this goal, we must build a better understanding of these issues through rigorous empirical research. Therefore, I must call for greater investment and research attention in these fields and countries. For the academic community, showing our value in practice and doing real things is the best way to fight the "anti-science" mindset as well.

This blog aims to fill this gap a bit through sharing my data-searching experiences. Nowadays, we have smarter and smarter LLM and AI tools to facilitate data work and data searching. This might be a change for empirical researchers to explore in more "urgent" topics. Nevertheless, the most important point is to improve the data collection capability in those developing countries. Independent and capable bureaus of statistics and democratic institutions are the only solutions.

References

Araujo, D., Barreto, Y., Castro, D., & Tigre, R. (2024). Illegal markets and contemporary slavery: Evidence from the mahogany trade in the Amazon. Journal of Development Economics, 166, 103177.

Bales, K. (2012). Disposable people: New slavery in the global economy, updated with a new preface. Univ of California Press.

Bogdan, L. (2025). Who are human traffickers? Evidence from Romanian prisons. Comparative Migration Studies, 13(1), 81.

Cao, G., & Rusch, H. (2025). Modern Slavery and Mistrust: A Conceptual Replication of Nunn & Wantchekon (AER, 2011).

Cepaluni, G., & Civitarese, J. (2026). Gun violence and the political economy of forced labor investigations in Brazil. World Development, 204, 107416.

Jang, J. S. D. L. (2014). Dear Leader: North Korea's Senior Propagandist Exposes Shocking Truths Behind the Regime.

Liu, X., Bai, Y., Li, Y., & Sun, Y. (2024). Highway havens for hidden horrors: Expressway connections and child trafficking in China. Journal of Economic Behavior & Organization, 228, 106765.

Monti, Datta, Spires Bob, and Arise (2021). Trust and Liberation. Arise. url: https://www.arisefdn.org/trust-and-liberation-report

Passos, T. S., Santos Santana, M. F., Cordero-Ramos, N., & Almeida-Santos, M. A. (2022). Profile of reported trafficking in persons in Brazil between 2009 and 2017. Journal of Interpersonal Violence, 37(11-12), NP8257-NP8273.

Rangel, F. C., & Schwarz, K. (2023). Are women not enslaved in Brazil? A data-driven analysis of gender dynamics in Brazilian antislavery efforts. The International Journal of Human Rights, 27(1), 117-143.

Rusch, H., Schmitt, M. L., Teufel, J., & Vater, T. M. (2026). Exploitation: Theory and practice. In Handbook of the Philosophical Foundations of Business Ethics (pp. 1-33). Dordrecht: Springer Netherlands.

Weitzer, R. (2015). Human trafficking and contemporary slavery. Annual Review of Sociology, 41(1), 223-242.