Modern Slavery—From Statistics to Prevention

As noted by Kevin Bales in the opening article of this issue, the practice of slavery—that is, the practice of treating other people as if they were property to be bought, sold, given, or exchanged—is probably as old as humanity itself. Today, most of us are fortunate to live in a time and place where it reasonable to expect that practices such as forced labor, human trafficking, and slavery (collectively, “modern slavery,”), will be both criminalized and subject to severe social sanction—but also, statistically speaking, are relatively rare.

While this assumption holds true for many people, as the research highlighted in this issue confirms, it is also the case that modern slavery persists on a perhaps-surprising scale. For example, as Joudo Larsen and Diego-Rosell note in this issue, recent survey research in Bangladesh confirms that 1.5 million people in a population of 160 million have been subjected to modern slavery in the past five years.

The research in this issue also confirms that different people and populations can have markedly different risk profiles. In particular, the impacts of conflict and environmental disasters, associated displacement, health crises and distress migration, malfunctioning justice systems, and deeply embedded discrimination against certain sub-groups in the population, can all combine to create the conditions that enable modern slavery to occur.

Statistics as a discipline has a critical role to play in helping policymakers understand the scale of this problem: a key requirement if sufficient resources are to be allocated to responses. Statistics can also help us understand vulnerability—which, in turn, helps inform our protection efforts. Understanding these factors is critical to the development of responses that either prevent modern slavery from occurring in the first place or, at a minimum, help reduce the severity and effects of the crime itself on victims: both fundamental goals of crime prevention.

While advances are happening rapidly in the statistical analysis of modern slavery, this remains a field that is relatively early in its development. Nonetheless, the early indications of value arising from this work are promising. This article draws on the body of research in the edition to highlight a few select examples of how statistics plays a fundamental role in answering key policy and practical questions related to preventing the crime of slavery.

Given the importance of data in finding solutions, this article discusses some of the measurement priorities for the next five years, as the field collectively seeks to improve our database and knowledge on this horrendous abuse of human rights.

From Understanding Risk to Identifying Responses

“There is nothing as practical as a good theory.”

Writing in 2000 but drawing on generations of different threads of crime prevention theory, particularly that of Cohen and Feison (1979), criminologists Paul Ekblom and Nick Tilley (2000) posit a theory that for any crime to occur, whether it is bicycle theft or drug trafficking, several enabling factors, or precursors, have to converge. They describe this as the “conjunction of criminal opportunity.” This theory has since been expanded and developed by other criminologists, but the fundamental tenets remain the same.

First, for crime to occur, there has to be a vulnerable and attractive target of the crime—the potential victim—someone who may be vulnerable because of individual characteristics, or lacking protection, perhaps because of their social situation.

Second, there has to be a person or people who are both willing to offend (bringing considerations of individual morality and psychology), and able to do so (bringing considerations of access, equipment, and resources).

Third, for crime to occur without interruption, there has to be an absence of willing preventers—the police, the community, and other good Samaritans must either be disempowered, looking the other way, or not even present in the environment.

Fourth, the environment itself must be logistically favorable to potential offenders (promoting concealment or inhibiting pursuit), and unfavorable to those who might seek to prevent a crime from occurring.

Finally, Ekblom and Tilley (2000) note the critical role of crime promoters—people in the larger environment who may simply be unwitting or careless, but who make the crime more attractive (for example, by providing a market for stolen goods or other proceeds of crime).

From this theoretical starting point, it is clear that improving our understanding of individual, systemic, and environmental risks (and protections) must be at the heart of any research effort to prevent crime, including modern slavery. As with any crime, statistical modeling is an important and useful method used to help improve our understanding of risk. As Joudo Larsen and Durgana describe in this issue, statistical testing undertaken for the Global Slavery Index suggests that vulnerability to modern slavery is affected by a complex interaction of factors related to the presence or absence of protection and respect for rights; physical safety and security; access to the necessities of life such as food, water, and health care; and patterns of migration, displacement, and conflict.

Specifically, the vulnerability model in the Global Slavery Index grouped 24 measures of vulnerability into four dimensions, covering: civil and political protections, social health and economic rights, personal security, and refugee populations and conflict (see Table 1).

Table 1—Vulnerability Model in the Global Slavery Index

As can be seen from the vulnerability model, conflict is one of the factors correlating to increased risk of enslavement. As Bales notes in the opening article in this issue, conflict has the effect of destabilizing entire populations and destroying people’s safety nets, removing the protection of the rule of law, and engendering a culture of impunity that enables atrocities to occur.

Consider the profound, institutionalized role of forced labor and slavery in the Second World War, only some instances of which were the subject of official investigation. In 1946, Martin Bormann was sentenced to death by the Nuremberg Military Tribunal for his role in transferring approximately 500,000 female domestic workers from Eastern Europe into slavery in Germany. The following year, the Military Tribunal found a number of corporations liable for their role in Nazi slave labor schemes, noting that a lack of torture or ill treatment does not preclude a finding of slavery: slavery can exist where people are “without lawful process…deprived of their freedom by forceful restraint.” In 1948, Fritz Sauckel, who was in charge of Nazi “labor deployment,” involving the forced recruitment of 500,000 people for slave labor, was sentenced to death.

Unfortunately, just as conflict remains with us, modern slavery in conflict situations is still very much a part of the world we live in today. In December 2016, the UN Secretary General, Ban Ki-Moon, noted to the UN Security Council:

Both ISIL and Boko Haram have engaged in the sexual enslavement of women and girls through trafficking. Yazidi girls captured in Iraq are trafficked into Syria and sold in open slave markets as if they were things, not people.

Indeed, in 2014, ISIL captured 3,000 women and children, mostly from the Kurdish-speaking Yazidi minority group—the largest single capture of women this century. ISIS propaganda claimed that the captured Yazidi women and girls were “spoils of war” to be divided among fighters. Publications released by ISIS provide an extreme interpretation of Sharia describing the legality and illegalities of dealing with slaves: “it is permissible to buy, sell, or give as a gift female captives and slaves, for they are merely property, which can be disposed of…” While some women and girls have escaped or their families have purchased their freedom, it is estimated that thousands remain in the hands of ISIS.

Does the correlation of conflict and modern slavery mean that we cannot end these crimes until we have world peace? Fortunately, policy solutions—just like problems—are rarely as simplistic as this. At a global level, certainly the role of conflict in enslavement points to the importance of peace efforts—but it also points to the immediate need for humanitarian actors to take account of risk of enslavement in conflict situations.

Indeed, in December 2016, the UN Security Council took a critical first step toward ensuring this occurs, through recognizing the role of slavery in conflict situations. Among other things, the UN Security Council Resolution calls for the UN and United States to implement robust victim identification systems to provide access to protection for victims, whether they are in refugee camps in Iraq or in relative safety in Europe. The resolution urges UN agencies involved in humanitarian crises to ensure that risk of modern slavery is considered in humanitarian needs assessments.

This is an approach the International Organization for Migration (IOM) has long called for, on the basis of its field experience of working with victims of modern slavery in conflict situations.

As the IOM and others have highlighted, there is too often a failure in emergency situations to see that responding to modern slavery is, in fact, lifesaving emergency aid. For the women and children who have escaped sexual enslavement at the hands of ISIS, responding to their situation is not an option that can be dealt with after the basics of food, water, and shelter have been handled.

The role of conflict in enslavement also suggests the need to develop innovative law and justice responses that find ways of providing access to justice outside the geographic areas where the rule of law has broken down. For example, in December 2016, 70 years after the Nuremberg Military Tribunals, it was reported that the German government issued the first prosecution warrant against an ISIS commander for genocide—a key part of which has been the brutal enslavement of minority men, women, and children for sexual abuse, as soldiers, or even as human shields.

Files from document raids done to rescue children from situations of slavery. The circled number represents the number of children rescued at a time.
Photo by Grace Forrest, Walk Free Foundation,

Since the effects of conflict are frequently felt well beyond the immediate area of fighting, so too does the risk of enslavement spread. At the end of 2015, the number of displaced people globally exceeded 65 million people—the highest number since records have been kept. Considered against the global population, UNHCR notes this means 1 person in every 113 people globally is now either an asylum seeker, internally displaced, or a refugee. Around half of these 65 million people are children, struggling to exist without some of the most basic human needs of food, water, and shelter; uprooted from education and social protection networks.

Again, the ideal policy result is not that we must stop migration to end enslavement—this would be as illogical as saying we should end the institution of marriage to end domestic violence. Rather, a clearer understanding of risk highlights the critical responsibility that governments have, particularly in countries that host large populations of displaced people, to factor in the heightened risk of enslavement when developing policy responses and services.

For example, policies that ensure displaced people can lawfully gain access to employment markets, housing, education, and health care are critical to enabling them to rebuild their own safety nets and social protections.

As the vulnerability modeling described above suggests, today, as in ancient times, slavery is not only about conflict and mass movement—it can also reflect and result in deeply ingrained discrimination, which may have persisted in the same locations since time immemorial.

The role of discrimination in enabling slavery can perhaps be seen most clearly in the operation of the caste system in South Asia.

In India, for example, economic growth has rapidly transformed the country over the past 20 years. In 1993, some 45 percent of the population was living in poverty; by 2011, that had been reduced to 21 percent. In addition to economic growth, ambitious programs of legal and social reform are being undertaken right across the board, from regulation of labor relations to systems of social insurance for the most vulnerable.

Nonetheless, even with such remarkable change, there are still at least 270 million people living on less than US$1.90 per day in India today. While laws, systems, and attitudes regarding key “fault lines” such as the caste system, gender, and feudalism are rapidly changing, social change of this depth and scale takes time.

In this context, it is perhaps unsurprising that recent survey research confirms that some age-old practices, such as debt bondage and agricultural feudalism, continue to persist. In 2016, the Walk Free Foundation commissioned random-sample surveys on the prevalence of modern slavery to be conducted by Gallup in 15 states of India. Given the use of a random sample and weighting for gender, age, and rural/urban breakdown, these survey responses account for nearly 80 percent of the Indian population. These survey data suggest that more than 18 million people, or 1.4 percent of the total population, are living in conditions of modern slavery in India. Industries implicated in these survey data include domestic work, construction, sex, agriculture, fishing, manufacturing, manual labor, and forced begging.

While bonded labor has been outlawed for decades, narratives available from 2016 survey responses confirm that inter-generational (inherited) bonded labor still persists.

Sir, it is the sin of my father that I have to repay the debts unless I shall have to beg. I have a threat against my family. I am prone to physical violence every day. (Survey respondent, 2016)

This is an old disease in the village that if you are not able to pay off your debts you will have to work as a bonded labor in the field of a powerful person. My husband was employed far from the village so that he can not run away. (Survey respondent, 2016)

In other cases, it appeared that bonded labor reflected debt/lending practices and continuation of a feudal mindset:

There are many people in the village who were working with me as a bonded labor. I was physically and sexually assaulted when I was working in the field. I had also threat on my life and on my family. I was also threatened that I had to leave the village. (Survey respondent, 2016)

It’s very common in this village that we have to work for repay the amount, our family borrowed. I was one of them. My motto was just repay the amount as soon as I could. They threatened to evict us from our homeland, shown their anger if I denied to perform any task. Sometimes the consequence extended to my family and they used physical torture in several times. (Survey respondent, 2016).

The policy relevance of a strong correlation between deep forms of discrimination, such as the caste system and feudal practices, and these traditional forms of slavery is profound. Clearly, simply legislating against discrimination—or formally prohibiting atrocities against people in so-called backward castes—is not enough. At a bare minimum, legislation must be backed up by clear strategies to ensure that these highly marginalized and disempowered populations can, in fact, access the protections that exist on paper. This is an approach demonstrated by the work of the Freedom Fund in its work in Northern India (Harvard FXB Center, 2016).

Advancing Measurement of Modern Slavery

As with any crime, measurement and the collection of relevant, sufficiently detailed data to allow disaggregation is a critical step toward formulating rational, effective prevention policies. As this issue of CHANCE confirms, researchers have made great progress in recent years toward improving the measurement of modern slavery.

From a starting point of estimating prevalence that necessarily relied on secondary sources, research has progressed to the situation today where the multiple methods described in this issue are being used to estimate prevalence in both developed and developing economies. In 2017, random sample, nationally representative, comparable surveys on prevalence of modern slavery have been conducted in more than 50 countries, and equivalent surveys have been undertaken in 15 of India’s states.

Looking beyond survey methodology, as described by Cruyff, van Dijk, and van der Heijden in this issue, methods that were initially developed to measure the number of fish in fjords have been adapted to measuring modern slavery in the UK and the Netherlands. Researchers such as Zhang and Vincent in this issue are looking to apply different sampling techniques, including respondent-driven sampling and adaptive sampling, to apply probabilistic methods to populations that are very hidden.

In short, we are seeing the emergence of increasingly robust and sophisticated methods of measuring the deeply hidden crime of modern slavery.

Progress in research is necessarily iterative. Over time, and thanks to the collective efforts of different institutions, variables have been refined and survey instruments that were initially many pages long have been tested, shortened, and refined. Different sample sizes and methods have been tested and refined, leading to increasingly cost-efficient methodologies. All of this has resulted in methods that are now simpler, shorter, and—as a result—more replicable and scalable than ever before.

Iteration has lead to improved methodologies and to a larger body of data, which improves our ability to validate pre-existing data. As Joudo Larsen and Durgana note in this issue, recent survey data is enabling validity checks of earlier results emerging from extrapolation and modeling, used in earlier editions of the Global Slavery Index. As all of this becomes repeated over time, we start to collectively build our capacity to monitor change.

While great progress has been made, it is equally important to keep looking forward to the next big challenge on the horizon. As the foregoing analysis highlights, any serious efforts to prevent modern slavery cannot be one-dimensional. Preventing modern slavery is deeply interconnected with the achievement of other social, economic, and environmental objectives.

It is in this regard that the framework provided by the Sustainable Development Goals (SDGs) becomes particularly relevant. The SDGs are a set of agreed-upon global priorities that governments will use to organize their efforts to ensure a fairer, healthier, cleaner, and safer world. The 17 goals cover key socio-economic goals, discrimination, law and justice, and environmental concerns.

A key feature is recognition that the SDGs are deeply inter-connections and dependent. A focus on one goal will necessarily affect other goals. Within the larger SDG of “decent work and economic growth,” one of the targets agreed on by governments globally is eradicating forced labor, ending modern slavery and human trafficking, and eliminating child labor (SDG Target 8.7).

Key to realizing the SDGs is a capacity to measure progress against the goals set. As Claire Melamed, inaugural executive director of the Global Partnership for Sustainable Development Data, has noted, data have never been so in demand, yet even some of the most basic data, such as births and deaths, are only partly recorded. Around 50% of the data for the SDG’s predecessors, the Millennium Development Goals, were estimated rather than based on actual counts.

Further, as Durgana and Brown note in this issue, the reality is that critically needed data are not only patchy (if they exist at all); data also can be blatantly political. While the data quality measures they describe will help provide insight into the varying levels of quality of data, this cannot fix gaping holes in global data requirements.

Arguably, the very nature of the challenge of tackling modern slavery—immense, interconnected, and necessarily global—also suggests its solution. Collaboration is critical to achieving comparable statistics at a global level. Whether this involves similar research instruments being used in different locations by different organizations, or pooling comparable data, collaboration has to be the key. For this reason, the Walk Free Foundation entered into a number of collaborative agreements in 2016. These include ones with the University of Nottingham, International Labor Organization (ILO), International Organization for Migration, and UN Office on Drugs and Crime. The intention of each of these agreements is to increase data sharing and capacity for secondary analysis of pre-existing data and standardization, and create efficiencies by pooling resources.

In 2017, the Walk Free Foundation and ILO are collaborating to generate the inaugural Global Estimate of Modern Slavery, a statistical effort to provide a starting baseline for SDG Target 8.7. This builds on earlier collaboration between the two organizations, including surveys in 26 countries conducted jointly in 2016. The collaboration will also draw in data from the International Organization for Migration (IOM). The IOM does not collect prevalence data, but it does collect case history and demographic data about its clients: victims of trafficking who are being provided with psycho-social support.

The Walk Free Foundation and ILO are examining the ways that this (de-identified) unit record data might be used; for example, to help inform estimation of average duration of time in forced labor, a critical component of the stock and flow measure. It is hoped that the collaboration will also draw in data from multiple systems estimation, the approach described in the article in this issue by Cruyff, van Dijk, and van der Heijden.

The Walk Free Foundation and the UN Office on Drugs and Crime also have a partnership on MSE that will result in up to eight more national prevalence estimates being generated through this methodology.

If collaboration is the goal, then standard setting—about operational definitions, indicators, and survey design—is a critical pre-condition of collaboration. In 2008, standard setting around the measurement of child labor by the International Conference of Labor Statisticians, led to many countries undertaking standardized measurement of child labor in their countries for the first time. It is important that the ILO is leading an effort to bring together practitioners working on the measurement of forced labor to identify the minimum requirements for such an undertaking.

The results will be considered by a meeting of the International Conference of Labor Statisticians in 2018. If adopted, the new standards will provide a common framework for measuring forced labor at the national level globally. This should result in an equivalent rapid update and increase in measurement of modern slavery by national statistics offices all over the world.

Next Steps

We are at an important turning point in research on modern slavery. As the articles in this issue show, far from being “uncountable”—new methods are being tested and developed to provide us with more data on prevalence and vulnerability to modern slavery than ever before. As the UN Secretary General has noted, data are critical to decision-making. For further progress to be made, measurement on modern slavery has to become the task of not just international organizations and private foundations, but also national governments. There is also a role for businesses that are operating in high-risk markets.

The methods described in this issue point to the emergence of ways for governments to measure the situations in their own countries. Collecting reliable statistics on a global level does not come cheaply. Understandably, in discussions about the need for improved global statistics, reference might be made to critical funding gaps in support of statistics and related capacity. Indeed, a 2015 needs assessment of capacity of the 77 world’s poorest countries to monitor progress on the SDGs found a funding gap of some US$1billion. Calls have been made for donor governments to increase the share of aid funding that they provide to national statistics offices.

While the funding gap in these 77 countries is critical, this does not explain why only two of the comparatively wealthiest countries in the world—the United Kingdom and the Netherlands—have even attempted measurement efforts of modern slavery in their own countries. It would seem, for example, that the United States, Germany, France, Italy, Japan, and China could follow the lead of the UK and Netherlands, and proceed to undertake their own measurement efforts without further delay.

Without everyone—statisticians, donors, civil society, highly developed countries, poor countries, and potentially even businesses—combining efforts to radically improve measurement, the Sustainable Development Goals, and the life-saving goals that they reflect, including on modern slavery, will remain largely symbolic. That would be a tragedy for millions of people.

Further Reading

Cockayne, J., and Panaccione, K. 2015. Fighting Modern Slavery: What Role for International Criminal Justice? UN University and Freedom Fund.

Cohen, L., and Felson, M. 1979. Social Change and Crime Rate Trends: A Routine Activity Approach. American Sociological Review 44; pp. 588–608.

Ekblom, P., and Tilley, N. 2000. Going Equipped: Criminology, Crime Prevention and the Resourceful Offender. British Journal of Criminology 40, pp. 376–398.

Harvard FXB Centre. 2016. When We Raise Our Voices: The Challenge of Eradicating Labor Exploitation (PDF download). An evaluation of community empowerment intervention in Uttar Pradesh, India.

International Labor Organization. 2012. Hard to See, Harder to Count. Geneva, Switzerland: ILO.

Joudo Larsen, J., Datta, M.N., and Bales, K. 2015. Modern Slavery: A Global Reckoning. Significance, pp. 32–36.

Ki-moon, B. 2016. Remarks at Security Council Ministerial Open Debate on Trafficking in Persons in Conflict Situations.

Melamed, C. 2015. The Revolution will be Analysed, Significance, pp. 24–26.

Silverman, B. 2014. Modern slavery: an application of multiple systems estimation. Gov. UK.

UNHCR. 2016. With 1 human in every 113 affected, forced displacement hits record high.

Van Dijk, J.J.M., and van der Heijden, P.G.M. 2016. Research Brief. Multiple Systems Estimation for estimating the number of victims of human trafficking across the world (PDF download). Vienna, Austria: UNODC.

Walk Free Foundation. 2016. Global Slavery Index.

About the Author

Fiona David is a lawyer and criminologist, and executive director of global research at the Walk Free Foundation. She has led the foundation’s research since its inception in 2012, including the Global Slavery Index (entering its fourth edition in 2018) and the forthcoming Global Estimate of Modern Slavery (co-authored with the International Labor Organization in collaboration with the International Organization for Migration).

Back to Top

Tagged as: , , , ,

Leave a Response

Please note: comment moderation is enabled and may delay your comment. There is no need to resubmit your comment.