Monitoring International Labor Standards: Quality of Information: Summary of a Workshop (2003)

Kevin Bales, Director of Free the Slaves, began his presentation by highlighting the challenges of identifying and measuring slavery (Bales, 2002b). Over the course of the 20th century, most nations made slavery a crime. This forced the problem underground, making it difficult to understand the dramatic changes in the economics of slavery that have taken place since the 1950s. Since that time, rapid population growth has led to “a glut of slaves on the market,” with the result that a slave can be purchased for an average price of $100. The declining price of a slave also reflects changes in how societies organize and view slavery. Today, slavery is “about the use and control of people,” rather than direct ownership. In some parts of the world, although slavery is legally a crime, it is not really considered to be a crime in the popular mind. This poses a challenge to research and measurement. Further complicating matters is the fact that slavery presents a “moving target,” as large criminal organizations ship slaves across borders to support widely dispersed economic activities.

In the past, Bales said, governments had much more control over whether an individual was a slave. For example, in the United States in the 19th century, a slave who left Mississippi and crossed the border into Illi

nois became free. But today governments have little influence, and slavery is based on relationships between people. Bales presented his working definition of slavery—an economic and social relationship that takes many different forms yet retains three core characteristics:

1. loss of the slave’s free will;

2. the use of violence, or the threat of violence, to control the slave; and

3. economic exploitation (the slave receives no recompense for his or her labor).

Bales acknowledged that this definition, based on his analysis of slavery over the past 5,000 years, “doesn’t necessarily match up with official definitions.” He presented a table illustrating how UN conventions have expanded the definition of slavery over the course of the 20th century (see Table 4-1).

Bales said that, after 1998, within UN organizations and activities, “slavery” has been used to describe a large list of relationships and activities, including incest. Explaining that he had argued against these expansions, Bales showed a second table that compares the three core elements of slavery in his working definition with these expanded definitions of slavery (see Table 4-2). He noted that the relationships and activities at the top of the table correspond most closely to his definition, but he would not consider those closer to the bottom of the table to constitute slavery.

Faced with this lack of agreement on how to define slavery, Bales asked, “Where does that all leave us [in terms of understanding slavery]?” He said it would take some time to answer that question, based on research at many levels. The beginnings of an answer will appear from micro-level research that illuminates the stories of individual slaves and slaveholders. This will provide a basis to understand the “mezzo-level”—how slavery fits into local economies and communities. This is the approach Bales took to his study of slavery in five countries (Bales, 1999). He spoke to individual businessmen who held slaves, and he tried to understand how the community accepted the use of slave labor, how the slaveholders understood slavery, how the people who were enslaved dealt with it, and how slavery fit into the lives of both slaveholders and slaves.

Bales contended that macro-level data, aggregating information at the

regional or national level, “are the weakest of all.” To illustrate this, he discussed the “dark figure,” a term from criminology. The dark figure is the distance or gap between the actual number of crimes committed and those that are reported to the authorities. British criminological studies, Bales

said, show that the largest dark figures are for the most frequent and wide-spread crimes, including theft of milk bottles from the doorstep and theft of bicycles. Because more serious crimes (like murder) are usually reported and investigated, the usual rule is that the more serious the crime, the smaller the dark figure. However, this is not the case with slavery and human trafficking, which have enormous dark figures.

Admitting that there are no “magic bullets,” Bales suggested “we step back” to qualitative, exploratory research. It may take longer to research slavery than to learn about violations of other international labor standards, because slavery is a “criminalized . . . labor violation.” Citing Robert Yin (1994), he said case study methods would be appropriate for this research.

The next step, Bales said, was to identify themes that emerge from this qualitative research at the individual and community levels. These themes can be used to conceptualize variables. However, he cautioned that the resulting data are loose, inexact, and unclear. Faced with such inexact information, researchers must decide whether to work with it and build on it or to declare “you won’t begin to work with any numbers unless you are absolutely certain that they are absolutely sound.” His choice, he said, is to go ahead with these variables, while making it very clear that they “are not so robust.”

As an example, Bales presented his test of an ordinal variable he had created to illuminate human trafficking from countries. He spoke with experts in the countries to create estimates of trafficking levels. Then he assembled independent data on country-by-country population growth, poverty levels, government corruption, and other factors that might influence the extent of human trafficking. He then tested the strength of the relationship between the independent variables (poverty, population, growth, etc.) and the dependent variable (estimated level of trafficking) for various countries. The test results suggested that government corruption “is one of the prime indicators of whether or not people are going to be trafficked from a country.” The relationship between government corruption and trafficking appeared to be much stronger than the relationship between trafficking and other variables, including the infant mortality rate and the proportion of the population under age 14. Calling the analysis more of a “sketch” than a real empirical test because of the weakness of the data, Bales said he looked forward to Ann Jordan’s reaction (see below), based on her real-world knowledge of trafficking. As a social scientist, Bales felt that in trying to meet the pressure to provide precise measures of slavery and trafficking, “we’ve been caught with our data down.” Although the primary criterion for judging the quality of social science research should be predictive validity, we are trying to test models with variables that may be invalid, and this limits our ability to test the models. Bales sees no clear answer to this problem, which is faced by all exploratory fields of inquiry.

Despite the problems and weaknesses in the data, Bales suggested that researchers move forward with research on slavery. He noted that, in publishing his recent article on slavery, the editors of Scientific American had coined the term “protoscience” to warn readers of the uncertainties in this field of inquiry (Bales, 2002a). To build this protoscience and increase our understanding of slavery, Bales said, we must first reach agreement on the definitions of key variables, including slavery itself; and we must also centralize information, “which is precisely what is happening in this room today.” In addition, he suggested involving business analysts in the research, because slavery is mostly located within small businesses, “and we

need to understand how those businesses work.” We also need to learn more about “the lived experience of slavery.” Most people, he said, think of slavery in terms of the pre-Civil War southern United States, but slavery takes many different forms today. He also suggested an approach often used by criminologists who want to learn about distribution of a crime— conducting victim surveys.

In conclusion, Bales said that learning about slavery “means we have to go back to first principles.” Researchers need to develop new approaches to advance the field of inquiry. Finally, “We have to think outside the box because I think we are building a new box.”

Ann Jordan, of the International Human Rights Law Group, commended Bales and other researchers for making progress in documenting the scope of slavery and trafficking. She said that Bales’s definition of slavery differed “markedly” from the definition found in UN conventions, which was based on forms of state-sanctioned slavery that no longer exist. Bales’s definition is closer to the definition of forced or compulsory labor in ILO Convention 29, which includes his three core elements—loss of free will, use of force or the threat of force, and economic exploitation. Her preference would be to retain the definition of slavery that focuses on the purchase and sale of human beings and a separate definition of forced labor that focuses on the imposition of force to extract labor.

However, Jordan did not agree with Bales’s identification of factors related to human trafficking, including government corruption, population pressure, and social unrest. Arguing that these factors could lead to a person being smuggled out of the country and arriving at his or her destination as a free person, Jordan proposed an alternative list of factors identified by individuals who have worked directly with victims:

1. lack of a means for legal migration;

2. lack of knowledge of conditions in the destination country (creating a need to depend on third parties to migrate); and

3. presence of friends, relatives, or other trusted individuals who are, in fact, criminals.

Jordan said that her research and activism focus on human trafficking. The recent UN Convention on Transnational Organized Crime includes a

definition of human trafficking in a Protocol to Prevent, Suppress and Punish Trafficking in Persons Especially Women and Children. As defined in this convention, human trafficking includes the crimes described in Bales’s research, as well as additional forms of forced service and labor. Jordan said she prefers to speak about human trafficking, rather than using a “singular definition of slavery,” because it more accurately describes the problem, and it helps ensure that “all persons subjected to these crimes are counted.” She argued that it is important to document all the different types of forced labor and slavery. This includes the buying and selling of human beings; government acquiescence to such practices; forced labor in businesses, homes, factories, fields, and elsewhere; and forced or servile marriages, forced pregnancies, and domestic servitude.

Jordan called for continued research into all aspects of human trafficking to ensure that all victims are counted. She agreed with Bales on the need for economic and business analyses of the problems of slavery and trafficking, as well social science research into the social and cultural forces that “make human trafficking possible.” Finally, she supported Bales’s suggestion that researchers “document the lived experience” of being a victim of human trafficking. She called for “honest and accurate information” in order to “develop effective interventions at all points in the trafficking chain.” She concluded by saying data are “desperately needed at every level and in every country.” Without more accurate and reliable data, she said, progress in combating trafficking would be slow, and assistance and protection for its victims would be scarce.