Policy Research Working Paper 9623 Human Trafficking Definitions, Data, and Determinants Stephen J. Winkler Europe and Central Asia Region Office of the Chief Economist & Social Sustainability and Inclusion Global Practice April 2021 Policy Research Working Paper 9623 Abstract This paper reviews the literature on human trafficking with shows how imprecise definitions of human trafficking a focus on understanding how it is defined and measured and a lack of data and analyses contribute to widespread and what factors contribute to or constrain the prevalence uncertainty regarding the relative effects of anti-trafficking of human trafficking. It finds a growing consensus among policies such as border and migration policies or laws on scholars on the importance of distinguishing between coer- prostitution. The paper suggests several avenues for future cive and non-coercive activity to prevent inflated statistics research that could help clarify these policy debates and and misguided programs and policies. The paper summa- emphasizes the need for additional micro-level data collec- rizes the individual, societal, and institutional explanations tion and analysis. for the prevalence of human trafficking. However, it also This paper is a joint product of the Office of the Chief Economist, Europe and Central Asia Region and the Social Sustainability and Inclusion Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The author may be contacted at swinkler2@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Human Trafficking: Definitions, Data, and Determinants ∗ Stephen J. Winkler+ JEL: E26, J47, O17, F22 Keywords: human trafficking, labor trafficking, sex trafficking, coercion, organized crime, public policy, review ∗ All findings and interpretations in this paper are solely those of the author and do not necessarily represent the views of the World Bank, its Executive Directors, or the countries they represent. I am grateful to Audrey Sacks for her support and guidance on this paper. I also thank Laura Bailey, Guy Grossman, Andrew Rosenberg, Audrey Sacks, Iván Torre, Andrea Woodhouse, and Austin Wright for their helpful comments. This paper was produced as part of a project on human trafficking, which is funded by the Human Rights and Development Trust Fund. + The World Bank. 1 Introduction Human trafficking affects millions of people worldwide each year. The International Labour Organization estimates there are 40 million people in forced labor or marriage, many of whom are also trafficking victims (ILO 2017). Despite contestation over the definition and measurement of human trafficking, as this paper will show, it is clear that human trafficking exists in both weak and strong states and that women, men, and children are trafficked for purposes including domestic servitude, forced labor, and forced sex work (Joarder and Miller 2014). The Sustainable Development Goals include a target to end human trafficking by the year 2030; however, the hidden nature of human trafficking makes it difficult to study and to identify effective policies to confront it. What is human trafficking? How is it identified and measured at the national and global level? What factors contribute to the prevalence of trafficking and what can be done to prevent trafficking? This paper addresses these questions by reviewing existing human trafficking literature in development, economics, political science, public policy, and sociology. The paper focuses on three themes: definitions, data, and determinants. First, despite widespread international agreement on the Palermo Protocol’s definition of human trafficking, the nuances between what is voluntary and coercive creates considerable debate with implications for both research and policy. In research, inattention to definitional nuances creates concerns about what is being measured and the generalizability of findings to other geographic regions and to different forms of trafficking. In policy, unclear definitions that conflate non- coercive and coercive behavior can result in inflated statistics and misdirected programs/policies. 1 Second, because of the definitional challenges and the illicit nature of human trafficking, it is difficult to develop accurate data on trafficking. The data that do exist include global and national efforts to monitor trafficking, qualitative studies that help to untangle the micro causes of trafficking, and a handful of program and policy evaluations. However, there is very little original, systematic, micro-level data on the topic. In turn, some studies rely on data that are collected in different ways across countries and then aggregated into a cross-national data set. Use of these data sets has drawn strong critiques, especially when the analysis based on these data could lead to major policy changes. 2 Third, and stemming from the definitional and data challenges, there is wide variation in what scholars and policy makers view as the key determinants of trafficking. They identify a variety of push and pull factors that drive human trafficking, including individual (i.e. gender), societal (i.e. social norms), and institutional (i.e. markets) forces. The literature reveals some especially strong debates about how different policy approaches, such as bans on prostitution, or changes in domestic and international markets affect overall rates of human trafficking. 1 For example, some argue that inflated statistics led many government and non-government actors to prioritize sex-trafficking despite the fact that labor-trafficking is likely much more prevalent. 2 For example, see Weitzer’s (2015) critique of Cho, Dreher, and Neumayer (2013). 2 The rest of this paper expands on these themes. Section 2 describes the key definitional tensions in the literature and discusses the important implications for both research and policy. Section 3 reviews the existing data sources and attempts to measure human trafficking before concluding with a discussion on the limitations and critiques of these data. Section 4 is dedicated to reviewing the primary causes of human trafficking, including individual (section 4.1), societal (section 4.2), and institutional (section 4.3) factors. Section 5 concludes by summarizing areas of agreement and disagreement in the literature and providing recommendations for future research. 2 Definitions and Their Implications Although the international community widely agrees on a general definition of human trafficking, scholars and policy makers disagree on important nuances within the definition. In particular, the literature emphasizes how difficult it can be to separate human trafficking from similar non- coercive activity. This section reviews the key debates on defining human trafficking and summarizes why these debates matter for research and policy. The Palermo Protocol defines human trafficking as: “recruitment, transportation, transfer, harboring or receipt of persons, by means of the threat or use of force or other forms of coercion, of abduction, of fraud, of deception, of the abuse of power or of a position of vulnerability or of the giving or receiving of payments or benefits to achieve the consent of a person having control over another person, for the purpose of exploitation.” Here, ‘exploitation’ means, “at a minimum the exploitation of the prostitution of others or other forms of sexual exploitation, forced labor or services, slavery or practices similar to slavery, servitude or the removal of organs” (UN General Assembly 2000). This definition emphasizes three key points: • Act: Human trafficking involves the recruitment, transportation, transfer, harboring and/or receipt of a person. • Means: It can include the threat or use of force, deception, abduction, the abuse of power or a position of vulnerability, or other forms of coercion. • Purpose: The purpose of human trafficking is exploitation, which can include the prostitution of others, forced labor, slavery or servitude, removal of organs. Although this is the internationally accepted definition of human trafficking, some states have adopted more expansive definitions and there is vigorous debate in the literature about what constitutes trafficking. For example, Israel includes trafficking for the purpose of surrogacy and the United States includes any commercial sex act of a minor. Recent reviews of the literature on human trafficking are especially critical of scholars’ imprecision in defining the concept (McCarthy 2014; Patterson and Zhuo 2018; Weitzer 2014). In particular, tensions arise in how the following concepts are defined: labor trafficking versus economic migration; migrant trafficking versus smuggling/migration; and sex trafficking versus sex work. 3 2.1 Labor Trafficking vs. Economic Migration First, there is a debate in the literature on the difference between labor trafficking and other, perhaps non-coercive, forms of economic migration. On the one hand, there is extensive documentation of coercive or exploitative labor trafficking in all regions of the world. Andrees and Belser’s (2009) edited volume documents examples of slave labor in Brazil, debt bondage in Latin America, bonded labor in Pakistan, slavery in Niger, and forced labor in Europe. However, some scholars are concerned that non-coercive forms of labor, or forced labor that does not involve trafficking, are too often conflated with labor trafficking. Chuang (2014) argues that the vague international definition of trafficking outlined in the Palermo Protocol allowed states to pursue broad anti-trafficking policies that led to ‘exploitation creep.’ “The definitional muddle has resulted in indiscriminate conflation of legal concepts, heated battles over how best to address the problem, and an expanding crowd of actors fervently seeking to abolish any conduct deemed ‘trafficking,’” writes Chuang (2014, pp 610). Chuang is especially critical of the United States’ criminal justice approach to trafficking, which Chuang argues was only justified because policy makers conflated forced labor and slavery. Several qualitative, and especially ethnographic, studies support the idea that some cases of economic migration are conflated with labor trafficking. Busza, Castle and Diarra (2004) argue that multinational, governmental, and non-governmental groups working to reduce trafficking often misinterpret the cultural contexts of migration. In interviews with Malian children that an NGO identified as victims of child labor trafficking, the authors find that very few of the children reported any form of deception. Similarly, in ethnographic interviews with youth trafficked into the United States, Goździak (2016) finds that most of the children voluntarily agreed to the trafficking as a way of helping their families improve their economic situation. Howard (2008) argues that this conflation contributes to a ‘pathological’ view of youth migration that automatically labels labor migration as abnormal and exploitive. Instead, through ethnographic research in Benin, Howard (2008) argues that children are economic and social actors that often view labor migration as a positive opportunity. 2.2 Migrant Trafficking vs. Smuggling/Migration Similar issues arise in how scholars and policy makers distinguish between trafficking and human smuggling. Some research assumes that any form of illegal migration for the purpose of finding work is a form of trafficking (Kara 2010), or that anyone who facilitates movement for the purpose of sexual activities, forced labor, or removal of bodily organs is a trafficker regardless of consent (Petrunov 2014). Similarly, some governments conflate human smuggling and trafficking in both their policies and their official data on trafficking (Zhang 2012). However, Bilger, Hofmann, and Jandl (2006) argue that the concepts of human smuggling and trafficking in human beings are, by definition, different. According to Bilger, Hofmann, and Jandl (2006, pp 61), “Human smuggling means helping with an illegal border crossing and illegal entry and, therefore, always has a transnational element. This is not necessarily the case with trafficking in human beings, where the key defining element is the exploitative purpose and which can involve cases where no borders are crossed or, in cases where borders are crossed, where entry takes place 4 legally as well as illegally. In the absence of exploitation, smuggled migrants are therefore not in the same situation as victims of trafficking, which in turn has legal and practical consequences on the treatment of each group by law enforcement authorities.” Moreover, Bilger, Hofmann, and Jandl (2006, pp 64) argue that, contrary to popular media reports, human smuggling is not run by mafia-like criminal structures, but by “a complex market for highly differentiated smuggling services offered by a multitude of providers from which potential migrants can choose.” 3 Although this distinction is nuanced, it has important implications for research and policy and underscores the significance of the ‘exploitation’ element of the human trafficking definition. 4 2.3 Sex Trafficking vs. Sex Work Finally, there is significant tension in the literature in how sex trafficking and sex work is discussed, measured, and policed. McCarthy (2014) provides a thorough review of this debate and its origins. McCarthy (2014) argues that the tension is rooted in the fact that the earliest anti- trafficking movements, along with the more modern re-emergence of anti-trafficking advocacy, was rooted in the connection between prostitution and trafficking. The conflation of prostitution with trafficking created a particular frame: “The innocent and helpless protagonist must […] be rescued by gallant law enforcement agents, creating a hero and rescue narrative that is appealing to both the popular imagination and policy makers because it draws a clear moral line between bad traffickers and innocent/good victims” (McCarthy 2014, pp 227). However, this ignores an important distinction between exploitation and agency. 5 As was true for the tension between economic migration and labor trafficking, many qualitative (and especially ethnographic) studies find that individuals who engage in sex work (and who are presumed victims of sex trafficking) often do so voluntarily. For example, Chin (2013) argues that Chinese migrants go to Kuala Lumpur for sex work both because of the income it provides and because of the ‘cosmopolitan lifestyle’ the additional income allows. Kempadoo (2007) argues that sex workers from the Caribbean are often described as victims of sex trafficking because United States anti- trafficking policies conflate sex work with sex trafficking. In a review of literature on sex trafficking into the United States, Schauer and Wheaton (2006) conclude that trafficking is not well defined and that the conflation of trafficking with sex work results in sexist policing of prostitution. Regardless, the literature makes clear that it is not always possible to draw clear lines between what is considered sex trafficking and what is considered consensual sex work. This is partially because individuals who voluntarily enter into sex work may be at higher risk of future exploitation. For example, through interviews with migrant sex workers in Europe, Andrijasevic 3 Bilger, Hofmann, and Jandl (2006) do acknowledge that this does not mean that the market for human smuggling is one with perfect information. This is discussed more in Section 4.2.2 below on information environments. 4 Kyle and Koslowski (2011) also provide an overview of definitions of human smuggling, trafficking, and illegal migration. 5 See McCarthy (2014) for more on the ‘victim/agent dichotomy’ in general. 5 (2010) finds that sex workers often enter into sex work consensually, but the work later becomes coercive. 2.4 Implications for Research and Policy What are the implications of these definitional tensions and how can they be resolved? A lack of clarity on definitions creates ramifications for both research and policy. First, research that relies on unclear definitions has led to sensationalism about the prevalence of trafficking (Keo et al. 2014; Zhang 2009), and contributed to the misperception that sex trafficking is the most common form of trafficking (Chang and Kim 2007; Weitzer 2014). 6 Scholars have been especially critical of Bales’ (2007, 2012) claim that there are more slaves today that during the Atlantic slave trades of the 18th and 19th centuries. 7 Second, policies that ignore the nuances in definitions of trafficking have contributed to a focus on criminalization and made it more difficult to understand real causes of exploitation (Chuang 2007; Kempadoo 2007). In response, scholars have called for more precision in how trafficking is defined and discussed. McCarthy (2014) concludes that the biggest challenge for future research on trafficking will be to establish clear definitions and improve measurement of actual human trafficking. Patterson and Zhuo (2018) emphasize that there is rarely a distinct line between what is and is not trafficking but argue that scholars must be explicit in defining what they are studying. They outline clear definitions of basic terms that are often conflated, including: servitude, forced labor, slavery, trafficking, smuggling, and abusive migration. Bélanger (2014) is an example of a study that pays special attention to the nuances in definition, showing that there is a ‘continuum’ between abusive working conditions, forced labor, coercion, and trafficking. 3 Data and Measurement There are two main challenges to establishing quality data on the prevalence or causes of human trafficking. First, the definitional issues described above make it difficult for governments and researchers to establish clear, universally agreed upon concepts to measure. Second, the illicit nature of trafficking makes it difficult to identify, and creates ethical challenges to collecting systematic data. Existing attempts to identify, measure, and understand trafficking fall into four broad categories: global and national efforts to track the prevalence of trafficking, micro-level (often qualitative) studies that provide case-specific analysis, cross-national (often quantitative) studies that focus on identifying causes of trafficking, and evaluations of program or policy interventions. This section reviews each of these categories and concludes with a summary of the ongoing debates about the quality of data on human trafficking. 3.1 Global & National Tracking Efforts 6 Most empirical studies now argue that labor trafficking is more pervasive than sex trafficking. 7 See Patterson and Zhuo (2018) for a summary of this critique or Bunting and Quirk (2017) for a more thorough critique. 6 Most of the research on human trafficking relies on a handful of prominent data sets used to track the rates of trafficking at the global and/or national levels. 8 The United Nations Office on Drugs and Crime (UNODC) releases an annual “Global Report on Trafficking in Persons.” The 2018 report included data from 142 countries. The report provides a global and regional overview of human trafficking, including rates of trafficking, forms of exploitation, trafficking flows, and institutional responses. UNODC’s findings rely on publicly available administrative data (such as national police reports, national trafficking reports, etc.) and a questionnaire that is distributed to governments. 9 The UNODC warns that its methodology does not allow for cross-country comparisons. The International Labour Organization (ILO) is also a prominent source for data on human trafficking. Their main report, the “Global Estimates of Modern Slavery”, focuses on forced labor (including forced sexual exploitation) and forced marriage. The ILO report relies on data from national surveys (to measure forced labor of adults and forced marriage), case data from the IOM (to measure forced sexual exploitation), and published and unpublished reports (to measure forced labor of children). The International Organization for Migration (IOM) has developed a global database on victims of human trafficking. This is the largest such global database. The database contains over 50,000 individual cases, with approximately 5,000 new cases added each year. IOM currently assists between 7,000 and 9,000 victims annually, collecting a unique source of data on victims of trafficking that is international in scope. Data captured include information about the victims’ backgrounds, trafficking locations and routes, how people fall into the trafficking process, associated forms of exploitation and abuse, sectors of exploitation, means by which victims are controlled, and some information on perpetrators. The IOM data are limited in scope to geographic regions with an IOM programming presence. National efforts to track human trafficking include the U.S. State Department’s “Trafficking in Persons” report. This report provides a global overview of human trafficking, including a brief summary of the prosecution, protection, and prevention actions taken in every country. Each country is placed into one of four tiers, as mandated by the Trafficking Victims Protection Act (TVPA). Tier placement is based on the extent of the governments’ efforts to meet TVPA minimum standards. 3.2 Micro-Level Case Studies Another set of studies focus on micro-level (often regions, cities, or country dyads) cases. These studies often apply qualitative, mixed-method, and ethnographic approaches to gather data on the causes and effects of trafficking. The ethnographic approach, in particular, has been critical in developing a more nuanced understanding of the factors that incentivize trafficking. The case studies also illuminate the volume and range of different industries associated with trafficking. 8 For more on the different estimation methods used to measure the prevalence of human trafficking, see Table 1 in Patterson and Zhou (2018). 9 For more on methodology, see: https://www.unodc.org/documents/data-and- analysis/glotip/GLOTIP16_Annex_I_Methodology.pdf 7 For example, Busza, Castle and Diarra (2004) conducted interviews with Malian children identified by an NGO as victims of child labor trafficking and with Vietnamese sex workers. In both cases, the authors find that very few of the assumed victims report any form of deception or exploitation. In ethnographic research in Benin, Howard (2008) finds that children are economic and social actors and that they often view labor migration as a positive opportunity. Howard argues that local social norms contribute to the supply of child labor. In the United States, scholars have applied ethnographic and mixed-method approaches to understand sex work and human trafficking. Goździak (2016) conducted ethnographic interviews with 140 youth trafficked into the United States and finds that many were trafficked by family members and trusted friends, not criminal syndicates. Farrell and Pfeffer (2014) study policing responses to human trafficking by combining data from 140 closed human trafficking case records with interviews with police, prosecutors, victim service providers, and other court officials or legislators. Farrell and Pfeffer argue that the culture of local police agencies and the perceptions held by police officials about human trafficking prevent the police from seeing a broad range of human trafficking cases. Edited book volumes also provide a number of country and regional case studies. For example, Andrees and Belser (2009) cover slave labor in Brazil, debt bondage in Latin America, bonded labor in Pakistan, slavery in Niger, and forced labor in Europe. Bales (2012) covers prostitution in Thailand, the charcoal industry in Brazil, brick-making in Pakistan, farm workers in India, and water transport in Mauritania. 3.3 Cross-National Much of the empirical literature on human trafficking uses cross-country regressions to investigate the determinants of aggregate levels of trafficking. Using UNODC data from various sources within 161 countries, Cho, Dreher, and Neumayer (2013) examine whether countries where prostitution is legal have better or worse human trafficking records than countries where prostitution is illegal. On average, countries where prostitution is legal experience larger reported human trafficking inflows. Similarly, Tallmadge and Gitter (2018) use EU longitudinal cross- national data to determine the factors that affect the rate of trafficking in a nation. They find that higher rates of trafficking are predicted in nations where prostitution is legalized, immigrants are a larger share of the population, there is access to the sea, the level of GDP per capita is low, and the level of unemployment is low. Toman and Trebesch (2010) analyze the economics of human trafficking and labor migration with unique household surveys from Belarus, Bulgaria, Moldova, Romania, and Ukraine. They find that individual trafficking risks are much higher in regions with large emigration flows. They argue that traffickers face lower recruitment costs in high emigration areas and in areas with illegal migration, which increases trafficking risks. While these studies are laudable, they are limited by their geographic scope (Toman and Trebesch (2010); Tallmadge and Gitter (2018)) or use aggregate, country-level data to make inferences about individual behavior (Cho, Dreher, and Neumayer (2013)). 8 Akee et al. (2007) also use cross-national data to examine how the incentives of traffickers are influenced by interventions that grant legal status to trafficked persons in host countries and tougher legislation targeting traffickers in the source countries. Akee et al. develop a model that focuses on the market imperfections and differential bargaining power of those in the market. Avdan (2012) combines cross-national data on bilateral visa restrictions with data on human trafficking from UNODC to examine how countries shape their migration policies in response to trafficking. Avdan argues that there is a 'virtuous' policy cycle between immigration policy and rates of trafficking – where states implement stricter visa policies in response to increased trafficking rates, and that these stricter policies reduce trafficking rates. 3.4 Program/Policy Evaluations Finally, there are some, but limited, rigorous impact evaluations of programs or policies aimed at reducing human trafficking. In a review of 49 program evaluations of anti-trafficking interventions, Davy (2016) finds that most programs have not been adequately evaluated and argues that more emphasis should be placed on rigorous monitoring and evaluation of anti- trafficking programs. Similarly, Van Der Laan et al. (2011) reviewed several studies aimed at preventing human trafficking for sexual exploitation and found that none of the studies met a level three research method of the Maryland Scientific Methods Scale – that is, a controlled design with both pretest and posttest measures and comparable control conditions. One exception to this is Archer, Boittin, and Hyunjung Mo’s (2016) RCT evaluation of a media campaign designed to raise awareness of human trafficking in Nepal. The campaigns delivered either negative or positive appeals about trafficking through both fact-based messages (posters) and narrative formats (novel, radio, and audio-visual dramatization). Findings suggest that the campaigns increase the ability of respondents to recognize incidents of trafficking among family and friends and increase respondents’ belief that trafficking is an urgent problem that requires government action. Mass media campaigns increase respondents’ commitment to take actual actions to combat human trafficking, including committing to volunteer time or money to anti- trafficking organizations and lobbying the government of Nepal to take greater action against human trafficking. However, the overall effects of the media campaigns did not persist in the long- term. 3.5 Conflicts about Data Quality Given the challenges in defining and identifying human trafficking, it is not surprising that there is conflict in the literature about the quality of data. Overall, there are very few empirical studies that use original systematic data (Laczko and Gramegna 2003; Zhang 2009). Instead, many studies rely on figures from the few established reports. This has sparked growing concern that some, especially cross-national, studies are drawing conclusions based on incomplete or inaccurate data. For example, Weitzer (2014) is sharply critical of Cho, Dreher, and Neumayer’s (2013) cross- national study of whether there is a relationship between prostitution laws and rates of human trafficking. Weitzer notes that the paper uses UNODC data to make cross-country comparisons about rates of trafficking even though the UNODC explicitly states that the data do not contain information on actual numbers of victims and should not be used for cross-country comparisons. 9 Moreover, Weitzer (2014, pp 13) argues, “Even more problematic, the authors relied on aggregate national human trafficking figures (lumping sex trafficking together with other kinds of human trafficking) in their attempt to assess whether legalizing prostitution increases or decreases the amount of all trafficking, not just sex trafficking. Thus, there is a glaring mismatch between the trafficking figures and their relationship to prostitution: The trafficking “data” are based on a compound of different types of trafficking, yet these generic figures are used to assess whether prostitution law is related to the incidence of all kinds of trafficking. More generally, Cho and colleague’s article is an object lesson in the dangers of attempting cross-national comparisons in the magnitude of human trafficking, yet it received wide publicity in the media.” To address this, scholars have called for more primary data collection that measures the incidence and prevalence of labor trafficking activities (Zhang 2012). There is also a call for an increased focus on micro-level studies instead of macro-level cross-national studies (Weitzer 2014, 2015; Zhang 2009). For example, Weitzer (2015) argues that macro-level studies do not control for all the push/pull factors and fail to take advantage of a longitudinal approach. 4 Determinants of Human Trafficking Most of the literature on human trafficking is dedicated to understanding the factors that contribute to its prevalence. Scholars acknowledge that there are several push and pull factors that drive human trafficking and that these factors are difficult to untangle. This section summarizes the literature on causes of human trafficking, grouping the explanations into three categories: individual, societal, and institutional level factors. These categories are broadly defined and there is significant overlap across some categories. For example, gender is discussed as an individual level determinant of human trafficking but there are societal (i.e. social norms) and institutional (i.e. market structures) factors that also shape why gender matters in human trafficking. Therefore, these categories are only meant to provide structure and should not be viewed as rigid boundaries. 4.1 Individual Factors Any focus on individual-level determinants of human trafficking must also acknowledge that nearly every individual-level factor is rooted in larger societal and institutional factors. For example, individual economic hardship is one of the most cited causes of human trafficking. However, an individual’s economic condition is shaped mostly by political and economic structures, which are discussed in more detail in Section 4.3. Therefore, this section is limited to factors such as gender, race, and ethnicity, that are not discussed in detail in other sections. 4.1.1 Gender Gender is a primary individual factor that contributes to trafficking. As Patterson and Zhou (2018, pp 248) note, “women move within their countries to engage in exploitative work in domestic service and sweatshop operations, and they make up the majority of the millions of international migrants across the globe, a significant number of whom end up in forced labor situations.” 10 As described in Section 2, scholars have called for increased attention to the distinctions between non-coercive migration and coercive forms of trafficking. However, in doing so, it is also important to note that women who enter a non-coercive market may be at a higher risk of later becoming exploited. For example, an Amnesty International report on Nepalese migrant workers finds that female migrant workers are trafficked into prostitution or domestic house work at higher rates than male migrant workers (Amnesty 2011). These female migrant workers also report facing violence at higher rates than male migrant workers. Similarly, in reviewing the nature of child trafficking in West and Central Africa, Dottridge (2002) argues that young girls are more likely to be trafficked than young boys. In many cases, the gendered nature of trafficking is also shaped by larger societal or institutional forces (which are covered in more detail in Sections 4.2 and 4.3). For example, women in Nepal are banned from migrating to Gulf states for domestic work unless they receive written permission from their families. This means that women who do want to migrate must do so through underground options that increase the risk of exploitation (Amnesty 2011). Social norms or inheritance policies that encourage investment in boys may also contribute to women being trafficked (Dottridge 2002). High female youth unemployment in origin countries (Danailova- Trainor and Belser 2006), along with demand for female domestic labor in origin countries (Dottridge 2002) also contributes to the gender differential in trafficking victims. Not only are women more likely to be victims of trafficking, but a lack of income or education opportunities also makes it more likely for women to work as traffickers. Through an analysis of administration data and interviews with individuals incarcerated for trafficking in Cambodia, Keo et al. (2014) find that 80 percent of incarcerated traffickers are poor, uneducated women who engaged in trafficking because they lacked other opportunities. Gender and other individual-level characteristics can also play a role in how individuals navigate through trafficking networks. Joarder and Miller (2014) analyze how individual-level characteristics affect the payments that migrants make to traffickers (migrant-debt). Through three field surveys conducted in Bangladesh from April 2009 to November 2010, Joarder and Miller (2014) find that migrants who are young, male, and married make higher payments to their traffickers than migrants who are older, female, or single. 4.1.2 Race & Ethnicity Race and ethnicity, although understudied in the literature, also play a role in trafficking. In a study of debt bondage in rural areas of Bolivia, Paraguay, and Peru, Bedoya et. al (2009) find that indigenous people are more likely to be indebted to their employers than non-indigenous people. The authors argue that this is likely due to systemic social biases that then permeate if and how labor regulations are enforced. Similarly, Kara (2014) finds that marginalized castes and ethnic groups are more likely to be subjected to bonded labor in South Asia. Finally, Dank et al. (2014) 11 provide descriptive evidence of racial differences in the makeup of clients and workers in the sex work industry across the United States. 10 4.2 Societal Factors Research shows that variation in societal factors, and especially differences in cultural context, can shape how trafficking is perceived and the prevalence of trafficking. This section focuses on what is broadly defined as the societal factors that motivate or mitigate human trafficking, including social norms, information environments, and demographic and migration patterns. 4.2.1 Social Norms The literature on how social norms affect human trafficking includes 1) arguments that local cultural norms affect perceptions about acceptable work, 2) studies suggesting that changes in international social norms might help to prevent trafficking, and 3) some early research on how norms affect the agencies tasked with policing trafficking. First, there are several case-specific studies showing wide variation in the social norms surrounding work – especially when it comes to child labor and sex work. On the one hand, this research argues that local social norms should be understood and respected rather than policed or altered. For example, Howard (2008) argues that existing research on migration and trafficking in Benin views the migration patterns among children (who migrate for work) as ‘abnormal.’ However, Howard suggests that local economic and ‘social forces’ – including family perceptions about child labor – make child migration for purposes of work an acceptable practice. In similar research, Howard (2014) examines child labor in West Africa by combining interviews and participation observations with teenage boys, their employers, and their communities. Howard finds that most young boys do not view themselves as victims of trafficking; instead, they view their labor as a ‘reasoned response’ in their social setting. On the other hand, some research suggests that it is important to shift social norms to prevent trafficking. For example, Fitzpatrick (2002, pp 1145) argues that, “preventing and punishing trafficking requires a multi-level game of coordinated norm development, communication concerning activities of transnational criminal groups, mutual assistance in law enforcement, provision of social services to trafficking victims, economic development in source countries, and reforms in migration policy, involving both state and non-state actors at the international, national, and local levels.” This approach suggests that governments, NGOs, and INGOs can work together to ‘mobilize shame’ to strengthen anti-trafficking norms. It also suggests that law enforcement and migration control alone may be ineffective and that, “significant changes in social attitudes and economies in lesser-developed and transitional states must occur […]” (Fitzpatrick 2002, pp 1167). However, other research shows that diffusing anti-trafficking social norms may not necessarily result in changes to local laws or rates of trafficking. Cho (2013) examines the relationship between 10 I include Dank et al. (2014) because it is often cited as an example of how race affects sex trafficking. However, the report is mostly descriptive and does not make strong causal claims about race and sex work or sex trafficking. 12 social globalization, women’s rights, and rates of trafficking in 150 countries from 1981-2008. Cho finds that increased levels of social globalization 11 does improve the economic and social rights of women with legal status in the country. However, increased social globalization may be detrimental to foreign women and could actually increase rates of human trafficking. Cho suggests that this could be because social globalization increases person-to-person contacts across countries, which could increase migration and therefore trafficking. Finally, new research shows that the norms that exist within enforcement agencies may shape incentives that drive trafficking. Using proxies 12 for police corruption, Jonsson (2019) examines how corruption among the police affects the incentives for others to engage in trafficking. Jonsson’s study includes over 100 countries and the results show that countries with more police corruption also experience higher rates of trafficking outflows. Jonsson argues that this is because police corruption lowers the costs for traffickers to engage in recruitment and transport in origin countries. Although this is not a direct study of social norms, it opens the door to future research on how the norms of institutions involved in policing and prosecuting trafficking may indirectly shape flows of trafficking. 4.2.2 Information Environments There are two general ways in which information environments may affect human trafficking: 1) lack of public knowledge about human trafficking or about any government programs intended to address it, and 2) asymmetric information in the human trafficking market. Several scholars argue that trafficking would be prevented if the public were more aware of its prevalence. Much of Bales’ (2007, 2012) work argues that the public is unaware of the extent of modern day ‘slavery.’ Toman and Trebesch (2010) provide some of the only strong empirical evidence confirming that a lack of information is related to rates of human trafficking. Using original household survey data from Belarus, Bulgaria, Moldova, Romania, and Ukraine, Toman and Trebesch (2010, pp 180) show that, “Households in regions with higher awareness of the phenomenon of human trafficking are less likely to have a trafficked family member. The same is true for households which use the TV as main source of information on social and political issues.” 13 Bettio and Nandi (2010) come to a similar conclusion in their analysis of IOM data on sex trafficking victims. They find that, “nationals of Belarus and Ukraine consider trafficking in human beings to be less important than do the nationals of the three other large source countries in the survey—Bulgaria, Romania and Moldavia. Lower awareness of, and sensitivity to, human trafficking often means a higher tolerance of rights violations; or it may simply foster ignorance about such violations and thus reduce the pressure on traffickers” (Bettio and Nandi 2010, pp 32). 11 Cho (2013) instrumentalizes social globalization with the KOF index. It includes information flows, personal contacts, and cultural sharing across countries. 12 These proxies are ‘reliability of the police’ and rates of ‘bribe seeking’ among police. 13 Toman and Trebesch (2010) operationalize information in two ways: “First, we code a dummy for households that use the TV, as opposed to other media and social contacts, to get informed about social and political issues. […] Second, we include a regional risk-awareness measure by taking the share of respondents who stated that they had heard of the phenomenon of human trafficking before.” 13 Others argue that victims would be better served if they had more awareness of government programs aimed at preventing trafficking and rehabilitating victims. For example, although the Government of Nepal established a way to receive complaints and issue compensation to victims, an Amnesty (2011) report shows that migrants were unaware of how to access these systems. The second way information is discussed in the literature is in regard to information asymmetries in the migration, labor, and trafficking markets. In many cases, migrants will not know that they have been exploited until they arrive in a destination country (Amnesty 2011). This is because the market suffers from incomplete information. Bilger, Hofmann, and Jandl (2006) provide a detailed analysis of the human smuggling industry that is best left in their own words. This includes a description of the information problem in the market: 14 “For various reasons (but primarily because smugglers want to conceal their activities from law enforcement authorities) potential clients have only imperfect information on the “quality” of the services they are about to buy from smugglers. This is advantageous for smugglers for several reasons: they have a lower risk of denunciation to the police, they can intimidate their clients and thereby gain better control over them during the journey, and they can charge higher prices to the unaware even if their services are of the same or lower “quality” than that offered by others. On the other hand, this situation of incomplete (“asymmetric”) information also has a drawback for the smugglers. Just like in the “market for lemons” […], it creates the problem of how to convince potential clients that one’s own product (in this case, the smuggling services) is superior to that of others and therefore merits a higher price (smuggling fee). The answer comes in several parts as smugglers, too, have devised various ways to “signal” their trustworthiness to potential clients,” (Bilger, Hofmann, Jandl 2006, pp 66). They also explain the subtle ways in which information asymmetry differentially affects the smuggling and trafficking industries: “The fact that reputation and trust in the smuggler-client relationship are such important elements in the human smuggling industry, distinguishes smuggling from other illegal activities – and particularly from trafficking in human beings and smuggling of goods. As mentioned, trafficking in human beings involves the exploitation of migrants, which in turn makes it difficult, if not impossible, for traffickers to rely on their “reputation” for gaining the trust of their victims. Thus, traffickers need other ways of luring their victims into their fold. Evidence from studies into the trafficking of human beings has shown that trafficking often involves “friends” or even relatives, who already have the trust of the later victims. Thus, traffickers often “buy” the trust of their victims by paying trusted middlemen and by offering “free” transport, while the profits are made through the later exploitation of the trafficking victims. This is not the case in human smuggling processes, where the initiative usually comes from the smuggled migrants and the profits accrue only for the illegal 14 Note that this passage references the smuggling market and not the trafficking network. Bilger, Hofmann, and Jandl (2006) distinguish between the two. The passage is included to highlight the information asymmetries, which also exist in some aspects of the trafficking market. 14 crossing of borders. The traffickers and the smugglers are, thus, in largely separate businesses. One involves the recruitment, transport, control, and exploitation of human beings, who were tricked or sold into this situation; the other involves only the contacting and transport of willing migrants. The first is mainly demand-driven, the second largely supply-driven” (Bilger, Hofmann, Jandl 2006, pp 87). In an attempt to address these information asymmetries, governments and NGOs often implement awareness-raising campaigns. These campaigns are often targeted to potential victims or perpetrators, or to the population as a whole. While there is some evidence that information campaigns can increase the public’s ability to identify incidents of trafficking among family or friends (Archer, Boittin, and Hyunjung Mo 2016), most of the information campaigns have not been rigorously evaluated (Tjaden, Morgenstern, Laczko 2018; Van der Laan et al. 2011). 4.2.3 Demographic & Migration Patterns Demographic factors may also contribute to the prevalence of human trafficking through two different channels. First, demographic trends are linked to migration patterns. Hatton and Williamson (2002) argue that some of the same demographic trends that drove mass migration from Europe in the late 19th century will likely drive much larger migration out of Africa in the near future. Hatton and Williamson (2002, pp 564) highlight three trends driving Africa’s emigration: “First, the living standard gap between Africa and Europe or America is now enormous, and it is unlikely to shrink in the near future. Thus, the huge incentive to move will persist. Second, while Africa is unlikely to record significant catch-up with Europe in the near future, it will still undergo some standard-of-living improvement, which will make it easier for potential emigrants to finance their moves. Indeed, even a modest recovery from the African growth disasters of the 1970s and 1980s will serve to raise living standards at home, thus helping finance the move abroad. Third, young adults—those most vulnerable to a move—will increase their share of the African population and their numbers will rise at a rate far exceeding that of the late nineteenth century or even anywhere else in the Third World during the late twentieth century, even adjusting for the AIDS epidemic.” These projected migration patterns are important because several studies show that trafficking is more prevalent in regions or countries with large migration flows (Cho 2015; Hernandez and Rudolph 2015; Toman and Trebesch 2010). Second, demographic gaps can create demands in labor markets that incentivize trafficking. The dearth of females in certain parts of India (in part due to female feticide), or in China as a result of the one-child policy, have been linked to trafficking of women and girls (US Department of State 2005). In one of the few studies covering internal trafficking, the UN Special Rapporteur on Violence against Women reports on the scale of internal migration and trafficking of women in China, largely from the poorer provinces to those where there are large gender imbalances (IOM 2005). According to the Special Rapporteur, such trafficking accounts for 30-90 percent of marriages in some Chinese villages (Coomaraswamy, 2003). 15 4.3 Institutional Factors Finally, there is a robust literature on the institutional factors that drive trafficking, including political and market structures. 4.3.1 Political Structures Much of the literature assumes that political structures and events are key determinants of human trafficking. In particular, scholars have emphasized the link between political instability and human trafficking. The collapse of the Soviet Union and rise of post-Soviet states (Wylie and McRedmond 2010), the proliferation of civil wars in Africa, Asia, and the Middle East (Patterson and Zhuo 2018), and the general lack of human security (Adhikari 2013; Yousaf 2018) have all been shown to drive displacement and increase the risks of trafficking. There is an implicit assumption in the research that variation in political institutions (especially at the country level) affects outcomes tied to human trafficking. While scholars often include country fixed effects in their analyses to account for this variation, there is very little empirical research on the political institutions themselves. 15 The research that does exist tends to focus specifically on variation in anti-trafficking legislation and enforcement. Therefore, this section focuses on the policy debates in the literature. It first summarizes the debate surrounding what is often referred to as the prosecutorial approach to human trafficking. Then, it reviews how this debate plays out with border and migration policy and anti-sex trafficking policy. 4.3.1.1 The Prosecutorial Approach First, there is a set of research that looks generally at the prosecutorial approach and finds some support for the idea that increased criminal crackdowns reduce trafficking. For example, Frank and Simmons (2013) examine whether increased enforcement of trafficking laws – within a single state or joint enforcement within a dyad of states – affects the prevalence of trafficking. They use the United States’ Trafficking in Persons Reports to code over 5,500 trafficking corridors between 150 origin states, 119 transit states, and 174 destination states between 2001 and 2011. Frank and Simmons (2013, pp 14) find that increased enforcement of trafficking laws within a single state results in significant reductions in trafficking, and that, “when two countries in a dyad both crack down on human trafficking, the result is a significant reduction in observed trafficking within that dyad that is well above and beyond the benefits of one-sided enforcement.” In other words, a prosecutorial approach works best when two states within a trafficking corridor coordinate their increased law enforcement efforts. However, Frank and Simmons (2013, pp 13) also find that increased law enforcement of trafficking laws in one state can create negative externalities in neighboring states. In particular, “external law enforcement among a source state’s contiguous neighbors is associated with an increase in observed trafficking from a source state. This relationship is consistent with the 15 IOM and World Bank (forthcoming) is a rare exception and finds that democratic institutions in general, and especially a commitment to the rule of law, help mitigate increases in human trafficking caused by economic shocks. 16 recruitment and exploitation of new labor supplies when neighboring alternatives become more costly. […] This implies that law enforcement in parts of the world with large potential pools of vulnerable labor can stimulate trafficking from new contiguous sources.” Others are highly critical of an increased law enforcement approach to human trafficking. Chang and Kim (2007) review the United States’ approach toward human trafficking and argue that it emphasizes sex trafficking over other forms of trafficking and therefore focuses too narrowly on criminalization of prostitution. Chang and Kim argue that this approach infringes on the rights of willing sex workers while ignoring abuses of trafficked persons in other industries such as agriculture, domestic service, and manufacturing. When federal governments pursue a prosecutorial approach, it may also open the door for sub- federal governments to use criminalization policies to perpetuate systemic biases against racial minorities and noncitizens. Chacón (2017) examines sub-federal approaches to trafficking by analyzing state legislation, newspaper content, and prosecutorial decisions in nine U.S. states (Arizona, California, Georgia, Illinois, Massachusetts, Missouri, New York, Texas, and Washington) from 2004 to 2014. Chacón (2017, pp 128) argues that, “the attitudes of state legislatures toward noncitizens have played an important role in shaping legislative and law enforcement responses to trafficking within states. Most notably, restrictionist jurisdictions like Arizona, Georgia, and Texas have enacted antitrafficking legislation as an explicit form of migration control, and have deployed antitrafficking laws discursively to justify a broader set of immigration enforcement efforts.” Finally, in a review of 78 push and 67 pull factors, Cho (2012) finds that increased law enforcement presence, and institutional quality as a whole, is not a significant predictor of trafficking inflows to destination countries. 16 Regardless of whether it is effective or not, the prosecutorial approach to anti-trafficking is increasingly adopted across the globe. Simmons, Lloyd, and Stewart (2018) show that between 2000 and 2015, the number of states with criminal statutes against trafficking in persons rose from 10 to 75 percent of countries. The authors argue that the rapid diffusion of anti-trafficking policy is driven largely by how the issue has been framed. By framing human trafficking as a problem of organized crime, instead of as a human rights challenge, states were able to develop policies that view human trafficking as a security threat. Simmons, Lloyd, and Steward (2018) use satellite data on roads to show that governments are more likely to adopt their neighbors’ anti-trafficking statutes when the issue is framed as organized crime. Similarly, Charnysh, Lloyd, and Simmons (2015) argue that there is growing international consensus that human trafficking is a security threat, and that this ‘securitization’ framing enables similar consensus on a prosecutorial anti- trafficking approach. The spread of criminal statutes against trafficking in persons could also be driven by powerful states exerting their social power through naming and shaming. Kelley and Simmons (2015) argue that the United States’ systematic monitoring of human trafficking across the globe, through the annual Trafficking in Persons Report, affects other states’ implementation of policy. Countries that 16 Importantly, Cho (2012) does find that institutional quality is important in source countries, as poor institutions tend to push people into risky situations that can lead to trafficking. 17 are included in the report and that are placed on a ‘watch list’ are more likely to criminalize human trafficking. 4.3.1.2 Border and Migration Policy A similar debate about the pros and cons of a prosecutorial approach exists with specific focus on border and migration policy. 17 On the one hand, there is evidence that tougher border control and law enforcement presence deters trafficking. 18 Tamura (2010) presents a theoretical model where changes in border and migration policy affect smugglers’ capacity to exploit smuggled migrant labor in the destination country. The model focuses on whether countries should invest their resources on increasing inland apprehensions or border apprehensions of smugglers. Tamura’s (2010, pp 541) model suggests that, “improved border apprehension decreases the incidence of smuggling attempts by causing existing exploitative smugglers to become unemployable in the market. Improved inland apprehension, on the other hand, either maintains or even increases smuggling by inducing exploitative and unemployed smugglers to take up nonexploitative smuggling.” Avdan (2012) provides empirical evidence of the relationship between migration policy and trafficking. Using cross-sectional data on visa policies for 192 states and human trafficking data from UNODC, Avdan analyzes the feedback mechanisms between more stringent visa policies and the prevalence of trafficking. Avdan (2012) argues that there is a ‘virtuous’ policy cycle between immigration policy and rates of trafficking: states implement stricter visa policies in response to increased trafficking rates, and these stricter visa policies reduce trafficking rates. 19 Similarly, Hernandez and Rudolph (2015) find that more liberal border control policies (i.e. less stringent visa policy for short-term stays) increase the likelihood of trafficking. On the other hand, scholars have critiqued the ‘securitization’ approach to anti-trafficking policy and argued that some border and migration policies actually increase the prevalence of trafficking. Friebel and Guriev (2006) present a model of illegal immigration that accounts for the intermediaries who help smuggle migrants in return for payments that are typically withheld from the migrants’ future wages (debt/labor contracts). According to this model, “These debt/labor contracts are easier to enforce in the illegal than in the legal sector of the host country. Hence, when moving from the illegal to the legal sector becomes more costly – for instance, because of stricter deportation policies – fewer immigrants default on debt. This reduces the risks for intermediaries, who are then more willing to finance illegal migration. Stricter deportation policies may thus, ex ante, increase rather than decrease the flow of illegal migrants” (Friebel and Guriev 2006, pp 1085). 17 Migration policy and anti-trafficking policy are linked for several reasons, including because studies have shown that human trafficking is more likely in countries with large immigrant populations (Tallmadge and Gitter 2018) or when countries are part of a prominent migrant or refugee corridor (Cho 2015; Hernandez and Rudolph 2015; Toman and Trebesch 2010). 18 See the working paper that Guy shared via email for the best micro-evidence that improved law enforcement capacity at borders leads to decreased trafficking rates. I do not cite or discuss the paper because I do not have permission to do so from the author. 19 Avdan (2012) acknowledges several caveats to this finding, including the fact that, for destination countries, on-site visa restrictions worsen the problem of trafficking. 18 Joarder and Miller (2014) provide an empirical test of the Friebel and Guriev model by analyzing payments that illegal migrants make to traffickers. Using data from three field surveys conducted in Bangladesh from April 2009 to November 2010, they find that payments to traffickers are more common when the costs of illegal migration are high. In other words, policies that increase the costs of illegal migration (i.e. strict visa or deportation policies) may increase the likelihood that migrants are exploited by smugglers. Other critics are more focused on the theoretical and normative reasons against the prosecutorial approach to anti-trafficking. Lobasz (2009) argues that the traditional security approach views trafficking as a threat to the state and borders and therefore emphasizes border security, migration controls, and international law enforcement cooperation. Lobasz presents a feminist approach that prioritizes the security of trafficked persons and therefore emphasizes social services, human rights, safe migration, and worker protections. Regardless, it is important to note that any changes to domestic migration or border policy are likely to create transnational externalities. Theoharides (2018) examines a policy change in Japan that imposed barriers on the migration of Overseas Performing Artists from the Philippines. Theoharides finds that the policy created many negative spillovers in the domestic labor market in the Philippines, including increases in child labor, underemployment, and the number of people engaged in short-term work. 4.3.1.3 Anti-Sex Trafficking Policy Finally, the debate on the pros and cons of a prosecutorial approach also plays out with regard to sex trafficking and prostitution. One of the most discussed studies on this topic is Cho, Dreher, and Neumayer (2013). They use data from UNODC to examine the effects of prostitution laws in 161 countries. Cho, Dreher, and Neumayer (2012) find that, on average, countries where prostitution is legalized experience higher rates of human trafficking inflows. 20 The authors suggest that this is because of a ‘scale effect’ where legalizing prostitution leads to an expansion of the prostitution market. Similarly, both Jakobsson and Kotsadam (2013) and Tallmadge and Gitter (2018) find that European countries where prostitution is legal report higher rates of trafficking victims than European countries where prostitution is illegal. Others find that bans on prostitution have either no effect or may actually increase the prevalence of human trafficking. Hernandez and Rudolph (2015) employ a gravity-type model to examine the causes of human trafficking victim inflows into European countries. They find that legislation regulating commercial sex services has no effect on the rate of identified victim inflows. Hernandez and Rudolph’s (2015) analysis suggests that overall institutional quality (i.e. quality of the legal system) is more important to reducing exploitation than is specific regulations of sex work. In Akee et al.’s (2014) analysis, there is a small positive link between prostitution bans in destination countries and the probability of trafficking. Although the estimates in Akee et al. (2014, 20 Despite this finding, Cho, Dreher, and Neumayer (2013) stop short of recommending a policy of banning prostitution. They acknowledge that such policies might ignore the other potential benefits of legalized prostitution, including improved working conditions for prostitutes. They also acknowledge that banning prostitution raises questions about ‘freedom of choice.’ 19 pp 371) are not statistically significant, the authors argue that, “increases in law enforcement related to illicit-sector activities in both host and source countries mutually reinforce one another and are likely to increase rather than decrease trafficking flows.” 21 4.3.2 Market Structures Another institutional factor that shapes the nature and intensity of human trafficking is market structures. Scholars often apply an economic approach of supply and demand to better understand human trafficking. This approach yields three lines of research that are discussed in the literature: studies that describe the organizational structure of the market, studies that examine the supply and demand factors that shape the market, and studies on how macro/international economic factors such as globalization affect human trafficking. 4.3.2.1 Organizational Structure First, the literature describes a market for human trafficking and provides insight on its organizational structure. Wheaton, Schauer, and Galli (2010, pp 117) “model human trafficking as a monopolistically competitive industry with many sellers (human traffickers) offering many buyers (employers) differentiated products (vulnerable individuals) based on price and preferences of the individual employers.” They argue that there are low barriers to entry or exit for traffickers and that traffickers have some control over the prices. The implication of Wheaton, Schauer, and Galli’s (2010, pp 130) model is that, “Increasing costs to human traffickers is the main way to affect the supply side of the market.” The authors suggest this could be achieved through increased legal coordination to improve punishments of traffickers or improved job opportunities to increase the opportunity cost of participating in the human trafficking market. Aronowitz (2001) also describes human trafficking as an illegal market but applies a more flexible description of the organizational structure of the market. Aronowitz’s approach considers how the market for human trafficking interacts with other illegal markets and large organized crime groups. In this model, the organizational structure may range from a single individual who facilitates trafficking to a segmented business arm with connections to global trafficking operations. Others focus on specific elements of the organization structure. For example, Joarder and Miller (2013) examine the additional payments that are made to traffickers to expedite the process of trafficking. These payments are often referred to as migration-debt contracts. Joarder and Miller use a game-theoretic approach to show that a migrant’s willingness to pay additional money is a function of the effort that traffickers will apply in the migration process. The implication is that, “Any policies […] that facilitate legal migration between countries will reduce the amount that potential migrants will be prepared to pay for illegal border crossings, and hence reduce trafficking” (Joarder and Miller 2013, pp 1341). 4.3.2.2 Supply and Demand Factors 21 Cho, Dreher, and Neumayer (2013) are critical of Akee et al.’s (2014) decision to estimate rates of trafficking at the dyadic country level rather than at the monadic level. 20 Second, the literature examines what factors affect the supply and demand functions in the market and ultimately the prevalence of human trafficking. Many of these studies focus on supply side factors at the community or country level. As described in Section 3.2, there are several case- specific studies that highlight how poor economic conditions and a lack of job opportunities fuel risky migration patterns (Di Tommaso et al. 2009; Dottridge 2002; Howard 2014). Others focus on demand side factors at the community or country level. Acemoglu and Wolitzky (2011) provide a theoretical model that is useful for understanding how labor scarcity and wages shape the incentives for an employer to utilize coercion (and therefore to engage in labor trafficking). In a study conducted for IOM, Anderson and Davidson (2003) examine employer demand for domestic workers and consumer demand for commercial sex services in Denmark, Italy, Thailand, India, and Japan. The authors combine data from semi-structured interviews with two small surveys and conclude that three factors are key to explaining demand for domestic labor and sex work: 1) the unregulated nature of the labor market segments in which these individuals work, 2) the abundant supply of exploitable labor, and 3) the power and malleability of social norms regulating the behavior of employers and clients. 4.3.2.3 Macro/International A final set of studies focuses more on the macro, international factors that shape supply and demand in the market for human trafficking. There is particular interest in how economic development and economic globalization affect the prevalence of human trafficking (Danailova- Trainor and Belser 2006). In a cross-national, longitudinal analysis of the determinants of human trafficking in the European Union, Tallmadge and Gitter (2018) find that countries with higher levels of GDP per capita experience lower rates of trafficking. The authors argue that this may be because richer countries have the resources to invest in reducing human trafficking. In contrast, Peksen, Blanton, and Blanton (2017) argue that economic liberalization leads to increased rates of labor trafficking. They use data from the Economic Freedom Index to examine three common elements of liberalization: business-friendly regulatory environments, reduced government size, and policies that favor economic openness. In a longitudinal analysis of 129 countries between 2000 and 2011, Peksen, Blanton, and Blanton (2017, pp 683) conclude that, “neoliberal policies in general […] are positively and significantly related to labor trafficking at the source and destination stages, and are also conducive to trafficking within a country.” Peksen, Blanton, and Blanton’s (2017) work aligns with other evidence suggesting that new market opportunities and reintegration of former Soviet countries into the global economy drives labor migration in Europe and Central Asia (Mansoor and Quillin 2006). However, it contradicts Cho’s (2013) finding that economic globalization has either a null or positive effect on women’s rights. It also contrasts with Jiang and LaFree’s (2017) finding of a curvilinear relationship between trade openness and human trafficking, which indicates that states in transition between high and low levels of trade openness are the most vulnerable as destination countries for traffickers. Regardless of trade and globalization, recent research shows that economic shocks in source countries are associated with increased trafficking flows (IOM and World Bank, forthcoming). 21 5 Conclusion This paper summarized existing research on human trafficking by drawing on literature from development, economics, political science, public policy, and sociology. It reveals that, although there is consensus in the literature on some topics, several questions remain unanswered. This conclusion summarizes the areas of agreement and disagreement in the literature and provides recommendations for future research. First, scholars tend to agree on the fact that not all alleged cases of trafficking involve coercion or deception. Although early research on human trafficking often conflated non-coercive forms of labor, including sex work, with human trafficking, today scholars recognize that norms regarding ‘acceptable’ work vary across contexts. This is largely due to a growing body of ethnographic research that uncovers the complicated motivations behind voluntary child labor (Busza, Castle, Diarra 2004; Howard 2014) and voluntary sex work (Chin 2013). There is also consensus in the literature that human trafficking exists in both strong and weak states and that it is not caused solely by violent conflict. While it is true that political instability and a general lack of human security have been shown to drive displacement and increase the risks of trafficking (Adhikari 2013; Wylie and McRedmond 2010; Yousaf 2018), the research clearly shows that a number of factors in strong, stable states also contribute to human trafficking. For example, the structure and (un)regulation of labor markets in strong states can incentivize employers to engage in labor trafficking (Acemoglu and Wolitzky 2011; Anderson and Davidson 2003; Bilger, Hofmann, Jandl 2006). Similarly, political institutions in strong states, such as border and migration policies, affect the prevalence and costs of human trafficking (Frank and Simmons 2013; Joarder and Miller 2014). In some cases, these political institutions, or lack thereof, fail to protect trafficked persons and instead infringe on the rights of voluntary workers (Chacón 2017; Chang and Kim 2007). Despite a consensus on these topics, the review also points to several unresolved disputes in the literature. Many of these disputes involve the relative effects of anti-trafficking policies. For example, while some researchers find that bans on prostitution effectively reduce trafficking inflows (Cho, Dreher, and Neumayer 2012), others find that these policies have no effect on trafficking (Hernandez and Rudolph 2015) and result in criminalization of voluntary sex work (Chang and Kim 2007). Similarly, existing research does not offer a clear conclusion on how different border and migration policies affect the human trafficking market. Some evidence suggests that increased border apprehensions and stricter visa policies deter human trafficking (Avdan 2012; Tamura 2010), while others find that these types of policies increase the costs of illegal migration and subsequently increase the likelihood that migrants are exploited (Joarder and Miller 2014). Finally, existing research does not fully address the role of information in the human trafficking market. Some scholars argue that human trafficking can be mitigated by increasing public awareness of the phenomenon (Toman and Trebesch 2010) and by reducing information asymmetries in the market (Bilger, Hofmann, and Jandl 2006). While there is some evidence that information campaigns increase the public’s ability to identify incidents of trafficking in the short- 22 term (Archer, Boittin, and Hyunjung Mo 2016), most information campaigns have not been rigorously evaluated (Tjaden, Morgenstern, Laczko 2018; Van der Laan et al. 2011). These conclusions yield important recommendations for future research. First, more work can be done to untangle the determinants of human trafficking and the factors that shape the rate of trafficking in both strong and weak states. Special attention should be paid to resolving existing debates about the effects of anti-trafficking policies and to improving our understanding of how social norms and information environments affect the supply and demand of trafficking. Second, to address these questions, there is a need for more systematic data collection and analysis at the micro-level. While existing macro-level analyses have made important contributions to theory and evidence, there is growing concern about the comparisons being made with data that are aggregated in different ways across countries (Weitzer 2014, 2015; Zhang 2009). Qualitative micro-level studies have contributed to a better understanding of the local social, political, and economic dynamics that incentivize trafficking (Howard 2008, 2014). Quantitative micro-level studies can build on this foundation. While there are ethical and logistical challenges to the systematic collection of micro-level data on victims of trafficking, IOM’s new data set on Victims of Human Trafficking is one example of advances on this front. Finally, because there is broad consensus that not all alleged cases of trafficking involve coercion or deception, scholars should use precise language in their research and be explicit about what concepts they are measuring. While there is rarely a distinct line between coercive and non- coercive activities (Patterson and Zhuo 2018), scholars should avoid conflating key concepts (i.e. labor trafficking vs. economic migration, sex work vs. sex trafficking) and should be realistic about the generalizability of their findings across concepts and geography. 23 6 References Acemoglu, Daron, and Alexander Wolitzky. 2011. “The Economics of Labor Coercion.” Econometrica 79 (2): 555–600. https://doi.org/10.3982/ECTA8963. Adhikari, Prakash. 2013. “Conflict-Induced Displacement, Understanding the Causes of Flight.” American Journal of Political Science 57 (1): 82–89. https://doi.org/10.1111/j.1540- 5907.2012.00598.x. Akee, Randall, Arnab Basu, Arjun Bedi, and Nancy Chau. 2007. “Determinants of Trafficking in Women and Children: Cross-National Evidence, Theory and Policy Implications.” https://www.researchgate.net/publication/228980483_Determinants_of_Trafficking_in_ Women_and_Children_Cross-National_Evidence_Theory_and_Policy_Implications. Akee, Randall, Arnab K. Basu, Arjun Bedi, and Nancy H. Chau. 2014. “Transnational Trafficking, Law Enforcement, and Victim Protection: A Middleman Trafficker’s Perspective.” The Journal of Law & Economics 57 (2): 349–86. https://doi.org/10.1086/675404. Akee, Randall K. Q., Arnab K. Basu, Arjun S. Bedi, and Nancy H. Chau. 2009. “Combating Trafficking in Women and Children: A Review of International and National Legislation, Coordination Failures, and Perverse Economic Incentives.” Journal of Human Rights and Civil Society, no. 2: 1–24. Amnesty International. 2011. “False Promises: Exploitation and Forced Labour of Nepalese Migrant Workers.” https://www.amnesty.org/en/documents/document/?indexNumber=ASA31%2f007%2f20 11&language=en. Anderson, B., and J. Davidson. 2003. “Is Trafficking in Human Beings Demand Driven? A Multi- Country Pilot Study.” MRS No. 15. International Organization for Migration. https://publications.iom.int/books/mrs-ndeg15-trafficking-human-beings-demand-driven- multi-country-pilot-study. Andrees, Beate, and Patrick Belser, eds. 2009. Forced Labor: Coercion and Exploitation in the Private Economy. Boulder: Lynne Rienner Publishers. Andrijasevic, R. 2010. Migration, Agency and Citizenship in Sex Trafficking. New York: Palgrave Macmillan. Archer, Dan, Margaret Boittin, and Cecilia Hyunjung Mo. 2016. “Reducing Vulnerability to Human Trafficking: An Experimental Intervention Using Anti-Trafficking Campaigns to Change Knowledge, Attitude, Beliefs, and Practices in Nepal.” Research and Innovation Grants Working Papers Series. USAID. https://www.usaid.gov/sites/default/files/documents/1860/DRG_Center_Working_Paper- Vanderbilt_2018.pdf. Aronowitz, Alexis A. 2001. “Smuggling and Trafficking in Human Beings: The Phenomenon, The Markets That Drive It and the Organisations That Promote It.” European Journal on Criminal Policy and Research 9 (2): 163–95. https://doi.org/10.1023/A:1011253129328. Bales, Kevin. 2007. Ending Slavery: How We Free Today’s Slaves. Berkeley: University of California Press. ———. 2012. Disposable People: New Slavery in the Global Economy. Rev. ed. upd. with a new preface. Berkeley, Calif.: University of California Press. Bedoya, Eduardo, Alvaro Bedoya, and Patrick Belser. 2009. “Debt Bondage and Ethnic Discrimination in Latin America.” In Forced Labor: Coercion and Exploitation in the Private Economy, edited by Andrees Beate and Patrick Belser. Lynne Rienner Publishers. 24 Bélanger, Danièle. 2014. “Labor Migration and Trafficking among Vietnamese Migrants in Asia.” The ANNALS of the American Academy of Political and Social Science 653 (1): 87–106. https://doi.org/10.1177/0002716213517066. Bettio, Francesca, and Tushar K. Nandi. 2010. “Evidence on Women Trafficked for Sexual Exploitation: A Rights Based Analysis.” European Journal of Law and Economics 29 (1): 15–42. https://doi.org/10.1007/s10657-009-9106-x. Bilger, Veronika, Martin Hofmann, and Michael Jandl. 2006. “Human Smuggling as a Transnational Service Industry: Evidence from Austria.” International Migration 44 (4): 59–93. https://doi.org/10.1111/j.1468-2435.2006.00380.x. Bunting, Annie, and Joel Quirk, eds. 2017. Contemporary Slavery: Popular Rhetoric and Political Practice. Vancouver Toronto: UBC Press. Busza, Joanna, Sarah Castle, and Aisse Diarra. 2004. “Trafficking and Health.” BMJ: British Medical Journal 328 (7452): 1369–71. Chacón, Jennifer M. 2017. “Human Trafficking, Immigration Regulation, and Subfederal Criminalization.” New Crim Law Rev 20 (1): 96–129. https://doi.org/10.1525/nclr.2017.20.1.96. Chang, Grace, and Kathleen Kim. 2007. “Reconceptualizing Approaches to Human Trafficking: New Directions and Perspectives from the Field(s).” SSRN Scholarly Paper ID 1051601. Rochester, NY: Social Science Research Network. https://papers.ssrn.com/abstract=1051601. Charnysh, Volha, Paulette Lloyd, and Beth A. Simmons. 2015. “Frames and Consensus Formation in International Relations: The Case of Trafficking in Persons.” European Journal of International Relations 21 (2): 323–51. https://doi.org/10.1177/1354066114530173. Chin, Christine B. N. 2013. Cosmopolitan Sex Workers: Women and Migration in a Global City. Oxford University Press. Cho, Seo-Young. 2012. “Modeling for Determinants of Human Trafficking.” Ibero-America Institute for Economic Research. http://EconPapers.repec.org/RePEc:got:iaidps:216. ———. 2013. “Integrating Equality: Globalization, Women’s Rights, and Human Trafficking.” International Studies Quarterly 57 (4): 683–697. https://doi.org/10.1111/isqu.12056. ———. 2015. “Human Trafficking, A Shadow of Migration – Evidence from Germany.” The Journal of Development Studies 51 (7): 905–21. https://doi.org/10.1080/00220388.2015.1010158. Cho, Seo-Young, Axel Dreher, and Eric Neumayer. 2013. “Does Legalized Prostitution Increase Human Trafficking?” World Development 41 (January): 67–82. https://doi.org/10.1016/j.worlddev.2012.05.023. Cho, Seo-Young, and Krishna Chaitanya Vadlamannati. 2012. “Compliance with the Anti- Trafficking Protocol.” European Journal of Political Economy 28 (2): 249–65. https://doi.org/10.1016/j.ejpoleco.2011.12.003. Chuang, Janie A. 2014. “Exploitation Creep and the Unmaking of Human Trafficking Law.” American Journal of International Law 108 (4): 609–49. https://doi.org/10.5305/amerjintelaw.108.4.0609. Coomaraswamy, R. 2003. “Report of the Special Rapporteur on Violence against Women: Trafficking in Women and Forced Prostitution.” Geneva: UNHCR. Danailova-Trainor, Gergana, and Patrick Belser. 2006. “Globalization and the Illicit Market for Human Trafficking: An Empirical Analysis of Supply and Demand.” Working paper. 25 http://www.ilo.org/global/topics/forced-labour/publications/WCMS_081931/lang-- en/index.htm. Dank, Meredith, Bilal Khan, P. Mitchell Downey, Cybele Kotonias, Deborah Mayer, Colleen Owens, Laura Pacifici, and Lilly Yu. 2014. “Estimating the Size and Structure of the Underground Commercial Sex Economy in Eight Major US Cities.” The Urban Institute. Di Tommaso, Maria L., Isilda Shima, Steinar Strøm, and Francesca Bettio. 2009. “As Bad as It Gets: Well-Being Deprivation of Sexually Exploited Trafficked Women.” European Journal of Political Economy 25 (2): 143–62. https://doi.org/10.1016/j.ejpoleco.2008.11.002. Dottridge, Mike. 2002. “Trafficking in Children in West and Central Africa.” Gender & Development 10 (1): 38–42. https://doi.org/10.1080/13552070215890. Farrell, Amy, and Rebecca Pfeffer. 2014. “Policing Human Trafficking: Cultural Blinders and Organizational Barriers.” The ANNALS of the American Academy of Political and Social Science 653 (1): 46–64. https://doi.org/10.1177/0002716213515835. Fitzpatrick, Joan. 2002. “Trafficking and a Human Rights Violation: The Complex Intersection of Legal Frameworks for Conceptualizing and Combating Trafficking.” Michigan Journal of International Law 24: 1143–68. Frank, Richard W., and Beth A. Simmons. 2013. “National Law Enforcement in a Globalized World: The Case of Human Trafficking.” SSRN Scholarly Paper ID 2299624. Rochester, NY: Social Science Research Network. https://papers.ssrn.com/abstract=2299624. Friebel, Guido, and Sergei Guriev. 2006. “Smuggling Humans: A Theory of Debt-Financed Migration.” Journal of the European Economic Association 4 (6): 1085–1111. Goździak, Elżbieta M. 2016. Trafficked Children and Youth in the United States: Reimagining Survivors. Rutgers University Press. Hatton, Timothy J., and Jeffrey G. Williamson. 2002. “Out of Africa? Using the Past to Project African Emigration Pressure in the Future.” Review of International Economics 10 (3): 556–73. https://doi.org/10.1111/1467-9396.00350. Hernandez, Diego, and Alexandra Rudolph. 2015. “Modern Day Slavery: What Drives Human Trafficking in Europe?” European Journal of Political Economy 38 (June): 118–39. https://doi.org/10.1016/j.ejpoleco.2015.02.002. Howard, Neil. 2008. “Independent Child Migration in Southern Benin : An Ethnographic Challenge to the ‘ Pathological ’ Paradigm *.” In . University of Sussex. ———. 2014. “Teenage Labor Migration and Antitrafficking Policy in West Africa.” The ANNALS of the American Academy of Political and Social Science 653 (1): 124–40. https://doi.org/10.1177/0002716213519242. ILO. 2009. “The Cost of Coercion - Global Report on Forced Labour 2009.” Report. Geneva. http://www.ilo.org/global/topics/forced-labour/publications/WCMS_106268/lang-- en/index.htm. ILO. 2017. “Global Estimates of Modern Slavery.” Report. Geneva. https://www.ilo.org/wcmsp5/groups/public/@dgreports/@dcomm/documents/publication /wcms_575479.pdf IOM. 2005. “Data and Research on Human Trafficking: A Global Survey.” Geneva: International Organization for Migration. https://publications.iom.int/books/data-and-research-human- trafficking-global-survey. IOM and World Bank. Forthcoming. “Economic Shocks and Human Trafficking Risks: Evidence from IOM’s Victim Surveys.” IOM and World Bank joint report. 26 Jakobsson, Niklas, and Andreas Kotsadam. 2013. “The Law and Economics of International Sex Slavery: Prostitution Laws and Trafficking for Sexual Exploitation.” European Journal of Law and Economics 35 (1): 87–107. https://doi.org/10.1007/s10657-011-9232-0. Jiang, Bo, and Gary LaFree. 2017. “Social Control, Trade Openness and Human Trafficking.” Journal of Quantitative Criminology 33 (4): 887–913. https://doi.org/10.1007/s10940-016- 9316-7. Joarder, Mohammad Abdul Munim, and Paul W. Miller. 2013. “A Theoretical Perspective on Human Trafficking and Migration-Debt Contracts.” The Journal of Development Studies 49 (10): 1332–43. https://doi.org/10.1080/00220388.2013.812195. ———. 2014a. “Empirical Evidence on Human Trafficking and Migration-Debt Contracts in Bangladesh.” The Journal of Development Studies 50 (3): 399–412. https://doi.org/10.1080/00220388.2013.858128. ———. 2014b. “The Experiences of Migrants Trafficked from Bangladesh.” The ANNALS of the American Academy of Political and Social Science 653 (1): 141–61. https://doi.org/10.1177/0002716213518722. Jonsson, Sofia. 2019. “The Complex Relationship between Police Corruption and Sex Trafficking in Origin Countries.” Journal of Human Trafficking 5 (2): 109–29. https://doi.org/10.1080/23322705.2017.1422090. Kara, Siddharth. 2010. Sex Trafficking: Inside the Business of Modern Slavery. 33661st edition. New York: Columbia University Press. ———. 2014. Bonded Labor: Tackling the System of Slavery in South Asia. New York: Columbia University Press. Kelley, Judith G., and Beth A. Simmons. 2015. “Politics by Number: Indicators as Social Pressure in International Relations.” American Journal of Political Science 59 (1): 55–70. https://doi.org/10.1111/ajps.12119. Kempadoo, Kamala. 2007. “The War on Human Trafficking in the Caribbean.” Race & Class 49 (2): 79–85. https://doi.org/10.1177/03063968070490020602. Keo, Chenda, Thierry Bouhours, Roderic Broadhurst, and Brigitte Bouhours. 2014. “Human Trafficking and Moral Panic in Cambodia.” The ANNALS of the American Academy of Political and Social Science 653 (1): 202–24. https://doi.org/10.1177/0002716214521376. Kyle, David, and Rey Koslowski, eds. 2011. Global Human Smuggling: Comparative Perspectives. Second Edition. Baltimore: Johns Hopkins University Press. Laan, Peter H. van der, Monika Smit, Inge Busschers, and Pauline Aarten. 2011. “Cross-Border Trafficking in Human Beings: Prevention and Intervention Strategies for Reducing Sexual Exploitation.” Campbell Systematic Reviews. The Campbell Collaboration. https://www.researchgate.net/publication/297689687_Cross- border_trafficking_in_human_beings_prevention_and_intervention_strategies_for_reduci ng_sexual_exploitation. Laczko, Frank, and Marco A. Gramegna. 2003. “Developing Better Indicators of Human Trafficking.” The Brown Journal of World Affairs 10 (1): 179–94. Lobasz, Jennifer K. 2009. “Beyond Border Security: Feminist Approaches to Human Trafficking.” Security Studies 18 (2): 319–44. https://doi.org/10.1080/09636410902900020. Mansoor, Ali, and Bryce Quillin. 2006. “Migration and Remittances : Eastern Europe and the Former Soviet Union.” 38426. The World Bank. http://documents.worldbank.org/curated/en/183131468024337798/Migration-and- remittances-Eastern-Europe-and-the-Former-Soviet-Union. 27 McCarthy, Lauren A. 2014. “Human Trafficking and the New Slavery.” Annual Review of Law and Social Science 10 (1): 221–42. https://doi.org/10.1146/annurev-lawsocsci-110413- 030952. Patterson, Orlando, and Xiaolin Zhuo. 2018. “Modern Trafficking, Slavery, and Other Forms of Servitude.” Annual Review of Sociology 44 (1): 407–39. https://doi.org/10.1146/annurev- soc-073117-041147. Peksen, Dursun, Shannon Lindsey Blanton, and Robert G. Blanton. 2017. “Neoliberal Policies and Human Trafficking for Labor: Free Markets, Unfree Workers?” Political Research Quarterly 70 (3): 673–86. https://doi.org/10.1177/1065912917710339. Petrunov, Georgi. 2014. “Human Trafficking in Eastern Europe: The Case of Bulgaria.” The ANNALS of the American Academy of Political and Social Science 653 (1): 162–182. https://doi.org/10.1177/0002716214521556. Schauer, Edward J., and Elizabeth M. Wheaton. 2006. “Sex Trafficking Into The United States: A Literature Review.” Criminal Justice Review 31 (2): 146–69. https://doi.org/10.1177/0734016806290136. Simmons, Beth A., Paulette Lloyd, and Brandon M. Stewart. 2018. “The Global Diffusion of Law: Transnational Crime and the Case of Human Trafficking.” International Organization 72 (2): 249–81. https://doi.org/10.1017/S0020818318000036. Tallmadge, Rachel, and Robert Jeffrey Gitter. 2018. “The Determinants of Human Trafficking in the European Union.” Journal of Human Trafficking 4 (2): 155–68. https://doi.org/10.1080/23322705.2017.1336368. Tamura, Yuji. 2010. “Migrant Smuggling.” Journal of Public Economics 94 (7): 540–48. https://doi.org/10.1016/j.jpubeco.2010.03.005. Theoharides, Caroline. 2018. “The Unintended Consequences of Migration Policy on Origin- Country Labor Market Decisions.” Journal of Development Economics, August. https://doi.org/10.1016/j.jdeveco.2018.07.012. Tjaden, Jasper, Sandra Morgenstern, and Frank Laczko. 2018. “Evaluating the Impact of Information Campaigns in the Field of Migration: A Systematic Review of the Evidence, and Practical Guidance.” Central Mediterranean Route Thematic Report Series. Geneva: International Organization for Migration. https://gmdac.iom.int/sites/default/files/papers/evaluating_the_impact_of_information_ca mpaigns_in_field_of_migration_iom_gmdac.pdf. Toman, Omar Mahmoud, and Christoph Trebesch. 2010. “The Economics of Human Trafficking and Labour Migration: Micro-Evidence from Eastern Europe.” Journal of Comparative Economics 38 (2): 173–88. https://doi.org/10.1016/j.jce.2010.02.001. UN General Assembly. 2000. “Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children, Supplementing the United Nations Convention against Transnational Organized Crime.” UNODC. 2018. “Global Report on Trafficking in Persons.” Vienna: United Nations Office on Drugs and Crime. US Department of State. 2018. “Trafficking in Persons Report.” Washington DC: U.S. Department of State. Weitzer, Ronald. 2014. “New Directions in Research on Human Trafficking.” The ANNALS of the American Academy of Political and Social Science 653 (1): 6–24. https://doi.org/10.1177/0002716214521562. 28 ———. 2015. “Human Trafficking and Contemporary Slavery.” Annual Review of Sociology 41 (1): 223–42. https://doi.org/10.1146/annurev-soc-073014-112506. Wheaton, Elizabeth M., Edward J. Schauer, and Thomas V. Galli. 2010. “Economics of Human Trafficking.” International Migration 48 (4): 114–41. https://doi.org/10.1111/j.1468- 2435.2009.00592.x. Wylie, Gillian, and Penelope McRedmond. 2010. Human Trafficking in Europe - Character, Causes and Consequences. London: Palgrave Macmillan UK. Yousaf, Farhan Navid. 2018. “Forced Migration, Human Trafficking, and Human Security.” Current Sociology 66 (2): 209–25. https://doi.org/10.1177/0011392117736309. Zhang, Sheldon X. 2009. “Beyond the ‘Natasha’ Story – a Review and Critique of Current Research on Sex Trafficking.” Global Crime 10 (3): 178–95. https://doi.org/10.1080/17440570903079899. ———. 2012. “Measuring Labor Trafficking: A Research Note.” Crime, Law and Social Change 58 (4): 469–82. https://doi.org/10.1007/s10611-012-9393-y. 29