Accountable to Whom? Accountable for What? Understanding anti-child trafficking discourse and policy in southern Benin

In Benin, anti-child trafficking discourse misrepresents the nature of youth labour migration, while anti-child trafficking policy fails to protect those “beneficiaries” in whose name it is officially designed. Despite this, both have remained stable for over a decade. This paper attempts to explain why. It argues that, in contrast to claims made by many other critiques of anti-trafficking work that policy makers are either ignorant or malevolent, here discourse and policy are hampered more by the conceptual, institutional and political structures within which they are developed and articulated by individuals, thereby ensuring discursive and policy stability despite inaccuracy and failure.


Introduction
The story behind this paper begins in 2005. Fresh-faced and na ve, I travelled to Benin to start working with a child rights NGO on the topic of child trafficking. Well-versed in the standard horror stories, I quickly learnt, through meeting young labour migrants, "victims of trafficking", and their communities, that all was not quite as it seemed. As with elsewhere in the world, 2 a disjunct existed between the way institutions represented and responded to "trafficking" and the "trafficked", and the way the trafficked understood and represented themselves. My research has been designed to explore this disjunct, and has sought to ask what sustains it. Where answers often point to Machiavellian politicians uninterested in the lives of the poor, in what follows, I will offer a different analysis. I will argue that inertia, stability and resistance to change within anti-trafficking discourse and policy result from the discursive, institutional and political constraints within which they are elaborated. The paper begins with a background to the emergence of child trafficking as an issue for concern in Benin and explains the research I have conducted into it. It briefly identifies the gap between narrative and reality and that between current and optimal 1 All names of individuals, institutions and villages have been either changed or anonymised to protect the identity of informants. Details can be provided on request. Unless otherwise indicated, all translations from French are my own. 2 See, for example, L Agustin, 'Migrants in the Mistress's House: Other voices in the "trafficking" debate ' Social Politics, vol. 12, no. 1, 2005, pp. 96-117 ' International Migration, vol. 43, no. 4, 2005; S Castle, and A Diarra, The International Migration of Young Malians: Tradition, necessity or rite of passage? School of Hygiene and Tropical Medicine, London, 2003;A de Lange, 'Child Labour Migration and Trafficking in Rural Burkina Faso' International Migration vol. 45, no. 2, 2007, 147-167;M Dottridge, 'Introduction' in M Dottridge (ed.) Collateral Damage: The impact of anti-trafficking measures on human rights around the world, GAATW, Bangkok, 2007. policy. Subsequently, it grounds its analysis in the work of discourse theorists such as Norman Fairclough and Michel Foucault and development ethnographers such as David Mosse.

Setting the Scene
Human trafficking exploded as an issue in Benin with the Etireno affair in 2001. The Etireno was a Nigerian trawler used by people-smugglers to illegally transport Beninese and other West African children 3 to Gabon. After a complex series of events involving law enforcement and diplomatic personnel from various countries, the ship and its passengers were left stranded at sea, prompting local child protection organisations to raise the alarm and the world's media to descend on the region in order to report on 'the slave ship' that heralded the uncovering of 'a modern-day slave trade'. 4 Benin was quickly identified as an epicentre of the traffic in children, and the government responded by ratifying the UN Trafficking Protocol, 5 opening its doors to an influx of anti-trafficking money, and jointly establishing various anti-child trafficking initiatives.
The definition of child trafficking in operation in the country broadly classes the crime as any part of the process of '(coerced) movement for the purpose of exploitation'. As per Article 3(c) of the Trafficking Protocol, a minor's consent to engaging in activities defined as 'exploitative' is deemed juridically irrelevant, such that these activities (and anything facilitating them) are thus always seen to constitute trafficking. Within both the Trafficking Protocol and Benin's national 3 Though I consider "children" and "childhood" to be situational concepts varying according to time, place, culture and social structure, amongst other things, the institutional literature and policy I am examining take "children" to be those under the biological age of 18 and "childhood" thus to be the period before one reaches that age. Since I am reflecting on the appropriateness of these institutional understandings and policies, I have decided to engage with them on their own terms, and have therefore used the same criteria throughout this study. In my study, however, I focused almost entirely on adolescent migrants, and the word "child" will be used interchangeably with "youth" unless otherwise specified. 4 The Independent, 18 April 2001, http://www.independent.co.uk/news/world/ africa/mystery-of-missing-children-as-slave-ship-docks-681766.html.

5
United Nations Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children, Supplementing the United Nations Convention Against Transnational Organised Crime, UN General Assembly, New York, 2000. legislation, exploitation comprises a non-exclusive list of situations, including forced labour, slavery or servitude, but also, in Benin, work which 'by its nature and/or the conditions in which it is exercised, may damage the health, safety or morals of the child'. 6 This draws on the International Labour Organisation (ILO)'s related anti-child labour framework, according to which those activities deemed inherently exploitative include work in: mining and quarrying; manufacturing; construction; electricity, gas and water; sanitary services; transport, storage and communication; and plantations and other agricultural undertakings mainly producing for commercial purposes, but excluding family and small-scale holdings producing for local consumption and not regularly employing hired workers. 7 In practice, this has led to the de facto criminalisation of most of the remunerated activity in which young people engage in Benin. My research into the appropriateness of this criminalisation and into the operation of anti-trafficking policy more broadly in Benin consisted of 14 months of multi-sited fieldwork, during which I interviewed more than 300 people and observed and worked with individuals and institutions at every level of the anti-trafficking policy chain. I focused at the institutional level on those bodies that are at the heart of producing and sustaining the anti-trafficking discourse in Benin, and at the forefront of forming and implementing anti-trafficking policy, both in Cotonou, Benin's de facto capital, and throughout the country more widely. These bodies include: UNICEF and the ILO, from among the international organisations working in this field; the US Departments of Labour and of International Development, the Danish Development Agency, the EU and France, from the donor community; the Family and Justice Ministries, from within the Beninese government; and a collection of national and international child-focused NGOs operating in Benin.
At the community level, I concentrated my research entirely in the south of the country, principally because, in Benin, the "hub" of child labour, trafficking and exploitation has been widely identified as the  ILO, Geneva, 1973. DOI: 10.14197/atr.201213 poor, agricultural Zou Department, from which teenage boys frequently leave to work in the gravel pits of Abeokuta in Western Nigeria and girls leave for domestic work in the major coastal cities. Within this department, two communes-Zakpota and Zogbodomey-have been identified as particularly affected by the above-mentioned phenomena. I thus selected four case-study villages, two from each commune, two with flagship "village child protection committees" and two without. In these villages, I used various participatory techniques, including principally open-ended individual and group interviews, the focus of which was on how people see their life-worlds, how they experience and view (hazardous) youth work and movement, and what they have to say about current anti-trafficking interventions. In total, I interviewed 109 adults and 61 adolescents from these villages. Additionally, I interviewed 15 adults and 28 adolescents from various other locations in southern Benin, during my Masters fieldwork in 2007. The current and former youth migrants I interviewed were predominantly male, since I was principally examining adolescent male labour migration to the gravel pits and surrounding areas in Nigeria; I did however interview a number of female former domestic servants in institutional care.
The data have been carefully coded and continually analysed with relevant actors over the course of the research's unfolding and in subsequent return trips. Though my sample does provide a good cross-section of deep, qualitative data, it is not statistically representative and should not be read as such.

Narratives, Policies and their Discontents
The dominant institutional narrative around trafficking in Benin holds that trafficking is a major problem in the country and involves unsuspecting children being forced away from home and into exploitative work, largely as a result of pathological cause-factors, ranging from grinding poverty to the corruption of traditional practices and parental irresponsibility. 8 This narrative is illustrated in the extract below, from an emblematic briefing paper written for the US Secretary of Labor, on the event of his visit to Benin as part of the US "war on trafficking": Child trafficking is a multifaceted phenomenon in West Africa. It started with parents placing their children with relatives. That placement is called in Fon, one of southern Benin languages [sic], "vidomegon". The culture of vidomegon, originally allowed more fortunate members of the community to receive the children of less fortunate members, in a climate of solidarity. The idea is that by confiding a less fortunate child in the home of someone who is better endowed economically, that child will be better taken care of. This practice is rampant all over West Africa. However, over time the practice has been abused by individuals who have sought financial rewards, resulting in a behavior where children are given to traffickers, who in turn sell them to agents in neighboring countries. Victims of this new practice are reduced to mere commodities that are bought, sold, transported, and resold according to market forces of supply and demand. Most trafficked children are threatened with physical and emotional abuse, and nearly all suffer from neglect or diseases. Poverty is one of the causes of child labor and trafficking in Benin. 9 Against such a discursive backdrop, it should come as no surprise that anti-child trafficking policy is highly interventionist in Benin, with emphasis placed on the prevention of exploitative work by the preemptive thwarting of the migration seen to lead to it. 10 Measures include the strengthening of border patrols, enhanced cooperation with Nigeria on trans-border surveillance, widespread anti-work and antimovement "sensitisation" or "responsibilisation" campaigns, the promulgation of the Law Regulating the Movement of Minors and Suppressing the Traffic in Children, and the establishment of village vigilance committees expected to prevent youth labour migration at the village level. 11 Poverty is also addressed, though only through individualised schemes involving the promotion of apprenticeships. 12 Though tens of millions of dollars have been invested in these initiatives, my qualitative research with young labour migrants and their communities strongly suggests that they fail. Young people still migrate and they still work. Indeed, these extracts from my notes are indicative of the answers I received when I asked them whether people pay heed to anti-trafficking efforts: Have some of you ever been away to do holiday work? Yes, every single one of us! That is what allows us to continue at school! You can go to Nigeria or Sav and earn 30,000 or 40,000 FCFA in a summer. What work do you do? In Sav , it is the land. In Nigeria, it is the land but also the quarries. How is it? It's hard but it's ok. Do NGOs, white people or the government come here and say that it's bad? Yes, loads. Why? Because they see that it can be hard, but they offer us no alternative. What do you say to them? When they come and speak to us, their words go in one ear and come out the other. We listen and then we ignore them. 13 How do you go about getting around the authorities? They said the state has set up village committees all over the place, but these are corrupt. We can easily turn them and take kids away, no problem. There are also many paths that you can take towards and across the border and the state has no idea about them all. The police sit there and guard the ones they know about and so we just take the others. The state is really stupid. Sometimes we even take kids down to Cotonou and then, under the nose of Yayi Boni [Benin's President], we take them on a boat to Nigeria or Gabon. 14 My research into community understandings of youth work and migration offers an indication of why this might be the case. In contrast to the institutional discourse which paints most work and migration as exploitation and trafficking, young labour migrants and their communities see their work and migration as both necessary and highly constructive. Central to this understanding is the well-founded belief that economic opportunity is concentrated away from the resourceand capital-poor village and that labour migration thus represents the only viable path to a better future, since it allows them to access money and bring it home. In the words of one adolescent former migrant labourer: We don't have the same view [as the authorities]. We think that leaving can be good, especially as it is a way to find money, or to come back and put a roof on your father's house. 15 In the absence of policies which directly target the very clear, structurally-influenced material underpinnings of youth labour migration in Benin, 16 these views are unlikely to alter.

And Yet Things Remain Unchanged…?
Despite the inaccuracies within the trafficking narrative and the failures of trafficking policy, little has changed over the past ten years. The major questions posed by my research, therefore, are: how is it that discourse and policy can get it so wrong? and why have they not developed? Research from similar contexts identifies policy-maker ignorance or politicised malevolence as an explanation, suggesting that relevant actors either fail to understand ground-level realities or remain uninterested in addressing them. 17 I believe this to be reductive. Indeed, it attributes too much agency to those in power and fails to take sufficient account of the power of the structures within which they operate. In this section, then, I seek to offer a more nuanced explanation for discursive and policy stability. I will do so by deploying discourse theory and by focusing on the importance of "orders of discourse" (and of powerful actors in structuring those orders) in framing and constraining both policy-maker understandings and policy choices.

Orders of discourse
According to Van Dijk, discourse is 'an essentially fuzzy concept' while, for authors such as Gee, it can include, at its most basic, any instance or form of communicative action. 18 In more complex usages, the term also refers to systems of meaning and histories of multi-semiotic interaction. In this paper, I will be following theorists such as Fairclough and Foucault in understanding discourse as short-hand for 'relatively stabilized configurations of discourse practices', 19 or 'orders of discourse', 20 by which is meant 'systems of thought composed of ideas, attitudes, courses of action, beliefs and practices that systematically construct the subjects and the worlds of which they speak', 21 as well as being constructed by them. As Candlin explains, this denotes 'a means of talking and writing about and acting upon worlds…which both constructs and is constructed by a set of social practices within these worlds, and in so doing both reproduces and reconstructs afresh particular social-discursive practices, constrained or encouraged by more macro movements in the overarching social formation'. 22 In other words, discourse, as it is understood here, is a set of relatively stable, linguistically expressed social understandings of phenomena, continually (re)constituting and being (re)constituted by the practices of individuals within time and space. In this section, I will reflect on two of the core social understandings I believe to lie at the root of antitrafficking narratives and anti-trafficking policy in Benin. These understandings are (western notions of) childhood and neoliberalism.

Childhood
Within the anti-trafficking community working in and on Benin, a very particular notion of what childhood is and should be persists. In contrast to localised understandings of childhood and child development, the policy establishment here betrays a very rigid, westernised conception of childhood that centres on the belief that children are inherently vulnerable, that they are in need of protection from work, and that their development can only be guaranteed at home and at school. Gidi, for instance, is a senior Beninese government official. In one interview we conducted, she explained that as a result of her country's anti-trafficking work, 'people are finally understanding what it means to be a child'. 23 In her view, this meant a vulnerable school-goer who should be protected from work. Similarly, Veronica, a local-level Beninese official, argued that all people under 18 are children, they are still developing and therefore they must be prevented from migrating for work even if they want to. 24 Why, one may ask. Because, in the words of Cyril, Veronica's colleague: 'When we send children away, they end up missing out, so it's better if a child stays here. That way he has more chance of developing correctly, of becoming a man tomorrow.' 25 As has been argued extensively elsewhere, 26 though these understandings reflect norms which emerged in, and are particular to, the socio-historical and economic context of the industrialising/ industrialised West, they have become globalised through their centrality to the operations of international child protection institutions. 27 As such, they frame the (understandings of the) individuals and bodies working across the anti-child trafficking field. And, as Dorte Thorsen has argued in her work on anti-child trafficking discourse and policy in Burkina Faso, they are 'instrumental at policylevel in condemning adolescents' migration and attracting funding for projects aimed to "protect" and repatriate rural children and youth on the move'. 28 This has, of course, also been the case in Benin, where anti-trafficking strategy relies on attempts to keep under-18s at home and in school, and to criminalise work that is viewed through the lens of international norms as de facto exploitative and de facto bad for children.

Neoliberalism
The second major framing discourse in operation within Benin's antitrafficking world is that of neoliberalism. Whilst debate abounds as to what exactly neoliberalism constitutes, there is general agreement that it represents 'a political philosophy of governance…premised on a mantra of market rationality and on the active encouragement of laissez-faire economic systems'. 29 It involves a reduction of the state's role as service-provider and labour protector, 30 privatisation of state resources, the acceptance of pre-existing distributions of wealth and power, and the placing of responsibility for economic well-being firmly on the shoulders of the individual economic agent-in Foucault's terms, the homo economicus, whose duty it is to maximise himself as a vehicle of capital and self-advancement. 31 That such a philosophy underpins the work of the Beninese state and its partners in their fight against child trafficking is best illustrated by an examination of the way in which poverty-apparently the prime cause of trafficking-is understood and responded to in the country. Almost always, poverty is constructed as a pre-existing, a-political fact that is just out there. In the National Anti-Trafficking Plan of Action, there is no discussion whatsoever of why people are so poor that they apparently allow their children to be exploited, or which 28 D Thorsen, ibid. p.7.

29
K Mitchell, 'Neoliberal Governmentality in the European Union: Education, training, and technologies of citizenship ' Environment and Planning D: Society and Space, vol. 24, 2004, pp. 389-407, p. 389. 30 See, for instance, V Sewpaul and D Holscher, Social Work in Times of Neo-Liberalism: A post-modern discourse, Van Schaik, Pretoria, 2004. 31 G Burchell, et al., The Foucault Effect: Studies in governmentality, University of Chicago Press, Chicago, 1991 relations of power underpin their poverty. 32 Similarly, in my interviews with policy makers, only one of over a 100 institutional employees independently elaborated on Benin's poverty beyond simply acknowledging that it was a reality. Indeed, when I suggested that we take time to reflect on where that poverty comes from, reactions ranged from blank stares to periods of silence. This is further reflected in the way in which anti-trafficking policy seeks to address poverty. Not only does it avoid directly targeting the structural underpinnings of youth work, movement and the poverty which frames it, but when it does "tackle poverty", it does so through the individualised promotion of apprenticeships, school bursaries or family-based income generating activities. Poverty is thus conceived here as an issue of personal responsibilisation, with individuals moulded to become the self-sufficient, independent and market-oriented economic actors embodied by the homo economicus who adapts himself to pre-existing economic conditions. 33

Bringing power back in -vertical constraints
In his seminal discussion of Critical Discourse Analysis, Norman Fairclough follows Foucault in arguing that although structures and agents do interact in forming orders of discourse, we should not be blind to the fact that some agents have more power than others in setting and maintaining the stability of these orders. He writes: The power to control discourse is…the power to sustain particular discursive practices with ideological investments in dominance over other alternative (including oppositional) practices.…Power is conceptualized both in terms of asymmetries between participants in discourse events, and in terms of unequal capacity to control how texts are produced, distributed and consumed (and hence the shape of texts) in particular sociocultural contexts. 34 Critical discourse analysis is thus a method for studying the 'connections between language, power and ideology', a means of deconstructing both 'how power relations constrain and control productivity and creativity in discourse practice, and how a particular relatively stabilized configuration of discourse practices ("order of discourse") constitutes one domain of hegemony'. 35 In the case of Beninese anti-trafficking policy, my research reveals very real exercises of power in both constructing and maintaining neoliberalism and western childhood as the base-level operative orders of discourse.
In the case of neoliberalism, this is again best illustrated by examining the way in which poverty is understood. In Benin, cotton is the major cash crop. It accounts for around 5 per cent of the GDP and almost 40 per cent of the country's export receipts. 36 Unlike in many other contexts, it is principally a household industry and provides income for thousands of families, including in the area where I conducted my ground-level research. When prices are high, people benefit, with more farming children in school, less migration for work and small-scale household projects being realised. Apart from the spike recorded over the past few months, cotton prices have been at record lows for over a decade, in large part due to illegal US subsidies, recognised as culpable by the WTO in a case brought by Brazil to its Appellate Body. 37 Many policy makers know this. Yet, when asked why they do not therefore advocate a removal of subsidies as part of the fight against the poverty that underpins trafficking, they admit that doing so is either beyond their organisational capacity or politically impossible. Rose, for example, is a senior UN figure working in Cotonou. She explained that her institution could do nothing about such structural issues because it was 'politically unacceptable'. 38 Matt, a key US government figure, said that while he would like to raise the issue, he was doubtful of success because there were 'big interests to fight'. 39 Sandra, a senior donor government employee based in Benin, offered a similar assessment, after which I made the following notes: What does this entail? Are the subsidies mentioned? Yes, she said, the effects of this are taken into account, but only at the ground level. They can't talk about the top level. Their last reference is the national level, the government. I mentioned the hypocrisy of what westerners do and she said yes, some of the foreigners here in Benin know that their policies cause poverty and trafficking. Many of them even want to change it, but they can 't. 40 In this policy world, then, discussion of the forces that create poverty is off-limits. As per the neoliberal order of discourse, pre-existing distributions of wealth and the relations of power that underpin them are taken not only as given but beyond questioning, with powerful actors ensuring that this remains the case.
The same is also true with the discursive order of western childhood. Though, as explained above, Benin's anti-trafficking legal framework demands the blanket abolition of work defined as exploitative, some policy makers are aware that this is impractical and accept that what constitutes exploitation will vary from place to place. Some therefore advocate the adoption of a regulatory as opposed to an abolitionist framework, based on the recognition that work can be positive for many children and that abolition often fails. 41 Despite this, however, when I asked senior figures from the anti-child labour bodies fighting trafficking in Benin why they did not pursue a regulatory instead of an abolitionist line, I was told flatly that this was impossible because it did not fit with institutional discursive norms. What if children have no alternative but to migrate for work? I asked Handel, who had global responsibility for his organisation's anti-trafficking programme. 'It doesn't matter,' he responded, 'we cannot finance anything that disagrees with our normative framework. We have to be consistent with our positions even if children are working. If not, in twenty years, where will we be?' 42 40 Interview with Sandra, 17/2/2010.

41
M Bourdillon, et al., 'Re-assessing Minimum-age Standards for Children's Work' International Journal of Sociology and Social Policy, vol. 29, no. 3/4, 2009, pp. 106-117. 42 Interview with Handel, 21/5/2009. DOI: 10.14197/atr.201213 The need for representational harmony Handel's rigidity points to a further issue of major importance. In this policy world, great emphasis is placed on the importance of representational harmony, such that policy and its justifying narratives are depicted as accurate, appropriate and coherent even where they are not. Thus Handel admitted to editing reports so that they fit with institutional discourse and organisational responses to poverty, portrayed as optimal even where they palpably are not. Why is this so?
Within the anthropological study of policy making and the ethnographic examination of international aid chains, authors such as David Mosse have argued that, in a top-down neoliberal policy world where "efficiency" is the major criterion for the spending of public money, it is crucial for actors to be able to represent what they do as a wellplanned and well-executed success. This leads to a reluctance on the part of institutions to publicly question their pre-existing narratives, and results in staff-members moulding their representations of reality to have them conform to those narratives. It further results in bodies lower down the disbursement chain having to contort their representations in order to have them fit the terms defined by those above, while those above enforce representational stability so as not to lose the symbolic capital that they are able to translate into government funding. 43 In the case of trafficking in Benin, this can be illustrated with two examples. The first is the Etireno affair, which sparked the explosion of interest in trafficking in the first place. Abidi, a UN employee working in Cotonou at the time, explained that 'when the Etireno happened, it was a huge thing for us and was amazing in terms of the mobilisation of resources'. 44 Ayala, the former Beninese government minister responsible for children at the time of the crisis, concurred: 'Some bodies used this for their own ends, to attract funding and attention.' 45 As Alexia, an international NGO operative working in Benin, explained: 'In child protection, you have to be fashionable to attract funding.' 43 D Mosse, 'Is Good Policy Unimplementable? Reflections on the ethnography of aid policy and practice' Development and Change, vol. 35, no. 4, 2004, pp. 639-671, p. 651. 44 Interview with Abidi, 12/1/2010.

45
Interview with Ayala, Cotonou, 13/3/2010. DOI: 10.14197/atr.201213 Fundamentally, this means eschewing the political economy of poverty or the reality that sometimes children must take sub-optimal jobs for want of better alternatives, and instead promoting the narrative of innocent and enslaved children in need of rescue through individualised policy responses.
The second example concerns government behaviour. Jeremia, another UN employee with an intricate knowledge of the various political figures working in Benin, explained that the heavily anti-movement antitrafficking law adopted by Benin's government was, in large part, the result of major pressure from a US embassy determined to see laws on the books as a sign of success in its push on the anti-terror and security agenda. Though many civil society representatives privately opposed the draconian nature of the law, arguing that it was un-enforceable given the mobility of the populace, US representatives were unmoved, and the Beninese government accepted that it needed to pass the law in order to ensure that the US would continue offering its bilateral assistance. 46 Beyond money, the Beninese government also derived symbolic, representational capital from the move. As Benin's senior civil servant working on trafficking explained: The law allowed us to move from Tier 3 to Tier 2 in the US's ranking of countries [in terms of the quality of their antitrafficking response] and it gave us a good image in Geneva, with the Commission on the Rights of the Child, and also with the donors. 47 By thus playing the game according to the rules of the discursive order established by those who sit above them in the hierarchical power and disbursement chain, actors lower down the pecking order are able to create and accumulate the symbolic capital which is subsequently cashed for the material capital that ensures their very institutional survival.

Conclusion
When reflecting on accountability within anti-trafficking policy making as it operates in Benin, it is clear that those above do not answer to those below. Indeed, accountability does not operate optimally from policy maker to policy beneficiary. Rather, policy and its justifying narratives are accountable more to the institutional and political logics, imperatives and orders of discourse within which policy makers exist. This is not to suggest that Machiavellian explanations are sufficient; far from it. Policy makers frequently recognise the drawbacks of what they do-they are merely constrained by the system and those who pull its most influential levers from offering any meaningful alternative.
Neil Howard is a doctoral student at the University of Oxford. He has been researching and working on human trafficking since 2005. Email: neil.howard@qeh.ox.ac.uk