HOW GENDERED IS THE EUROPEAN MIGRATION REGIME? A Feminist Analysis of the Anti-Trafficking Apparatus

51 Unaccustomed Alliances: Round Tables on the Trafficking of Women I am standing in front of a five-star hotel at the outskirts of Istanbul, only a few minutes from the airport, waiting for my lift. Once again I had endured a two-day political spectacle together with delegates from governments, secret service organizations, the European Union, international and supranational organizations such as Europol, the Organization for Security and Co-operation in Europe (OSCE), and the International Organization for Migration together with national and local non-government organizations (NGOs). For the past six months I had been doing ethnographic fieldwork, applying a multi-sited approach, on policy development within the so-called Budapest Process – a series of ongoing conferences and workshops begun in 1991 on the implementation of EU migration and border policies in areas outside of EU jurisdiction – and held the role of “independent researcher”, that peculiar person who somehow belonged at the far end of the table (see Hess 2009, 2010). The Budapest Process and the ICMPD, the International Centre for Migration Policy Development located in Vienna, which acted as Secretariat for the Process, had caught my attention many years earlier as a key political institution in the area of migration and border management. The ICMPD informs EU Member States, especially new Member States, about policies in these areas and also targeted adjacent countries through workshops and “capacity building programmes” (see Düvell 2002; Hess 2009; Georgi 2007; Geiger & Pécoud 2010). Based on a two-year ethnographic research project on the making of European migration policy, this article explores the ways in which gender is deeply inscribed in the articulations, practices, and rationalities of the new European migration regime. It focuses on the area of “anti-trafficking” policies at national and transnational levels, showing how and why an “anti-trafficking dispositif” has been created over the last twenty years. Anti-trafficking policy, which targets women in particular, has become one of the main pillars of a restrictive, Europeanized migration and border regime. The article offers theoretical and methodological approaches to this gendering of migration policy, and asks what such a co-optation of feminist discourses and practices means for reflexive feminist cultural theory, research, and practice.

Hosted by Turkey, a broad cross-section of very different organizations -including the Budapest Process, the ICMPD, the United Nation Office of Drugs and Crime (UNODC) and the Black Sea Economic Cooperation Organization (BSEC) -had extended the invitation to the Istanbul conference. The issue at hand was "Trafficking in the Black Sea Region" (UN. GIFT 2007). Surprisingly, despite their diverse interests, the participating institutions had reached a formulaic compromise to rally around the issue of "anti-trafficking", which meant that in the end funds would flow in this direction. Also surprising were the two foci on which participating actors had been able to agree after numerous rounds of negotiation: 1) "transnational cooperation between law enforcement and NGOs for the referral of victims of trafficking" and 2) "data collection and information management" (emphasis by the author).
In my eyes, this conference provided an extraordinary example of the far-reaching impacts and effects of the anti-trafficking discourse (re)produced for years by the EU, the USA, and many international intergovernmental organizations -as well as by regional and local NGOs. At this conference, interagency alliances could be observed that had previously been considered impossible in this region. Concurrently, spaces of possibility had also been opened to an unforeseen extent for formerly marginalized groups and positions, and for the discourses of non-government organizations. For example, estranged but nonetheless at the same table sat the chief of the Jandarma, Turkey's paramilitary law enforcement unit, and representatives of the country's two feminist battered women's shelters. They discussed the limits and difficulties of their therapeutic social work, whereby all NGO representatives repeatedly identified restrictive migration management policies as a major cause of the problem. The OSCE delegate also held an impassioned speech and, after distancing herself from the term "illegal migration" in a very differentiated manner, made clear that restrictive immigration policies and tightened border controls pushed more and more women into the hands of so-called human traffickers 3 in their efforts to cross the borders as "irregular migrants ". 4 After completing the departure process, once on the plane I wrote in my fieldbook: "The grotesqueries born out of European governance policy are incredible!" As an example: with me in the Jandarma's VW van being driven to the airport were two battered women's shelter workers, clearly anxious and fidgeting nervously in their seats. When we got out I asked them about the cause of their discomfort, and they answered that such vans usually brought them to another destination -namely into police custody after participating in demonstrations. In the context of anti-trafficking policy, these women are now committed members of civil society and not only sit with, but also co-operate with the very same law enforcement institutions. They are unequal actors in the field of anti-trafficking policy, although they understand full well the negative impact of restrictive border policies. Increasingly these alliances, although entry into them may be merely strategic, piqued my interest. In the course of my research on European migration and border policy (2004)(2005)(2006)(2007)(2008)(2009), I ran into them on the community, regional, and European level -in particular in the context of antitrafficking policy. I wanted to better understand these alliances and take a closer look at this style of setting government policy. I was further interested in the role and function of anti-trafficking policy for the European border regime from a gender theoretical and cultural anthropological perspective. My ethnographic and discourse analysis research therefore focuses on the European context.
The key focal point of this research interest is the international policy field of anti-or counter-trafficking in human beings -especially in women -in the main because it is this migration policy terrain that "works" with discursive positions situated within feminism and women's rights. It is a field which over the past fifteen years has stepped out of the political niche of the international women's movement and become a dominant apparatus within European and globalized migration management. Not only does it legitimize the discourse of tightening migration control, it provides a foundation for this discourse and has a massive public impact. For example, "MTV" and another music TV channel "Viva" take part in international campaigns against human trafficking and trafficking in women, and international celebrities such as Angelina Jolie act as international extraordinary ambassadors (MTV 2012; http://news.change.org/stories/latest-big-names-inhuman-trafficking-fight-liu-streep-jolie). 5 This discourse has also freed very large sums of money for the international fight against organized crime and produced new coalitions of actors such as the one outlined above in which -under the aegis of governments -members of civil society, feminist NGOs, international organizations such as the omnipresent International Organization for Migration (IOM), and national and supranational law enforcement and secret service agencies gather around the same table. Even if the IOM and the International Labour Organization (ILO) have made attempts in the past few years to expand the discourse to include all forms of the exploitation of labour, at the centre of the anti-trafficking apparatus are still the images and discourses of trafficking in women and forced prostitution. Like almost no other international policy area, it focuses on the female body and female sexuality -in particular on the topos of the helpless female "victim". 6 For this reason, an analysis of the anti-trafficking apparatus can also reveal how deeply the category of gender is inscribed in the procedures, technologies, articulations, and rationales of modern European migration policy.
Based on Michel Foucault's ideas, I speak of the "anti-trafficking apparatus" (dispositif). 7 Foucault, in Reiner Keller's estimation, introduced the concept of the apparatus following his discourse theory work in order to connect studies of the said with studies of the unsaid: of materials, infrastructures, and institutional actions (Keller 2008: 93ff.). In this vein, I am interested in an analysis of the linkages of policies and discourses, especially of their rationales and their impact in terms of power and discourses of truth. I am also concerned with the function and relevancy of these linkages in relation to European migration policy. To this end, I look at the way in which the apparatus of "anti-trafficking policy" plays a central role in policies governing tightening border controls and the externalization of the European border regime. It shall become clear that the current turn towards "governance" within the migration regime is also built upon these discourses of victimization and uses discursive positions of the human rights' and women's movements in its argumentation (see also Hess & Karakayali 2007). I shall also show how new governmental political practices and discourses are connected to this apparatus -such as the politics of round-table meetings -with very ambivalent effects on, among other things, feminist political practices. I am particularly interested in understanding how discourses of women's rights -in particular positions against violence against women -came to fit so well with the restrictive rationalities of the border regime. This leads me to ask in the final section what it means when feminist theory and practice act on the terrain of migration and border policy. I also try to pinpoint those discursive positions -including research on women and migration -which help(ed) produce this adaptability. 8 In the following I draw mostly from the interdisciplinary approach of Anglophone queer and genderaware research on immigration and border management such as the work of Bridget Anderson (2009) and Rutvica Andrijašević (2009Andrijašević ( , 2010, as well as from Foucauldian and post-colonial debates on the refashioning of Europe. I also draw from the results of my own research on European migration and border policy such as the work mentioned above (Hess 2009(Hess , 2011, the interdisciplinary research project Transit Migration (2003)(2004)(2005), and research conducted within the Network of Critical Migration and Border Regime Research (Hess & Kasparek 2010). Within the framework of the collaborative Transit Migration research project 9 , we conducted ethnographic research at the southern and eastern borders of the EU on the Europeanization of migration policy. Building on this research, we developed what we called an "ethnographic border regime analysis" (Hess & Tsianos 2007). This ethnographic border regime analysis connects more recent approaches to ethnographic field research -in particular variants of multi-sited ethnography (Marcus 1998) -with genealogical and discourse analysis approaches (Hess & Tsianos 2010). In terms of content and theory, it attempts to look at the developments of migration policy from the perspective of migrants' movements and also to develop a form of praxeological policy research capable of analysing "doing border" (see Hess, Karakayali & Tsianos 2009;Ferguson & Gupta 2005;Shore & Wright 1997). In our Transit Migration fieldwork, we identified the discourse of asylum and the discourse of anti-trafficking -and the actors and practices associated with them -as the central discursive and legitimizing pillars of European migration and border management policy. Eva Bahl and Marina Ginal, in their ethnographic research on the local adaptation of anti-trafficking policy in Munich, Germany, were able to show how this discourse has a strong normative effect on what could be expressed and what was unspeakable. For example, it makes it almost impossible to talk about migrant sex work unless it can be subsumed under the topos of sex trafficking and forced prostitution (see Bahl, Ginal & Hess 2010). 10 However, before going into the interplay of the Europeanization of immigration policy and the discourse on trafficking in women, I would like to first outline the genesis of the anti-trafficking apparatus. Below I sketch its varying disciplinary components and dominant dichotomous discursive strategy in an attempt to identify the factors leading to the "boom" in this sector.

Genesis of the Anti-Trafficking Apparatus
There are very few other political discourses and arenas of praxis in which the female migrant body is so explicitly the central object not only of numerous discourses and fears, but also of exploitative relations, of violence, and of a package of quite contradictory countermeasures. It is at the centre of feminist anti-violence politics advocated by activists who have successfully devoted decades to creating a public scandal of the issues of forced prostitution and sex trafficking of women and to promoting feminist protective measures developed to provide concrete support and care. These support practices run the gamut from charitable church measures to radical feminist and anti-racist approaches, from abolitionist to empowerment positions (see Andrijašević 2010: 2, 14f.;Karakayali 2008), and are clearly divisible according to whether migrant women are given the status of objects or subjects/active agents (see, e.g., Lindqvist 2007).
Historical research on the genesis of the discourse of human trafficking such as Serhat Karakayali's work on the politics of irregular migration (2008) or Eval Bahl, Marina Ginal, Bernd Kasparek's joint work on urban migration policy discourses (2009) reveal through their reconstructions a series of thematic shifts and, in Germany, three clear time periods: In the first period at the beginning of the twentieth century, as mass immigration into and out of Germany was the order of the day, the discourse was widespread and also focused in the main on immigrant women. Warnings about the danger of forced prostitution were published in leaflets and on posters. During this period for example, the Bahn hofsmission (church travellers' aid association) was founded for the protection of young women travelling alone (see Hess 2005). This discourse again became virulent in the 1960s and 1970s at the height of labour migration in the context of the so-called guest worker (Gastarbeiter) system. However, in this period, neither the immigration of women nor the sex industry was at the centre of the narrativization. Rather the conditions of self-organized and irregular labour migration were criticized. Working conditions in factories and on construction sites were termed "slavery" and the exploitative practices of labour recruitment agencies were condemned as "human trafficking". Serhat Karakayali points out that at this time too, left-wing poeple -advocated in this case mostly by trade unions -portrayed the situation as a scandal in order to give the discourse more impact (2008: 235). Anti-trafficking discourse died down again in the 1980s as Germany presented itself as a "nonimmigration country" and reappeared in the 1990s as a discourse about women as debates increased on the increase of immigration, in particular from Eastern Europe and the global South.
The renewed interest in this issue by policy makers in the 1990s was the fruit of decades of campaigning by feminist and women's groups to have sex traffick- ing taken seriously. The struggle against the trafficking of women and children and the struggle against forced prostitution, closely linked on the discursive level, were placed on the political agenda in particular because of the focus on these issues at the UN human rights conference in Vienna in 1993 and the UN World Conference on Women in Beijing in 1995 (Schwenken 2006;Düvell 2002). On the EU level, the political scientist Helen Schwenken (2006: 111) has underscored the importance of what she termed the "velvet triangle" of so-called femocrats (feminist bureaucrats), scientists, and activists/lobbyists in establishing this discourse in the 1990s. The European Union, after receiving the mandate to become active in the areas of justice and home affairs and legal issue in the treaties of Maastricht (1992) and Amsterdam (1995), worked closely together with NGOs and other lobby groups and profited from their expertise and knowledge. These new EU structures gave political women's networks in particular a chance to establish a hegemony for their positions -mostly through the Women's Office (later the Equal Opportunity Unit) -especially if these positions were presented as gender equality issues. The issue of trafficking in women profited greatly from this window of opportunity and feminist groups were increasingly able to put it on the political agenda during this period. The first EU action programme against trafficking in human beings was the 1996 STOP programme (Commission of the European Union 1996). In this programme, the EU stresses that it brings together a broad spectrum of actors including "universities, NGOs, police and immigration services, government and parliaments" (ibid.). The next communication about the programme also underscored the broad spectrum and heterogeneity of the actors with whom the EU commission was able to work in this area, naming in particular cooperation with "nonmember countries and specialised organisations" (Commission of the European Union 1998). This communication refers to other action programmes and forums within which the issue had been dealt with in the previous years: The fight against trafficking in human beings has also been tackled in the United Nations (special Protocol to the International Convention against Organised Transnational Crime), the G8 (action plan), the Council of Europe, the International Organisation for Migrants (IOM) (regional surveys and information campaigns), Interpol, the Organisation for Security and Cooperation in Europe (OSCE) and in the context of the transatlantic dialogue. (Ibid.) A further EU action programme, Daphne, ensued in 2000 (Commission of the European Union 2000). Daphne provided funding for NGOs that offered support to women and children who had been victims of trafficking. And two years later, in 2002, the Council of the European Union adopted a framework decision to establish a common legal framework on human trafficking within the Member States (Council of the European Union 2002). 11 Daphne I and II were extremely instrumental in strengthening the anti-trafficking discourse and the relevant actors. Where no NGOs had previously existed, they were created and, as our research within the Transit Migration project showed for the example of Turkey and the Balkan states, influence was exerted on the political programmes of existing NGOs. In an interview with the Berlin anti-racism organization Anti-rassismusBüro, Marion Böker from the Federal Association against Trafficking in Women and Violence against Women in the Migration Process (KOK) criticized the fact that the EU's grant programmes contributed greatly to splits in the NGO community between those initiatives which were willing to work together with the EU on victim protection measures and round table politics and those that rejected cooperation or collaboration with government law enforcement agencies and therefore did not receive financial aid (Böker 2004).
Concretely, within EU migration and border policies, two arenas of political practice emerged in the area of anti-trafficking policies. This can be seen in other areas of the world as well, for example in the so-called Bali Process in the Asia-Pacific region. Anti-trafficking politics thus act in an arena of high- profile politics of representation, and establish a very specific regime of the gaze of the female body. Concurrently, especially in the area of security, these policies are enacted as a part of the overall increasing efforts to combat irregular migration and organized crime. We should first examine the politics of visualization which create a racialized and sexualized regime of the gaze and has a negative impact in particular in female migrants' home countries, because it places the migration of women in general under the moral misgivings surrounding prostitution and stigmatizes women's migration as, among other things, a risk which cannot be taken (see Hess 2005). This takes place mostly within so-called awareness-raising or educational campaigns that warn of the dangers of human trafficking on large billboards and other advertising media in the home countries. The International Organisation for Migration (IOM), one of the few global all-around agencies in the field of migration control, is particularly active in this area (Düvell 2002;Georgi 2010). 12 Rutvica Andrijašević conducted an analysis of the images used by the IOM in their 2001-2002 campaign in the Baltic States. As in other IOM campaigns, this Baltic campaign made use of images of usually naked female bodies posed as victims. One poster, for example, portrays a halfnaked female body hanging in the air on hooks and ropes. These images are meant to warn potential migrants of the dangers of migration and prostitution (Andrijašević 2005). IOM's Head of Mass Information put it as follows: The nakedness was meant to show the helplessness and vulnerability of trafficked women. The hooks are visual metaphors used to convey an essential aspect of trafficking, namely the manipulation and exploitation to which trafficked women are subjected (…). Most trafficked women find themselves treated as slaves with no control of their lives whatsoever. This is the idea we wanted to convey. (Cited in ibid.: 35-37) Stressing the dangers of migration, in Rutvica Andrijašević's analysis, is meant to keep women away from informal labour migration and "implies that the safest option is to remain home" (ibid.: 31). Furthermore, Andrijašević shows that the male voyeuristic gaze is inherent to these images, constructed as they are of women's suffering and victimized bodies. This degrades them to objects of voyeuristic eroticization, thereby reproducing the stereotype of Eastern European women as "beautiful victims". In light of this, Andrijašević shows that the campaign is not about, but rather against, women and women's migration and transports the message: stay at home! Global anti-trafficking policy, alongside large public relation campaigns such as this one -campaigns in which Viva and MTV also sometimes participate -has otherwise been rooted mostly in the terrain of security and migration management policy. Combating sex trafficking takes place mostly in connection with the fight against organized crime, thus correlating the two. The UN protocol, considered the central document of anti-trafficking policy, illustrates the way in which sex trafficking was made into a security issue. The protocol was celebrated -also by feminist activists -as a milestone in bringing the issue onto the official European and global political agenda. The so-called Palermo Protocol was adopted by the United Nations Office on Drugs and Crime in 2003 as a supplement to the broader Convention against Transnational Organized Crime (UNODC 2004). It is thus primarily a political instrument to improve international cooperation and develop a new international legal regime in the fight against organized crime. The protocol differentiates between trafficking (forced trade) and smuggling (voluntary migration). On the one hand this is laudable, because it allows for differentiated interpretations -not all transport and information services in this area should be automatically defined as "trafficking". On the other hand, the Convention strengthens and formalizes the central dichotomy between voluntary migration and force which characterizes this discourse. Christina Hahn conducted a reconstructive discourse analysis of the positions of various NGO-affiliated lobby groups in this area. She showed how abolitionist positions increasingly gained hegemony that treated women migrants, es- pecially sex workers, as defenceless victims without a voice (Hahn 2007). However, biographically-oriented ethnographic research in this area has shown that the reality of these women is different. In their daily lives, voluntary actions and direct and structural violence intersect in contradictory manners and are judged and negotiated in myriad ways, both in migrants' interpretations and in their actions. Ramona Lenz for example, in her study of mostly Filipina and Eastern European sex workers in Cyprus, shows how these women arrived in the country on their own with an official artist visa and only later, as a result of having no rights, drifted into relations characterized by dependency and coercion (Lenz 2009). On the other hand, women interviewed by Rutvica Andrijašević for her study about Eastern European sex workers in Italy, describe how they chose to work as prostitutes and themselves turned to trafficking networks and at the same time say they were forced. Nevertheless, the women interviewed in this study clearly developed strategies to deal with dependency and coercion (see Andrijašević 2010).

Boom Sector Sex Trafficking: Migration Policy as a Catalyst
Numerous studies have shown that over the past years, demand has increased for sexualized and racialized women's bodies and emotion-laden sexualized work in the context of globalization and economic restructuring (see Andrijašević 2010: 5ff.;Hochschild 1983). Feminist scholars search for the causes in the heteronormative, patriarchal, capitalist, and racialized world order. Sex trafficking, the trade of women and children for the purpose of sexual exploitation, is now, according to international law enforcement agencies such as Europol, the fastest growing criminal sector. After the drug and arms trade it is also, according to some, the third most profitable sector (see for example Andrijašević 2010: 7). However, since statistics are problematic in this area, 13 because of the informal character of the phenomenon and the different statistical methods used by different authorities, the authors of the study "Human Trafficking in Germany" cite Europol as follows: Europol assumes that "hundreds of thousands" of victims are trafficked into the EU Member States each year for the purpose of sexual exploitation and labour exploitation and notes that the number of victims has increased considerably over the past few years. (Cited in Folmer & Rabe 2009: 20) Feminist studies with a biographical and ethnological orientation have pointed out that stereotypes of the impoverished woman from the global South and East together with increased male demand in the North create a push-pull model that does not do justice to the complexity of this field, which is not only economic, but also social, cultural, and political. These studies show that subjective, emotional, and ethical aspects such as the women's own ideas about sexuality and a better life as well as their rationalizations must be given more importance in order to place more weight on the agency of the migrant women themselves (see for example Andrijašević 2010: 12;Hess 2005).
Andrijašević also shows how the spread of the sector is connected to the general increase in underpaid labour in deregulated and informal sectors following the flexibilization of labour policies in the Western and Northern countries. This went hand in hand with an increased demand for an inexpensive and flexible female migrant labour force, leading to the feminization of migration which can be observed in many southern and eastern countries (see for example Anthias & Lazaridis 2000). However, for most of these migrants, because of restrictive immigration and labour policies in most EU countries -and in connection with racialized images of (still) female (domestic) workers -the only field open to them is the large field of caretaking; including care for the sick and elderly, cleaning and ironing, domestic and childcare services, and also sex work (see Lutz 2007). This labour segment of the three Cs (cooking, cleaning, caring) 14 is very gender-specific. It often falls under the "private" sphere and is thus not subject to public scrutiny or formalized labour contracts (ibid.). Ethnographic and biographical research has shown that this is also a factor which migrants often use to their advantage (see Hess 2005;Lutz 2007 Alongside these subjective biographical and political-economic factors, researchers also point out a genuinely political regulatory factor which has led to the increase in the nexus of migration-trafficking-sex work: official European immigration policy itself. Without this regulatory influence, which to this day can only be described as restrictive and gendered, the phenomenon of sex trafficking could not possibly be as widespread as it is today. For example, most countries allow very little official immigration and immigration laws reflect a view of migration as a male undertaking focused on male labour (see Hess 2005). Feminist and queer research on migration has been able to demonstrate the way in which immigration policy is based on heteronormative concepts of gender and sexuality, in particular that of the dependent wife who is not herself actively seeking migrant status. Policies geared towards the reunification of families or regulations granting the right of residence as a result of marriage -two ways in which women often gain legal immigrant status -reproduce heterosexual patterns and female dependency (Kofmann & Sales 1998;Andrijašević 2009: 394). Because of these immigration and labour market policies, women migrants are overrepresented in precarious and informal migration and labour sectors. In the end, policies that grant only few residency and work permits -which are moreover temporary and uncertain -create marginalized groups of migrants without any meaningful access to the formal labour market and without recourse to law. Within this logical framework, "illegal" immigrants are not an abnormality or an outgrowth of the system, but rather a constructed, regulatory category of immigration.
In order to further this restrictive migration policy, which creates hierarchies and puts migrants in precarious situations, over the past two decades European border management has been harmonized and continually intensified using diverse (military) apparatuses and practices of knowledge and technology. Images of drowned immigrants in the Strait of Gibraltar and the Aegean and Libyan Sea or of military patrol boats belonging to the European border control agency Frontex are only the apex of a technological/military apparatus that, in the name of "migration management", works towards the selection of immigrants to maximize benefits (see Ghosh 1997; for a critical analysis see also Mezzadra 2009;Geiger & Pécoud 2010;Hess & Kasparek 2010). In sum, it can be said that it is European immigration and border management policies themselves -as has been acknowledged at EU conferences as well -which have helped further the phenomenon of sex trafficking, because migrants are dependent upon the services of trafficking and smuggler organizations in order to cross the border at all. When migrants do manage to cross the external borders, 15 immigration policy paired with nationalist, racist labour market policies 16 and heterosexist gender policies push women towards informal (sex) work sectors and produce dependencies and disenfranchisement.

The European Border Regime and the Role of Anti-Trafficking Policies
In the public mind, the Europeanization of migration policy is closely connected to the name of a small village in Luxembourg. In Schengen, five Member States of what was then the European Community, including Germany, met in 1985 on a multi-lateral level and outside of EU structures to establish a common market. As a so-called compensatory measure, they agreed to shift border controls to the outer borders and to harmonize policies regarding refugees, foreigners, and visa practices. But it took over ten years before the Schengen Agreement became official EU policy in the 1998 Treaty of Amsterdam (see Hess & Tsianos 2007;Hess 2010). 17 Although the harmonization of EU migration policies is a sluggish process, harmonization and consensual immigration control practices can be observed, known in EU jargon as "integrated border management" (see Geiger & Pécoud 2010).
A central characteristic of this policy is the externalization and internalization of borders. "Borders" have now been placed far beyond the actual borders of the EU and have also been defined within EU territory, producing flexible, fragmented border areas and border corridors which are enacted in a highly . This approach no longer focuses on the crossing of national borderlines, rather it is the migratory movement itself -in particular the routes taken by migrants -that is at the centre of these control policies. This leads to a massive imperial extension of European immigration policy into migrants' home countries. In this context, the discourse on smugglers and human traffickers takes on a key operative function that is not limited to exposing and shutting down smuggling rings and human trafficking routes. This act of putting immigrants under general suspicion, or rather the linkage of migration and criminality, provides regulatory bodies with the legitimation they need to detain all migrants and interrogate them -using violence at times -about their routes. Pro Asyl has documented this practice in Greece (Pro Asyl 2008). In this way, large data sets are generated and visualizations of migration patterns are produced such as the so-called i-Map, a digital map of migration routes created by Europol, Frontex, the United Nation High Commissioner (UNHCR), the UNODC, and the ICMPD (see i-Map 2012). Even if the harmonization of EU migration policy has faced many setbacks (see Birsl 2005), the Europeanization of migration management has nevertheless developed an internal dynamic. As we were able to show in our research project Transit Migration, this can be traced in the main to the formation and activities of innumerable international, regional, and local actors who both support and flank government policies, and who are also miles ahead of them and are following their own political agendas. This multiplication of actors is known in international border studies as "privatization", and is called the politics of "remote control". European border policy is thus also made outside of government institutions and far from national borders (see Lahav & Guiraudon 2000;Guiraudon 2001). Within Transit Migration, we termed this process "NGOization" in order to make clear that this method of governmental policymaking takes place mostly through and in cooperation with NGOs, and itself takes on elements of an activist repertoire (see Hess & Karakayali 2007).
The two largest intergovernmental organizations in this field are the UNHCR, whose existence goes back to the Geneva Convention on Refugees (see Ratfisch & Scheel 2010), and the above-mentioned IOM. In contrast to the UNHCR, the IOM has no foundation in international law, although it makes great efforts to performatively generate such a foundation. As our ethnographic research in Transit Migration made clear, the exploitation and hegemonization of the anti-trafficking discourse played a key role in the IOM's ability to secure its dominant position in the global arena. Combatting human trafficking is, alongside refugee policy, one of very few areas of migration policy which uses a discourse promising protection and is pushed forward in the main by the USA (see US Department of State 2012), which in turn set the course of the IOM. As we have seen in the example of the Istanbul conference, antitrafficking discourse is also reproduced by further large and small institutional actors in the migration management arena from the OSCE to small NGOs on the local level (see Migmap 2006).

New European Governance -Turning the Border Regime into a Human Rights Issue
This multiplication of key players and inter-agency cooperation correlates directly with changes in the European Commission's political practice as laid out in detail in the White Paper on European Governance (Commission of the European Union 2001). As a response to global challenges, the Commission writes, politics should be "decentralized" and shaped at "various levels" within "networks", include the strategic "involvement of civil society" and the use of "expert advice". The findings of the Transit Migration project on the political practices and discourse of immigration policy as expressed by the EU within the preliminary rounds on the accession of Turkey, show that the politics of the "Round Table" is the incarnation of the politics of governance (see Bahl, Ginal & Hess 2010). On a discursive level, these new coalitions -as I have described using the example of the Istanbul conference -adapt the discursive positions of human rights activists and feminists to a restrictive border regime. William Walters, a researcher of Europeanization and border management, speaks in this context of the "Birth of the Humanitarian Border" which does not necessarily mean that actual practice becomes more humanitarian (Walters 2010). However, local ethnographic research such as Marc Speer's work on the implementation of the EU border regime in the Ukraine (2010) shows that human rights interventions and activists can in fact challenge law enforcement structures. But the price is often high, as players must accede to the logic of a human rights discourse which, among other things, seems to demand the choice of a policy approach that victimizes the migrant other. Eva Bahl and Marina Ginal come to the same conclusions in their study on the effects of the anti-trafficking discourse on negotiation options and positions on the municipal level. They clearly show the way in which this discourse delegitimizes empowering positions and only an abolitionist, victimizing discourse remains as the legitimate manner of speaking about, or having knowledge about, migrants in sex work (see Bahl, Ginal & Hess 2010).
Thus during the Istanbul conference, all key players spoke of a "victim oriented approach". This is a significant shift in the discourse and moves the focus away from law enforcement and concrete practices of control. Rather, within this discursive shift it seems that all institutions of political control are interested in protecting the victims and posit themselves as biased, almost activist players concerned primarily with the well-being of the women involved. Yet the measures taken in the name of victim protection include stopping and identifying trafficked women, taking them into police custody and interrogating them about their migration routes, placing them at women's shelters for the protection of victims, having them appear as court witnesses, and deporting them back to their home countries. The ICMPD, an organization which has played a key role in furthering this discourse, published interviews with women who had undergone these procedures in a study entitled "Listening to Victims" (Surtees 2007). In this study, the ICMPD -which does not have a reputation for doing pro-immigrant work -cites many women who tell of inappropriate and poor conduct by both the police and NGOs who scared them, threatened them, mocked them, and did not explain their situation to them. Many passages also cite migrant women clearly stating that the measures taken in order to protect them were not what they wanted, and that actions which go by the name of victim protection are actually against their interests. The following passage is one example: "It happened for the first time when I heard that I could not have residence papers and must leave from the centre. I fainted and when I was awakened I started to scream… I did not want to leave the centre; I had no place to go" (ibid.: 110). Many quotes in this section describe the women's distress at and rejection of their repatriation. In this context, the work of NGOs, women's shelters, and the IOM is particularly politically charged, since they support the detained women throughout all procedures up until repatriation. They may try to alleviate the migrants' situation by giving them financial start-up grants or arranging contacts to women's shelters in their home countries, but this is often not what the women themselves want. They want to remain as migrants in the new country.

Anti-Trafficking as the Motor of a Restrictive Border Regime
All in all it can be stated that this package of measures, encoded as victim protection, makes the female migrant body the central target of diverse and connected high-tech practices of knowledge within the migration and border regime, up to and including local NGOs and ending in operative mappings of migration routes such as the so-called i-Map. The i-Map is the brainchild of ICMPD, Frontex, and Europol. On the surface it shows a digital map that claims to portray current migration routes based on databases (i-Map 2012). The UNODC and other law enforcement agencies continually create data sets and visualizations of this type in the vain hope of being able to make accurate prognoses about migration movements. Victim protection measures objectify the migrant body; they make it penetrable and computable, subject to categorization and selection. More than anything, they limit the migrants' histories to their status as victims. We imagine "victims" as defenceless women, reduced to their corporeality and no longer able to speak for themselves. They must be protected from themselves -in our Transit Migration research we even learned of a women's NGO in Belgrade which locked women in "for their own protection". Migrant sex workers are only interesting for social work organization as victims. Migrants who voluntarily become prostitutes and are then confronted with a situation in which they are subordinated and exploited often have no recourse to support. Looking at all of this together, we see a convergence between anti-trafficking polices and anti-immigrant policies. As a result, even policies that operate in the name of helping victims in the end act against the interest of women from the European periphery. In sum, the central discursive impact of the anti-trafficking discourse is that it 1) not only legitimizes, but also demands more stringent control of migration to protect possible victims; 2) on a practical level, legitimizes the creation of so-called screening centres which are set up along migration routes in order to filter out "vulnerable" trafficking victims and refugees from the mass of irregular labour migration. As a result, the anti-trafficking discourse furthers a policy of selection, and concomitantly the criminalization and disenfranchisement of most immigrants without papers, who can be categorized neither as trafficking victims nor as refugees. In the end, antitrafficking policy thus plays a key role in the ascendency of this key biopolitical mechanism of hierarchizing and disciplining migrant passages by means of border management policy (see Mezzadra 2009). 18 Finally, the anti-trafficking discourse 3) legitimizes police practices of detainment and thorough interrogation meant to uncover migration routes and trafficking networks, and ending in numerous processes of knowledge acquisition as exemplified by the i-Map. For this reason, Rutvica Andrijašević comes to the conclusion that "trafficking discourse and anti-trafficking policies sustain and normalise a differential regime of mobility through which the EU hierarchically organises access to its labour market and citizenship" (Andrijašević 2010: 4). Researcher and sex worker activist Jo Doezema therefor asked at a 2002 conference on trafficking at the university of Ghent: "As trafficking is increasingly being used by governments and even by NGOs as an excuse for repressive policies, NGOs are left wondering: where did we go wrong?" (2003: 1).

On the Biopolitical Logic of Feminational Discourses 19
Against this background, I would like to end with an examination of how it was possible that wellmeaning feminist theories and practices -such as those surrounding trafficking in women and forced prostitution -were able to cooperate so easily with restrictive migration and border management policies or rather be coopted by the same without greater friction. I draw on the one hand from the work of the feminist migrant anti-violence activist and theorist Esra Edem, who has conducted an analysis of a similar migration issue -integration policies (Erdem 2009: 190, 192). I also draw from the work of the American queer theorist Jasbir Puar, in particular her book Terrorist Assemblages: Homonationalism in Queer Times. Puar observes a similar discursive coalition in the US "war against terror" in which the "terrorist assemblage" integrates discursive positions of the gay and lesbian movement. Puar shows how the normative figure of homosexuality is integrated into a post-liberal, post-colonial construction of the "modern" Occident contrasted with a pre-modern, traditional, and patriarchal Orient. She also speaks of an emerging "homonationalism" 20 which is able to build discursively upon the new-found national unity in the war against terror (see Puar 2007).
Esra Erdem uncovers  German integration policy. She describes the way in which gender equality positions have been used in past years in order to legitimize restrictive integration policies. 21 In this case, a feminist anti-violence against women discourse -expressed by German and immigrant women's rights activists such as Alice Schwarzer or Seran Ates as well as by conservative politicians -is used to suddenly impart the equality of women the status of a Western cultural value (Erdem 2009: 189). Erdem asks what it means for the women's movement (and the theory thereof) that it has become a central actor on the terrain of foreigners' rights and that it encourages a discourse and images which place migrant women predominantly in the context of violence, as passive victims who are incapable of action. Erdem goes on to argue that this feminist-inspired anti-violence policy not only chases after a liberal feminist illusion that jurisdiction and stricter immigration laws could protect women (ibid.: 191), but also that such policies are too reductionist and in the final analysis, racist. They ignore the intersections of other power relations such as policies on immigration and "aliens" or the colonial geographies of the gaze in which this discourse and the speakers' positions are located. Within the context of the debate on integration, as in the context of the US war against terror, the primary topoi, albeit radicalized, are colonial and Orientalist images of the modern Occident and its counterpart, the traditional Orient. From the very beginning of the colonial project, the relation of the sexes, in particular the figure of the subjected wife and the narrative of honour killings, have been used to prove the myth of Western modernity and Eastern (and Southern) traditionalism (see von Braun & Mathes 2007). Chandra Talpade Mohanty (1998) for example has shown that the production of the topos of the oppressed woman of the global South was constitutive for the production of the "emancipated Western woman". Gayatri Chakravorty Spivak argues similarly (1990) when she positions work done by Northern feminists for Southern women within paternalistic mission work. Thus national Western feminist discursive positions have a long tradition of legitimizing both the busi-ness of German colonialism and the project of the nation-state, and of putting itself at the service of these projects. Esra Erdem calls those feminist discursive positions which today are inscribed within restrictive debates on integration "racialized gender equality politics" which support the "codification of a racialized hierarchy (…) established by laws on aliens" (Erdem 2009: 191, 195). She speaks further of the "feminist disciplining of the migrant subject" (ibid.: 194), which found legitimatization and a quasi-scientific source of knowledge in, as I shall show in my final section, women's studies and gender migration studies.

The Discourse of Victimization and Feminist Migration Studies
In contrast to the ubiquitous thesis of the invisibility of migrant women, especially in feminist migration studies, in general migration studies show how migrant women receive specific visibility as an object of knowledge and are central to processes of positing migration in cultural and ethnic categories. An unsystematic genealogical look at the history of the analysis of migrant women demonstrates the way in which certain discourses -in particular the topos of women as victims and the discourse of victimization -have a long tradition in feminist migration studies.
Not only Ernest George Ravenstein noted in the 1885 Journal of the Statistical Society: "females are more migratory than males" (Ravenstein 1885: 196). In The Polish Peasant in Europe and America, a work that is influential in the social sciences to this day, the Chicago sociologists W.I. Thomas and Florian Znaniecki also include the experiences of women immigrants. However, Helma Lutz was able to show that, in contrast to their portraits of male immigration, portrayed as actively breaking out of the past, women are described quite differently in The Polish Peasant. As Lutz says: "For women, they write, migration is a calamity" (Lutz 2008). The topos was born of the female migrant as a "victim" dependent upon her family, a topos which -Lutz claimsmade an impressive comeback in labour migration research in the 1970s. One of the pioneers of research on women migrants, Mirjana Morokvasic, noted as In important works on migration, the symbolic reference to women as migrants' wives and their stereotypical presentation as wives and mothers has led to a conceptualization of migrant women as followers, dependents, unproductive persons, isolated, illiterate and ignorant. (Morokvasic 1984: 16) In German research, Umut Erel and Eleonore Kofman have shown that the dominant topoi are the "passive wife and mother" and the "victim of patriarchal honour codes" (see Erel & Kofman 2003). From the very beginning of research on labour migration, these topoi, and the concomitant images of the "defenceless female body" and "women as carrier of culture" played an epistemological role in the general theoretical concept of "migration". These topoi were taken up in processes of cultural and ethnic identification within the social sciences and used to link labour migration with backwardness and patriarchal structures. This early image of women's migration has a further discursive effect which influences academic and everyday thinking to this day: it implies coercion and dependency and ignores subjective aspects of women's plans for migration -their hopes and desires. The research on women's migration which began in the 1980s, initially meant as a supplement to academic research on migration, helped to continue this ambivalent pictorial history. In particular in the addition of the issues of (forced) prostitution and migration via marriage (see for example Hummel 1993;Schöning-Kalender 1989), this research sector took up the metaphor of the passive victim until biographical studies began to bring the women themselves and their narratives to the centre of study (Karrer & Turtschi 1996).
This victimizing approach was however continually criticized by feminist migrants such as, for example, Sedef Gümen (1996) beginning in the 1980s. Meanwhile, refutation of victimization approaches is standard anti-racist, feminist rhetoric. Neverthe-less, this approach, as Helen Schwenken has shown in her dissertation on strategies for the mobilization of domestic workers on an EU level, it is still considered as a successful means of framing political demands. Thus some types of institutionalized feminism still consciously work with the topos of victimization in their political actions, even if the migrants themselves, as Ramona Lenz has shown in her research on Cyprus, need to be convinced of their victim status. One Cypriote women's right activist even uses these very words: "We have to convince them that they are victims" (Lenz 2009). Migrant women themselves also, if not always intentionally, make use of an image of themselves as victims, because they know that the apparatus of victimization is one of their few options within the Western, "humanist" sphere. Paradoxically, sometimes claiming the status of victims and framing themselves as victims is the only way to become a subject and develop a position as a protagonist.
Given the way in which the feminist discourse and practices I have looked at in this essay are almost tailored to fit the racist discourse Philomena Essed has termed "Europism" (1995), and their increasing visibility within and co-optation for racialized policies of regulation, I would like to end by concurring with Esra Erdem's final thoughts. She has identified the recognition of the "struggles of migrants and of feminism" and building upon the same to create a progressive vision for a society open to migration as one of the central challenges for the German women's movement and feminist research. In cultural anthropological research this would mean rethinking the knee-jerk association of migrant women with misery and suffering, because, as we have seen, despite researchers' best intentions these images have a clear function in a hegemonical, restrictive border regime. Rather we should begin anew and ask which knowledge and which images would allow a critical analytical reflection on this policy.

Notes
1 The Budapest Process is a so-called regional consultation process or "informal dialogue" in which mostly eastern and southern European countries -and more recently also Caucasian countries -bring international ). 3 In the area of women's migration, and occasionally also in the area of undocumented labour migration, the terminology used most is "forced" trafficking in humans; in the area of refugees, there are "smugglers" and "labour brokers", terms that clearly do not imply coercion and trickery (see Karakayali 2008). However, within the public discourse on refugee and asylum policy, investigative practices, technologies, and expert knowledge aimed at reconstructing travel itineraries are also common in dealings with refugees in order to judge applicants' "right to asylum" and whether they are telling the "truth". 4 The term "irregular migration" seems to have gained ascendency in the European political arena as the "politically correct" phrase. However, in research on migration in the cultural and social sciences, a more thoroughgoing process-oriented debate about terminology is taking place which makes it clear that the situations of un(der)documented migrants -that is those without identification documentation and legal status -are in flux. Most common is a continuum of legal and "illegalized" statuses which can take different forms in terms of irregularities in residency and working permits. Migrants themselves deal flexibly and tactically with these statuses (see for example Vasta 2008;Karakayali 2008;Hess 2005). 5 This entrance of anti-trafficking policy into popular culture can be found in other forms of pop culture as well, in particular TV cop shows. 6 To avoid any misunderstandings, I would like to stress that trafficking in human beings and the resultant slave labour and forced prostitution are abominable and must be combated. This is an informal economic sector, often criminal, that is quickly growing around the world. However, legal action against this sector would be quite possible under existing legal frameworks and without migration management policies. 7 According to Reiner Keller, Foucault uses the term "dispositif " to refer to "an infrastructural apparatus made up of varying elements" and meant "to serve a particu-lar purpose" (see Keller 2008: 93), whereas Foucault himself defined it as the "system of relations that can be established between these elements" (Foucault 1980: 194). On the other hand Foucault understood the dispositif as a "formation which has as its major function at a given historical moment that of responding to an urgent need. The apparatus thus has a dominant strategic function" (ibid.: 195, emphasis in the original). These are the meanings which I have ascribed to the dispositive/apparatus. 8 It is this discussion of the (susceptibility to) co-optation exhibited by feminist theory and practice as well as the search for a progressive feminist politics of knowledge -leaning on Donna Harraway's postulate of "situated knowledge" -which positions this study in a continued process of feminist query as to how feminist theory and practice today can be articulated in an intersectional manner, critical of power and hierarchies (see Hess, Langreiter & Timm 2011). 9 Members of the research group were Serhat Karakayali, Vassilis Tsianos, Manuela Bojadzijev, Rutvica Andrijašević und Efthimia Panagiotidis. 10 I am indebted to this research project, one element of the research and exhibition project I directed in 2009, Crossing Munich, for many of the ideas in this essay on the negative impact of this discourse even at the municipal level. 11 "It is necessary that the serious criminal offence of trafficking in human beings be addressed not only through individual action by each Member State, but by a comprehensive approach in which the definition of constituent elements of criminal law common to all Member States, including effective, proportionate and dissuasive sanctions, forms an integral part" (preamble, paragraph 7). 12 I call the IOM an "all-around agency" since the IOM is involved in almost all facets of migration policy and practice from voluntary repatriation to public relation campaigns, disaster response, capacity-building measures such as building immigrant detention centres, and restitution payments for former slave labourers (see Georgi 2010). 13 Folmer and Rabe write: "The second report by the UN Office on Drugs and Crime (UNODC) compiles world data on human trafficking from criminal prosecution authorities, but also states that more can be learned from this about the activities of the criminal prosecution authorities than about the real numbers of crimes and victims " (2009: 20). 14 Analogous to the three Cs are the three Ds of men's migration: dirty, dangerous, and dull (see Favell 2009). 15 Alongside border controls outside national borders, there are now also various "borders" which have been drawn within countries as a result of the Schengen Agreement. For example, train stations and expressways are now considered "border areas" in which border police may control identities at any time. For migrants, this multiplication of borders means they can be confronted with a "border control" at any time. 16 German labour law still has a priority system according to which an open position must be filled first by a "German", second by an "EU foreigner" if no qualified German can be found and third, only if again no one is found, by a so-called third country party. 17 One catalyst was the crises of the national system of asylum in the late 1980s when legal immigration became more difficult, and more and more migrants used asylum as an immigration loophole, "overstretching" the right to asylum, as Sadako Ogata saw it, then director of the UNHCR (see Ogata 1997). Leading functionaries and bureaucrats demanded the Europeanization of migration policy and a shift towards "migration management". The IOM, which pushes global migration management, played a central role in this policy shift (see Ghosh 1997;Hess 2010). 18 Phillip Ratfisch and Stephan Scheel (2010) identified a similar mechanism within the context of asylum policy and practice. In this case too, the rhetoric of protection allows the UNHCR not only to speak for refugees, but also allows them to selectively weed out refugees in need of protection from the vast mass of irregular migrants. 19 This conflation of "feminist" and "national" is inspired by Encarnacion Gutierrez Rodriguez's thoughts on Jasbir Puar's ideas about the "biopolitical logic of homonational discourses" in the context of the US war against terror (Rodriguez 2011: 95). 20 In this case, it is the "Muslim man" in particular who is presented as the homophobic, fundamental "other". 21 Whereby here too it is the "Muslim man" who is portrayed as patriarchal and violent.