Stranded Migrants and the Fragmented Journey

Transcription

1 Journal of Refugee Studies Vol. 23, No. 3 ß The Author Published by Oxford University Press. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org doi: /jrs/feq026 Advance Access publication 5 August 2010 Stranded Migrants and the Fragmented Journey MICHAEL COLLYER Sussex Centre for Migration Research, University of Sussex, Falmer, Brighton BN1 9SJ m.collyer@sussex.ac.uk MS received January 2010; revised MS received May 2010 Long and dangerous fragmented journeys have become a common feature of global migration systems. These are more than simple responses to stricter migration controls and are associated with related developments in technology and communications. They are therefore a structural change in migration systems which require a re-examination of the ways in which protection is offered to migrants. This has become an important theme of international discussion since 2005 but the humanitarian situation of international migrants on particular routes remains of concern to aid groups and many in the international community. This paper, based on research in Morocco, considers the protection needs of three groups of stranded migrants: yet to be recognized refugees, previously recognized refugees and those with other protection needs. It examines migration histories to identify reasons why this third group of individuals may be unable or unwilling to return and the nature of humanitarian assistance they require. Keywords: forced migration, Morocco, European Union, protection It is now widely accepted that migrants who are not recognized as refugees may still be considered as persons in need of protection (UNHCR 2008). There is, however, continuing uncertainty about how to define this expanded category and what this might mean in terms of practical policy responses. Protecting vulnerable international migrants who fall outside the scope of the international refugee regime has generated significant international concern since at least the early 1990s. This concern was formalized in 2007 with UNHCR s 10-point plan on Refugee protection and mixed migration (UNHCR 2007) and the High Commissioner s dialogue on Durable solutions in the context of international migration (UNHCR 2008). The number of undocumented migrants who are killed or injured during lengthy and extremely risky overland or maritime journeys initially motivated these concerns. Data on the number of migrant fatalities is increasingly robust,

2 274 Michael Collyer particularly in the Mediterranean (Carling 2007). Responding to this humanitarian and human rights challenge is now a subject of discussion at the highest levels. The first element of the 10-point plan is cooperation with key partners and there is evidence that these international discussions have had some positive results. They have helped to identify limited areas where international cooperation can make a genuine difference, such as laws around interception at sea; UNHCR signed a new agreement with the Moroccan government in 2007, which has resulted in greater respect for UNHCR-issued documentation; there are new priority areas for continued focus, such as protection sensitive border controls, and the discussions have drawn international attention to the human rights of undocumented migrants, which may well have resulted in a reduction of state sponsored abuse of international migrants; there has been no repeat of the Moroccan police force s decision to abandon an estimated 2,000 migrants on its remote desert border with Algeria in October 2005, for example. Yet, although progress is inevitably slow and some signs of minor changes are evident, it is difficult to escape the impression that the tremendous focus of international attention at very high levels over a period of several years, has not accomplished very much. The most obvious explanation for this is that the problems posed by these types of movements are particularly difficult to solve. Until relatively recently these migration systems were characterized as the asylum migration nexus though the term appears to have fallen out of use following its rejection by UNHCR as focused too clearly on south to north movement (Crisp 2008). The European Commission, which considers Europe as the intended destination for many of these migrants, continues to refer to transit migration (EC 2009) which also emphasizes south to north movements. The International Organization for Migration (IOM) has, since 2005, institutionalized the idea of the stranded migrant, in its Stranded Migrant Facility (IOM 2009). The label mixed migration has become popular more recently and although this initially seemed to offer few new insights more recent analysis, particularly following the High Commissioner s dialogue, has involved a recognition of the greater complexity of motivations, timing, legal status, resources and locations involved in such movements (UNHCR 2009). This article uses the terms stranded migrant and fragmented migration, both key elements of mixed migration which capture the essential character of the protection requirements of migrants in this situation. Stranded migrants is a broad term, with a variety of uses dating to 1921, when it was used to describe the plight of African Americans moving from the rural south of the US to the cities of the north without resources (Henderson 1921). Although IOM s Stranded Migrant Facility provides an invaluable service to many migrants who wish to return and would otherwise not be able to, return is obviously not the only potential solution to being stranded. There are a significant proportion, the terminally stranded, who do not wish or are not able to do so. I follow Dowd (2008) in her clear analysis of the term.

3 Stranded Migrants and the Fragmented Journey 275 The migrants of concern in this paper are distinguished from other stranded migrants, such as domestic workers who have fled abusive employers (Human Rights Watch 2008), by the fragmented nature of their migration, broken into a number of separate stages, involving varied motivations, legal statuses and living and employment conditions. It is often not the case that entire journeys are planned in advance but one stage may arise from the failure of a previous stage, limiting future options and draining resources. This may also be a key contributory factor in migrants vulnerability and protection needs (Collyer 2007). This paper draws on a series of research projects with undocumented migrants in Morocco: between 2003 and 2005, when 42 migrants were interviewed mostly in Rabat and Tangier; in 2005 and 2006 when a further 100 migrants were contacted, all over the country, and in the Spanish enclave of Ceuta; and on subsequent shorter visits to Rabat and Oujda (on the border with Algeria) in 2007 and It builds on the growing literature on these forms of migration, but rather than exploring why migrants came to Morocco or where they are planning on going next it develops an analysis of the nature of their protection requirements. Framing the situation of these migrants as a failure of protection, rather than a failure of control as it is usually presented by states, follows an important change of focus in research on these issues (e.g. Betts 2008) that I wish to pursue in this paper. The paper falls into three sections. The following section considers the development of fragmented migration, particularly across the Sahara. The second section examines the protection needs of these migrants, considering refugee protection, irregular secondary movement and other protection needs of stranded migrants. Although the research on which this paper is based did not set out to determine or verify exact circumstances of departure there are a number of ways of estimating the proportion of undocumented migrants in Morocco who may qualify for international protection. The paper finally turns to existing or possible future policy developments in providing this protection. Fragmented Migration across the Sahara Migration across the Sahara was first recognized as a significant issue in the late 1990s (Goldschmidt 2004), much later than comparable overland movements to the European Union from Eastern Europe or overland migration into the United States. It was not until December 1999 that illegal migration of non-moroccans into the Spanish enclaves of Ceuta and Melilla provoked the construction of a double high tension chain link fence along the entire length of each town s border with Morocco. In 2000 the Moroccan news magazine Demain ran an article on sub-saharan Africans in Tangier who made the crossing of the straits of Gibraltar in the rickety wooden pateras (Lmrabet 2000), illustrating the immediate impact of the increased difficulty of entering Ceuta and Melilla. Changing patterns of controls at Morocco s

4 276 Michael Collyer land or sea borders continue to push potential migrants towards whichever is the less controlled route. In 2002 the first academic investigation into transit migrants in Morocco was published with the support of the ILO (Barros et al. 2002) which provided an estimate of 10,000 undocumented migrants in Morocco. Other small studies followed covering not only Morocco (Collyer 2006; Alioua 2005; Lahlou 2005) but Algeria (Bensaad 2005), Tunisia (Boubakri 2006) and Libya (Hamood 2006). These investigations reached varying conclusions on the nature of protection needs of transit migrants, largely due to the necessarily small, non-representative samples. The French NGO Cimade published a study of 95 migrants (Wender 2004), whom they classified into two groups: those fleeing persecution or war (57.8 per cent), who could not be returned and therefore required some form of protection, and those migrating for economic reasons (42.2 per cent) who did not. In 2005 Médecins sans Frontières (MSF) released composite data on 9,350 medical consultations they had carried out over the previous 24 months in the north of Morocco in their work with undocumented migrants (MSF 2005). This report provided a disturbing testament to the dangers faced by migrants in Morocco, especially the significant role of violence in their lives, much of it at the hands of Spanish or Moroccan security forces. Almost a quarter of all individuals treated by MSF were suffering injuries incurred as a result of physical violence. To some extent the development of fragmented migration may be seen as a response to increasingly effective immigration controls in the most attractive destinations: Europe, North America or Australia. Restrictions on legal means of accessing European territory have progressively increased since the late 1980s. These developments effectively deny access to European asylum regimes, a fact that has been widely commented on (ECRE 2007). Such restrictions provide the motive for overland travel but by themselves they are not sufficient to explain the increasing importance of fragmented forms of migration. Such lengthy overland migrations depend on the availability of the necessary communications and other technologies that facilitate movement, such as instant international money transfers along the route or the availability of cheap mobile communications. The prominence of trans-saharan migrations since 2000, when significant restrictions on migration to Europe had been in place for more than a decade, but mobile phone access, money transfer facilities and widely available were just beginning to roll out across the region, suggests that technological change is perhaps more significant in explaining the recent development of fragmented migrations than policy restrictions. If this is the case, fragmented migration may well become a longer term component of many global migration systems, almost irrespective of any particular policy configuration in wealthy states. This also helps explain the extension of long overland journeys even away from assumed prime target destinations. If these migrations are provoked by technological developments, rather than simply increased controls, the nature of the potential migrants changes.

5 Stranded Migrants and the Fragmented Journey 277 If increased controls alone are to blame, those making the trip across the Sahara would be mostly relatively wealthy individuals who have been priced out of the market for smuggling directly by air by newly developed controls. Technological changes allow much poorer individuals previously restricted to sub-saharan migration networks to take advantage of the opportunity to travel a little further using many of the same means. This research found evidence of both profiles; the wealthy middle class individual, well educated, often with family in Europe, but also the much poorer, serial labour migrant who may have been moving around the Sahel region for years but was initially forced to move by poverty, war, or both. As elsewhere, international mobility still appears to be beyond the reach of the very poorest, but this research found evidence that it is no longer the sole preserve of the wealthy. Technological developments in the Saharan and Sahel regions have stretched sub-saharan migration networks so that they now reach to North Africa and for some even into Europe. Dangers on the Journey The source of the dangers faced by migrants on these fragmented journeys varies. In some cases they are posed by natural obstacles, such as seas, deserts or mountain ranges, which must be crossed in order to avoid the most intensely policed stretches of border or in some cases, such as the long desert borders of Algeria or Libya, simply to cross the border at all. These difficulties are typically exacerbated by the intervention of traffickers and smugglers whose involvement may vary from simple misinformation in the hope of increasing their own profits, to extreme cases of physical violence, armed robbery or enslavement. In addition to smugglers, which most migrants interviewed for this research had used for only short stretches of the journey, if at all, many migrants reported the existence of relatively well organized gangs of other migrants who were not systematically involved in smuggling or trafficking but simply extorted money from other migrants. This remains a particular problem at the Moroccan Algerian border where gangs know that many migrants have money and can be robbed with impunity, since they are unlikely to go to the police (and even if they did, would be unlikely to receive much sympathy). Initially it seemed that gangs developed to ensure their members had enough money to reach Europe themselves, but reportedly the key figures now involved are sought by police in Europe and they have become more permanent as their activities have moved from a subsistence to a professional basis. Finally, there are a number of incidents where the greatest threat arises from the practices of the agents of states. The most significant effect is probably indirect: the displacement impact of greater border controls at easily accessible border points obliges migrants to travel across more hostile terrain. This has been noted for some time at the US Mexican border (Cornelius and Lewis 2007) but more recently it has also been identified as a potential

6 278 Michael Collyer consequence of the Spanish Integrated System of External Vigilance (SIVE) which now covers most of the coasts of southern Spain and the Canary Islands (Carling 2007; APDHA 2009). There are also regular incidents in which migrant deaths result from the much more direct interventions of border control officials, though responsibility is habitually denied by border control agencies (e.g. MFA Egypt 2010). Where they occur at sea, as is most often the case, it is very difficult to determine responsibility. Surviving migrants often allege that boats were sunk as a result of collisions with border patrol boats. Border patrols themselves typically have a different account: that boats were sunk by smugglers in order to create confusion as a border patrol boat approaches. 1 On more hazardous crossings, such as from West Africa to the Canary Islands, it has become common for migrants boats to head towards patrol boats rather than away, in the hope of rescue, yet border control crews are not sufficiently equipped or trained for large scale rescues at sea, particularly if seas are rough. Given the difficulty of the task, the lack of training and the absence of apparent motive of either side to deliberately sink the boats, it seems more likely that accidents result from mistakes than from deliberate sabotage on the part of the crew of either boat. Incidents on land are easier to account for, such as the deaths at the borders between Ceuta and Melilla and Morocco. In October migrants were shot and killed by border patrol agents. This incident provoked substantial change in the orientation of European migration policy, the Global Approach to Migration, which remains a central element in harmonizing EU approaches to migration (EC 2009), though there has still not been an effective enquiry into the incidents and on at least one occasion since migrants have been shot dead at the same border (AI 2006). The dangers of the journey itself are common to all migrants on these routes, whether they have broader protection needs or not. Information campaigns to alert migrants to these dangers were initiated in the early 1990s in Morocco, targeted at Moroccan citizens, and are now in use through diverse mediums beyond, such as school textbooks in Senegal. Most migrants interviewed for this research, however, were not ignorant of the journey s dangers; the awareness of the potential hazards did not stop them leaving. Indeed research with potential migrants in West Africa suggests that such journeys are seen as important rites of passage, an indication that young men, in particular, have the necessary courage to get on in life, and their danger is integral to that (Fall 2007). This suggests that dissuading individuals from leaving will have only limited success under current circumstances. Longer term efforts should obviously continue to be made to resolve the initial root causes of these movements. In the interim, renewed efforts should be directed to resolving protection needs at one of the many stages which punctuate these fragmented journeys, where migrants may become stranded.

7 Stranded Migrants and the Fragmented Journey 279 Three Groups of Stranded Migrants Most classifications of migration depend on the circumstances of departure. This is certainly the case for the classic binary of forced versus voluntary migration, as well as for legal distinctions between refugees and others. Even though from a sociological perspective it is now widely accepted that such binaries are more accurately reflected by a continuum (e.g. Van Hear 1998), it is still the circumstances governing an individual s decision to leave their place of citizenship or at least long term residence, that dictate the ways in which that migration will be subsequently classified. This is obviously never entirely appropriate; the characteristics of fragmented journeys highlight the drawbacks of viewing migration as a relatively rapid transition between defined points of origin and destination. Fragmented journeys are continually prolonged and migrants may be far removed in both time and space from their experiences of departure so that their reasons for leaving no longer have the relevance that they once did. At the same time, migrants do not have the benefit of arrival at an intended destination and the reinterpretation and justification of the migration that is permitted, even encouraged, by its termination. Rather, they are caught in-between since neither the logic of departure nor that of arrival can explain their movement adequately. Their migration is still very much in the present and it carries the logic of the journey in that it can only be justified by some uncertain future or an increasingly distant past. The length and complexity of many migrants fragmented journeys across the Sahara undermines any clear identification of points of origin and destination. From a sociological perspective their fluctuating experiences on the journey take on a much greater importance in understanding their migration. This is also reflected in migrants own descriptions of their movement. However, translating this perspective into policy terms raises a number of supplementary questions. Three groups of individuals present clear yet distinct protection needs, from the perspective of research in Morocco, which for most of them was only the most recent stop on a lengthy journey. The first group qualify for protection under the 1951 Convention or the 1969 OAU supplementary definition. It is likely that a significant proportion of stranded migrants, probably a minority, though a sizeable one, are eligible for refugee status under a strict application of these two conventions. In most of the countries they pass through they have little chance of receiving status, though following UNHCR s move to Rabat in 2005 and particularly since the agreement with the Moroccan government in July 2007, refugee recognitions in Morocco have increased (ICMC 2010). A second significant group of Saharan transit migrants have already been recognized as refugees, usually by UNHCR in a country of first asylum. They have frequently spent many years, accounting for most of their lives in some cases, in protracted refugee situations, often living in camps. Their departure

8 280 Michael Collyer arises from despair that their prolonged exile will allow any opportunity to develop their lives in any meaningful way. Much of this type of migration would be considered irregular secondary movement in the sense of UNHCR s Convention Plus discussions (see UNHCR 2005). The third group are those whose humanitarian situation is extremely urgent and who cannot turn to their states of citizenship for assistance, but are unlikely to qualify for refugee status. Until recently there had been no real attempt to provide protection for any of the three groups. Correct application of existing legislation and enforcement of state obligations would meet the protection needs of the first two groups and there are places, such as Morocco, where there are signs of progress in this area. Yet correct application of existing frameworks would offer little support for the third group. Even if these legal protections were all correctly implemented, stranded migrants with no case for asylum would remain vulnerable. It is for this reason that this obviously heterogeneous group of people poses a particular challenge for international policy. Despite the increasing attention this form of migration has received from various sources, limited information on illegally resident foreign nationals in Morocco suggests their numbers do not appear to have significantly increased and remain at about 10,000. This is a small fraction of the numbers of undocumented migrants regularized in Spain alone since Clandestine migration is a fairly insignificant means of entering Europe that receives a degree of media and policy attention far out of proportion to its numerical significance (de Haas 2008). Although data is scarce and uncertain, studies of illegal residence in Europe suggest that the large majority arrive with visas and simply overstay (Black et al. 2006; de Haas 2008). The only thing that has changed since the 1990s is the increased duration of the Moroccan stage of a journey on to Europe. Describing the situation in Morocco in the 1990s Goldschmidt (2002) writes of a relatively unproblematic passage to Europe for those that so desired, with few delays, typically through Ceuta or Melilla and on by boat to Spain. A decade later the average stay of undocumented migrants in Morocco had increased to 15.4 months with individual stays lasting as long as seven years (Collyer 2006). This suggests that some migrants are actively making a life for themselves in Morocco and indeed for some this may have been the objective of their migration to begin with (de Haas 2008). Over this period the international discourse surrounding undocumented migrants in Morocco has also changed as an initial focus on repression has given way to a broader interest in questions of protection and asylum. EU policy documents first referred to transit migration in relation to Morocco in the 1999 Morocco Action Plan, drawn up by the High Level Working Group on Asylum and Migration (EC 1999). This document called on Morocco to criminalize undocumented migration into and out of Morocco. At the time Morocco rejected this call but much of the 1999 Action Plan found its way into Moroccan Law 02/03, passed in November This was

9 Stranded Migrants and the Fragmented Journey 281 Morocco s first, post-independence migration legislation, which amongst other things makes undocumented migration a criminal offence (Belguendouz 2003). In December 2004 the EU released a further Action Plan for Morocco, in the context of the European Neighbourhood Policy. The 2004 Action Plan introduced asylum for the first time, expressing intentions to assist the Moroccan government to develop capacity to meet its international obligations. Morocco is a signatory to both the 1951 Convention and the 1967 Protocol and UNHCR had a relatively insignificant presence in Casablanca from 1959 onwards conducting occasional status determination. In January 2005 the UNHCR office in Morocco was transferred to Rabat, where contact with the institutions of the capital was easier, and began to conduct individual status determination with much greater urgency. The signature of the agreement between UNHCR and the government of Morocco in 2007 introduced an important new stage in this relationship. In early 2010 UNHCR was still doing all status determination work but there were plans to pass some of this work over to the Moroccan government when appropriate legislative frameworks can be implemented. Research on which this paper is based spans this transition in UNHCR s role in Morocco, covering the period Protection Needs The three groups of individuals considered in the previous section (previously recognized refugees, as yet unrecognized refugees and others with no apparent claim for refugee status but nonetheless with distinct protection needs) reported clear protection needs during research in Morocco. These three groups have distinct objectives and previous experiences of UNHCR and have different implications for policy interventions. The majority of individuals interviewed were men, mostly from West Africa (Tables 1 and 2), though the contrast between the profile of individuals interviewed in Morocco itself and across the border in Ceuta suggests that there are different migration systems operating. Migrants registered in the open detention centre (CETI) in Ceuta included a large minority of South Asians who have not been identified in this or any other research project in Morocco, yet who reported that they travelled through Morocco with smugglers and were kept in closed houses rather than staying in the migrant neighbourhoods of Rabat or Casablanca or the informal camps near the Algerian border. This suggests a number of parallel undocumented migration systems operating through Morocco, which may well have implications for those who are denied access to UNHCR by smugglers during their passage through Morocco. This section considers these three groups in turn.

10 282 Michael Collyer Table 1 Nationality of Undocumented Migrants Interviewed in Morocco during Two Research Projects, and Country of origin Men Women Total under over 40 Total under over 40 Total DRC Coˆte d Ivoire Nigeria Liberia Cameroon Congo (Brazzaville) Ghana Sierra Leone Senegal Mali Togo Benin Sudan Chad Total Table 2 Nationality of Undocumented Migrants Interviewed in Ceuta in December 2005 Country of origin Men Women under over 40 Bangladesh Liberia Guinea Bissau Guinea Conakry Uganda Gambia Cameroon Total As yet Unrecognized Refugees This group is perhaps the hardest and most controversial to identify and this research did not aim to assess individuals asylum claims, preferring to focus on everything that had happened since they left their home country, rather

11 than before. About 10 per cent of those interviewed, by their own admission did not require protection, and in some cases were actively searching for means of returning home. These included a Liberian student whose studies in Morocco had ended but who never received money from his government for his flight home, a Nigerian sex-worker whose boyfriend was recently repatriated and wished to follow him due to her concerns about continuing work without his protection, and a 29 year old engineer from Cameroon who, after five years sleeping rough in the forests of northern Morocco and countless assaults on the chain link fences reported that I no longer have the courage to fight any more. These individuals are ideally suited to the services of IOM s Stranded Migrant Facility, keen to return, but unable to do so with their own resources (though that is not to say that they would qualify). This 10 per cent also included others who fitted the European Commission s classic image of the transit migrant, such as a 25 year old former student from Benin, interviewed near the border with Melilla in January 2006 who was particularly clear in his goals: I m not interested in asylum, or UNHCR or Rabat, or any of that. I m aiming for Europe. Beyond this 10 per cent group, however, individuals were not so easy to classify. During the research project the majority had applied or reported that they intended to apply for asylum (Table 3). Before UNHCR operations shifted to Rabat in 2005 the Casablanca office did not fulfil UNHCR functions particularly effectively. Lindstrom (2002) found that none of the migrants she spoke to in 2001 had registered a claim. This was still the pattern during the research reported here, when only one of the 42 individuals had registered a claim and had waited six years without a response. The rise in applications and intended applications after 2005 is likely caused by an increase in confidence in UNHCR, following the move to Rabat and a fall in the feasibility of other possible options for remaining in Morocco. Nevertheless, it is surprising that even in the context where asylum offers the only possible protection against deportation from Morocco, a significant minority had no intention of applying (Table 3), so it is not the case that individuals automatically claim asylum. From 2005 to 2009 inclusive, UNHCR Rabat received 6,362 applications for asylum; 318 applications were pending at the end of Over this Table 3 Stranded Migrants and the Fragmented Journey 283 Numbers of Migrants Interviewed in Morocco in Who Had Registered an Asylum Claim with UNHCR Rabat or Intended to Do So (n¼100) Yes No Have already registered asylum claim in Rabat Of those who have not registered: intend to register NB: UNHCR only began operations in Rabat in 2005.

12 284 Michael Collyer period 867 individuals were recognized as refugees, a recognition rate of approximately 15 per cent (ICMC 2010). The review of UNHCR Morocco conducted by ICMC found no basis to criticize those decisions, in fact it identifies UNHCR Morocco as having had considerable influence in shaping UNHCR policies on mixed migration and on urban refugees. It also commends UNHCR s role in conducting outreach operations. Nevertheless, UNHCR cannot itself offer protection, and since it lacks the reach of the state there will inevitably be those who are unable to access this protection. One of the most significant outstanding problems resulting from mixed migration in Morocco is the fate of those who are not recognized as refugees. A key insight arising from research into mixed migration is that many of those who are not refugees still have legitimate protection needs. The Moroccan government s policy is that all should be deported and various large scale deportations reveal an ability to reach this goal. In contrast, once individuals reach Europe they have various other options. Most EU Member States operate one or more categories of subsidiary protection. Even those who are refused all forms of legal status are in practice unlikely to be deported, a fact well known to most participants of this research. Interviews in Ceuta confirmed that only about 10 per cent of applicants in Ceuta and Melilla are deported directly. The remainder of rejected asylum seekers are given an invitation to leave Spain, usually within three months, with which they are able to travel to the mainland, from where the chance of them being forcibly repatriated, even if caught, is extremely low. 2 Given the continual harassment and appalling conditions these individuals typically face this cannot be considered a solution, but it provides a safety net to those who cannot return that is currently lacking in Morocco. This has important consequences for the other two categories of migrants. Previously Recognized Refugees A small but significant minority of individuals interviewed had already been granted asylum under mass recognition systems operated by UNHCR in the context of one of a number of West African conflicts. The refugee status they had already received had a specific, locally defined meaning, usually based on the OAU Convention, rather than the 1951 Convention, but they did not always appreciate this and considered themselves as refugees when they arrived in Morocco. Applying to UNHCR a second time seemed an unnecessary step, though in the language of Convention Plus their migration across the Sahara would be considered irregular secondary movement since they had already received protection. A short migration history of one of the individuals interviewed for this research illustrates the confusion of their position: Paul 3 is Liberian. Following the deaths of both his parents he left Liberia in 1989 at the age of nine, and travelled with a family friend to Guinea. He was recognized as a refugee by UNHCR and spent the next 14 years in Kola

13 Stranded Migrants and the Fragmented Journey 285 refugee camp. In 2003, when he was 24, the family friend, who had brought him to Guinea and become his only remaining family, died. This motivated his decision to leave Guinea. He travelled overland taking up small jobs along the way with the intention of financing his trip to Europe. It took him two and a half years to reach Morocco. He arrived in August He soon realized that his Guinean refugee card was meaningless to Moroccan police and after his first removal to the Oujda border he applied for refugee status again in Morocco. He was confused that UNHCR had so little influence in Morocco, a radically different situation than he was used to in Guinea, where it effectively operated as a surrogate ministry. He described his first short interview at UNHCR s Rabat offices: I replied to all the questions, and they told me that, really, I deserve to have the paper, but UNHCR is not really valued in Morocco, so I said what will I do? I don t have other parents, I don t have anything else, even when I was in Guinea, my parents, my mother and my father, was UNHCR because it was them who gave us food to eat, they took care of us, of everything. His confusion at UNHCR s presence in Morocco is indicative of many other irregular secondary movers who had learnt to depend on the organization for everything. These forms of dependence have been widely criticized elsewhere, particularly in the context of long term encampment (e.g. Turner 1999). Migrants with previous experience of UNHCR often expected some form of material support on arrival in Morocco, as they had been used to elsewhere. A Congolese man, Peter, who I interviewed soon after he crossed the Algerian border had heard that he had to make the 600 km journey to Rabat before he was able to identify himself as a refugee to UNHCR, which he already found strange, and he wanted to confirm what facilities would be available once he got there: Peter: I suppose they ve got tents set up for people once they get to Rabat, yes? No, no, there s nothing like that. Peter: No shelter? With winter coming on? How do they expect people to manage? Peter s incredulity at the lack of refugee camp facilities in Rabat illustrates the mode of operation of UNHCR which he had become accustomed to. Officially provided shelter for asylum seekers, in the form of a refugee camp or other housing, is treated with suspicion by the Moroccan authorities. If applications for asylum are accepted in such a location they would come to resemble what some EU Member States have had in mind in discussions of extra-territorial processing centres, particularly if resettlement to EU Member States is also made available. Both the Moroccan government and migrants rights organizations have strongly resisted such developments. UNHCR Morocco is now using a significant proportion of its budget to support income generating activities for recognized refugees, so there is material

14 286 Michael Collyer support available (UNHCR 2010). These forms of support are very different from the basic necessities that many are used to receiving. Previously recognized refugees were also confused at the idea of UNHCR conducting individual status determination, rather than mass recognition. Since they had already been recognized as refugees, and many had the identity documents to prove it, they could not see why that recognition should not be international. These changes in material support and recognition processes reflect the distinctions between the traditionally imagined rural refugee context and the increasingly significant urban displacement. Mass recognized refugees are expected to remain in the context of asylum in which they have been recognized, unless they are resettled, however protracted the situation has become. Fragmented journeys in this context illustrate a relatively new development of self-resettlement as the new infrastructure and information technologies make trans-saharan journeys feasible even for relatively poor refugees. Protection and the Journey: the Terminally Stranded The third category of individuals have not been recognized as refugees and, either by their own admission or on the basis of a short description of their migration background, are unlikely to be so. Yet for often very good reasons they cannot return to their country of citizenship. Most of the migrants interviewed during this research would be considered as stranded, yet for the previous two groups their situation would be resolved by the correct application of states existing international obligations: the first group through access to European territory to register a claim or respect for the international refugee conventions signed in Morocco or in any of the countries passed on the way, the second by the creation of options other than long-term warehousing of refugees in the protracted situations in which they were first recognized as such. In contrast this third group have extremely limited protection. There may be some protection under the 1990s Migrant Workers Convention, 4 though even in the limited number of countries which have ratified it, it is not clear that it would be applicable to individuals without authorization to enter, stay or work. Other human rights conventions would, if applied, prevent serious human rights abuses on the part of states, but as discussed earlier, it is only under rare circumstances that the threats these migrants face are posed directly by state agents. Enforcement of existing agreements would therefore have only a marginal effect on these migrants. The difficulties of return under these circumstances are illustrated by a second extended migration history, of a 42 year old woman from Coˆte d Ivoire. Mrs Kwembe is from Bouake in Coˆte d Ivoire. Soon after the fighting began in September 2002 her husband and eldest son were killed in the street. She did not know if they were singled out, but soon afterwards their

15 Stranded Migrants and the Fragmented Journey 287 house was attacked while she was out. She returned to the house briefly to collect their limited savings and travelled north with her younger son and daughter to escape the fighting. They crossed into Mali and eventually reached Bamako where she thought they would be safe until they could return home. They remained in Bamako for one year but when there was no sign of resolution to the conflict, and it was clear that Bouaké was on the front line, she made plans to use the remainder of their savings to move on, rather than use it up waiting in Bamako. She paid smugglers for passage to Tamanrasset for herself and her two children, passing through Gao, Kidal and Sibris. They spent several months in Maghnia, a camp on the Algerian side of the border with Morocco, over the summer of 2004 and came across the border direct to Rabat in November. Morocco was not at all as she imagined it: I thought I may be able to get work, that my daughter would be able to re-enter school. She had not worked at all in the year that they had been there and had only managed to live by borrowing money from friends at home and charity from various church groups. In October 2005 she visited UNHCR to claim asylum and was given an interview date in December but two days later she was stopped in a large police operation in the neighbourhood where she lived. The police officers told her they did not recognize her UNHCR document and forced her and her children onto a waiting bus. They were on the bus for many hours and were eventually told to get out, in the desert. Her 16 year old son became extremely agitated and ran off through the crowds. She followed other people walking back through the desert, in the darkness, looking for him, but she was unable to find him and she has not seen him again. When the buses returned to collect them again, she was not taken to the military camps, but directly to the border near Oujda, with her daughter, from where they were able to return, but she still had had no news of her son. Mrs Kembe said that she would be prepared to return to Coˆte d Ivoire if only she could find her son, even though she knows that her house has been destroyed and she remains unclear about the reasons that provoked the killing of her husband and eldest son. Her daughter had been out of school for three years, she was exhausted by the difficulties she had endured and was terrified at the prospect of being separated from her daughter in a subsequent police raid. On the basis of this story it is probably unlikely that Mrs Kembe and her daughter would be granted asylum in Europe. She admitted that she left Bamako, where she was in no immediate danger, in search of work and the move from Bamako was not motivated by a search for protection. Nevertheless, absence of violence is not the same as protection, and she is hardly an economic migrant. In addition to the complexities of motivations of her flight, the difficulties she faced were as much to do with the circumstances of her journey to and residence in Morocco, as with anything connected to Coˆte d Ivoire, which she had left more than three years earlier. There are many such stories to explain a reluctance to return, even once the difficulties of living in Morocco have become clear. Mrs Kwembe may well

16 288 Michael Collyer be considered safe on return to Coˆte d Ivoire, certainly once hostilities have ended, but she is not prepared to go back without her son. Even given the option of return under a programme such as IOM s Stranded Migrant Facility, she would not return voluntarily. For many the most significant barriers to return resulted from financial obligations to family, friends or worse (such as credit agents) which they would never be able to repay if they returned. Even those without debts described their journey in terms of an investment. James, a 19 year old Nigerian, had taken over a year to come overland from Benin City, but had left home with barely 100 dollars and taken small jobs in exchange for food along the way. He was sleeping rough, near the border at Oujda. When I asked if he would be interested in returning, as the Nigerian government had begun to organize mass repatriation flights, he replied No, I don t wish it. I ve suffered too much. I deserve better. Few expressed it so clearly but this understanding of suffering as somehow earning a right to expect more was widely reported, particularly amongst those who had been injured during their migration, a connection also reported in research in Tunisia (Laacher 2005). The powerful, almost violent commitment to an imagined future goes beyond the widely reported debt obligations and poses a particular challenge for policy to engage with migrants in this situation. Providing Protection: Policy Responses Mixed migration poses obvious policy challenges. Quite apart from the special category of refugees, it recognizes that migrants may have valid reasons for movement which at times may be irregular. As the illustrations in this paper have shown, some migrants have good reason to resist return to their country of origin and it may be unacceptable, from a human rights perspective, to return them. These forms of migration occur all over the world and only a small minority of migrants even try to reach wealthier parts of the world, such as the EU. Yet those that do attract a disproportionate level of scrutiny. It is easy to mistake the characteristics of migrants engaged in these fragmented journeys if only the most recent stage of their movement is considered. At this point they become stranded, unable to continue but refusing to return. Yet the reasons for their current situation can only be understood by a longer term perspective of the previous stages of their journey. There is no question that the full application of the 1951 Convention in Morocco and respect for that process by the Moroccan government authorities would improve conditions for many transit migrants in Morocco who are eligible for international protection. This is the thrust of the current EU partnership discussions with Morocco as well as UNHCR negotiations to establish some kind of protection space, however limited, in which recognized refugees will be guaranteed to be free of harassment from the Moroccan authorities. There is some ambiguity here as this development would also help to legitimize the gradual closure of access to asylum systems

17 Stranded Migrants and the Fragmented Journey 289 within the EU. The role of the EU in implementing mechanisms to provide international protection beyond its borders, as opposed to simply controlling migration, is a comparatively new idea. It entered the EU s agenda at the Thessaloniki European Council in June 2003 and appeared in the EU s Mediterranean policy in the Morocco and Tunisia Action Plans, published in December It was the events of September 2005, when 11 migrants were shot by security forces while trying to scale the fences around Ceuta and Melilla, that indicated the potential gravity of continued migration of this nature (AI 2005). These events were an important catalyst for the shift in orientation of the EU s approach to migration from East West to South North (EC 2008). The first draft of the EU s Global Approach to Migration appeared in December The immediate result of the policy dialogue was the first Euro African summit on migration and development, held in Rabat in July 2006, followed by a full African Union European Union meeting in November 2006 in Tripoli. Follow-up ministerial level meetings in Madrid in June 2007 (in the Rabat process ) and the EU AU summit in Lisbon in December 2007 (in the Tripoli process ) as well as the first ministerial meeting devoted to migration in the Euro Mediterranean Barcelona process in November 2007 all focused on related themes (Collyer 2009). The Mediterranean Union initiative of 2008 has combined these with a Euro African process structured along the lines of the Rabat priorities, though the Mediterranean Union has had an extremely low profile since then. These multi-lateral approaches to migration in the Euro Mediterranean area are all strongly influenced by the EU s Global Approach to Migration. This considers migration in three sections: legal migration, illegal migration and migration and development. Asylum is missing an obvious place in this overview, though the European Council has continually emphasized that asylum is central to its concerns (such as at the Luxembourg JHA European Council in June 2008). Although the boundary between refugee movement and other forms of migration is certainly blurred, a line must be drawn at some point and UNHCR maintains that refugees are a clearly identifiable group of people (UNHCR 2009). The policy challenge is then to link the first two sections of the global approach (legal migration and illegal migration) in ways that recognize that people of concern, from a protection point of view, are not always refugees, and that protection may be found through access to legal forms of migration just as easily as through access to asylum. Yet this appears to be some way off. There is currently little enthusiasm within the Moroccan government to institute a realistic system of protection for individuals that they assume will continue in their attempts to reach Europe. It is not in Morocco s interests to attempt to integrate groups of individuals who have no desire to be there, especially when unemployment among nationals is officially 20 per cent and unofficially very much higher. UNHCR is reporting some progress in discussions with the Moroccan