This thesis has been submitted in fulfilment of the requirements for a postgraduate degree (e.g. PhD, MPhil, DClinPsychol) at the University of Edinburgh. Please note the following terms and conditions of use: This work is protected by copyright and other intellectual property rights, which are retained by the thesis author, unless otherwise stated. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the author. The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the author. When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given. Standing on one leg Mobility, money and power in East Africa’s Somali social networks Gianluca Iazzolino PhD The University of Edinburgh 2016 i Declaration This thesis has been composed by myself from results of my own work, except where stated otherwise. It has not been submitted in any previous application for a degree. July 2016 ii Abstract My thesis examines dynamics of inclusion and exclusion within Somali social networks in East Africa. It focuses on Somali mobility patterns and financial practices to draw insights on the maintenance, reproduction, and transformation of both solidarity ties and inequalities. By examining Somali communities in Kenya, host of the largest Somali refugee population outside of Somalia, and Uganda, an increasingly important recipient of Somali refugees and migrants, this thesis seeks to understand how mechanisms of social stratification rooted in Somali socio-cultural structures are reproduced in mixed migration flows encompassing both forced and voluntary migrants. It analyses sets of relationships whose continuity and changes are regulated by the interaction of structure, agency, and institutions, and argues, on the one hand, that networks are dominated by groups who hold sway over economic and political resources, precluding others from accessing key assets that may help challenge relations of subordination. On the other, that pre-existing inequalities hinder on the capability to move across both physical and institutional categories. These inequalities can be traced back to asymmetric clan relationships shaped by Somali historical trajectories before and after the implosion of the state. However, this thesis suggests also that kin relationships only partially explain why and how bonds are sustained and forged. Instead, by observing the mechanisms that animate networks, reproducing both solidarity and marginalisation, this thesis teases out how new linkages are created and how Somalis communities accommodate to specific institutional settings, either adapting to narrowing windows of opportunity or maximising the benefits that may be yielded from their widening. The thread running throughout this thesis is the argument that mobility contributes not only to accessing and mobilising strategic resources but also to shaping processes of social stratification. By using ethnographic methods of data collection, this thesis seeks to shed light on rifts in Somali social networks often masked by the veneer of trust. iii Acknowledgments Tom Molony and Jamie Cross, my supervisors, have helped me fight off the demons of procrastination and see things more clearly along the way. Tom Molony, in particular, was the first one to welcome me in Edinburgh when, in 2011, I handed in a draft for a research project much different from the one which was to become this thesis. Lucy Lowe, Robert MacDonald, Peter Chomka, Diego Maria Malara, Mohamed Hersi and Leila Sinclair-Bright have offered precious comments on the draft. I am grateful to the Economic and Social Research Council for funding my PhD and to the Institute of Money Technology and Financial Inclusion (IMTFI) University of California Irvine for supporting my fieldwork (Subaward No. 2013-2929). I thank in particular Bill Maurer and Jenny Fan. I was lucky to first meet, by chance, Neil Carrier in a restaurant in Eastleigh in 2011 and have a first insightful conversation with him over a goat soup. The conversation has continued ever since. In Nairobi, Mark Bradbury at the Rift Valley Institute has encouraged me to write a briefing for the RVI which later became chapter 5 of this thesis. Julie McDowall-Lawson at FAO Somalia provided me with precious contacts to acquire information in Somalia and Somaliland. iv During my fieldwork, I have collected many debts with my informants, and some of them have become friends. I am referring in particular to Omar Abu Omar, Abdi ‘Senegal’, Adams Mohamoud Hussein and Mohamed Abdullahi in Nairobi and Shukri Islow and Hussein Mohamed in Kampala. In Eastleigh, I am thankful to Sayyed ‘grazie maestro’ Ali, Sheik Durman, Mohamed Kidogo, Abdikadir Osman and his beautiful family, Muhubo Salad and his cousin Liban, Ferdowsa Abdir and her family, Saba, Fatuma, Bashir and the ‘gang’ of the exchange market, Jamal Yussuf, Hamda Sala, Adams Mohammed, Abdul Xadir, Abdel Nassir, Hassan, Abdi Fatah, Abdiqadir Ali, Prof Ibrahim Farrah at the University of Nairobi, Edmund Langata of the Refugee Consortium of Kenya, Abdi Aziz, Feisal Abdul Axmed, Yussuf Isaq, MJ, Axmed Abdullahi. Fieldwork in the Kakuma refugee camp would not have been possible without the support of Refugee United, in particular Alexandra Aparicio, and the help of Ibrahim Mual, Dahir Abdullah, Omar Ahmed Mohammed, Kero Abdi Isaq, Akhyiar Mustaf and Aisha Mahmoud. Prof Omar Eno and Dan Van Lehman have shared with me their knowledge and their passion over the plight of the Jareer people. Eventually, I thank my family, and Veronika, for bearing with my absences and being such a great partner. v Prologue 1 Introduction 8 Rationale of the study and research questions 9 Methodology 11 Choice of the sampling 13 Multi-sited ethnography 15 Interactions with the informants 16 Organisation of the thesis 17 1 Literature review and theoretical framework 19 Discussing mobility 19 The significance of ‘sticky’ places 21 African Migrations in Africa 23 Diasporas between security and development 26 Remitting trust 28 Digital diasporas 30 African nodes of low-end globalisation 31 Urban livelihoods between the formal and informal 33 Social networks as theoretical framework 37 What we talk about when we talk about networks 37 Variety of social networks 41 Mechanisms of social exclusion 45 Culture, agency and power in networks 50 2 Clanship in Somali history through colonial and postcolonial times 55 Introduction 55 An encounter 56 Framing clanship in Somalia 59 The Somali grand narrative 60 Islam and clanship 63 Somali minorities 65 Somalia before Independence 72 vi Somalia under Italian rule 75 Somali Independence 77 The 1969 Somali revolution 79 The rise of the remittance industry 81 The Ogaden war and the beginning of the end 82 Conclusions 86 3 Conflict, crisis, and mobilities 88 Introduction 88 Of games, fashion, and war 89 War, Humanitarian Intervention 92 Havoc and famine in Gosha 94 Questioning the failing state narrative 98 Strategic convergences 99 The TFG and the ‘4.5 formula’ 100 The 2006 Ethiopian invasion and the rise of Al-Shabaab 101 Famine and relief 103 Somalis in Kenya 104 Refugee flows to Kenya 106 Conclusions 107 4 Hustle and bustle in Little Mogadishu 111 Introduction 111 The road to Eastleigh 112 ‘Openly informal’ 117 A time of uncertainty 120 Fear and desire 121 Little Mogadishu rises 125 Changes and consolidation 127 Eastleigh’s communicative ecology 131 Eastleigh’s voices 135 Digital remittances 140 vii The emergence of an ethnic niche 141 Conclusions 144 5 Follow the money 145 Introduction 145 Spinning the spindle 146 Unpacking trust 149 The ‘myth’ of trust 151 Obligations and debts 152 Money-go-round Somali style 154 Social capital as collateral 157 Remitting the past and the future 158 When kinship is not enough 161 Converting capital, producing value 164 Intersecting platforms 167 Taking financial choices 169 Informal interoperability 170 The exchange market 172 Preserving multiplicity, cultivating alternatives 176 Conclusions 178 6 Crossing borders in troubled times 180 Introduction 180 Troubles in Eastleigh 181 An emerging ‘migration system’ in a volatile context 183 “Should I stay or should I go?” .. 186 The route to Kampala 190 A new Little Mogadishu takes shape . 193 Youth mobility 196 The role of student associations 198 Regional connections 200 Strategic synergies 202 viii Conclusions 204 7 Power and class in a humanitarian space 206 Arriving in Kakuma 207 History and topography of Kakuma 212 Social differentiation, space, economy, security 219 Biometric systems and foodstuff sale 224 “Thanks to America, I became a Bantu” 227 A vicious circle of poverty 233 Conclusions 237 Conclusions 239 References 246 1 Prologue We hung out after iftar. The drifting chant of the muezzin was a spiritual watershed between the day’s atonement and the evening’s respite. Silence fell on the streets of Eastleigh, and the entire neighbourhood remained suspended for the time of the prayer, the noises of Nairobi buzzing at a distance. Then the Sura Qadr ended, and Somalis of all ages, mostly men, poured out of the mosques, rushing home or to other venues where they would break the fasting of the Ramadan. The restaurants filled quickly. I usually met with Hassan on the balcony of the Andalus Hotel, where local notables sat around large plates of pasta and rice and camel meat. In their pockets, they kept Kenyan identity cards, foreign passports, UNHCR cards or, sometimes, all of them together. Yet, Little Mogadishu, as this area of the Kamkunji constituency was known across Kenya, was the mirror of their sense of belonging. A crucible where they conjured up their troubled homeland, by speaking Somali in the streets, recognising Somali faces at every corner, and ‘living a Somali life’, as many said. Eastleigh was a mooring to Somalia in the heart of Nairobi, for Somalis coming from both the homeland and the diaspora. Hassan, for instance, was a Kenyan citizen from Mandera, in the North-Eastern Province (NEP), where the border with Somalia is just a bureaucratic formality, a reminder of past colonial empires. He was employed at the local branch of Barclays Bank and his job, as the Somali face of an international bank, coupled with his flamboyant attitude, made him very popular among businessmen in Eastleigh. His salaried job was a poor fit for his entrepreneurial ambitions. His favourite topics of conversation were business and the Premier League, and I had little to say on both. But I enjoyed listening, because he was an enticing 2 storyteller, like most Somalis, and more so when Ali and Mahmud were around. They were also from Mandera – or, at least, so it was written on their papers. Mahmud’s past, in particular, was as blurry as the Kenyan-Somali border: not a clear cut line, but a boundary that could be redrawn according to the circumstances, provided that you have the means to do it. And Mahmoud was wealthy (“He has two wives and he’s prowling for the third one”, his friends said jiggling) and well connected. To explain it better, he was wealthy because he was well connected. Glimpses of a previous life in Benadir region, around Mogadishu, sometimes surfaced in our conversation here and there; accounts of the war were only briefly mentioned and put aside. The life worth telling began in Mandera, on the Kenyan side of the border, where Mahmud had become a Kenyan citizen. Small talk typically accompanied the dinner. Then, in Hassan’s car, we ventured through the traffic of Nairobi, as far as the upscale Westland neighbourhood. The nights in July were crisp, and, sipping cappuccinos, we squeezed around the terrace heaters that punctuated the dehors of the Art Cafe, on the ground floor of the Westgate shopping mall. Only two months later, that same place was to be awash with blood, and the glamorous corridors of the mall strewn with corpses. As many around the world, we would watch aghast the images of the building turned into a battle ground for militants of Al-Shabaab, an Islamist organisation, and Kenyan security forces, survivors rushing to shelter, plumes of smoke billowing from windows and terraces. The Westgate shopping centre attack lasted three days, between the 21st and the 23rd September 2013, claimed at least 67 lives and left many questions unanswered, triggering a heavy backlash on the Somali population in Kenya. A security crackdown, with mass roundup and detentions, disrupted businesses, prompted the deportation of many refugees back to Somalia and gave corrupt police free rein to extort money from any ethnic Somali they could get hold of. Encounters with predatory security forces were not a completely new thing for Somalis in Kenya, but the post-Westgate situation marked a new low in the relations between the Kenyan state and its Somali population, both citizens and refugees. However, back in July, everything seemed smooth. The Kenyan presidential elections, which had stirred fears of a return to the political violence seen in 2008, had passed without major incidents. Westgate was a monument to the sense of self- 3 righteousness of a Kenyan elite that fuelled an ‘Africa rising’ rhetoric and of an international community almost equally split among those who enthusiastically embraced this narrative (the ‘startupper’ population of Nairobi was becoming larger that of the international NGO (INGO) workers) and those from the acronym-rich constellation of international organisations. For the latter, the Art Cafe was the perfect venue to chill out after a mission to Juba, or to Mogadishu. It was not much different for Hassan and his friends. They too enjoyed the cosmopolitan flair of that part of Nairobi and to share travel stories. Ali, who supplied clothes to many shops in Eastleigh, recounted in awe his recent trip to China, with the Yiwu trading centre “so big that you need a car to move around.” He would spend hours describing and comparing Hong Kong and Guangzhou – Mumbai was not even worth discussing – and praising Dubai, the usual stopover on his way to the Far East, where he regularly relaxed at some relatives’ place. And then there was Mogadishu. Ali had just returned from his first ever trip to the Somali capital. He had looked at the scarred landscape of the city with the eye of a businessman, daydreaming rows of cafes along the waterfront – the Lido – and hotels and restaurants to welcome back the diaspora and future investors. He was invited by his uncle, who had bought a long strip of coastal land and was seeking finance for development. Ali had already contributed 5,000 USD, and Mahmud was about to follow suit. Business, they all agreed, was thriving, particularly for those who knew how things worked in Mogadishu, or, at least, whom to trust. They all fell in this group. It helped to be, as it was their case, from the Hawiye Abgal clan, the same of President Hassan Sheikh Mohamud, elected in September 2012. It was silent collateral: perhaps your powerful clan member would never have to step in to rescue you. But, in case of need, everybody knows that he would not let you down. The success of the African Union Mission to Somalia (Amisom) troops in reclaiming large swathes of Somali territory from Al Shabaab had cleared the way for a wave of investments. Yet, the political situation remained precarious. The militants had only retreated in the shadows and violence could flare up at any moment. For this reason, other opportunities were being explored. In East Africa, Uganda was drawing a growing number of Somalis. Some were refugees, others were Kenyan citizens, and others were Somali citizens who eventually registered as refugees. They flew directly from 4 Mogadishu to Entebbe airport, where they were let in by just producing their brand new blue Somali passport. They would travel back to Kenya or to neighbouring countries, such as South Sudan, Rwanda or the Democratic Republic of Congo (DRC), where they envisaged emerging business opportunities. Alternatively, they would stay in Uganda, benefiting from the lower cost of living compared to Nairobi and relatively friendly refugee policies, and often enrol in the local universities. Somali student associations in Kampala were mushrooming, mirrored online by a proliferation of Facebook pages, where each association fed prospective students (and migrants) with information on how to relocate to Uganda. Clanship also shaped digital spaces. Dispersed Somalis found online their community of belonging and traded information. A broad repertoire of identification devices – ID papers, refugee cards, visas, student cards, and online profiles – widened the range of possibilities. And keeping the door open to possibilities seemed a primary concern in Somali decision-making processes. In both Nairobi and Kampala, I often sat before my Somali friends as they were showing proudly their multiple, often contradictory, papers, like they would spread a deck of cards. “Waxaan ku hal lug ku taagan,” Hassan usually said. “’We stand on one leg”. It’s how it goes: you never know what is going to happen. And if you want to do business you have to move. You have to be ready if you don’t want to get stranded.” I liked Hassan’s metaphor, and I jotted it down in my notebook. Standing on one leg evokes the constant risk of falling, but there was no hint of fatalistic acceptance in his words. The precarious balance engenders a ceaseless tension, akin to the labour of a ‘nervous system’, to borrow Taussig’s (1992) suggestive image: an entity – or, to say it better – a reticular structure suspended between “disordered order” and “ordered disorder”, sensitive to external signals in “the normality of the state of emergency” (17). Standing on one leg enshrines a movement that has not been performed yet, but also the rationale for the act of moving: the awareness that the next step has to be anticipated, in order to find a steady foothold and thus keep standing. While, as they insisted, the Fate is firmly in the hands of God, it was still possible to walk the tightrope dividing a risk from a gamble, and what is halal, allowed, from what is haram, forbidden. Sharing was the key, for one’s jump into the unknown is tantamount to challenging the Almighty. But 5 collectively partaking in the risk is to abide by His will, through weaving ties, forging bonds, (re)creating community. “It is all about trust”: this was a constant refrain. Hassan, Ali, Mahmud and many others insisted that trust permeates Somali social relations. Imbued with Islamic values, it turns ‘gambling on people’ into the fulfilment of a broader project: keeping Somalia alive in a transnational space. And yet, I heard ‘trust’ mentioned so many times, during my fieldwork, that, at some point, the very word began to sound hollow. It did sound hollow to Omar’s ears. “There is no trust among us,” once he said sharply. “When one of us gets resettled, he forgets about those left behind. And he disappears.” Omar was not used to going out to Westland to sip cappuccinos. He never stepped foot into the Andalus, although he also lived in Eastleigh with his wife and five children. He was from Qoriooley, in Lower Shabelle, near Marka, and had only two papers: a UNHCR refugee mandate and a Kenyan alien card. When I first met him, he was teaching Somali and English in a language school in Eastleigh’s 4th Street. Later, he took over another small school and became a supervisor. The school was in a rundown detached house in the backyard of the electronic repair workshop of his father-in-law. We usually met there among disassembled TVs and radios. He had received a basic education while living for eight years in the Dadaab refugee camp and earned a living by teaching English. Most members of his community in Kenya were in Kakuma refugee camp, and he had no one to rely upon for support or relatives in business to partner with. He was from Somalia, as most in Eastleigh. But the ‘us’ he referred to pinpointed simultaneously his group of belonging and the racial category in which he was entrapped. He was a Shidle from the farming communities located along the rivers Shabelle and Juba in Southern Somalia. In the 1990s, USAID and UNHCR identified the members of these communities as a particularly vulnerable group among the larger Somali refugee population. They lumped them together under the label Bantu, and targeted them for a special resettlement program to the US. Instead of Bantu, Omar referred to himself as a Jareer, a term meaning ‘hard’ or ‘curly hair’, as opposed to Jileyc, or ‘soft hair’. The latter term refers to the physical feature through which he, and the other members of his 6 community, identified Darood, Hawiye, Isaaq, and Dir – the pastoralist groups that had historically wielded power across the Somali territories. Although he had many students from dominant Somali clans – especially young women waiting to be reunited with their spouses who had already resettled abroad and young men with a yearning to travel and do businesses – he had no strong ties with any of them in Eastleigh. This became clear to me two months into my fieldwork, as he assisted me in my research. Outside his language school, he hung out only with other Somali Jareer, all refugees with very limited access to financial resources. For them there were few borders to cross and few boundaries to blur: they were pigeonholed into rigid humanitarian categories that constrained their mobility. Omar did not share Hussein, Ali and Mahmud’s enthusiasm for emerging business opportunities in East Africa. He had no capital to invest, either in Somalia or elsewhere, and considered Uganda suitable only for those who received support from the diaspora. ‘Standing on one leg’ had for him a different meaning: his balance was precarious and his room for manoeuvre very restricted. His idea of mobility was constrained by a condition of ‘bare life’ (Agamben 1995), in which desires were silenced by contingent needs. The contrast with the life experience of other Somalis in Eastleigh was glaring. One day, I came to his school with another Somali friend. His name was Mohammed and, although he was officially a Kenyan citizen, once he told me that he was born in Galkayo, Somalia. Earlier, I had accompanied him to a money transfer operator (MTO) where he withdrew 2,000 USD ahead of a business trip to Mozambique. Once the money was in his pocket, he explained to me that it came from his cousin, who ran an NGO in Somalia. It was simple information, yet puzzling enough: I had always assumed remittances to flow in the opposite direction. Later, when I introduced them to each other, Omar added further details, and I wondered if I had made a mistake to bring them together. Sitting in a room surrounded by piles of pupils’ chairs, I witnessed one of those moments in which Somalis ‘read’ each other without producing any paper, and almost without words, but by simply resorting to signs invisible to a stranger’s eye. They scanned each other, identified their respective complexions and physical features, tapping into an unwritten knowledge that defined the perimeter of their belonging, rooted less in a common present that in a shared past in which survival could depend on understanding whom you were facing. Omar 7 lingered on Mohamed’s teeth. He traced the dark pigmentation of the dental enamel to iron-rich sources of water, which Somalis associate with Mudug region. This geographical hint helped him navigate through the map of Somali clan structure, and the underlying power relations. “These people are occupying our land. They are killing us!” he said, looking at me and pointing a finger at Mohammed. He referred to him using the plural form, as Mohammed was standing there not as an individual but as a representative of his clan, Habr Gedr. Mohammed laughed as he had just heard a joke. Omar grinned, breathing slowly. I said that I had to rush somewhere else and asked Mohammed to join me, lest the situation became further inflamed. Reflecting later on that moment, I realised that it was not just awkward. It was for me what Elyachar (2010a) calls an “aha” moment, “when something goes against what your theory tells you should be there, when you know something is important but you don’t know why or have the concepts to explain it” (1). In the course of my fieldwork, I would come across other aha moments that created a cognitive dissonance, begging reflection on the way conflicting narratives of solidarity and exclusion relate to each other. Not to produce sweeping narratives, but to shed light on the way legacies rooted in the Somali pre-colonial space, linkages with heterogeneous actors and localities defined by specific institutional configurations conspire to shape social networks and to reproduce and transform both relations of solidarity and subordination. 8 Introduction My thesis examines dynamics of inclusion and exclusion within Somali networks in East Africa. It focuses on Somali mobility patterns and financial practices to draw insights on the maintenance, reproduction, and transformation of both solidarity ties and inequalities. By examining Somali communities in Kenya, host to the largest Somali refugee population outside Somalia, and Uganda, an increasingly important recipient of Somali refugees and migrants, this thesis seeks to understand how mechanisms of social stratification rooted in Somali socio-cultural structures are reproduced in mixed migration flows encompassing both forced and voluntary migrants. It analyses sets of relationships whose continuity and changes are regulated by the interaction of structure, agency and institutions and argues, on the one hand, that networks are dominated by groups who hold sway over economic and political resources, precluding others from accessing key assets that may help challenge relations of subordination; on the other, that pre-existing inequalities hinder on the capability to move across both physical and institutional categories. These inequalities can be traced back to asymmetric clan relationships shaped by Somali historical trajectories before and after the implosion of the state. However, it suggests that kin relationships only partially explain why and how bonds are sustained and forged. Instead, by observing the mechanisms that animate networks, reproducing both solidarity and marginalisation, it teases out how new linkages are created and how Somalis communities accommodate to specific institutional settings, either adapting to narrowing windows of opportunity or maximising the benefits that may be yielded from their widening. My thesis unfolds along two converging routes: one clarifying how members of groups who were able to accumulate strategic resources before the collapse of the Somali state in 1991, and to entrench their position of dominance during the ensuing civil war, 9 succeeded not only in surviving but also in developing a greater resilience to changing circumstances and even thriving; the other, explaining how members of Somali minorities have failed to escape a condition of economic and political marginality, exacerbated throughout Somali colonial and post-colonial history and laid bare by the eruption of a civil war. Therefore, this thesis examines the calculative rationalities and the decision- making processes underlying recent Somali mobility patterns, unpacking practices and institutions aimed at fostering social ties and expanding trade networks, and embedded in a culture that holds in high esteem adroitness in business. Simultaneously, it fleshes out the relations of subordination and, often, exploitation and control lurking behind narratives of trust, which punctuate most discourses on migrant networks, to relate the access to resources that enable mobility and escape from marginalisation to intra-Somali racial dynamics. The central issue of my thesis reflects Massey’s (1991) observation that It is not simply a question of unequal distribution, that some people move more than others, and that some have more control than others. It is that the mobility and control of some groups can actively weaken other people (240) To paraphrase the title of a recent collection of essays on mobility patterns in sub- Saharan Africa (hereafter Africa), “Mobility makes States” (Vignaswaran and Quirk, 2015), the thread running through this thesis is the argument that ‘mobility makes power’. Mobility contributes not only to accessing and mobilising strategic resources but also to shaping processes of social stratification. This thesis explores how a plurality of flows (Somali forced and voluntary migrants, money and information) interact, outlining the contours of socio-economic spaces in which transnational dynamics intersect with national and international migratory regimes. Rationale of the study and research questions This topic boils down to three key questions:  What are the dynamics shaping Somali social networks in East Africa?  What is the relationship of mobility and social stratification in these 10 networks?  Why have some groups fared better than others in securing access to resources that enable entrepreneurial activities and mobility across borders? Answering to these questions sheds light on the rift in Somali networks between dominant clans - key units in Somali politics and society - and groups with a long history of social and political marginalisation in Somalia. The reference is here to farming and agropastoralist communities from the middle and central Somalia riverine regions of Juba and Shabelle, known in Somalia as Jareer and lumped together during the 1990s under the category of Somali Bantu. My multi-sited ethnography was conducted in two locations in Kenya (Eastleigh’s neighbourhood in Nairobi and Kakuma refugee camp in Turkana county) and one in Uganda (Kisenyi’s neighbourhood in Kampala). It examines the everyday practice of living ‘standing on one leg’, asking in which way mobility patterns are linked to processes of social stratification. Mobility has a specific importance for Somalis, in the light of their unique socio-cultural features, their protracted humanitarian situation and the events which have unfolded in the region during my fieldwork. Since the collapse of the Somali state in 1991, Somali networks have spread across several countries, in refugee camps and urban sites. Factors at macro-level - such as the so-called War on Terror - have interacted with national refugee regimes and local dynamics, shaping the refugee experience. At the same times, actors have not simply reacted to external circumstances but have furthered their desires and life projects. These strategic choices, finalised to fulfil vocational aspirations and projected into mobility patterns, are buttressed by the far-flung Somali money transfer infrastructure, based on a set of financial practices known as hawala. These practices provide a critical lifeline for the Somali diaspora, which enables the transfer of remittances and business capital even in areas cut off from formal financial circuits, such as Somalia and refugee camps. The significance of hawala, however, goes beyond its economic dimension: this institution 11 allows the cultivation of relationships that are vital to survival and advancing aspirations. It is often described by Somali as the backbone of a community scattered across the world and, yet, still rooted in the homeland and steeped in ideals of ethnic and faith-based solidarity. As this thesis argues, this rosy narrative is often window dressing of a more complex reality. However, capturing power relationships and the reproduction of asymmetries is often neglected in studies of diasporas and migrant networks. Methodology This thesis stems from a contingency plan. This is a useful preamble to understand how the context shaped the development of the research project and its roll out. I had conducted a preliminary fieldwork in Eastleigh in May-June 2011, before deciding to focus on Somaliland for my doctoral research. At my transition board, I thus discussed a paper on the technological innovations in the remittance industry in Somaliland, with the aim of looking at the impact of mobile money on established informal financial systems. At the end of the discussion, I was asked whether I had envisaged a contingency plan to face whatever circumstance could impinge on my project. I did not and, to be honest, back then, it did not seem such a pressing issue to me. However, two months ahead of the planned beginning of my fieldwork, the British Foreign and Commonwealth Office (FCO) issued advice against all travel to Somaliland due to the threat of terrorist attacks. The University of Edinburgh ethical committee contacted me and made clear that, given the circumstances, I had to opt for a different field of investigation and rethink my project. I needed a plan B. Returning to my original project on Eastleigh, to study the impact of mobile money on the local trade, at first looked like a makeshift to me. Although my preliminary fieldwork had provided me with an insightful introduction to the fascinating reality of Somali money transfer operators, I had subsequently decided to focus on the heart of the Somali money transfer business in Somaliland. But, over time, it dawned on me that it might be a prism through which I could revive my project. Terrorism struck 12 (as a matter of fact, not in Somaliland but in Kenya, where I was supposed to be safer), triggering a clampdown by the Kenyan security forces on the Somali population in Kenya, and having a plan B became a universal need for Somalis in Eastleigh. I went back to Nairobi to spend at least one year there and, during that period, as I looked at the way people navigated uncertainty, I gradually realised that I had to accommodate my own strategy to the changing scenario. Eventually, my fieldwork was in total thirteen months, divided into three stints; not only in Eastleigh, where I was based for nine months, but also in Kampala, Uganda, where I stayed for three months, and Kakuma refugee camp, Kenya, where I spent one month. I like to think of my strategy as a constant compromise between agency and serendipity, punctuated by turning points which acted as ‘instigators to thought’, resounding Guyer’s (2013) epistemology of surprise. I often retrospectively looked at what Kesserling (2015) defines ‘stumbling’ (4), moments which prompted me to call into question my assumptions and readjust my methods as I realised that I was unable to anticipate what was going on around me. In the course of my thesis I describe the events that precipitated the securitisation of the Somali population and thus faced my informants with some crucial dilemmas, whether to stay or go, and, if the latter, where to. It was not much different for me, as a spate of terror attacks, an indefinite threat looming on Eastleigh ahead of the Kenyan elections in March 2013 and eventually a major attack, namely the raid of the Westgate shopping mall in September 2013, forced me to reconsider my initial idea to investigate how mobile money interacted with hawala to reshape Somali refugees’ livelihoods in Eastleigh. This aim encapsulated three main assumptions: that refugees could be easily sifted out from the broader Somali population in Eastleigh; that Nairobi’s ‘Little Mogadishu’ was a bounded field of investigation; and, naively, that Somali mobile money agents and hawala operators would have willingly disclosed their books to me. However, these assumptions were shattered by the encounter with the empirical reality. This lead to the decision to rethink my initial focus on refugees; to adopt a multi-sited ethnographic approach; and to use qualitative methods of data collection based on interpretative anthropology to grasp how 13 the different populations within the broader Somali community at the centre of my investigation make sense of their world. Choice of the sampling Unlike most studies conducted among Somalis in Eastleigh, my approach was defined by a minor emphasis on the separation between Kenyan Somalis and refugees. My preliminary proposal was drafted on the basis of previous scholarly research (Horst, 2006; Campbell, 2006; Lindley, 2007, 2010a; Lambo, 2012) and grey literature (Pavanello et al. 2010; Simpson, 2010). Most of this literature was focused on refugees (with the exception of Carrier and Lochery, 2013) and driven by policy concerns. However, as Bakewell (2008) points out, having academic researchers relying too much on policy categories “tends to obscure and render invisible some population groups, causal relationships, and questions that are methodologically difficult to capture” (433). Instead, given the importance of interpersonal relationship beyond formal juridical status, I deemed necessary to question the institutional and social significance of the labelling process through which, while making readable its population of concern (Zetter, 1991, 2007), the humanitarian apparatus turns labels into resources to pursue according to the context. And the one in which I was collecting my data was indeed a context which challenged institutional taxonomies. The procedure to select a sample of refugees in Eastleigh usually entails to rely on national NGOs, such as the Kenyan Refugee Consortium (KRC) or the Nairobi Urban Refugee Protection Network, or international NGOs (INGOs), such as Refugee International (RI) or Danish Refugee Council (DRC), in order to have access to refugee grassroots associations which are ubiquitous in Eastleigh. However, some people who initially introduced themselves as Kenyan citizens eventually turned out to be born in Somalia. Others were Kenyan citizens in Kenya but became Somali refugees upon moving to Uganda. Significantly, I began questioning the usefulness of rigidly separating Somali refugees from Kenyan Somali citizens and Somalis with foreign 14 passports as I grew aware of another kind of split within the broader Somali community, and one I had no anticipated. I was first introduced to the field by my Somali teacher, Omar, a Somali refugee who identified himself as Jareer and also became my first research assistant. He helped me single out six key informants and assisted me in the meetings with interviewees, identified through snowballing. I was initially reluctant to think in terms of clans, fearing to essentialise clan membership, and also because the aspect of inequalities among clans seldom featured in the relevant literature on Eastleigh. However, a month into my fieldwork, I realised that Omar was linking me up almost exclusively with other Somali Jareer, living on menial jobs and petty trade, with limited access to remittances and to business opportunities. The only moments I interacted with non-Jareer businesspeople were at the English classes Omar held in the school in which he worked. I had the feeling that his reluctance was less related to an interest in advancing a specific clan agenda than to a general uneasiness in dealing with Somali Darood and Hawiye in public places. He was mostly intimidated by them, but sometimes, as in the vignette I describe in the prologue, he could become quite outspoken. Subsequently, in March and April 2013, I contributed as a research assistant to a project of The Programme in Comparative Media Law and Politics (PCMLP), University of Oxford, “Talking politics: a comparative analysis of Somali radio call-in programmes”. The project, focusing on Radio Star FM, which I will tackle in chapter 4, enabled me to reach out to a part of the Somali community that I had been hitherto unable to access. It was the Somali upper class of Darood and Hawiye entrepreneurs and encompassing refugees, Kenyan citizens, and Somalis with foreign passports. As I began wondering whether there was a relationship between the greater diversity of juridical statuses within a subgroup and its economic performance, I decided to draw inspiration from the approach to formality and informality and focus on the linkages among Somalis with refugee cards, Kenyan, and foreign passports. This approach highlights the difference between units of observation and units of analysis, two often confused notions. As Sedgwick (2014) points out, The unit of observation, sometimes referred to as the unit of measurement, is defined statistically as the “who” or “what” for which data are measured or collected. The unit of analysis is defined statistically as the 15 “who” or “what” for which information is analysed and conclusions are made (1) In this thesis, the units of observation are Somali refugees, Kenyan Somali traders and politicians, Ugandan Somali traders and Somali passport holders. The units of analysis are Somali networks. Multi-sited ethnography The volatile context in which I conducted my research, and the awareness that heterogeneous juridical statuses coexisted and shaped different mobility patterns, also influenced my decision to adopt a multi-sited ethnographic approach. My pathway towards this decision echoes what Falzon (2009) calls the multi-sited “little ‘road to Damascus’ storyline”: “As originally planned, my fieldwork was to be conventionally single-sited; after some time on site, however, an epiphanic moment revealed to me that this was inadequate; I therefore chose to move around” (12). Only once in Eastleigh I took the decision to include in my fieldwork also Kampala and Kakuma refugee camp. Initially theorised by Marcus (1995), the multi-sited approach has become in the last decade a fixture in the ethnographic literature on migration, as it moves beyond the ‘methodological nationalism’ that, according to Wimmer and Glick-Schiller (2003), tends to regard the nation-state as the natural unit of analysis to study social phenomena, despite the transnational reality of migration routes. There is no universal consensus on the usefulness of doing fieldwork in various sites linked to each other by social relationships and observing exchanges and flows, as prescribed by the multi-sited approach: Hage (2005) for instance, argues that the notion of a single geographically discontinuous site proves analytically more useful than the idea of multi-sitedness. Reflecting on “being there... and there...” (Hannerz, 2003), however, suggests that the sites are not bounded spaces but, at the same time, they keep their specificities. In the light of the local differences, strategies of transformations and adaptation are laid bare. Thus, teasing out the links between Eastleigh in Nairobi, Kisenyi in Kampala and Kakuma refugee camp highlights both continuities and discontinuities in Somali networks. 16 Interactions with the informants The particular period of my fieldwork impacted upon security constraints for me and influenced my relationship with the informants, some of whom were particularly wary of being on the radar of media and intelligence. This was mainly the case for hawala agents, whose activities, often unlicensed, regularly arouse suspicions of money- laundering in financial regulators. Upholding confidentiality as a professional virtue, especially as a large number of Somalis were using hawala to transfer money abroad (to Uganda, Somalia or Dubai), they were quite reluctant to disclose the figures of their business, keeping their ledgers well away from my gaze. However, after a laborious trust-building process, financial operators and traders eventually became willing to discuss their views on Somali financial habits. In some case, my own Italian background unexpectedly struck a nostalgic chord, as many senior Somalis from Southern Somalia were particularly keen to retrieve the Italian language they had learned in their early years, when they were part of the school system of the Somalia Italiana. They played a crucial role in negotiating access on my behalf to their more suspicious and less sentimental countrymen. Interactions with Somalis in Kampala were easier, during my fieldwork, scores of Somali refugees were arriving daily in Uganda, a flow that characterised a situation in constant flux. Like other studies on networks straddling international and institutional boundaries, which often operate beyond the oversight of the law, my research was exclusively based on qualitative methods of data collection (MacGaffey and Bazenguissa-Ganga, 2000). Because of the peculiarities of my field of investigation and of the time of my fieldwork, the data were collected piecemeal over the course of several protracted conversations with financial operators, conducted mostly outside their working locations, in cafés, restaurants, hotel lobbies but also cars and shared cabs. Significantly, the only informants who agreed to meet in their homes were Somali Jareer, who felt uncomfortable discussing their specific situation with a gaal (a white foreigner) in a public space. However, in some circumstances, we also met outside Eastleigh. Some 17 meetings were planned, others were accidental, and I should admit that the latter were often the most fruitful ones. Therefore, rather than tallying the number of the interviews that were conducted, it would be fairer to say that, among all the persons I have engaged with during my fieldwork, around 120 of them offered the contributions that best fitted my topic of investigation. These contributions often took forms other than face-to-face interactions. The usage of different ICTs during the data collection and analysis is also worth mentioning. The conversations with my informants continued at a distance, via mobile calls, SMS, emails and messages on Facebook and Skype. During my fieldwork, I made extensive usage of mobile communication to coordinate my movements and to cross-check data. Interestingly, often face-to-face meetings were followed by SMS exchanges in which the interviewee expressed intimate opinions or disclosed information that he or she did not feel comfortable to communicate in person. This often happened with the young women I interviewed, mostly under the supervision of a male relative, or in sensitive settings such as Kakuma refugee camp. Organisation of the thesis My thesis is structured along the organising principles of legacies, linkages, and localities. Chapter 1 presents the strands of literature I draw from to fashion my theoretical framework, revolving around the concept of networks. Chapter 2 and 3 discuss the social and historical background of the Somali networks in East Africa. Chapter 2 illustrates Somali social structures and history until the collapse of the state in 1991. It describes how inequalities in Somali networks were the outcome of processes of social stratification that can be traced back to pre-colonial Somalia and were accelerated by political-economic factors during colonial rule and the post-Independence era. Chapter 3 examines how the eruption of Somalia’s civil war further entrenched relationships of dominance, and focuses on the interplay of conflict, crisis, and mobilities. It argues that the conflict that broke out in the 1990s widened the gap between dominant 18 clans, in particular Darood and Hawiye, and minorities, especially Jareer farmers. It ends with an overview of the refugee policy framework in Kenya. Chapter 4 and 5 emphasise the importance of creating linkages with new actors and explore the human and financial infrastructure of the neighbourhood of Eastleigh, Nairobi, the main hub in East Africa’s Somali networks. Chapter 4 in particular analyzes the emergence of a Somali community in the neighbourhood and argues that the dynamic economy of Little Mogadishu is the product of an ongoing conversation among diverse business classes, including traders cultivating transnational connections but also street- level entrepreneurs. Chapter 5 integrates this argument by examining the financial infrastructures which facilitate interactions and partnerships and argues that Somali financial institutions have to be looked at as sites in which different forms of capital are transformed into each other. Chapter 6 and 7 examine the significance of localities in defining the orientation of socio-economic networks and driving mobility-related decision-making processes. Chapter 6 focuses on the recent emergence of a Somali route to Uganda. It also discusses synergies between the state and Somali networks to highlight the convergence of interests between institutional actors and Somali community organisations. Chapter 7 discusses a contrasting case and, by focusing on Kakuma refugee camp, in northwestern Kenya, illustrates how institutional frameworks can be hijacked and manipulated by stronger members of the network to preserve positions of dominance and control. It illustrates how the ‘racialised space’ that marginalised Somali farmers fled in Somalia is being reproduced in humanitarian spaces. 19 1 Literature review and theoretical framework My thesis locates the analysis of power relationships in Somali networks in East Africa within the broader context of mobility in Africa. Therefore, this chapter reviews theories and evidence from the scholarship on mixed migrations and low-end globalisation hubs in urban Africa and narrows the focus on migrant livelihood strategies between formality and informality. It also considers the current approaches to informal networks in order to define a theoretical tool that enables to account for continuity and changes and shed light on the reproduction of power relationships. In sifting through these different strands of literature, the gaps addressed in the course of the thesis are identified. Discussing mobility The recent interest in mobility (or, better, mobilities) stems from the interdisciplinary debate on globalisation that has grown popular in the past twenty years (Bauman, 2000, 2002; Harvey, 2006; Hardt and Negri, 2000; Inda and Rosaldo, 2002). This literature often verges on a ‘nomadic metaphysics’ (Cresswell, 2001) that emphasises forms of deterritorialisation (Deleuze and Guattari, 1972) and matches the dichotomy between domination and emancipation with the one between immobility and mobility, framing mobility as an act that challenges spatialised structures of power. It takes stock of the reflection on major political, economic and technological changes at the intersection of the global and the local, and explores how flows and places shape each other, what synergies and tensions emerge from this interaction, and how differences and inequalities within and between flows are (re)produced. 20 Transnational studies have emerged as a scholarly field to tackle the sedentary bias ingrained in social science, which frames mobility as the disruption of a social order associated with dwelling a place within fixed national boundaries (Hannerz, 1996). Instead, the transnational discourse explores how technological and institutional assemblages shape a ‘common - however virtual - arena of activity’ (Vertovec, 1999: 447) in which ‘transnational habitus’ are cultivated (Vuorela 2002). As summarised by Faist (2010), transnationalism thus “refers to sustained cross-border ties, events and processes across the borders of several national states” (1667), and contemplates the engagement of both state and non-state actors. The notion of ‘flows’ has become a trope in the transnational literature on mobility (Appadurai, 1996; Castells, 1989; Hannerz, 1997; Rockefeller, 2011), conveying the sense of liquidity and ephemerality that permeates the postmodern debate and fascinates popular accounts. This debate raises questions on how to make sense of the variety of interactions between local peculiarities and global dynamics. Appadurai (1996), for instance, suggests that these flows shape five different and coexistent types of landscape revolving respectively around people (ethnoscapes), media (mediascapes), technology (technoscapes), finance (financescapes) and ideas (ideoscapes). Hannerz (1997) points out that the very notion of flow implies both mobility and transformation in time, the latter not necessarily corresponding to moving from one location to another. Hence, a transnational approach should aim at explaining, on the one hand, how new social formations come into being and keep on reproducing themselves; on the other, how pre-existing institutions are re-defined by collective actors emerging across borders (Faist, 2010). These flows overlap and sustain each other, suggesting interrelations and mutual shaping (Heyman and Campbell, 2009). Particularly relevant to my study is the interweaving of financial and human flows, and the way their interactions are embedded in places, or ‘spatial fixes’ (Harvey 2001; Sheppard, 2002) acting as both spatial footholds and stepping stones of migration and capital accumulation projects. Therefore, although the emphasis on mobility has contributed to the “theoretical shift from place to space” (de Bruijn and Brinkman, 2012: 47), localised institutions remain preponderant in modulating flows, both financial (Thrift 1996; Leyshon 1995, 1997, 1998; Hall 2012) and human (Van Hear, 21 2010; Bakewell et al., 2012; Åkesson, 2013; Glick Schiller and Salazar, 2013; Flahaux and De Haas, 2014). In these spaces, new cultural and power arrangements emerge from “heterogeneous and unequal encounters” (Tsing, 2005: 5). What makes these encounters possible, in Tsing’s view, is the presence of ‘productive frictions’, attritions derived from socio-political configurations that impinge on mobility. The significance of ‘sticky’ places Challenging the oversimplified image of a seamless world of flows, Tsing (2005) instead suggests looking at impasses and interruptions that specific populations experience in ‘sticky places’. She hints at the significance of ‘waiting areas’, or spaces of transition (ibid.). Indeed, a thorough understanding of new mobility paradigms cannot leave aside the examination of the factors underlying the coexistence of “zones of connectivity, centrality, and empowerment in some cases, and of disconnection, social exclusion, and inaudibility in other cases” (Sheller and Urry, 2006: 210). Global flows proceed along the ‘fast and slow lanes of social life’ (ibid.: 213 ), an effective metaphor for the inequalities in the way different populations benefit from mobility opportunities. Institutional frameworks play a crucial role in shaping mobility patterns, defining “new forms of movement control and unequal rights of mobility” (Heyman and Campbell, 2009: 134). Curdled by ‘control’, ‘inequality’ and ‘state power’, the flows at the centre of the transnational discourse thus appear in constant tension with geographic locations and national sovereignties (Salter, 2008). While technological innovations help improve coordination and the circulation of both people and information, they also play a critical role in preventing or restraining mobility, as it is evident in the increasing use of biometric border management systems and passports (Cunningham, 2004). National policies or supranational entities craft different mobility regimes, and contribute to the construction of mechanisms of social stratification that reflect the complex internal dynamics of mobility phenomena, where multiple regimes, speeds and degrees of freedom coexist and underpin each other. Glick Schiller and Salazar (2013) suggest examining the articulation between “the privileged movements of some and the 22 co-dependent but stigmatised and forbidden movement, migration and interconnection of the poor, powerless and exploited” (188). Authors who regard mobility as a relational process, embedded in the material reality, reject sweeping narratives of time-space compression and point at both the differences in the experience of mobility and the underlying relations of dominance and subordination (Cresswell, 1996; Adey, 2006). Massey (1993), for instance, stresses the fact that not only the resources enabling mobility are unequally distributed, but also that, among those who experience mobility, “some are more in charge of [mobility] than others; some initiate flows and movement; some are more on the receiving end of it than others; some are effectively imprisoned by it” (61). Cresswell (1996) adds that mobility is not just relational but also contextual, depending on particular socio-political settings and best understood in the light of a specific culture and history. The entanglement of heterogeneous ‘regimes of mobility’, where the great freedom of movement of some actors contrasts with the limited agency of others, thus points to the need to look at social processes that produce enclosures, constraints and control and at the institutional devices, such as visa and passports, used by states to bestow legitimacy to mobile actors as they cross international borders (Wang, 2004). Ports of entry are places in which the rhetoric of unbound flows, which permeates the literature on globalisation, is empirically dispelled as inequalities in movement become more evident (Kearney, 1991; Salter, 2008). Acting as a catalyst of aspirations and fears, particularly in the post-9/11 discursive framework (Cunningham, 2010; Andreas, 2003), borders have become the object of growing concern in political arenas. As such, they can be regarded not only as markers of political sovereignties but also as sieves to separate the included from the excluded, and as projections of policies that act “as classificatory devices with contextually-encoded meanings; as narratives that justify certain courses of action while disabling others; and as discursive formations that empower some voices while silencing those of others” (Baba, 2013: 3; see also Shore and Wright, 2011). This is achieved through the socio-political construction of categories such as ‘illegal’ or ‘deportable’ (De Genova 2002) and translated into the empirical realm through the deployment of apparatuses of surveillance, detention and 23 deportation (De Genova and Peutz, 2010; Fassin 2011). The regulation of flows at borders is thus defined by the combined effect of material and juridical classificatory devices, which brings to the fore the gulf between a ‘kinetic elite’ (Salter, 2008) and a ‘deportation class’ (Walters, 2002): an elite able to travel freely and a population forced to move across shadow circuits. Interestingly, the humanitarian sector often dovetails with state security apparatuses through supranational organisations (such as the UNHCR) or international NGOs (INGOs) by managing identification and detention camps, which act both as filters and as places of ‘permanent precariousness’ (Agier, 2011: 71). The contradictions intrinsic in the contiguity of security and humanitarian institutions stem mainly from the fact that the institution of the refugee camp, and in general all practices of encampment, fulfil an ambiguous role, as they simultaneously provide basic assistance and control movements (Harrell-Bond, 1986). In so doing, humanitarian agencies participate actively in redrawing borders that are “everywhere that an undesirable is identified and must be kept apart, ‘detained’ and then ‘expelled’” (Agier, 2011: 50). Fuelling politicised anxieties and tensions over migration flows, the literature on borders has increasingly been tackling the impact of migration policies on social exclusion (Baba, 2013), highlighting how enclosures not only signify inequalities but actively contribute to their reproduction, as the right to mobility paves the way to a broader set of rights. Unearthing the mechanisms through which this takes place contributes to addressing a key question in the entire debate on connectivity and control in a global setting: “which groups are benefiting, which losing?” (Susser, 1996: 40). African Migrations in Africa Although mobility has been a core feature of African livelihoods and coping strategies since pre-colonial times, the last decades have seen an intensification of internal and transnational flows (Vigneswaran and Quirk, 2015). Recent research on African migrations (primarily motivated by European policy concerns) has challenged many assumptions about trajectories and drivers of human flows originating in the continent (Flahaux and de Haas, 2014; Bakewell and Jόnsson, 2011; Berriane and de 24 Haas, 2012; Bredeloup and Pliez, 2005; De Bruijn, van Dijk and Foeken, 2001; Lessault and Flahaux, 2013; Schielke and Graw, 2012; Whitehouse, 2012). Particularly significant are the works that examine intra-African migratory flows (Schoumaker et al., 2013; Sander and Maimbo, 2003; Bakewell and Bonfiglio, 2013; Bakewell and De Haas, 2007). This literature emphasises the complexity of migration flows in Africa in terms of trajectories, mobility patterns, and migration drivers. It also challenges a popular narrative, underpinned by the grim images of migrant shipwrecks and fear mongering rhetoric warning of a ‘swarm’ of migrants, to use an expression of the British Prime Minister David Cameron (BBC, 2015), from the Southern shore of the Mediterranean sea, according to which African migration flows are all directed to Europe, and driven by despair. As Schoumaker et al. (2013) show, while is true that the number of Africans on the move has steadily increased in the past fifty years, still in 2000 the overwhelming majority of the African migrant stock, around 70% of the 17.5 million African migrants, remained in Africa (4). Conceptual dichotomies, such as the one between internal and international flows (King and Skeldon, 2010), or voluntary and forced migrations (Zetter, 2007; De Haas and Bakewell 2007; Bakewell, 2008, 2010, 2011) have become more problematic. The notion of ‘mixed migration’ reflects the growing difficulty in disentangling refugee and migrant flows (Linde, 2011; Martín and Bonfanti, 2015). This difficulty is mostly due to the fact that these flows are read in terms of ‘population of concern’ by international organisations and institutional actors.1 This distinction often fades at analytical and empirical level, not only because refugees 1 According to the 1951 Refugee Convention, a refugee is someone who "owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality, and is unable to, or owing to such fear, is unwilling to avail himself of the protection of that country." (UNHCR, 2014) As for the term ‘migrant’, the International Organisation of Migrations (IOM) acknowledges that “at the international level, no universally accepted definition for "migrant" exists. The term migrant was usually understood to cover all cases where the decision to migrate was taken freely by the individual concerned for reasons of "personal convenience" and without intervention of an external compelling factor; it therefore applied to persons, and family members, moving to another country or region to better their material or social conditions and improve the prospect for themselves or their family.” (IOM, 2014) 25 and migrants mostly travel along the same routes, but also because they may change status in the course of their journey, as it is the case of refugees travelling from the host country towards another destination as migrants, or migrants seeking shelter in humanitarian facilities, or simply concealing their status. However, according to Bakewell (2011), the designations of migrant and refugee do have an impact on the material life of the people falling in each category as they define, at least formally, their relationship with institutional actors. Therefore, while the policy relevance of these definitions is not denied, greater attention should be paid to the way people move between these two categories. As I shall argue, in Kenya the situation is further complicated by the possibility for many migrants and displaced persons from Somalia to access the Kenyan citizenship by mobilising pre-existing social networks. In general, a major rift emerges in the way different groups negotiate their position within what Malkki (1995) calls ‘the national order of things’. Protracted displacement, defined “both in terms of a prolonged subjective experience of disenfranchisement in exile and as a reality with juridical implications” (Horst and Grabska, 2015: 3), is not experienced by all displaced persons in the same way. Moreover, despite media accounts placing emphasis on conflict as the major mobility driver in the continent, recent research has made a case for a more nuanced view of migratory phenomena in Africa. As argued by Bakewell and Bonfiglio (2013) in their study on migrations in the Great Lakes region, conflict-related displacement is one among many factors which lead to the decision to migrate. In fact, according to the UNHCR, refugees and people living in “refugee-like situations” account for only 14% of the over 17 million African migrants in Africa (UNHCR, 2011). Other factors, including work, study and marriage are considered when decisions to move are made and pursued. At the same time, even the drivers of ‘survival migrations’ (Betts, 2013) triggered by conflicts, persecution or environmental catastrophes or degradation may evolve as a result of changing circumstances, both at home and in the place of resettlement. Highlighting the coexistence of different forms of migration movement and the possibility, especially in the African context, for many to blur the boundaries and move across categories (King, 2012), this literature has stressed the need to account 26 for multilayered decision-making processes and for the variegated range of factors that drive people’s decisions to migrate. As Bakewell (2010) points out, a main issue of debate in migration studies is on the articulation of agency and structure. I will return to this point later in this chapter when discussing different perspectives on networks. But it is important to note here that, from interpretations of migrations based only on push-pull factors and considering the “rational comparison of the relative costs and benefits of remaining at home or moving” (Castles, 2010: 1572), the so-called Harris-Todaro model, these explanations have increasingly taken into account the way “desires, lifestyles, resources, and needs” (Cohen 2004: 19) determine migrant strategies. As Carling (2014) suggests, and as I argue in the course of my thesis, aspirations, generally overshadowed in the analysis of migratory phenomena, play a critical role. Introducing the notion of aspirations into the reflection on migrations is surely challenging because of their non-measurable and often elusive nature. Yet, fine-grained analysis, based on ethnographic accounts, are crucial to understanding how ‘spatial aspirations’ are shaped, or inhibit, by social practices and discourses. At the centre of the discussion there is thus the interaction of “the structural forces promoting emigration in areas of origin and enabling immigration in destinations, the motivations, goals and aspirations of the people who migrate, and the social and economic structures that connect areas of inward and outward migration” (Bakewell, 2010: 1703). Diasporas between security and development The challenge of reconciling these aspects is particularly evident in the discussion on one of the most popular topics within the broader research on transnational migrations: diaspora. Tracing the contours of diaspora studies with respect to the broader scholarship on migrations, Nyberg-Sørensen (2007) points out that “migrants’ transnational practices have been understood to dissolve fixed assumptions about identity, place, and community, whereas diasporic identity-making has been understood to revolve around attempts to ‘fix’ and closely knit identity and community” (7). Although, according to a much-quoted definition by Tölölyan (1991), “diasporas are the 27 exemplary communities of the transnational moment” (4), the concept of diaspora predates the notion of transnationalism (Faist, 2008). From initially referring to the dispersed Jewish community, the use of the term diaspora has gradually proliferated in the second half of the 1900s (Safran, 1999), prompting many scholars to outline the boundaries of the concept. Wahlbeck (2002) identifies in dispersal, settlement in multiple locations and the idea of a ‘homeland’ the elements that qualify a specific migratory phenomenon as a diaspora. Drawing from Cohen (2008), Bakewell (2008) provides a list of four criteria defining what a diaspora is: “Movement from an original homeland to more than one country [...]; A collective myth of an ideal ancestral home; A strong ethnic group consciousness sustained over a long time [...]; A sustained network of social relationships with members of the group living in different countries of settlement” (11). Initially stressing the notion of ethnic identities, and thus inspiring a cultural studies literature exploring processes of ‘hybridization’ (a buzzword in the studies on diasporas in the 1990s) (Clifford 1992), diaspora and transnationalism scholars have increasingly considered diasporas as heterogeneous and dynamic social realities (Pirkkalainen and Mahdi, 2009; Werbner, 2002). To this regard, Sökefeld (2006) rejects “essentialising concepts of diaspora” based on the assumption that migrants are naturally rooted and connected to the places of origin, arguing instead that diaspora identity and the imagination of a diaspora community is also an outcome of mobilisation processes. The development of diaspora identity is not simply a natural and inevitable result of migration but a historical contingency that frequently develops out of mobilisation in response to specific critical events. Diaspora is thus firmly historicised (280). The acknowledgment of diasporas’ agency has paved the way to their ‘securitisation’ and ‘developmentalisation’ (Isotalo, 2009). Diasporas have been increasingly regarded through the lens of security (of both the origin and the host countries) and development, as critical actors in shaping the relationship between the conflict-stricken and impoverished areas of origin and the countries of resettlement (Pirkkalainen and Mahdi, 2009). Driven by policy concerns, these studies have focused mostly on diasporas in the West, neglecting research on African diasporas in Africa 28 (Bakewell, 2008), and conceptualising migrants and diasporas either as “agents for promoting peace and development”, or spoilers, or as simultaneously ’peace-makers’ and ‘peace-breakers’ (ibid., 5). This latter approach is particularly recurrent in studies on the Somali diaspora, “perhaps one of the most widely studied diaspora communities in the West” (ibid.). These studies take stock of the different forms and directions Somali transnational mobilisations may assume in perpetuating conflict or peacebuilding and development (Horst, 2007; Farah et al., 2007; Hoehne et al. 2011). Somali diaspora groups in Western countries, for instance, have proven able to lobby on issues related to Somalia. This was the case of the Somali Canadian Diaspora Alliance (SCDA), Somali American Peace Council (SAPC) and Somali Diaspora Network (SDN), which have applied political pressure to their governments to ask for the withdrawal of Ethiopian troops from Somalia after the 2006 invasion and for an increase in humanitarian aid. Another example is offered by Somali diaspora groups in Scandinavia organising regular fundraising to support development programs in Somalia (Pirkkalainen and Mahdi, 2009). However, Lindley (2010) suggests moving beyond a ‘conflict/peace dichotomy’, which focuses on “narrower, more formally institutionalised transnationalism” (9), to look also at the rich repertoire of practices which buttress living conditions in conflict areas and keep alive social ties in a transnational space. One of the most important practices is indeed sending remittances to family members to sustain livelihoods or investments. Remitting trust The topic of remittances – the money sent by migrant workers to relatives or communities in the place of origin – is a central concern in the literature on diaspora in general and African diaspora in particular (Oucho, 2008; Pieke et al., 2007; De Haas, 2008; Horst and Van Hear, 2002). According to a report by the Overseas Development Institute (ODI), in 2013 USD 32 billion were transferred to Africa and USD 41 billion are projected to be transferred in 2016 (Watkins and Quattri, 2014: 10). The same study calculates that the region has the world’s highest cost of sending money, to the extent 29 that “Africa’s diaspora pays twice as much as its South Asian counterpart when sending money home” (19). By the admission of the authors, establishing the reasons for such high charges is difficult. However, they identify three main factors driving up fees: limited competition; ‘exclusivity agreements’ between money transfer operators (MTOs), agents and banks; financial exclusion and poor regulation (7). Although, since the end of the 1980s, banking systems across Africa have also been upgraded and integrated, to different extents, into the global financial circuits (Beck and Cull, 2013), the region still lags behind other world’s regions in terms of the rate of people with access to formal financial institutions (Demirgüç-Kunt and Klapper, 2012). However, it is useful to point out that a precise estimate of the amount and costs of remitting money to and across African countries is particularly challenging due to the opacity of the channels through which most remittances are funnelled (Ezeoha, 2013). As Ratha and Mohapatra (2011) and Page and Plaza (2006) remind us, the African remittance market is dominated by informal MTOs, likely as a consequence of the limited reach and the high costs of formal providers of financial services, and resulting in the fact that the region has the highest share of unrecorded remittances in the world. I will further develop this topic when discussing Somali financial infrastructures. Here, it is worth mentioning that remittance patterns are largely discussed in the research on Somalia and Somali communities abroad (Lindley, 2007, 2009, 2010; Hammond, 2013d; Horst, 2002, 2006b; Orozco and Yansura, 2013). The reason lies mostly in the economic significance of remittances for one of the worst protracted displacement situations in the world, where the impact of the contributions, estimated at approximately USD 1.3 billion a year (Orozco and Yansura, 2013), is felt on both short-term humanitarian needs and long-term development initiatives. These studies have argued that remittances are not ‘just money’, but stem out of a rich social texture and encapsulate social relations that are produced by an intense ‘relational work’ (Lindley, 2010a: 141). As the expression of a collective obligation (Page and Mercer, 2012), remittance flows serve as “expressions or claims of membership” (Goldring, 2004: 820). Moreover, as suggested by the concept of social remittances (Levitt, 1999; Levitt and Lamba-Nieves, 2011), migrants remit to the communities of origin not only money but also ideas and 30 behaviours. Therefore, remittances may assume different meanings for both senders and remitters, arising from and eliciting different expectations (Carling, 2014). Other authors, such as Bréant (2013), raise questions on the “depoliticised and consensual rhetoric linking migration and development” (105). His argument – that remittances often benefit individuals and households rather than ‘communities’ – resonates with other views according to which migrant networks and practices do not just provide support but can also accelerate economic stratification and exclusion of some groups within the broader diaspora. At the same time, the relationship of remittances and socio-economic inequalities is shaped by the interplay of other forms of income and immigration regimes in host countries (Åkesson, 2013). Grenier (2011) contends that “the way in which the process of migration, as a key factor of diversification, actually contributes to this growing stratification has rarely been scrutinised in detail.” (607). In general, there is a lack of research examining the mutual strengthening of mobility and remittances on the one hand and pre-existing inequalities on the other. I will return to this topic, which is crucial to my overall argument, below. Digital diasporas Eventually, a fledgling strand of literature within the broader domain of diaspora tackles the role of information and communication technologies (ICTs) in facilitating material and symbolic exchanges at a distance (Conradson and McKay, 2007). The diffusion of ICTs has been a major driver of transformation for diasporas (Gueron and Spevacek, 2008). However, this literature should be seen within the broader discussion on the impact of digital media, particularly mobile telephony, on people's communicative practices in middle-income and developing countries (Horst and Miller, 2006; Ling and Horst, 2011; Molony, 2008, 2012; de Bruijn, 2008, 2010). The sector is booming in Africa more than anywhere else in the world: according to the International Telecommunication Union (ITU), there were 502 million active SIM connections in the region by mid-2013, with a penetration rate of 61% (2013). Mobile telephony has leapfrogged fixed telephony, characterised by poor landline infrastructures, and has 31 spread in a very short span of time through urban and rural areas. This diffusion, driven by the decreasing costs of mobile devices and services, is often seen through a developmental lens (Etzo and Collander, 2010; Economist, 2011), imbued of both Tilley’s (2011) view of Africa as a ‘living laboratory’ and Prahalad’s (2004) notion of the ‘fortune at the bottom of the pyramid’. In general, since telecoms have made their foray in the region, mobile phones have become popular in the most diverse socio- economic settings, with implications for the way people communicate, make a living and move. As “the telephone has grown legs” (de Bruijn, 2008: 19), internal and international migrants have enjoyed greater possibilities to cultivate relations at a distance, rethinking consolidated ways of doing business and staying in touch with relatives (Molony, 2012). Recent ethnographies (De Bruijn et al., 2010; Archambault 2011; Schaub 2012) have discussed the significant impact of mobile telephony on human mobility. Moreover, the recent emergence of mobile money, “a network infrastructure for storing and moving money that facilitates the exchange of cash and electronic value between various actors” (Kendall et al., 2012: 49), is bringing about profound changes in the circulation of people, information and value (Porteous, 2006; Kumar, McKay and Rotman, 2010). This aspect is at the centre of a fledgling scholarship examining how these communicative and financial innovations reshape physical spatial arrangements by interacting with pre-existing practices. I discuss this topic more in detail in chapter 4 and 5. African nodes of low-end globalisation Since Saskia Sassen published “The Global City” in 1991 to chart the emergence of the financial hubs underpinning the neo-liberal globalisation, and concentrated in the rich world, major global transformations, such as the rise of China as a commercial powerhouse, and the diffusion of ICTs, have rendered more complex the geography of economic flows. Following the increased awareness of these changes, the concepts of “non-hegemonic globalisation” (Lins Ribeiro, 2009), “globalisation from below” (Portes, 1997) or “low-end globalisation” (Mathews, 2006) have gained currency in 32 recent years. This paradigm examines South-South connections, and links among trading hubs servicing a low-income global population through the provision of low- cost goods and services. An example of low-end globalisation hub is Hong Kong’s Chungking Mansion, on which Mathews (2011) concentrates his attention, where traders from Africa and South Asia meet and partner for business, catering to consumption patterns at the low-end of the market. Although African entrepreneurs are at the centre of several ethnographies on transnational trading connections (MacGaffey and Bazenguissa-Ganga, 2000; Schapendonk, 2013), there are not many studies on sub- Saharan Africa trading hubs, with recent exceptions such as Lee’s (2014) study on a Chinese trading posts in Southern Africa, and Carrier’s (2015) work on Eastleigh. However, African urban spaces have increasingly emerged as nodes of transnational circuits of migration and commerce (Bakewell and Jόnsson, 2011). Today, Sub-Saharan Africa is the fastest urbanising region in the world, and is likely to double its urban population in the next two decades (Pieterse, 2009). It features a rapidly changing urban landscape, which provides access to services that fulfil the needs and the aspirations of diverse migrant populations, thus enabling strategies to overcome constraints to physical and social mobility. By examining the creative dimension of the exchanges and interactions through which the socio-economic texture of urban Africa is reproduced, a rich scholarship explores the role and the limits of survivalist networks in coping with situations of institutional neglect (Lourenço-Lindell, 2001; Trefon, 2004; Myers, 2011), the interplay of grassroots associative forms and institutional actors in shaping urban governance (Prag, 2010; Lindell, 2010; Meagher, 2011; Goodfellow and Titeca, 2012) and the importance of diversity in economic and cultural processes (Simone, 2004, 2015; Mbembe and Nuttall, 2004; Coquery-Vidrovitch, 2005). Moreover, as Omata and Kaplan (2013) remark, “urbanisation is a major emerging trend in the contemporary world of forced migration” (11; see also Crisp et al., 2012). According to the UNHCR (2013), urban refugees account for more than half of the world’s 11.7 million refugees (64), and the acknowledgement of this reality has elicited a number of studies on how refugees cope with uncertainty outside humanitarian shelters. The importance of these hubs lies in the range of services they offer, from communication centres to MTOs, 33 through which refugees are able to communicate with relatives abroad and receive financial support through remittances (Lindley 2006, 2007, 2010; Horst 2006), thus cultivating pre-existing networks; but also in the possibility to establish linkages with new networks, such as business associations, religious organisations, INGOs, and state actors, expanding individual agency and enhancing the opportunities to access the market or other domains. The thread running through this literature is a warning against reducing the complex reality of urban Africa to taken-for-granted dichotomies – “the rural and the urban, the formal and informal sectors of the economy, tradition and modernity, and the margins and the centre” (Staples, 2007: 11) –, calling instead for a fine-grained approach to grasp nuances and blurring. As Mbembe and Nuttall (2008) point out, rather than opposing the ‘formal’ with the ‘informal’, or the ‘visible’ with the ‘invisible’, we need a more complex anthropology of things, forms, and signs in order to account for the life of the city in Africa. Analytically as well as in people’s daily experience, simplistic oppositions between the formal and the informal are unhelpful. (8–9) Urban livelihoods between the formal and informal Urban spaces are ideal sites to observe the processes through which conceptual boundaries are blurred. With reference to my research, disentangling the formal from the informal in the African city is a challenging task, as it is made clear in the literature on urban livelihoods. This is particularly the case when it comes to groups, such as refugees or illegal migrants, who are mostly relegated to informal market channels but rely on linkages with formal economic actors and institutions. This applies, for instance, to Somali refugees partnering with relatives who are Kenyan citizens and operate within the formal economy, as I explain in chapter 4, or trading within humanitarian spaces (chapter 7), or to the interplay of formal and informal financial actors (chapter 5). Although I will return to the formal/informal duality later in the discussion on networks, it is, however, useful to provide the definitions of livelihoods and informality here and outline the debates from which they emerge. Chambers and Conway (1992) are credited with the most widely accepted definition of livelihood, as a set of “capabilities, assets 34 and activities required for a means of living” (7; see also DfID, 1999). Carney (1998) suggests the notion of ‘livelihood framework’ to account for the broader socioeconomic and physical context within which livelihoods are constructed. According to Sorensøn and Olwig (2000), approaches to livelihoods should take into account also how people make sense of their way of making a living. Staples (2007) highlights the social embeddedness of strategies to face material deprivation using ethnography “to find contextually specific ways of learning what particular ‘assets’ mean to particular people in particular situations”(15). He devotes his attention to socioeconomic groups experiencing a condition of vulnerability, defined as “a high degree of exposure and susceptibility to risk of stress and shocks, and little capacity to recover” (13), and, drawing from Collinson (2003), stresses the need to tease out the historical roots of vulnerability. This condition of vulnerability is ascribed to the ‘margins’. Yet, the margins, to stay with the spatial metaphor, feature a variegated landscape, shaped by the tension between constraint and creativity, in which economically disenfranchised people sit alongside refugee entrepreneurs and the poor are reformatted as consumers through the category of ‘the bottom of the pyramid’ (Cross and Street, 2009). Ethnographic accounts from urban slums or refugee camps, quintessential fields of observation in studies on livelihoods and vulnerability, emphasise that the margins are a social construct, rather than physical spaces neatly separated from the centre (Staples, 2007: 16; Perlman, 2010; de Bruijn, 2008). The interdependence (not void of tensions) of margins and centre is brought to the fore by the broader scholarly debate on the informal economy. Indeed, defining the formal and informal divide has given way to examining the linkages between economic actors operating inside and outside the regulatory framework of the state, according to a conventional definition of these two categories (Meagher, 2013: 2; Castells and Portes, 1989). This shift reflects a growing interest in harnessing the policy implications of these linkages to improve governance and economic performance. But it also related to the acknowledgement that informality is a phenomenon on the rise, not only in Africa, often at the centre of studies on informalisation, but globally, including in rich economies (Meagher, 2010). The informal economy has been initially seen either through the lens of the modernization 35 theory as a relic of a pre-modern, rural past and destined to disappear under the pressure of modernization and urbanization, or, in Marxist terms, as a pre- and extra-capitalist niche shaped by the petty exchanges of a lumpen proletariat (Elyachair, 2003). A highly influential conceptualisation of informality was developed by Hart (1973) and further expanded by a report of the International Labour Organisation (ILO, 1972). Doing research among migrants from the Frafra ethnic group in Accra, Ghana, Hart examined the constellation of “petty enterprises of all types ranging from the most trivial activities to major business” (Hart, 1973: 67), suggesting that the distinction between formal and informal is based on the difference between ‘wage-earning’ and ‘self-employment’ (68). The ILO report was based on an investigation conducted in Kenya and identified the distinctive feature of informality as a certain way of doing business, rather than in the scale or nature of specific productive sectors. Subsequent studies have further delved into processes of informalisation, observing that this is not a specifically African phenomenon and linking it to global economic changes (Portes, Castells, and Benton, 1989). Also, a more recent debate recasts informality as a historical process and dependent on the degree of flexibility in labour regulations (Meagher, 1995). It is worth pointing out that informal phenomena do not necessarily stem out of economic recessions as a response to the crisis of formal institutions. However, informalization strategies emerge from and are sustained by specific conditions, both economic and socio-political. Among these conditions, Weiss (1987) mentions a multitude of small- scale operators homogeneously distributed in space, a pool of marginalised workers and the presence of long-established social networks (218). Other studies have challenged the assumed link between informality and unemployment, arguing instead that are often those already employed in the formal sector who have access to connections and resources (or, to anticipate a concept I will examine in the next section, to different forms of capital) (Pahl, 1985: 398; Meagher, 1995: 268). While many scholars have argued that the difference between formality and informality is so blurred that these categories have lost their analytical significance (Little, 2003; Elyachar, 2003; Hart, 2007), others have suggested examining 36 informality, not in terms of lack of regulation but in the light of regulatory frameworks alternative to the state (Hansen and Vaa, 2004; Titeca, 2012; Meagher, 2009). Informal activities, although extra-legal, may be perceived as legitimate within the local moral economy, either because they are responding to the failure of the state in dealing with major crises, or because they contest the legitimacy of the state, or both. Moreover, when examined against the formal institutional setting, the informal sector is appraised according to dominant ideological perspectives in specific historical circumstances. The emphasis is thus placed either on the dynamism and beneficial contribution of the informal sector to the broader economy, or on its dysfunctionality and backwardness, as a space in which inequalities are reproduced and which, by depriving the state of tax revenues, lays bare the limited reach of the state and further erodes the rule of the law. It may also be helpful to consider the articulation of formality and informality in terms of visibility and invisibility to the state, and therefore as a continuum, rather than a dichotomy (Palmer, 2004: 12). Along this spectrum, economic actors may graduate their visibility according to the circumstances, concealing or camouflaging their activities in some cases or, on the contrary, striving to dispel “suspicions (which may be particularly intense during years of economic crisis) that criminal activities and fraud are the sources of newly acquired wealth“(Muñoz, 2010: 149). A fine example is offered by Somali MTOs, which I discuss in detail in chapter 5. This case also shows how this ambiguity may draw opposing views from institutional actors, who regard hawala either as a lifeline for displaced people or an underground channel for money laundering and terrorism financing. In general, the analysis of the mechanisms and practices upon which the hawala is based offers insights that challenge assumptions on the formal/informal dichotomy: hawala companies may abide by the law in some countries but not in others and they can funnel unrecorded capital into formal businesses, linking formal and informal livelihood activities. They may keep a low profile and stay invisible, or, as in the example of Dahabshiil that I will discuss in chapter 5, burnish their credentials as a legitimate business and a reliable partner of development and humanitarian organisations. The case of hawala thus highlights the need to shift the attention from the definition of informality to its relational dimension. In what it 37 follows, I turn to the concept of social networks as a key to explaining how informal organisations work. Social networks as theoretical framework In this section, I will clarify the threads knitting the theoretical framework that I apply to my study. I am referring specifically to the ideas of networks, informality, and social capital. Although interweaving, these key concepts, recurring throughout my thesis, should not be conflated, but should be rather seen as illuminating each other and, yet, preserving their differences as they hail from different debates and stem from different concerns. As highlighted in the literature review, social networks are a recurrent image in scholarship on both migrations and the informal economy. The main reason is that, in both cases, they are framed as modes to organise the social through regulatory mechanisms outside the framework of the state, articulating the relationship of formal and informal and facilitating the circulation of people, goods, and information. This particularly applies to relationships straddling international borders, where institutional boundaries divide polities featuring different, and often diverging, regimes that shape the range of possibilities of both voluntary and forced migrants. In crafting my analytical tool, my concern is twofold: to understand how relations of dominance and subordination are reproduced across time and space, straddling physical, institutional and conceptual boundaries; and to put under scrutiny the role of national and transnational actors in reinforcing socio-economic inequalities among a refugee population in which some groups, enjoying greater agency than others, are better positioned to access resources to improve living conditions and escape marginality. What we talk about when we talk about networks A preliminary step to clarify the relation of social networks and informal institutions is to put order into a conceptual jungle in which, as a consequence of the 38 mainstream popularity of the term ‘network’, differences in interpretations and analytical implications have been glossed over, thus flattening the analytical richness of the concept. The multiple approaches to social networks, and the theoretical patchworks made up of concepts uncritically assembled together, risk neutralising their effectiveness in highlighting the dynamics that influence the forms and the contents of networks’ relationships, and to understand how these sets of relationship adapt to, and induce, external changes. Since the 1990s, the notion of network has been seeping from the social sciences debate into popular discourses, particularly in its plural form ‘networks’, as a metaphor, or even an ideology (Beal and Lavin, 2011), of a world system increasingly digitally interconnected. This has contributed to building a sense of false familiarity with a term whose definition, along with its theoretical implications, remains contested, and whose relevance predates the global spreading of ICTs in the last twenty years, going back to the “pre-digital poststructuralist effort to theorize non-hierarchical modes of representational and political connectivity” (ibid.: 3). The concept of social networks has gained centrality in the literature on informal institution by helping explain the balance of stability and change in a labour market that, although not regulated by state laws, rests on “a set of social connections, personal and kinship allegiances, which ensure the flow of labour market information and which connect the employer to a supply of ‘trustworthy’ labour” (Weiss, 1987: 218-19; see also Cooper, 2001; Meagher, 2005). Social networks have thus a critical role in facilitating the recruitment of employees, particularly among populations, such as recent migrants and refugees, lacking recognized credentials or familiarity with the local labour market. This case is discussed by Stepick and Portes (1986) in their study on the differences in the possibilities of incorporation of Cuban and Haitian refugees in the American economy. They argue that it is “where they settled, whether they were male and married, and how much support they extracted from their kin networks which determined their chances for entry into some form of regular employment.” (512) Here, Stepick and Portes suggest that, besides the refugees’ new residence and their specific demography, kin networks are crucial to facilitate the access to the formal economy. The importance of networks in providing “efficiency and effectiveness” (MacGaffey and Bazenguissa-Ganga, 2000: 13) 39 is even greater in situations, such as informal cross-border trade, “where there is need for efficient, reliable information which is not otherwise readily available” (ibid.) MacGaffey and Bazenguissa also draw a distinction between structured networks, based on channels of communication regularly kept open by members of the same ethnic trading community, and personal networks, which are activated by individuals according to the need and the situation and are used for a broad range of purposes (12). These networks are based on reciprocal obligations that create a state of permanent indebtedness (a topic that I will develop further in chapter 5) and are driven by investment strategies aimed at reproducing and expanding trusted relations within and across pre-existing communities. This has a crucial role in contexts of crisis, in which social networks act as safety net when, to use an expression quoted by Lourenço-Lindell (2002) in her study of informality in Guinea-Bissau, someone ‘cai na balanso’(14) – falls off balance – after losing means of support. Before discussing the nature of the capital at the centre of these investment strategies, it is useful to dwell on a theoretical perspective that, in the last decades, has proven to be influential in the discussion on networks, particularly when their material infrastructure, based on goods, artefacts and technological innovation, is considered. The reference here is to the actor-network theory (ANT), an approach mainly developed by Law (1992) and Latour (2005) and primarily concerned with the maintenance of networks made by both human and non-human actors. By equaling human and non- human elements of a network, ANT draws the attention to the linkages among the actors rather than to their nature. ANT provides a theoretical lens for reading the social reality as a fabric of threads connecting “hybrid entities” and in which no distinction is made when approaching the social, the natural and the technological (Tatnall and Gilding, 1999). ANT seeks to explain the emergence, the stability or the failure of social systems by looking at ‘social engineering processes’ that mostly consist in the ‘translation’ of social, technical, conceptual and textual bits. The concept of translation dictates the modality according to which the engagement of all elements occurs and, indeed, ANT is 40 also dubbed ‘sociology of translation’ (Law, 1992). According to this perspective, power is not an entity ascribed to a specific actor but refers to the capacity to enroll other elements of the network in the production of knowledge. This is done, according to Latour, “by ‘interesting’ others and then getting them to follow our interests, so becoming indispensable to them” (Latour, 1986 in Tatnall and Gilding, 1999). The interest is elicited by bestowing qualities, desires, visions and motivations to the newly enlisted actors (ibid.). These alliances can be volatile and shifting but they prompt all actors, whether human or non-human, to translate their resources giving shape to a ‘black-box.’ In this black box are recorded all the negotiations, or the translations, that have led to the creation of the network, and this explains why ANT is considered appropriate to explain ‘how’ a social order works, sustains itself and in which conditions collapses. Opening these black boxes helps also acknowledge the fact that an actor is a network in itself, or, to quote Law (1992), is “a patterned network of heterogeneous relations, or an effect produced by such a network” (384). The emphasis placed on the alliance of human and technological elements is both a strength and a weakness of ANT. While the theoretical premises of this approach are useful to engage with social phenomena in which technology has a prominent role in the production and circulation of value, such as the networks of money transfer operators discussed in this thesis, the articulation of heterogeneity and the endless possibility of expansion of networks is a problematic aspect well captured by Strathern (1996) as she reverses the problem and asks not how networks expand, by involving, or ‘interesting’, others, but how they stop, or are being cut. She points out that the divide between those included in and those excluded from the networks is a result of special social arrangements, such as the claim of ownership, both “a matter of belonging and of property” (531), and that “the kinds of interests, social or personal, that invite extension also truncate it, and hybrids that appear able to mix anything can serve as boundaries to claims.” (ibid.) So, while ANT can be a useful tool to examine the symmetric interaction of human and technological elements, it shows its limits when it comes to analysing the mechanisms through which social and political identities are crystallised to legitimise hierarchies, and forms of dominance and control, such as the one at the centre of this study, are brought to the fore and reinforced. 41 Instead, in what follows, I will discuss other perspectives on networks that enable to account for the reproduction and the entrenchment of power relations. In particular, I will delve into theories of social capital and argue that, despite the popularity gained by the concept in the literature on informality, a few clarifications are necessary to understand the different meanings and implications of social capital as a lens to observe networks as both a space of inclusion and exclusion. Variety of social capital Since the 1990s, a ‘social capitalism’ paradigm has come to dominate the discussion on informal institutions and actors, imposing a normative bias upon the examination of “social networks as ‘social capital’ to the extent that they promote economic efficiency and accumulation, and tends to portray them as ‘social liabilities’ if they do not.” (Meagher, 2006: 219). It is worth noting that this normative perspective on social networks betrays the ideological thrust driving the consolidation of neoliberal globalisation after the end of the Cold War and the demise of the Socialist Bloc, when a certain idea of social capital rose to prominence among policymakers and academics. In a political context animated by voices calling for limited state regulation and unfettered market economy on a global scale, the idea of social networks as organisational structures alternative to state institutions found particular appeal. This view was theoretically grounded in a concept of social capital framed as a conduit, at the same time, of economic development and civic participation, and owed particularly to Putnam’s work on how trust is produced and creates bonds among the members of a community (Putnam, 1993; 1995; 2000). Defining social capital as “the features of social life — networks, norms, and trust — that enable participants to act together more effectively to pursue shared objectives” (1995: 664–665), Putnam’s interpretation aligned with the grand narrative permeating that historical moment and encompassing the belief in the mutual buttressing of free-market and democracy and the pivotal role of civil society in promoting the advance of multiparty democracy in Eastern Europe and 42 Africa. This explains why this idea of social capital gained momentum in policymaking circles, sidelining other interpretations that drew from the social conflict scholarship – above all, Bourdieu’s notion of social capital, to which I shall return later. The social science debate on the multifaceted concept of social capital has a long history (Farr, 2004), intersecting analysis of typologies of ties (Granovetter, 1973), norms and sanctions (Coleman, 1988), and forms of trust (Williams, 1988; Newton, 1997). Particularly influential is Granovetter’s (1973) distinction between strong and weak ties, where the former refer to kinship, or primary ties, while the latter consist in acquaintances or secondary ties. In the light of this classification, Granovetter foregrounds the key role of weak ties in building broader communities beyond the restricted kin group. Coleman (1990) points at the presence of norms and sanctions as a crucial element to formulate decisions on the allocation of trust. The topic of trust in relation to social capital also features central in the analysis of Williams (1988), later developed by Newton (1997), in which a distinction is drawn between ‘thick’ and ‘thin’ trust, according to which the former is engendered by regular interactions among members of homogeneous communities, related to each other through strong ties, and able to enforce sanctions in case of default, while the latter binds members of a community sharing weak ties. Although this distinction applies mostly to interpersonal interactions, relationships among individuals may unfold within institutional frameworks, or involve individuals and institutions. Institutional trust acquires specific relevance in economic relationships (Mishler and Rose, 2001; Williamson, 1985), although it is often rooted in interpersonal trust, such as when an individual embodies a specific institution (as will become clear when the significance of remittance companies is discussed). Meagher (2005) identifies three variants within social capitalist approaches that have shaped the discourse on social networks. They revolve around embeddedness, bridging and bonding and governance. I will first go through these perspectives and then discuss their pitfalls. The embeddedness approach highlights the importance of bonds reproducing thick trust in settings in which members can monitor each other and enforce norms. According 43 to this perspective, social networks are regarded “more as a property of groups and communities than of individuals” (ibid.: 220). Studies on trading diasporas focus on the central role of kinship or religiously based social norms in increasing the efficiency of economic networks operating in loosely regulated settings (Cohen, 1969; Ben-Porath, 1980; Portes and Landolt, 2000; Al-Sharmani, 2007). The emphasis is placed on strong ties and thick trust, the structure of the networks and the way they are being defined by a culture from which both a shared sense of belonging and regulatory norms are derived. In contrast, the ‘bridging and bonding’ angle highlights the role of weak ties and engages with the notion of agency as the key to explaining the functioning of social networks. This approach is concerned with understanding how individuals use social roles rather than “roles using individuals, and [...] the crossing and manipulation rather than the acceptance of institutional boundaries” (Hannerz, 1980: 175, in Meagher, 2005: 220). Social scientists embracing this perspective do not deny the importance of strong ties but argue that a stable and economically efficient social fabric stems out of a balance of strong and weak ties. Bonding and bridging thus refer to the integration of social relationships among people with similar and different social backgrounds, in order to cross-cut networks and overcome social gaps to counteract inward tendencies and create intergroup solidarity. This approach resonates with Putnam’s work and it is helpful to explain democratic governance. Drawing on Woolcock (1998) and Szreter (2002), Titeca and Vervisch (2009) add the notion of ‘linking social capital’, which differs with the previous two views in its vertical, rather than horizontal, dimension. In Szreter and Woolcock’s (2004) words, linking social capital comprehends “norms of respect, and networks of trusting relationships between people who are interacting across explicit, formal or institutionalised power, or authority gradients in society.”(6). Titeca and Vervisch apply this triadic framework to their study of community associations in Uganda, and suggest that linking social capital might breed the risk of social networks being hijacked and turned into neo-patrimonial structures. Eventually, the governance approach regards social networks as a non-hierarchical form of organisation that, 44 featuring a high degree of flexibility and internal cooperation, renders state regulation unnecessary (Meagher, 2005). An interesting synthesis of different perspectives on the significance of social capital in migrant entrepreneurship is the ‘mixed embeddedness’ approach (Kloosterman et al., 1999, 2001; Kloosterman, 2010), which rests on the assumption that an enabling regulatory environment is crucial for business to be established and develop (see Ibrahim and Galt, 2003; Thornton and Flynn, 2003). Kloosterman et al. (1999) elaborate on Granovetter’s concept of social embeddedness, arguing that the significance of social networks for aspirant businesspeople can be better understood by “taking into account not only their embeddedness in social networks of immigrants but also their embeddedness in the socio-economic and politico- institutional environment of the country of settlement” (254). The “time-and-place specific opportunity structure” (Kloosterman, 2010: 26) therefore plays a crucial role in enhancing the effect of a migrant entrepreneur’s network and his capacity to mobilize resources. Migrants’ economic survival and success are thus affected by the legal and political framework of the “context of reception” (Portes and Rumbaut, 1990), a concept that places emphasis on the way national legislations affect the trajectories of different communities of migrants by providing opportunities to some while raising barriers for others. This topic is explored for instance by Stepick and Portes (1986) in their study on Haitian and Cuban migrants in Florida, as they observe that migrants from the two communities experience the arrival at destination in very different ways since the reliance on their respective networks engender diverging trajectories of occupational mobility and aspirations. Schwartz et al. (2014) argue that the experience of the local context of reception may vary according to the individual migrant’s level of acculturation or his access to personal, social and economic resources. They therefore speak of perceived context of reception as “an immigrant’s perception of welcomeness, opportunity structure, and availability of social supports in the receiving community”.(1) It is worth pointing out that, often, the benefits some migrant entrepreneurs reap by operating in specific contexts of reception are derived precisely from their ability to 45 overcome the restrictions imposed upon their broader community or other groups of migrants and facilitate the access to otherwise unattainable resources, such as goods or remittances. Indeed Kloosterman’s analytical framework proves particularly useful to explain local economic changes by taking into account how different ‘openings’, corresponding to turns in livelihood trajectories, are made possible by the mobilization of different forms of capital and by the “systematic coupling of the general market characteristics of the model to the characteristics of individual entrepreneurs” (33). I will delve further on the notion of mixed embeddedness in chapter 6, when discussing the emergence and the transformation of the Somali community in Uganda, and in chapter 7, in which I will discuss a peculiar context of reception, a refugee camp. Mechanisms of exclusion Social networks based on ethnicity and religion can serve not only as a gateway but also as a gatekeeper, controlling or denying entry for some members or groups to more lucrative informal activities (Meagher, 1995: 265; Fernandez-Kelly and Garcia, 1989: 260; McGee et al., 1989: 274; Stepick, 1989; 117, 120). Social networks act as an interface between their members and the market, creating an environment encouraging cooperation among small firms and between employers and employees in the name of solidarity (Portes et al., 1989b: 305). The effect, however, is a blanket over class divisions, with the result that “what appears as ethnic or religious solidarity conceals a mechanism of control and accumulation, but one in which the basic antagonisms are ideologically sublimated” (Meagher, 1995: 274). The concern to explain inequalities and class dynamics in social networks has led other scholars to consider the ‘downsides of social capital’ (Portes and Landolt, 1996, 2000; Woolcock, 1998). Portes and Landoldt (2000) in particular have questioned the rosy narrative of community solidarity by stressing potential negative effects such as patronage, limitations to individual freedom and the burden that falls upon group members. Portes’ line of criticism sheds light on the intrinsic risks of close-knit 46 associations, by bringing the notion of power back into the picture, after it was eclipsed by the popularity of Putnam’s model among policymakers. By arguing that “sociability cuts both ways”, Portes (1998: 18) suggests that networks are regulated simultaneously by logics of inclusion and exclusion. This argument resonates with the criminalisation approach that has largely been applied to the African context to explain why social networks not only have not achieved diffused economic development and democratic governance but have degenerated in the opposite direction (Bayart, 1993; Bayart et al., 1999; Chabal and Daloz, 1999). This strand of Africanist literature has examined the dark side of social capital, arguing that, in Africa, most social networks have turned from being safety nets to cultivate relationships of mutual support into spaces of patrimonialism, elite capture, social control and corruption. In doing so, though, this perspective reproduces a primordialist view of associative forms in Africa, that explains the potential and challenges of social networks mostly in terms of ethnicity and tribalism. While kinship can act as a powerful mobilizer of support in settings of cutthroat competition, it is not sufficient to explain how inequalities in accessing resources persist in networks as the political and economic conditions mutate. To move beyond a short-sighted focus on kinship, or, better, to integrate socio-cultural structures in a theoretical underpinning that accounts for individual agency and institutional peculiarities, I want to reprise Portes’ distinction between the source and consequences of social capital. Although membership allows benefits (in terms of privileged access to contacts and resources, members need to continue investing in social capital by fulfilling specific obligations to preserve their position within the group and eventually increase material and social rewards. Network solidarity, therefore, does not naturally emanate from blood ties but is instead constantly reproduced through participation, engagement and contribution. To retain access to the benefits attached to their membership, members of a network must constantly prove their reliability and usefulness to their group. This argument is directly inspired by Bourdieu (1980; 2014; Bourdieu and Wacquant, 2013) and his view of social capital that, although predating Putnam’s formulation of the same concept, addresses many of the gaps highlighted by critiques of the neoliberal interpretation of social capital, particularly Putnam’s apparent obliviousness to power 47 dynamics. Bourdieu’s definition of social capital differs substantially from Coleman and Putnam, in the premises and the conclusions. According to Bourdieu (1980), social capital can be defined as the aggregate of the actual or potential resources which are linked to the possession of a durable network of more or less institutionalised relationships of mutual acquaintance and recognition or in other words, to membership in a group, as an aggregate of agents who are not simply endowed with shared features (perceivable by the observer, by others or by themselves) but are also bound to each other through stable and useful ties (2, emphasis in the original French version, my translation). This view suggests that the solidarity within the group is based on the benefits that members reap from being part of it. This entails that the relationships are not fixed, but are constantly being reworked. The social network is, therefore, the “product of social investment strategies” (Bourdieu, 1986: 250-251) aimed at reproducing social relationships fungible in the short, medium and long-term. The access to and accumulation of resources is made possible through the conversion of social capital into other types of capital, namely economic (financial and material resources), cultural (skills, knowledge and education), and symbolic (referring to the symbolic legitimisation of position of dominance) (Bourdieu, 1986). There are three main aspects of Bourdieu’s ideas of social capital that deserve consideration: the importance of inter-relations of social capital and a variety of capitals; the role of social capital in the (re)production of socio-economic differentiation and intergenerational (dis)advantage; the embodiment of agency in practices and habitus, as in the case of transnational remittances discussed in the literature review (Holt, 2008). Moreover, Bourdieu’s emphasis on access implies a constant struggle involving different parties to gain the upper hand at political, economic and military level – three aspects that, in the Somali context, are intertwined. This point is further highlighted by Guyer’s (1981) remark that social capital is a perishable item, and, as such, should constantly be accessed and nurtured. On the crucial role of access and the intertwining of social, political and material resources, Guyer observes that where resources are not commodities, people have to invest in the mechanisms that assure them continuing access. One needs social and political resources to get material resources. Relatively high social expenditure is not only to 48 do with thresholds in investment possibilities but with the necessity of investing in the system which alone guarantees the resources (17). This argument also points at the need to take into account the way the institutional setting can either facilitate or impinge on the conversion of diverse forms of capital into each other, particularly when the resources upon which people invest are accrued through mobility. This explains why Bourdieu’s approach has proved influential on migration scholars and geographers focusing on human flows (Cederberg, 2012; De Haas, 2010; Holt, 2008; Van Hear, 2014). De Haas (2010) draws on Bourdieu’s concept of social capital to explain the reproduction of inequalities within migrant networks, which he conceives as a “form of location‐specific social capital that people draw upon to gain access to resources elsewhere” (6), suggesting that some migrants can act as bridgeheads for their own group and gatekeepers for outsiders (ibid.). He points out that, in the context of migration studies, Bourdieu’s view of social capital helps understand how, by migrating, people can gain access to other forms of capital that can eventually increase their value within the networks in which they are already embedded. It is worth noting, though, that network membership can be a necessary but not sufficient condition to facilitate migration, as migrants should be “both [1] able to mobilise the necessary resources (or capitals) to facilitate migration, and [2] willing to help prospective migrants because of moral obligations, altruism, self‐interest or a combination thereof” (20). Related to this point, another of De Haas’ observations, which shall emerge in the course of my thesis, is that social capital is not the only enabler of mobility, but other forms of capital can also facilitate migration – although, in the Somali context, social networks remain crucial to acquire financial resources or education and skills (or economic and cultural capital) that agents can leverage to increase their room for manoeuvring. Kauffman and Bergman (2004) borrows from biology the concept of ‘motility’, which refers to “the capacity of entities (e.g. goods, information or persons) to be mobile in social and geographic space, or as the way in which entities access and appropriate the capacity for socio-spatial mobility according to their circumstances” (750). Originally referring to Western urban mobility patterns, motility is a theoretical tool which allows the analysis 49 of movements simultaneously across geographical spaces and social categories and, in so doing, to individuate the factors that enable it. According to Kauffman and Bergman, these factors are related to: (1) access, referring to the set of options and conditions, often determined by social networks and local dynamics, that make possible different forms of mobility; (2) competencies, which refers to specific skills and abilities acquired and certified by permits and documents (such as laissez-passer and passports), but also to the capacity to acquire information and further competencies; and (3) appropriation, informed by aspirations and needs, and related to decision-making processes and strategies. The elements shaping mobility patterns are “fundamentally linked to social, cultural, economic and political processes and structures within which mobility is embedded and enacted (ibid.).” Discussing the nexus between social capital and migration, Van Hear (2014) points out that the issue boils down to a key difference: who can move and who cannot, considering that capital (in its various forms) is both a precondition and a driver of mobility. He argues that the form of migration and its outcomes are shaped by the resources – economic or network-based – that would-be migrants can muster. The capacity to mobilize those resources is largely determined by socioeconomic background or class, which, drawing on Bourdieu, can be conceived in terms of the disposal of different amounts and forms of capital. [...] Holding combinations of such capital shapes the routes would-be migrants can take, the channels they can follow, the destinations they can reach, and their life chances afterward (102). The concept of social class, evoked by Van Hear according to Bourdieu’s interpretation, has a long history in social science. It features central in the so-called ‘conflict tradition’ (Collins, 1994) and is rooted in Marx and Engels’ theorisation of class as a specific type of social relationship defined by the capacity to control means of economic and cultural production. Van Hear, drawing from Bourdieu, embraces a view of class as a social group identified by the common position of its members, sharing similar views of the world and interests, in a space in which various forms of capital are unequally distributed (Bourdieu, 1987). Van Hear underlines the ‘fungible quality’ of 50 social classes, alluding to the fact that social relationships, economic resources, skills and competencies and legitimisation discourses can be turned into each other to further migration projects, shape career trajectories and protract relations of dominance and subordination. By positing the possibility of converting one form of capital into another, Bourdieu and those drawing from his work provide a more dynamic view of social networks than those regarding them as an alternative to the state, or focusing on strong and weak ties. Bourdieu’s approach underlines the importance of securing access to resources – resources that are paramount to define livelihoods and mobility strategies. It also helps explain inequalities both across and within classes in Somali networks in East Africa, where ‘class happens’ through a double process: on the one hand, Somali dominant clans have generally easier access than minority groups to networks that facilitate the conversion of one type of capital into another. On the other, processes of social differentiation are at work also within clans as some members prove to be more able than other able to use their networks to muster assets and maximise the benefits they can extract from the local opportunity structure. This is why a balanced approach to networks should consider the integration of structure, agency and context. The integration of the institutional approach with Bourdieu’s conceptual model helps thus explain how dynamics of power between dominant clans and minority groups, although rooted in historical contingencies, have been further entrenched through the interface with institutional and humanitarian actors in given contexts of reception. But also how, among members of the same clans, relationships of obligation and indebtedness, that I will discuss in chapter 5, may create inequalities hidden by the veneer of clan solidarity. As Debora James’ (2015) points out in her study on the indebtedness of South African black bourgeoisie, though, the issue of class can be tied to group identity not only on the basis on objective criteria but also subjective criteria, grounded in forms of self-identification and therefore in the adoption of a lifestyle or collective practices that emphasise the difference with a group deemed inferior. This aspect will become evident when I will examine the hegemonic discourse that counter the ‘real’ Somalis, heir of a pastoralist tradition, to the former slaves living on 51 agriculture in Southern Somalia and lumped together in the category of Somali minority. Interclass inequalities are thus reproduced through the control not only of the economic resources, but also of the dominant narratives. Culture, agency and power in networks As Meagher argues, the literature on the role of social networks in informal trade gravitates towards either culture, agency or power as principles to explain the regulatory framework of networks (Meagher, 2009). The emphasis has been placed in turn either on structure, concentrating on the nodes and the flows that constitute networks, or agency, and how networks are used by the actors to overcome structural constraints, or institutional influence on economic behaviours (ibid.). As a result, most attempts to illustrate the functioning of informal networks fail to craft “theoretical tools for exploring the complex intertwining of cultural institutions, social agency and political authority in shaping governance effects of informal economic networks in different empirical contexts” (ibid.: 9). An attempt in this direction is made by Grabher and Stark (1997) in their work on post-socialist economies in Eastern Europe, and whose analytical structure revolves around the principles of legacies, linkages, and localities. Grabher and Stark’s model is based on the assumption that networks do not evolve in a vacuum, but unfold along pre-existing paths, expand and accommodate to new scenarios. In Grabher and Stark’s words, “legacies are not simple residues of the past but can serve as resources for the future” (ibid.: 534). They do not refer to the primordial character of social relationships, but rather to the rationalities that permeate networks and turn them into repositories of heterogeneous practices. Central to Grabher and Stark’s reasoning, are “variations and mutations emerging from the recombination of the inherited forms with emerging new ones” (ibid.: 536). This diversity of forms is organised according to a principle of ‘compartmentalisation’, which allows an increase of the available selection by avoiding encroachment and mitigating competition. As Hannan (1986) points out, 52 Having a range of alternative ways to produce certain goods and services is valuable whenever the future is uncertain [...]A system with greater organisational diversity has a higher probability of having in hand some solution that is satisfactory under changed environmental conditions (85). The importance of the preservation of the alternative shall emerge clearly in my discussion of Somali financial practices, where the allocation of resources among different monetary circuits, devices and currencies is regarded as a strategy to minimize risk. In general, legacies pave the way to career trajectories. For instance, Grabher (1996) suggests that the continuity between the socialist and the post-socialist elites in Eastern Europe originated from the fact that the former technocratic elite, endowed with a greater amount of cultural capital, was able to convert it into forms of capital that better suited the new order. Similarly, members of Somali ruling groups prior to the collapse of the Somali state were more able to cope with the outbreak of the civil war by virtue of the capital previously accrued (such as education or financial assets). The second principle, linkages, refers to the establishment of loose couplings, which better respond to the need to adapt to the context, to better ‘sense’ the environment and to preserve a greater number of variations and alternative solutions than in a tightly coupled network (Grabher and Stark, 1997). Loose relationships leave room for strategic ambiguity, which guarantees access to a broader pool of resources. This explains, for instance, the success of some Somali money transfer companies that were able to create profitable working relationships with a very diverse array of actors, from international organisations to businesspeople to private militias. The third principle, localities, regards the political setting as a crucial breeding ground for the development of networks and posits that networks are shaped by the interactions of formal and informal institutions. The state has thus a key role in either supporting networks by creating synergy (Evans, 1996), neglecting or even hindering the development of networks. As ‘ecologies of social logics’ (Grabher and Stark, 1996: 541), localities act as a catalyst for innovation and cooperation strategies “not because they are a locus of shared meanings but because they are sites of interdependence among different social groups and different social logics.”(ibid.) The significance of localities as spaces of organised diversity appeared 53 clearly in my sites of investigation, where, for instance, the label Little Mogadishu conceals what is, in fact, a multifaceted reality that encompasses heterogeneous actors and institutional rationalities. These three principles not only shape social networks’ trajectories, but also preside over the co-construction of capitals, defining the possibility of converting social capital into cultural or economic capital, and creating the conditions that favour some while limiting others. As I shall illustrate in the course of my thesis, legacies, linkages and localities set the coordinates along which networks, and the power relationships they enshrine, accommodate to encounters with state institutions and adapt to a transnational reality. At the same time, they help explain why some groups are more economically successful, and politically influential, than others. 54 Legacies 55 2 Clanship in Somali history through colonial and postcolonial times Introduction In this chapter, I provide an overview of Somali social structures and history through the dissolution of the state in 1991. My argument is that current inequalities in Somali networks are the result of processes of social stratification rooted in Somalia’s pre-colonial modes of production based on slavery and accelerated by political- economic factors during colonial rule and the post-Independence era. The aim of this chapter is thus to highlight the ‘portent of things to come’, explaining the historical origin and the ideological underpinning of the dominance of pastoralist groups of alleged Arab descents over ‘African’ farming communities. My initial vignette brings to the fore the significance of legacies in the mechanisms regulating Somali networks. I also sketch a map of the Somali clan structure to help the reader navigate through the recurrent clan references that surface throughout my thesis. Acknowledging the importance of kin ties in Somali culture and society, I nevertheless caution that clanship is a contested topic in both Somali studies and society, and, as such, it should be handled carefully. Then, I describe the historical backdrop against which elites from dominant clans first competed and then bargained over strategic assets, excluding so-called minorities and entrenching relations of dominance. 56 An encounter I first met Mohamed at Massai’s bar, where you could find a good Ethiopian injera and a shadowy privacy to indulge in a cold Tusker beer without being seen and frowned upon, or worse, by Eastleigh’s Muslim population. The restaurant was located on the first floor of a building overlooking the corner of 1st and 7th streets. You reached it through a small door squeezed between two shops overflowing with home accessories, and a tight flight of stairs. At first, you were blinded by the contrast with the outside. But then the eyes adapted to the dimly lit environment, and the shapes of people and objects began emerging around the thin blades of light seeping through the shaded windows. The TV was constantly tuned to a channel of Ethiopian music videos, but I had never seen any Ethiopian or Eritrean inside, with the exception of Massai, the owner, his wife, and a male waiter. The other two waitresses were Kenyan, like most of the customers who stopped for lunch during their business trips to Eastleigh. Once in a while, there were also some Somalis, sitting alone behind small plastic bags of khat and a coke, chewing and spitting in silence in their palms. On some rare occasions, they were drinking beer, and I surprised myself by being quite uncomfortable with the picture, like I was looking at them through the moral contempt of my Somali friends. This, no doubt, affected my judgement when, one day, a thirty-something with clear Somali features caught my attention. Wearing a white slim-fit shirt and black cigarette pants, he was standing behind a table, sipping beer and loudly cracking jokes with a waitress in a blend of English and Swahili. He noticed me and, as I foreboded an approaching nuisance, I tried to keep a low profile, but in vain. He addressed me from across the room and, after dispelling his supposition that, for some obscure reason, I could be a German, I revealed myself as Italian. He cheerfully replied that, back in pre-Independence Somalia, his family used to hate the Italian colonialists. As the hatred had clearly subsided since then, he now felt a sort of bond with the ancient foes, and joined me at my table, apparently fulfilling my concerns. Instead, despite my initial resistance, our conversation gained 57 momentum – and a pleasant one. Mohamed (so he introduced himself) was inquisitive, but not suspicious: he was curious about the reason of my presence there, which I strove to explain clearly and simply. He listened quietly for a while, then picked upon something I had just said and exclaimed ‘yes!’, and erupted into a flood of words. He spoke good English, which he had learned when he was a computer science student in Cairo. He had spent three years there, amidst a close-knit Somali community, and had returned to Nairobi two years earlier. During this period, he had travelled to Somalia a few times, working as a consultant for an NGO. He was born on the outskirts of Galkayo, in the Somali region of Mudug: a place, he said, where “everybody loves camels”. Indeed, on the screensaver of his smartphone, he had a picture of a herd of camels, their necks branded by a blue stripe. It was his family’s mark. He had grown up among camels. He loved camels, and when I asked him whether he knew of songs that pastoralists devoted to their animals, as finely described by Abokor (1987) and a recurrent topic in a rich oral tradition (Sheekh, 2009), his exuberance faded away, and he turned almost shy as he begun mumbling a melody. It was a typical hees-geel, or camel watering song, which herd-keepers use to soothe the animals at the water borehole. He explained that all the camels in the picture were female and that, while the males are castrated when they are sold to Saudi Arabia, she-camels are never on sale to foreigners. Money is invested in camels, source of prestige. Camels, then horses and cattle, and then, at the very bottom, goats and sheep. “A man would never milk a goat or a sheep”, he stated. “But you need camels if you want to get married.” He was given his first camels as a child and continued to purchase more even from Egypt, sending money to his relatives back home. Once it was different. “We loved to steal camels”, he added with renewed complacency. His father was a Hawiye Habr Gedr and his mother a Darood Majarteen. I did not immediately grasp the implications of this combination, but his smug expression suggested that there were many. The coexistence of the two clans in the Galmudug area was punctuated by frequent skirmishes, mostly over camel and cattle thefts. Mohamed’s father was a skilled raider. “Then, one day, he saw my mother. She was very young and very beautiful, so he kidnapped her. He took her on his horse and rode back to his village. She spent time there, and then I was 58 born. She could have been disgraced by her family, but she wrapped me in a blanket and walked to her village, all on her own for an entire day. Her brothers saw me and immediately loved me. I brought peace between the two families.” His account was fascinating, but, at times, it sounded far-fetched. Was he selling me some cheap exoticism? Camel raids, family feuds and woman abductions made a perfect package to tease one’s orientalist fascination. Yet, more beguiling was how contradictions merged into each other in the shadows of Massai’s restaurant, their borders smoothed by objects that flickered in the dark. The TV screen on which Teddy Afro, an Ethiopian pop star, was singing, the light bouncing on the bottle of beer, just beside the smartphone in which Mohamed kept the contact numbers of his far-flung relatives. He grabbed the handset and, to corroborate his story, opened his Facebook page. Scrolling down his gallery, among pictures of the Manchester United crest and snapshots from news items, he picked and enlarged a black-and-white portrait of a young woman, or maybe a teenager, staring intensely at the camera, a bare shoulder, two large earrings and a light scarf wrapped around her brow. That picture seemed to speak of a bygone Somalia, a country that did not exist anymore, and it did tease my orientalist fascination. But then, on second thought, it was not quite like that. The separation was illusory. Not only because, a few days later, Mohamed would take me to his mother, the same eyes, although framed by a web of wrinkles and a tight hijab, in her shop in Mandera mall, few blocks from Massai’s bar. But, also, because the past was still very present to Mohamed and his fellow Somalis, as it was reproduced in their interactions and enshrined in their genealogies, thus shaping their trajectories in the world. The past was conjured up in their narratives and, thus, recreated by them. Mohamed was performing his identity – an identity which, as a text, was a patchwork of words, images, objects (images encapsulated within objects!), and rife with idiosyncrasies – in the concealing shadow of an Ethiopian restaurant in the Somali neighbourhood of the Kenyan capital. And yet also these enclaves were illusory: most people in the restaurant were not Ethiopian, and Eastleigh was much more integrated into Kenyan politics and economy, and much more diverse, than the label Little Mogadishu might suggest. But what was the underlying narrative of Mohamed’s text? And could not clanship itself, I wondered, be an illusory enclave? 59 Framing clanship in Somalia “A conflict of narratives” is indeed how Kusow (2004: 1) suggests looking at the roots of the Somali social conflict as he identifies the characteristics of the master narrative that have set the guidelines for Somali studies. This narrative revolves around the self-same nation theory and “establishes two ontological points: 1) an original, Muslim and non-indigenous founding ancestor, and 2) an original dispersal point.” (ibid.: 2). Both aspects refer to the strand of Sunni Islam that spread across the Horn in the years following the death of the Prophet Mohammed, in the 7th century CE, carried by Arab ancestors whose shrines, along with those of other holy men, have been worshipped throughout Somali history. A further tenet of this narrative, as also illustrated by the vignette with which I opened this chapter, is the concept of maandeq, referring to both a specific livelihood, camel husbandry, and to the pastoralist lifestyle derived from it (Lewis 1994, 1998; Laitin and Samatar 1987; Abdi 1979). This narrative dominated Somali studies until the late 1980s, when the offensive launched by the central government against guerrilla movements in the north of the country and eventually the collapse of the Somali state seemed to give dramatic evidence to the intractable ancestral rivalry among Somali clans which is still today often regarded as a central feature of the Somali culture (Luling 2006; Mohamed 2007). Yet, discussing clanship requires careful consideration. Assumptions about the centrality of segmentary ties in Somali politics and society are often taken at face value, thus producing sweeping generalisations and neutralising differences and crucial nuances. It is, therefore, useful to unpack the diverging (and conflicting) narratives that shape academic and policymakers’ approaches to the discussion on Somalia. It is worth pointing out that these narratives are entangled in historical processes as both outcomes and drivers of social and political changes, and, as such, critical to shape dynamics of inclusion and exclusion inside and outside Somalia. While the importance of kinship as a source of solidarity and protection in Somali politics and society has been evident to explorers and scholars since colonial times, less noticeable (and, consequently, less scrutinised) was the complex architecture 60 of power relations and positions of subordination underlying the Somali segmentary lineage system. This is mainly due to the context in which Somali studies emerged and to the influential anthropological corpus of research of Ioan Lewis, the author of a seminal in-depth study of pastoralist traditions and institutions in Northern Somalia. Conducting his fieldwork mostly in British Somaliland, Lewis derived concepts that, according to other scholars of Somali affairs (Besteman, 1999; Cassanelli, 1982) were arbitrarily extended (particularly in popular and policy discourses) also to Southern Somalia, where different social structures were in place. Nevertheless, his work remains a vital rope to climb up and down Somali lineage trees, although, as I realised in the course of my research, a very slippery one. The Somali grand narrative Lewis identifies in the notion of tol, literally lineage solidarity, the tenet of an agnatic, or patrilineal, descent group. This group solidarity operates on different, concentric, levels, corresponding to sub-units that situate an individual within a relational map. According to Lewis (1994) “in the sphere of politics (Somali genealogy’s) significance lies in the fact that it represents the political affiliations of individuals and groups… As Somalis themselves loosely put it, a person’s address in Europe is his pedigree in Somaliland” (97). Mohamed (2007) distinguishes between genealogy (abtirsimo) and group feeling (casabiya, or tolnimo) and argues that colonial anthropology mistakenly conflated the two terms, thus suggesting that the chart to navigate the fluid Somali reality was immutably inscribed within genealogical trees. This distinction provides a lens to examine the constant and intense reproduction of sociality that animates Somali transnational migrant practices. As I will discuss later, while clan membership locates Somalis in networks initially spanning across the Horn of Africa and increasingly global, specific institutions enable the fostering of the ‘group feeling’ underlying solidarity ties. However, it is worth noting here that the segmentary structure which is being 61 extensively discussed (and a major reason for the lack of agreement, by scholars and Somalis alike) is a work in progress, in which the articulation of the different segments mutates against a changing background. In order to clarify how Somalis strike a balance between the stability of the system and the fluidity of its parts, it is necessary to take into account the concept of xeer, translated as social contract or customary law (Mohamed, 2007; Van Notten, 2005). The xeer provides the framework according to which kin connections are mobilised, regulating the making and the unmaking of alliances and establishing comparisons and commensurations in the case of payments on which settlements and partnerships are based. Alliances are forged to respond to immediate situations, but may cease to be of any usefulness as new situations arise (Mohamed, 2007: 98). Unlike what Mohamed claimed in his biographic snapshot, it was probably not his birth per se that brought closer two conflicting groups, but rather an agreement based on the social contract regulating the disputes arising from women abductions. A payment paved the way for his introduction to his cognatic family. Mohamed was positioned at the convergence of two strong lineages, which had been for a long time fighting each other over the control of local strategic resources. Lewis’ work on Somali pastoralists has been highly influential in the broader field of Somali studies, with two main implications: feeding a grand narrative of the Somalis as an ethnically, culturally, and religiously homogeneous population; and reinforcing the ideological thrust of the clan politics with a ‘primordialist’ view of the origin of the segmentary structure (Menkhaus, 2010). The two points are intertwined. Lewis’ ethnography focused mainly on Cushitic-speaking groups deriving their nomadic lifestyle from a pastoralist livelihood better suited to the barren environment of the northern Horn of Africa. The map of the Somali clans reproduced below has been sketched by Abbink (2009: ) on the basis of a simplified chart by Lyon and Samatar (1996) and reprised by several authors. 62 In the map drawn above, mostly based on Lewis’ work, Abbink outlines a ‘total Somali genealogy’, structured in six “levels of genealogical depth” (7): clan-family; clan moieties; clans; sub-clans; lineages; dia-paying groups (where dia, or diya, stands for blood money). In Lewis’ (1994) words, “the standard question, posed to a stranger in order to decide how to place him and so react to him, is ‘Whom are you descending from’, segmentary genealogical distance being… a primary criterion in defining socio- political identity” (83). The complexity and sometimes inconsistency in Somali studies over how to label and order segments reveals a lack of consensus on the terminology (among other issues) (Bradbury, 1994). Given the fuzziness of the intermediate levels, Somalis adopt as identifiers, particularly with non-Somali interlocutors, only the upper and the lowest level: what is referred to as reer (people of) or qabila, in Arabic, generally translated as clan; and the immediate agnatic circle, dia-paying group or jilib, which is mobilised to face outsiders, particularly in cases of payment of blood compensation (Lewis, 1994: 20). Indeed, I will use throughout my thesis a binomial 63 nomenclature, consisting of the clan and, when possible, the subclan, both terms referring to agnatic kin, typically considered the main source of one’s position within the larger clan system. The map above highlights the main bifurcation between Samaale and Saab groups, roughly corresponding to the differences in their core livelihoods, based on pastoralism in the case of the former and a mix of animal husbandry, farming and fishing in the case of the latter. Geographically, Saab groups are concentrated in Southern Somalia, while Samaale groups are traditionally located in Northern and central Somalia although, during pre-colonial times and later, nomads belonging to these groups roamed across the Horn seeking for greener pastures. This is particularly true of Darood subclans, which are still largely represented beyond the Somali borders, especially in Kenya and Ethiopia. However, despite living a nomadic lifestyle, each reer is connected to an ideal homeland, often coinciding with the shrine of the mythical founder or places invested with particular significance for the story of the lineage. With the exception of the Isaaq, concentrated mostly in the former British Somaliland, since 1991 the unrecognised Republic of Somaliland, the three main Samaale clans, Darood, Hawiye and Dir, are scattered across Somalia. While following grazing routes, these pastoralists have been thus expanding their kin network across the Horn of Africa. In historical perspective, it is, therefore, possible to see how the greater mobility of some groups, derived from their key livelihoods, has helped create ties that would survive the colonial partition and the subsequent independence, weaving a safety net that would prove vital after the outbreak of the civil war. Islam and clanship Islam plays an important role both as a cohesive force at the roots of the Somali common identity and as a legitimising discourse that frames hegemonic roles in religious terms.The relation between Islam and the clan system is complex, and it is testament to the changing political and social significance of Muslim precepts and institutions in the Somali territories since the teachings of the Prophet Mohammed first reached the coast of the Horn in the very early years of the Islamic era (Mukhtar, 1995; Samatar, 1992; 64 Lewis, 1999). Islam spread across the Red Sea and through the Horn of Africa and the coast of East Africa through fath (conquest), hijirah (forced and voluntary migration) and tijarah (commerce) (Mukhtar, 1995). Since the beginning, Islam fulfilled a regulatory function by establishing hierarchies, underpinning the belief in the factual existence of bloodlines running through clan lineages, but also creating protected niches for minorities. The Somali oral tradition repeatedly affirms the continuity of Somali clans and the very root of Islam, the Qurayshitic lineage of the prophet Mohammed, and, by tracing the origin of the forefathers of the dominant Somali reer back to the Arabian Peninsula, enshrines a clear political claim (Mansur, 1995). This paradigm “construct[s] a social boundary of Somaliness by enacting ‘a process of simultaneous exclusion and inclusion.’Certain groups are explicitly and permanently placed outside his boundary, while others are included or excluded depending on the prevailing power arrangement or power structures” (Kusow, 2004: 2) According to Kusow, lineage-based narratives unfold through polarised categories that stem directly from the association between Islam and an ideal Somali identity. A cosmological distinction between noble and non-noble groups is thus established by mythically linking the origin of the formers to the birth places of Islam in the Arabian peninsula while confining the latter among the relics of a pre-Islamic (and therefore unholy) African past. Articulated in “’the degrees of distance and/or closeness’ from the common ancestor or the supposed original dispersal point” (Kusow, 2004: 2), the segmentary-based system provides for mechanisms to differentiate between full and partial membership, establishing hierarchies within the same clan. Accounts from the early 1900s report the opposition of aristocrats (gob) and outcasts (sab) (Kirk, 1904, in Hohne, 2014), these latter usually bondsmen engaged in activities considered degrading, such as weaving, pottery, tanning and blacksmithing, or associated with the materiality of both human and animal bodies, and traditionally living in symbiosis with dominant pastoralist patrons, who offered them protection (Lewis, 1999). Besides providing Somali scholars with a schematic device to render legible the complex Somali social reality, the structure of the Somali clans thus contributes to reinforcing the boundaries of 65 an idealised Somaliness, drawing a line between those who locate themselves on the map and those who do not. Using Strathern’s critique to the ANT mentioned in the literature review, I would say that the constructed boundary of Somaliness thus cut the network of solidarity ideologically informed by shared Islamic values. Somali minorities The alternative approach to the dominant narrative focuses on what lies outside the Somali clan structure. Mansur (1995) suggests a constructivist take on the origin of the clans, arguing that, instead of being based on blood relationships, they are mainly “the fruit of the nomadic pastoral life” and that “the necessity of defence, and the movement to new territory necessitated by a constant search for pasture and water have resulted over time in the formation of new alliances and later, new clan identities” (122). The term ‘minorities’ is generally used by scholars and humanitarian practitioners to lump together the groups falling outside the ‘Somali ethnic stock’ represented in the Somali definitive genealogy (Cassanelli, 1995; Besteman, 1996, 1999; Luling, 1984; Menkhaus, 2010). With the term ‘minorities’, Somali studies scholars typically refer to heterogeneous communities engaged with a variety of livelihoods, including trade and agriculture, who speak variations of Af-Soomaali or even a different language, Maay- Maay, widespread across the South of the country, and whose members are not fully integrated into the clan system (Luling, 1984; Menkhaus, 2002; Besteman, 1996, 1999, 2012; Hoehne, 2014). The humanitarian usage of the category ‘minority’ also entails political disenfranchisement, socio-economic marginalisation, and vulnerability in Somalia and in humanitarian spaces. This label is highly problematic for both scholars and humanitarian practitioners. Indeed, the estimated size of the Somali population falling in this category ranges between 5% and 33%. For instance, a 1935 Italian census conducted in Southern Somalia fixed the number of non-ethnic Somalis to 6.2% of the total population (Perouse de Montclos, 2008: 315). According to Lewis (2002), minority groups account for “no more than an eighth of the total Somali population” (7). However, more recent accounts suggest revising this figure upward. According to Rutter 66 (2006, in Perouse de Montclos, 2008), for example, 20% of the population in the whole country and 40% in Mogadishu belong to minority groups. The definition of minorities is also an object of disagreement among scholars. Rahanweyin Digil and Mirifle, for instance, encompass communities located, for several reasons, outside the clan system and comprising outcasts and migrants from other areas of Somalia joining the existing communities to seek protection (Hoehne, 2015). According to Menkhaus (2002), however, this clan ill fits the definition of ‘minority’. Menkhaus observes that this group occupies “an ambiguous position in Somali social hierarchy. On the one hand, it is treated as a low-status group by the dominant Darood and Hawiye clans in southern Somalia. On the other hand, it possesses its own low-caste groups” (Menkhaus, 2003: 328). Expanding through the practice of shegaat, or adoption, described above, Rahanweyin groups reached a strategic convergence to counter the offensive of more powerful Darood and Hawiye clans during the civil war and formed militias to protect their members (Bakonyi, 2013). As Luling remarks (1994), until recently there has been scant research on the origin and the social positioning of these minority groups. Lewis’ emphasis on concepts such as ‘pastoral democracy’ (Lewis, 1999) has contributed to essentialize the ‘nomadic character’ of the Somali society and culture, and to subsume the historical significance of groups whose institutions were based on agro-pastoralist and agricultural livelihoods. These ideas have not only become very popular, but have also been embraced by Somali political actors to underpin hegemonic discourses and provide legitimacy to superiority claims and, consequently, to the entrenchment of power roles. Moreover, this has fostered assumptions that have permeated discussions on Somali refugees’ coping strategies and institutions based on trust, overlooking the need to delve into the political- economic dimension of this often-mentioned concept. Xeer and mandeq were originally studied in their specific social and ecological milieu of origin (mostly among northern Horn pastoralists) but these institutions were later used to explain social dynamics in the whole post-colonial Somalia resulted in the perpetuation of the dominant Somali narrative and exacerbated the invisibility of Somali minorities, both in Somalia and outside. During the 1980s, more systematic research in Southern Somalia started 67 undermining the segmentary-based perspective buttressing the hegemonic Somali narrative. The ethnographies of Luling (1976), Besteman (1999a), and Menkhaus (2003) were crucial in shedding light on the historical victimisation of agricultural communities, and in their construction as ‘minorities’. These studies also highlighted the structural conditions that exposed their particular vulnerability during the violence that engulfed Somalia in the 1990s, and their current marginalisation, both inside and outside of Somalia. However, there is no universal consensus, among Somali scholars, on which groups fall into the broader category of ‘Somali minority’. Below, I provide a tentative map, summarising the literature on this subject and drawing mostly on Luling (1984) and Hill (2010) – the former, one of the first scholars to conduct in-depth examination of Somali groups located outside the dominant clans, the latter an aid worker whose study has brought the issue of Somali minorities to the attention of humanitarian practitioners operating in Somalia. Sources: Luling (1984), Hill (2010). Somali minorities Sab Gabooye/Midgan Madhibaan Muuse Dhariyo Howleh Hawraar Same Habar Yaquup Yibr TumaalBantu Rer Hamar Biymaal Sheikhal Asharaf Coastal groups Benadiri Barawan Bajuni Jaaji 68 Luling (1984) was the first to draw a distinction between occupational castes and farming communities from the inter-riverine areas of southern Somalia. Among these groups of outcasts, she singled out Midgaan, Yibir, and Tumaal (ibid: 41). Lewis adds other occupational groups, such as Booni hunters and Bajuni fishermen in southern Somalia. There are also minorities to which Islam grants a special status as they have a significant position within the Islamic tradition: such is the case, for instance, of the Ashraf, reputedly tracing their ancestry to the Prophet Mohammed’s daughter Fatima, or the Sheikhal, descendants of the first caliph Abu Bakr (Lewis 1999). Members of these groups were incorporated in major clans as clients. Dominant clans also used to forge patron-client ties with non-Cushitic farmers, as I explain later. Yet, in the case of religious groups, adoption, and even intermarriage were particularly valued by pastoralist clans, as a way to ennoble the blood line. Their social prestige did not spare Ashraf and Shaikhal families from victimisation during the civil war, although their networks, which span the Islamic world, enabled many of them to seek refuge outside Somalia. Outside the clan system, there are also the commercial communities that thrived on the cosmopolitan Somali coast around key trading cities such as Mogadishu, Brawa, and Marka. These groups emerged in the context of intense commercial exchanges between the coast of the Horn, the Arabian Peninsula and the Persian Gulf (Kassim, 1995). Called Benadir from the name of the area around Mogadishu (Benadir derives from the Arabic word bandar, port), these mercantile groups have different names according to the urban centres, or even the specific neighbourhoods in which they established their presence, such as Reer Hamar in Mogadishu (the name derives from the historical Hamar district of the Somali capital). Identified by other Somalis from the light-skin and often pale eyes, the members of these communities trace their ancestors back to seafarers sailing across the Indian Ocean, mostly of Arab and Persian origin, but also of Indian and Portuguese. According to Eno (2008), unlike the Barawan, from the city of Brawa, who were included in the broader Digil confederation, Reer Hamar were considered a separate group. They had accrued wealth through their far-flung 69 commercial networks across along the East African coast and the Persian Gulf and commanded respect since most ulema, or experts of Islamic jurisprudence, hailed from this community. Through their history, they were able to leverage economic influence and social prestige to forge ties, often through marriages, with larger and stronger clans However, during the initial stage of the civil war, they were particularly targeted by the armed militias ravaging Mogadishu and only later, as I explain in the next chapter, were able to mobilise their resources to buy the protection of local strongmen (Hoehne, 2015). The second large minority group referred to by Luling comprises heterogeneous communities located in Southern Somalia, practicing, to a various extent, agro- pastoralist livelihoods and speaking Maay-Maay, a Cushitic language that differs from standard Somali. It comprises two macro-categories, the Rahanweyin clan and the farming communities of descendants of East African slaves and pre-Somali inhabitants of the area who mostly self-identify as Jareer. This category is of particular interest to my overall argument. Their protracted marginalisation outside Somalia, in Kenyan urban centres and humanitarian spaces, foregrounds the legacy of pre-colonial slavery, colonial subordination, post-colonial political disenfranchisement and eventually extreme victimisation after the collapse of the state in shaping power relations among clans, reinforcing the dominant position of Darood and Hawiye political and economic elites. The farming communities of Southern Somalia were the worst hit by the civil war, which broke out in 1991. The origin of these communities is composite. The Jareer population is indeed a mix of descendants of slaves (runaway, manumitted and emancipated after the formal abolition of slavery) from East Africa, mostly from the areas which became Tanzania, Malawi and Mozambique, kidnapped and brought by Arab slave traders in the late 1700s and 1800s to work in coastal cash crop plantations, particularly in the Juba region of Southern Somalia (Cassanelli 1982; Luling 1984; Del Boca, 1992); members of non-Somali pastoralist populations of the Horn such as Oromo and Boran (Besteman, 1999: 59); and, according to a more contested theory, farmers who settled in the valley of the Shabelle river before Somali speaking pastoralists spread across the area (Eno, 2008). Luling (1984), who was among the first scholars to conduct 70 systematic research on Somali minorities, observed that pastoralists and farmers occupied different ecological niches (the tzetze fly infested banks of the Juba and Shabelle rivers were unsuitable for the nomads’ herds). Relationships between the two populations, although changing across time and place, were mostly characterised by forms of domination and serfdom with which members of dominant clans established protection and control upon Jareer communities. Focusing on the maroon communities along the river Juba, Besteman (1999) describes how the complex relationships between farmers and pastoralists were shaped by a constant tension between practices of assimilation, through conversion to Islam and incorporation within Somali social structures as shegaad or clients, and processes of othering that remarked racial and hierarchical differences. These inequalities were enshrined in the xeer, the customary law, by establishing for instance that the compensation for the killing or the rape of a Jareer was to be lower than the one for members of dominant clans. Jareer were forced by their patrons’ clan to contribute to diya payment, but could not benefit of compensation in case of injury or death (Hoehne, 2015). Jareer negotiated access into Somali pastoralist families through trading partnerships and support during the frequent skirmishes pitting one clan against the other. As a result, Jareer communities continued to depend on their patrons for protection and not only failed to develop an overarching political and social identity, but, when the civil war broke out in Southern Somalia in the 1990s, were caught in the crossfire, suffering from the looting of their lands by the militias of both their patrons and their rivals. Processes of othering are evident in the nomenclature used to lump together this constellation of communities. Their livelihoods, physical and cultural features, although distinct from those of Somali pastoralists, are often different from each other. Yet, these communities share a social status rooted in a common legacy of slavery and in cultural constructs present along the frontier of the Islamic pastoralist world and inner Africa. Drawing from historical accounts, Besteman (1999) argues that in pre-colonial Somalia, as in other Muslim pastoralist societies across North African where black Africans were subdued, enslaved and forcibly converted, a shift gradually occurred in the way slavery, from being initially associated with paganism, became discursively connected to blackness, thus creating a “potent 71 classification system” (ibid.: 117). The entanglement of slavery and blackness reverberates in a broad repertoire of terms that includes references to both racial markers and condition of subordination. For instance, the term Jareer means ‘hard hair’ and refers to a prominent physical feature that defines a category “equated with “African” – and thus slave – ancestry, as distinguished from the (mythical) “Arabic” ancestry of Somalis” (ibid.: 116). This latter category is instead identified with the name Jileec, or ‘soft hair’, referring to Cushitic pastoralists who claim a direct bloodline to the Arabian Peninsula. Interestingly, Jareer is the term that has met the largest favour among members of these farming communities and is the one that I have heard the most during my fieldwork in Eastleigh2. However, as I explain in chapter 7, in the past twenty years, the term Bantu, which emerged in media and humanitarian circles, has gained increasing popularity. Other terms with which these farmers are known among other Somalis reflect pre- and colonial relations of subjugation and have an offensive connotation: they include the Italian word Oggi, ‘today’, used by the Italian colonialists to refer to the supposed Jareer inability to long-term planning; Adoon, the Somali word for ‘slave’; and Gosha, Somali for ‘forest’, also in the variation wa-Gosha or Rer-Goleed, people of the forest, with reference to the wooden banks of the river Juba, along which many runaway slaves found refuge already in the 1800s (Menkhaus, 2003; Besteman, 1999). However, as Besteman (1999) points out, “despite the abolishment of slavery, the Gosha/Jareer were persistently considered to be of lower status as ex-slaves, ‘blacks’ and former infidels (converted to Islam only by their masters)” (80). A ‘racialised space’ was thus constructed in the farming areas of Southern Somalia, in which marginalisation of the local Jareer population endured despite changes in local, national and global dynamics. Indeed, colonial and post-colonial rules entrenched the hierarchic relationships between dominant pastoralist clans and subordinate farming groups. Since the purpose of this thesis is to illustrate how these dynamics are reproduced in Somali transnational networks, the remainder of this chapter examines the historical construction of inequalities in Somalia. 2 According to Menkhaus (2003), embracing this term Somali Jareer people proudly emphasise the resilience of their character. 72 Somalia before Independence Somali pre-colonial history is characterised by the diffusion of Cushitic-speaking groups from the Northern Horn to the rest of the region over a nine-century period, between the Tenth and the Nineteenth century; the consequent displacement of pre- Somali agriculturalists and their confinement to the area between the rivers Jubba and Shabelle; and regular exchanges between the hinterland and the more commerce- oriented and cosmopolitan coast (Lewis, 2003). This region was punctuated by emporium cities evolving from Arab and Persian settlements that predated the Islamic era and played a crucial role in the diffusion of Islam since the very early years, as traders paved the way to proselytism (Kassim, 1995; Mukhtar, 1995), similarly to what observed along trade routes in West Africa (see Cohen, 1969 on the role of Hausa traders in the diffusion of Islam). The coast and the hinterland followed different historical trajectories, but remained economically linked (Hashim, 1997). Organised as Sultanates, coastal cities acted as a commercial interface between the hinterland, over which they had no political control, and foreign maritime powers. The Somali political and economic heartland thus regularly oscillated between the coast and the interior (Lewis, 2003), a fixture through Somali recent history. Among pastoralists, governance remained fragmented: customary law and kinship structures provided the framework to sort out disputes, either through mechanisms of conflict resolution or by resorting to limited violence (Menkhaus, 2014). As mentioned above, Somalia’s southern coast was the recipient of an high number of East Africans brought by slave traders to work initially in Arab/Swahili and Somali-owned plantations of grain and other cash crops, cultivated for export to other markets of the Omani Empire in East Africa, Southern Arabia and the Middle East (Eno and Eno, 2007: 14). These slaves were also employed as porters or in other forms of heavy labour in Mogadishu, Brawa and Marca. Most escaped slaves found sanctuary in the Gosha, the wooded areas along the river Juba, where they were absorbed into existing non-Somali farming communities (Eno, 2008), or established maroon settlements across Southern Somalia. According to estimates by Besteman (1999) and Menkhaus (2003a), between 1865 and 1895 over 20,000 slaves escaped from their Somali masters. A “racialised space” (Besteman, 1999) began thus 73 taking shape, defined by a legacy of slavery, readable racial markers (which, as previously said, although heterogeneous, were perceived as clearly distinguished from those of Somali pastoralists) and livelihoods based on agriculture. The growing market for grain in the Middle East increased demand for labour in cash crops production and boosted the slave trade to Southern Somalia. But this was not the only global phenomenon that contributed to shaping this racialised space. Jareer communities also grew larger as a result of the anti-slavery campaign that gained momentum in the course of the 1800s, as Italian colonialists helped runaway slaves to reach established sanctuaries (Robecchi-Brichetti, 1904) and British naval patrols released slaves on shore (Eno and Eno, 2007). The colonial scramble for the Horn of Africa accelerated in the last quarter of the century. European powers, each driven by different geopolitical agendas and adopting different ruling styles, gradually penetrated the region from the coast, while, in the interior, the Ethiopian kingdom consolidated its grip on the Haud, a traditional grazing area for a variety of pastoralist groups, including some Somali communities. The colonial partition had a limited impact on the mobility of pastoralists. However, cutting across migratory routes and kin and cultural ties, the initially invisible state boundaries created facts on the ground for future irredentist claims to ‘reunite’ the Somali territories (even though they have never been politically united). While France gradually asserted its rule over Djibouti, with its large Somali-speaking population, the territories that would eventually merge into the independent Somalia fell under the control of the British and the Italians. The former had been expressing their interest in the Northern Horn following the 1839 conquest of Aden, where a military presence was established to oversee the short strategic route to India through the recently opened Suez canal. The relevance of the region, known as Somaliland, rested on the abundance of livestock, which made it a vital meat provider to the garrison in Aden. Yet, Britain opted for a limited engagement in local affairs, relinquishing control initially to Egypt, which had fallen under British control in 1882. Only when it saw its interests being threatened by the gradual encroachment of its European rivals (France and Italy), Britain established the Protectorate of Somaliland, agreeing its Western borders with French Djibouti in 1888. British rule was indirect and 74 based on negotiations with different clans. Xeer remained the dominant legislative framework in the hinterland, and local authorities were not disrupted, even during the 20-year struggle (between 1900 and 1920) to quell the so-called Derwish insurgency led by Sayyid Muhammad ‘Abdille Hassan, a charismatic sheik (Lewis, 2003). In 1895, Britain took control of Jubbaland, a region comprising today’s Lower Juba, Middle Juba and Gedo regions, and whose administrative centre was Kismayo. Previously a possession of the Zanzibari Sultanate, it marked the northern border of the Consulate of Kenya in British East Africa and was transferred to Italy in 1924 as a reward for the help during World War I and for dropping claims over British Somaliland (Hess, 1966; Little, 2003). Clarifying these colonial manoeuvres is important to understand the origin of the forces shaping both Somali migratory outflows, the fragmentation of the Somali society and the reproduction of power imbalances, particularly in Southern Somalia and on the border with Kenya (Abbink, 2003). Tensions and rivalries were further exacerbated, after the 1969 Somali socialist revolution and obviously after the demise of the state in 1991, but already “British and Italian colonial experiences in the region sharpened existing ethnic and clan tension that remain today” (Little, 2003: 26). Jubbaland was characterised by a strong presence of Darood, Ogaden and Marehan pastoralists, who had penetrated from Western Ethiopia through North-Eastern Kenya during the 1800s, forcing out or subduing as slaves or clients pre-existing Oromo herders or Jareer farmers. Along with the British army, traders and herdsmen from Northern Somalia settled in Kismayo and the surrounding area. They were mostly Harti, a Darood sub- group that includes Majarteyn, Warsangeli, Dishilhe and Dulbahante clans and, thanks to their trade links to the Arabian Peninsula, had developed mutually benefiting relations with the British (Turton, 1972). As I will explain later, the same dynamics brought the first Somalis to Uganda, at the time also a territory of British East Africa, with the difference that, while in Uganda the position of the Harti groups among other Somali clans was undisputed, in Southern Somalia their interests (especially in trade and the control of grazing lands) collided with those of the Ogaden. However, resettlement of Harti traders and herders to the area was encouraged by the British and, as noted by Turton, the Harti “gained a virtual monopoly of government jobs” (ibid., 122), including 75 in the police force. Harti clansmen were also recruited as Askari (soldiers) by the British against the Ogaden, particularly from the Mohamed Zubeyr subclan, who regularly disrupted local trade routes. Harti groups retained their dominant position in the area after Jubbaland was relinquished to Italy. Somalia under Italian rule The Italian colonial project in the area between the river Jubba and Shabelle was driven by the aim to establish permanent settlements. However, the Italians faced a fierce local resistance when, after initial neglect, they tried to enforce an anti-slavery policy (Hess 1966). The provision enraged, in particular, the pastoralist groups of the hinterland of the Benadir coast, especially Hawiye subclans (ibid.). The abolition benefitted the local Jareer communities, from which the pastoralists drew their servants, but Hawiye sheiks framed the resistance to the ban in religious terms (Eno, 2004) and the row escalated into an open conflict that aggrieved the local nomads against the Italians for years to come. Local farming communities as well rose against the occupiers, who, meanwhile, had set up large estates in the regions along the rivers Juba and Shabelle (Besteman, 1999). To provide labour for these estates, the Italian colonialists tapped into the settlements of emancipated slaves. With the coming to power of Benito Mussolini in Italy, the effects of the Fascist turn on the ruling of the colony were particularly harsh. A system of coerced work was imposed upon the local population and, given the resistance of the nomads to work in the plantations, only Jareer farmers were forcibly conscripted, and their former masters, especially Darood kinsmen, became instrumental in overseeing them on behalf of the colonial rulers (ibid.). Thus, despite the formal abolition of slavery, the Italian colonial administration contributed to exacerbating the position of subordination of Jareer farmers to other Somali groups, on the ground of a presumed proclivity for agriculture and a greater docility. According to Declich (1987, 2000), the Italians started using the term Bantu to indicate the farmers from different communities, playing a critical role in institutionalising the difference between Jareer farmers and Jileec pastoralists because the former, unlike the latter, could 76 be forcibly recruited to work in the colonial plantations (see also Menkhaus 1989; Besteman 1999, 2012; Del Boca, 1992). As the lands inhabited by Jareer communities became once again a pool of workforce, the colonial subjugation reproduced the historical association between Africans and forced labour. Another effect of the shift to cash crop production under Italian control was the increase in the importance of the livestock sector, with the arrival of Darood Harti herders from the centre and the north of the country (Little, 2003). In other nomad-dominated regions of Somalia, the relationship between the Italian colonialists and the local population deteriorated to open conflict, such as in Mohamed’s city, Galcayo, where the Italians faced a strenuous resistance from Darood Majarteen militias, who were eventually defeated at the price of thousands of dead (Issa-Salwe, 1996), the tragic past to which Mohamed lightheartedly alluded when we first met. After WWII, with the end of the ill-fated and short-lived Italian colonial endeavour and its former territories falling under British administration, the prospect of independence for a state merging former British Somaliland and Somalia Italiana began to take shape. Examining the Somali state building process is beyond the scope of this thesis, but it is important to point out that elite formation and Somali nationalism were entwined since before independence. In 1943, the Somali Youth Club (SYC) was established to champion the cause of a greater Somalia (Lewis, 2002), an agenda aligned with Britain’s. Indeed, when in 1947 the SYC became the Somaly Youth League (SYL), the British military administration (BMA) recruited cadres and officers from among SYL members and Darood kinsmen, who held sway over the party (Castagno, 1964). About 85% of the administrative and police posts were assigned to members of dominant communities, who seized key positions in the nascent state machinery (Castagno, 1959, in Balthasar, 2014: 228). As I will show, the alliance between an urban elite of civil servants and a fledgling class of livestock traders related by kinship would prove significant in Somalia’s modern history. In 1950, the BMA relinquished control to the Amministrazione Fiduciaria Italiana della Somalia (AFIS, Trust Territory of Somaliland under Italian Administration), who received an UN mandate to administer a trusteeship to “foster the development of free political institutions and to promote the 77 development of the inhabitants of the territory towards independence” (in Lewis, 2002). Under AFIS administration, the entrenched elite of civil servants, mostly hailing from Darood and Hawiye clans, increased its leverage to influence the path towards independence. Indeed, because of the lack of a truly national party, competing national projects emerged (Samatar, 1993). In Southern Somalia, the political arena was almost equally split between the SYL, a strong proponent of independence and unification with the lost Somali territories, and Hizbia Digil-Mirifle Somali (HDMS). Initially enjoying the sympathies of the Italians, HDMS was popular among farming communities in the south (Rahanweyn and Jareer) and minorities living along the Benadiri coast “who presciently feared the threat of hegemony and loss of land at the hands of dominant pastoral clans” (Menkhaus, 2014: 564), whose interests were represented by the SYL. HDMS championed the establishment of a separate political entity in the farming areas of Southern Somalia with links to Ethiopia (Lewis, 2002: 157) but, as AFIS struck a deal with SYL in 1955, members of HDMS remained cut-off from the political process. Moreover, ahead of the proclamation of independence, on 2 March 1960, the lawmakers of the Somali Republic (Lewis 2002) approved the criminalisation of public disclosure of clan affiliation (Hoben, 1988: 206). HDMS changed its name to avoid any reference to the Digil Mirifle clan and became Hisb al-Dastuur Mustaqil al-Somal (the Constitutional Party of Independent Somalia). Despite the ostensibly nationalist ideology promoted by the Darood and Hawiye, social differences rooted in clan affiliation endured, even after the proclamation, on the 1st July 1960, of the independent Republic of Somalia including the former British Somaliland. Somali Independence The first decade of the Somali Republic may look like a democratic lull compared to the authoritarianism of the 1970s and the gradual institutional decay of the 1980s, which precipitated the eruption of the civil war. However, phenomena of elite capture and neo-patrimonialism emerged shortly after Independence (Lewis, 1972; Samatar, 1993; Adam, 1995). Behind the facade of a vibrant multi-party system lurked unresolved 78 tensions and widespread corruption. The two aspects were entangled. Balthasar (2014) argues that “it was strategic calculus, rather than some form of intrinsic identity that bound Somalis together at the time of independence and unification” (224). This helps explain why, initially, northern livestock traders supported the idea that a unified Somalia could leverage a stronger case to take control of the Haud, now part of Ethiopia but still considered a lost territory crucial for herdsmen’s livelihoods (Geshekter, 1985; Markakis, 1987). The irredentist hopes of the former British Somaliland, though, were dashed in the aftermath of independence and turned into disillusion towards the government in Mogadishu, creating an environment of resentment that led to a coup attempt in 1961 (Adam, 2008). The crisis was temporarily solved through a pact of northern and southern elites, which brought an apparent stability to the country during the 1960s. Somalia became the African country with the highest rate of foreign aid per capita but, with the increase of rent-seeking opportunities, the public sector ballooned (Laitin and Samatar, 1984). Widespread corruption contributed to exacerbating the inherent fragilities of the Somali Republic as predatory behaviours soon became the norm among elected representatives and bureaucratic cadres alike (Adam 1999: 262; Lewis 2002: 206). As a result, urban–rural divide and social stratification increased (Laitin, 1976), forging a rapacious ruling class that, despite the formal ban on clanship, continued to mobilize kin ties to control access to institutions and economic assets. The combination of massive flows of foreign aid and graft established a pattern in Somali postcolonial history, and one in which, despite political upheavals, the arena where dominant clans competed to capture the state continued to preclude Jareer farmers and minorities in general. As Menkhaus (1996) points out, the farming area in the Gosha witnessed the rise of a commercial class engaged in cash-crop production (mostly grain and cotton) and including a few Jareer villagers. And yet minorities continued to be underrepresented in the national assembly and “virtually absent from ministerial positions, diplomatic posts, and national and cultural institutions” (Cassanelli, 1996: 17). In a context of growing inequalities, this “first generation of millionaires” (Laitin, 1976: 452) fuelled the nationalist call to carve a ‘Great Somalia’ in the Horn of Africa. This aspiration was embodied in the adoption, at Independence, of a flag adorned with a star 79 with five points (each one standing for a fragment of the Somali ‘nation’ scattered among the former Italian Somalia, British Somaliland, Djibouti, Ethiopia, and Kenya). The Somali state’s endorsement of this ideology stoked irredentist tensions in neighbouring countries with large Somali ethnic populations, such as Kenya. Ahead of Kenyan Independence, as British and Kenyan authorities dismissed all secessionist claims advances by local Somali leaders, protests started brewing in the Northern Frontier District, a semi-arid area in the northeast of the country that was once part of Jubbaland before this latter was ceded to the British by the Italians. In the wake of the proclamation of the Republic of Kenya in 1963, two groups of Somali Kenyans merged in the Northern Province Peoples Progressive Party (NPPPP) and later in the Northern Frontier Districts Liberation Army (NFDLA) and launched an insurgency in the area (Branch, 2014: 645). One group was supported by Hawiye clans and was rooted in the northern borderland areas of Wajir, Moyale and Mandera, while the other hailed from the Darood clan and was active in the surrounding of Garissa and across the south- western border with Somalia. Following attacks against Kenyan troops, Kenya’s president Jomo Kenyatta declared a state of emergency in the entire district, a momentous decision not only for Kenyan Somalis, but also for Somalia, and one that would have implications for the Kenyan attitude towards Somali refugees and more generally for the Kenyan refugee policy framework. The ensuing four-year war, remembered as the shifta war after the local word for bandit that Kenyatta used to label the insurgents, left about 4000 dead and was bound to leave an open sore between Kenya and its citizens of Somali origin (Africa Watch, 1991). Although the hostilities between Kenya and Somalia ceased in 1967 with the signing of a Memorandum of Understanding in Arusha, Tanzania, the area continued to be marred by tension and violence and remained under a state of emergency until 1991. The 1969 Somali revolution The outcome of the Shifta war was seen as a defeat and precipitated instability in 80 Somalia, sowed discontent among the political forces and paved the way for a military coup, which took place on 21 October 1969. The takeover lead to the establishment of a Supreme Revolutionary Council (SRC), presided by General Muhammad Siyaad Barre and a 24 member military junta, which proclaimed the Somali Democratic Republic. The SRC officially embraced scientific socialism, bringing the country into the orbit of the Soviet Union. On the first anniversary of the revolution, the Somali government launched a program which included a nation-scale anti-corruption campaign, the nationalisation of the economy, a ‘cultural revolution’ to extend literacy to the countryside and the ‘de-tribalisation’ of the Somali society (Samatar, 1988). Its declared target was the lineage system; xeer and dia-paying were declared illegal and clanship was publicly associated with nepotism and corruption. As for the other pillar of Somali identity, Islam, the regime initially treaded carefully: although espousing a form of socialism inspired by Chinese, North Korean, Soviet and Nasserite experiences, President Siyaad Barre repeatedly underlined the compatibility of socialism and Islam and, at least during the first years of the regime, avoided direct confrontation with high- profile Islamic leaders. Despite publicly chastising tribalism, the regime was largely perceived as an expression of clans connected to the president, namely Marehan, Ogaden and Dulbahante, which Somalis, forbidden to use clan nomenclature publicly, referred to with the acronym MOD (Lewis, 2010). The SRC became increasingly staffed with members of these clans Members of the paternal clan of Siyaad Barre, Darod Marehan, were widely regarded as direct clients of the regime. The modernisation programs of the government were thus met with suspicion by pastoralists from other clans. The expansion of water facilities and market infrastructures funded by the World Bank lead to frequent encroachments into pasture lands of other clans, with resulting skirmishes. In the confrontation between Marehan and Ogaden herders in Southern Somalia, the former typically had the upper hand, as the security forces were reluctant to act against Siyaad Barre’s clansmen (Little, 2003). The situation was different in the farmlands of southern Somalia. In the course of her fieldwork in the Jubba Valley in the 1980s, Besteman (2012) observed that “the 81 Jareer farmers only avoided re-enslavement by pastoralists because of national laws against slavery” (289) Barre’s official policy of national solidarity at least provided some protection to minorities (Lewis, 2010), and, until the mid-1970s, Jareer farmers enjoyed some prosperity. However, Jareer continued to be “effectively excluded from political power and most state resources” (Hoehne, 2015: 796). Moreover, the adoption of northern af- Soomaali as the country’s official language contributed to marginalize further Maay- Maay speaking Jareer. The situation deteriorated with the introduction of land reforms and the creation of a registration system for land titles. It is necessary, at this point, to describe the regional and global backdrop against which, in the 1970s and 1980s, class stratification in Somalia accelerated, with lasting effects on Somali transnational networks. The rise of the remittance industry During the 1970s, thousands of Somalis migrated to the Gulf countries to meet the growing workforce demands of the local oil industry. Since most migrants remained undocumented, there are no precise data on this ‘muscle-drain’, but estimates range between 67,000 and 300,000 in the 1980s (SomConsult, 1985; Lewis, 1994), mostly male and from the former British Somaliland, where a drought and subsequent famine in 1974-1975 forced many to emigrate (Lewis, 1994; Marchal et al., 2000). It was during this period that the Somali remittance sector, crucial until the present day for the dispersed Somali diaspora, began taking shape. Data on the total amount of the remittances at the time are also vague. According to Lewis (1994), they accounted annually for two or three times the Somali Republic’s earnings (122). Drawing from an ILO report, Geshekter (1997) suggests that Somali migrants in the Gulf in 1985 earned 700 million USD, 40% of which (280 million USD) were remitted (see also Bradbury, 1997; Kleist, 2004). The former British Somaliland was the gateway of the remittance flows to Somalia, thanks to its geographical proximity to the Arabian Peninsula and a long history of trading routes linking the Horn of Africa to the Arab world (Pankhurst, 82 1974). These connections evolved into a system designed to “stop underground trading activities and to cover the importation of certain local needs with the remittances” (Qassim, 2004: 186) and authorised in 1976 by the government in Mogadishu. Known as franco-valuta system, it responded to the growing demand for money transfer services from the Arabian Peninsula and the Gulf, allowing traders to collect foreign hard currency, especially US dollars, extremely sought-after in a country, such as Somalia, exposed to chronic inflation. Based on networks of agents from the same clan, this system was considered more trustworthy and efficient that the state-run banking system managed by the Somali Commercial and Saving Bank, characterised by a lack of efficiency and higher fees. Interestingly, the pursuit of hard currency is a feature which has been playing a central role in Somali traders’ decision-making for a long time, as already observed by Burton (1943: 289) and Lewis (1994: 129) – and as I observed in the course of my research. Import licenses were granted to businesspeople having access to foreign currency outside Somalia. Tapping into the growing demand for remittance services from the Gulf States, these businessmen collected the remittances in USD from the workers. The cash could either be delivered directly to the migrants’ families or used to purchase goods to be sold in Somalia. The corresponding amount in Somali shillings, or in goods, was handed to the recipients, according to an exchange rate previously agreed (Lewis, 1994; Qassim, 2004). Trust was a crucial element in this business, and the agents were connected with each other along clan lines. However, remittances in foreign currency continued outside the scrutiny of the regulators, while only a small amount was funnelled through the official banking system and converted into local currency. In 1982, the franco valuta system was eventually banned by the regime, yet it continued to play a central role in Somali household financial strategies, evolving into a sophisticated money transfer industry, which I describe in Chapter 5. The Ogaden war and the beginning of the end In 1977, Siyaad Barre ordered the invasion of the eastern part of Ethiopia, hoping 83 to take advantage of the political turmoil engulfing the country since two years earlier, a Marxist revolutionary movement, the Derg, had seized the power in a coup. The goal was to conquer, or redeem, according to the nationalist ‘Great Somalia’ rhetoric, the Ogaden region. After initial military successes, Barre’s offensive came to a halt as the USSR and Cuba intervened in support of Ethiopia. In less than a year, the Somali forces were in disarray. As Menkhaus (2014) points out, in the aftermath of the Ogaden defeat the pre-existing social contract “between state and society and Somali political elites was replaced by a ‘republic of fear’” (566). After losing the support of the USSR, the regime turned to the West for both military and humanitarian aid. According to Rawson (1994), because of its geopolitical relevance, between 1980 and 1989 Somalia drew about 2.5 billion USD from international donors (171). These funds were funneled into development and military aid, fuelling corruption (according to Laitin and Samatar (1987), 40% of every development contract ended up in the pockets of high ranking civil servants) and flooding the country with small arms (Bradbury, 1997). Lewis (2010) argues that a major consequence of the defeat by Ethiopia was the blurring of the differences in the regime’s “two-dimension political structure, with clandestine clan politics at the core and surface socialist nationalism for everyone else” (213). Moreover, the conflict drove over 1,5 million people from the region and across the border to Somalia. In the former British Somaliland, dominated by members of the Isaaq clan, hostility against the regime in Mogadishu was rife after the arrival of 700,000 displaced from Ethiopia. The government was largely perceived as having a strong anti-Isaaq bias for giving preferential treatment, in terms of assistance and business license, to refugees from Ogaden, from the Darood clan-family of the President (Bradbury, 2008: 55). The government’s policies were regarded as favouring Mogadishu’s mercantile class against northern traders, despite the modernisation changes through which the livestock sector in the North underwent during the 1970s and 1980s. Supported by a widespread network of agents (which helps explain why, after 1991, the Somali financial sector re-emerged first in the North, still today the epicentre of Somali financial flows), the sector became increasingly export-oriented. At the beginning of the 1980s, 75% of Somalia’s recorded foreign currency income came from livestock exports through the port of Berbera, in 84 northern Somalia (ibid.). Yet, despite their contribution to the national economy, northerners saw no benefits from the central government, particularly when, in 1983, Saudi Arabia imposed a ban on the import of Somali livestock following an outbreak of rinderpest, which had serious repercussions on local livelihoods. As initially pacific protests to demand more representation were repressed by the authorities, grievances mounted. In the course of the 1980s, tensions escalated and the Isaaq-based Somali National Movement (SNM) launched an armed insurgency. The Somali government responded with repression and the collective punishment of the entire population. The lethal mix of grievances, rent-seeking opportunities and abundance of weapons triggered a proliferation of armed movements across the country. In Central and Southern Somalia, rivalries ran roughly along clan lines, pitting the Marehan, the president’s clan, against the Ogaden and the Hawiye. During the 1980s, the competition among dominant clans for grabbing land in the fertile area between the rivers Jubba and Shabelle escalated (De Waal, 1997). Taking advantage of the 1975 Land Reform Act, well- connected businessmen, politicians, civil servants and army officers rushed to register parcels of land, obtaining titles from either the central or the regional offices of the registry, with or without the consent of village councils, sometimes, but not always, in exchange for a nominal fee (Besteman, 1999). As De Waal (1997) explains, “a group of capitalist landowners established itself, creating in the process a class of indigenous smallholders and day labourers” (163; see also African Rights, 1993; Besteman and Cassanelli, 1993). The process of social stratification proceeded at a faster pace in Jareer areas, where farmland became particularly alluring for urban elites and members of clans with governmental connections. According to Besteman, this is due, on the one hand, to the fact that those who had reaped profits from refugee relief, remittances and livestock exports needed to invest in durable assets to weather inflation, which had skyrocketed, and diversify investments after the 1983 Saudi ban on Somali livestock had disrupted the main industry of the country. On the other, because of the pressure from international donors, which attached great importance to agricultural programs, the Somali government had liberalised grain prices (Besteman, 1999). The Jubba valley was zoned by the World Bank as the site for the construction of Africa’s second biggest dam 85 and other big aid donors, such as USAID, the German development agency and the European Economic Community funded development initiatives in the area (ibid.). As De Waal (2007) explains, this rush to register properties was not just to acquire arable land for cash-crop plantations, but also to obtain land titles to use as collateral in applications for loans from aid donors. Jareer farmers lacking the education and political access found themselves dispossessed of the land they had been cultivating for decades. In Menkhaus’ (2003) words, “From the 1920s to 1990, the state was used first and foremost to control and exploit Bantu agricultural labor. In the 1980s it was also a vital instrument with which to dispossess the riverine Bantu of their most valuable commodity, their farmland. The state, and the laws which it wielded, were never a friend to the Bantu.” (324) The escalation of political violence in the North precipitated the decay of the state, hollowed out by the rampant corruption, cronyism and inefficiency. In 1988, the regime’s “war with its own people”, as it was called by an Africa Watch report (1990), culminated with the bombing of Hargeisa, the largest city in North-Western Somalia. Between May 1988 and January 1989, the fighting displaced an estimated 300,000- 500,000 people mostly to Ethiopia, but also Djibouti and Kenya, a situation complicated by the Ogaden refugees in Somalia (Gersony, 1990). While the war in the North further weakened the regime, in the South several armed factions grouped into two main armed movements, the Hawiye-dominated United Somali Congress (USC) and the Ogaden- dominated Somali Patriotic Movement (SPM) (Lewis, 2002; Samatar, 1992). Despite the formal ban on clanship disclosure, Siyaad Barre had cynically manipulated and nurtured clan rivalries to underpin the foundations of his regime and divert the attention from political, economic and military failures. As a result, clanship became a politicised issue and kinship networks, traditional sources of solidarity and mutual support, were increasingly regarded as associations to penetrate the state and extract resources (Reno, 2011: 188). As grievances mounted across different segments of the Somali population, Barre’s attempts to control the elites of different groups either by confrontation or cooptation, or often a mix of the two, backfired. A telling example is provided by General Mohamed Farah Aidid, a leading figure of the USC and a prominent member of 86 the Hawiye Habr Gedr clan, appointed by Siyaad Barre as ambassador to India but later entrusted by his clan council to lead the armed opposition to the regime (Shoumatoff, 1994). Launching his military campaign in 1989 from central Somalia, he drew support from Hawiye pastoralists aggrieved by the frequent intrusions of the regime into rangeland controversies with encroaching Marehan kinsmen from the president’s clan. Aidid gave a structure to an already existing insurgency and turned clan based bandits into a militia (De Waal, 2007: 4). In January 1991, the USC and the SPM marched on Mogadishu, prompting the flight of Siyaad Barre and precipitating the collapse of the regime. Conclusions In this chapter, I have examined how Somali power relationships were shaped by hegemonic discourses and historical trajectories. My starting point was to discuss the competing narratives that frame the Somali population either as a monolithic ‘nation in search of a state’ (Laitin and Samatar, 1987) or as ‘people of distinct cultures, customs, traditions, languages, values, as well as destinies’ (Eno, 2007: 132), and to reflect on how the racialised space of the so-called minorities, particularly Jareer farmers, has been carved out through modes of production based on slavery and socio-political marginalisation in colonial and postcolonial times. My argument is that current inequalities within Somali networks are rooted in both mechanisms of social stratification and discourses that legitimize hierarchies. As a result, a large part of the Somali population was prevented from accessing different forms of capital (in the form of financial and non-financial assets, education, political connections, international ties) that would prove crucial when the conflict broke out. The hollowing out of the Somali state was the result of the increasingly violent competition between Darood and Hawiye groups over asset transfer and market control. Later, the civil war and the ensuing statelessness period saw the reinforcement of pre-existing power relations. As I will illustrate in the next chapter, access to weapons, money and international connections exacerbated inequalities among those who were able to reduce uncertainty and even reap 87 benefits from the protracted lack of a central government and those who had limited or no access to resources. 88 3 Conflict, crisis, and mobilities Introduction With the collapse of the Somali state, the power relations inscribed in Somali culture and social structures, and sharpened during the 1960-1990 period, were laid bare (De Waal, 1997). This chapter examines how the eruption of civil war in Somalia further entrenched relationships of dominance. Groups connected to the previous regime with access to remittances or other profitable livelihoods enhanced their military capability to secure their position in the conflict. The humanitarian intervention created looting and new rent-seeking opportunities, such as the management of the aid distribution and the control of the internally displaced people (IDPs) camps. During the years of statelessness, refugee flows changed according to periods of calm and fresh outbreaks of violence. As new internal actors emerged and external players intervened, the Somali conflict went through profound transformations, with repercussions on Somali networks abroad. The restoration of a central state authority in 2012, far from readdressing historical inequalities, legitimised pre-existing ones. The dynamics of the conflict in Somalia had implications for two intertwined aspects: the transformation of the role of clanship and the relationships among clans and displacement patterns3. This chapter describes and analyses these dynamics, with specific attention to the interaction of conflict and mobility patterns, and ends by looking at the emergence of Somali mass displacement to Kenya. 3 A third important change concerns the transformation of political Islam, which is beyond the scope of this thesis. 89 Of games, fashion, and war It is a lazy afternoon in Eastleigh, and I am sitting in the workshop of a tailor, in front of a checker drawn on cardboard. Two players are staring at rows of coloured buttons deployed on it. Each player has his own colour, but the buttons have all the same size and shape. Yet, sometimes a single button wipes out an entire row of opposing buttons; other times, the game ends only after several single moves. Despite all my efforts, I cannot grasp the rules, and I ask my host Sayied Ali to explain. Sayied Ali is the tailor, and also the imam of a small local mosque in a mall. He is also a fluent Italian speaker and, since we first met, he keeps on inviting me in his shop to drink tea and practice the language he learned as a young boy in Mogadishu, and then used as a promising fashion designer in Milan in the 1980s. As I see him with his beard dyed with henna while he heads to the mosque to issue the call to prayer, I cannot really imagine him backstage at the Milan Fashion Week in the 1980s among models ready to hit the catwalk. But Sayied Ali grew up in the cosmopolitan Reer Hamar community of the Somali capital, among traders used to travelling the world and, once back, recounting its differences and contradictions. He lived in Italy for three years, and, before moving to Milan, he had spent one year in Turin, where he attended a vocational training at the headquarters of the International Labour Organisation (ILO). The first time we met he was excited to hear that Turin was my city. He enjoys asking me whether the places where he stayed while in Italy are still how he remembers, and our conversations typically develop from the mnemonic maps of the country he visited when he still had a country to return to. Sayied Ali has been living in Kenya as a refugee since the early 1990s, when anti-Siyaad Barre militias conquered Mogadishu and then fought over its rubble. He has owned this workshop since 2000 and has been able to see Eastleigh changing just by leaning out on the rail of the balcony where we are sitting as we spoke. His workshop overlooks the courtyard of one of Eastleigh’s main business centres, the exchange market, where USD, Kenyan Shillings and gold are traded daily. It is thanks to Sayied Ali that I have been granted access to it for a while, and I have been able to observe how money changes hands and circuits and meaning – as I describe in chapter 4. As for now, though, I am interested in the game. The name is Shantirad, he says. 90 Typically, there are stones instead of buttons, and the checker is drawn on dust instead of cardboard. In Shantirad, the movement of each piece depends on the position of all the others, but Sayed Ali is not very familiar with the rules. It is only later that I find more information on the game, in an article that Eugene Gaspard Marin, an eccentric Belgian anarchist, published in 1931 in the Journal of The Royal Anthropological Institute (Marin, 1931). While travelling in the Horn of Africa, Marin had compiled a catalogue of Somali games, including Shantirad. Related to the Arabian Siga, from which it is likely to originate, Marin observes that the game has many analogies to the Japanese Go - which reminds me of something. In their fascinating essay ‘Nomadology. The War Machine’, the philosopher Gilles Deleuze and the psychoanalyst Felix Guattari (2010) resort to games to draw a comparison between the State and what they call ‘the war machine’, irreducible to State apparatus, “pure and immeasurable multiplicity, the pack, an irruption of the ephemeral and the power of metamorphosis”(4). A substantial difference between the State apparatus and the war machine lies in the greater fluidity of this latter compared to the rigidity of the former, which entails different reactions to and engagement with the surrounding space. Their difference is embodied by the contraposition of two games, chess and, indeed, Go. In chess, ‘a game of State’, pieces are coded, “have a nature and intrinsic properties from which their movements, situations, and confrontations derive” (ibid.: 5). Chess pieces have a structural functioning, entertaining univocal relations with each other and with the adversary’s pieces. Chess is about occupying space, and imposing sovereignty, a sovereignty expressed through the coding and decoding of space. Go, on the contrary, is made by pieces that entertain only relational properties. A Go piece assumes properties according to the circumstances, “it has only a milieu of exteriority, or extrinsic relations with nebulas or constellations, according to which it fulfils functions of insertion or situation, such as bordering, encircling, shattering” (ibid.). Go, they explain, “is pure strategy, whereas chess is a semiology”(ibid.). What is radically different in these two games is the way they conceive space: ‘striated’ in the case of chess, ‘smoothed’ in the case of Go. Go is about “arraying oneself in an open 91 space, [...] holding space, [...] maintaining the possibility of springing up at any point: the movement is not from one point to another, but becomes perpetual, without aim or destination, without departure or arrival” (ibid.). These descriptions reminded me of what happened when the Somali state collapsed, and a myriad of factions started fighting along, and then against each other, splintering in sub-factions, and repositioning themselves as new actors entered the pictures. I begin entertaining the idea that Shantirad encapsulates the logics of the warring parties in the Somali conflict, their need to preserve the “possibility of springing up at any moment”. I later found an apt lens to look at the balance between consolidated rationalities and fluid configurations in North’s (2016) concept of “rules of the game”, with which he describes processes of institutional change. The rules of the game are, in his view, “the humanly-devised constraints that structure human interaction [...] composed of formal rules (statute law, common law, regulations), informal constraints (conventions, norms of behavior, and self-imposed codes of conduct), and the enforcement characteristics of both.”(ibid., 74) These are the institutions regulating the competition among organisations, which are the players of the game. I saw an analogy with the interaction of xeer, the customary law mentioned in the previous chapter, and Somali clans, meant as organisations vying for scarce resources and adapting to the context through shifting alliances. The outbreak of the war in the early 1990s brought these groupings to stiffen their external boundaries, and the rules of the games that had traditionally enabled to peacefully solve conflicts among clans with equal strength, and to guarantee the protection of minority groups through mechanisms of adoption, did not apply to a scenario in which the major competing parties had all interests in maximizing the usage of violence. These rules, however, re-surfaced as the initial wonton violence abated and forms of governance, even in the absence of a government, gradually came into being. I expose to Sayed Ali my views on the capacity of Somali clans to accommodate to the mutating environment. He partially agrees. But he suggests another ludic metaphor, referring this time to a game that was very popular in the alleys of Mogadishu where he spent his childhood. The game is called Shax, and it was a favourite pastime for young and old men around Baraka, Medina and Hamar Weyne markets, and in the 92 shade of the port of Mogadishu where porters rested during the hottest hours of the day. Shax is played on the same board of Nine Men’s Morris, but the goal of the game is to align three pieces in a row (a move called jare) and prevent the opponent from doing the same by restricting his mobility. However, it remains open to negotiation, so that Shax matches looked like loud debates, where even onlookers contributed. A player in a difficult situation can resort to poems and anecdotes to convince the other to break the chain and let him pass. Or he can remind the player who is having the upper hand of the power of his family, and thus keep on moving. Sayed Ali is suggesting that Shax revolves around finding safe escapes in critical situations. It engages with the player’s background and his capacity to adapt to changing scenarios among shifting alliances. This was the case during the humanitarian crisis that followed the fall of the Barre regime. War, Humanitarian Intervention The scholarly debate on the collapse of the Somali state revolves around whether clan rivalries were the cause or the consequences of the conflict. This approach derives directly from the discussion on opposing narratives of Somaliness tackled in the previous chapter, and is reflected in a large literature that Mohamoud (2006) roughly summarised in two distinct interpretations, labelled as ‘traditionalist’ and ‘transformationist’. The former view is embraced by scholars, such as Lewis (1994), who emphasise the role of segmentary ties in Somali politics. It sees in the outbreak of the civil war the re-emergence of a pattern of violence that has been pitting antagonist pastoralist groups, particularly Darood versus Hawiye, through generations. As for the second interpretation, the continuity with the constructivist approach to the Somali clan structure is evident. Authors such as Samatar (1992), Cassanelli (1982) and Besteman (1999) stress the political and economic factors that contributed to the stiffening of clan identities and the mobilisation of clan allegiance to capture state institutions. As De Waal (2007) argues, evoking an interpretation of class that resonates with Bourdieu’s definition of the concept, 93 the description of the Somali civil war as a war between clans obscures the very important ways in which control of resources lies at the heart of the conflict. We must develop a class analysis of the origins and development of the crisis, locating it in the growth of state-mediated capitalist relations in both agriculture and pastoralism, and the key role that control of the state apparatus played in allowing capital accumulation among certain sections of the mercantile class in the 1980’s. (1) The leading figures of the forces that ousted Siyaad Barre in 1991 were members of the elite, close to the regime, that had grown powerful through extraction and asset exploitation. The interweaving of political, economic and military networks shaped a class of strongmen labelled with the mediatically successful, although much contested, name of ‘warlords’ (Adam, 1992). Financial resources and kin ties were mobilised to forge private armies of armed pastoralists. General Mohamed Farah Aidid was the most notorious among the leaders who emerged in a fluid scenario of shifting alliances and rivalries. Reno (2011) mentions also Ali Mahdi Mohamed, a Hawiye at the helm of the Abgal faction of the USC (the other main faction was dominated by Habr Gedr and lead by Aidid) who had amassed wealth particularly through the connections of his wife, a legal advisor to the President; General Mohamed Said Hersi ‘Morgan’, belonging to the Darood Majeerteen Abdirahim, who was an army officer (and Barre’s son in law) and lead the advance of the USC on the strategic port city of Kismayo, in the South, before taking the command of the Somali National Front (SNF), which included former soldiers; and Col. Ahmed Omar Jess, a Darood Ogaden in command of the SPM. Special reference needs to be made to Osman Hassan Ali Atto, ‘the most skilled and pragmatic’ of all the Somali players, according to Western diplomats quoted in his portrait in the New York Times (Lorch, 1995). A wealthy transport tycoon with long experience as a contractor for Western companies, he was from the same Hawiye Habr Gedr clan of General Aidid, of which he was initially the financial backer, before setting up his own militia. He was one of the most prominent among a group of ‘entrepreneurs of violence’ during the early years of the conflict, moving through ‘shadow networks’ (Nordstrom, 2004) straddling national and international, and formal and informal boundaries. The importance of competition over resources for defining alliances and rivalries 94 became evident after the USC and the SPM took control of Mogadishu, fell out with each other, and went through internecine struggles. The USC split between the faction of General Aidid and the one of Ali Mahdi Mohamed, who proclaimed himself the new president. The Somali capital was divided into sectors and became a battleground among warring factions. Militias swelled by recruiting masses of young petty criminals and drifters, often organised in gangs and known as mooryan, whose number had boomed in Mogadishu during the 1980s (Marchal, 1993). Particularly violent was the treatment of Barre loyalists at the hands of USC militias, which was compared by Kaptejin (2013) to an ‘ethnic cleansing’, in which Hawiye targeted Darood clan members and their clients from minority groups. Fierce fighting erupted around strategic assets, such as Mogadishu airport, which was located in Ali Mahdi’s sector. The economic value of these assets increased as the conflict escalated. The ensuing humanitarian crisis drew in new international actors and drove out droves of displaced people, who sought sanctuary in camps in Somalia or across the Kenyan border, where, in 1992, 285,000 Somalis registered as refugees (Lindley, 2013). The emergence of the refugee phenomenon to Kenya will be discussed at the end of this chapter. Havoc and famine in Gosha To suffer the most during the conflict were the minorities who, because of protracted political and social marginalisation, could not rely on militias for protection. However, responses to the war varied across different groups and changed over time. For instance, during the early stage of the conflict, the affluent mercantile communities of the Benadir region suffered "large-scale organised violence" (Mohamoud 2012: 11): Reer Hamar women were assaulted and forced into marriage with militia leaders, while their properties were looted. However, according to Gundel (2009), with the passing of time, some Reer Hamar traders established strategic alliances with key political and military players and, leveraging their commercial links outside of Somalia, succeeded in securing protection. Agropastoralist Rahanweyn villages were also occupied and pillaged by Hawiye and Darood militias after Siyaad Barre established his headquarter 95 in Bay region, on the border with Ethiopia, which was considered a strategic location. The occupying forces singled out and ravaged villages suspected of collusion with the enemy. The inhabitants fled in droves to urban centres, leaving fields uncultivated. According to Menkhaus (2003), the common experience of victimisation forged a shared sense of belonging that paved the way, in 1995, to the creation of an armed movement and a pocket of governance in Bay region with Ethiopian patronage, forging a common Rahanweyn political identity (Bakonyi, 2013). Those who paid the highest price were once again the farming communities of southern Somalia. On their retreat from Mogadishu, militias led by officers of the former regime and splinter factions of the SPM wrought havoc in the riverine regions of Jubba and Shabelle, adopting a scorched earth tactic and creating the conditions for a devastating famine (Reno, 2011). As described in the previous chapter, these areas were settled by communities of former slaves at the margin of Somali politics and society, with no access to weapons to defend themselves, who had also been victims of landgrabbing in the previous decades as Darood and Hawiye elites had seized property of the best riverine locations to create cash-crop plantations. This ‘agricultural proletariat’ (De Waal, 2007) of dispossessed farmers initially welcomed the USC as ‘liberators’. However, new masters replaced the old ones. When the militias of the previous landowners took back the areas, Jareer farmers suffered retaliation for having supported the enemy. In general, far from being a guarantee of protection, their affiliation to dominant clans, or even working on the land of someone associated with a specific warring party, proved to be a liability. Since their ill-defended properties presented looting opportunities to both their patrons and their enemies, Jareer ended up between a rock and a hard place. Ethnic Somalis living in proximity of Jareer communities acquired weapons to assault farmers, rob them of their properties and abduct them to ransom for food (Besteman, 2012). Villages were divided between those who could get hold of weapons, because of their clan affiliation or their resources, and those who were left with no option but to abandon their farms. The disruption to agricultural production due to the constant looting and the systematic diversion of humanitarian aid worsened the effects of the conflict, provoking the death of an 96 estimated 250,000 people between 1991 and 1992 from the combined effects of war and famine (Sommer, 1994). The famine, as summarised by De Waal (1997), “was the legacy of long-standing power relations, not poverty as such” (162). Jareer villages were at the centre of a fierce competition between the Hawiye dominated USC militias and the mostly Darood SPM and SNF, not for territorial gains but for the possibilities of looting and exploitation (Menkhaus, 2003). For two years, the farming area between Mogadishu and Juba was turned into a ‘shatter zone’ (ibid: 334) in which Jareer were regularly assaulted and forced to work for the warring factions. As the agricultural production in the Shabelle and Juba valleys south of Mogadishu stopped, the repercussions were felt elsewhere in Somalia, not only because these areas were the breadbasket of the country, but also because the plantations gave work to farmers from other regions. Yet, the famine was selective. The available food was distributed in order of clan priority. The racial classificatory device forged by decades of exploitation and marginalisation became a system to attribute the right to survive. The appalling situation of Jareer farming labourers is rendered by Menkhaus (2010), at the time working with the UN in Southern Somalia: The virtual holocaust visited upon low-status groups such as the Jareer in 1991 and 1992 was not just a tragic result of warlords and young gunmen run amok; it was also the result of conscious decisions by clan elders and militia leaders over who lived and who died, an ‘allocation of pain’ which reflected the ethics and logic of the existing social order in crisis, and which betrayed the fact that low-status members of the clan simply did not matter enough to live (99). The response to the famine that unfolded in 1991-1992 came initially from grassroots relief organisations, but, as the crisis escalated, international actors stepped in. The Somalia Red Cross Society (SRCS) and the International Committee of the Red Cross (ICRC) partnered to deliver humanitarian aid on a large scale, but the deliveries required adequate protection, which entailed buying the protection of local strongmen. Soon INGOs became entangled in the war economy in different ways: they were forced to pay mafia-style protection, hiring armed guards who actually facilitated the diversion of the aid. ‘Unaccounted losses’ in the ICRC’s balance sheets ranged between 10% and 50% (De Waal, 1997: 169). ‘Ghost villages’ were registered by local intermediaries to 97 divert aid to the militias. Protection costs were accounted for as ‘technical expenses’, from which armoured pick-up trucks that become iconic images of the Somali civil war in media accounts derived this name (Drysdale, 1994: 46-47). Ahmad (2012) traces the boom of the fake NGO industry back to this period, a lucrative sector epitomised by the so-called ‘suitcase NGO’ business model. This model involved three parties: a businessman, a militia leader and a Somali official employed by an international aid organisation. The businessman and the militia leader agreed to hire a skilled professional to draft a request for funding for a bogus IDP camp in an area which international aid workers could not access for safety reasons. A conniving Somali official at the donor organisation would facilitate the approval of the proposal in exchange for a fee or percentage of the profits from the sale of the aid. The businessman created a local decoy NGO to receive the aid, which would then be diverted to the market, and then paid a percentage or split the profits with the militia leader. As a result, while the humanitarian situation worsened, competing factions increasingly fought over the provision of protection to INGOs and access to crisis hotspots, thus triggering a vicious circle. The UN drew widespread accusations of neglect and inefficiency, until the growing media attention helped build up the momentum for the creation of the multinational United Nations Operation to Somalia (UNOSOM) in April 1992, later joined by the US military’s Operation Restore Hope, defined by De Waal (1997) “an extraordinary experiment in philanthropic imperialism” (178). The goal of the mission was to guarantee security conditions for the delivery of humanitarian aid. In mid-1993, the mandate was extended with the establishment of UNISOM II. The goal was to protect relief INGOs by taking ‘necessary measures’ against armed gangs responsible for attacking and looting humanitarian convoys, but the mission quickly turned into a manhunt against General Aidid. It culminated with the notorious Battle of Mogadishu on the 3rd of October 1993 between US and Aidid’s forces, in which the killing of eighteen American Rangers prompted the withdrawal of international forces and accelerated the disengagement of the international community 98 Questioning the failing state narrative Although the botched humanitarian operation helped make Somalia4 the quintessential ‘failed state’ in media accounts (Nenova and Hartford, 2004), the turbulent years that followed state collapse gave way to a period of relative stability (Powell et al., 2006; Little, 2003; Menkhaus, 2002, 2003; Lindley, 2010; Kurtulus, 2012). From 1995 to 2006, with the exception of a short flaring up of violence in 2002, human security improved across the country and for some clans in particular (Lindley, 2010). Pockets of governance were restored in some regions of Somalia, bringing beneficial effects to strategic business sectors, such as livestock, trade, remittances and telecommunication (Van Nooten 2005; Menkhaus 2005, 2007; Lindley, 2010). At the origin of this stabilisation there was a reduction in looting opportunities and in incentives to join armed factions. Most businessmen stopped funding warlords and moved to semi-legitimate trade, particularly in import-export, telecommunication and transport (Menkhaus, 2002). With the combination of statelessness and transnationalisation of the Somali networks, war became a less profitable activity. Cross- border livestock trade in the Kenya-Somalia borderlands, for instance, performed better than during the last decade of Barre’s regime thanks to a set of factors: the absence of coercive policies, a lower tax burden (to non-state actors), the restoration of communal spaces of negotiation for the management of water and grazing lands, the rebuilding of an informal financial system and the flow of remittances from those who had found refuge abroad (Little, 2003). The livestock sector proved to be particularly resilient to the dynamics of the conflict (ibid.). Pastoralist groups had a higher level of livelihood diversification than farmers (for instance, through the commercialisation of milk), and were, therefore, better equipped to minimize risks by moving their assets. Both the activity of the port of Kismayo and the livestock trade in Southern Somalia were mostly in the hands of Darood Ogaden traders whose ties straddled the border, and therefore also had repercussions on the Kenya-Somalia borderland economy. 4 Since, in 1991, the former British Somaliland unilaterally proclaimed independence, from this point on the term Somalia refers to the former Italian Somalia. In 1998, also the northern tip of Somalia, known as Puntland, became an autonomous, although not independent, state. 99 Particularly interesting is the case of the Somali telecommunication sector (Hesse, 2010). Pioneering mobile telephony companies were established in the mid-1990s by businessmen, mostly from the Hawiye clan, who had accumulated financial and social capital in the last years of Barre’s regime. Telcom Somalia (also called Telecom Somalia), established in 1994 by Said O. Olow, a Hawiye Galje’el from Mogadishu, first introduced mobile communication in the country, although the high cost of the technology limited the reach of the service to communication centres (where customers had to share a single mobile handset). The mid-1990s also saw the rise of BGC (Barakaat Group of Companies), which soon became Somalia’s leading company and largest employer. Established by Ali Ahmed Nur Jim'ale, a Hawiye Habr Gedr then based in Mogadishu, at its peak in 2001 BGC was divided among over 600 shareholders (BBC, 2001). On 7th November 2001, in the wake of the 9/11 Terror Attacks, the company appeared in a list of 62 organisations and individuals suspected by the Bush administration of connections with Al Qaeda. All the assets of the group in the US were frozen and thousands of account holders of the BGC financial services had no longer access to their savings. The demise of Barakaat Money Transfer and Barakaat Bank of Somalia paved the way for the emergence of Dahabshiil, which soon became the leading Somali financial institution and money transfer operator, and lead to an overhaul of the entire Somali telecom sector, with far-flung consequences. Strategic convergences The 1990s also saw the convergence of telecommunication companies and MTOs. Over the years, Hawiye subclans, particularly Abgal, Habr Gedr and Dududle, had expanded their influence over the telecom business. The telecommunication sector developed in parallel with the money transfer business, which, especially in South- Central Somalia, was dominated by Hawiye businessmen (while Isaaq-dominated Dahabshiil was popular in Somaliland and Darood-dominated Amal in Puntland), who proved particularly adroit to adapt to frequent power shifts among conflicting factions. Moreover, in the course of the 1990s, the Somali business class forged a strategic 100 alliance with armed Islamist movements (Marchal, 2011). The most prominent organisation, al-Itihaad al-Islaamiyya (Islamic Unity), was established back in 1983 with contributions from the Somali diaspora in the Gulf, from which it derived its ideological Wahabi tenets. Because of its approach that subordinates clan allegiance to the adherence to Islam, it was sidelined throughout the 1980s and during the struggle for power in the early 1990s. It was only in the mid-1990s that succeeded in the establishment of courts led by Qadi (Islamic judges) and entrusted with the implementation of the Shari’a (Islamic law). Businessmen such as the above mentioned Ali Ahmed Nur Jim’ale were more that mere funders: Jim’ale, for instance, played an active role in supporting the creation of Islamic courts across Southern Somalia. The mission of these institutions was to provide a regulatory framework to solve disputes according to Islamic precepts, without denying the importance of customary law but with the declared aim to overcome clan logics. However, they initially failed because of the enduring primacy of clan allegiance within the Somali society. It is worth noting that the benefits of this period of stability were not equally distributed. Between 1996 and 2006, while some groups enjoyed greater security and freedom of movement across Somalia, others became “stuck in protracted displacement” (Lindley, 2011: 16). This made a crucial difference in the capacity of different clans to negotiate the terms of reconciliation in the series of conferences that, over the years, have been promoted by international actors. The TFG and the ‘4.5 formula’ The humanitarian sector played a crucial role in legitimising the claim of dominant clans to speak for the whole Somali population. For INGOs and UN agencies, clans increasingly became a reality to be reckoned with in funding allocations and political negotiations. The humanitarian actors’ engagement contributed to recast clan identities and affiliations, a phenomenon common in conflict areas, and influenced the self- designation of collective actors vying for control over aid (Little, 2003). As Little (2003) points out, 101 The clan system is amazingly adaptable to the changing demands of the international community, as well as to the challenges of statelessness and pastoralism. In fact there is little doubt that the proliferation, fragmentation and – in some cases – consolidation of clan identities were strongly influenced by the presence of outside, resource-rich groups, such as the United Nations and Western development agencies (46) The twelve major conferences that were organised between 1991 and 2002 saw the number of groups skyrocketing from seven in the first meeting to over 800 at the latest (Reno, 2011: 194). These conferences produced first a short-lived Transitional National Government (TNG), and later, in 2004, a Transitional Federal Government (TFG) strongly supported by Ethiopia. Both initiatives, however, failed to create governments of national unity and, instead, distributed power among narrow coalitions based on clan relationships. Particularly relevant to my argument is that the TFG was built on a quota system which attributes equal shares of power to the representatives of the four major clan families: Hawiye, Darood, Dir and Rahanweyn, the latter eventually recognised no longer as a cluster of agropastoralist groups but as a unique clan thanks to the cohesion and greater military capability acquired during the 1990s. Half share (‘0.5’) was attributed to minorities, including Jareer groups (ICG, 2011; Eno and Eno, 2009). The ‘4.5 formula’ recognised clans as legitimate political actors and institutionalised the asymmetries between dominant and minority clans (Eno and Eno, 2009). The 2006 Ethiopian invasion and the rise of Al-Shabaab The establishment of the TFG should be seen against the global backdrop of the so-called War on Terror launched by the US and its Western allies in the aftermath of the 9/11 attacks conducted by Al-Qaida. Ethiopia had every interest in framing volatile Somalia as an ideal breeding ground for other designated terrorist organisations in order to win US support against Somali Islamists backed by Eritrea, Ethiopia’s regional enemy. The US and Ethiopia sponsored a counterterrorism policy based on targeted killings of radical Islamist clerics outsourced to local strongmen, and encouraged an alliance between a new generation of warlords and the TFG, a move that elicited the 102 resurgence of al-Itihaad al-Islaamiyya as al- Itihaad al-Mahaakim al-Islaamiyya (Islamic Courts Union - ICU). The clashes between the Ethiopia and U.S.-supported warlords and the ICU escalated into open war leading to the latter seizing Mogadishu and most of the south and central Somalia in 2006. Despite enjoying large popular support for having restored security, the ICU takeover prompted the Ethiopian armed intervention in the succour of the TFG with the backing of the US, which saw the ICU as a sort of African version of the Afghan Taliban (Hoehne, 2006; Mantzikos, 2010). As a result of fighting pitting Ethiopian and TFG forces against a united front of clans and Islamist armed groups, a fresh humanitarian crisis erupted, with over 300,000 Mogadishu residents displaced (Menkhaus, 2007). After the ICU fragmented and withdrew, Ethiopian troops occupied Somalia. They were shortly joined by a regional peacekeeping force of the African Union’s Peace and Security Operations Division, mandated by the UN Security Council. The force, known as the African Union Mission to Somalia (AMISOM), was composed of only Ugandan and Burundian troops in the first four years and had the task to protect GoS officials and infrastructure and guarantee the delivery of humanitarian aid. However, soon AMISOM troops were mired in an increasingly hostile situation. Widespread anti-Ethiopian feelings fuelled a violent insurgency that brought to prominence the militant group Ḥarakat ash-Shabaab al- Mujāhidīn, commonly known as Al-Shabaab, hitherto the most radical faction of the ICU (Hansen, 2013). Ethiopian forces departed from Somalia in 2008, leaving the TFG weakened, protected by government and AMISOM troops whose reputation was marred by allegations of human rights abuses (HRW, 2014). At the same time, the TFG was increasingly threatened by Al-Shabaab, which kept seizing control of large swathes of territory in Southern Somalia. Al Shabaab presented itself in these areas, mostly inhabited by Jareer farmers, as an inclusive Islamic movement that was not entangled in clan logics, and it succeeded in recruiting many minority members. However, its role in these areas was characterised by a mix of harsh punishment, forced recruitment of the youth, taxation on remittances and strict control of communication with the outside that soon alienated the sympathies of most locals. Many people, especially the youth, fled Al-Shabaab held areas, particularly as the humanitarian situation deteriorated. 103 Famine and relief In 2011, the worst drought in the past 50 years hit Southern Somalia, affecting four million people. The effects of the drought were amplified by the combination of different factors: Al Shabaab’s ban on most international relief organisations; constraints on food distribution related to counter-terrorism regulation; lack of preparedness because of the prolonged absence of humanitarian organisations from the affected areas; aid diversion by armed militias linked to the TFG; and selective distribution of food, which, as during the previous famine in 1991-1992, had an impact particularly on farming and agropastoralist communities, with the result that most victims of the famine were Rahanweyin and Jareer (Menkhaus, 2012; Majid and MacDowell, 2012). This latest factor is particularly relevant to my overall argument because points at the “‘‘privilege gap’’ in Somali society, in which low-status groups lacked the social capital to access relief aid, remittances, and lateral transfers from fellow Somalis” (Menkhaus, 2012: 34). However, throughout statelessness, Somalia has not constantly been mired in conflict. After 1995, as I have previously explained, the violence abated, to return with the rise of the ICU and the Ethiopian invasion. Massive displacements were elicited by the intensified fighting in 1991-92 and again in 2006-07, and by the 2011 famine. The conditions and the trajectories of the displacement varied according to the populations involved. Majid and MacDowell (2012) suggests that important factors to account for are the kind of assets and the level of diversification of the livelihoods of the different groups. Thus, the degree of vulnerability was greater for groups whose key livelihoods were based on fixed assets, such as land, and unable to access portable assets such as cash or gold. Repeated crop destructions and displacement enhanced the vulnerability of Rahanweyin agropastoralists and Jareer farmers, leaving them able to move only short distances and thus limiting their capability to reach food distribution centres. According to a 2013 Human Right Watch report on IDPs camps in Mogadishu, While an accurate picture of the famine-affected population is not available, 104 it is believed that the majority of IDPs from southern Somalia displaced as a result of the famine in mid-2011 were from the Rahanweyn and Bantu communities (24- 25). The report also points to the role of gatekeepers in furthering situations of protracted marginalisation, noting that “in Mogadishu the most powerful clans [...] are from the Hawiye clan group” (ibid., 25). The report also points at the “more limited international reach of the Rahanweyn and Bantu, including fewer links in the diaspora and neighbouring countries” (ibid.) as a reason for the poor social support mechanisms, thus suggesting a connection between limited mobility and vulnerability. The combination of discrimination ingrained in the Somali culture, livelihood constraints and a lack of external support is thus an important aspect to consider when trying to understand inequalities in Somali networks in Kenya, where pre and post- colonial legacies influenced national refugee policies and shaped the trajectories of the Somali diaspora. Somalis in Kenya According to Millner (2009), it is necessary to adopt a historical perspective to understand the factors that conspired to shape Kenyan refugee policy. I would also add that historical legacies help explain how different groups coped with the refugee experience in Kenya. As mentioned in the previous chapter, the 1963-1967 Shifta War and the ensuing state of emergency in the Northern Eastern Province had left a scar on the relationship of the Kenyan state with its Somali citizens. Lochery (2012) suggests that, since Independence, the Kenyan state has repeatedly attempted to institutionalize “the marginality of its Somali citizens” (615) through processes not only “to differentiate Kenyan Somalis from ‘Somali Somalis’ but also to make differences among Kenyan Somalis more visible – or, to use James Scott’s term, legible (ibid.: 616). The presidencies of Jomo Kenyatta and Daniel Arap Moi were characterised by the intense militarisation of the NEP, with frequent round-ups and episodes of state violence. Border control efforts intensified in the 1980s as the Barre regime in Somalia was unravelling 105 and a growing number of Somalis crossed into Kenya. Former officers of the Somali regime were granted asylum, while thousands of refugees were turned away (Milner, 2009). Having relatives in Kenya became crucial to acquire a Kenyan passport, an increasingly valuable asset. Indeed, events in the 1980s highlighted how the relationship between the Kenyan state and the Somali population presented both an inter-ethnic and intra-ethnic dimension. This relationship was shaped by political events that consolidated the clout of members of the Darood Ogaden clan through a series of political and military promotions. After a botched coup attempt in 1982, president Moi rewarded senior army officers who had remained loyal and increasingly relied on members of ethnic minorities, including Somalis. The brothers Mahmoud and Hussein Maalim Mohamed, from Garissa district and the Darood Ogaden clan, quickly reached top posts, the former in the Army, first as commander of the Air Force ’82 Unit and then Chief of General Staff, the latter as Minister of State in the Office of the President in 1983. Also from the Ogaden clan family was the influential Rift Valley provincial commissioner, Yusuf Haji (Lochery, 2012). It is with this process of Darood Ogaden elite formation in mind that the episodes of state violence that occurred during this period should be viewed. Among these, it is worth remembering the massacre perpetrated by Kenyan security forces at the military airstrip of Wagalla, in the proximity of Wajir, where several hundred Kenyan Somalis from the Hawiye Degodia clan were tortured and killed in 1984 (Anderson, 2014). As Anderson points out, Whether the Mohameds acted in a partisan manner or not, after 1982 all clan politics in northern Kenya tended to be viewed through this politicised lens. In Wajir, their shadow would hang darkly over the Degodia. Moi, an admirer of all things British, had learned the tactic of ‘divide and rule: his policies, carried out by the Mohameds would effectively intensify intra-ethnic conflicts in the north. (ibid., 662) Ogaden politicians and top army officers thus used the state apparatus to benefit their own lineages, such as during a screening in 1989 to assess citizen claims among ethnic Somalis in Kenya. In the course of this exercise, the Darood Ogaden elite hijacked the state apparatus to neutralise economic and political competitors, “mirroring conflicts expressed along clan lines in neighbouring Somalia” (Lochery, 2012: 617). Interestingly, the consolidation of a Darood Ogaden elite in Kenya occurred while the 106 demise of the regime in Somaliaaccelerated the flight of those who could rely either on ties with relatives or money to relocate across the border. As many purchased Kenyan IDs, the effect became visible in the size of the Kenyan Somali population, an increasingly politicised issue. The 1979 census put the number of Kenyan Somalis at 516,385, or 3.3 to 3.4% of the Kenyan population of approximate 15.1 to 15.7 million. In 2010, however, the Kenyan Somalis population had reached 2.3 million, or about 5.8%, of the total Kenyan population of 39,423,264. These figures were leaked following allegations of doctored census data at the National Bureau of Statistics and the Ministry of Planning (Chau, 2010: 304). The sensitivity of the topic spoke volume of the anxiety that gripped the Kenyan political establishment as the refugee population was growing. Refugee flows to Kenya Kenya is a signatory to the 1951 UN Convention Relating to the Status of Refugees and the 1969 Organisation of African Unity (OAU) Convention Governing Specific Aspects of Refugee Problems in Africa. However, prior to 1991, refugees were treated as Immigrant and regulated according to the 1963 Citizen Act, the 1972 Immigrant Act and the 1973 Aliens Restrictions Act (Verdirame and Harrell-Bond, 2005). As Verdirame and Harrell-Bond (ibid.) point out, “Kenya never enacted domestic legislation on refugees but, until 1991, its open-door laissez-faire policy was, for the most part, consistent with international refugee law.” (26) The situation changed with the dramatic increase in the flows of displaced Somalis following the eruption of large- scale violence in Southern Somalia. Between 1991 and 1994, over 400,000 people crossed the border into Kenya (Millner, 2009), prompting Kenyan president Daniel Arap Moi to request the intervention of international donors and lift the state of emergency imposed on the NEP in 1964 (ibid.). The UNHCR became responsible for registering asylum applications and determining refugee status, which was thus granted on the basis of the UNHCR mandate. However, hundreds of thousands of Somalis were recognised 107 as prima facie5 refugees, on a group basis and offered temporary protection in camps. This approach overturned the pre-1991 refugee policy, according to which freedom of movement was granted to refugees to facilitate their integration in the national economy.6 Although relinquishing responsibilities to UNHCR for the management of humanitarian affairs, the Kenyan government officially adopted a politics of “abdication and containment” (Millner, 2009: 88), ordering refugees to reside in a camp while awaiting a durable solution (RCK, 2003). Refugee camps were located in the Kenya- Somalia borderland. The main shelter was established on the outskirts of Dadaab village, in the Garissa district at about 100 Km from the Somali border. It soon expanded into three UNHCR facilities, Ifo, Dagahaley and Hagadera, turning Dadaab into one of the largest refugee complexes in the world and an icon of the protracted Somali refugee situation (Horst, 2006; Agier, 2011). Other camps were established across Kenya, including on the coast. An interesting case refers to Southern Somali mercantile communities such as Reer Hamar from Mogadishu and Barawans from Brawa, who escaped by ship and resettled in the outskirts of Mombasa. Thanks to their local commercial partners, often hailing from their same families, many Reer Hamar, and Barawan businesspeople turned the Swaleh Nguru refugee camp, also known as Utange, into a platform from which they successfully rebuilt their regional and transnational trading links (Eno, 2005). It was a windfall not only for the broader Mombasa’s economy but also for members of other Somali groups who, as recounted by Eno, “benefited from the opportunity by venturing into businesses with the Barawans and the Reer Hamar Banadiris.” (252) The camps in Mombasa, along with other humanitarian facilities on the Coast, were eventually closed in 1998. Many traders resettled overseas, but others moved to Nairobi, where the dynamics described above gave momentum to the emergence of a Somali business hub in the middle on the Kenyan capital (see next 5 The prima facie refugee status applies to applicants for asylum coming from regions affected by generalised conflict, such as Somalia. 108 chapter). By 1993, the mortality rates among the refugees decreased and the number of new arrivals dropped. Although the UNHCR declared that the emergency was over (Lindley, 2011), repatriation from Kenya to Southern Somalia was hindered by security concerns. The most vulnerable refugees were unable to reap the benefits of statelessness described above and thus many repatriated Somalis ended up in IDP camps in the surrounding of the urban centres (Hammond, 2014: 6). In 2006, Kenya passed a new refugee law that established a Department of Refugee Affairs (DRA), in charge of the administrative processes of refugee status determination. The law drew a distinction between statutory (those who have gone through the status determination process) and prima facie refugees and stated the right to seek protection within 30 days from the entry, whether by legal or illegal means, into Kenya. It also upheld the principle of nonrefoulement, or the prohibition to forcibly repatriate those fleeing countries in which their lives or freedom may be threatened because of race, religion, nationality, political opinion or membership in a particular social group. An important provision of the Refugee Act referred to the residence requirement within camps, although this encampment policy was mostly loosely implemented (Grant et al., 2012). New peaks in the refugee flow were reached in 2007, when 180,000 displaced people poured into UNHCR refugee camps, and especially 2011, when the refugees reached the staggering number of 420,000 (UNHCR, 2012). The immediate security concerns merged with a more creeping anxiety that the growth of the Somali population might subvert the demographic, religious and political balance of the country. Nevertheless, exposed to violence and harassment from both bandits in humanitarian facilities and Kenyan authorities, most displaced Somalis had little scope for improving their life conditions, unless they resorted to their pre-existing social relationships in Kenya in the NEP or in the main Somali hub in East Africa: Nairobi’s neighbourhood of Eastleigh. Conclusions This chapter has described the events in Somalia since the collapse of the state, arguing that conflict, statelessness, and humanitarian crisis have further entrenched power 109 relationships among Somali groups. It has argued that processes of stratification and accumulation precipitated by the predatory tendencies of the Somali state and then by the fragmentation of the central authority have led to a stiffening of clan identities. Clan identification and racial relations have been reinforced (and in some case reshaped) by conflict dynamics. Pre-existing key livelihoods have informed the way different social groups coped with the conflict and the ensuing humanitarian crisis, consolidating Somali networks. Dominant clans succeeded in mobilising and diversifying resources, simultaneously restricting the agency of traditionally low-status groups. The development of the money transfer and telecommunication sectors were crucial to the construction of cross-border trading networks through which assets were moved and converted. Also, this chapter has described the dynamics of Kenya’s postcolonial history that, after the eruption of the civil war, will pave the way for the emergence of a Somali trade hub in Nairobi. 110 Linkages 111 4 Hustle and bustle in Little Mogadishu Introduction The previous chapters highlighted the mutual shaping of Somali sociocultural structures and historical trajectories and argued that the conflict that broke out in the 1990s widened the gap between dominant clans, in particular Darood and Hawiye, and minorities, especially Jareer farmers. Displacement to Kenya further foregrounded inequalities between those who could leverage on previously available resources and those who lacked movable assets and kin ties in the host country. This chapter and the next one engage with the notion of linkages by focusing on the neighbourhood of Eastleigh, Nairobi, a major hub in East African Somali networks. This chapter in particular examines the emergence of a Somali community (which is referred to as ‘Little Mogadishu’) in the neighbourhood, describing how, over the past two decades, Eastleigh has drawn together a heterogeneous population of migrants from across the region; refugees from Somalia and elsewhere; Kenyans, both of Somali descent and otherwise; and Somalis with foreign passports travelling to Somalia on reconnaissance trips or elsewhere for business. It also discusses the changes through which the area went as a Somali entrepreneurial class emerged over time, turning Eastleigh first into an ethnic market and then into a transnational business hub. The centrality of Eastleigh in the Somali diasporic cartography is underscored by the high number of studies on the neighbourhood (Campbell, 2006; Lambo, 2012; Lindley, 2007, 2010). This chapter expands the focus from the Somali refugee population, at the centre of most studies, to the broader socio-economic texture of Eastleigh, regarded as an interface, a term with which Guyer (1995) refers to “a point of meeting where difference [is] maintained, albeit on changing bases and with changing terms” (8). By retracing the 112 development of the Somali community in the neighbourhood, this chapter illustrates the emergence of a rich communicative ecosystem, shaped by the interweaving of infrastructures of people (Simone, 2004) and technologies, through which information and value are funnelled. This communicative ecosystem has strengthened the position of Eastleigh within the Somali transnational space as a mooring from which mobility projects are enacted, and which facilitates the circulation of capital and the creation of linkages between formal and informal institutions. The road to Eastleigh During the rainy season, the journey from the centre of Nairobi to Eastleigh would take over one hour. But if it was also the beginning of the month, when salaries are paid, and more Nairobians than usual have money to refuel their cars, then the matatus were unlikely to make it through the jammed streets of the Kenyan capital in less than two hours. At the time of my fieldwork, the road to Eastleigh was punctuated by large billboards, mostly with faces of political candidates and slogans in English and Swahili. Some of them were sporting just the national colours, exalting Kenyan diversity and trumpeting the word Amani, “Peace” in Swahili. Ahead of the political elections, due on the 13 March 2013, there were fears across Kenya that that could remain just a word on billboards. The shadow of the violence that had marred the previous elections, between December 2007 and February 2008, loomed largely, and many feared fresh turmoil. This anxiety was not groundless: the International Criminal Court (ICC) in The Hague, Holland, was investigating the role of two top politicians and presidential contenders, Uhuru Kenyatta and William Ruto, in fanning the violence of five years earlier, in which 1,500 were killed and over 600,000 were displaced. Despite the approval of a new constitution in 2010, many issues remained unresolved. In 2007, the violence had concerned the Rift Valley and Western Kenya particularly, pitting the mostly ethnic constituencies of the main candidates. But, this time, it began simmering in the NEP and threatened to engulf some areas of Nairobi again. Eastleigh was among these. Unlike during the previous elections, when the Somali community was left unscathed by the 113 political violence, there were fears that, in the case of upheaval, the neighbourhood’s shopping facilities could be the target of looters. Eastleigh was on a fraught line, bordering the slum of Mathare, the stronghold of the dreaded Mungiki militia. Conspiracy theories ran wild. There were rumours that Indian businessmen were hiring Kikuyu gangsters to arson or loot Somali businesses. And then there was the terrorist threat. Between August and December 2012, 18 people were killed and at least 41 were injured in five attacks attributed to Al-Shabaab7. On at least one occasion, the attack had triggered an anti-Somali riot and, in the ensuing battle between Kenyans and Somalis, local Somali-owned shops were pillaged (Goginieni, 2012). This had exacerbated the sense of insecurity for locals and visitors. Against this background, I was advised by my Somali research assistants and friends to reside outside Eastleigh for the period ahead and immediately after the elections. I thus found accommodation in the South of Nairobi and every morning I took a bus to the city centre and then the matatu N.9 from Tom Mboya Avenue to the South-West end of First Street, just around the corner of Juja Road. Travelling two kilometres north-east from the centre gave me the sense of the distance, and yet of the integration of the neighbourhood in the urban fabric. The passengers on the matatu changed according to the time of the day and the direction. Early in the morning, they were mostly young Kenyan men in working clothes travelling to Eastleigh in search for menial jobs, particularly as porters. Later, there were Kenyan traders. Many of them were women. I would see them again at the end of the morning or in the afternoon, going back to the city centre, holding large plastic bags on their laps or loading bulky bags on the roof of the matatu. There were shoppers and Somali students 7 On the 3rd of August, the eve of a visit of then US Secretary of State Hillary Clinton, a grenade detonated in the area, near the headquarter of the Kenya Airforce, killing one person and injuring six (Al Jazeera, 2012); on 1st October, another grenade was hurled to the St. Polycarp ACK church, off Juja Road, and killed a nine year old boy and wounded nine more (The Star, 2012); on Sunday 18 November, an explosive device ripped through a matatu, killing ten and injuring 24 (Reuters, 2012); on 5th December, an improvised explosive device (IED) went off in Eastleigh’s Joska area, killing one and injuring six (The Standard, 2012); on Friday 7 December, a grenade was hurled in the crowd leaving a local mosque, killing five and injuring eight, including the area MP Abdi Yusuf Hassan (BBC, 2012); on 16 December, a similar attack to the same mosque left one person injured (The Standard, 2012); on 19 December, two people remained injured in two explosions outside the local Al Amin mosque (Al Jazeera, 2012) 114 travelling throughout the day, and they were most of the time very eager to chat. Occasional shoppers usually ended the conversation suggesting extra caution when in Eastleigh. Yet customers continued to flock to this subdivision of the Kamukunji constituency stretching from Juja Road to Jogoo Road and divided into three sections. The area is a grid, bisected by First Street, the commercial heart of the neighbourhood, and intersected by smaller roads crammed full, during the day, of lorries, handcarts and porters carrying heavy yuta bags on their shoulders. Most businesses are concentrated in the central section II, particularly along First Street and its parallel, Second Street. The visible Somali presence has earned the area the name ‘Mogadishu Kidogo’, Swahili for Little Mogadishu, although there are no precise data on the number of Somalis (both refugees and Kenyan citizens) living in Eastleigh. It is highly plausible that most of the 32,435 Somali refugees (as of April 2014) counted by the UNHCR in Nairobi reside in the neighbourhood but, as I have personally observed, and also as it was confirmed in interviews with humanitarian workers, Somali refugees are a very mobile population: many of those living and working in Eastleigh officially reside in Dadaab and Kakuma refugee camps, while others, registered in Nairobi, have moved to different locations without notifying the UNHCR8. Wealthy Somali traders may work in Eastleigh but live with their families in the better-off South C estate of Nairobi. Moreover, the area is home also to Borana Oromos, from the Kenya-Ethiopia borderland, and refugees from Ethiopia and Eritrea. However, the label Little Mogadishu has proved very successful in both Kenyan and international public discourses, conjuring up clear-cut, yet contradicting ideas: good deals, cheap stuff, a large variety of items ranging from textiles to electronics; but also money laundering from illicit businesses and a Somali haven where Islamist militants from Somalia can blend in and plot against Kenya. An article in the Financial Times describes this business district as “built on trade, tax 8 The Refugee Consortium of Kenya (RCK) has calculated that 43% of refugees in urban areas are self- employed, 36% are unemployed and dependent on remittances, and 21% are employed by others (Pavanello, Elhawary & Pantuliano, 2010). According to UNHCR & Danish Refugee Council (2012) 35-45% of Eastleigh residents earn between 16,000 Kenya Shilling (KSh) and 20,000 KSh per month (between GBP 100 and 127 GBP per month), while 10-20% declare an income between 8,000 KSh and 9, 500 Ksh per month (between GBP 50 and GBP 60 per month). 115 evasion, smuggling and regional connections” (Manson, 2012). Abdulsamed (2011) points out that “Eastleigh is at the centre of a network of trade that connects the Arabian Peninsula, Somalia, Kenya and East and Central Africa, with the Somali business community as the common thread”. These far-flung links are rendered visible even by the signs with the names of some of the neighbourhood’s shopping malls: Mandera mall, Bangkok mall, Hong Kong Mall, Emirates Shopping Centre, Medina mall. Other names, such as Bilal, al-Habib or Towfiq evoke Islam, while others, such Amal, or Olympic, celebrate the mundane success of the enterprises (mostly money transfer companies) behind their construction. In the past 20 years, over 40 malls have been built in the area, becoming a landmark of the neighbourhood (Carrier and Lochery, 2013). Most of them are located along First Street and are the economic engine of Eastleigh, challenging the dominance of the Central Business District (CBD), Nairobi’s financial heart. These large multi-storey complexes comprise hundreds of shops (Amal shopping centre, one of the biggest malls in the area, contains almost 500 shops) where textiles, leather, electronics and clothes, mostly imported from the Gulf, China and South-East Asia, are on display along narrow corridors, alternated with tailors’ workshops. Moreover, these malls provide space also to hotels, restaurants, mosques and remittance companies, located on different floors. Besides the shops in the shopping centre, Eastleigh also teems with street vendors, mostly retailers of clothes purchased from local wholesalers. On the Northern end of First Street, strategically located near the terminal of the buses travelling to Dadaab and Kakuma refugee camps, to the Kenyan Coast and Uganda, there are rows of khat sellers9, and clusters of Somali women selling camel milk, a typical Somali delicacy. 9 Khat is a plant native of the Horn of Africa and Yemen. It contains a monoamine alkaloid which makes it a mild stimulant when chewed. Khat consumption is widespread among Somali men, although it is enjoying increasing popular among non-Kenyan Somalis and Ugandans. The leaves should be consumed fresh to give a feeling of euphoria. Therefore, logistics is very important to promptly transport batches of leaves from the producers to the retailers when the twigs are still fresh. 116 Eastleigh’s atmosphere brings back to life what the Somali writer Nureddin Farrah (2010) describes as “one of Mogadishu’s best kept secrets”, the Tamarind market shopping complex, looted and destroyed as the conflict broke out in 1991: This was always abuzz with activities, its narrow alleys filled with shoppers. You could see entire families pouring into its alleys and plazas soon after siesta time, some shopping for clothes, others wishing to acquire what they could find in the way of gold or silver necklaces, many made to order. Stories abounded in which you were told that some of the shoppers came from as far as the Arabian Gulf to strike bargains, well aware that they would pay a lot more for the same items in their home countries in the Emirates or Saudi Arabia. (10) Akin to the Tamarind market, Eastleigh is a crossroad of flows of goods and information shaped by logistic infrastructures (Rinelli and Opondo, 2013). Physical and digital networks interweave in plain sight: the real backrooms of clearing houses, shipping companies and most tiny shops in the neighbourhood are the mobile phones in every trader’s hand. Cutting-edge smartphones and cheap mobile handsets are on display in the showcases of the electronics shops. Concealed from view under woman sellers’ nijab, golden and silver jewels purchased in the Emirates or from other women in need of cash are carried and then traded on low tea tables in the shade of the Exchange Market, behind First Street, where circuits of foreign currencies, cash and digital money (this latter ushered in by the foray of M-Pesa, the Safaricom mobile money service whose green signs dot Eastleigh’s street landscape) merge in intricated fashion – a subject I tackle in the next chapter. Furthermore, the fabric of this new Tamarind market in Nairobi connects worlds geographically and institutionally apart: the interweaving with humanitarian structures is brought to the fore by the UNHCR logo on tents reinvented as sun screeners for stalls or on the bags of rice that find their way from the refugee camps. The mix of social relations that defines the uniqueness of Eastleigh is thus “by no means all included within that place itself. Importantly, it includes relations which stretch beyond – the global as part of what constitutes the local, the outside as part of the inside” (Massey, 1994: 5). This vibrant and variegated landscape contrasts starkly with the poor infrastructures of the area. As soon as the rain season starts, from the matatu stop, right across the church of Saint Therese, Eastleigh’s First Street turns into a broad waterway in rainy days and a muddy field right after. Piles of rubbish clog the 117 roadsides, and people make their way squeezed between the water and the improvised stalls of hawkers. It is a perplexing sight in an area from where, according to Ahmed Mohamed Hassan, the Secretary General of the Eastleigh Business Community (EBC), the leading Somali business association, the City Council and the Kenya Revenue Authority collects over KSh 2,5 billion (GBP 17,8 million) each year, or 25% of the overall Nairobi city council’s revenues10. In general, the Kenyan state maintains an ambiguous relationship with the area: apparently conspicuous for its absence, as suggested by the depleted infrastructures, yet regularly materialising in both the military fatigue of the General Service Unit (GSU), a paramilitary unit of the Kenyan police usually deployed in the neighbourhood in the aftermath of terrorist attacks and, during political campaigns, in the attires of the politicians praising the dynamism of the local business community (Carrier and Lochery, 2013). ‘Openly informal’ Despite the widespread idea among Kenyans that Eastleigh’s prosperity relies on smuggling, formal and informal economies coexist in the neighbourhood in complex ways. Little (2003) claims that “Eastleigh is ‘openly informal’, neither hidden from authorities nor entirely consistent with an official, public place of business” (164). However, this view risks glossing over a difference that, although blurred in practice, still makes sense to Kenyan institutions and Somali economic actors. A closer look at the reality on the ground, in fact, suggests that Eastleigh is not completely ‘openly informal’, but it is rather a space characterized by different degrees of visibility. Institutions recognised by the state, such as the ECB, are in charge of rendering legible the neighbourhood to the state, acting as an intermediary between local businesspeople and political institutions. The ECB, for instance, requires its over 3000 members to 10 In 2010, the EBC had brought the local government to court for negligence, calling into question its authority to levy taxes on businesses and arguing that, despite the taxes paid by local shop-owners, public services were not delivered to the area (Abdullahi, 2010). In June of the same year, the Nairobi High Court had accepted the claimant’s argument, barring the City Council from collecting taxes from over 3000 local business until it fulfilled its obligations (ibid.; Hiiran, 2011). 118 operate under a regular licence, which, for small shops, costs up to 50,000 KSh (350 GBP) per year. This does not apply to those shop owners or hawkers who are not ECB members. As it emerged in the course of my investigation, though, the licence may sometimes refer to a decoy business, which is the case of currency exchange kiosks officially registered as water sellers – a case that I examine in the next chapter. Also, while wholesalers and retailers in Eastleigh are licensed, some of their suppliers may often operate through underground channels. Staple foods such as flour, rice and cooking oil, are delivered from the Dadaab and Kakuma refugee camps, as I discuss in chapter 7. Consumer items, sold at a lower price than elsewhere in Nairobi, may be shipped through the port of Mombasa, and therefore taxed by the Kenyan state; or imported through the Somali port of Kismayo, and taxed by local informal authorities. Somali traders may thus decide whether to receive some or all the goods they deal through the former or the latter route, depending on the possibility to negotiate a better deal at any given moment. In some case, Somali wholesalers relinquish part of their stock to unlicensed retailers without recording the transaction but, particularly in the case of relatives or close acquaintances, creating obligations that would then be repaid in another form. Trade misinvoicing, a system to move unrecorded financial capitals across borders, is widespread. Through the discrepancy between the market value of the received goods and the invoiced amount, the recipient can thus receive payments eluding the scrutiny of tax authorities. This system, as discussed in the next chapter, is widely used by money transfer companies. Being inconspicuous is not a concern unique to some businesspeople in Eastleigh, but it is common to other settings in which “economic actors resort to practices that allow them to remain more or less invisible to the tax authorities” (Muñoz, 2010: 155). Moreover, in the specific case of many unofficial MTOs in Eastleigh, this concern is amplified by the fact that they often operate in a grey area. The effort to pass unnoticed requires thus the investment of labor, which, as already observed by Muñoz in Cameroon, “stems from tactical improvisations and ad hoc calculations” (157). 119 Eastleigh’s socio-economic fabric appears characterised by the intertwining of the formal and the informal and by the preservation of the difference between the two domains, in order to draw from both formal and informal channels. Adopting a formal structure vis-a-vis the state is a strategic move to maximize the leverage of local businesspeople on institutional actors, international organisations, and non-Somali businesses. It is important to consider, however, that formal organisations such as the ECB are strongly influenced by informal institutions, particularly clan assemblies which regulate the relationships among its members, provide support and enforce sanctions. Reflecting the clan composition of the businesses in the neighbourhood, the most influential clans within the ECB are Darood Ogaden and Hawiye Abgal, although the organisation is reluctant to provide data on the clan composition of its members. The ECB indeed embraces a ‘trickle-down rhetoric’ that, emphasising the beneficial effect of business on the whole Somali community in Eastleigh and beyond, makes it a legitimate interlocutor of the state, international business partners, and advocacy organisations. This emerged clearly during my frequent conversation with Ahmed Mohamed, the ECB’s spokesperson. Faced with the question of whether clan matters within the organisation, he regularly skirted the issue. Business was for Ahmed Mohamed a universal language, although spoken with a much emphasised Islamic inflection. He proudly boasted that Chinese businessmen were striving to learn Somali to partner better with local entrepreneurs, and even an Israeli company had recently been awarded a contract to refurbish the local infrastructures. Vigorously denying the existence of market entry barriers for non-Somali businessmen, he interpreted the economic success of the area as a clear sign of the Somali resilience and initiative. For Ahmed Mohamed, trust was once again the key: in his speech a well-rounded concept that spoke for itself. As the spokesperson of the EBC, his role was clearly to ‘sell’ Eastleigh. A well-educated Somali-Kenyan born in Mandera, in the NEP, Ahmed was a regular speaker on Radio Star FM, the most popular radio broadcast in Eastleigh and, in general, among Kenyan Somalis. The radio studios were located on the top floor of the Eastleigh Mall, owned by Sheik Mahmoud Shakul, an influential Darood Ogaden cleric, and businessman from the NEP. On the radio, Ahmed, an Ogaden himself, commented Kenyan politics from a 120 Somali perspective. In those days, his task was mostly to voice the concerns of the Somali business community. Despite his selling attitude, he could not deny that business was stale, and not only because of the recent terror attacks. A time of uncertainty On 12 December 2012, the Kenyan Department of Refugee Affairs (DRA) issued a directive ordering the ceasing of all urban refugee operations and the relocation of Somali refugees to Dadaab (UNHCR 2014). The relocation was subsequently put on a hold on 23 January 2013 by a High Court judge who had accepted to evaluate a constitutional petition of one of UNHCR’s legal partners challenging the directive (ibid.). The rationale for the directive was consistent with the ‘containment and abdication’ approach that inspired Kenya’s policy towards refugees and that was officialised with the Refugee Law approved in 2006 and rendered operative in January 2007. However, this directive stemmed from more contemporary anxieties, brewing at institutional and street level as a result of a combination of factors. The rise of al- Shabaab in Somalia had not only ushered in a new phase of the conflict, bound to have heavy repercussions at the regional level, but impacted upon the popular perception of Somalis in Kenya. Increasing insecurity in the borderlands and in the humanitarian facilities foreshadowed the risk of a spillover of the Somali conflict into Kenya. In October 2011, the Kenyan government launched Operation Linda Nchi (Protect the Country), officially to strike al-Shabaab strongholds in southern Somalia and prevent terrorist infiltrations.11 Instead, the operation achieved the opposite, triggering a spike in terror attacks claimed by al-Shabaab in the Kenya-Somalia borderland areas and in Nairobi, and fuelling suspicions of the Somali refugee population. This added to the increasing outpouring of displaced Somalis from Somalia and the growing presence of 11 The months ahead of the invasion had seen a number of attacks against Kenyan police, foreign humanitarian workers and tourists. However, suspicions of vested interests of the Kenyan government over Southern Somalia have accompanied the KDF operation since the beginning. See Branch 2011; Rawlence, 2014) 121 Somali refugees in urban areas (Rinelli and Opondo, 2013). Fears of terrorist infiltrations added up to the deep-seated suspicion against the large Kenyan Somali minority that, as previously mentioned, has accompanied Kenyan post-colonial history. Security concerns increasingly seeped into refugee policies, to the point where public discourses in Kenya conflated the two issues and Somali refugees were increasingly portrayed by Kenyan politicians and media as a potential fifth column of al-Shabaab (Okungu 2011). Eastleigh was largely perceived as a vital node in a transnational web of terror, as vividly illustrated by the words of a former Assistant Minister for Internal Security Orwa Ojodeh before the Kenyan Parliament, as he compared the war against the Islamist group to the fight against ‘a big animal with its head in Eastleigh, Nairobi and the tail in Somalia’ (Opiyo and Githinji 2011). He then announced “a massive operation to get rid of Al Shabaab and Al Qaeda here in Nairobi" and ended with the promise: “I know Kenyans might be uncomfortable with this, but after two weeks, things will be okay." This was on 20 October 2011. It was not the first time and, as I shall discuss later, it would not be the last. Indeed, it was a pattern: key events were regularly followed by a security crackdown on the Somali refugee population in Eastleigh (Campbell, 2006). This was the case, for instance, in 1999, when repeated security raids following the US Embassy bombing in 1998 culminated with a major ‘police crackdown on illegal aliens’, mostly Somalis and Ethiopians suspected to pose a threat to Kenyan security (IRIN 1999). During 2002, 1,800 “illegal foreigners” were arrested in two roundups (The Nation, 2002; East African Standard, 2002). It occurred again in 2003 with the arrest of over 100 Somalis after the Kenyan intelligence warned of a terrorist threat to the US Embassy in Nairobi (The Nation, 2003). Fear and desire Amalgamating issues of cross-border smuggling, illegal immigration and terrorism, these media reports contributed to reviving the never completely faded ghost of the Somali Shifta, the internal yet foreign bandit threatening the Kenyan state and its Kenyan (non-Somali) citizens. Furthermore, the securitisation of the Somali population 122 in Kenya ought to be seen against the background of the broader global agenda that, in the post-September 11 scenario, strengthened the security-development nexus (Faist, 2005). For a country such as Kenya, located on the front line of the so-called War on Terror because of its proximity to Somalia, a ‘failed state’ in policy discourse, and therefore equalled to a breeding ground for terrorism, donors’ conditionalities were attached to the commitment to fight against the terrorist threat (Bachman and Hönke, 2010)12. What made the Somali ‘alien’ threat more sinister in 2013 was its financial wealth. Piracy provided a clear-cut explanation, although never corroborated by any hard evidence, of the source of capital fuelling Somali entrepreneurs. While the contribution to Kenya’s exchequer of the Somali business community in foreign currency is estimated at USD 780 million per year (Barnes, 2014), the opacity of the financial channels through which these capitals move is not only a widespread source of suspicion, but also softens the popular opprobrium for the harassment and extortion of which Somalis (both refugees and Kenyan citizens) are often victims. Since the onset of the Somali exodus, advocacy organisations have criticised the Kenyan authorities for the treatment of Somali refugees (Otunnu, 1992; HRW, 2010). Despite the political prominence acquired by many Kenyan-Somalis during the Moi and Kibaki presidencies, the business success of the Somali diaspora in Kenya made refugees increasingly appealing to corrupt police, which indiscriminately targeted their victims as ‘walking ATMs’, according to an expression very popular among Somalis in Eastleigh. However, the period following the launch of the Operation Linda Nchi saw an exacerbation of these abuses (Teff and Yarnell, 2013). The Pangani police station on Juja Road became sadly familiar to many refugees, particularly the youth. 12 Right after being sworn in, the Kibaki Government presented to the Parliament a controversial Suppression of Terrorism Bill. Beside allowing security forces the extensive power to arrest and detain in a judicial limbo suspects of terrorism, it included clauses which appeared to be targeting Muslims and Somalis in particular. Among these, there was the provision that allowed the arrest of whomever ‘in a public place wears an item of clothing . . . in such a way or in such circumstances as to arouse reasonable suspicion that he is a member or supporter of a declared terrorist organisation’ (cit. in Bachman and Honke, 2010). The bill was subsequently withdrawn. 123 The widespread belief in the murky origin of Somali money permeated the views of many of my Kenyan interviewees asked about the targeting of Somalis by the police. “That is not their money. It comes from piracy and smuggling”, said Jane, a female passenger I met during one of my matatu rides. However, neither her moral condemnation nor the fear of attacks did prevent her from continuing to travel regularly to Eastleigh from Karagita, a settlement in the surroundings of Naivasha. There, she owned a tailoring shop, and she could not do without visiting her regular supplier of fabrics, “which you must touch and see with your own eyes because the real colours may be different on picture”, she explained. There was no better place than Eastleigh to find good bargains and fine quality fabrics. Only there it was possible to purchase large quantities of goods, because businesspeople were connected to each other, and they could promptly join forces to fulfil even large and short notice orders. This view embodied an apparent contradiction. The Somali presence in the midst of the Kenyan capital was largely acknowledged as a primary conduit for bargain goods. Yet, for many Kenyans, a threat lied behind the piles of fabric rolls and electronics on display. Examining anxious public discourses in South Africa over alien invasion and border policing, Comaroff and Comaroff (2001) well capture a conundrum in which “the border is a double bind because national prosperity appears to demand, but is simultaneously threatened by, both openness and closure” (636). In the specific case of Eastleigh, the coexistence of Kenyan Somalis and refugees further blurs the borders, not only between two sovereign states, but also between two postcolonial state-building projects with diverging trajectories: one, Kenya, economically successful (although regularly marred by political tensions and chronic corruption); the other, Somalia, in shatters. In Eastleigh, the elusiveness of this border, and the risk of an eventual contagion, are embodied by the ultimate tradable good: Kenyan IDs, accessible through the capitals that make the area so alluring to customers (Somalia UN Documents Security Council Report, 2012: 25). For Kenyan visitors, this cognitive dissonance, in which desire and fear are entangled, is elicited by the broadening of consumption opportunities and the threat of a looming havoc. There was a widely held assumption among Kenyans that both goods and extremist menaces moved along the same underground channels. Many 124 Kenyan traders did not conceal their resentment for what considered unfair competition from their Somali counterparts. The Somali capacity to mobilize their social relations to gain a competitive edge on the other traders emerges particularly in the case of the payment of ‘goodwill’. The term may refer to both tangible assets, such as money, and intangible assets, such as reputation and social relations. These assets are used to secure a specific space in a mall or the right to rent one, and are unrelated to the capital required to pay the rent, the deposit, and the stock. The payment of goodwill is a business practice widespread across Kenya since Independence (Carrier and Lochery, 2013), but, in the Somali context, is permeated by logics of clan allegiance spanning national borders. Renting a shop in an Eastleigh’s mall requires the payment of very high goodwill, even before the construction of the mall. As described by Carrier and Lochery (ibid.), the entire process begins with a sign announcing that a shopping complex will be built soon and urging would-be shop owners to secure their space. However, according to my sources, the call for investors is essentially based on word-to-mouth and spreads from the investment companies raising capital for the construction of the building. These investment agencies mostly act as the financial arm of money transfer companies, facilitating the pooling of money, also from a distance. Funds for goodwill are raised with the help of kinsmen in the diaspora or even in Somalia, and can often amount to several thousand Kenyan Shillings, depending on the location of the shop. Shops at the street level or on the first floor are likely to receive more customers and therefore require a higher goodwill. Among Somali entrepreneurs, the payment of high goodwill is possible by mobilizing connections to raise large amounts of money. At the same time, the pooling of capital contributes to strengthening pre-existing ties. Testimony to the high demand for commercial spaces in the area, according to Ahmed Mohamed, is the fact that, while ‘For sale’ signs were widespread throughout Nairobi, in Eastleigh only signs announcing that a plot is not for sale were on display. However, many ‘not for sale’ signs were present in other areas of Nairobi to prevent conmen claiming to have the deeds to the property. His explanation, though, was that “everybody wants a piece of Eastleigh. Instead of luring in developers, we have to fend them off”. The request for a high goodwill payment to secure a shop in a strategic location sets a market barrier to 125 the entrance of non-Somali entrepreneurs, vying for space in Eastleigh but unable to access the required amount. But it also excludes Somalis from minority groups who lack strategic linkages through which they could access financial capital. The few thousand Somali Jareer (according to very vague estimates from the informants, since no data are available) who lived in Eastleigh mostly survived on menial jobs in large Somali shops, or, as I will explain in the next chapter, started small entrepreneurial activities only thanks to the support of INGOs. Among Kenyans thus dominates a blanket view of the Somali community in Eastleigh, which oversimplifies a more nuanced and uneven reality. It enshrines an idea of trust ingrained in a primordial (and, I earlier argued, to some extent fictional) unproblematic Somali identity and standing for both mutual help and connivance. Nevertheless, the Somaliness inscribed in the streets of Little Mogadishu is multilayered and multifaceted, shaped by the ebbs and flows of human tides displaced by conflicts or lured by opportunities, since before the beginning of the Somali conflict. Little Mogadishu rises The dramatic increase of the Somali refugee population in Kenya during the 1990s was an important factor in the development of a key Somali migrant hub in Nairobi’s Eastleigh neighbourhood. Yet, it was not the only factor. In fact, other elements at global, regional, national and local level contributed to boosting the significance of the area for Somali refugees and businessmen. The convergence of these factors increased the centrality of this neighbourhood within trade networks spanning across East Africa and as far as China through the Gulf States and India. The origins of Eastleigh are rooted in colonial urban zoning. At the beginning of the 1900s, Somalis from the British-ruled Northern Somalia followed the British army to Nairobi to build the Mombasa-Uganda Railway along with Asian workers brought from India. Mostly Isaac, they clung on their pastoralist lifestyle and kept livestock in the initial area of settlement, near Ngara. However, fearing an outbreak of zoonotic diseases, in the 1910s the colonial authorities ordered the transfer of the Somali population to the 126 Eastleigh area, home at the time of many Asians who had settled on land purchased by an Indian businessman (Goldsmith, 1997; Carrier and Lochery, 2013). Keeping alive regional cross-border connections following Kenyan Independence, these first Somalis occupied a market niche by providing logistic services, mostly working as truck drivers for Indian traders operating across East Africa. During the 1970s and 1980s, Eastleigh became thus a pole of attraction for Somali Ogaden from the NEP, who moved from their impoverished area to Nairobi in search of a job. Small groups of refugees from Somalia and Ethiopia, harbingers of the waves to come, reached Eastleigh on the trucks and buses travelling from the borders and settled there. It was at the beginning of the 1990s, as the unravelling of Somalia precipitated, that the flows of refugees intensified. As summarised by Carrier and Lochery (2013), “With these refugees came ‘pieces’ of politics, economics, culture, and social relations from back home, and much capital that would drive investment and radically transform the urban landscape” (336). As the war in South and Central Somalia broke out, most displaced people sought assistance at the UNHCR humanitarian facilities in the borderland, particularly Dadaab. However, urban areas in the NEP and Eastleigh became a magnet for those who had still access to financial assets and could rely on relatives. For many Somalis, Eastleigh was initially a transit space where to wait for resettlement to the West or the end of turmoil before returning to Somalia. As the conflict endured, and the influx of refugees increased, though, new opportunities arose. The growing Somali presence in the estate spurred the proliferation of hospitality structures, thus creating job opportunities that drew more Somalis both from the NEP and from the refugee camps. The opening of the Garissa Lodge, Eastleigh’s first shopping mall, traces back to that period. The building was since the 1980s a Kenyan-owned hotel where refugees could find affordable accommodation. The rooms were turned during the day into market stalls where different items were on display, particularly textiles, leather goods and khat, but also goods smuggled from Somalia and foodstuff diverted from the refugee camps and on sale at a competitive price. In 1992, a Somali businesswoman purchased the building, and the first mall was established. It comprised 92 shops (Carrier and Lochery, 2013) and, soon, the reputation of the Garissa Lodge spread across Nairobi and 127 the country, luring both customers and investors. According to Marchal (1996), among the first big traders were some Reer Hamar from Mogadishu, who had previously liquidated their business in Somalia and, as mentioned earlier, initially resettled among business partners and relatives on the Swahili Coast. Local sources also suggest that the impetus for the development of the neighbourhood was given by the financial capitals that members of the Somali elite who had resettled in Europe or Gulf States injected into the local economy. Darood businessmen, mostly from the Ogaden subclan and able to rely on Kenyan relatives, were pioneer investors in the area. A logistic industry flourished to cater to the needs of those who sought resettlement to a third country and therefore needed communication centres and MTOs, where it was possible to use phones and, later, the internet to communicate with relatives abroad and solicit the sending of remittance. Changes and consolidation A broader range of goods flowed into Eastleigh during the 1990s, as liberalisation policies were implemented in Kenya (Read and Parton, 2009). These policies also interested the meat market, upon which Somali brokers in the NEP held sway, and whose profits also contributed to fuel Eastleigh’s trading sector and construction spree (Little, 2005). Transformations at local level interacted with changes in global trade, notably the emergence of China as a global manufacturer of cheap items and therefore a driving force behind a process of low-end globalisation (Mathews, 2007). Somali entrepreneurs tapped into this circuit of low-cost goods through Dubai. As explained in chapter 2, a Somali community had grown and established business relations in the Gulf during the 1970s oil boom. Other resourceful Somalis had transferred assets and resettled to the Emirates in the 1980s, as the Barre’s regime was falling apart, and after 1991. Wealthy Somalis from the Darood and Hawiye clans found themselves living, interacting and doing business with long-established Arab, Indian and Iranian traders, but also with members of other conflict-displaced diasporas, such as Lebanese and Palestinians. According to Marchal (2005), the aggregation of business classes from 128 different backgrounds has been a key factor in the transformation of Dubai into a global financial hub, together with the role of the Emirates as a neutral player in a complex geopolitical scenario, bridging between the two regional rivals Saudi Arabia and Iran, and the business-friendly attitude of the country’s leadership (98). Dubai often features in the accounts of Somali traders, and the frequent references to Eastleigh as a new Dubai are steeped in the fascination for the city and its momentous development. For Eastleigh’s business community, Dubai represents an ideal model to aspire to, to the extent that the second shopping mall inaugurated in 1993 was indeed named after the Emirates’ capital. Initially, the goods were shipped to Somali ports (particularly Kismayo) and then smuggled through the border. In the course of the 1990s, though, the decrease in customs tariffs at Mombasa port rerouted the shipping of textiles from the Gulf (Carrier and Lochery, 2013). This route did not replace the cross-border smuggling completely, but rather provided a viable alternative when travelling through Somalia became too expensive because of the proliferation of warring factions and, therefore, of checkpoints. Somali businessmen relinquished the second-hand clothes market (known as mitumba in Swahili) to Kenyan traders in the Gikomba neighbourhood, adjacent to Eastleigh, and took for themselves the commerce of “fine quality” counterfeit designer goods. The business success of Eastleigh thus shows two interlocked dimensions: an internal, as it is related to a specific ethnic market niche; and an external, as it builds upon pre-existing transnational networks. The dynamics leading to the emergence of consolidation of ethnic markets are explained by Kloosterman (2010) as he points out that the formation of spatially concentrated settlement of (mainly first generation) immigrants in urban areas strongly contributes to the articulation of this [ethnic] demand. Immigrant entrepreneurs are usually much better positioned to benefit from these opportunities as they tend to have the required knowledge of products, suppliers and consumers. They have, moreover, the necessary credibility to cater for these niche markets of co-ethnics. (33) The trajectory of Eastleigh’s development is consistent with the ones of other ethnic economies in migration clusters, starting with “the use of a common cultural bond for economic survival and advancement” (Portes and Stepick, 1985: 499). Somali business in Eastleigh thrived thanks to the combination of an ethnic market and a pool of 129 cheap labour, mostly related by family ties, recently arrived from Somalia or the refugee camps. These workers were willing to accept low wage because excluded from the formal secondary market and because an entry level job would have helped prove their skills and trustworthiness to the business community. However, as already argued by Portes and Stepick (1985) in their study on Cuban and Haitian immigrants in Florida, while some “immigrants may spell a tour of duty at the worst jobs [and then] they often move quickly into self-employment through the support of preexisting ethnic networks” (494), other groups, lacking the support of an entrepreneurial community, may remain stuck at the lower level or, as in the case of many refugees from minority clans, alternate periods of work in town with reliance on humanitarian aid in the refugee camps. In the early 1990s, resourceful refugees partnered with kinsmen with Kenyan ID to give a legal façade to their business. Internal and external migration flows reinforced each other, as the Somalis who made their way to Nairobi directly from Somalia, from the refugee camps or cities in the NEP, forged partnerships based on clan ties. In 1997, Somali traders began using the Eldoret Airport, recently opened, as an entry point for their goods. The introduction of the East African passport in 1999 contributed to streamlining border crossing for East African citizens, thus facilitating the inflow of suitcase traders from neighbouring countries, particularly Uganda and Tanzania, and the transport of a wide range of goods outside of Kenya. As Little (2013) summarizes, “those with capital and informal businesses were able to take advantage of Kenya’s neoliberal reform program, and the Somali business community fit this bill” (166). Moreover, the development of Eastleigh continued to resonate the changing political and humanitarian situation in Somalia. Fresh waves of refugees arrived in 2007 and 2011, triggered by new humanitarian crises in Somalia (see the previous chapter). The booming Somali population increased the local demand for housing and other services, such as money transfer and telecommunication, and enlarged the co-ethnic pool of cheap workforce. This allowed selling goods at a competitive price and thus expanding beyond the boundaries of the ethnic market. At the same time, the improvement of communication, discussed below, enabled a better coordination between Somalis in Kenya and Somalia, leading to the emergence of new commercial routes. 130 The case of MJ, a businessman dealing in gentlemen’s suits and shirts with a shop on the ground floor of the Bilal shopping mall, was particularly telling of the capacity to adapt to a shifting scenario, at local, regional and global level. A Darood Dulbahante from northern Somalia, MJ arrived in Kenya for the first time in the late 1990s. He hailed from a family of traders and his kin network stretched from the Gulf to Uganda, passing through Somalia. He referred to Somali refugees with some disdain and mentioned with pride that he never had to live in a refugee camp. However, he had a UNHCR refugee card and a Kenyan alien card. From the resources he relied upon, I had the impression that his status of refugee was less a constraint than a strategic choice. He did not actually perceive himself as a refugee, since he attached that label to people living in camps and relying on humanitarian aid – and he was not: he was residing in a comfortable hotel in Eastleigh and regularly travelled to Somaliland and Uganda. On top of this, his refugee card did not prevent him from expanding his business. His shop was registered under the name of a ‘cousin’ (ina abtida) who was a Kenyan citizen, and until 2012 his suppliers were in China and he received goods through Dubai. In 2012, though, another of his cousins in Mogadishu received a scholarship from the Turkish Embassy to study in Ankara13. He mentioned that his application was ‘strongly supported’ by the Somali agency that assigned the scholarship, and, once in Turkey, his cousin found new suppliers offering better quality products. MJ’s case brings to the fore how cross-border links with Somalia not only persisted, but adapted over time and in fact turned dispersion into “a spatial resource” (Ma Mung, 2004). Defined as “a spatial arrangement that is available for the subjects of dispersion to use for their personal benefit” (ibid.), this notion can also be expanded to another kind of dispersion, across juridical statuses. This resonates with what observed by economic anthropologist Rhoda Halperin (1990) in her study of livelihoods in North- Eastern Kentucky as she notices that members of the same kin networks make ends meet by engaging in a variety of productive activities and with different institutions, thus “crossing spatial, institutional and cultural boundaries” (139). Also MacGaffey and Bazenguissa-Ganga (2000), in their study of Congolese trading networks between Central Africa and Europe, observe that the traders tend to maximize business 13 Since 2011, after a historic visit of then Turkish Prime Minister Recep Tayyp Erdogan, Turkey had strengthened its ties with Somalia, funding development initiatives and offering scholarships to Somali youth (ICG, 2012). 131 opportunities by establishing links with multiple institutions and stretching institutional participation to its limit (99). Similar dynamics can be observed in Somali trading networks: relations spanning different institutional arrangements guarantee access to resources allocated from different institutions – a point that is particularly relevant to my argument on the preservation of power relations in transnational spaces. Driven by a policy-relevant orientation, most studies on the Somali population in Eastleigh tend to isolate the refugee population, focus on their interplay with humanitarian or state institutions (Campbell, 2006) and gloss over clan differences and power relations (Goldsmith, 1997). Nevertheless, the links of this population with Kenyan Somalis, Somalis with foreign citizenship or Somalis in Somalia matter. There are differences among refugees who can tap into multiple sources of support and those who can draw only from humanitarian agencies or from precarious, pauperised communities. Mirroring the Somali situation, the changes which Eastleigh went through since the 1990s seem to have entrenched structural differences among dominant and minority groups. There are no data on the relation of clans and business ownership in Eastleigh. However, a survey conducted by one of my field assistants for a Western embassy in a sample of 200 shops in prime locations (ground and first floors of the main shopping malls) found that Eastleigh’s largest companies are almost equally divided among Hawiye Abgal and Habr Gedr (55%) and Darood’s main subclans Ogaden, Mareexan, and Majeerteen (40%), while members of other clans own the remaining 5%. Although I had no possibility to verify these figures, it was generally held that Somalis hailing from clans controlling social and political resources in Somalia appeared better positioned to access assets such as documents and money, which are crucial to set and expand businesses outside Somalia. These relations are critical to raising capital for goodwill and, in general, to integrate with the local economic fabric. Eastleigh’s communicative ecology To understand how this integration is worked out in practice it is necessary to consider how the presence and the development of ICTs simultaneously acted as a 132 powerful pull factor for both Somali refugees and internal migrants and contributed to turning Little Mogadishu into a ‘translocality’, a concept which refers to “the emergence of multidirectional and overlapping networks created by migration that facilitate the circulation of resources, practices and ideas and thereby transform the particular localities they connect” (Greiner, 2011). The discussion on translocality stems out of the broader debate on transnationalism as a reaction to what is seen as an excessive emphasis on notions of deterritorialisation and ‘uprooting’ while eclipsing the place- based processes that shape spatial relationships and identities (Oakes and Schein, 2006; McFarlane 2009; Brickell and Datta, 2012). It suggests instead an approach that takes into account how transnational relationships are grounded in specific localities and “deterritorialised networks of transnational social relations take shape through migrant agencies” (Brickell and Datta, 2011: 3; see also Conradson and McKay, 2007; Gielis 2009) The notion of translocality is a useful theoretical lens through which to look at Eastleigh not as a bounded space but rather as a crossroad of networks, reproduced through daily practices linking migrants to their areas of origin and destination. The interaction of these networks reshapes the economic and logistic organisation of the urban space (Schmoll, 2012). In their investigation of translocal linkages between Italy and France, Schmoll and Semi (2013) remark that translocality gives prominence to specific urban areas within cities. Indeed, when talking about the nodes of their transnational networks, Somalis mention Eastleigh rather than Nairobi (and, as I show in chapter 6, Kisenyi instead of Kampala). As previously suggested, a key factor in the transformation of urban space into crossroads of networks straddling borders is the concentration of logistic services, from communication centres to MTOs to travel agencies, which have gone through important changes following the diffusion of ICTs. However, to fully grapple with the way ICTs contribute to the mutual shaping of translocal connections and people’s actions (Steinbrink, 2009, in Grenier, 2011), it is necessary to take into account the broader local ‘communicative ecology’ of Eastleigh. This concept refers to the complex and multilayered system of institutions, codes, technological artefacts, that makes possible symbolic exchanges among the members of a community (Slater and Tacchi, 2004: 3; Panagakos and Horst, 2006). Regular 133 interactions play a crucial role in the creation of a ‘community of exchange’ (Humphrey and Schmitz, 1998: 37; Overå, 2006). As Overå notices in her study on telecommunication and urban traders in Ghana, “communication, screening, observation, and control are particularly complicated when the process of network building extends widely across space” (Overå, 2006: 1302). The Somali spatial clustering in Eastleigh simplified this process. A key role of Eastleigh’s “communicative ecology” was to craft a sense of belonging to a shared imagined community. “What attracts Somalis to Eastleigh? Other Somalis”, explained a Kenyan Somali. Among refugees, Eastleigh came to be defined by a peculiar balance between two apparently conflicting feelings: buufis, or longing for resettlement (Horst, 2006), and fadikudire (Lambo, 2012), called by someone also hiloowid, or nostalgia for the homeland. Horst (2006) has extensively discussed buufis as an obsessing decision- making driver. The relevance attached to this concept in informing Somali strategies echoes other studies that have explored how mobility aspirations are nurtured and projected in space and time (see Chu, 2010). Like in the case of Chinese Fuzhounese would-be migrants at the centre of Chu’s study, also the concept of buufis is framed as a ‘disease’ that would drain one’s energies and time (Horst,2006), and animates the intense exchange among asylum seekers and relatives already resettled and able (or believed so) to provide economic support or act as sponsors. On the other hand, young Somali refugees envisage an undetermined future in which peace (nabad) would be restored in Somalia, thus paving the way for the return (Lambo, 2012). This yearning is permeated by a nostalgic feeling for a country that emerges from the conversations as a socio-cultural construct, rather than a historical reality. In the course of my fieldwork, I was invited to attend several student association gatherings in Eastleigh, in which fadikudire was expressed through paraphernalia, such as Somali flags, banners with the clearly recognizable shape of Somalia, starred hats or t-shirts, and the initial singing of the Somali national anthem. However, every meeting ended with several students wishing to exchange phone number, email and Facebook contact and enquiring about possibilities to be sponsored for scholarships abroad and, possibly, more durable 134 solutions. There was no contradiction between the idea of obtaining resettlement to and, hopefully, citizenship in a Western country, preferably in North America or Northern Europe, and the wish to contribute to the rebuilding of Somalia. The tension between these two feelings is a thread running through the migration projects of many young Somalis. It also informs a heterogeneous community that draws a collective identity from sharing, at the same time, mobility aspirations (although formulated in different ways) and ties to Somalia. The concomitant influx of people and capitals to Eastleigh enabled the expansion of the technological infrastructure upon which a rich communicative ecology has flourished. Long-time Eastleigh residents still remembered, back in the 1990s, the long queues outside the first communication centres. At the time, refugees relied mainly on a two-way radio system used in camps, urban centres and Somalia called taar (Horst 2006). In Dadaab, the taar was the only channel of communication with family members resettled or still in Somalia, although very inefficient. Besides the long queues, an often mentioned problem of the taar was the lack of privacy in the conversations, which was a serious hindrance when refugees needed to ask for a financial contribution to relatives abroad: this could affect the security of the requester, exposing him to the risk of thefts by other eavesdroppers. Like many others in Eastleigh, Yussuf and his brother Omar, both Jareer from Lower Shabelle, had spent their first period in Kenya in the refugee complex of Dadaab. According to Omar, apart from the long wait people had to go through, a major problem of the taar was the lack of privacy in the conversations. “If one was without money and was complaining to his mother, everybody could listen to his plight. That could be embarrassing. Moreover, people nearby could understand that you were requesting money and so they could decide to rob you” Taar posts were widespread across Eastleigh, where they lingered in a juridical limbo, officially illegal but tolerated by the police in exchange for bribes (Horst, 2006). With the passing of time, the radio equipment was replaced by VoIP adapters that connect landline phones to the internet. Taar posts gradually gave way to 135 communication centres, or ‘branches’, how they are colloquially called14, where a system made of several telephone VoIP adapters allowed to make calls at a very low cost. The branch was the cheapest way to communicate with relatives in Somalia or refugee camps. Yussuf pointed at the amount of time required to reach family in Somalia, where there was often a single communication centre for clusters of villages. “You had to call the branch on the other side, speak with the operator and tell him that you needed to talk with your brother or your mother and that you would call back the day after. At this point, he had to send someone to look for them and tell them to come at the branch the following day. So the day after you called again and they were there or maybe not and you had to wait. In general, it took it a lot of time.” As of May 2014, only two communication centres survived in Eastleigh, mostly used because the lower fees allowed for longer conversations15, for instance to offer condolences for the funeral of a relative or greetings for a wedding. The decision to resort to communication centres, in these cases, was related to the nature of these specific events, where an open display of grief or felicitations is socially commanded and therefore communications have to be made public. In general, the decrease in the number of communication centres coincided with the growing diffusion of mobile phones across Kenya. Mobile handsets had profoundly changed the way Somalis, not only in Eastleigh but in the entire world, circulated information and money, with repercussions on mobility patterns and decision-making. Eastleigh’s voices The popularity of mobile telephony in Eastleigh is part of a broader trend in Africa and in the Global South that has spurred a significant amount of academic, grey and 14 The name ‘branch’ derives probably from ‘private branch exchange’ (PBX), a telephone exchange which creates a network of several devices and connects them to a public telephone network. 15 Calls to Somalia costed 20/25 KSh for ten minutes while calls to the US costed 2 KSh for ten minutes. 136 popular literature, some of which imbued with a ‘mobile phone revolution’ narrative (Giridharadas 2010; Etzo and Collander 2010; The Economist 2011). However, some of these studies offer useful insights to understand how people engage with mobile technology through patterns that are consistent with specific cultural features (Horst and Miller, 2006; Ling and Horst 2011). These works have highlighted how mobile telephony reshapes, rather than replaces, previously established forms of communication. The implications of this integration are particularly significant for the construction and the maintenance of social ties. Several studies have examined the repercussions of mobile technologies on traditional power structures, intensifying existing social networks (Palackal et al. 2011), through the practice of ‘linking-up’ (Horst and Miller 2006), helping improve coordination among the members of a group and abate transaction costs (Aminuzzaman et al., 2001; Overa, 2006). Some scholars claim that mobile telephony plays a limited role in expanding one’s social sphere (Ling and Horst 2011; Ito and Okabe, 2005; Ling, 2004, 2008), while others underline its potential for increasing social contacts, thus allowing categories that have limited social interactions, such as women, to explore new forms of sociality, and to improve the efficiency of market and business relations (Molony, 2008; Tenhunen, 2008). The popularity of mobile telephony in Eastleigh reflected the boom of the Somali telecom sector, an ‘unlikely’ success story (Sheikh and Mohamed 2009; The Economist 2011; Gagliardone and Stremlau 2012) which had thrived in a regulatory vacuum to meet the needs of an increasingly globalised diaspora. As Lindley (2009) points out, there is a mutual strengthening relation between money transfer and communications industries: [The financial] sector rapidly incorporated new technologies to relay payment information, from radios to fax, mobile and satellite phones, to email and dedicated company software. Money transmitters’ substantial investments helped drive Somalia’s telecommunications revolutions. (523) The lack of a central authority has given free rein to businessmen to set their own companies without the burden of fees, taxes and legal constraints which characterize the telecom sector elsewhere in Africa. Fierce competition among entrepreneurs has kept prices low and today it is possible to call cheaply to and from Somalia. According to Harper (2012), the Somali oral culture has also contributed to making radios and phones 137 the favourite means to circulate information in a context in which the sound of the voice is deemed very important to seal agreements. Along with the financial industry, the Somali telecommunication sector is often mentioned as evidence of ‘what works in Somalia’ (Hesse 2010). Clanship plays an important role on both the technological and marketing side of the business: trusted relations make possible the purchase of the required technology and the retention of customers (ibid.). The impact of mobile telephony on livelihoods was conspicuous in Eastleigh. Scratch cards were sold in bulk in large shops and by retailers, and most malls hosted several phone shops selling brand new and second-hand devices and offering services of repair and unlocking. In the last years, mobile phone shops had cropped up to meet growing demand, not only from Somali customers: customers arrived from all across Kenya and even from Tanzania and Uganda to buy mobile phones (on sale at a lower price than anywhere else) in bulk. Somalis claimed that, in Eastleigh, mobile phones spread earlier than in the rest of Nairobi. The first models were brought in the neighbourhood by Somalis living overseas. Some were given to family members while others were put on sale. Some returnees purchased several second-hand devices in their countries of resettlement and re-sold them in Eastleigh, either at home or in the shops of their relatives. This first generation of phone dealers learned to repair and unlock phones through word-to-mouth and, later, videos on Youtube. Trading in mobile phones, accessories and services became thus a source of income for many who could rely on family members overseas and were able to intercept the growing demand. The decreasing prices of both handsets and services, with the diffusion of pay-as-you-go schemes, have been a primary driver in the massive uptake of mobile phones, particularly in a context characterised by poor landlines. Mobile phones have smoothed the flow of information through pre- existing social networks and have enhanced the coordination of distant business partners, contributing to turning dispersal into a resource, for both coping with unpredictable events or expand businesses (such as in the case of MJ). 138 The case of Bilal, an 18-year old boy who left Mogadishu with his family in 2011, was particularly of about the importance of mobile communication in a highly volatile environment. The whole family travelled to the Kenyan border in a crammed taxi. The driver used a mobile phone to receive updated information along the route on fighting or the presence of bandits. When all the passengers safely reached the border, the driver collected their telephone numbers in order to expand his pool of contacts, the critical resources on which he both survived and made a living. The moral of this anecdote is highlighted by Abdullah, a young Somali who found himself in Nairobi after having lost his closest family members on the way from Southern Somalia: “In a group you survive, alone you die.” After struggling to find a job in Eastleigh, he asked for help to relatives in Kakuma and, after receiving money via mobile phone, travelled to the refugee camp to wait for better times. Mobile phones have also facilitated communication to Somalia. Omar recalled that, before owning a mobile phone, every summoning to the branch was a very distressful experience, as he worried about both the news coming from Somalia and the fee he had to pay also for his counterpart. The possibility to speak from anywhere anytime is still a source of amazement (and one that deserves a God’s blessing upon Safaricom). In the follow-up of an interview, for instance, Yusuf Abubakar sent me the following SMS: [...] Myself i just pray the inventer of mobile communication system and is mobile is the settler of the heart recently i know the updates of the family I sleep with the cell phone like my id no more warry about the fmily (sic) Mobile telephony further expanded the range of the sources of information upon which people take decisions. This point was stressed by Mohammed, a 23-year old Somali from the Rahanweyn clan who left Baidoa, in Somalia, when he was 4. After 14 years in Ifo refugee camp, he had moved to Nairobi and he currently worked in an English school in Eastleigh. He communicated almost daily through SMS with his brothers who lived in Canada and the UK and regularly with the ones who were living in Somalia, Ethiopia, and the NEP. Mohammed claimed that this regularity had contributed to developing a sense of presence in the lives of his brothers in the West. 139 “I speak with them almost every day. They tell me what they go through, the difficulties they find, the lack of a job. Everything is very expensive in Canada and Britain, they tell me that every day. Getting a visa is very difficult. And also going there can be dangerous. Everybody understands that over there money does not grow on the trees” According to Mohammed, the possibility to receive regular updates from trusted sources in the diaspora influenced migration expectations and strategies. Mohammed, for instance, had recently secured some funding from his brothers to start a business. He planned to open a convenience store in Juba, South Sudan’s capital, where some of his relatives had moved just after the Independence from Sudan and informed him of opportunities over there. Mobile phones were also used as radio transmitters. Since 2011, the above mentioned Radio Star FM broadcasted on demand by dialling a phone number. The radio was a pervasive presence in shops and cafe. It broadcasted in English and Somali and its programmes in Somali were widely popular among Somalis in Somalia and the diaspora, as they encouraged the participation of the listeners, who used their phones not only as primary means of reception (because they were easy to carry and, unlike transistor radios, did not require a periodical change of battery) but also to interact with the journalists. The radio covered Somali or Somali-related issues, devoting a great deal of attention to inform on the security situation in Somalia. The radio was part of a broader communicative ecosystem which also encompassed social networks, interpersonal communication, and the internet: relevant news aired over the radio were often crosschecked online or through phone calls with relatives or acquaintances living in the areas mentioned in the news. Sometimes, the facts of the day were commented or discussed on Facebook. Some interviewees used the phone to communicate information to relatives living in areas not reached by the radio signal. Radio Star FM extended Eastleigh as a space of encounter and discussion, enabling to speak to parties which were not easily reachable ‘offline’: Somalis with investments in areas under the control of al-Shabaab spoke to militants or youth at risk to be forcibly recruited; women spoke to men. 140 Digital remittances In 2007, Safaricom, a Kenyan telecom company, launched M-Pesa, a mobile money service developed with the financial support of the British Department for International Development (DfID). Previously, Safaricom users purchased air-time and traded it or sent it as a gift by phone, as if it were an alternative currency. M-Pesa enabled users to remit KSh among each other within Kenya. With the passing of time, M-Pesa became increasingly used as a wallet in which subscribers could store money. The system revolves around a mobile platform and a physical network of agents. The application is available to all Safaricom users and must be activated by registering with an agent. The agent is in charge of verifying the ID for the registration and manages the ‘cash-in’ and the ‘cash-out’ by converting the customer’s cash into digital money and reverse. Users pay a fee according to the amount transacted for the deposit and the withdrawals, while the agents earn from the commissions on the operations. It is therefore in the agent’s interest to fulfill the customers’ cash-out requests (this point will turn out particularly relevant when I will be discussing the interplay of M-Pesa and other financial institutions in Eastleigh). Since its launch, M-Pesa has steadily grown in diffusion at home and popularity abroad, becoming a best practice for other mobile money deployments across Africa. Today, M-Pesa has around 12,2 million active users, relies on 81,000 agent outlets and has an average daily transaction volume of over 1,800,000,000 KES (13,692,857 GBP) (Safaricom 2014). These figures have galvanised academics, development practitioners and business analysts interested to assess the transformational impact of mobile money and the possibility to replicate such a profitable experience in other African contexts (Hughes and Lonie 2007; Morawczynski, 2009; Mas and Morawczynski, 2009; Mas and Radcliff, 2010). The popularity of M- Pesa has been assessed against a background in which a gift culture informs social interactions (Duncombe, 2011; Kusimba 2013). A major perk of M-Pesa, according to many women users, is the privacy allowed by mobile accounts. Mobile phones facilitate the carving of a virtual niche where actors subject to a strict social control can cultivate new social relations (a case often mentioned by young Somalis, both females and males, when referring to shukansi, a practice that can roughly be translated with “flirting” and 141 that is possible behind their relatives’ back thanks to mobile phones and the internet), but also independent businesswomen mention the possibility of storing the money earned through their business without their husbands or household monitoring their incomes. The institutional trust towards M-Pesa, though, is not homogeneously distributed: some interviewees, in fact, vent the fear that the Kenyan police might respond to a national security threat by accessing Somali accounts and questioning the provenance of the money. As Omar says “The government might wonder how a refugee can have all this money.” At the same time, M-Pesa enables new ways to deal with unpredictable events, making more flexible the reception of remittances. A telling example is provided by Ahmed, a 23-year-old refugee from Mogadishu who was stopped in December 2012 by the GSU. “It was just before sunset, few days after the attack in which the MP (Yussuf Hassan was injured (7 December). I was returning from Nairobi and on my way to the mosque, and the GSU was waiting in a car, just on Juja Road. They ordered me to come closer and show my documents (the UNHCR mandate and government issued alien card). They took them and told me to get in the car. I was sitting in the middle and was very scared because they were asking why I was not in Dadaab, and saying that I was in Nairobi because I wanted to do something bad. Then we arrived at the Pangani police station, and they said there was just one way to avoid be sent back to the camp, or worse, and it was to pay. But I did not have any money on me and therefore I was told to request money on M-Pesa to my friends and relatives. They wanted KSh 10,000 (GBP 50), and I sent SMS to some friends asking for money. I spent the entire night on the floor waiting for the money but it was too much, and I could raise only KSh 8,000 on my M-Pesa account. In the morning, they allowed me just to go to an M-Pesa kiosk across the street to withdraw but they retained my UNHCR mandate. At the M-Pesa kiosk, I produced only my alien card and withdrew the money. Once I paid, they let me go.” Despite recent years have seen the foray of other mobile network operators in the Kenyan mobile money market, the dominance of M-Pesa remains unchallenged. I will discuss more in detail the changes induced by mobile money into the local financial system in the next chapter. The emergence of an ethnic niche A major factor that turned Eastleigh into a predominantly Somali space in Nairobi was the growing availability of consumption goods charged with a special significance 142 for Somalis, such as camel milk (caano geel) and meat, fabrics for hijab, incense, and jewels. The function of consumer goods in forging or strengthening collective identities has been widely discussed (Douglas and Isherwood, 1978; MacGaffey and Bazenguissa- Ganga, 2000) and, as MacGaffey and Bazenguissa-Ganga (2000) point out, “goods are constituents of self-hood: the practice of identity encompasses a practice of consumption” (50). In Eastleigh, these consumption practices underpinned an ‘affective economy’ (Ahmed, 2004) in which feelings of belonging are mobilised ‘to do thing’ by aligning “individuals with communities—or bodily space with social space—through the very intensity of their attachments” (ibid.). However, this attachment should be understood within a context emerging from a situation of displacement in which social and spatial relations are critical to the reproduction of sociality (Hammar, 2014: 9). The visible infrastructure of Eastleigh, through which goods and information circulate, thus relies on a hidden infrastructure of people, which enables configurations (either regular or ephemeral) of urban actors engaged in shaping an ethnic market by creating economic and symbolic value (Simone, 2004). This is particularly evident in the trade of camel milk, a commodity whose value chain not only integrates Somali-Kenyan pastoralists and women, both refugee and Kenyan (Elliott, 2014), but also brings to the fore the channels through which trust is built and maintained. I was introduced to a group of camel milk sellers by my friend Mohamed, already introduced in chapter 3, but it was his mother, Muhubo Salad, who actually vouched for me. Already the protagonist of her son’s passionate tale of raiders of camels and women, she ran a textile shop in the Mandera Lodge. The textiles arrived from the Gulf and were purchased ahead of family celebrations or holidays. Her customers were mostly Somalis. She stressed that managing a textile shop was not an easy task. It was about choosing good quality fabrics, advising her customers on the right tailor for their dresses and sealing the deals. And no deal could be sealed without a cup of Somali tea, for which camel milk was the defining ingredient. Somali tea was sold by female hawkers. They were mostly refugees and each one had her regular customers among Eastleigh’s shop owners. But Muhubo Salad was not a customer like others (as I will explain later) and, therefore, she had a say on many issues, including her 143 tea seller’s suppliers. There were ten women stationing on a corner of Seventh Street, near the stage of the bus to Isiolo. There were two other selling areas in Eastleigh where clusters of camel milk sellers, all women, gathered, strategically located on Twelfth Street, near the stage of the bus to Garissa, and on Jam street, a few blocks from the neighbourhood’s main mosque. Almost 4,000 litres of milk were delivered daily at 11 am by buses coming from Isiolo, Garissa, and Namanga. The sellers could be easily spotted from their bright yellow jerrycans with their name written on, sitting on stools and chatting. Holding in their gloved hands the unmissable mobile phone from which they listened to Radio Star FM, the sellers were constantly in touch with suppliers and customers to coordinate the delivery of the milk to tea sellers, private customers and at several restaurants. From Eastleigh, the milk was then shipped to all major Somali communities across the country, including in Kakuma refugee camp (the camel milk to Dadaad was shipped directly from Garissa), and as far as Kampala and, sporadically, to Turkey (Musinga et al., 2006). Over the past decade, Eastleigh had become the national hub for camel milk trade. Although the sector employed at least 50 women permanently, it often provided an entry job to newcomers to Eastleigh from the NEP of the refugee camps. In some case, milk was given in advance to would-be tea sellers and then paid back. However economically relevant, the significance of camel milk trade goes beyond the income it generates. Gathering at their selling points, in the hustle and bustle of the neighbourhood, the sellers were engaged in what Elyachar calls ‘phatic labour’, aimed at the (re)production of communicative channels through which reputation is built and information flows, thus creating a “social infrastructure on which other projects oriented around the pursuit of profit could be constructed” (Elyachar, 2010: 453). In her assessment of microfinance initiatives for women in Cairo, she argues that the motives behind the cultivation of these channels go beyond the “empowerment framework” regulating the implementation of the programmes and shifted between a social and a monetary scale of value (ibid.). At the same time, the human infrastructure of Eastleigh consists of nodes in which practices of sociality establish equivalences between financial and non-financial values and, in doing so, engender resources useful to respond to a shifting environment. 144 Conclusions Sayied Ali often repeated that “all Somalis, soon or later, pass through Eastleigh.” Reprising concepts already mentioned in the literature review, Little Mogadishu is a spatial fixture crossed by flows of people, goods, and capitals. It is thus not just a haven for Somali refugees, but also a space of consumption and an interface between the formal and the informal. By charting the transformation of Eastleigh from an Asian- dominated area in Nairobi into a well-known hub in Somali transnational networks and a pole of attraction for traders and shoppers from all across Kenya and beyond, this chapter has discussed the human infrastructure of the neighbourhood. This infrastructure is shaped by pre-existing networks and facilitates the creation of interactions, “not as motionless things, frozen in time, [but as] processes” (Massey, 1991: 244). The communicative ecosystem of the neighbourhood facilitates the circulation of information, a critical resource in volatile times. But, as it will be discussed in the following chapter, it also enables practices through which different capitals are converted into each other, reproducing power relationships and shaping mobility projects. 145 5 Follow the money Introduction The previous chapter has examined how Eastleigh’s Little Mogadishu emerged at the intersection of local and translocal flows. This chapter focuses on the financial infrastructures through which these flows travel and interact. It analyzes processes of capital conversion, emphasising the significance of Somali financial institutions for the transformation of different forms of capital into each other in order to get a glimpse of that “delicate web made up of obligations” (Graeber, 2011: 122) lying underneath the neighbourhood’s reality of rundown roads, concrete buildings, cheap consumption goods and crowded sidewalks. By describing the functioning of two crucial Somali financial institutions: the self-help groups known among Somalis as ayuuto, and the hawala system. It examines how these institutions interact with formal banking services and the mobile money service M-Pesa. An argument is made that the mechanisms regulating ayuuto and hawala preside over the cultivation of social interactions through which risk is spread. People participate in multiple financial socio-technical arrangements in order to engage with multiple locations and temporalities as a strategy to cope with unforeseen circumstances. However, the different ways in which various groups engage with these institutions highlights the fact that Little Mogadishu is “the product of power-filled relations” (Massey, 1994: 21), in which Darood and Hawiye traders dominate the local commerce and are better positioned than Jareer and other minority refugees to diversify strategic resources. This chapter sets off by situating the concept of trust and obligation within Somali socio-cultural structures. It discusses how these concepts underpin the functioning of ayuuto groups and the flows of remittances circulating through the hawala system. After describing the interactions of these forms of organisation with formal providers of 146 financial services, it narrows the focus on Eastleigh’s exchange market, a space in which the intertwining of different circuits and processes of circulation and transformation of value are made visible through soft and hard currencies, gold, mobile money – and hand gestures. Spinning the spindle I was with Mohamed in the shop of his mother in the Mandera Mall. There were always people there, mostly women, often patrons, but also hawkers taking a break or passers-by. Muhubo Salad’s shop was a place of intense traffic: located on the corner of Fifth and First street, near other large shopping malls, it suggested that the payment of a substantial goodwill had been made to secure that location. A debate was pouring from a mobile phone, tuned to Radio Star FM. It was on police harassments and security in Eastleigh. The speakers mentioned the arson of a shop a few months earlier. I asked Mohamed’s mother if she had any kind of insurance. She plainly reminded me that insuring one’s life and properties is haram, against Islam, and rotated her right hand in the air. It was the gesture you would make to move stagnant water in a basin. “If things go bad, people will help,” she said16. She was referring to the neighbouring shops and to a larger circle of clients and relatives. But the movement of her hand meaningfully evoked a spindle, whose spinning merges strings into threads, which eventually interweave and create fabrics. This is how she weaved her safety net, for the present and the future. In her shop, as sips of Somali tea marked the passing of time, she traded in textiles and information. You wondered whether the gossip was the corollary of her business, or the other way around. I frequently stopped by with Mohamed and asked each time a few questions. She was reluctant to give interviews. Her life trajectory 16 According to Khorshid (2001), insurance in Islam is not thoroughly forbidden, but it is rather strictly regulated according to religious precepts. In Islamic studies, a difference is drawn among (1) faith as insurance (al-Ta', n n al Imani), (2) insuring the Hereafter (al-Ta'min al Akharuwi) and (3) worldy insurance (al-Ta'min a! Dunyawi) (ibid., 26). The latter form of insurance applies to mundane issues, but, to abide by religious precepts, the corresponding product must comply with the prohibition of any kind of interest. 147 remained elliptic. The topics of discussion in her shop were always about someone else’s life, the situations in Somalia and in Kenya, her wishes, or the way they were curtailed by her faith. “Going back to Somalia? Inshallah” Her past was confined to the picture her son showed me when we first met. Fragments of her present life floated on the surface. She had left Somalia for Kenya in 2007 with two daughters, but, she said with a hint of pride, she had never spent a single day in a refugee camp. In Eastleigh, she was later joined by Mohamed, before he moved to Egypt to study. She received a regular income from Somalia. In Mogadishu, she still owned a house for which tenants paid 50 USD per month. Sometimes the money was wired through hawala, but ofter it was passed on to some of her relatives in need in Somalia. She was not worried about receiving the payments of the rent since she could rely on relatives who could enforce her entitlement as owner in Mogadishu, where indeed entitlements did not exist for those unable to enforce them. But that, Mohamed specified, was a little money. Besides her shop, Muhubo, along with 15 Somali women, also refugees like her, ran a khat export business to Somalia. The leaves were brought to Eastleigh by lorries from the Meru area early in the morning, bundled by some of the women in a local warehouse and partly sold in the neighbourhood, partly loaded on a chartered aircraft flying to Mogadishu. Mohamed referred to a couple of khat sellers on Fourth Street as ‘aunt’ (eedo). Muhubo declined to provide details about the profits she earned from that business. She preferred to shift the discussion on her other sources of income. She was the treasurer of six self- help groups, known in Somalia mostly as ayuuto, but also as hagbed or shalongo. The term ayuuto, a relic from the colonial past, derives from the Italian word aiuto, which means help. The groups she managed included traders and businesspeople from Somalia and the diaspora. As her son proudly underlined, “She is very good at keeping books”. Sometimes she had to manage up to a million dollar, and she repeatedly proved her reliability and carefulness in keeping track of credits and debts. Her ability, though, went beyond numbers. Standing by the rolls of fabric on display, her phone attached to a cord and dangling like a necklace, she skilfully kept on spinning her spindle, so that the web would not wear out – unless she decided so. “She does waasta”, Mohamed once explained. Perhaps because of his long Egyptian stay, in 148 speaking about his mother Mohamed resorted to the Arabic word for what Somalis usually call dallal, or dalleel – intermediary. Among Somali pastoralists, the dallal brokers transactions involving livestock. The concept of waasta has been extensively investigated in informal economies, particularly in the Middle East. The term stands for both the person acting as intermediary and the connections to which she may grant access (Cunningham and Sarayrah, 1993). Although waasta is often conflated in Western scholarship with nepotism or corruption (Bellow, 2004), the concept refers to a principle that articulates loyalty and access to power (Mann, 2014). In her study of informal networks in Cairo, for instance, Singermann (1995) argues that, by increasing the coordination, these networks create a convergence of interests among members. Drawing from her work, Elyachar (2010) suggests that the waasta is central in performing the ‘phatic labor’ mentioned in the previous chapter by facilitating the opening of “communicative channels” that can potentially transmit not only language but all kinds of semiotic meaning and economic value” (453). Highlighting the key role of interactions finalised to create and maintain infrastructures through which ‘reputation, information and emotion’ (ibid.) flows, also in the form of remittances, she points at the nexus between these communicative channels and financial practices that allow the transfers of affective and material value, such as in the case of the money sent by migrants back home or to relatives in need somewhere else. Within Cairo’s informal economy, the waasta plays a critical role as a conduit to vertical networks, bridging deprived Egyptian women and political elite. In the Somali context at the centre of my analysis, the waasta interweaves both horizontal and vertical networks, creating complex ledgers of social obligations and extracting influence out of this role. The story of Ferdowsa, a young woman I was introduced to by Muhubo, illuminates how a waasta works in practice. Originally from Galkayo, Somalia’s Mudug region, Ferdowsa had spent two years first in Dadaab and then in Kakuma with her mother, four sisters and a little brother. They had no one in the diaspora to rely upon and were living only on the food rations distributed by the World Food Program (WFP). However, in 2011, they decided to leave Kakuma because of the constant risk of sexual violence, and moved to Eastleigh, despite the fact that, at the time, they did not know anyone there. After selling some of her mother’s golden jewels salvaged when fleeing Somalia, they had money to 149 pay for the bus ride and enough to pat the rent for a few months in a basic accommodation. At Eastleigh’s Jam Street mosque, Ferdowsa was told about Muhubo, who came from her same region back in Somalia. They were not directly related, but Muhubo decided to give her a chance and, through Mohamed’s father, who worked as a police officer in Galkayo, she traced Ferdowsa’s background and collected information on her family. She eventually decided to vouch for her with a group of camel milk sellers, who started supplying Ferdowsa with milk. Ferdowsa began selling tea on the street, and soon became Muhubo’s favourite tea seller. Shortly, Muhubo introduced her to her supplier of textiles in Mombasa. She quit selling tea in the street and moved to textiles, which she sold from home to customers met at the mosque or referred to her by Muhubo. At the end of 2012, she partnered with a Somali man living in Kampala to sell her textiles to Uganda. Ferdowsa felt compelled to also involve Muhubo in the business, as a way to thank her for the support enjoyed so far. Ferdowsa thus alternated the sales of her own textiles to Muhubo’s, receiving the orders via mobile phone and sending the rolls by bus. Then, she received the payments through hawala. Recently, she obtained financial help from Muhubo to open a small shop in Eastleigh’s Mandera Lodge. But that, Ferdowsa pointed out, was a loan that she committed to paying back. This short life story highlights Muhubo’s role as a waasta: she does not just connect Ferdowsa to other market actors, but she also vouches for her. She thus grants “entry tickets” into the marketplace (Overa, 2006) to a young trader by entrusting her. This entails the need to call into question some notions that typically accompany analyses of Somali financial institutions, starting with the idea of trust as ingrained into kinship ties. I argue instead that a more problematic approach is required, and one that disassembles a discourse that often couches power relationships into the comforting language of trust and mutuality. Unpacking trust The recurrent emphasis on solidarity among Somalis in Eastleigh, expressed particularly by members of dominant clans, clashes with the obvious acknowledgement of the reality on the ground in Somalia, where indeed the political system has been characterised, for the past twenty years, by political and territorial fragmentation among a myriad of sources of governance, sporadically allied to face contingent situations but often in conflict with each other. To understand how financial flows reflect and are oriented by power trajectories originating in Somalia, it would be useful deconstruct the 150 very notion of trust, especially among kinship associations operating as networked organisations. The scholarly discourse on trust concentrates on the relational nature of the concept (Brann and Foddy, 1987; Dasgupta, 1988; Humphrey and Schmitz, 1998; Mayer et al., 1995) and it is intertwined with the discussion on social capital that I have previously tackled. Indeed, trust is a recurrent concept in explanations of socio- economic relations in Eastleigh and remittance flows among the Somali diaspora. The notion of trust conveys a sense of reciprocity among peers that glosses over differences of status, class and clan. Yet, the popularity of the notion seems due to the ambiguity lying behind its self-evidence. The Somali word for trust is aammin, or aammindo. This term was the most referred to during my fieldwork. However, its many nuances were sometimes rendered by other terms, such as the above-mentioned tolnimo, literally clan solidarity, and midnimo, which may be translated as unity, or solidarity, particularly with reference to Somalia as a political entity. These three terms roughly match Humphrey and Schmitz’s (1998) dissection of the concept of trust on three levels: micro-, meso- and macro- level. The first is between two individuals, the second is between an individual and his group of belonging (such as kinship), the third is between individuals and political institutions. They also define “trustworthiness as membership of a group considered trustworthy’’ (p. 39), an argument that resonates with a large literature exploring the crucial role of trust in solidarity and trading networks regulated by shared norms based on religion (Cohen, 1969) or kinship (Evers and Schrader, 1994; Hart, 1988). Nevertheless, Humphrey and Schmitz argue that in developing transition economies the erosion of the macro- and meso- level of trust is matched by the growing importance of interpersonal relations (Humphrey and Schmitz, 1998: 54). This conclusion echoes Hart’s (1988) reflection on growing informalisation and multiplication of “free-floating social relationships formed by the expectation of mutuality” (178). The rationale of this ‘expectation of mutuality’, and the way it differs intra- and inter-clan relations, is a problematic issue. 151 The ‘myth’ of trust Trust may be indeed a veneer behind which pre-existing power asymmetries and inequalities are perpetuated. (Meagher, 2009). Carrier and Lochery (2013) point out that “Somalis in Eastleigh and elsewhere speak so much of how their businesses rely on ‘trust’ that the concept appears not so much descriptive as proscriptive, perhaps allowing other aspects of these relationships – especially power asymmetries – to be hidden” (339). At the centre of a discursive construction in Eastleigh, there is thus a ‘mythical discourse’ reproduced by regularly claiming a sense of belonging to a common Muslim and Somali community. To understand how this myth is constructed, though, it is useful to recognize trust and power as “functional equivalent” (Hardy et al., 1998). Reviewing the organisational literature on trust, Hardy et al., argue indeed that one of the main interpretations of trust as predictability – or the “probability with which an actor assesses that another actor will act in a certain way” (ibid.) – implies often a relation of interdependence between the actors, in which the weaker party has no other options but to rely on the stronger one, who can, therefore, anticipate the other’s course of action. Power advantage is not neutralised but, instead, “trust is a derivative of power- dependence relations” (ibid.). Hardy et al. acknowledge that also the other dominant view of trust as goodwill – a notion that matches the homonym business practice mentioned in the previous chapter –– fails to explain how reciprocal expectations of mutuality emerge in a first instance, and suggest an approach that, drawing from both interpretations, is based on communication as a sense-making process to which a variety of actors participate in order to create shared meanings. These meanings, and the participatory way in which they are produced, avoid opportunistic behaviours – or, at least, this should be the goal in a situation of mutual trust. The problem, though, occurs when the communicative process is ‘hijacked’: the resulting ‘spurious trust’ (Fox, 1974, in Hardy et al. 1998) conveys “a facade of trust, without any consideration of reciprocity, but in order to consolidate power” (ibid.). 152 Obligations and debts The key to the understanding of how both symmetric and asymmetric relations are produced and maintained is the concept of obligation, which is thrown into sharp relief in Muhubo and Ferdowsa’s case. The notion of obligation is intrinsic to the indissoluble debt/credit dyad and is entangled in the broader moral debate on the ties of an individual with the group of belonging (Robbins and Akin, 1999; Bloch and Parry,1989; Gudeman, 2001; Guyer, 1995; Peebles, 2010). Since Mauss’ (1925) study on the importance of gift in the weaving of social fabric, a rich literature has explored the principle of reciprocity underlying social obligations (Bell, 1991; Kolm, 2008; Sykes, 2005), arguing that, at the centre of the trade in relations of obligation, there are not objects but sociality. The creditor and the debtor are tied into relations of mutual support that can be mobilised when a need arises. If anything happens to the debtor, the creditor will lose her credit; but if the creditor is in need and the debtor fails to show to her rescue, the debtor would lose face, and will see her possibility to mobilize support in future severely compromised (Graeber, 2011). According to Graeber, the difference between debt and an obligation lies in the fact that the former is quantifiable (and, possibly, monetizable), while the latter establishes a bond that is potentially open in time and in the form in which the reciprocity is returned. However, often the reciprocity is indirect and the obligation to pay back may be passed on to other members of the network. The commitment to fulfil expectations of reciprocity is willingly vague and indeterminate, and creates a state of permanent indebtedness, which may be bequeathed to future generations (Shipton, 1995, 2009; see also Gouldner, 1960). Exploring the mismatch between Western and local ideas of credit and mortgage among Luos in Western Kenya, for instance, Shipton (1995) links entrustment, or the act of placing trust onto someone, to obligations, exploring the moral dimension of a commitment which involves not just economic, but also religious transactions, thus transcending the material reality of the exchange. Exchanges varying by terms and conditions, and depending on identity and intimacy, weave the social fabrics among the livings, and between these and the dead, creating an ideal, or an ideology, of mutualism, as “a sense of something held, owned, or just thought or felt in common – or even (in the ‘ism’) just a notion or belief that there 153 ought to be something of the sort” (Shipton, 2014: 511). Also, as Guérin (2014) suggests in her study on microcredit in India, these exchanges are imbued with the permanent tension—which varies greatly according to personal and family histories—between the desire to get out of these hierarchical relationships where debt is the most visible expression and the social and moral values of debt, starting with the trust the creditor grants to his/her debtor. Debt, as debtors are very well aware, is a potential source of social relationships, employment, resources, and respect (44) The literature on trading diasporas has discussed obligations in relation to self- enforcing mechanisms regulating informal commerce on a transnational scale, established in order to reduce transaction costs and mitigate risk in informal markets (MacGaffey and Bazenguissa-Ganga, 2000; Menkhoff and Labig, 1996; Overa, 2006). The embeddedness of economic transactions in close-knit communities of “merchants who monitor one’s another’s behaviour closely” (Granovetter, 1985: 492) enables the build-up of one’s reputation. As discussed in chapter 2, reciprocity in mobilising group solidarity to face hurdles is consistent with the moral framework regulating intra-clan relations. Among Somali traders in Eastleigh, trust dynamics are therefore complex and combine predictability and goodwill, domination and mutual help. As Muhubo and Ferdowsa’s story illustrates, their partnership emerged from the intersection of symmetric and asymmetric relationships, having Muhubo at the centre: on the one hand, she related to her peers, clan members to whom she asked to verify Ferdowsa’s credentials and to help her with goods on consignment; on the other, she bound Ferdowsa in obligation, vouching for her. Moreover, no interest was included in the deal, thus complying with the riba prohibition. However, Muhubo lent the money in KSh but established that the payoff would be made in USD. Although she did not make explicit the rationale for this decision, other interviewees suggested that this calculation is often made when lending money, and reflects the constant need to purchase USD. This strategy resonates with Maurer’s (2005; 2008) reflections on the ‘quality of quantification’ (2008: 72) and, framing the specific currency in which the payback will be done as a sign that creates a vague link between the present and the future, it enables to go around the ban on interest. Muhubo accepted the risk that, in the future, the USD 154 could depreciate against the KSh, and this shielded her financial operation from the risk of incurring in a haram transaction. However, the pursuit of hard currency followed a strategy aimed at securing a resource that, in the long term, would prove resilient against inflation and tradable across national borders. The terms of payment of the loan also oriented Ferdowsa’s choices: for instance, she requested her business partner in Uganda to be paid through hawala, which operates mostly in USD, and not through M-Pesa, which operates only in KSh. Failing to fulfill the obligation to Muhubo might make Ferdowsa unable to access other “lines of credit”. Under this light, a waasta is not simply an intermediary, but a social capital entrepreneur tapping into her pre-existing ‘wealth-in-people’, intended “as accumulation of social relations or [...] as kinship, religious clientelism or political prebendalism” (Guyer 1995). Muhubo’s power, therefore, lies not only in the capacity to initiate and expand networks but, especially, in the (unspoken) threat to break them. She draws from her established social capital to bind in obligation people whose trustworthiness she considers vetting before engaging in business relationships. She masters the art of ‘juggling’ not only temporalities, but also practices that “often reflect deliberate choices, strategies, or tactics for multiplying and diversifying social relationships and strengthening or weakening the burden of dependency ties” (Guérin, 2014: 47). As it emerges from the initial vignette, two crucial mechanisms to receive support in Eastleigh are the ayuuto and the hawala systems. They remind us that the access to material resources is contingent upon not only the existence but also the capacity to mobilize social networks. How this is translated into practice will be clarified by describing what they are and how they work. Money-go-round Somali style The ayuuto is a financial arrangement popular among Somali across the Somali territories and the diaspora. It is a typical rotating credit and saving association (Rosca), usually including 20/30 members, often women, pooling together money in a common fund (Hesse 2010; Horst 2006). The group is typically initiated by someone with a clear 155 business idea or a specific need of money, by involving initially people the initiator is more familiar with, who, in turn, may refer others. Then ayuuto members establish the amount to be pooled every week for an established period of time. The money is borrowed by each member in turn and usually earmarked for setting up small businesses (such as camel milk commerce), paying for school and health fees, weddings or funerals. Loans are thus accessed without incurring in the Muslim prohibition of interest (riba), which permeates financial practices in conservative Islamic Somali societies. However, this mechanism is widespread also among non-Islamic societies, as illustrated by a large literature on informal financial institutions. The repertoire of these institutions, particularly significant where the banking system fails to reach large swathes of the population, is extremely diverse (Ardener and Burman, 1995; Ghatak, 1999; Johnson and Zarazua, 2011; Okurut and Botlhole, 2009). Although the names vary according to the place - tontine, money-go-round, esusu or osusu - these ROSCAs present similar basic characteristics and feature a high degree of flexibility, in terms of amount to collect, the number of the members, the periodicity of the contributions, rotation criteria and possible usages of the loans. Focusing on the ROSCAs’ capacity “to organize traditional relationships in such a way that they are slowly but steadily transformed into non-traditional ones” (Geertz, 1962: 260-1), Geertz described these forms of organisation as vectors of social transformation, finalised to accommodate traditional ties to an increasingly integrated global economy, but bound to give way to formal financial institutions. Yet, the evidence contradicts this prediction. This is not just the case in Somalia, where only recently there have been attempts to rebuild a banking system after over twenty years of statelessness, but also in Kenya, where local informal financial institutions, known as chamas, still play a crucial role in smoothing households’ consumption and coping with unexpected expenses (Zollman, 2014). Rather than organising ‘traditional relationships’, ayuutos are thus finalised at reducing complexity by creating ties based on mutual expectations. New members are vetted prior to their acceptance into the group through exchanges of information that occur in phatic spaces and see the facilitation of a waasta, such as in the case of Muhubo. As Ferdowsa’s story showed, she was gradually vetted and entrusted by Muhubo. The 156 significance of preliminary ‘risk assessment’ in the establishment of a partnership emerges clearly when considering the following case, which refers to a situation in which kinship is not sufficient collateral to access credit. Among the sessions of ayuuto groups I attended while in Eastleigh, I was particularly puzzled by a gathering in an unusual venue, at least for local Somalis. While ayuuto meetings are typically held in private houses, this was taking place in St. Therese church. The reason why the women, all dressed in nijab, met in a Christian church, became clear when it emerged that the 27 female members of the ayuuto were mostly from Jareer and other Somali minorities (a third of them were Migdan), unable to rely on other collaterals but the ones provided by a micro-credit initiative set up by an American NGO hosted by the church. With an initial loan from the organisation, they were thus able to set up small livelihood activities as hawkers of second-hand clothes, honey, tea or porridge. All single mothers, the women did not rely on any support from relatives resettled abroad and struggled to make ends meet by renting small rooms in the surrounding of the church, located on the periphery of Eastleigh, near Juja road. Being unable to meet in the rooms where they lived or other public venues in the neighbourhood, they experienced the ayuuto gatherings as socialising moments in which they could interact with each other. The treasurer changed periodically and was in charge of collecting the established amount, storing the money and keeping track of the transfers. If one could not pay, she would be asked to double her quota in the following session. The contributions were collected in cash during the monthly meeting, a much- awaited moment. In general, showing up was a central aspect of the ayuuto, because, as one of the members said, “You only trust a person by her face.” Indeed, the gathering revolved around the pooling of the money but it did not end with it: they discussed current events, shared life experience, advertised their commercial activities and forged exclusive business partnerships. Moreover, kinship should not be overestimated when considering the creation of an ayuuto circle. As a woman jewel seller suggested, pooling money into an ayuuto often provides an excuse to refuse frequent money requests from greedy relatives. 157 Social capital as collateral Ayuuto groups are crucial to the cultivation of social networks as sites in which information is strategically disclosed and circulated, enabling to weight the degree of reliability of potential business partners (Dowla, 2006). The communicative process occurring in these spaces is thus aimed at monitoring each other’s economic situation, reminding the immanence of social and financial capital. Challenging the mainstream developmental discourse on social capital and discussing power inequalities in microfinance groups, Geleta (2014) applies Bourdieu's interpretation of social capital to argue that the acceptance of a new member into a group is contingent upon that member’s access to specific assets acting as collateral. This implies that belonging to the same clan is often a necessary but not sufficient requirement to establish a business relationship: shared kinship ties may facilitate the collection of information for running background check but, in the absence of collateral, it may prove difficult for someone to enter into a symmetric agreement. As in Muhubo and Ferdowsa case, though, the stronger party may decide to take a higher risk to endorse someone without a regular source of income but to burnish her credential as a charitable person, or because of personal feelings such as empathy or fondness. Receiving financial help from relatives in the form of international remittances is often mentioned as a crucial requirement to get accepted into an ayuuto. Besides providing the initial capital for setting up a livelihood activity, remittances are also a guarantee against the risk of payment defaults. Somali refugees draw indeed a line between those who have access to remittances and those who do not. This is obviously a broad brush distinction: amount and regularity of the remittances may determine what kind of self-help group one is eligible to. But, as in the case of the single mothers group gathering in Eastleigh’s St. Therese church, discourses of clan solidarity are ineffective when they are not grounded in avenues of access to economic resources. This is why looking at the hawala system, through which remittances from relatives resettled overseas, but also capitals from Somalia, pour into Eastleigh, is critical to understand how both solidarity relationships and inequalities are reproduced. 158 Remitting the past and the future Transnational financial flows shape Eastleigh’s space and time. Many malls, including Amal shopping centre, the neighbourhood’s largest shopping complex, bear the name of the MTO behind its construction. As said in the previous chapter, inflows of capital have played a crucial role in the expansion of Little Mogadishu, and the variety of branches of Somali MTOs are still today a major conduit of financial assets. At the same time, the arrival of the remittances, especially at the beginning of the month and during holidays, dictates the timeframes of the ayuutos and defines the local business turnover. Remittances are largely recognised as a lifeline for Somalis both in the Somali territories and in humanitarian spaces in East Africa (Hammond, 2010, 2011; Horst, 2006, 2007, 2008; Lindley, 2007, 2009, 2010). According to a conservative estimate, a total amount of USD 1,2 billion per year is sent each year to the Somali territories (Somalia, Puntland and Somaliland) (FSNAU, 2013). This is higher than the amount of aid, which account averagely to USD 834 million per year between 2007 and 2011, Foreign Direct Investments (FDIs), which account to USD 102 million in 2011, and exports, USD 516 million in 2010 (ibid.). Yet, there are no figures on the amount of financial flows circulating among Somali communities in the diaspora, particularly because most of the money move through hawala circuits, which mostly escape the oversight of international financial regulators. In fact, my previous analysis of the concepts of trust serves as an introduction to the discussion of the hawala system: trust relationships are the affective material upon which Somali MTOs thrive and are made of. A starting point to understand how plurality lies at the core of the hawala system can be found already in the word used by Somalis to refer to MTOs: shirkaad (pl.shirkooyn). The term is usually translated as ‘firm’ but, without further specification, it is commonly used for MTOs. The prefix shir means literally conference or convention, and suggests the act of coming together and participating in an initiative that is, at the same time, political, economic and social. In chapter 2, I have described the development of the hawala system in Somalia 159 during the 1970s in a context of increased transnational mobility, as a growing number of Somalis migrated to the Gulf States to work in the local oil industry, and institutional decay, particularly during the 1980s, when state-managed financial institutions proved increasingly inefficient. Grabowski (1997) notices that “the coordination role of the state may be partially subsumed within the activities of such ethnic groups. That is, ethnic ties can serve as a conduit through which credit, technical advice and other critical elements in the development process can be allocated” (399). In Somalia, the emergence of clan-based trading networks matched the withering of the state, taking over its function of coordination. The term hawala refers to a repertoire of practices and arrangements popular across the Muslim world, and developed along the coasts of the Indian ocean, at the centre of a heterogeneous scholarship spanning from forced and voluntary migration studies (Ballard, 2003, 2004; Lindley 2009, 2007; Sander and Maimbo, 2003; Monsutti 2004) to self-enforcement mechanisms in trade (Ballard, 2005; Schaeffer, 2008) to terrorism financing (Passas, 2006; Raphaeli, 2003). In the literature on informal finance, hawala is mostly examined through either a security or a development lens (an approach that, meaningfully, replicates the terms of the debate on diaspora, as previously discussed). Informal MTOs are thus regarded either as a conduit of money laundering or as a vital channel for delivering funds to areas with no access to international banking institutions. This discussion has framed the linkages between formal and informal finance mostly in terms of competition or complementarity, emphasising either the disruptive effects of informal MTOs on development, or their role as necessary interlocutor of formal financial institutions and aid agencies delivering services to areas excluded from transnational financial circuits (Meagher, 2013). Passas defines hawala as an informal value transfer financial system (IVTF), a category encompassing “any system or network of people facilitating, on a full-time or part-time basis, the transfer of value domestically or internationally outside the conventional, regulated financial institutional systems” (Passas, 1999). The Somali hawala figures alongside other institutions such as the Indian hundi and the Chinese fei chien. From Afghanistan (Monsutti, 2004; Thompson, 2008, 2011), to DRC (MacGaffey and Bazenguissa-Ganga, 2000), the conjunction of fragile state regulators, 160 lack of access to formal banking institutions, displacement and migrations have created fertile ground for financial systems based on interpersonal relations and reputation to flourish. Furthermore, IVTF networks play a central role for humanitarian and development agencies, enabling to funnel funds to areas of conflict cut off from international banking circuits (such as Somalia and Afghanistan) (Hammond, 2013; Thompson 2011). The Somali hawala is based on the trading of credits and debts among agents, or hawaladar. A hawala transaction involves an hawaladar A (HA) and an hawaladar B (HB): HA collects the amount, inclusive of a fee, from the sender, records information about the recipient, such as his location, name, telephone number, ID or, in alternative, place of birth, and transmits the data via phone or enters them in an online system; and issues a receipt to the sender. HB receives the information, contacts the recipient and, after verifying his identity, releases the credit, while HB is left with an outstanding debt.. If unable to produce a valid ID, the recipient can claim the money providing the information shared by the senders and mostly related to their relationship. When the same branch or a cluster of branches cumulate several debts, the overall debt is settled by misinvoicing goods moved through clearing houses (Cockayne and Shetret, 2012; Schaeffer, 2008). Hawala shirkaad are indeed providers of multiple financial services catering to the needs of the different populations that conflate in the broader Somali diaspora. Vlcek (2010) suggests that, as opposed to the hierarchical and rigid structure of Western banking institutions, the hawala is legible through Deleuze and Guattari’s epistemological model of rhizome to the extent that it is “malleable because [...] non- hierarchical, regenerating from its rootstock in multiple directions (along multiple lines of flight)” (430). The organic nature of IVTF networks stems out of the interactions of interpersonal ties and technological devices (from letters of credit in the past to emails, SMS and phone calls today) through which information on specific financial transactions and the reputation of each agent travels and reaches each node. Schaeffer (2008) emphasises the importance of reputation and the threat of sanctions such as ostracism, drawing an important distinction between ex-ante and ex-post reputation mechanisms to establish trustworthiness. The ex-ante mechanisms may include the 161 presence of a business that communicates credible reputation to customers, and indeed the hawaladars are mostly engaged in trade on the side of their money transfer activity; ex-post reputation mechanisms refer to the sanctions in which an agent may incur in case he fails to comply, particularly a blow his reputation and a social ban on future commercial partnerships (ibid.). Therefore, the entire system of hawala rests on a collective condition of mutual indebtedness as the one described above. As Schaeffer points out, “the staggering of clearing debts among hawaladars lowers the probability of defection of any one member within the group” (14). This is an important aspect, which puts into perspective the role of clan membership and reprises my previous argument: common kinship ties may be a necessary but not sufficient condition to build trust. This is consistent with what Monsutti (2004) argues about MTOs in Central and South Asia, as he identifies four types of actors within the system: close relatives; distant relatives and friends; people from the same locality; and eventually, and most importantly, people located outside the ethnic community. While dealing with clansmen can mitigate uncertainty and reduce the risk of cheating, reaching out to other networks enables to expand business and inject fresh capitals into the system. When kinship is not enough At a basic level, though, shared clan membership is a crucial factor in both recruiting agents and orienting individual and household decisions regarding which hawala company to resort to for sending remittances (Medani, 2011). Somalis in Eastleigh are generally able to associate a particular hawala company with a clan. Just to mention the largest MTOs, Amal, based in Bosasso, Puntland, is known to be owned by a cluster of Darood Harti (mostly Dulbahante) businessmen, while Tawakal is linked to a group of Hawiye Abgal businessmen from Mogadishu. The largest and most popular hawala company, Dahabshiil, deserves a separate discussion. Established in 1970 by Mohamed Saeed Duale, an Isaaq from Burao, Somaliland, it is managed by his son Abdirashid Duale as for June 2015. In 2002, after the demise of the hawala company Barakaat, it became the largest Somali MTO and in 2014 employed 24,000 agents in 144 162 countries all over the world. Dahabshiil is a privileged partner of the Somaliland state (for which it acts as official money vendor, trading Somaliland Shillings for USD) and of a diverse constellation of humanitarian and development actors operating in the Somali territories. To negotiate its access to insecure and potentially hostile areas across Somalia, Dahabshiil employs respected members of local ruling groups. According to Dahabshiil operatives and executives I met in Kenya, Uganda and Somaliland, the company adopts a pragmatic policy in critical areas, particularly in Southern Somalia, recruiting most branch directors among the dominant Hawiye and Darood clans. By reckoning local dynamics of power, Dahabshiil is granted permission to operate in areas under the control of certain groups, whose position as gatekeepers is thus reinforced. However, the attitude towards Dahabshiil is influenced by two interlocked considerations: 1) the shirkaad is associated with the Isaaq clan that, established in Somaliland, does not have direct political interests at stake in South and Central Somalia;17 2) Dahadshiil enjoys an unrivalled prestige among Somalis, compared to other MTOs, for being a privileged interlocutor of UN agencies and INGOs as a conduit of aid money. Since it is widely perceived as above inter-clan feuds in the volatile Somali context, Dahabshiil has secured a role as a legitimate interface between the Somali hawala system and international humanitarian and development actors and has thus established a widespread presence in Somalia. The extent of Dahabshiil’s network, straddling territories and types of actors, is a key reason for its popularity, and it proves vital for moving and tapping into financial capitals in critical moments. For instance, the case of Abdirizak, a 38-year-old refugee from the Rahanweyin clan who made a living as keeper of a mosque in Eastleigh, illustrates how hawala networks are intertwined and shape Somali mobility routes, across the region and beyond. Abdirizak arrived in Eastleigh in 2012 after fleeing Lower Shabelle, a region south of Mogadishu under the control of Al-Shabaab, in 2011. Since he had savings for about USD 900, he decided that the safest way to take the money with him was to deposit the sum with a hawala agent. He opted for 17 Some interviewee, particularly from minority groups, suggested a link between Dahabshiil and Al Shabab since the late leader of the Islamist group, Ahmed Abdi Godane, was, such as Dahabshiil’s executives, an Isaaq from Hargeisa. 163 Dahabshiil because the local agent was a former pupil of his. Moreover, he knew that it was the company with the most capillary presence in both Somalia and Kenya. Abdirizak was able to withdraw little amounts (between USD 50 and 80) to pay for each leg of the journey by simply disclosing his details at each agent along the route to Eastleigh. As evinced from this case, the hawala system serves both to transfer and to store value. The two sides are interlocked and this explains why, in Eastleigh, hawala companies are used as an alternative to banks by Somali refugees. Although most formal banking institutions in Kenya accept alien cards as IDs (but not UNHCR mandates), most refugees still rely only on accounts with hawala companies because it is there that they receive remittances. In the past years, Dahabshiil has started issuing cards with photos, fingerprints and unique ID numbers to speed up and improve the identification of the clients. Aisha, a young Rahanweyn woman, expressed her relief for the possibility to produce a document that spared her from disclosing information on her background, thus granting her a comfortable (and surprising, in the light of Western privacy concerns) anonymity in her biometric data. In general, hawala is a cheap (compared to international money transfer companies such as Western Union or MoneyGram), fast, reliable and accessible (even in refugee camps and conflict areas) system to send or receive money: according to the estimates of the International Monetary Fund (IMF), hawala transactions cost between 2 and 5 percent (El Qorchi, 2003: 16), while equivalent transactions in the formal sector range between 10 and 20 percent (Ballard, 2005). Since M-Pesa made its foray, hawala is seldom used for sending money inside Kenya. Many Kenyan Somalis have started transferring money through hawala since they have partnered with relatives from Somalia who are familiar with the system. These Kenyan Somalis are mainly clients of formal banks and regularly withdraw cash from their bank account and deposit it in their hawala accounts. As I mentioned in the prologue, this was the case of Ali, a Somali Kenyan wholesaler of clothes and accessories. While he used his Kenyan Commercial Bank (KCB) account for transactions within Kenya, to withdraw petty cash and pay bills, he relied on an account opened at Tawakal for sending payments to Yiwu, a city in central Zhejiang Province, China, dubbed also ‘Chocolate city’ for being a popular destination among African 164 traders. Ali was one of them, and considered hawala as the fastest and more efficient system for sending money. The local Tawakal agents also acted as brokers and, according to Ali, were able to purchase a large amount of Chinese Yuan Renminbi to pay the suppliers because, despite the fact that the hawala system is based on USD, local Chinese traders “don’t like USD”. Tawakal was used for money transfer to China, where IVTF were illegal, because it bore the risk to operate clandestinely there, while Dahabshiil, concerned to preserve an untarnished reputation, did not. Converting capital, producing value There is thus a qualitative difference between the value ‘produced’ either through the engagement with the formal banking system or through the hawala circuits. Featuring low margins of profit and high flexibility, and screened from the scrutiny of state regulators, a particularly important factor for Somalis in Kenya, the hawala system is much about financial value as about information, circulating through linkages that cut across institutions and geographical spaces. Information that, as Maurer (2005) points out, “is money, and power, and everything else that goes with them. At the same time, recognising that money is only information, mere words or sign, allows us to seize it and make it in our own image, to do good and re-create community and trust” (166, italic in the original). The elusiveness and opacity of hawala companies puzzle international regulators. The limited usage of paper records and the confidentiality surrounding the customers’ transactions are a matter of concern for regulatory bodies and national security agencies, raising suspicions of money laundering and terrorism financing. Wheatley (2005), for instance, argues that "hawala facilitates money laundering in three stages: by discreetly introducing criminals' funds into the financial system ("placement"); manipulating them to appear legitimate ("layering"); and making them available for further use ("integration")" (357). Medani underlines that “the criminalisation of informal financial transfers has had the unintended consequence of potentially undermining state building efforts in the very region of the world which is in most need of building and 165 strengthening formal institutions” (Medani 2011). This was particularly evident in the case of Al Barakaat, as explained in chapter 3, suspected of money laundering and links with militant Islamist organisations. Indeed, links between Somali money transfer operators and political Islam exist, particularly in the way precepts of Islamic solidarity are embraced by financial practitioners to legitimize their work and thus bridge among different actors (ibid.). Al Barakaat’s bankruptcy brought major changes to the Somali money transfer sectors, prompting some companies to register with Western regulators and embrace the set of rules and practices to comply with international money transfer standards. However, examining the organisation and functioning of Somali MTOs through the lens of the formal/informal dichotomy may be misleading (Pieke et al., 2007). The framing of Somali hawala companies as ‘informal’ is based on normative assumptions modelled on Western regulatory frameworks. This discrepancy is particularly evident in the Somali MTOs’ implementation of ‘know-your-customer’ (KYC) policies which, although considered very effective by the same UN agencies and INGOs partnering with hawala companies, do not meet international anti-money laundering requirements. Once again, an exemplary case is offered by Dahabshiil. In May 2013, Barclays Bank announced the closure of the accounts of Dahabshiil (along with other 250 remittance companies) in the UK because of insufficient guarantees to ensure anti-money laundering compliance. The decision elicited the reaction of aid workers, activists and scholars involved in Somali affairs who called for the need to keep open a ‘vital lifeline’ for Somalia (BBC, 2013; Tran, 2013). The resulting discussion illuminated the polarisation between two contrasting narratives: one emphasising informal financial institutions as a breeding ground for illicit activities ranging from piracy to terrorism; and another that stresses the crucial role of the hawala sector for development actors operating in Somalia. In September 2013, I attended a conference in the Kenyatta conference hall, in Nairobi, in which a group of high-profile representatives of UN agencies and INGOs gathered to discuss the implications of Barclay’s decision. Sitting on the side of Abdirashid Duale, the members of the panel were indeed advocating the importance of Dahabshiil, rather than of the broader constellation of hawala companies. 166 Offering their interpretations of the dispute with Barclays, the panelists stressed that, unlike other MTOs from other countries (such as Pakistan and Colombia) whose accounts would have been closed down by Barclays, the shutting down of Dahabshiil would have had dire consequences for Somalia since no Western remittance companies operate there. To counter the argument that hawala companies are conduits for money laundering and terrorism financing, it was also reminded that transfers to Somalia usually range between 100 and 500 USD per month and are used for basic expenses. However, during the discussion, I was set aback by the comment of a representative of a well-known INGO, who claimed that the problem derived from a misperception of the issue at policymaking level as “a commercial issue, not as a humanitarian and developmental one.”18 But is it really possible to draw such a clean-cut line between commerce and business and a humanitarian and developmental approach? The case of Dahabshiil in particular seems to support the argument that the resilience of the hawala system is located in its capacity to blur boundaries between countries and categories, to integrate new actors and accommodate them to what Grabhner and Stark (1995) refer to as a logic of “organisation of the differences”. Acting as an interface, in Guyer’s terms previously described, between the formal and the informal, Dahabshiil does not conflate but rather accommodates to each other the two domains. The company prides itself of complying with international financial standards, by implementing KYC and anti-money laundering policies and, in Western countries, operating through accounts in formal banking institutions, such as Barclays Bank, as a guarantee of transparency (the irony of which is that, in recent years, these formal financial institutions have proved far from being transparent and law-abiding, as the 2014 money-laundering and tax diversion investigations involving the banking giant HSBC have glaringly shown). Upholding to international financial standards is a strategy that pays off, as it renders Dahabshiil an acceptable contractor of UN agencies and INGOs for the implementation of cash transfer programmes to the Somali territories. In return, Dahabshiil is able to leverage its 18 On 16th April 2014, Dahabshiil and Barclays settled the dispute. According to a press communiqué of Dahabshiil, “as part of the agreement, there will be a transition period to allow Dahabshiil to end its banking relationship with Barclays and move to alternative arrangements"(www.dahabshiil.co.uk) 167 unique position to mobilize high-profile support in defending its role as a lifeline to Somalia. Moreover, this partnership with international humanitarian actors is a marketing asset for Dahabshiil, creating among its Somali customers a perception of the company as more solid than other MTOs because of the institutional network in which it is embedded. Aware of the need to preserve a respectable facade, Dahabshiil treads carefully on informal terrains: although it does not officially operate in countries that ban MTOs altogether, such as China, as previously mentioned, it services spaces, such as refugee camps in Kenya, where MTOs are not regulated, as I explain in chapter 7. Operating in the Somali context, Dahabshiil adopts a pragmatic approach in brokering the linkages between the formal and the informal. This approach foregrounds the importance of the human infrastructure that makes possible the intersection of different circuits. Intersecting platforms Since the 1990s, when the first MTOs began operating in the area, the local financial landscape has become much more variegated. In the last decade, following the economic boom of the district and the inflow of Somalis with Western passports, 12 major Kenyan and international banks have opened up branches in Eastleigh: spanning from Kenya Commercial Bank (KCB) to Barclay and Gulf Arabic Bank, they offer Shariah-compliant financial products and vie for a growing market, which encompasses Somalis and Kenyan Muslims in general. Although hawala is still the institution of choice for refugees who are in need to monitor assets in Somalia or regularly resort to remittances, a growing number of young Somalis refugees are signing up for a bank account. Many admit that they do not plan to deposit any money in the near future but, since opening an account is free, and lest a deterioration of their living conditions in Kenya that would make it more difficult, they have decided to seize the moment. The rationale behind the choices lies in the need to allow multiple options, particularly in relation to financial behaviours that are so critical in determining the position of someone vis-a-vis his community. What if one wants to distance himself from the 168 ‘community’ – or want to resort to it on his own convenience? Abdillahi, a young Darood Dulbahante refugee studying in Kampala, clearly explained the anxiety deriving from belonging to a too tight ‘community’: “If you have some savings, an account in a hawala will do the job. It is cheap and with a company such as Dahabshiil you won’t have problems in finding a branch. Then it is also easy to reach because hawala companies are always in the middle of Somali neighbourhood. But this can turn into a problem if you make a lot of money. Because people will see you constantly going to the office, and the voice will spread around and after a while you won’t have a life anymore, because everybody will start asking for help. In that case you also need a bank account, so when you go to a downtown bank nobody will know.” The usage of a specific financial institution may also facilitate or hinder career trajectories. Particularly telling is the case of Senegal, a 30 years old Jareer from Lower Juba who learned proper English after living four years in South Africa. Barred from most livelihoods in Eastleigh because unable to access capital to launch his own business, he resorted to his language skills and his familiarity with the NGO jargon acquired during his stay in Dadaab. He set up an account on Proz.com, a translation website, and started a career as a freelance translator from English to Somali and Somali to English working mostly with not for profit organisations. He then opened an account on Paypal, an online payment system through which he received the payments on his bank account. The diffusion of Safaricom M-Pesa has had an impact on the way Somalis make sense and usage of hawala. As mentioned in the previous chapter, mobile money has reshaped the relation between Eastleigh and the refugee camps. To grasp the importance of this route for many Somali refugees, it is worth reminding that moving from the refugee complex of Dadaab or Kakuma to Nairobi is the outcome a collective decision in which both potential costs and expected benefits for a larger group are taken into account. Often, the choice to move to Eastleigh falls on the better-educated member of the group, who is likelier to find a job and support others. Humanitarian facilities provide shelter and free food staples and other basic items but are overcrowded and 169 highly insecure. The diffusion of mobile telephony has both opened a constant channel of communication between Eastleigh and the camps, bypassing taar, and changed modes of remittances, which are transferred through M-Pesa instead of hawala. Remittances are often sent to the relatives in the refugee camps to set up businesses like small stores selling items not provided by the UNHCR and delivered from Eastleigh on the daily buses to Dadaab and Kakuma. Mobile phones are used to place orders and transfer payments between the city and the camp. Yet, as it will become clearer in the chapters on Kakuma refugee camps, money flows in both ways: encamped refugees running little businesses can send money to their relatives in the city to pay for school or medical fees. Taking financial choices The choice between M-Pesa or hawala was taken against a set of factors in specific circumstances. Preferences towards either hawala or M-Pesa reflect attitudes towards currencies (USD/KSh): while hawala companies mainly deal with USD, the M- Pesa platform enables to transfer and store only KSh. This emerged clearly during the pre-electoral period, when, among fears of forced repatriation and widespread profiling of the Somali population, there was in Eastleigh a rush to cash out from mobile wallets and convert KSh into USD, as I will explain later. A widespread concern was in fact that, in the case of forced repatriation, Somali M-Pesa account holders would not be able to access their savings from Somalia or elsewhere. While Safaricom requirement to provide an official ID (either a Kenyan ID, a UNHCR mandate or a Kenyan alien card) can often inhibit recently arrived refugees or individuals without documents from using M-Pesa, in the case of hawala identity vetting is a more flexible process, done through both ID or a guarantor. Hawala is still favoured by elders, who often encounter technical and linguistic barriers in using M-Pesa (whose instructions are in Swahili), or refugees recently arrived from Somalia, where money transfer and currency exchange shops often coincide, being the country a de facto dollarised economy. Also businesspeople claim to use mainly hawala companies because they have no amount limits, while M-Pesa allows maximum daily transactions of 140,000 KSh 170 (around 1000 GBP). Officially, M-Pesa operates only through the Safaricom network in Kenya. However, in the next chapter I will explain how groups of young Somali entrepreneurs have achieved to turn M-Pesa into a service of cross-border money transfer. Furthermore, M-Pesa is mostly used to transfer money, rather than to store it. For traders relying on long-distance connections, such as camel milk sellers shipping their jerry cans as far as Uganda, M-Pesa has successfully replaced previous methods of payment such as handing the money to truck drivers (for small businesses, this system was much cheaper than hawala), but the general trend is to withdraw as soon as a payment reach one’s account. This behaviour suggests limited institutional trust towards M-Pesa, trust that ebbs and flows in coincidence with Kenyan political turbulences and periodical targeting of the Somali population. However, it also hints at the need to cash out in order to have paper money that can be deposited into hawala accounts, used to purchase hard currency or pooled into ayuutos. Informal interoperability Cash remains the means of payment that most facilitate capital conversions because of its ‘universal interoperability’. In the mobile money industry, the term interoperability refers to the possibility, based on a commercial and technological agreement among telecoms, for two or multiple platforms to exchange data related to their customers and, therefore, to enable monetary transactions among subscribers to different companies. Cash is the ‘missing link’ between financial providers: one has to cash out from a network and then cash in into another. The materiality of money plays a crucial role in mediating among heterogeneous circuits, associated with different notions of time and space. Most Somalis in Eastleigh acknowledged the complementarity, rather than the opposition, of M-Pesa and hawala, where the former was mainly used for small transactions in Kenya, using KSh, while the latter was still the favourite option for large international transactions in USD. A particularly telling example was the case of Abdicadir Cali, a 26-year-old Rahanweyin from Baidoa, in Somali Bay region, who reached Eastleigh in 2009. 171 After fleeing his home because of the fear to be forcibly recruited by Al Shabaab after they seized the region, he crossed the border into Kenya and arrived in Mandera. He spent few months in a cousin’s house and then, driven by the desire to get a formal education, moved to Eastleigh. Once again, some relatives offered him hospitality and a little financial help with which he enrolled in a technical institute. After a year, though, his funding dried up, and he was forced to drop out from school. Yet, he had discovered that he was particularly talented in resurrecting laptops and computers, and his friend had it too. “Let’s ask Abdi” became the immediate reaction to a crashed operating system, and his name – and mobile number – started travelling via word-to-mouth or SMS. It thus spread outside Eastleigh and eventually Kenya, spurring a flood a job offers from as far as Mombasa and Kampala. His customers included mostly Somalis and few Kenyans and Arabs from the coast, all wanting him to fix their apparently dead devices, packed and shipped by bus to Eastleigh. Abdiqadir Cali became busy dismantling and reassembling laptops and PCs, constantly informing his customers about this or that spare part to purchase, payment upfront. The money was sent through M-Pesa and cashed out, so that Abdicadir could convert it into USD and deposit a part in a hawala account, and thus easily authorise transfers to his family in Baidoa. “I keep tab of my transfers and my mother takes care of my money”, he said. “One day, I will use it to start a business.” In the while, though, his savings were shoring up his family’s livelihood and fostering his siblings’ mobility aspirations: one of them was set to reach Abdiqadir Cali in Nairobi, tapping into his older brother’s little wealth to fund his journey. The functional specialisation of hawala and M-Pesa acquires a meaning as we relate these financial institutions to the currencies that circulate through their networks. The complementarity of soft and hard currencies reflects different configurations of space and time in relation to social and individual action: while the local currency (KSh) dominates short distances and term horizons, the hard currency (USD) evokes a transnational dimension and an extended temporality. The hard currency preserves its value when crossing borders, defying inflation and thus mitigating the volatility of the economic and political scenario. M-Pesa had largely replaced hawala companies operating mostly on a national scale, with some significant exception, as it emerges from the following case study. Rashid was a 30-year-old Darood Ogaden from Mandera, in the NEP, and had been working in the hawala business for the past two years. He worked in a family run company called Warsan Express located on 7th street. He landed the job after being referred by a relative. He claimed that “Nobody would employ you if they are not sure who you are and whether you can handle money in a trustworthy way”. Warsan – a female name for a company whose most customers are women – worked only in the NEP and operated only in KSh. So far, Warsan had survived M- 172 Pesa because it anticipated money to customers. More specifically, they authorized the payment to the recipient and then charged the sender. So, on the transfer of 1000 KSh, people usually paid 10 KSh fee. Thus Warsan acted as a lender, morally eschewing the riba prohibition by separating the beneficiary of the loan from the direct debtor. Most of their customers were relatives who needed to send substantial amounts that would not be possible to send via M-Pesa. However, they offered their customers the possibility to withdraw via M-Pesa from their accounts. In that case, the agent would receive a call from a customer who is unable to travel to the office and be asked to send money through M-Pesa. The recipient would get the money on his mobile wallet and cash out at an M-Pesa kiosk. Access to cash float is a common issue for M-Pesa agents in busy commercial areas across Kenya. Agents in need are able to withdraw cash from a Safaricom account at the nearest branch of the Central Bank of Kenya (CBK) but this would entail a lapse between the request and the response that result in losing a customer and therefore a profit (since, as previously explained, agents earn a fee for each transaction). This seldom happened among Somali M-Pesa agents in Eastleigh. A hawaladar might establish a mutually beneficial partnership with an M-Pesa agent. When an agent needed cash float, he usually borrows the amount from a trusted hawaladar. As Mohamed, a Somali Kenyan M-Pesa agent, pointed out, “Maybe you come to me and need KSh 50.000 but I don’t have it. So instead of going to the bank (CBK) and get the money, I go to my relative and borrow the money, which I return at the end of the day. I don’t want to lose the customer.” This accommodation of hawala and M-Pesa is at odds with assumptions that the “network power” (Grewal 2008; Donovan 2012) engendered by the diffusion of M-Pesa in Kenya would lead to an obliteration of alternatives. Yet, in Eastleigh, alternative standards are preserved and creatively combined. The exchange market Mobile money has reshaped the local repertoire of means of transfer and storages of value and their configuration. A vivid glimpse of this variegated and complex environment can be grasped at Eastleigh’s “exchange market”, called also Suk Dahab, or Gold Market, near Jam Street, where is located the neighbourhood’s main mosque. The 173 market is reachable through a gate with security guards, beyond which there is a large room where women wearing niqab sit along the walls in front of display cases on low tables, upon which stand small electronic scales, calculators and mobile phones. The cases contain golden jewels such as earrings, necklaces and rings. The customers are mostly Somali and Kenyan, but also Ugandan and Sudanese travel there to purchase gold ahead of a wedding, or as an investment. The rooms are filled with the constant buzzing of the ongoing trading. Then, the market opens on a courtyard along whose sides and in the middle are shops where groups of men sit behind counters, chatting or counting stacks of cash. In the corners stand an M-Pesa shop and a small stall where a man smelts gold and silver with a blowtorch. A stairway leads to the upper floor where there are three tailor shops, two travel agencies, a freight company and a small mosque. The market teems with people from 8 am to 6 pm, businessmen about to leave or just returned from a journey and women purchasing golden jewels as dowry or converting USD just received through hawala from relatives overseas into KSh. In the exchange market, a Somali businessman can purchase his plane ticket to Honk Kong while waiting for its tailored suit to be ready. Gold traders constantly check with the money vendors the current gold exchange rate (jewels are sold by weight), while the money vendors buy foreign currency (mainly USD) and sell it to hawala companies. There are young men standing near the entrance and offering to sell USD to newcomers, and other sitting behind counters of little kiosks. On my first visit to the exchange market, I did not immediately notice what filled the back shelves on these kiosks. However, my eye subliminally recorded information that, later, I found quite puzzling. Rows of bottles of water were on display on the shelves, from top to bottom. The money vendors initially replied with giggles to my questions, saying, tongue-in-cheek, that, if I was thirsty, they could sell me a bottle. Only after a while it dawned on me that the presence of bottles of water justified the licence as water sale points, rather that as a money exchange, for which another, more difficult to obtain, permit was required. It was a strategy to dissimulate, or, better, negotiate visibility according to a pattern already observed elsewhere, where most businesses that need identifiable locations to operate on a regular 174 basis tend to present a front that does not fully reveal backstage operations (Muñoz, 2010: 156). At the end of the working week, the owners of the shops drop by to review the ledgers. Money vendors take a percentage of the day profits, but trading hard currency is not a livelihood as another. The way the young males who enter the business talk and display their trading skill is steeped in ideas of masculinity. During the afternoons I spent with them, they often proudly exhibited the strongly gendered adroitness in handling and counting the cash that embodies, more broadly, the importance attributed to trade in the Somali (and Islamic) culture. Once, Asad, one of them, gave me a demonstration of the Somali money-counting style: he pinched a folded wad of notes between the second and the third finger of the right hand and flicked through them using only his toe. He proudly pointed out that that counting method starkly contrasted with what he called the clumsy ‘Kenyan style’, in which both hands are required to count the money. It was a wholly sensorial exercise in which finger counting evoked the ontological and intimate connection between being Somali and commerce, but also ‘religion and cosmology’ (Maurer, 2010). Maurer suggests that dactylonomy, or the practice of counting with fingers, is associated with one of the expressions of the Islamic concept of bayan, or ‘manifestation of meaning and its elucidation’ (Bernand, 1985, in Maurer, 2010: 4). Sometimes, I even noticed seasoned money vendors counting money with their fingers while lingering with their eyes on the flows of people passing by, in the same way they would count the beads of a misbaha, an Islamic prayer rope, during the utterance of a tasbih. It was meaningful, in my view, that the transactions of the exchange market occurred under the eyes of Sayed Ali, the Italian-speaking imam that I introduced in chapter 3, whose tailor workshop was on the first floor of the exchange market, beside the small mosque of the market. Islamic morality permeated the way money vendors reflected on their activity. Drawing a sharp line between cooperation and competition among these money vendors groups was problematic: often they lent money to each other, even when they had the possibility to drive other clusters of vendors out of the market. There was a subtle balance between cooperation and competition akin to what observed in other contexts in which fortune reversals always loom on the horizon 175 and safety nets are created by binding others in moral obligations. Interactions among money vendors in the exchange markets recall what Meagher (2010) observes among producers in Nigeria who are concerned with avoiding ‘‘getting a bad name’ in their home communities for refusing to help a ‘brother’ (or sister)” (132). Some dynamics among money vendors in the exchange market resonate with the findings of other studies on similar population, such as De Herdt and Marysse’s (1999) ethnography of the world of Kinshasa’s cambistes (informal money exchange brokers), or Bolt’s (2014) study on wage and sociality among rural workers in Zimbabwe. Most anthropological works on money agree on the relation of money and the construction of social identity. In the exchange market, it seemed that this task rested not on money per se but on the handling of money, and in the way money was not accumulated but moved around. The fingers of the young money vendors of the exchange market were another spindle in a site of production of value, linking hawaladars, M-Pesa agents and gold sellers. I initially struggled to understand why hawala companies, trading almost exclusively in USD, refer their customers to money vendors for converting into KSh. Yet, the entire architecture of the exchange market seemed to suggest that it was exactly through the interactions among these heterogeneous actors that value was produced. I drew a map of the financial relations in the Suk Dahab. It looked like the one below: Retailers KSh KSh Hawala M-Pesa agents Money Vendors Gold sellers Clearing Houses Customers KSh KSh KSh KSh KSh KSh KSh USD USD USD USD Investors KSh USD KSh USD Loans USD KSh KSh 176 This map is an attempt to capture the complex interactions among the different actors operating in the market and mediated by USD and KSh. The solid lines indicate a payment, while the dashed lines indicate a loan or deposit. The money vendors purchase USD with KSh, also drawing from the amounts that gold sellers, all women who are paid in KSh, deposit in their safes at the end of the day. The money vendors resell USD to hawala in exchange for KSh and make a profit. Hawala companies host deposits in USD from investors, retailers and private customers and, in return, invest in clearing houses in order to mobilize their credits. M-Pesa agents operate exclusively in KSh and interact mostly with customers who visit the market to purchase gold and wish to avoid carrying cash. In case of need, they may receive a loan from a related hawaladar. Preserving multiplicity, cultivating alternatives The example of the exchange market illustrates how hybrid arrangements emerge from the intertwining of multiple networks. The debate between a hegemonic and a pluralistic view of monetary standards has a long history in African studies and traces back to Bohannan’s (1959) pioneering study on the impact of Western multi-purpose money on African traditional institutions. Focusing on the Tiv society in Nigeria, he outlines a system based on spheres of exchange, whose distinction is regarded as a guarantee of social stability. Drawing on her own study on the coexistence of multiple monetary realities in Atlantic Africa, Jane Guyer (1995, 2004) challenges this model and the idea that these moral spheres are rigidly separated, and instead describes a complex repertoire of financial instruments and practices to move across scales, ranks and calculative logics. She argues that, by navigating through different monetary regimes, people seek to absorb the uncertainty in volatile economic situations. Commenting on Guyer’s “combinatorial juggling of multiple independent scales – for things and people, in quantities and qualities”, Shipton (2014: 501) links this performance to the need to weather through a crisis. The importance of preserving a heterogeneous repertoire of 177 financial institutions becomes particularly evident in contexts in which crisis is not an exception, but the norm (Vigh, 2008; Taussig, 1992). The coexistence of cash in multiple currencies, mobile money, hawala and ayuuto in a world experienced as shifting and unpredictable is consistent with the need of preserving a range of alternative ways”, already emphasised in my theoretical framework, and engages also with Bourdieu’s notion of ‘social alchemy’, based on the “endless reconversion of economic capital into symbolic capital” (1977: 195). The chart above highlights the role of different currencies (USD and KSh, hard and soft currencies) as ‘interfaces’ regulating the interactions of alternative scales and modes of commensuration. However, rather than ‘a cognitive map’, the visualisation of the interactions in the exchange market is rather a ‘repertoire, the elements pegged to each other in performance’ (Guyer, 2004: 60). It is indeed a performative model that makes visible the production of value through the constant circulation of credits and debts and the creation of obligations that bridge the quantifiable and the unquantifiable time, the realm of the market and the realm of the biosocial reproduction, as in the case of the remittances sent to fulfill kinship duties and group solidarity. The pursuit of USD is a tenet of this strategy, given the capacity of hard currency to hedge against inflation and retains value across borders. Yet, this model rejects a totalizing standard. In the exchange market, each node is a gateway to a separate network: the money vendors are related to each other, the women gold-sellers are members of ayuutos, the hawaladars are employees of transnational companies, the M-Pesa agents work for Safaricom. This complex financial architecture is not simply aimed at transferring money, but, through processes of conversion and commensuration, fulfils the aim of spreading risk across heterogeneous circuits (Guyer, 2004). The model of the exchange market reproduces, once again, the spindle that keeps on weaving the social fabric. 178 Conclusions This chapter has examined the mutual shaping of human and financial flows within Somali networks. It has argued that the strategic combination of informal and formal sets of financial practices responds to the need to bridge different circuits through which value is reproduced. While formally complying with the moral imperative to surrender in the hands of God, through the acceptance of uncertainty as a test of faith, the strategy to cultivate multiplicity seems finalised to the forging of social bonds by atomising and partaking uncertainty. The ambiguity between the acceptance of risk and the attempt to avert it is thus sorted out through practices aimed at creating and nurturing relationships of reciprocity and interdependence, both on a local and translocal level. If remittances are stories, as Lindley (2010) points out, then it is from the constant reproduction of these stories that value is extracted. Little Mogadishu is a stage on which people are engaged in juggling – currencies, devices, identities, temporalities, needs and aspirations. And the way they juggle – ‘standing on one leg’ with more of less adroitness, steady or wobbly – depends on both the solidity of their foothold and their knack to move to another, as dark clouds gather on the horizon 179 Localities 180 6 Crossing borders in troubled times Introduction The previous chapters have highlighted the interplay of human and financial infrastructures that enable the forging of new linkages for the purpose of spreading risk and thus coping with situations of uncertainty. This chapter and the following one discuss how this strategy is played out within specific institutional settings by different Somali groups. This chapter in particular looks at the way linkages and practices of conversion are entangled with mobility strategies. It engages with the literature on migrant networks to examine the emergence of a ‘migrant system’ (De Haas, 2009) between Eastleigh and Kampala’s Kisenyi, shaped by endogenous and contextual feedback mechanisms. But it also draws a theoretical underpinning from Kloosterman’s concept of ‘mixed embeddedness’, mentioned above and here further developed to explain how the interaction of individuals’ resources, opportunity structure, and macro- institutional framework breeds migrant entrepreneurship (Kloosterman, 2010: 27). This chapter examines the combined effect of the deterioration of the security situation for Somalis in Kenya, Ugandan progressive refugee policies and the greater ease to travel between Uganda and Somalia, and argues that to understand the rise in the volume of Somali migrant flows to Uganda we need to clarify the articulation of the broader institutional framework and the opportunity structure. It also argues that mobility is both an indicator and a source of social differentiation, as the access to certain resources (hard currencies, IDs that allow international travels, but also social relations and political capital) enables to explore alternatives that may prove vital when political circumstances limit the range of available options. 181 Troubles in Eastleigh At the beginning of 2013, a long stretch of First Street had been turned into an open construction site. Hawkers had placed their stands among trenches and earth mounds, while workers operated bulldozers and excavators. Lobbied by the EBC, and pushed by the local MP Yussuf Hassan, the Kenya Urban Roads Authority (Kura) and Kenya Rural Roads Authority (Kerra) had launched a 3 billion KSh refurbishment plan for the Kamukunji Constituency, had Although some local businessmen held high expectations that paved roads would boost the economy of the area, ahead of 2013 Kenyan political elections a sense of foreboding loomed over the neighbourhood. While shoppers and goods continued to flow in and out of Eastleigh, the Somali ‘nervous system’ was highly alerted. The combination of terror attacks, the encampment directive, and the imminent elections had created a situation of deep uncertainty. Many Somalis were pondering the possibility to leave Eastleigh, or even Kenya, for the time being. While Kenyan Somalis or refugees from the Darood Ogaden clan were moving to Garissa, or to other locations in the NEP where they could be temporarily hosted by relatives, others were heading to the Kenyan coast, hoping to find protection among Muslim fellows. A few wealthy Kenyan Somalis had left to the United Arabs Emirates (UAE) for the time being. There were rumours of flight of capital, and local money vendors complained about a shortage of hard currency. The official exchange rate in March 2013 was 87 KSh per 1 USD, but, in the exchange market, the offers were above the rate. The Star newspaper reported in January that “such irrational mass victimisation of the Somali entrepreneurs” (Mohammed, 2013) was costing Kenya between 10 and 40 KSh billion (71 to 280 million GBP). These figures are based on estimates, but the Somali traders and financial operators I spoke to agreed on the fact that this shortage was due to the buying frenzy of hard currency caused by the fear of both a fall in the value of the KSh in the case of turmoil and a sudden need to move somewhere else. Cashing out from their bank and M-Pesa accounts, they thus purchased USD to deposit in hawala accounts. According to some hawaladar, hawala companies had learned the lesson since the closure of Al Barakaat in 2001 and immediately used the money to invest in goods that were loaded on Somali-owned trucks in Mombasa and delivered across East and 182 Southern Africa. Abdirizak, the employee of an MTO whose name he preferred not to disclose, suggested that it was in those moments of uncertainty that ‘money has to keep on moving, because when it stays idle the government can grab it more easily”. Moreover, hawala companies allowed withdrawing from Somalia, where, according to many travel agencies and freight companies in Eastleigh, since December 2012 a growing number of Somali refugees were travelling. Indeed, security in Mogadishu had improved over the previous year as a result of the surge of AMISOM troops against Al- Shabaab. Moreover, the election of Hassan Sheikh Mahmoud as president of Somalia in September 2012 was internationally hailed as a milestone on the way to recovery, boosting expectations that Somali could rise from the rubbles (Hammond, 2013). However, the improvement was mostly limited to the Somali capital. Travelling across Southern Somalia remained a perilous journey. This is why many families travelled to Mombasa and from there, by sea, to Mogadishu. The manager of an agency on the first floor of the exchange market explained that January 2013 had seen the daily departure of 4 fully loaded flights to Mogadishu. Entire refugee families were flying back to Somalia after being issued a temporary travel document by the Somali embassy in Nairobi. The price of one-way tickets was around 260 USD, payable only in cash (which helped explain the dollar rush). Sitting in front of a large map of the world with a one dollar bill pinned to it, pictures of aircraft and views of the Burj Khalifa in Dubai on the background, a travel agent confirmed that recent months had seen an intensification of the returns to Somalia, although travelling to Mogadishu was an exacting endeavour, which required contacts and financial resources to secure adequate protection (see also Hammond, 2014). Most journeys were reconnaissance trips, rather than definitive return. Indeed, a key factor in the decision to travel to Mogadishu was, for many refugees, the possibility to go back to Kenya if the need arose (for lack of security or of livelihood), not through the Kenyan-Somali border, which remained officially closed, but flying directly from the airport of Mogadishu to the airport of Entebbe, in Uganda. “It is all thanks to the blue passport!”, the travel agent explained, producing his brand new document, which sported the Somali star on the front cover. It was issued by the Federal Government of Somalia, established in August 2012 to 183 replace the Transitional Federal Government (TFG), and, since then, it replaced the previous green passport, dating back to the time of Siyaad Barre and defined by the Economist (2004) one of the ‘dodgiest passports in the world’ because no country would accept it. Following the election of Hassan Sheikh Mohamoud, Uganda had begun accepting the blue passport at its ports of entry. Reassured by the existence of a way out in case the unstable situation there would flare up again, those who could afford the risk had thus decided to venture to Mogadishu. Once in Somalia, though, many had reconsidered their decision. At the end of summer 2013, in Kampala, I met a Hawiye Abgal businessman who had travelled in February to Mogadishu with his wife and six children. His initial plan was to return definitely Mogadishu, fly periodically to Uganda and travel overland to Kenya in order to oversight his shop in Eastleigh. Later, though, the quick deterioration of the security situation in both Somalia and Kenya forced him to revise his plan and to move his family to Kampala. Like him, other Somalis were relocating from Eastleigh to Uganda. As I did the same, I found that the official figures corroborated the rumours in Eastleigh. According to Uganda's Office of the Prime Minister (OPM), the main administrative body which deals with refugee affairs, the number of registered Somali refugees in the country soared to 41,515 in the year to March 2014. The influx rose from 27,143 in December 2012 and 8,239 in 200819. The overwhelming majority of the new arrivals were previously refugees in neighbouring Kenya, but had re-applied for refugee status upon reaching the Ugandan border. Others had applied as refugees in Uganda for the first time after illegally venturing through Kenya from Somalia. Moreover, this number did not account for Somalis with foreign and Somali passports based in or regularly travelling to Uganda. An emerging ‘migration system’ in a volatile context The events described above depict a volatile landscape. However, as pointed out when discussing the eruption of the Somali civil war, the onset of the refugee crisis and the transformation of Eastleigh, Somali networks have developed by adjusting to a 19 Source: Uganda’s Department of Refugees - Office of the Prime Minister 184 background in which, far from being a temporary interruption of normality, crisis is a chronic situation. Vigh (2008) uses the notions of ‘fragmentation’ and ‘normalisation’ as conceptual tools to grasp how social life becomes “adjusted and attuned to volatility” (13). The terms capture the practices through which the actors read the social environment and adjust their movements accordingly, able to recreate a ‘normality’ (although unpleasant) and, at the same time, to imagine an alternative (ibid.). Living ‘on shifting ground’ (Vigh, 2009), people ‘navigate’ through uncertainty by seeking to anticipate changes in the political and economic context. They do so by participating into the local communicative ecology to acquire information which plays out in the shaping of migratory projects against a social, cultural and political background. The importance of information, ideas and consumption patterns is emphasised in De Haas’ (2009) conceptualisation of ‘migration system’ to refer to “a set of places linked by flows and counterflows of people, goods, services, and information, which tend to facilitate further exchange, including migration, between the places” (8). De Haas argues that migration systems are shaped by endogenous and contextual feedback mechanisms. Endogenous feedback mechanisms refer primarily to the ‘network effects’, which resonate the process already identified as ‘chain migration’ (MacDonald and MacDonald, 1964), according to which prospective migrants take mobility-related decisions on the basis of information provided by their related predecessors and rely on their support and help for acclimatisation during the early period in the country of destination. In the case of Somali networks, these mechanisms are embedded in genealogy and group solidarity, as discussed in chapter 2. However, De Haas (2009) suggests that the excessive focus on internal dynamics of Somali networks have diverted the attention from the contextual feedback mechanisms through which migration flows bring about transformations for sending and receiving communities and societies at social, cultural and economic level. These transformations range from widening social and economic gaps to fostering entrepreneurial ambitions and achieving cultural changes. For De Haas, it is necessary to integrate the comprehension of the endogenous mechanisms with the understanding of the contextual factors that may either facilitate or inhibit specific forms of migration to understand how migration 185 processes constantly interact with the broader context. Moreover, this interaction may engender ideas and behaviours of migrations that, as I have mentioned in the literature review, Levitt (1999) calls ‘social remittances’, and that contribute to shaping a ‘culture of migration’, in which social prestige is attached to mobility and migrating is largely perceived as a rite of passage (Massey et al. 1993: 453). The evidence suggests that this is the case for many Somali migrants, for which a greater mobility is at the same time an enabler of agency and a strategy to achieve aspirations. This attitude resonates with what Chu (2010) observes among aspirant migrants from Fuzhou, China, for which mobility is a quality embodied in behaviours and lifestyles, or, according to her definition, “a privileged qualisign of modern selves and relations” (14, italic is mine). Aspirations are seldom considered in the scholarship on African migrations, dominated by an excessive focus on coping strategies, particularly in discussions on forced migrations (Carling, 2014). However, as already mentioned in the literature review, the reality on the ground appears much more variegated than the explanations based on conflict-induced displacements seem to suggest (Bakewell and Bonfiglio, 2013). Van Hear (2014) links the scope for aspiration allowed by migration to power and class relations, reprising Hyndman’s (2000) observation that “those with money can take advantage of time-space compression. Those who are uprooted from their homes and forced to flee their country with few resources experience migration in a very different way “(37). Mobility strategies often may change according to both fresh motivations and contextual feedback mechanisms. It is worth noting, though, that the notion of individual aspiration may be problematic. Consider for instance the above-mentioned notion of buufis, or longing for settlement to a third country. It may reflect the desire to enhance one’s agency and pursue aspirations that are constrained in the host country. At the same time, the resettlement of a family member entails the commitment to support the relatives left behind 186 “Should I stay or should I go?” March elections passed without major incidents, and, during the spring and summer 2013, local tensions abated. Many Somalis who had temporarily moved out of Eastleigh returned, but others had eventually decided to settle in Uganda. As suggested by the above-mentioned cases of Ferdowsa, the fabrics trader, and Abdicadir Cali, the computer fixer, the growth of the Somali presence in Uganda had boosted the demand for goods and services. Nevertheless, a sense of unease remained, influencing mobility- related decisions. The case of Abdi, a 20-year-old born in Kismayo, but now a Kenyan citizen, was particularly interesting. When I first met Abdi, he was selling mattresses in the shop of his uncle on Eastleigh’s First Street. He was planning to move to Kampala by autumn 2013 and enrol in the Kampala International University at the suggestion of his father. His family had already lived in Kisenyi in the 1980s, when his father, a Darood Dulbahante, was importing sugar from Brazil to Somalia and then across East Africa (as it shall become clear later, his clan of belonging mattered). Abdi attended primary school in Kampala and then his family moved to Kenya. Later on, his father established a construction company in Thika, a town 40 Km north-east of Nairobi but, given the current situation in Kenya, he had come to the conclusion that the best solution for his son would have been to attend university in Uganda. Abdi’s plan was to study medicine in Kampala and then move to China. We agreed to meet in the Ugandan capital, where he was supposed to arrive at the end of September. His plans were dramatical to change as, on 21 September 2013, a commando of Al-Shabaab gunmen raided the Westgate shopping mall in Nairobi's Westlands, an area mostly frequented by upper-class Kenyans and foreigners working for humanitarian agencies and in the private sector, and in the ensuing three days siege more than 60 people lost their lives. In the wake of the siege, the uncertainty of the time and the modes of the government’s response was such that there were many discordant positions on how to brace for the future. By phone, few days after the attack, Abdi told me that he was still resolute to travel to Kampala, but he had decided to stay put for the time being because crossing the border would have drawn suspicions. Following the Westgate attack, also the Ugandan intelligence raised the terror alert level. At the beginning of 187 October, the streets of Kampala were plastered with pictures of an alleged Al Shabaab terrorist of German origin who, according to witnesses, had entered the country a few days earlier by bus through Kenya (Sudan Tribune, 2013). In the pictures, Ahmed Khalid Mueller, this was the name of the suspect, sported very short hair and a scruffy beard. This was enough for an agent of the Ugandan police to find resemblances between him and me, and stop me as I was hanging around the Kisenyi mosque during the mid-day pray, waiting for a Somali friend to come outside. The issue was quickly sorted out at the Old Kampala police station before an inquisitive, yet very professional, young police officer, who went through my passport and then, before letting me go, shared with me his concerns. He lamented a lack of coordination with the Kenyan intelligence and expressed general criticisms over the lack of community policing of the Kenyan police. The differences between Kenya and Uganda in handling the terrorist threat became starker in the months to come. Faced with controversies in the aftermath of the Westgate siege, the government of Uhuru Kenyatta pandered widespread anti- Somali feelings among the general public, and, on 10 November 2013, signed a tripartite agreement with its Somali counterpart and the UNHCR. The document established the legal framework for the repatriation of more than 500,000 Somali refugees currently living in the country, to be carried out according to an unspecified timeframe and in a ‘voluntary’ way (UNHCR, 2013).20 The Kenyan government’s announcement of repatriation sowed apprehension among refugees. Young Somali refugees, even from better-off families and torn between buufis and fadikudire, claimed that times were not ripe yet to return to Somalia. The most vulnerable refugees from the Jareer and other minorities feared the return to the areas in Southern Somalia they had fled. They were aware that, once back to Somalia, they would have been again at the mercy of either bandits, clan militias or Al Shabaab (well aware of the thin line separating each group). Moreover, the 4.5 formula regulating the Somali political system, described in Chapter 3, was perceived as the officialisation of the discrimination they had suffered throughout 20 UNHCR, Tripartite Agreement Between the Government of the Republic of Kenya, the Government of the Federal Republic of Somalia and the United Nations High Commissioner for Refugees Governing the Voluntary Repatriation of Somali Refugees Living in Kenya, 2013, 10 November 2013 188 Somali history. Omar, my Somali teacher, was constantly in touch with his mother in Marka, a coastal city still controlled by Al Shabaab despite the presence of Amisom troops, and was aware that, in the case of return, he would have been a primary target. “I am too old to be recruited, I won’t have any value for them and I will be returning after a long time in Kenya. They would think I am a spy and kill me immediately”, he said. Returning to Somalia would have been entrapped him, and other Jarer, again in a situation of political disenfranchisement, social marginalisation, and economic destitution. While staying in Kenya, the hope of resettlement to a third country, intended as a ‘gateway to a brighter future’ (Nyamnjoh and Page, 2002: 627), was still alive, although dimming. However, the most immediate effect of the provision was to exacerbate the vulnerability of Somali refugees further by propelling the ethnic profiling of the Somali population and implicitly giving the Kenyan police free rein to harass and extort money (HRW, 2014). Stoking pre-existing communal tensions, particularly in the coastal region (BBC, 2014a), and fuelled by belligerent op-eds (Daily Nation, 2014) and diplomatic rows between the Kenyan and the Somali government (BBC, 2014b), the security crackdown escalated between March and April 2014 when, following attacks on a church in Likoni, Mombasa, and at a bus stop in Eastleigh, the government launched the Operation Usalama Watch (Peace Watch). In the course of a massive security swoop across Kenya, the police arrested over 4,000 people, restricted the movement of refugees to camps and began the refoulement of dozens to Somalia (BBC, 2014c). The operation not only was punctuated by gross violations of human rights (HRW, 2014), but also proved ineffective to curb the violence: large-scale attacks were staged, allegedly by Al Shabaab, in the NEP again in June and then in November and December 2014. The response of the Kenyan government to the deterioration of the security situation in the country put repatriation at the centre of a debate involving policymakers and humanitarian agencies (as often happens, the voice of the refugees themselves was conspicuous by its absence). In policy discourses, repatriation is usually framed as one of the possible solutions to protracted displacement situations, along with integration in 189 the country of asylum and resettlement to a third country (Black and Koser, 1999; Horst, 2006; Omata, 2013). Recent years have generally seen narrowing possibilities for integration and resettlement, leaving repatriation (or its announcement) as the option that best fits populist political agendas in the host countries. The emphasis is placed on the voluntary character of the return in order to formally abide by the principle of nonrefoulement. Yet, as Omata (2013) points out in his study of decision-making processes among Liberian refugees in Ghana, refugees’ return decisions are far from linear. Indeed, humanitarian organisations and host governments often attach repatriation to a vague feeling of nostalgia for the homeland left behind and therefore speak of ‘homecoming’ (Omata, 2013; Bakewell, 2000). In so doing, though, they neglect “the dynamic nature of return and ‘home’ among forcibly displaced populations caused by changes over time” (Omata, 2013: 1282), overlooking that physical safety is not the only reason refugees take into account when considering to return to the country of origin. Omata also mentions “refugees’ current personal and familial situations”, “pursuit of future objectives, [...] the possibilities of other durable solutions, and [...] access to support networks to re-establish their lives in the country of origin” (ibid.). By stressing the fact that refugees’ return migration is a ‘situated concept’ (Oxfeld and Long, 2004: 6), he suggests that it is necessary to consider the “untidy nature of return decision-making” (Omata, 2013: 1292): current circumstances in which refugees live and available resources play a role in refugees’ decision to go back home. At the time of my fieldwork in Eastleigh, travelling to Somalia on reconnaissance trip was a key strategy to assess the level of security, but also the possibility to continue current businesses or establish new ones upon arrival. In general, the go-and-see visit is a common risk-reduction practice among forced migrants planning to travel back to the homeland. Typically, a member of the family would be entrusted to check the state of the family properties and assess whether there are the conditions for returning home (Afghan Research and Evaluation Unit, 2006: 46; Korac, 2009: 127). This strategy, though, is expensive because it entails the arrangement of security measures, mostly with clan members in the country, who provide logistic assistance. Among Somalis travelling from Kenya, other aspects were considered: business opportunities, or safe 190 escape routes in the case of a worsening situation. Moving to Uganda was an option for those who had resources to invest in the exploration of new routes, both in terms of migratory patterns and career trajectories. However, it is necessary to consider how the significance of Uganda as a key destination for Somalis depends on legacies that have endured since colonial times and linkages that have been strengthened by major changes at the regional level. The route to Kampala The Somali presence in Uganda traces back to the early 1900s and coincides with the arrival of Somali soldiers from northern Somalia recruited in the British army. However, it was only in the late 1970s that the community found eventually its heart in Kisenyi, a shabby and ill-famed area of Kampala's Mengo Municipality, mostly inhabited by menial workers at the nearby Owino market. There, a Darood Dhulbahante Imam from Northern Somalia, Sheik Abduhani, well connected to local authorities, was granted a concession to build a mosque. Sheik Abduhani was a leading figure among Somalis — mostly from Darood Harti clans — interested in setting up entrepreneurial activities in Uganda, or using the country as a stepping stone, mainly to neighbouring Zaire. During this period, entrepreneurs of Somali origin began consolidating their control over strategic business sectors such as transport. Among them, there was Ahmed Omar Mandela, owner of the Mandela Group — a car and truck spare parts company established in the 1980s — and later of the fuel supply company City Oil and the Java Café chain. In 2004, Ahmed Omar Mandela was appointed Minister of State for Management and Development of the Royal Treasury of the Buganda Kingdom, an achievement which still today young Somalis, particularly from northern Somalia, in Uganda regard with pride. Another Somali Ugandan notable was Hussein Shire, born in the 1930s in Tororo, where he started building up his transport company by ferrying passengers on a blue Peugeot 504 between Tororo and Malaba, on the border with Kenya. In the 1980s, his transport business expanded into a 100 bus fleet bus company, Gateway, and an oil company, a store chain, several buildings across the country and 191 around 809 ha of sugar cane plantation supplying two sugar factories21. Somali business networks across Africa thus began taking shape prior to the collapse of the Somali state, lubricated by clan-based trust, and straddling the border with Zaire, where Somali traders interacted with counterparts from other corners of Africa in places such as Ariwara, a Zairian market town close to both the Ugandan and Sudanese borders (Meagher, 1990: 73). Transport tycoons and truck drivers of Somali origin developed ties which proved helpful also during volatile times, such as in 1989, when a screening to determine citizenship status in Kenya prompted the flight of almost 15,000 Somalis to Uganda, who obtained temporary political asylum from the Ugandan government (Daily Nation, 1988). After 1991, most of these Kenyan Somalis returned to Kenya, as the incipient refugee flow heralded future business opportunities, while the first refugees from Somalia arrived in Uganda. In 1994, the Ugandan government allowed the UNHCR to settle recently arrived Somali refugees in Nakivale refugee camp, but not in urban centres (Bagenda, Naggaga and Smith, 2003). Refugees from Darood Harti and Isaaq clans with connections and resources settled in Kampala and Jinja, although many among those who moved to town retained their registration in Nakivale because of the widespread belief, yet unfounded, according to UNHCR officers, that being registered in a refugee camp would speed up the resettlement process. In 1994, as a testament to the growing economic success of Somalis in Uganda, a larger mosque with an annexed madrasa (college for Islamic learning), named Al- Tawheed, was inaugurated by Sheik Abduhani, who had been raising funds among the local business community. The first Somali hotels and restaurants were established around the new Kisenyi's mosque thus accelerating the spatial clustering of the Somali community. Despite featuring rundown infrastructures and widespread petty crime, the area started attracting young refugees in search of job opportunities and encouraged by cheap accommodation and its proximity to Kampala’s largest open market and one of the 21 Hussein Shire eventually passed away in South African in July 2014 ("Gateway bus founder dies in South Africa". New Vision) 192 city’s main bus and truck terminal. In the 1990s, the Ugandan government started implementing liberalisation policies that further boosted Somali businesses in capital- intensive activities such as fuel supply. Somali remittance companies became increasingly important and, at the end of the 1990s, a branch of Amal, a money transfer operator from Puntland, was opened in Kisenyi by the same Sheik Abduhani. By bestowing his prestige over the company, he turned Amal into the main conduit of financial capitals that, through the 1990s, allowed Somali businessmen to seize control of the strategic oil and transport sectors. The oil was usually purchased by clusters of entrepreneurs, who arranged its shipping from the Gulf to Mombasa. From there, it was funneled to Eldoret through the Western Kenya Pipeline Extension (WKPE)22 and transported by truck to Uganda and further on, to DRC, Rwanda and Burundi. Large fuel distribution companies emerged in the 1990s and, as in other Somali business sectors, they were able to cooperate to a certain degree. They were all related to Darood Harti (Dulbahante and Warsangeli) entrepreneurs from Northern Somalia but with Ugandan citizenship. Most of these companies, such as the market’s biggest players, CityOil, described above, and Hass, established by Abdinasir Alidheere, are still operating in 2015,. Other companies, working mostly in logistic and transport, are Hashi Impacts, a Somali-owned company that manages fuel storage depots in Mombasa and a fleet of trucks, and operates from Nairobi and Kampala; Ainu Shamsi, whose core business is the shipping of refined oil from the Gulf; and Banoda petroleum, a freight company that manages an average of 50 trucks a day and a few stations across the country.23 These companies became major employers for Somalis arriving in Kenya. On top of that, the introduction of the East African passport in 1999 eased border crossing for East African citizens, thus facilitating the transit of suitcase traders, notably from Eastleigh. Through the 2000s, Uganda continued to attract mostly refugees with pre- existing connections. The Somali presence increased around Kampala International 22 The 446 Km WKPE was opened in 1994 to expand the existing Mombasa to Nairobi pipeline and connect it to Nakuru, Eldoret and Kisumu. 23 I owe this information to Mohamed Hersi. 193 University, Cavendish University, and St. Lawrence University, in areas such as Kabalagala, Kasanga, Mengo Rubego. However, Kisenyi remained the centre of the Somali community in Kampala, which became more conspicuous with the opening of more shops and eateries with a clear Somali connotation. At the same time, daily bus connections between Nairobi's 'Little Mogadishu' and its Kampala's counterpart, and the possibility to use M-Pesa in roaming, sustained a growing Somali refugee community, earning a living by trading a wide range of goods, from camel milk to clothes. Yet, the high fees of commercial licenses slowed the proliferation of business activities. The cost of a business license was around USD 2,000 and only those able to raise the necessary capital could rent a shop. The country was indeed considered only a profitable market for big businesses, and most refugees still saw Kenya as an easier place for Somalis to live. However, this disadvantage was gradually offset by the increasing cost of living in Nairobi and by a changing regional configuration. A new Little Mogadishu takes shape The development of the Ugandan ‘Little Mogadishu’ mostly mirrors the evolution of Eastleigh, but it brings to the fore the significance of the national institutional framework for the local opportunity structure (Kloosterman, 2010) and of the broader regional context. Uganda sits at the centre of what Crisp (2000) calls one of the two “principal sub- regions of displacement” in Africa, which is “the vast area of central Africa which… encompass[es] the Democratic Republic of Congo [DRC], Congo Brazzaville, Burundi, Ethiopia, Kenya, Rwanda, Somalia, Sudan, Tanzania, Uganda and Zambia” (2). This area is, at the same time, a producer and a recipient of displacement flows. In the past decades, Uganda has maintained its primary role as a place of asylum for refugees mostly from South Sudan, DRC, and Rwanda. Because of this central position in a conflict-ridden region, refugee issues are central in Uganda’s internal politics, to the extent that, according to the Uganda’s First Deputy Prime Minister and Minister for Refugees, Moses Ali, “our geographical position has had some effect on us. You see, 194 we are more or less in the centre of the continent, and I don’t think there will be a time when we will be without refugees” (Meyer, 2006: 5). For this reason, Uganda features in many academic studies and policy papers on humanitarian spaces hosting mostly refugees from DRC and South Sudan. (Dryden- Peterson, 2006; Dryden-Peterson and Hovil, 2003). Other studies have examined Uganda’s innovative humanitarian framework, which adopts a development-based approach that promotes the ‘self-reliance’ of refugees (Betts, 2013), in stark contrast to the encampment policies of neighbouring countries. This framework in based on a Refugees Act approved on the 24 May 2006, with which the Ugandan parliament repealed the previous Control of Alien Refugees Act (CARA), in place since 1964 and object of criticism from the UNHCR and activists. Hailed as ‘progressive [and] human rights and protection-oriented’ by advocacy organisations (RLP, undated), the new law embraces a refugee definition in conformity with the 1951 Convention, of which Uganda is a signatory. The 2006 Refugee Act grants refugees the right ‘to an identity card or travel document, to remain, to non-discrimination, to administrative assistance, to freedom of religion, to freedom of association, to access the courts, and to freedom of movement’ (ibid.). Moreover, it recognizes ‘rights in respect of which refugees must receive treatment in accordance with “aliens generally in similar circumstances”: rights regarding movable and immovable property, the transfer of assets, public education above the elementary level, self-employment, liberal professions, and wage-earning employment’(ibid.). As for the right of association, the Refugee Act clearly states that refugees 'have a right of association as regard non-political and non-profit-making associations and trade unions' (ibid.). This provision paved the way to the vibrant Somali student activism that I will discuss later. It is important to point out here that, although the Act was set to enter into force in 2008, its approval contributed to building trust between the Ugandan state and the refugee population. Moreover, it stood in sharp contrast to the more restrictive Kenyan refugee Act passed in December 2006. Also, as in 2006 Somalia experienced a renewed outburst of violence following the Ethiopian invasion, with a resulting fresh wave of displaced people, both within Somalia and outside, a large number of refugees poured into the Dadaab refugee complex and 195 Eastleigh. Ugandan media began turning the attention to the local urban Somali community,24 and increasingly so as the Ugandan government of Yoweri Museveni was discussing the deployment of the Ugandan People Defence Force (UPDF) to Somalia. The Somali community in Kisenyi took indirect part to this discussion, staging demonstrations to support the Ugandan intervention (Musamali, 2007). In March 2007, the first Ugandan troops were deployed to Somalia under the command of AMISOM. As a result of the intervention, Al-Shabaab threatened to retaliate against targets in Uganda (Jaramogi, 2008). In response, the Ugandan army publicly called on the Somali population to cooperate with security agencies (ibid.), but, on 7 November 2008 a bomb detonated in Kisenyi, killing two (Jaramogi and Wagamulo, 2008). The Ugandan government, particularly the OPM and the Ministry of Internal Affairs, intensified the collaboration with local Somali organisations. In particular, the Somali Community Association in Uganda (SCAU) became a key partner of the government, in charge of monitoring the burgeoning Somali refugee population and collecting intelligence on the Somali theatres of operation of the UPDF troops. It should be noted, though, that the authority of the SCAU was not acknowledged by all Somali refugees in the country. Indeed, the association was dominated by Somalis from South-Central Somalia. Northerners were active in large businesses and were mostly connected to the Somali Ugandans. Following the implementation of the Refugee Act in 2008, a growing number of Somali refugees crossed the border from Kenya at the Ugandan ports of entry of Busia and Malabo. There, they were required to declare their status to the border authorities which would allow them 30 days to register either at Old Kampala police station or at the Office of the Prime Minister’s Department of Refugee Affairs. If they had a sponsor in an urban area, they were allowed to reside in a city, otherwise, they would have been resettled in Nakivale refugee camp. Upon arrival in Kisenyi, where Somali guest houses were located, they would have had to register at the SCAU in order to be pre-screened by the security officer of the association. The cooperation between the Somali community—often outspoken in supporting the Ugandan mission in Somalia 24 ‘Uganda: Kisenyi - a Haven for Somali Tranquility Or Scams?’ The New Vision, 27 September 2009 196 (Kajoba, 2009) —and the Ugandan government was tested again in the aftermath of the 11 July 2010 terror attacks, when 74 people were killed by three bombings in Kampala—one in Kabalagala's Ethiopian Village, a venue packed with football fans watching the World Cup Finals, and twin blasts in Nakawa's Kyadongo Rugby Club (BBC, 2010). The attacks were claimed by al-Shabaab and stirred fears of an escalation of police harassments of Somali refugees (Wesonga, 2011). The impact, though, was limited to stricter vetting procedures for new arrivals from Somalia: applicants to refugee status who wanted to live in Kampala were required to register at the Old Kampala police station with a local guarantor.25 This cooperation yielded a number of arrests on suspicion of terrorist activities but the considered response of the Ugandan authorities strengthened the positive perception of Uganda as a context of reception for Somalis. Youth mobility The 2006 Refugee Act made Uganda particularly appealing to many Somali youths, attracted by the greater freedom of movement and by the possibility to pursue higher education in Ugandan institutions. Refugees were entitled to pay the same fees allowed to Ugandan nationals in both public and private universities, whose number had skyrocketed since the late 1990s to fulfill a booming demand for enrolment. Moreover, student cards had the same legal value of state or UNHCR issued IDs, thus providing an additional document which many young refugees were particularly eager to produce instead of a refugee card. In terms of self-perception, many students felt that, through university enrollment, they were distancing themselves from the condition of refugee. The lower cost of living in Kampala and the lower risk of harassment by the police than in Nairobi were an incentive also for young Somalis with Kenyan passport. The ease of purchasing a university degree and the greater freedom of movement in Uganda shaped the local opportunity structure, thus raising expectations in young refugees to move away from ethnic businesses in a relatively short time, and towards other sectors, and 25 This provision was eventually dropped in 2013. 197 greener pastures. Kloosterman (2010) significantly points out that the institutional setting has a key role in articulating accessibility of markets, referring to the level of human capital necessary to start a business, and growth potential. He identifies three typical market openings along these axes, coinciding with “high threshold, low growth potential (less common unless there is strong discrimination); relatively little human capital and high growth potential (‘post-industrial/low-skilled’) and relatively high human capital and high growth potential (‘post-industrial/high-skilled’)”(40). The integration of young Somalis recently arrived in Uganda and with no specific qualification, would proceed gradually: they would be initially supported by relatives to facilitate their integration in the ethnic economy and, later, enhance their human capital to take full advantages of the opportunities offered by a dynamic regional market, as I will explain later. For many Somali students, the decision to enroll in a local university was driven not only by the wish to pursue a degree but to also by the ambition to build a reputation as a reliable entrepreneur or as a leader, creating a base of followers among the fellow students or drawing the attention of established businessmen. A group of Somali refugee students at the Kampala International University (KIU), a private institution, were among the first to set up an informal MTO that adapted M-Pesa to the need of many Somalis in Kampala. The system is worth describing because it illustrates how a formal technological infrastructure is appropriated and reinvented. Indeed, Safaricom does not rely on a network of official M-Pesa agents in Uganda and does not have a license to operate in the country as a mobile money operator. However, Safaricom operates in Uganda as a mobile network operator (MNO) through a tie-up network agreement with MTN, a local MNO. M-Pesa transactions thus occur ‘over the counter’: users cannot send money from their MTN SIM card, but instead pay the agents, who have a Safaricom SIM card and Kenyan Shillings in their M-Pesa accounts, to perform the payment in roaming on their behalf. The Somali students working as money transfer agents simply ‘rent out’ their M-Pesa mobile wallets to colleagues who need to send money but want to bypass official money transfer agencies. This informal usage of M- Pesa is preferred in case of small amounts. The freelance agent is handed by the 198 customer the cash in Ugandan Shillings (UGX), including a small fee for the transaction, then calculates the amount in KSh (the currency in his M-Pesa account) and, if he has the corresponding amount in his M-Pesa wallet, sends the payment to the recipient in Kenya. At the same time, he can to receive payments from Kenya on his M- Pesa wallet associated with his Kenyan phone number and then pay the corresponding amount, minus a fee, to the recipient. These freelance money transfer operators travel regularly across the Kenyan border to top up their M-Pesa mobile wallets, often bringing back to Uganda over 100,000 KSh (1000 GBP) in a handful of SIM cards. This implies a strong level of interpersonal trust in the agent because, in the case of misplaced payment, there is no way to complain to Safaricom. Working as money transfer operator is often a way of gaining a reputation among other youth and businesspeople. This social prestige may then be invested and enhanced by joining a student association and thus start building a political career. The role of student associations Student associations are key organisations to expand social networks and accrue social capital. They have animated Somali political life since colonial times, particularly in Northern Somalia. Indeed, the first unions in Uganda were established at the end of the 2000s initially by Darood Harti students to help prospect and current students from the same clan. These associations fulfilled two main functions: 1) providing support to prospect and current students, especially for getting information ahead of relocation in Uganda and university enrollment and coursework; 2) establishing connections with the Somali business community. This latter aspect is particularly relevant: Somali businessmen support student associations or offer scholarships to students from their own clan, mainly to individuate trustworthy individuals to employ, but also to create loyalty bonds and groom future political leaders. The funds are deposited in the student associations’ accounts in hawala branches (where they can also be monitored by the sponsor of the association) and are used to pay for the associations’ offices and to cater at student conventions. Student associations’ elections held in the various Kampala’s 199 universities are major political events in which ambitious young refugees would display their oratory skills and political adroitness, by evoking the enduring links with the homeland, exhorting their fellow students to get an education to rebuild Somalia, and acknowledging the contribution of the sponsors, in front of an audience waving the Somali flag. These spaces are expanded online, on the Facebook pages of the associations. Indeed, since 2011-2012 Facebook has become very popular among Somali students, and Somali student associations pages have proliferated on the social network. In these online communities, prospect students can stay in touch with friends and relatives and request information before moving to Uganda. Although clan names are not made explicit, social relations are mostly along clan lines, intersected with more mundane forms of memberships based on universities or football clubs. The link of these Facebook pages to a specific clan is often revealed by ‘liking’ the pages of Somali politicians whose clan’s affiliations are well known, or referring to events that characterize the history of a specific lineage. The fact that the most popular degrees are in business, development studies and public policy reflects the likeliest career trajectories for Somalis in Somalia and East Africa. Student associations enable to gain access not only to potential Ugandan Somali employers, but also to INGOs operating in Uganda and working with Somalis, or, through the Somali Embassy, with entrepreneurs or international donors funding programs in Somalia. Forms of ‘tactical cosmopolitanism’ (Landau and Freemantle, 2010) are thus deployed by instrumentally embracing a rhetoric of universalism and commitment to pluralism in order to achieve tactical targets. Igarashi and Saito (2014) provide an interesting contribution by linking cosmopolitanism to Bourdieu’s reflection on cultural capital and the reproduction of social stratification. The idea is that cosmopolitanism can be institutionalised and accumulated as cultural capital, and, as such, converted into other capitals. They thus suggest that while cosmopolitanism as openness to foreign others and cultures can be acquired as part of habitus (an embodied state) and through consumption of foreign commodities (an objectified state), we hypothesize that it becomes most clearly integrated into stratification, struggles for dominant positions within 200 society, when it takes the institutionalised form of academic qualifications (ibid., 223; see also Lareau and Weininger, 2003). Their argument helps explain how Somali student associations in Kampala are functional to dynamics of both inclusion and exclusion. They facilitate the access to ‘cosmopolitan’ competencies, related to the capacity to interact with foreigners and their cultures, through for instance the acquisition of the “ability to speak foreign languages and navigate through foreign cultural environments” (ibid.) which are often related to the availability of economic resources. Travelling to Uganda is thus for many Somali youths the opportunity not only to acquire cultural capital through academic qualifications but also to make it profitable on the labour market and turn it into “the condition for legitimate access to a growing number of positions, particularly dominant ones” (Bourdieu, 1983: 254). By underlying that the education systems is integral to stratification, Igarashi and Saito thus suggest that cosmopolitanism, which education systems around the world increasingly define as cultural capital, also yields profits on labor markets, allowing its owners to gain access to the growing number of positions that require extensive interactions with people of multiple nationalities. (Igarashi and Saito, 2014: 231) Regional connections This aspect is particularly important when looking at how rising business opportunities in neighbouring countries have turned Kampala into a regional hub to operate across borders. Recent years have seen the emergence of a business route between Uganda and South Sudan, where the foray of aid organisations, INGOs, and foreign companies after the Independence in 2011 had created tender opportunities paid in USD. Somali businessmen invested in construction, transport and trade, particularly in the capital Juba and in Bor, Jonglei State and Kisenyi became a transit point for the trucks coming from Mombasa through Nairobi and travelling to South Sudan. Mostly dealing in supplying fuel, foodstuff and building material, they were also contractors for road construction, and provided logistic support to the Somali and other African embassies in South Sudan. In June 2013, the South Sudanese Central Bank limited the 201 provision of hard currency to commercial banks, prompting many Somali businessmen to reduce their investments in the country. The decision was subsequently reversed in November but, after violence broke out in South Sudan in December 2013, the flight of Somali businessmen accelerated. Around 500 Somali citizens were flown to Mogadishu by the GoS (Dalsan Radio, 2014). In mid-2014, many Somali entrepreneurs started returning to South Sudan. Somali passport holders flew directly to Kampala to obtain a south Sudanese visa. The majority of applicants for visas were truck drivers for domestic fuel services and heavy machinery operators. Since only Kenyan Somalis with Kenyan driving licenses could drive fuel trucks through the South Sudanese border they earned more than Somali passport holders. The story of Abdulqani Umar Jamac, a young Somali Dulbahante refugee, was particularly illuminating. Born in Las Anood, an area still disputed between Somaliland and Puntland that he still identifies as Somalia, Abdulqani arrived in Kampala in December 2009. He had family in Uganda who helped him get there and, when he first arrived, landed a job at a petrol station. According to his initial plan, Kampala was supposed to be a transit stop on the way to the West, either Europe or North America, but when he realised that his resettlement application did not advance he decided to look for work. In 2012, right after independence, he moved to South Sudan with a few others for work. He decided to invest his savings in business in Uganda and, thanks to some family contribution, he opened a shop in Kisenyi. Currently, he made a living selling second-hand phones and accessories, and managed a money exchange. Moreover, he took evening business classes at a local college. His ambition was to break into the fuel supply business and explore the ample business opportunities across East Africa, venturing as far as DRC and Tanzania, where he has valuable contacts. Somali entrepreneurs control fuel distribution also in Rwanda and the DRC, despite the local linguistic barriers (in both case the dominant language is French) and— particularly in the case of the latter—insecurity. Therefore, the fuel business is less centralised in order to spread the risk. Some Somalis only own petrol stations with others supplying them from Kenya. However, companies are starting to invest in the whole supply chain. In the case of Rwanda, Paul Kagame's government adoption of English as another official language of the country is facilitating the access for Somali entrepreneurs. Kisenyi is located at the centre of an extensive network of services, mainly long- 202 haul transport and fuel supply. This network straddles as far as Kisangani, in the DRC, where some are engaged in timber extraction and trade, and Zambia. Strategic synergies The Somali community plays a pivotal role in the relationship between Uganda and Somalia, a relationship which further Uganda’s geopolitical ambitions in the region. As mentioned earlier, after the election of Hassan Sheikh Mohamud as president on 10 September 2012, Uganda became the first country since 1991 to accept the Somali blue passport at its borders. The document was issued either in the Somali capital or at the Embassy of the Republic of Somalia in Kampala. On 8 July 2013, Air Uganda began operating between Entebbe and Mogadishu, the second national air company, after Turkish Airlines, to re-establish the connection with the Somali capital. This facilitated many in the Somali diaspora—with either Somali or foreign passports— to travel from Entebbe to Mogadishu and back for the first time in many years, often on a reconnaissance trip to meet relatives, check on family's properties and pursue business opportunities. Improved security and growing economic expectations had started fuelling a building spree across Mogadishu (Hammond, 2013b) and boosted land acquisition in the most fertile areas of the country, particularly along the rivers Shabelle and Juba. Yet, the situation on the ground remained volatile, and Somali entrepreneurs were reluctant to move back to Somalia with their families. Instead, even refugees based in Kenya benefited from the greater ease to travel back and forth between Entebbe and Mogadishu, travelling to Uganda as refugees and then to Somalia as Somali citizens. When going back to Nairobi as refugees became more difficult, some decided to resettle in Kampala. The Somali population became more heterogeneous, encompassing not only students but also families. As Kenya intensified the deportations to Somalia, many flew back from Mogadishu to Entebbe, this time not as refugees but as Somali citizens. Uganda is the top troop contributor of AMISOM, deploying at the time of writing 6,223 soldiers in some flashpoints such as Banadir, Lower and Middle Shabelle, Bay 203 and Bakool26. At the deployment of UPDF troops in 2007, Uganda was viewed by Somali warring parties as unbiased, not having a large Somali minority, nor direct stakes in Somali politics (Hansen, 2013). Discussing the reasons for the Ugandan participation in the AMISOM mission, some analysts have talked of a blend of concerns of a 'domino effect' of instability in East Africa; a way to appease the army, whose officers and privates would benefit from a higher pay than at home—therefore lowering the pressure of the armed forces on the regime; and the will to strengthen the relations with the US and in general to bolster its international prestige (ibid.). Following the 2010 Kampala bombings, Uganda's commitment against al-Shabaab on Somali soil became a matter of national security. In pursuing this strategy, the Ugandan authorities have been unwavering in support for both legitimate institutions in Somalia and Somali grassroots organisations in Uganda. The partnership between the state and the SCAU resonates Evans’(1996) reflection on how the synergy between state and society can be beneficial for both parties, especially as top-down and bottom-up efforts converge in a relationship of embeddedness in which “networks of trust and collaboration that are created span the public/private boundary and bind state and civil society together. Social capital inheres, not just in civil society, but in an enduring set of relationships that spans the public-private divide” (ibid.: 1122) The SCAU successfully leveraged its capacity to collect intelligence both in Uganda and in Somalia on potential threats to Ugandan civilian and military targets Members of the community were regularly employed as consultants for ‘cultural awareness’ training courses, funded by the US Embassy and aimed at UPDF troops being deployed in Somalia. Additionally, Uganda has been able to leverage its leading role in the mission at an international level. For instance, in November 2012, Uganda threatened to pull its troops out of Somalia in retaliation for a UN report that claimed that its government was supporting the M23 rebel group in the DRC. In the following days, the Ugandan media reported the concern of Somali refugees in the country — particularly those affiliated to the SCAU — over this possibility. As the diplomatic 26 http://amisom-au.org/uganda-updf/ 204 outcry over the report abated, Uganda continued to project its influence over the international mission in Somalia. On 15 May 2014, a United Nations Guard Unit (UNGU), formed by 410 UPDF soldiers, was deployed for the first time to protect UN personnel and installations in Mogadishu.27 At the same time, the edgy relation between Kenya and Somalia has increasingly made Uganda a privileged regional partner of Hassan Sheikh Mohamud's government, as became clear on 12 May 2014 when a large Somali delegation, including the President, the Minister of Defence and other top generals, travelled to Entebbe upon invitation of Ugandan President Yuweri Museveni. The Somali population is largely perceived—by Ugandans and Somalis alike—as being under the personal protection of President Museveni and therefore its security is seen as inextricably bound to the President's or to his son—often tipped as a likely successor to his father—Muhoozi Kaneirugaba’s political future. This belief is a source of anxiety for many long-term members of the Somali diaspora, who admit having feared the end of the Museveni era during the 2011 elections. This is also a reminder that the current Ugandan sanctuary could come to an abrupt end. Aware of this possibility, the SCAU has been lobbying since 2007 to have Somalis recognised in the Constitution as a Ugandan tribe, supporting this claim with the argument that Somalis have been living in the Ugandan territory for a century. On top of that, a growing number of refugees, backed by the advocacy organisation Refugee Law Project, are seeking to become Ugandan citizens by challenging the Department of Immigration in court. They demand the implementation of a provision in the 2006 Refugee Act which grants Ugandan citizenship to those who fulfill a specific set of requirements.28 Conclusions This chapter has traced the emergence of a migrant and trading hub in Uganda in the light of recent developments at the regional level. While foregrounding the 27 http://www.un.org/apps/news/story.asp?NewsID=47820#.U3seAfmSySo 28 “16.(5) The qualifications for naturalisation are that he or she-- (a) has resided in Uganda for an aggregate period of twenty years; (b) has resided in Uganda throughout the period of twenty-four months immediately preceding the date of application; (c) has adequate knowledge of a prescribed vernacular language or of the English language; (d) is of a good character; and (e) intends, if naturalised, to continue to reside permanently in Uganda (Uganda Citizenship and Immigration Control Act” (Walker, 2007). 205 importance of agency, which is often overlooked in discussions which regularly dehumanize refugees and deny their role as ‘purposive actors’ (Turton 2003: 12), the situation here described illustrates the complexity of migratory flows, which change across time and space and are shaped by endogenous and contextual factors. As also suggested by other studies (Hovil 2010; Kaiser 2010; Long 2010), even in protracted refugee situation there is not a sharp dichotomy between return and exile. Refugees carve their path according to the repertoire of available options. The opportunity structure is crucial to diversify assets, as I have illustrated when discussing the importance that young Somali attach to pursue a higher education degree, but also to participate in the political activities of university student associations. The next chapter brings this to the fore, as it examines the political economy of a humanitarian context of reception. 206 7 Power and Class in a Humanitarian space This chapter describes how the issues of domination and subordination, power and marginality that I have explored in the rest of the thesis are also blatantly inscribed in a context of reception whose specific administrative structure is shaped by the interaction of Kenyan and international humanitarian policies. In Kakuma refugee camp, a humanitarian complex in Northern Kenya’s Turkana district, the threads running through my thesis are drawn together. Here I tease out the dynamics of inclusion and exclusion that distinguish a Somali elite made mostly by members of the Darood Ogaden clan, who have greater access to international remittances and institutional actors, from a Jareer population, diverse by background yet lumped together in a category, Somali Bantu, constructed by the humanitarian apparatus, and located on the physical and institutional margins of the camp. The issues examined in the previous chapters are brought to the fore by the ‘power geometry’ of the camp (Massey, 1992). In Massey’s words, this concept refers to space’s intrinsic “nature full of power and symbolism, a complex web of relations of domination and subordination, of solidarity and cooperation” (264-265). In Kakuma refugee camp, Somali Jareer are the largest segment of a broader, and diverse, Somali refugee population. As a form of 'urbanisation in the making' (Perouse de Montclos and Kangwanja, 2000), the refugee camp displays processes of social differentiation that unfold in a space in which pre-existing power relations are reproduced. However, also the humanitarian apparatus actively creates conditions of access and marginalisation through spatial arrangements, such as the proximity or remoteness to major infrastructures and centres of governance, in particular the UNHCR compound and the headquarters of the humanitarian agencies operating in the camp, the access to humanitarian aid and the possibility to diversify resources. 207 Arriving in Kakuma At the beginning of November 2013, a shared taxi packed with seven passengers and overloaded with boxes, jerrycans and yuta bags dropped me at the entrance of Kakuma town in Turkana county, north-western Kenya. The car stopped at a lay-by beside an airstrip upon which a small aircraft with the insignia of the World Food Program (WFP) was parked, its white paint glaring in the sun. That, and the better quality of the tarmac on that stretch of road, signalled the presence of humanitarian facilities nearby. A Turkana pastoralist leant against the net surrounding the airfield, and held the front legs of a goat in one hand. More goats dotted the barren landscape, trying to find some green in the endless khaki expanse. I had just arrived from Lodwar, the regional capital, and had travelled on the road that bisected the town and continued to the South Sudanese border, 100 km away, and led further on, to the capital of South Sudan, Juba. Along the road, heavy trucks sputtered along, in contrast to the tidy white humanitarian vehicles that silently cruised past. The refugee camp was still out of sight, one kilometre away, but its tall communication tower was visible, soaring high above the plains. In front of me, Kakuma town began and ended: a handful of low buildings lined up along the road from which the agglomerate derived its very own raison d’etre. I was supposed to call my host organisation as soon as I arrived. I had the contact number of the local coordinator of Refugee United (RU), a Danish NGO. RU runs a web and mobile-based platform to help refugees and displaced people in 11 countries (mostly in Central and Eastern Africa, the Middle East and the Philippines) track lost relatives. I discovered they had a presence in Kakuma refugee camp and, since I needed the support of an organisation to gain access to the field, I visited their Nairobi office and explained my research project and my plan to do fieldwork in the humanitarian complex. Eventually, they agreed to facilitate my access to the camps. However, at the moment of my arrival, I had nothing but a phone number. On top of that, early that morning in Lodwar, I had realised, to my dismay, that three hundred dollars I kept in the hidden pocket of a belt were missing. As my backpack was already in the trunk of the taxi, 208 buried under bags of pasta ‘made in Turkey’, I convinced the taxi driver to take me to an ATM, which he did. Indeed, it was the only ATM in Lodwar. And it was out of service. My only hope was to travel to a refugee camp, and find a bank there, a remote possibility. But I did not need a bank, I needed cash. While the taxi speedily left the city – the driver wanted to prevent me from changing my mind – I called a Somali friend in Eastleigh, and simply asked: “Is there a hawala in Kakuma?” “Of course”, he answered. Only four hours later, as I stared at the goat whose front legs were clamped in the grip of the shepherd, and she stared back in return, I recalled the number of times during fieldwork on which I had been given affirmative answers out of politeness. The problem, though, was mine. I had always asked the wrong question. It was quite likely that there was a hawala in the camp. According to what I had learned so far, from my readings and my own interviews, there had to be a hawala for the refugees to receive remittances from abroad. The question I failed to ask was whether I would be able to use it to receive the money I needed. I sidelined the worry as I took a piki piki to reach the gate of the UNHCR compound, where the RU coordinator, a Colombian man, waited for me. The compound lays near the dry bed of the Tarach river. In sharp contrast to the parched outside, strips of lawn surrounded the residential blocks where the humanitarian personnel was housed. To prevent me from imagining that the coming weeks might be spent in one of these air-conditioned houses, the RU coordinator immediately informed me that there was no place for me in the compound. “Anyway, here there are plenty of snakes”, he said, as if to alleviate my disappointment. Instead, he directed me to a guesthouse in town. He added that he would introduce me to their field worker and then leave for holidays. As for my most pressing concern – a money point to cash out – he could not be of any help. The day after, we met early at the UNHCR compound to set off on a field trip to Kakuma 3, the sector hosting the Somali Jareer population of the camp. We drove past Kakuma 1, the oldest sector of the refugee camp, adjacent to the compound and stretching along the bed of the river. Despite being the most diverse area of the humanitarian complex, it was tellingly called Mogadishu. It was the economic core of 209 Kakuma, both the refugee camp and the city, its significance radiating out across the entire area. Few shade trees stood out above the low housing units, benefitting from the underground moisture, but as we diverted towards Kakuma 2 and then Kakuma 3 the trees increasingly gave way to the shrubbery. The empty landscape was sporadically crossed by piki piki and lorries, the only means of transport for the refugees to travel between camps and get to the UNHCR compound. It took over forty minutes to reach the centre of Kakuma 3, the WFP distribution centre, and as we drove we passed stacks of sun-baked mud bricks on the sides of the dirt track. The plan of the camp was a grid, with wide streets separating the blocks. The houses facing the streets were made of bricks, while those inside the quarters were often simple shacks of a variety of assembled materials, including plastic sheeting and tents with the UNHCR logo. The shades barely mitigated the heat. I was introduced to the local Somali Jareer leaders, some of whom I had already met in Eastleigh, and to Hussein, the RU outreach volunteer, a 25-year old refugee from Lower Juba who would work as my assistant in the coming weeks. After greeting the men and explaining the terms of my presence there, I sat down with Hussein to draw out a working plan. I told him of my financial concerns. Hussein acknowledged the existence of a hawala in Kakuma 1, but remained vague and was clearly reluctant to get involved in that specific matter. Instead, he suggested resorting to a Western Union agency in town or to one of the many M-Pesa outlets scattered across the complex. His elusiveness surprised me. Later, I would recognize in his disinclination the awareness that Kakuma 1 and Kakuma 3, and their respective Somali populations, followed two different trajectories. The gap separating Kakuma 3 from the commercial hub in Kakuma 1 was not only evident in the physical distance between the two sectors, and in immediately visible differences, such as the isolation of the former as opposed to the latter’s proximity to the main road and the UNHCR compound, and small but telling signifiers of better living conditions, such as shade trees. There were also more subtle reasons for his unease. For the moment, I decided to handle the issue myself and, back to the main compound, I ventured through Kakuma 1 on a piki piki and then by foot. The camp was divided into sections, its perimeters marked by thorn bush fences connected by gates. The humanitarian workers 210 had previously pointed out that each section matched a specific refugee population, but, as I shall explain later, the reality on the ground was more fluid and complex. Although I was told by humanitarian workers that the only authorised jobs were with the INGOs operating in the camp, and that the refugee employees were given ‘incentives’ rather than a wage (Pavanello, Elhawary & Pantuliano, 2010), along the main street of Mogadishu several stalls displayed a great variety of goods, ranging from canned and packed food to solar lamps and electronic material. A couple of shops provided access to the internet through old PCs, and power sockets to recharge mobile phones. Refugees idled on the side of the street, while Turkana women with children squatted in the shade. The humanitarian logos were ubiquitous on bags and dustsheets. Asking around, I came across a piki piki driver who was willing to take me to the local hawala office. We rode through a maze of narrow alleys, moving down inside the camp. A tall wall of corrugated iron eventually opened on a court. At the centre sat a brick construction with metallic grids to the windows and a spacious porch where five women donning niqab stood. I drew their puzzled glances, as well as that of a man holding a club which he discreetly swung behind his back. They were all Somali, as was the man sitting in the bare-walled room inside. I asked whether I would be able to receive some money from the UK in his shop. He was unperturbed by my question and my presence. He simply nodded and handed me the details of the transaction. Only then I realised that it was a Dahabshiil operator. I called a friend in Bristol and explained the situation. It took him 15 minutes to locate a Dahabshiil branch in Saint Pauls, a mostly African and Caribbean suburb of the city, and one hour to travel there and send two hundred dollars via the details I had shared with him through SMS. But it took only five minutes for the money to travel between England and that spot in the desolated North Western Kenya where I was sitting. The hawaladar summoned me to his office and, after I produced my passport and he verified my identity, he handed me the amount already converted in KSh. I was relieved: I had money. But I was also incautious, and made a false step. Abruptly, I switched my role from customer to researcher. I revealed the reason of my presence in Kakuma and asked for an interview. He was set aback by my question and became wary. He kindly asked me to wait in the office and left the room. I realised my 211 mistake, and increasingly so when he returned with another man. The newcomer wore a green uniform that was not a proper military fatigue but nonetheless an outfit that commanded authority. He introduced himself as the camp manager, one mister Abdi, and demanded I produce my passport. I also handed him a badge I had been given by RU, which he barely looked at. I had assumed that my host organisation had arranged all my clearances, which it had, but with the UNHCR. Mister Abdi was an officer for the Department of Refugee Affairs and was visibly annoyed that I had not been officially introduced to him. Akin to the police officer who had questioned me in Kampala, he adduced the recent Westgate attack and the security concerns looming over Kenya. He finally handed back my passport and requested I pay him a visit in his office, accompanied by an RU representative, before I carried out any investigation. After leaving the hawala office, I reported what had happened to Hussein. He replied with increasingly alarmed SMSs, particularly when I asked him further information about Mr. Abdi. He is the Camp manager. Ogaden Somali Kenyan who hv attachments with very stuff that happens in the camp. from officer, business pple and lowest gruop in the camp (sic) Another SMS followed suit. It spoke volume of his anxiety for my unexpected encounter: i fear now fr my lyf man. insecure here. Talk later. I think important to finish up with hm (sic). I felt guilty, although I did not know exactly why. Later, once Hussein’s panic had abated, I read my encounter with Mister Abdi not so much as an impingement, but as a preliminary sketch of the a map of governing authorities and power relations that overlayed the official plans of the camp provided by the UNHCR. However, in the following days, I focused my efforts on making myself recognizable in an unambiguous way. I paid a visit to mister Abdi’s office, accompanied by an RU Kenyan staff member. He was absent, but his secretary duly noted that I had abided by the order, and this seemed to satisfy him. As long as I had acknowledged his authority over the camp, I had earned preliminary clearance. Moreover, during my stay, he often phoned to inquire 212 about my progress, and even accepted a short interview. Upon Hussein’s request, I carefully expunged any reference to the terms Somali Jareer or Bantu from the cursory description of my research project that I provided to Mr Abdi. Instead, I couched my interest on livelihoods and trading dynamics within Kakuma in the humanitarian and development jargon, resorting to terms such as ‘empowerment’ or ‘resilience’, and emphasising the value of solidarity among Somalis. Interestingly, his concerns were all focused on Kakuma 1. He was informed of my travels to Kakuma 3, yet these did not seem to draw his interest. I wondered why. I concluded that as long as I stuck to the official cartography and nomenclature of the humanitarian complex I would have no problems with Mister Abdi. The UNHCR maps had clean cut demarcations between the camps and the outside. However, while they mapped the humanitarian facilities, and the main access roads, they eclipsed the complex grassroots architecture within the camp: the narrow alleys, the thorn bush fences, the grid and the iron walls that created real gated communities within the humanitarian spaces. History and topography of Kakuma Kakuma refugee camp sits on an arid plain surrounded by low sun-scorched hills, off the road linking Kitale, that runs through Lodwar and Lokichoggio, to Nakodok, on the South Sudanese border, and continues to the South Sudanese capital Juba. It is jointly managed by the UNHCR and the Camp Manager’s Office of the Ministry of Refugee Affairs of the Government of Kenya (GOK). The relief operations are provided by the WFP and the Lutheran World Federation (LWF), in charge of the distribution of food rations; the International Rescue Committee, which provides health services; the National Council of Churches of Kenya, in charge of housing; the International Organisation for Migration, tasked with the resettlements; and the Jesuit Refugees Service, which focuses on education, care for people with special needs and assistance to survivors of sexual and gender-based violence. UNHCR, WFP and LWF manage one compound each, all located in the proximity of Kakuma 1. 213 In November 2014, the complex hosted approximately 180,000 refugees from 20 different countries in four camps: Kakuma 1, the first camp to be established, Kakuma 2, Kakuma 3 and Kakuma 4 (UNHCR, 2014). The latter was set up in late December 2013, after my fieldwork, to accommodate South Sudanese displaced by the outbreak of the civil war. Indeed, since then, South Sudanese refugees have outnumbered the Somalis, until this time the largest population in the camp. Most refugees arrive from South Sudan (86,800), Somalia (55,825), Sudan (9,150) and DRC (8,800). The entrance of Kakuma 1 is located 1.2 km from Kakuma town, the largest urban centre in northern Turkana county and the poorest district in the country according to the Kenyan Commission on Revenue Allocation (CRA)29. It has a population of 31,962 in the urban core and 33,852 in peri-urban and rural areas. The town is a place of transit for long-haul transports (Oka, 2004) and consists of houses and shops mostly clustered along a 100- metre stretch of road. The businesses facing the road include a couple of mechanics garages, a retailer of spare parts, a few warehouses and M-Pesa points, a Western Union agency, and a couple of ‘hotels’, small restaurants with an annexed butchery luring passing truck drivers with the promise of goat meat, cold beer and the company of sex workers. The few secondary roads departing from the main one are all unpaved and wind through shabby wooden shacks, a few houses built of concrete (which, according to local informants, comes from the refugee camp), and some churches from different Christian faiths. Other houses are scattered across the barren landscape, their corrugated iron roofs shining in the dazzling sunlight. Turkana pastoralists strolling along the road, leading herds of goats or squatting in the shade of parked trucks, are a common sight. They mostly come to town to sell their livestock to the local slaughterhouse and buy basic goods. All businesses are managed by Kenyans from other regions of the country. One warehouse is owned by a Kenyan Somali businessman but most workers are Kenyan. However, some shops sporadically hire refugees for menial jobs and one restaurant employs refugees from DRC as cooks and waitresses. The proximity to the 29 http://siteresources.worldbank.org/INTAFRICA/Resources/257994- 1335471959878/Kenya_County_Fact_Sheets_Dec2011.pdf, accessed on 14 May 2015 214 South Sudanese border has been the main economic asset of Kakuma town since the 1960s, when some pioneering businessmen established a trading post among impoverished nomadic Turkana, catering to truck drivers travelling between Kenya and Sudan (Crisp, 1999; Jok and Hutchinson, 1999; Verdirame, 1998; Grayson, 2014). Among these, there were Somali traders, such as Abdullahi Ogle, a Somali Isaaq who opened a store along the road in 1968, trading cattle for manufactured products during periods of pastoralist transhumance (Pérouse de Monclos and Kagwanja, 2000). Prior to the establishment of the camp in the early 1990s, these traders had consolidated their position as major suppliers of goods and ancillary services for the entire area, whose population had become increasingly diverse as a result of the mounting conflict in Southern Sudan. The launch in 1989 of the UN Operation Lifeline Sudan (OLS) to deliver basic aid across the border turned the route between Nairobi and Lokichoggio into a crucial humanitarian transport corridor. Their clientele of truck drivers and relief workers rapidly expanded after 1992, when the UNHCR set up facilities to host about 17,000 young men displaced from the conflict between the Sudanese government and the Sudan People’s Liberation Army (SPLA) in Southern Sudan, (Jok and Hutchinson, 1999; Otha, 2005; Verdirame, 1998; Grayson, 2014). In the following years, the camp grew to shelter a growing refugee population, becoming increasingly diverse, and assuming urban features. In 1996 there was a single camp made of 76 blocks. A second camp was added shortly after (Pérouse de Monclos and Kagwanja, 2000). In the course of the 1990s, Sudanese refugees were joined by refugees from Ethiopia, Rwanda, Burundi, DRC, Eritrea (since 1994) and Somalia. Most of the latter were transferred from the camps on the coast closed down by the Kenyan government between 1995 and 1999 (Jamal, 2000; RCK, 2012; Pérouse de Montclos and Kagwanja, 2000; Verdirame, 1998). In the mid-1990s, for reasons I shall explain below, the issue of discrimination and violence against Somali minorities in Dadaab entered the humanitarian agenda and Kakuma was designed as a sort of sanctuary within a sanctuary to host several thousand Somali refugees from groups, mostly Jareer, Madiban, Benadiri or Ashraf, those deemed more vulnerable. The aggregation of these groups in the Bantu category has been crucial in shaping space and internal dynamics of the refugee complex. In 2002, 12,000 Somali 215 Bantu were relocated from Dadaab by the UNHCR and the IOM ahead of a planned resettlement to the US (IRIN, 2002; RCK, 2012; USCR, 2004; Grayson, 2014). In 2003, another 15,000 Somalis Bantu were fast-tracked to resettlement to the US. In 2006, a second wave of resettlement to the US targeted refugees who had lived in the camp for a long time. At the same time, as Somalia entered a new phase of the conflict, there were new arrivals from Dadaab. Since then, most refugees were periodically relocated by the UNHCR to decongest Dadaab and occupy the places left vacant by those who had been resettled. 13,124 refugees were transferred in 2009 in multiple rounds of a Relocation exercise (Kanere, 2009). Also, many small groups of refugees arrived on their own from Dadaab or directly from Somalia, drawn by Kakuma’s reputation as a gateway to resettlement and a humanitarian shelter with higher standards of living than in Dadaab (Grayson, 2014). In the same years, following the signing of the Comprehensive Peace Agreement between the SPLA and the Government of Sudan in 2005, around 38,000 Sudanese began returning to Sudan (ibid.). Between 2006 and 2009, the balance of the refugee population tipped from being predominantly Sudanese to being mostly Somali30. A new batch of 2000 refugees was relocated from Dadaab between 2012 and 2013. Following the relocation directive of the Kenyan government, small groups of refugees were transferred by the DRA. Finally, with the launch of the operation Usalama Watch, between April and June 2014, 3,771 refugees were relocated from urban centres and Dadaab (UNHCR, 2014). At the same time, a significant slowdown in resettlement processing, coupled with the massive inflow of South Sudanese refugees displaced by a fighting surge after December 2013, led to a peak of over 180,000 registered refugees in June 2015. Tallying the exact number of refugees living in the camp at any given moment is a daunting task. Yet the official process of identification of the Somali refugees also reveals all the challenges humanitarian agencies face in making populations of concern readable. To give an idea, according to the data I obtained from 30 According to UNHCR figures, at the beginning of 2006, Sudanese refugees were 74,000, 80 % of Kakuma’s population, while Somalis were 12,600 or 14% of the population. By the end of 2009, the Sudanese population had shrunk to 17,000 (27% of the population), while the Somali one had risen to 38,000, or 60% of the overall refugee population (Grayson, 2014) 216 the Kakuma’s registration office, there were 53,890 Somalis (26,609 female and 27,281 male) in November 2013. According to a breakdown ‘by ethnicity’ [sic] of the Somali population, Somali Bantu are the largest population (13,711 units), followed by Darood (10,728) and Hawiye (9,350). However, the breakdown also includes puzzling data, such as one ‘Armenian Somali’, four ‘Hutu Somali’, one ‘Macedonian Somali’, three ‘Romanian Somali’ and one ‘Turk Somali’. In its sloppiness and redundancy, the UNHRC dataset interestingly tells the limits of a humanitarian classificatory device designed to put order, or to make legible, its population of concern. This format is indeed unable to capture a reality that, despite the assumption of limited mobility encapsulated in the very function of the refugee camp as a space of containment and segregation, allows a great degree of internal and external mobility. On the one hand, the figures are invalidated by the fact that, as already discussed in the previous chapter, an undetermined number of refugees are registered in the camp but live in urban sites in Kenya or elsewhere, in order to remain recipients of food aid. On the other, ethnic labels are problematic since they either essentialize and institutionalize categories that are in fact elastic, or are used instrumentally, as I go on to explain. The ethnic lens applied by the UNHCR to the refugee population has repercussions on the organisation of the space of the camp and on the way humanitarian governmentality is deployed. The concept of governmentality is a central tenet of Michel Foucault’s theoretical architecture devoted to the technologies of rule and control enacted by the state and other institutions engaged in practices of biopower (Foucault, 2007). One of the definitions of the term refers to the ensemble formed by institutions, procedures, analyses and reflections, calculations, and tactics that allow the exercise of this very specific, albeit very complex, power that has the population as its target, political economy as its major form of knowledge, and apparatuses of security as its essential technical instrument. (Foucault, 2007: 144) Foucault’s work has gained large currency in the literature on humanitarianism as an apparatus aimed at control and assistance through rules, practices and categories (Malkki, 2002; Fassin, 2007; Rabinow and Rose, 2006). Produced as epistemological tools to read the empirical reality, these categories do not account for the notion of class, 217 and for the processes of class formation that take place within the camp. These dynamics emerge clearly as we consider the intertwining of security and economic development. In a report for the UNHCR in the late 1990s, Crisp (1999) wrote that the level of violence in Kakuma was ‘unacceptably high’ (16). As noted by several scholars and humanitarian rapporteurs, brawls and clashes were frequent between different national groups and between different ethnic groups from the same countries (Crisp, 1999; Jamal, 2000; United Nations General Assembly, 2001; Verdirame, 1998). Particularly violent were the fights between Somalis and Sudanese. In a clash that erupted in Kakuma 1 in September 1998, two Somalis were killed and over 50 houses and shops were burned down (Crisp, 1999). In 2000, more than 700 houses were destroyed in four arson attacks in the same area (USCR, 2001). On top of that, there were several reports of sexual violence perpetrated by Turkana pastoralists on Somali women who had ventured outside the camp to collect firewood. These tensions between refugees and hosts often featured in Somali accounts. The “geographical concentration of the violence” (Crisp, 1999: 19) was triggered by the unbalance between the living conditions within and without the refugee camp, where the UNHCR’s population of concern, unlike impoverished Kenyan pastoralists in the area, enjoyed access to basic services. As Crisp (1999) points out, “there are simply more items to steal, more people to rob and more women to rape in and around the camps than in other parts of the two provinces” (19). Moreover, among humanitarian workers and camp residents there is a widespread awareness that, notwithstanding the effective risks of assault, in many instances allegations of sexual violence have been instrumental in advancing protection claims (Jansen, 2008). Several scholars agree that improved security and expansion of trading and social relations within the camp are interlocked (Grayson, 2014; Otha, 2005). In general, peaks in violence coincide with massive population reshuffles. For instance, Otha (2005) suggests that the local Turkana pastoralists intensified robberies against refugees in coincidence with the return of thousands of Sudanese as a way of compensating for the loss of income from the sale of goats. Most studies on Kakuma suggest that decreases in violence are matched by increases in the remittances that, since 218 2006, had started flowing into the camp. Grayson (2014) quotes a Somali refugee saying that “if something was going on within Mogadishu, people would fight here. It stopped in 2001. People realised they had the same interests and problems. That is when they started saying that they all were brothers and sisters and had to help one another” (117). Grayson also objects to Hyndman’s idea that a refugee camp is not a community, despite temporal convergences of interests and ephemeral cooperation grounded in nationality, ethnicity or neighbourliness (Hyndman, 2000). Instead, she argues that while the lack of voluntary association hinders the possibility of the camp becoming a whole community, “multiple communities are likely to come to co-exist in such a setting, probably not from the onset, but as time passes and people develop social ties, share experiences and a common space” (Grayson, 2014: 111). This suggests that security in the camp had improved over time as commerce had gained pace and different communities had developed weak ties. However, many interviewees (Somali, Ethiopian and Congolese refugees, and humanitarian workers) suggested that, towards the mid- 2000s, security improved in some areas of the camp but not in others, as an influx of money coming from outside (including Somalia) redesigned the topography of the camp. Not only physical barriers were erected, but money became available to some businesspeople to hire private security to protect their businesses. Violence did not disappear altogether, but became more localised. From the second half of the 2000s on, violence erupted mostly on Kakuma 2 and 3, leaving Kakuma 1 unscathed. The latest riots occurred in 2014, when a fresh wave of refugees from South Sudan arrived in the recently established Kakuma 4 and clashes erupted between members of the Dinka and Nuer communities, whose factions were fighting in the home country (UNHCR, 2014). In the course of my interviews, many residents of Kakuma 3 (mostly Somali Bantu) claimed that brawls, rapes and murders, often escalated from petty thefts, have continued despite the improvement of living conditions in Kakuma 1. Yet, this violence went almost completely unnoticed since the distance between the camp and the closest UNHCR compound (and the cost of the journey) often discouraged victims from reporting the facts to the protection officers (unless the case was solid enough to push 219 forward an application for resettlement). Recounting how a class of refugee entrepreneurs emerged in Kakuma thus entails the need to understand how goods and money started flowing in and still do; where they flow and where they do not. Social differentiation, space, economy, security According to an entrenched myth among policymakers and the general public, refugees are an economic burden to host states (Smith, 2004, 46; Kibread, 1989). Alongside security concerns, encampment policies are therefore framed as a solution to mitigate negative repercussions on the host society. However, a number of empirical studies dispel the assumption that refugees are passive recipients of aid and suggest instead that, even under duress and in humanitarian camps, refugees strive for agency in order to pursue livelihoods (Betts, 2012; Dryden-Peterson, 2006; Grabska, 2006; Zetter, 2012). In her study of Somali refugees in Cairo, for instance, Grabska (2006) points out that, despite the “popular image of refugees as passive, dependent and an economic burden for the host society [...], refugees are active agents making an economic and social contribution.” (289). While, as I have argued so far, the extent to which refugees can shape their own life and career trajectories varies according to the resources they are able to mobilize, particularly in a humanitarian space, the economic dynamism of Kakuma refugee complex and its positive impact on the region is immediately noticeable. Oka (2014) has calculated that the refugee complex generates a monthly turnover of USD 350,000–USD 400,000 for local wholesalers, while negatively affecting Kenyan retailers. The establishment of the camp in 1992 drew entrepreneurs from other regions of Kenya, who opened shops. However, with the passing of time, they were driven out of business by the competition with traders within the camp, who do not pay rent and a trading licence of KSh 1,970 per year which applies to traders in Kakuma town, and, as I explain below, have easier access to food rations re-sold by the refugees. The few retail shops still operating in Kakuma town are supplied by the local wholesalers and by traders within the camp (ibid.). Despite the very concept of encampment implies physical and institutional segregation, the boundaries of Kakuma 220 refugee complex ebb and flow, revealing de facto continuity with the surrounding area. The area known as Mogadishu, in Kakuma 1, is the economic hub of the entire northern Turkana. Linkages to external trading networks are crucial to overcoming the limits of the humanitarian assistance and foreground the significance of refugee camps, particularly in situations of protracted displacement, not just as sanctuaries but also as spaces of consumption. The importance of consumption patterns and materiality of daily experience in contexts defined by “the permanence of temporariness” (Brun and Fabos, 2015: 6), such as a refugee camp, is seldom accounted for in studies on refugees, where the emphasis is mostly on the dimension of coping at the level of basic needs. However, as Dudley (2010) remarks in her study on humanitarian camps for Karenni refugees on the Thai-Burma border, the material dimensions of everyday life (objects, devices, food) are crucial to recreate a sense of “being at home”. Over the years, a variety of shops sprouted in Kakuma’s commercial district. In present times, the offer of goods and service ranges from first and second-hand clothes to mobile phones and electronic shops, to TV booths where one can watch football matches, and drinking dens where men, mostly from South Sudan and Ethiopia, consume changaa31. Among Somali refugees, products that are particularly sought-after are pasta, meat and tea, not provided by the WFP, yet frequently described as symbols of the Somali food culture32. The meat is purchased from Turkana pastoralists, who trade their goats for rice and flour. The demand for meat is particularly high ahead of religious celebrations and is often fulfilled by resorting to a creditor in exchange for work. This aspect points to a crucial transformation that occurred gradually in the camp and has been accentuated through the entrenchment of positions of dominance: a process of social differentiation along class rather than ethnic lines. As a city in the making, Kakuma has seen the gradual 31 Changaa is a home-brewed spirit distilled from different grains which is particularly harmful because of the frequent addition of substances such as gasoline or battery acid. 32 Until November 2014, the recommended emergency ration provided by the WFP was 2,100 kilocalories per person per day and included cereals, pulse, vegetable oil, a nutrient-rich maize-soya blend and salt. However, on mid-November, rations were cut by 50 percent, to 1,050 kilocalories per day, due to donors fatigue (Kanere 2014) 221 emergence of an elite. Agier (2002) argues that “the humanitarian device of the camps produces cities, ‘de la ville’, if one considers the city from the point of view of its essential complexity” (32). The “sketches of a symbolic of space, social differentiation and identify change” (ibid.) surfacing in Kakuma refugee complex points to a tension between the homologation practices of the humanitarian apparatus and the “heterogeneous ensemble of identity resources” (ibid.). Although a few M-Pesa kiosks and small retailers are present in each camp, shops are concentrated in the Mogadishu area of Kakuma 1 and are owned by Somalis and, in a smaller part, Ethiopians. As observed by Pérouse de Montclos and Mwangi Kagwanja (2000) writing on Sudanese refugees arrived in Kakuma in the late 1990s: “these Sudanese had arrived from Ethiopia with nothing. Unlike the Ethiopians and Somalis, they did not have capital or a diaspora to rely on to start a business” (213). As I go on to show, this argument also applies to the gap between Somalis from dominant groups and Jareer. However, with the passing of time, the number and the diversity of Somali and Ethiopian-owned businesses has increased. Among these, there is for instance an Ethiopian restaurant in Kakuma 1 which, on Sundays, also draws humanitarian staff as patrons. In general, livelihoods established by tapping into pre-existing networks (refugees’ legacies) enable people to create linkages with members of other networks and thus increase the level of agency they have in a humanitarian space, a locality that, while hindering some group’s room for manoeuvre, carves out fresh opportunities for others. Location within the camp is a crucial asset. The commercial hub in Mogadishu was shaped through the gradual taking over and trading of plots, which re-arranged the refugee complex’s landscape. Although plots are assigned by the UNHCR on a first come first served basis, informal commercialisation reconfigured the space of the camp. The area from which Mogadishu emerged is still particularly sought after: its proximity to the UNHCR compound allows greater access to the humanitarian personnel and therefore greater work opportunities (Grayson, 2014). At the same time, the proximity to the main road facilitates the movement of people and the shipping of goods. As I have already mentioned, the hardships derived from the dry local climate are mitigated in Kakuma 1 by the presence of greenery, despite the risk of occasional flash floods during the wet season (however, 222 the shopping area of Kakuma 1 is located at a safe distance from the river bed). The trade of plots has enabled refugees to cluster, mainly in order to fulfil security needs. In most cases, the contours of these clusters were defined by the basic source of collective identification, such as clanship or origin. However, the reconfiguration of the humanitarian space, achieved through the mobilisation of financial resources, led to the emergence of new communities sharing common interests. In the commercial district of Kakuma 1, the time-specific opportunity structure created the conditions for a trading elite of refugees from Somali Darood Ogaden and Hawiye clans, and, to a lesser extent, Ethiopians, to gradually emerge and coalesce. While in Somalia members of Darood and Hawiye clans vied for control over humanitarian aid, in the camp they pragmatically set aside their rivalry and shared the cost of common security to protect their commercial establishments. As the camp population grew, and the resettlements slowed down, the demand for consumption goods increased. Innovations in technology and finance, as mentioned in chapter 5, enhanced the possibility of communicating and sending money to refugees in the camp. A crucial witness of the changes that occurred over the past 15 years was Barre, a Somali Darood trader who was unanimously referred to by my informants in Kakuma as the wealthiest man in the refugee complex. I met him on the margins of a meeting of Somali elders outside a building that acted as a community centre in Kakuma 1. Barre owned a car: this was the first thing I heard about him. And this information was confirmed by many refugees. As a matter of fact, he owned several cars: his personal means of transport and a fleet of at least five taxis. He also owned an undisclosed number of moto taxi, and a large shop in the Mogadishu area of Kakuma 1, where he sold electronics and other goods delivered twice per week from Kisumu and Eldoret. He was also a business partner of the only hawaladdar in the complex, the Dahabshiil agent I had resorted to. He remained vague about the terms of his involvement in the hawala business, but the convergence of his trading and money transfer activities seemed consistent with a business model I had already observed in Eastleigh and Kampala, where large traders and hawaladar's coincide. However, Barre was very elusive when it came to detailing the nature of his businesses. But he explained with pride that, since he had first arrived in Kakuma in 1999, he had learned Turkana and was therefore able to deal directly with the locals. He said that he often traded goods from his shop for meat and then sold this to other refugees. He had a wife and sons living in Eastleigh. He specified that had never lived in Dadaab, and mentioned that he came generically from Southern Somalia. In Kakuma, he partnered with Abdullahi Ogle, the Isaaq trader who first opened a shop in Kakuma town in 1968, and, once he had purchased his first car, he 223 started selling items within the camp. He confirmed that, with the growth of the Somali population of the camp after 2006, the demand for consumer goods increased as did the need to protect the items and the cash earned from their sale. He hired people for building warehouses and security. I was aware that he had a good relationship with the camp manager (some people told me that he was ‘in business with him’, whatever that meant), but I decided to skip any questions about Mr. Abdi. His name suggested ties with the Darood clan of the late Somali dictator Siyaad Barre, but when I casually asked about his genealogy , he dismissed my question with a gesture and, politely but resolutely, said that it did not matter. My chat with Barre was ridden with gaps. Despite officially being a refugee, he seemed to consider Kakuma less a shelter than a trading post. It dawned on me at one point that he could be a Somali Kenyan and that, in order to establish his business in the middle of the refugee camp, had embraced the status of refugee. However, his capacity to mobilize resources was crucial to his movement between categories. The widespread fascination with the fact that he owned a car was particularly interesting. While his private car was a powerful signifier of agency for other refugees whose mobility was limited by financial constraints, his dominant position in the camp (at an economic and, as I will explain, political level) derived from and, at the same time, was expressed through his control over mobility inside and outside the camp. As previously observed, in his case the convergence of trade and logistics enabled him to gain a position of dominance over competitors. Even as Barre dismissed my questions about his genealogy, Somali Bantu attributed his proximity to the camp manager and his business partnership with the only hawaladdar in the camp to the fact that they belonged to the same clan, Darood Ogaden. Barre deployed his entrepreneurial talent not only by leveraging structured networks, but also by weaving personal networks involving refugees from other communities and Kenyans in the outskirts of the camp, connecting for instance the Turkana pastoral economy to the camp economy. At the same time, his connections to urban centres across Kenya enabled him to access supplies to sell in the camp. His reputation as the main private employer in a context dominated by widespread joblessness was unanimous. His core businesses were in local transport and trade. He employed at least 40 people as drivers of his fleet of moto-taxis, mostly from Kakuma 3. According to many informants, Barre was also among very few traders in Kakuma 1, all 224 Somalis, who purchased foodstuff from refugees and shipped them outside the camp. This issue deserves specific attention, since it casts a light on the political economy of the humanitarian space. Biometric systems and foodstuff sale I met Barre outside the venue where a meeting among elders had just been held. He was an authoritative voice in those meetings, and definitely so as a very controversial issue was being debated among refugees and between refugee leaders and humanitarian personnel: the rollout of a biometric identity checks pilot project. The initiatives had been implemented one month earlier in one food distribution point (FDP) in Kakuma 3 and in another in Dadaab and, in both cases, had been met with stiff resistance from camp residents. The rationale of the pilot project was made explicit in a budget revision proposal submitted to the WFP Deputy Executive Director, Operations Department.33 This proposal, which was given to me by a UNHCR staff member, drew from the recommendations formulated in the report of a joint mission conducted by the WFP, the UNHCR, the Kenyan government and donors in Dadaab and Kakuma refugee camps in September 2010. The proposal demanded an increase of the WFP budget from USD 417,574,294 to USD 433,795,280 to implement a system that would have enhanced the cost-efficiency gains by reducing the number of beneficiaries and enabling “WFP to reduce its food requirements.”34 The document thus demanded an additional USD 16,220,986” to put in place “a pilot test of biometric identity checks in Dadaab and Kakuma camps to ensure accurate targeting of beneficiaries receiving general food rations.” The biometric system was based on fingerprint readers “to verify the identity of refugees collecting GFD (general food distributions) rations” and thus “accurately identify genuine cardholders”. In particular, the interest of the document lies in the 33 Kenya Food Assistance to Refugees in Kenya PRRO 200174 B/R No.: 3 34 The requested 4 percent budget increase aimed at Landslide, Transport, Storage and Handling (LTSH) operations (USD 5.3 million); Other Direct Operational Costs (ODOC) (USD 3.5million); Direct Support Costs (DSC) (USD 6.3 million); and Indirect support costs (ISC) (USD1.1 million). 225 insights it provides on the underlying logic framing both the humanitarian intervention and its population in that it drew a distinction between legitimate and illegitimate recipients. According to the Kenyan encampment policy, only refugees who are resident in the camp are legitimate ration card holders (and these are identified by a serial number). However, according to the new proposal, The lack of a rigorous identification method leaves the process open to abuse. Ration cards may be passed on and used by non-refugees, or by refugees already holding a card. Some refugees unofficially return to their home countries and most deaths are unreported until verification exercises are held; in the meantime, the ration cards continue to be used to access services in the camps, including food. Ration cards also risk being stolen or taken from their legitimate owners and used to receive food rations until reporting and de-registering is completed. These abuses result in the sale of a portion of relief food on local markets. The implementation of a biometric identity check system to verify who is legitimately entitled to receive food rations was expected to have “significant cost- efficiency gains”, a decrease of 5-10 percent in the number of refugees eligible for food assistance, and thus a reduction of the food requirements by 6,000-12,000 mt per year, with an estimated saving of USD 6-12 million (calculating average cost of delivering 1 mt of food of US$1,000). Thus the initial implementation costs of the system were expected to be recouped in the first year following the roll-out, with substantial savings in subsequent years. Anticipating “contestations by non-legitimate cardholders”, the document further recommended practical provision “to ensure the protection of UNHCR and WFP staff, equipment and infrastructure at food distribution points (FDPs)”, it includes recommendations from security officers: a strong perimeter fence, internal security control within the verification halls, and adequate emergency exits for staff.” It concluded by stating that, The implementation of the biometric identity check system may result in some risks, including violent reactions from non-legitimate cardholders. To mitigate against the risk of violence, the FDP structures in Dadaab and Kakuma will be rebuilt to ensure that they are robust and meet the safety requirements for both the refugees and staff working in the camps. Contrary to the fears of the rapporteurs, no attacks to the FDP occurred. However, the document and the approach to the implementation of the biometric identity check 226 reproduced a critical framing of refugees as either ‘cunning crooks’ or ‘vulnerable victims’ (Horst, 2006). This distinction, underpinned by the notion of ‘legitimate’ versus ‘illegitimate’ recipients, also highlights the gulf between the normative assumptions of the humanitarian apparatus and grassroots practices enacted by recipients of aid to increase their agency. In Kakuma, Jansen (2008) observes, refugees were used to UNHCR and NGO employees and police officers accepting or asking for bribes – ‘TKK’ was a well-known phrase: toa kitu kidogo in Kiswahili, meaning ‘bring something small’: it was part of the system as they knew it. Perhaps this explains why some people thought that the only way of organising resettlement was by cheating (575) Practices of ‘refugee cheating’ have been an object of study in several humanitarian settings (Kibreab, 1991; Pérouse de Montclos and Mwangi Kagwanja, 2000). Kumsa (2006) interprets refugee cheating in terms of entrepreneurship, while Harrell-Bond (2004) suggests that ‘perhaps the ubiquitous welfare cheating of the aid regime that occurs in refugee camps could be interpreted as evidence of the re- development of “solidarity” among these populations’ (27). On the record, practitioners hold a clear cut view on the issue of cheating, although they admit a more nuanced perspective in private. A protection officer in Kakuma expressed to me his frustration with the stalemate in discussions with refugee representatives over the issue: in his opinion, the fact that they would negotiate the acceptation of the biometric system in exchange for an increase in the food rations meant that the refugees had failed to grasp the core problem, which was that there was not enough money. However, the request of the refugee delegates is understandable if we consider it in terms of broader livelihood strategies. Indeed, a common strategy in the camps was to amass food rations by stockpiling ration cards of absent household members and then trading additional food rations in exchange for cash or items not provided by humanitarian agencies. The cash might be used for diet diversification (like purchasing goat meat from Turkana pastoralists or firewood/charcoal) or invested in furthering small entrepreneurship initiatives, such as setting up a stall. Cash is also commonly used to travel to urban centres to undergo medical visits, or to pay for the journey of the breadwinner of the family, typically the most educated young male, to 227 travel to urban centres, usually Eastleigh, from where they could send remittances back to the camp and help their families establish businesses by shipping goods on request. As Somali women in Kakuma 3 confirmed, 1 Kg of rice or wheat flour provided by the UNHCR could be resold to traders such as Barre for 30 KSh. In the market of Nakuru or Eastleigh, the same amount sold for 150 KSh. At the same time, cash was necessary to pay for purchasing phone credit, and travelling across the camp, which cut the distance between the UNHCR compound and the camps. The implementation of the biometric was thus regarded as a means of further marginalisation for a large swathe of the refugee population that could not rely on external support. Instead, it was likely to exacerbate the condition of subordination of Somali Jareer to Somali Darood and Hawiye. To understand how this occurred, it is necessary to explain how a racialised space that reproduced the one in Southern Somalia, described in chapter 2 and 3, was shaped in Kakuma. “Thanks to America, I became a Bantu” Retracing the momentous “ethnogenesis at the intersection of international humanitarianism and diaspora politics” (Besteman, 2013) of the Somali Bantu identity is important to my overall argument. The intervention of humanitarian institutions to redress the consequences of the protracted marginalisation of Somali minorities from Southern Somalia, while unmasking a deeply entrenched façade of culturally-based homogeneity and ethnic solidarity, also stoked further tensions among Somali groups and turned an ethnic label into yet another resource to be preyed upon. A few months ahead of my journey to Kakuma, I was introduced to a group of Somali Jareer in Eastleigh who had travelled to Nairobi to undergo some medical check-ups. As we spoke in a coffee shop, they recalled their trajectories from Southern Somalia to Kenya’s refugee camps, stressing both how they managed to navigate the uncertainties of the current political situation and their plans for the future. I met the group again in Kakuma 3, and realized a fact that I had previously overlooked: throughout our conversation in Eastleigh, they had introduced, and referred to themselves, as Somali 228 Bantu. This identification was seldom used by Omar, Senegal and other friends in Nairobi, who referred to themselves specifically as Jareer. I have discussed how relationships of subordination derive and are encapsulated in the Somali clan architecture and how historical trajectories have shaped a racialised space in Southern Somalia. This space, coinciding with the constellation of farming communities along the rivers Juba and Shabelle, has become the epicentre of a humanitarian crisis with the outbreak of the civil war, targeted by rivalling Darood and Hawiye factions, thus attracting humanitarian aid that has further exacerbated competition over access to the areas where Jareer farmers lived. As mentioned in chapter 3, these areas, and the Jubba Valley in particular, were turned into what human rights reports called ‘one big graveyard’ (Prendergast, 1994:12, in Besteman, 1999: 18). The appalling violence perpetrated against Jareer communities, and recounted by those who succeeded escaping to Mogadishu or to Kenya, struck humanitarian and media workers, who crafted a new ethnic category – Bantu – to describe the Somali population who was suffering the most (Menkhaus, 2003, 2010). According to Besteman (2013), The widespread violence against Jareer Somalis during 1991–1992 precipitated a growing political consciousness amongst Jareer that began to fluoresce within Somalia over the next few years. Somali Bantus now living in the U.S. recall that a UNOSOM officer from Southern Africa stationed in Mogadishu encouraged the use of the term Bantu to describe Somali farmers. When Somali minorities complained to him about their treatment by other Somalis, he reportedly suggested that they should self-identify as Bantu (292). The term Bantu was quickly embraced by the humanitarian apparatus, to the extent that displaced Somalis who arrived at the camps in Kenya and were unable to provide clear information on lineage, village of origin and clan were uniformly identified as Bantu, even if they did not speak Maay Maay or other languages widespread in the regions along the rivers Juba and Shabelle (Declich, 2000). In Kenyan refugee camps, this process of ethnogenesis was further fuelled by the economic marginalisation and the constant targeting by members of other groups. The adoption of the Bantu identity provided a resource both strategic and affective (Besteman, 2013). Those identified as Bantu refugees in Kakuma 3 stressed the fact that, despite the provision of basic aid, life conditions in Dadaab were not much better than in Somalia. They also pointed out that 229 the perpetrators ‘were the same’, meaning that they belonged to the same clans, especially Darood Ogaden and Hawiye Abgal. Members of Darood Ogaden and Hawiye Abgal groups were able to rely on support from clan members in Kenya or abroad: Darood Ogaden in particular shared kinship ties with the Somali elite that, as I explained in Chapter 2, emerged in the NEP in the 1980s during Moi’s presidency. Jareer’s protracted marginalisation, on the other hand, was exacerbated by their lack of external assistance. In Dadaab, they lived separately from pastoralists, and were continuously exposed to the risk of abuse and violence, as well as being forced to work as porters, latrine cleaners or wood collectors in exchange for small amounts of money and protection (Barnett, 2003; Van Lehman and Eno, 2003; Smith, 2013). However, a window of opportunity opened up when reports written by protection officers that described the conditions of Somali Bantu in the camps struck a chord with policymakers, particularly in the U.S. A UN field officer in Dadaab’s Dagahalley camp, Dan Van Lehman, became a vocal advocate of the Bantu cause. He realised that a racially based ‘caste system’ was in place among Somali refugees, in which those with more prominent African features were regularly victimised by other refugees and Kenyan Somalis. They were barred from accessing water pits, confined to the most unwelcoming areas of the camp and physically and verbally abused with terms - ooji and adoon – which, he later discovered, meant ‘slave’35. He also found that some of these refugees spoke a language similar to Swahili, Mushunguli. As he became closer to these Somali Mushunguli, he became aware of what he defined as “an underground railroad between the refugee camps and Tanzania”, where many Bantu Mushunguli married Tanzanian women and then disappeared into Tanzania. He drafted a plan for a mass resettlement to Tanzania, but, as he recalls, “then Rwanda blew up, and Tanzania stopped.” Another possible location for resettlement was Mozambique but this option was also set aside. Mushunguli refugees in Dadaab were invited to perform songs and dance from their own communities in front of Tanzanian and Mozambican delegates who were called on to recognize cultural affinities between their populations. During 35 Interview with Dan Van Lehman, 12 December 2013. 230 this preparatory stage, Van Lehman also compiled a list of names for fast-track resettlement with the help of Omar Eno, an educated Somali Jareer who went on to become a leading scholar and Jareer activist. In the mid-1990s, Eno’s framing of Jareer discrimination as “Somali apartheid” (Eno, 1997) played a crucial role in catalyzing the attention of public opinion in the US, still horrified by the shocking images of rangers dragged along the streets of Mogadishu in 1993. Following a mission of the US Congressional Black Caucus facilitated by American NGOs to eastern Africa’s refugee camps, including Dadaab, pressure mounted for granting resettlement to groups falling in the Priority 2 (P2) status of the US Refugee Admissions Programs (USRAP), which applies to people and groups “of special humanitarian concern” (Boas, 2007, in Besteman, 2013). Entrusted with the identification of these groups, the UNHCR selected the Sudanese so-called ‘Lost Boys’ and the Somali Bantu. For the Jareer, adopting the Bantu label became crucial to accessing the fast-track to resettlement (Jansen, 2008). In this sense, “Bantu ethnicity in Somalia is not unlike Somali clan identity - a flexible tool designed principally to manage risk in a very dangerous environment, to maximize personal security and access to resources in a context of scarcity, violence, and lawlessness.” (Menkhaus, 2003: 325). This explains why, when a process of identifying those eligible for P-2 resettlement was launched, the Bantu label became a valuable commodity, to trick, bribe and kill for. As Van Lehman recalls, ‘All of a sudden, everyone was a Somali Bantu’ (in Besteman, 2013)36. Riots erupted among Somalis from dominant clans, who felt doubly offended by their exclusion from the resettlement program in favour of those who were traditionally considered inferior. At least 30 Jareer were killed in the camp. Many were threatened and stripped of their IDs, or forced to include Darood and Hawiye in their families. In this climate of fear, in 1999 the 36 Compiling a list of refugees for resettlement became an exercise in anthropometry akin to early XX century racial pseudoscience, based on the measurement of physical features that were attributed to a stereotypical Bantu identity, such as the size of the nose or the hands, or the thickness of the hair. A set of ploys were enacted to ‘look Bantu’, such as “dressing in shabby clothing and rubbing dirt on their skin, as if to appear poor, and into their hair, to make it appear ‘harder’” (Besteman, 2012). 231 UNHCR decided to relocate 12,000 Somali Bantu to Kakuma, ahead of resettlement to the US. 13,000 more arrived in 2002. Among these was Abdi Rage Kerow, a 50-year- old resident of Kakuma 3 who was held in great esteem by other Bantu refugees. Abdi comes from Buale, in Middle Juba. His farm was on the bank of the river, and he proudly remembers that he had 155 mango trees.. At the beginning of the war, his father was strangled by a gang of Darood Ogaden, one of his three wives was raped, and his property was looted and devastated. He arrived in Dadaab from Somalia in 1991 with two wives and his children. In Dadaab, he landed a job as a porter for Care International. He earned 3000 KSh per month, and his wives were launderers. They did not have relatives abroad and could not rely on remittances. He well remembers the riots sparked by the decision to target only Somali Bantu for resettlement. His family was included in the list compiled by the verification team, but this exposed him to the fury of other Somalis. He was threatened with death unless he included a Darood Ogaden girl into his family. In 2002, she moved with him, along his wives and kids to Kakuma, the last stage of the resettlement to the US. Abdi recounts that she received some money from her family, but kept it all to herself. However, during the process, it became obvious that they were not related and the consequences fell on the entire family, who failed their resettlement application. The girl moved to Eastleigh, where she now lives, while Abdi and his family remain in Kakuma. Many refugees still remembered widespread allegations of bribes being paid to the Kenyan UNHCR staff to be included in the resettlement lists. Others pointed out that, since the UNHCR recruited Kenyan Somalis who spoke the language, and they were mostly Darood Ogaden, they favoured their kin and included them on the lists. Van Lehman quotes estimates according to which at least between 10 and 20% of the Somali Bantus resettled to the US were in fact Darood Ogaden37. Abuse and humiliation were recurrent in accounts by refugees who had experienced the window of opportunity for resettlement narrowing in the last few years. The harrowing stories I collected in Kakuma 3 suggested that the abuses people suffered were part of a longer pattern, and that the displacement and the encampment they experienced simply exacerbated a situation of structural violence that had a long history in Somalia. In other words, there was continuity between the victimisation they experienced in Somalia, during their flight and eventually in the camp, particularly in Dadaab, where they all resided before coming 37 Interview with Dan Van Lehman, 12 December 2013 232 to Kakuma. Gender added a further dimension to relations of dominance and subordination that were grounded in race and class. Looma Ooyaan – No one cries for them – was a common Somali expression through which Jareer women referred to their lack of social recognition. One of the women I met in Kakuma 3, Fatuma, provided a dramatic account, of her displacement and the process of relocation to Kakuma, highlighting how conflict and displacement did not create, but laid bare and aggravated the condition of vulnerability Jareer women already faced before they left Somalia. Originally from Lower Shabelle, Fatuma moved to Mogadishu after she married. She lived in a rented house in Madina, Mogadishu (the house belonged to a Somali who lived in the diaspora), until, one day in 2006, her husband did not return home. Since she was unable to afford the rent, she and her six children were evicted. Along with others, she found shelter in an abandoned school complex, but while there she lost three of her sons. Two of them died from malnourishment and disease, and one died of injuries suffered after being hit by a piece of artillery during a street battle. She rushed to the hospital but at first they would not admit her son. . The people in the hospital then seized her son’s body and demanded a ransom for its return for the burial. She was unable to pay the request of several hundred dollars but a foreign photographer took a picture of her. The picture drew the attention of an NGO that helped her raise the amount of money to retrieve the body and arrange a funeral, but the funeral feast was then looted by a Jileec gang. She flew to Mogadishu with her three children and first sought shelter with some relatives in Merca and Brawa, in Lower Shabelle, but they were also in a dire situation. However, she obtained a little money with which she reached Dhobley, in Gedo region, crossed the border and arrived in Dadaab. Along the way, she was harassed by Kenyan police, including ethnic Somalis. In Dadaab one of her children was employed as a porter, while she collected firewood. They were robbed many times and eventually succeeded in being relocated to Kakuma. In Kakuma 3 she felt safer, although she had several health issues, worsened by the arid environment, and she lacked money to travel to the IRC compound to undergo medical treatment. All her children now work as porters at the distribution centre. She is held in high esteem in the community because she also takes care of 13 motherless children. However, she does not receive any outside support. The ‘transnational connections’, so recurrent in the literature on Somali refugees, were conspicuous by their absence in the accounts of the residents of Kakuma 3. This sector of the camp was cut off from the flows of goods and capital that passed through Kakuma 1, and that even I, endowed with nothing but my European passport (and, of course, a social network stretching as far as Bristol), had tapped. These refugees had embraced a made up Bantu identity as the only asset that might grant access to resettlement. Yet, that asset had lost value with the passing of time and now needed to be 233 replaced with other resources, which continued to be out of their reach. Remittances did not feature as central in Jareer livelihood strategies as they did in those of Darood and Hawiye Somalis, who drew on their transnational social networks to acquire skills and competencies, set up businesses, pursue migration projects and thus establish new linkages outside their clan fellows which in turn allowed them to adapt better to changing scenarios and maximize benefits from favourable institutional settings. This was the case in Eastleigh and Kisenyi, at least for those who could turn their social capital into financial capital, and the other way around. The solidarity that Darood and Hawiye businessmen emphasised rested on solid collaterals. This stood in opposition to the disillusion of many Jareer, who instead stressed the weaknesses of kin ties once a family member had been resettled abroad. The story of Abdi, a 29-year-old from Kamsoma, in Lower Juba, is particularly telling. Abdi was 13-year-old in ’99 when he flew with his aunt [one of his father’s wives] to Dadaab. He was in school when the verification exercise to select the Bantus eligible for resettlement started. Unlike most Somali Jareer, he has good memories of Dadaab, saying that, despite the discriminations from Somali Jileec, he had hoped he would soon leave the refugee camp. He was relocated to Kakuma in 2002 as a preliminary stage to resettlement to the US. Since, as he said, he looked ‘very Kenyan’, he found a temporary job as a cook in a hotel in Kakuma town. In 2004, a group of relatives, including him and his aunt, were given resettlement to the US and were flown to Nairobi, where they went through the last interview at the American consulate. There, however, he and his aunt were crossed out from the list since they were not biologically related and therefore not eligible. One of his uncles was granted resettlement and told him to keep on hoping. He promised him: “I will support you if you marry my daughter.” So Abdi married his cousin, but then his uncle disappeared. Now, Abdi is responsible for his wife, two kids and an older brother. At the moment, thanks to his good education, he is working with the Lutheran World Service. His Bantu identity is all that he has left. “I am thankful to the US for teaching me who I am”. A vicious circle of poverty Refugees in Kakuma 3 constantly emphasised their lack of solidarity with other Bantus who had been resettled abroad. I strove to understand the reason for this ‘broken link’. Most Somali Jileec I asked vaguely pointed out that Jareer did not share the same values of solidarity and mutual help that they claimed were ingrained in the Cushitic 234 pastoralist tradition. Others suggested that, because Jareer solidarity ties are based on being part of a farming community, they existed as long as there was physical proximity among the members of the community. All these hypotheses were based on sweeping generalisations that overlooked the role of individual agency and the specificities of the context of reception. Obviously, the affective dimension of interpersonal relations played a crucial role, but solidarity acted as a powerful force, independently of people’s clans, when mechanisms of social control and incentives were in place. A Bantu leader I met in London helped me see the issue in a different light. Being resettled was experienced as both a privilege and a burden, because the resettled person was supposed to look after the relatives they left behind. The problem, though, was that limited support in the country of resettlement was often only just enough to make ends meet. As Besteman (2013) explains, with reference to the resettlement program for Somali Bantu launched by US State Department in the early 2000s, “the policy of rapid self-sufficiency condemns refugees to an intergenerational cycle of poverty” (18). Many resettled Bantu thus cut links from family left behind out of shame because they were unable to care for all their relatives in the refugee camps and Somalia and, instead, opted to support only some of them. To integrate the scarce flow of remittances, refugees in camps combined basic WFP provisions with small incomes from menial jobs. They put to use skills that, back in Somalia, were Jareer’s prerogative, such as masonry, woodwork and agriculture. Many, for instance, were recruited by aid agencies and NGOs in the camp to build structures such as warehouses or pit latrines. They also worked for private refugees, especially Somali Hawiye and Darood, and Ethiopians in Kakuma 1 and 2. In 5/6 days, if there was enough mud available, during the wet season between April and June, groups of 2/3 Somali Bantu could make up to 3,000 mud bricks worth 2-3 KSh each. Typically, a Somali Jileec family would buy bricks ahead of weddings, and then pay Jareer for the labour to build a house for the newly married couple. Jileec also provided timber with which Bantu carpenters built furniture. A few dozen Jareer men worked as piki piki drivers. The vehicles belonged to Darood businessmen who took 700 KSh out of the 1000 KSh earned. However, Bantu refugees in Kakuma 3 largely acknowledged that most workers were women. Bantu women cultivated scuma, okhra, tomatoes and 235 watermelons in small orchards near water catchments next to water distribution points. Most of this produce was then brought to Kakuma 1 and sold to wholesalers. Because of irregular precipitation in the region, these cultivations were only feasible for a few months per year, during which farmers sought to earn and save as much as possible. Many Bantu women were forced to seek work with Somali Darood and Hawiye families in Kakuma 1 and 2, cleaning latrines, plastering house walls and collecting firewood. Depending on the task, women were paid up to 150 KSh per day, but often they worked to repay debts for money they had already borrowed. Money was used to purchase a wide range of basic products, such as canned food, pasta, meat, and charcoal; but also to purchase airtime, recharge phones or buy medicine, very often from the same Somali traders the women ended up working for. 21 out 25 Bantu women interviewed claimed to have debts with Somali Darood in Kakuma 1. One woman reported having worked for three months to repay a 1000 KSh (7 GBP) debt. Since they had to rely on loans from better-off Somalis, most Jareer admitted that they often chose not to report abuse and harassment, lest they got a reputation for being troublemakers which might affect their possibility of accessing future loans. Caught in a vicious circle of poverty and marginalisation, the Somali Bantu population of Kakuma 3 was bound to members of dominant clans residing in Kakuma 1 and 2 in a relation of dependence akin to the one, described in chapter 2 and 3, historically experienced in Somalia, where the dynamics of the conflict and the interaction of non-state armed actors with the international aid sector entrenched the influence of gatekeepers. As Dan Van Lehman, the former UNHCR field officer that first brought the plights of Somali minorities in refugee camps to the attention of policymakers, told me, “the Somali Bantu are the oil of Somalia: whoever controls them controls aid, money and political influence. They are kept in enclosures like goats.”38 The resettlement program for Somali Bantu to the USA has apparently failed to reddress this condition. Studies among Bantu resettled in the USA thus found that ‘social isolation and marginalization continue in America, as most do not belong to 38 Personal communication, 24/11/2014 236 organizations in the broader community and most do not interact with their non-Bantu neighbors” (Smith, 2013: 178). Applying Bourdieu’s conceptual toolbox to her research on Somali Bantu living in urban areas in the Western US, Smith observes a widespread inability to transfer the communal support experienced in Somalia in refugee camps and sites of resettlement. As she reports, “One young Bantu male said, ‘‘We all Bantu. We have each other.’’ But as much as the Somali Bantu would like to support each other as they did in the past, physical distance from family members and others in their community, along with a lack of useful resources in Western society, have made it difficult, if not impossible, for them to maintain many of their former social relationships.” (176) The breaking of traditional ties and the isolation of the Somali Bantu community are adduced as major reasons that lie at the roots of their large occupational deprivation. Apart from the US, where indeed Somali Bantu are treated as a distinct category within the broader Somali community, in other Western resettlement countries there are no disaggregated data and it is therefore not easy to individuate processes of social stratification within the diaspora. For instance, a 2013 article in the Economist stirred a debate among Somalis in Britain by claiming that, among the 101,370 Somali-born people identified by the 2011 census in England and Wales, just one in ten is a full-time worker and over 50% of British Somalis (more than any other foreign-born population) live in council houses39. An open letter from a group of British Somali professionals in response to the Economist article highlighted analytical flaws (mixing national and local data and drawing comparison with longer established diasporas, such as Bangladeshi and Nigerian) and gross generalisations (such as “without work, Somali men while away their days chewing khat” (Economist, 2013))40, and pointed to the achievements of British Somalis. No studies have yet been conducted on class dynamics amongst the 39“The Road is Long”, The Economist, 17 August 2013 available at: http://www.economist.com/news/britain/21583710-somalis-fare-much-worse-other-immigrants-what- holds-them-back-road-long, accessed on 18 August 2015 40“The short road to demonization”, 23 Sept 2013, available at: https://hawashaven.wordpress.com/news-2/media-releases/, accessed on 18 August 2015 237 Somali diaspora in Britain and elsewhere in Europe, although the diasporas are framed as increasingly important political interlocutors of Western governments. Further research is therefore required to understand how social stratification works in the resettlement countries, and what role clanship plays in determining social advancement in the diaspora. Also, the investigation on which this thesis is based has offered much food for thought about the way the reproduction of inequalities in the diaspora mirrors the political situation in Somalia, the absent but always present site in my multi-sited ethnography. Three years after being elected as first President of Somalia since 1991 with the mandate to lead the country to democracy, at the end of July 2015 Hassan Sheikh Mohamud announced on the Somali presidency Twitter account that "'One person, one vote' will be not possible in 2016" (Reuters, 2015). With this short announcement, Mohamud surrendered to the logic of clanship, which has hitherto dominated Somali politics. This entails that the 4.5 formula will continue not only to confirm the centrality of the major Somali clans in politics, but also to sanction, through halved political representativeness, the lower social status of Somali minorities. Conclusions In this chapter, I have shown that a refugee camp is a context of reception fashioned by spatial and regulatory configurations in which entrepreneurial opportunities are present but are not homogeneously distributed. By examining power relationships in Kakuma, a humanitarian complex suspended between exception and protractedness, I have argued that deep-seated social structures, connections between the local and the global facilitated by financial and communication infrastructures, and the collusion of dominant groups with the humanitarian apparatus and state institutions conspire to reproduce a racialised space, akin to the one discussed in chapters 1 and 2 with reference to the farming communities of Southern Somalia. Considering “humanitarian action as an arena where actors negotiate the outcomes of aid” (Hilhorst and Jansen, 2010: 1120), I have discussed how pre-existing inequalities 238 are projected into such negotiation, thus enabling entrepreneurs embedded in both clan- based structured and personal networks to preserve, and reproduce, their position of dominance. This is done by controlling the circulation of goods and value within the camp and between the camp and the outside. The implications of the camp’s spatial arrangements limit the agency of members of Somali minorities. Comprehending these asymmetries is essential to grapple with the complex reproduction and interweaving of networks underlying the apparently homogeneous surface of the camp. But my ethnography of the Somali population in Kakuma is also a further step in closing a circle begun with the description of the Somali clan system and its crystallisation throughout colonial and post-colonial history. It is an empirical immersion into an experience in which not all refugees were confined to a state of ‘bare life’, to use Agamben’s idiomatic expression (Agamben, 1998). In fact, while this was the case for many (if not all) Somali Jareer, the refugee camp was for others a place where to leverage organisational legacies, translocal linkages and privileged access to local institutional actors, and where to accrue capital and thus preserve a dominant position. This specific locality, being a ‘space of paradox’ (Grbac, 2013), was simultaneously closed and open, a space of deprivation and a source of assets, a humanitarian shelter in the middle of nowhere and a financial terminal connected to a specific neighbourhood in the UK. As I experienced during my stay in Kakuma, some sectors of the camps enjoyed greater integration to the rest of the world, being connected to infrastructures (human, technological, financial) through which information (to be converted into money, knowledge or whatever resource could be harnessed to increase mobility) travelled. Even in an institution, such as a refugee camp, that epitomises segregation and marginality, some areas were better off than others, in terms of security and livelihoods. In Kakuma refugee camp, power relations relied on spatial infrastructures. Yet, their roots went much deeper, to the legacies from which my journey began. 239 Conclusions A key point of my study was the entanglement of capitals and mobility, and the significance of crossing borders for converting different forms of capital against a volatile background. The strategy of coping with the unpredictable is based on the preservation of alternatives, alternatives that are produced by creating new ties, consolidating pre-existing networks, bridging categories and institutions in order to spread the risk. No unique financial infrastructure is perceived as reliable, no unique locality is viewed as a definitive safe haven. Strategies are driven by the need to broaden the range of possible options. The practices that have been described here aim at accommodating different financial circuits and securing the possibility of escaping and moving on in case of need. Hard currencies, passports, and relations of obligation are the key resources that enable movement across space and time, hedging against inflation, as in the case of the much-sought-after USD, or allowing people to travel as a legitimate businessman to Dubai or China, or even Somalia. In this case, there is a stark difference between those who can rely on resources to buy protection in loco, or a safe return to Kenya or Uganda in the case of a serious threat, and those unable to do so for fear of not being able to return to safety. The significance of mobility appears thus contingent upon the historical situation of Somalia and the Eastern African contexts of reception (Eastleigh, Kisenyi, Kakuma) I have examined in this thesis. At the beginning of this thesis, I used the literature on transnationalism to question the ‘sedentary bias’ that permeates social sciences. I have consciously tried to avoid falling in the opposite – a mobility bias. However, a ‘mobility bias’ was, indeed, ingrained in the narratives that buttress the sense of entitlement of the diaspora elite. Members of dominant groups essentialised what was a pastoralist strategy to cope with a harsh environment and turned it into an identity boundary to legitimise (to themselves and to external actors) entrenched inequalities. During a drifting conversation in 240 Muhubo’s shop, Mohamed once explained that camels and cattle could travel over a long distance, and this was why they “pertain to men”. Goats and sheep, on the other hand, always roam nearby, and this was the reason why only women – and farmers, and slaves – herded them. In his narrative, the gendered difference between camels and cattle on the one hand, and goats and sheep on the other, mirrored the one between hard currency, the money handled by men, that can travel far away in space and time, and local currencies, with which the women provide for the daily needs of their families, and upon which Jareer simply survive. His mother, who had built a reputation out of her talent in managing complex ledgers and handling thick wads of USD, was clearly an exception to this rigid polarization. Gender was not the real issue. Power was: the power to produce and store value in an uncertain world. The power that pastoralists, above all Darood and Hawiye (Mohamed maternal and paternal clans) continued to wield, and Jareer farmers, dispossessed of their land in Somalia and mostly forced to live on humanitarian aid in Somalia and Kenya, had never held. This power was legitimised by discourses aimed at reproducing a view of the social reality that emphasises the dimension of mobility, and the possibilities attached to it, as a marker of the difference of status. The risk-taking, path-finding attitudes that were commonly viewed as a key to succeed in business were ascribed to the traditional lifestyle and livelihood of Darod and Hawyie clans. The ‘abandoned roads on an old map’ that, according to Bourdieu (1977: 38), are genealogical trees, were brought back to life by the postcolonial state and the period of statelessness and used to make sense of the balance of cohesion and dispersal, of Somali solidarity and of inequalities between ‘real’ Somali pastoralists and ‘non- Somalis’ farmers. These discourses continue to preside over the way Somali networks expand and accommodate to shifting scenarios. To explain why, and under which circumstances, these networks make opportunities available to some individuals but not to others, I have highlighted the importance of transforming different forms of capital into one another, for instance converting social capital into business opportunities, or investing in education, which may land a job with an INGO or a humanitarian agency, and thus 241 facilitate the creation of new linkages. I have also suggested that social networks, a necessary requirement for inclusion into business partnerships, should be regarded as the collateral to access loans, further binding the beneficiaries into a relationship of obligation. The argument I have advanced is that it is necessary to adopt a historical perspective to regard both ties of trust and current inequalities as legacies of the past in which current Somali socio-cultural structures took shape. This thesis has also suggested that trust is not just enshrined in traditional networks based on kinship, but has to be constantly reproduced. I have tried to steer away from sweeping generalisations and avoid structural determinism to eschew a circular line of argument, in which those with resources have the capabilities to access more resources. It is worth reminding that not all members of dominant clans enjoy a condition of privilege, although membership and display of group solidarity facilitate the access to resources to start a business, provided that this activity, and its eventual success, reflects positively on the broader clan. At the same time, although being a Jareer is considered both a racial and social marker that hinders social mobility, in the course of this thesis I presented examples in which attempts to break the loop were made, with mixed results. The cases of some Jareer refugees in Eastleigh highlighted the importance of individual agency in carving routes out of marginality through the skills and competencies acquired in refugee camps, as the examples of Omar and Senegal illustrated. Initiatives such as the resettlement plan launched by USAID in the 1990s, which defined the boundaries of the Bantu category, highlights the role of humanitarian and development actors in formatting a group identity identifying marginalised groups, buttressing livelihoods and designing adequate protection programmes. However, even resettlements to the West have had a limited impact on the structural factors that continue to bar Somali Jareer/Bantu from acquiring the means to improve their living conditions. There is, however, another major concern underpinning this study. Forced and voluntary migrations are currently among the most emotionally-charged and divisive 242 political issues in Europe. Media outlets constantly show images of migrants from Africa, the Middle East and Central Asia, adrift in the Mediterranean off the shore of Italy, or stranded in dismal camps in the Balkans. Metaphors of ‘floods’, or ‘swarms’ stoke xenophobic tensions and achieve the effect of both dehumanising and homologating what is in reality a multifaceted phenomenon. Africa continues to be framed as a gigantic pool of migrants gathering on the coasts of North Africa, ready to set off to Europe (Hughes, 2015; Roberts, 2014). Sporadic journalist reports underline the significance of “class systems based on money, ethnicity and religion” (Stevis and Mesco, 2015) that determine the fate of migrants, although this aspect remains under- investigated. This thesis has shown that mobility in Africa is a complex phenomenon, shaped by aspirations, life projects and relationships of solidarity and subordination. 243 Appendix 1 Amal Shopping Centre, Eastleigh, Nairobi An M-Pesa kiosk in Eastleigh 244 Smelting gold at the exchange market, Eastleigh View of Old Kampala from Kisenyi main street 245 The Lucky Star, the first Somali shopping mall in Kampala Outside Kakuma refugee camp 246 References Abbink, J. 2009. The Total Somali Clan Genealogy (second edition), African Studies Centre, Leiden, Abdi, S. Y. 1979. ‘Self-Determination for Ogaden Somalis’, Horn of Africa 1, 1: 20–25. Abdulsamed, F. 2011. Somali Investment in Kenya, Chatham House: Independent thinking on international affairs. Available at http://www.chathamhouse.org/publications/papers/view/109621 [accessed 28 February 2015]. Adam, H. 1992. ‘Somalia, Miltarism, Warlordism or Democracy’, Review of African Political Economy, 54. Adam, H., 1995. ‘Somalia: A Terrible Beauty Being Born?’, in W. Zartman, ed. Collapsed States – The Disintegration and Restoration of Legitimate Authority, Boulder, CO: Lynne Rienner Publishers, 69–90. Adam, H. 1999. ‘Somalia: Problems and Prospects for Democratization’, in K. Menkhaus and C. K. Daddieh, eds. State Building and Democratization in Africa – Faith, Hope and Realities, London: Praeger, 261–277. Adam, H. 2008. From Tyranny to Anarchy – The Somali Experience. Asmara: The Red Sea Press. Adey, P. 2006.’ If Mobility is Everything Then it is Nothing: Towards a Relational Politics of (Im)mobilities’, Mobilities, 1,1: 75–94. Afghan Research and Evaluation Unit (AREU).2006. Afghans in Quetta. Settlements, Livelihoods, Support Networks and Cross-Border Linkages, January. Available at www.unhcr.org/refworld/docid/47c3f3c412.htm [accessed 13 May 2015]. Africa Watch. 1990. Somalia: A Government at War with Its Own People, London: Africa Watch. Africa Watch. 1991. Kenya: Taking liberties. London: Africa Watch. African Rights. 1993. Somalia operation 'Restore Hope': a preliminary assessment. London: African Rights. 247 Agamben, G. 1995. Homo Sacer I: Il potere sovrano e la nuda vita. Torino: Giulio Einaudi editore. Agier, M. 2002. ‘Between War and City. Towards an Urban Anthropology of Refugee Camps’, Ethnography 3, 3: 317–41. Agier, M. 2011. Managing the Undesirables: Refugee Camps and Humanitarian Government. Cambridge: Polity Press. Ahmad, A. 2012. ‘Agenda for peace or budget for war? Evaluating the economic impact of international intervention in Somalia’, International Journal, Spring: 313–331. Ahmed, S. 2004. ‘Affective Economies’, Social Text 22, 279: 117-139. Al Jazeera. 2012a. Kenyan capital hit by grenade attack. 3 August. Al Jazeera, 2012b. Uganda threatens Somalia troop withdrawal. 02 November. Al Jazeera. 2012c. Two injured in Kenya explosions – Africa. 19 December. Available at http://www.aljazeera.com/news/africa/2012/12/20121219172758769966.html [accessed 19 January 2015]. Al-Sharmani, M. 2007. ‘Contemporary Migration and Transnational Families : The Case of Somali Diaspora ( s )’, in Migration and Refugee Movements in the Middle East and North Africa. The Forced Migration & Refugee Studies Program The American University in Cairo, Egypt. Åkesson, L. 2013. ‘The queue outside the embassy: Remittances, inequality and restrictive migration regimes’, International Migration, 51, 1: 1–12. Aminuzzaman, S., H. Baldersheim, & I. Jamil. 2003. ‘Talking back: empowerment and mobile phones in rural Bangladesh: a study of the village pay phone of the Grameen Bank’, Contemporary South Asia, 12, 3: 327-348. Anderson, D.M. et al., 2012. ‘Camel herders, middlewomen, and urban milk bars: the commodification of camel milk in Kenya’, Journal of Eastern African Studies, 6, 3: 383–404 Anderson, D.M., 2014. ‘Remembering Wagalla: state violence in northern Kenya, 1962–1991’, Journal of Eastern African Studies, 8, 4: 658–676. 248 Andreas, P. 2003. ‘A Tale of Two Borders: The U.S.–Canada and U.S.– Mexico Lines After 9–11’, in P. Andreas & T. J. Biersteker (eds.) The Rebordering of North America: Integration and Exclusion in a New Security Contexts New York: Routledge. Anju M. P. 2011. ‘Stepwise International Migration: A Multistage Migration Pattern for the Aspiring Migrant’, American Journal of Sociology, 116, 6: 1842–86. Appadurai, A.1990. ‘Disjuncture and Difference in the Global Cultural Economy’, Public Culture, 2, 2: 1–24. Appadurai, A. 1996. Modernity at Large: Cultural Dimensions of Globalisation. Public Worlds, Minneapolis, MN: University of Minnesota Press. Archambault, J.S. 2011. ‘Breaking up “because of the phone” and the transformative potential of information in Southern Mozambique’, New Media & Society 13,3:.444–456. Baba, M.L., 2013. Anthropology and Transnational Migration: A Focus on Policy. International Migration, 51, 2:.1–9. Bachmann, J. & J. Hönke. 2010. 'Peace and Security’ as Counterterrorism? Old and New Liberal Interventions and their Social Effects in Kenya’, African Affairs 109: 434, 97-114. Bagenda, E. A.Naggaga & E. Smith. 2003. Land Problems in Nakivale Settlement and the Implications for Refugee Protection in Uganda, Refugee Law Project Working Paper No. 8 Ballard, R. 2005. ‘Coalitions of Reciprocity and the Maintenance of Financial Integrity within Informal Value Transmission Systems: the Operational Dynamics of Contemporary Hawala Networks’, Journal of Banking Regulation 6, 4: 319 – 352. Balthasar, D., 2014. Somalia: unable to escape its past? Revisiting the roots of a fractured state and elusive nation. Critical African Studies, 6, 2-3: 223–239. Bakewell, O. 2000. ‘Uncovering Local Perspectives on Humanitarian Assistance and its Outcomes’, Disasters 24, 2: 103–16. Bakewell, O., & H.de Haas. 2007. ‘African migrations: Continuities, discontinuities and recent transformations’, in L. de Haan, U. Engel, & P. Chabal (eds), African Alternatives, Brill: Leiden, 95-118. 249 Bakewell, O. 2008. ‘Research Beyond the Categories: The Importance of Policy Irrelevant Research into Forced Migration’, Journal of Refugee Studies 21, 4: 432–453. Bakewell, O. 2010. ‘Some Reflections on Structure and Agency in Migration Theory’,.Journal of Ethnic and Migration Studies 36, 10: 1689–1708. Bakewell, O. 2011. ‘Conceptualising Displacement and Migration: Processes, Conditions, and Categories’, in K. Koser & S. Martin (eds.) The Migration- Displacement Nexus: Patterns, Processes, and Policies. New York and Oxford: Berghahn Books, 14–28. Bakewell, O., & A. Bonfiglio. 2013. ‘Moving Beyond Conflict: Re-framing mobility in the African Great Lakes region’, Working paper for the African Great Lakes Mobility Project. IMI working paper 71. Oxford: International Migration Institutes, University of Oxford. Bakewell, O.& G. Jόnsson. 2011. Migration, mobility and the African city, Oxford: IMI: International Migration Institute. Bakewell, O., N. Van Hear & K. Long. 2012. Drivers of Migration. Migrating out of Poverty RPC Working Paper 1. Migrating out of Poverty Consortium, University of Sussex, Brighton, UK Bakonyi, J., 2013. ‘Authority and administration beyond the state: local governance in southern Somalia, 1995–2006’, Journal of Eastern African Studies 7, 2:.272–290. Barnes, C. 2014. Losing Hearts and Minds in Kenya. International Crisis Group. 16 April. Available at http://blog.crisisgroup.org/worldwide/2014/04/16/losing-hearts-and-minds-in- kenya/ [accessed 20 Jan. 2015]. Barnett, D. 2003. Out of Africa: Somali Bantu and the paradigm shift in refugee resettlement.Washington, DC: Center for Immigration Studies. Bayart, J. F. 1993. The State in Africa: The politics of the belly, Longman, London Bayart, J. F., S. Ellis, & B. Hibou, (eds.) 1999. The Criminalisation of the State in Africa. Bloomington: Indiana University Press. Bauman, Z. 2000. Liquid Modernity. Cambridge: Polity Press. 250 Bauman, Z. 2002. Society Under Siege. Cambridge: Polity Press. BBC. 2001. US shuts down Somalia internet. 23 November. Available at http://news.bbc.co.uk/1/hi/world/africa/1672220.stm [accessed on 10 February 2015]. BBC. 2010. ‘'Somali link' as 74 World Cup fans die in Uganda blasts’, 12 July BBC. 2012. Nairobi's Somali area Eastleigh hit by fatal blast. 7 December. Available at http://www.bbc.co.uk/news/world-africa-20645415 {Accessed on 12 Jan. 2015] BBC. 2014a. Kenya terror charges after Mombasa police raid mosque. 03 February BBC. 2014b. Somali PM angry over diplomat’s arrest in Kenya’, 28 April BBC, 2014c. Kenyan Nairobi blasts : Police arrest 627 in Eastleigh. 1 April. BBC 2015 David Cameron: 'Swarm' of migrants crossing Mediterranean. 30 July Beal, W. & S. Lavin. 2011. Theorizing Connectivity : Modernism and the Network Narrative. Digital Humanities Quarterly Theorizing, 5, 2. Available at: http://digitalhumanities.org/dhq/vol/5/2/000097/000097.html. Beck, T. & R. Cull, 2013. ‘Banking in Africa’, Policy Research Working Paper Series 6684, The World Bank. Bell, D. 1991. ‘Reciprocity as a Generating Process in Social Relations’, Journal of Quantitative Anthropology 3: 251-260. Bellow, A. 2003. In Praise of Nepotism: A Natural History. Westminster, MD: Doubleday Publishing. Ben-Porath, Y. 1980. ‘The F-Connection: Families, Friends, and Firms and the Organisation of Exchange’, Population and Development Review 6,1: 1-30. Berriane, M., & H.de Haas. 2012. African migrations research: innovative methods and methodologies. Trenton: Africa World Press. Besteman, C. 1996. ‘Violent politics and the politics of violence: the dissolution of the Somali nation-state’, American Ethnologist 23, 3: 579–596. 251 Besteman, C. 1999a. Unravelling Somalia: Race, Violence, and the Legacy of Slavery. Philadelphia: University of Pennsylvania Press. Besteman, C., 1999b. ‘Somali Bantus in a State of Refuge’, Bidhaan, 11–33. Besteman, C., 2012. ‘Translating Race across Time and Space: The Creation of Somali Bantu Ethnicity’, Identities: Global Studies in Culture and Power, 19, 3: 285–302. Betts, A. 2012. Self-reliance for refugees: a view from Kyangwali settlement, Humanitarian Innovation Project. Available at http://www.humanitarianinnovation.com/blog.html Betts, A. 2013. Survival Migration: Failed Governance and the Crisis of Displacement.Ithaca: Cornell University Press. Black, R. and Koser, K. eds. 1999. The end of the refugee cycle? Refugee repatriation and reconstruction. Berghahn, Oxford Bohannan, P. 1959. ‘The impact of money on an African subsistence economy’, Journal of Economic History 19: 491-503. Bolt, M. 2014.’ The sociality of the wage: Money rhythms, wealth circulation, and the problem with cash on the Zimbabwean-South African border’, Journal of the Royal Anthropological Institute, 20, 1: 113–130. Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge University Press, Cambridge. Bourdieu, P., 1980. Le capital social. Actes de la recherche en sciences sociales, 31, 1: 2–3. Bourdieu, P. 1983. ‘The forms of capital’, in J.G. Richardson (ed.) Handbook of Theory and Research for the Sociology of Education. New York: Greenwood, 241–258. Bourdieu, P. 1987. ‘What Makes a Social Class ? and Practical On The Theoretical Existence Of Groups’, Berkeley Journal of Sociology 32:.1–17. Bourdieu, P. & L. Wacquant, L. 1992. An Invitation to Reflexive Sociology. Chicago: University of Chicago Press. Bourdieu, P. & L.,Wacquant. 2013. ‘Symbolic capital and social classes’, Journal of Classical Sociology 13, 2: 292–302. 252 Bradbury, M. 1994. The Somali Conflict: Prospects for Peace. Oxford: Oxfam Research Paper no. 9. Bradbury, M., 2008. Becoming Somaliland. Oxford: James Currey. Branch, D., 2014. ‘Violence, decolonisation and the Cold War in Kenya’s north-eastern province, 1963–1978’. Journal of Eastern African Studies, 8, 4: 642- 657 Branch, D. 2011. Why Kenya Invaded Somalia. Foreign Affairs, 15 November. Brann, P. & M. Foddy 1987. 'Trust and the consumption of the deteriorating common resource', Journal of Conflict Resolution 31,4: 615–630 Bréant, H. 2013. ‘What if diasporas didn’t think about development?: a critical approach of the international discourse on migration and development’, African and Black Diaspora: An International Journal, 6, 2: 99–112. Bredeloup, S. & O. Pliez. 2005. ‘Migrations entre les deux rives du Sahara’, Autrepart (special issue on trans-Saharan migration) 4, 36. Brees, I. 2009. ‘Refugee Business: Strategies of Work on the Thai-Burma Border’, Journal of Refugee Studies, 21, 3: 380–397. Brettell, C., 2006. Introduction: Global Spaces/Local Places: Transnationalism, Diaspora, and the Meaning of Home. Identities, 13(3), pp.327– 334. Bricchetti, Robecchi, G. 1904. Dal Benadir, Lettere illustrate alla societa Antischiavista d’Italia. Milano. Brickell, K. & A. Datta. 2011.’Introduction - Translocal geographies’, in Brickell, K. & A. Datta (Eds.) Translocal geographies - Spaces, places, connections, Ashgate Verlag, Farnham,S. 3–20 Brun, C., 2015. ‘Active waiting and changing hopes. Toward a Time Perspective on Protracted Displacement’, Social Analysis, 59(1), pp.19–37. Brun, C. & A.Fábos. 2015. ‘Making Homes in Limbo? A Conceptual Framework’. Refuge, 31, 1: 5–17. Burton.R. 1943 First Foots teps in East Africa. London: Dent & Sons- New York: Dutton & Co. (orig. 1856) 253 Campbell, E.H. 2006. ‘Urban refugees in Nairobi: Problems of protection, mechanisms of survival, and possibilities for integration’, Journal of Refugee Studies, 19, 3:,396–413. Carling, J. 2014. ‘Scripting Remittances: Making Sense of Money Transfers in Transnational Relationships’, International Migration Review, 48, S1: 218–262. Carling, J. 2013. The role of aspirations in migration. Paper presented at Determinants of International Migration, International Migration Institute, University of Oxford. Carney D. 1998. Sustainable Rural Livelihoods. What Contribution can we make? London: DfID. Carrier, N. 2015. Little Mogadishu. Eastleigh, Nairobi's Global Somali Hub. London: Hurst. Carrier, N. & E. Lochery, 2013. ‘Missing states? Somali trade networks and the Eastleigh transformation’, Journal of Eastern African Studies 7, 2: 334–352. Cassanelli, L. V. 1982. The Shaping of Somali Society. Philadelphia: University of Pennsylvania Press. Cassanelli, L. V. 1996. ‘Explaining the Somali Crisis’, in C. B. and L. Cassanelli (eds.) The Struggle for Land in Southern Somalia – The War Behind the War, London: Haan Publishing, 313–326. Castagno, Alphonso. 1964. “The Political Party System in the Somali Republic.” In Political Parties and National Integration in Tropical Africa, edited by James Coleman and Carl Rosberg, 512–559. Berkeley: University of California Press. Castells, M. 1989.The informational city: infor mation technology, economic restructuring, and the urban-regional process. Oxford: Basil Black. Castells, M. & A. Portes. 1989. ‘World Underneath: The Origins, Dynamics, and Effects of the Informal Economy’, in A. Portes, M. Castells & L. Benton (eds). The Informal Economy. Baltimore: Johns Hopkins University Press. Castles, S.,2010. ‘Understanding Global Migration: A Social Transformation Perspective’, Journal of Ethnic and Migration Studies 36, 10: 1565–1586. 254 Cederberg M. 2012. ‘Migrant networks and beyond: Exploring the value of the notion of social capital for making sense of ethnic inequalities’, Acta Psychologica 55, 1: 59–72. Chabal, P. & J.-P. Daloz 1999. Africa Works: Disorder as a Political Instrument, Bloomington: Indiana University Press. Chambers, R. and G.,Conway. 1991. Sustainable rural livelihoods: Practical concepts for the 21st century. IDS Discussion Paper 296, IDS, United Kingdom Chau, D. 2010. ‘The fourth point: An examination of the influence of Kenyan Somalis in Somalia’, Journal of Contemporary African Studies 28, 3: 297–312. Chu, J. 2010. Cosmologies of Credit. Transnational Mobility and the Politics of Destination in China. Durham: Duke University Press. Cockayne, J. & L. Shetret. 2012. Capitalizing on Trust: Harnessing Somali Remittances for Counterterrorism, Human Rights and State Building. Washington, DC: Center on Global Counterterrorism Cooperation. Cohen, A. 1969. Custom and politics in urban Africa: a study of Hausa migrants in Yoruba towns. London: Routledge & Kegan Paul. Cohen, J.H. 2004, The culture of migration in Southern Mexico. Austin: University of Texas Press. Cohen, R. 2008. Global diasporas: an introduction. London: UCL Press. Coleman, J. 1988. ‘Social Capital and the Creation of Human Capital’, American Journal of Sociology 94: 94-120. Coleman, J. 1990. Foundations of Social Theory. Cambridge, Mass: Harvard University Press. Collinson, S. (ed.). 2003. Power, Livelihoods and Conflict: Case Studies in Political Economy Analysis for Humanitarian Action HPG Report 13, Overseas Development Institute: London. Comaroff, J. & J.L. Comaroff. 2001. ‘Naturing the Nation : Aliens , Apocalypse and the Postcolonial State’, Journal of Southern African Studies 27, 3: 627–651. Comaroff, J.J. & J.J. Comaroff. 2009. Ethnicity, Inc, Chicago and London: University of Chicago Press. 255 Conradson D. & D. McKay. 2007 ‘Translocal subjectivities: mobility, connection, emotion’, Mobilities 2: 167–74 Cooper, F. 2001. ‘What is the Concept of Globalization Good for? An African Historian’s Perspective’, African Affairs 100, 399: 189-213. Coquery-Vidrovitch, C. 2005 ‘Introduction’, in S. J. Salm and T. Falola (eds.) African Urban Spaces in Historical Perspective. Rochester: University of Rochester Press. Cresswell, T. 2001. ‘The production of mobilities’, New Formations 43, 1: 11–25. Cresswell, T. 2006. On the Move. Mobility in the Modern Western World. London: Routledge. Crisp, J. 1999. ‘A state of insecurity: the political economy of violence in refugee- populated areas of Kenya’, New Issues in Refugee Research, edited by Evaluation and Policy Analysis Unit, United Nations High Commissioner for Refugees, Geneva. Crisp, J. 2000. Africa’s refugees: patterns, problems and policy challenges, New Issues in Refugee Research Working Paper No. 28. Geneva: UNHCR. Crisp, J., T. Morris & H. Refstie. 2012. ‘Displacement in urban areas: New challenges, new partnerships’, Disaster 36, S1: S23-S42 Cross, J. & A. Street. 2009. ‘Anthropology at the bottom of the pyramid’, Anthropology Today 25, 4: 4–9. Cunningham, H. 2004. Nations Rebound: Crossing Borders in a Gated Globe. Identities 11, 3: 329–350. Cunningham, H. & J. Heyman. 2004. ‘Introduction: Mobilities and Enclosures at Borders’, Identities, 11, 3: 289–302. Cunningham, R. and Y. Sarayrah. 1993. Wasta--The Hidden Force in the Middle Eastern Society. Westport, Connecticut: Greenwood Press. Daily Nation. 1988. 'Somalis flee to Uganda', 15 December. Daily Nation. 2014. ‘Are we just going to sit around and wait to be blown to bits by terrorists?’, 20 March. 256 Dalsan Radio. 2014. ‘Third flight carrying 200 Somali citizens arrive Mogadishu [sic] from South Sudan’, 15 January. Dasgupta, P. 1988, ‘Trust as a commodity’ in D. Gambetta (ed.). Trust: Making and Breaking Cooperative Relations, Oxford: Basil Blackwell, 49-72. de Bruijn, M. 2008. The telephone has grown legs: Communication and social change in the margins of African society. Leiden: African Studies Center (Inaugural Address). de Bruijn, M. 2010. ‘Africa Connects: Mobile Communication and Social Change in the Margins of African Society. The Example of the Bamenda Grassfields, Cameroon’, in M. Fernández-Ardèvol and A. Ros Híjare (eds.). Communication Technologies in Latin America and Africa: A Multidisciplinary Perspective. Barcelona: IN3, 167-191. de Bruijn, M. & I. Brinkman. 2012 ‘Research practice in connections: travels and methods’, in: M. de Bruijn and R. van Dijk (eds.). The Social Life of Connectivity in Africa. Palgrave/Macmillan, 45-63. de Bruijn, M., R.van Dijk & D. Foeken. 2001. Mobile Africa: changing patterns of movement in Africa and beyond. Leiden: Brill. Declich, F. 1987 “I Goscia del Medio Giuba. Una etnia di origine Bantu”, Africa, 4. Declich, F. 2000. ‘Fostering Ethnic Reinvention: Gender Impact of Forced Migration on Bantu Somali Refugees in Kenya’, Cahiers d’Etudes Africaines 157: XL-1. De Genova, N.P. 2002. ‘Migrant ‘illegality’ and deportartability in everyday life’, Annual Review of Anthropology 31, 1: 419–447. De Genova, N.P. & N. Peutz (Eds.) 2010. The Deportation Regime: Sovereignty, Space, and the Freedom Movement. Durham, NC: Duke University Press. De Haas, H. 2008. ‘Migration and Development: A Theoretical Perspective’, International Migration Review 44, 1: 227–264. De Haas, H. 2009. Migration System Formation and Decline: A theoretical inquiry into the self-perpetuating and self-undermining dynamics of migration processes, IMI, Oxford. 257 De Haas, H. 2010. ‘Migration System Formation and Decline: A theoretical inquiry into the self-perpetuating and self-undermining dynamics of migration processes’, Journal of Ethnic and Migration Studies 36, 10: 1587–1617. De Herdt, T. & Marysse, S. 1999. ‘The invention of the market from below : Reinvention the Kinshasa Monopoly’, Review of African Political Economy 26, 80: 239–253. Del Boca, A. 1984. Gli Italiani in Africa Orientale Vol.4: Nostalgia delle colonie. Bari: Laterza. Deleuze, G. & F. Guattari. 1972. L'Anti-Oedipe. Paris: Les Editions de Minuit. Deleuze, G. & Guattari. F. 1986. Nomadology: The war machine .., Demirguc-Kunt, A., & L. Klapper. 2012. Financial Inclusion in Africa An Overview. Policy Research Working Paper 6088, World Bank, Washington, DC. Department for International Development.1999. Sustainable Livelihoods Guidance Sheets, DFID, United Kingdom. De Waal, A. 1997. Famine Crimes: Politics and the Disaster Relief Industry in Africa. Bloomington: International African Institute. De Waal, A. 2007. ‘Class and power in a stateless Somalia’, Social Science Research Council 20: 1–14. Douglas, M. & B. Isherwood. 1996. The World of Goods. London: Routledge. Dowla, A. 2006. ‘In Credit We Trust: Building Social Capital by Grameen Bank in Bangladesh’, Journal of Socio-Economics 35, 1: 102–122. Dryden-Peterson, S. 2006. ‘”I Find Myself as Someone Who is in the Forest”: Urban Refugees as Agents of Social Change in Kampala, Uganda’, Journal of Refugee Studies, 19, 3: 381–395. Dryden-Peterson, S. & L. Hovil. 2003. Local integration as a durable solution: refugees, host populations and education in Uganda. New Issues in Refugee Research, Working Paper no. 93, UNHCR, Geneva. Drysdale, J. 1994. Whatever Happened to Somalia? London: Haan Associates. 258 Dudley, S., 2011. ‘Feeling at Home: Producing and Consuming Things in Karenni Refugee Camps on The Thai-Burma Border’, Population, Space and Place 17, 6: 742–755. Duncombe, R. 2011. Researching impact of mobile phones for development: concepts, methods and lessons for practice, Manchester: University of Manchester, Institute for Development Policy and Management. East Africa Standard. 2002. 'Eastleigh: Kenya's Somali Capital', 4 November. Elliot, H. 2014. ‘Somali displacements and shifting markets: camel milk in Nairobi’s Eastleigh Estate’, in A Hammar (ed.). Displacement Economies: Paradoxes of Crisis and Creativity in Africa. Uppsala, London and New York: Zed Books, 127-145. Elmi, A. 2010. Understanding the Somalia Conflagration. Oxford: Pambazuka Press. Elyachar, J. 2003. ‘Mappings of Power: The State, NGOs, and International Organisations in the Informal Economy of Cairo’, Comparative Studies in Society and History 45, 03: 571–605. Elyachar, J. 2010. Phatic labor, infrastructure, and the question of empowerment in Cairo. Critical Theory Institute, UCI, February 24, unpublished paper. Eno, M.A. 2008. The Bantu—Jareer Somalis: Unearthing Apartheid in the Horn of Africa. London: Adonis and Abbey Publishers Ltd. Eno, O. and M. Eno. 2007. ‘The Journey Back to the Ancestral Homeland: the Return of the Somali Bantu (Wazigwa) to Modern Tanzania’, in A. Kusow & S. R. Bjork (eds.) From Mogadishu to Dixon: the Somali Diaspora in a Global Context. Trenton, NJ: Red Sea Press, 13–43. Eno, M.A. and O. A. Eno. 2009. ‘Intellectualism amid Ethnocentrism : Mukhtar and the 4 .5 Factor’, Bildhaan 9: 137–145. Etang, A. 2010. ‘Analysing the radius of trust in rural Cameroon’, Journal of African Economies 19, 5: 691–717. Etang, A., D. Fielding & S. Knowles. 2011. ‘Trust and ROSCA membership in rural Cameroon’, Journal of International Development 23,4. 259 Etzo, S. and G. Collender. 2010. ‘The mobile phone “revolution” in Africa: Rhetoric or reality?’ African Affairs 109, 437: 659–668. Evans, P. 1996. ‘Government action, social capital and development: Reviewing the evidence on synergy’, World Development 24, 6: 1119–1132. Evers, H.-D. & H. Schrader. 1994. The Moral Economy of Trade: Ethnicity and Development Markets. London, New York: Routledge. Ezeoha, A.E. 2013. ‘Financial Determinants of International Remittance Flows to the Sub-Saharan African Region’, International Migration 51: e84–e97. Faist, T. 2005. The Migration-Security Nexus: International Migration and Security before and after 9/11. Working Paper, No. 9. University of Bielefeld. Centre on Migration, Citizenship and Development (COMCAD). Faist, T. 2010. ‘Towards Transnational Studies: World Theories, Transnationalisation and Changing Institutions’, Journal of Ethnic and Migration Studies 36, 10: 1665–1687. Falzon, M.-A. 2009. Multi-Sited Ethnography: Theory, Praxis and Locality in Contemporary Social Research, Farnham: Ashgate. Farah, N. 2010. ‘Of Tamarind and Cosmopolitanism’, in N. Edjabe & E. Pieterse (eds.) African Cities Reader I: Pan- African Practices. African Centre for Cities & Chimurenga Magazine, 9,16. Farah, O., M. Muchie & J. Gundel (Eds.). From Mogadishu to Dixon: Conceptualizing the Somali Diaspora in Somalia: Diaspora and State Reconstitution in the Horn of Africa. London: Adonis & Abbey. Farr, J. 2004. ‘Social capital: A conceptual history’, Political Theory 32, 1: 6–33. Fassin, D. 2007. ‘Humanitarianism as a politics of life’, Public Culture 19, 3: 499–520. Fassin, D. 2011. ‘Policing Borders, Producing Boundaries. The Governmentality of Immigration in Dark Times’, Annual Review of Anthropology, 40: 213–226. Flahaux, M.-L. & H. De Haas. 2014. African Migration. Exploring the role of development and states, IMI Working Papers Series No. 105 International 260 Migration Institute (IMI), Oxford Department of International Development (QEH), University of Oxford. Food Security and Nutrition Analysis Unit, 2013. Family Ties: Remittances and Livelihoods Support to Puntland and Somaliland. Nairobi: FSNAU/Food and Agricultural Organisation, www.fsnau.org. Foucault, M. 2007. Security, territory, population: lectures at the College de France 1977–1978. London: Palgrave Macmillan. The Economist. 2011. ‘Mobile telecoms in Africa: Digital revolution’, 7th April. Available at http://www.economist.com/node/18529875. [accessed 11 May 2015]. Gagliardone, I. & N. Stremlau. 2012. Mapping Digital Media: Digital Media, Conflict and Diasporas in the Horn of Africa. Open Society Foundations. Geertz, C. 1962. ‘The Rotating Credit association: A `middle rung’ in development’, Economic Development and Cultural change 10: 241-263. Geleta, E. B. 2014. ‘Social Capital as Collateral: Banking on the Poor’, American Journal of Economics and Sociology 73, 1: 108–125. Gersony, R. 1990. ‘Why Somalis flee a synthesis of conflict experience in northern Somalia by Somali refugees, displaced persons and others’, International Journal of Refugee Law 2, 1: 4–55. Geshekter, L. 1985. ‘Anti-Colonialism and Class Formation : The Eastern Horn of Africa before 1950’, The International Journal of African Historical Studies 18, 1: 1–32. Ghatak, M. 1999. ‘Group lending, local information and peer selection’, Journal of Development Economics 60,1: 27-50. Giridharadas, A. 2010. ‘Where a Cellphone Is Still Cutting Edge’, The New York Times, 9 April Glick Schiller, N. & N.B. Salazar. 2013. ‘Regimes of Mobility Across the Globe’, Journal of Ethnic and Migration Studies 39, 2: 183–200. Gielis, R. 2009. ‘A Global Sense of Migrant Places: Towards a Place Perspective in the Study of Migrant Transnationalism’, Global Networks 9, 2: 271- 287. 261 Goldring, L. 2004, ‘Family and collective remittances to Mexico: a multi- dimensional typology’, Development and Change, 35, 2: 799–840. Goldsmith, P. 1997. ‘The Somali Impact on Kenya 1990-1993: The View From Outside the Camps’, in H.M. Adam and R.Ford (eds.) Mending Rips in the Sky: Options for Somali Communities in the 21st Century, Lawrenceville, NJ: Red Sea Press: 461-483. Goodfellow, T. & K.Titeca. 2012. ‘Presidential intervention and the changing “politics of survival” in Kampala’s informal economy’, Cities 29, 4: 264–270. Goginiani, R. 2012. ‘Police Raids Alienate Kenya's Largest Ethnic Minority’, Voice of America, November 23. Available on http://www.voanews.com/content/violent-eastleigh-police-raids-alienating-kenyas- ethnic-somali-minority/1551714.html [accessed 12 April 2015] Gouldner, A. W. 1960. 'The norm of reciprocity', American Sociological Review, 25: 161-78. Grabher G. 1996. ‘Adaptation at the cost of adaptability? Restructuring the East German regional economy’, in Grabher, G. and D. Stark. (Eds). Restructuring Networks in Post-socialism: Legacies, Linkages and Localities Oxford: Oxford University Press, 107- 34. Grabher, G. and D. Stark. 1997. ‘Organizing Diversity: Evolutionary Theory, Network Analysis and Postsocialism’, Regional Studies 31, 5: 533–544. Grabowski, R. 1997. ‘Traders’ dilemmas and development: A variety of solutions’, New Political Economy 2, 3: 387–404. Grabska, K. 2006. ‘Marginalisation in Urban Spaces’, Journal of Refugee Studies 19, 3: 287–307. Graeber, D. 2011. Debt: the first 5,000 years, Brooklyn: Melville House Publishing. Grant, K., R.Wangui Kimotho & E. Gerstner. 2012. Asylum under threat: Assessing the protection of Somali refugees in Dadaab. Refugee Consortium of Kenya, June. Granovetter, M. 1973. ‘The Strength of Weak Ties’, American Journal of Sociology 78, 6: 1360-1380. 262 Grayson, C.-L., 2014. Civilising Kakuma: shared experience, refugee narratives and the constitution of a community. International Journal of Migration and Border Studies, 1, 1: 109–126. Grbac, P. 2013. Reimagining the refugee camp as the city, Working Paper Series 96. Refugee Studies Centre, Oxford. Grenier, C. 2011. ‘Migration, Translocal Networks and Socio-Economic Stratification in Namibia’, Africa, 81, 4: 606–627. Gudeman, S. 2001. The anthropology of economy: Community, market, and culture. Malden: Blackwell. Guérin, I., 2014. ‘Juggling with Debt, Social Ties, and Values’, Current Anthropology 55, S9: S40–S50. Gueron, J. and A. M. Spevacek. 2008. Diaspora-development nexus: The role of ICTs. USAID Knowledge Services Center. Available at: http://www.unitar.org/ny/sites/default/files/Hannafin%20ICT%20Diasporas.pdf [accessed 18 January 2015] Gundel, J. 2009. Clans in Somalia. Austrian Centre, December. Available at: http://www.refworld.org/docid/4b29f5e82.html. [accessed 21 April 2015]. Guyer, J.I. 1981. ‘Household and community in African studies’, African Studies Review 24, 3/4: 87-137. Guyer, J. 2004. Marginal Gains: Monetary Transactions in Atlantic Africa. Chicago: University of Chicago Press. Guyer, J. (ed.) 1995. Money matters: instability, values and social payments in the modern history of West African communities. London: Currey. Guyer, J.I. 2013. ‘“The quickening of the unknown”: Epistemologies of surprise in anthropology’, HAU: Journal of Ethnographic Theory, 3, 3: 283–307. Hage, G. 2005. ‘A not so multi-sited ethnography of a not so imagined community’, Anthropological Theory 5, 4: 463–475. Hagmann, T. and M. Höehne. 2009. ‘Failures of the state failure debate: evi- dence from the Somali territories’, Journal of International Development 21, 1. Hall, S. 2012. ‘Geographies of money and finance III: Financial circuits and the “real economy”’, Progress in Human Geography 1–8. 263 Halperin, R.H. 1990. The Livelihood of Kin Making Ends Meet "The Kentucky Way" Austin: University of Texas Press. Hammar, A. 2014. ‘Displacement Economies: Paradoxes of Crisis and Creativity in Africa’, in A. Hammar (ed.). Displacement Economies: Paradoxes of Crisis and Creativity in Africa. Uppsala, London and New York: Zed Books, 3-34). Hammond, L. 2013a. ‘Somali Transnational Activism and Integration in the UK: Mutually Supporting Strategies’, Journal of Ethnic and Migration Studies 39:6, 1001-1017 Hammond, L. 2013b. ‘Somalia rising: things are starting to change for the world’s longest failed state’, Journal of Eastern African Studies 7, 1: 183–193. Hammond, L. 2014. History, overview, trends and issues in major Somali refugee displacements in the near region. New Issues in Refugee Research Working Paper No. 268. Geneva: UNHCR. Hannan, M. T. 1986. Uncertainty, diversity, and organisational change, in Behavioral and Social Sciences: Fifty Years of Discovery. Washington, DC: National Academy Press. Hannertz, U. 1980. Exploring the City: Inquiries Toward an Urban Anthropology. New York: Columbia University Press. Hannerz, U. 1990. ‘Cosmopolitans and locals in world culture’, Theory, Culture, and Society 7, 2/3: 237-252. Hannerz, U. 1996. Transnational Connections: Culture, People, Places. London: Routledge Hannertz, U. 1997. ‘Fluxos, fronteiras, híbridos: palavras-chave da antropologia transnacional’, Mana, 3, 1: 7–39. Hannertz, U. 2003. ‘Being there . . . and there . . . and there! Reflections on multi-site ethnography’, Ethnography 4, 2: 201–216. Hansen, K. T., & M. Vaa. 2004. Reconsidering Informality: Perspectives from Urban Africa, Uppsala: Nordiska Afrikainstitutet. Hansen, S. H. 2013a. Al-Shabaab in Somalia: The history and ideology of a militant Islamist group, 2005–2012. Hurst, London. 264 Hansen, S. H. 2013b. ‘Opposing interests? The geopolitics of the Horn of Africa’, in E. Leonard & G. Ramsay (Eds.). Globalizing Somalia: Multilateral, International, and Transnational Repercussions of Conflict. New York, NY: Bloomsbury Academic. Hardt, M. & A. Negri. 2000. Empire. Cambridge, MA: Harvard University Press. Hardy C., P. Phillips & T. Lawrence. 1998. ‘Distinguishing Trust and Power in Interorganizational Relations: Forms and facades of Trust’, in C. Lane & R. Bachmann (eds). Trust within and between Organizations Oxford: Oxford University Press, 65-87. Harper, M. 2012. Getting Somalia Wrong?: Faith and War in a Shattered State. London: Zed Books Harrell-Bond, B. 1986. Imposing Aid: Emergency Assistance to Refugees, New York: Oxford University Press. Harrell-Bond, B. 2004.‘Response to Kibreab’, Journal of Refugee Studies 17, 1: 27–8. Hart, K. 1973. ‘Informal Income Opportunities and Urban Employment in Ghana’, Journal of Modern African Studies 11:61–89. Hart, K. 1988, ‘Kinship, contract, and trust: the economic organization of migrants in an African city slum’ in D. Gambetta (ed.) Trust: Making and Breaking Cooperative Relations, Oxford: Basil Blackwell. Hart, K. 2007b. ‘The African revolution: urban commerce and the informal economy’, Economia Global e Gestão 12, 3: 93–104. Harvey, D. 2001. ‘Globalisation and the “Spatial Fix”’, Geographische Revue 2: 23–30. Harvey, D. 2006. Spaces of Global Capitalism. London: Verso. Hashim, A. 1997. ‘Conflicting Identities in Somalia’, Peace Review 9, 4: 527–532. Hess, R. L. 1966. Italian Colonialism in Somalia. The University of Chicago Press. 265 Hesse, B. 2010. ‘Where Somalia works’, Journal of Contemporary African Studies 28, 3: 343–362. Heyman, J.M. & H. Campbell. 2009. ‘The anthropology of global flows: A critical reading of Appadurai’s `Disjuncture and Difference in the Global Cultural Economy', Anthropological Theory 9, 2:131–148. Hilhorst, D. & B.J. Jansen. 2010. ‘Humanitarian Space as Arena: A Perspective on the Everyday Politics of Aid’, Development and Change 41, 6:1117–1139. Hill, M. 2010. “No Redress: Somalia’s Forgotten Minorities,” Minority Rights Group International, November, Available at http://www.minorityrights.org/?lid=10370 [accessed 25 June 25, 2015]. Hoben, A. 1988. ‘The Political Economy of Land Tenure in Somalia’, in R.E. Downs & S.P. Reyna, (eds.), Land and Society in Contemporary Africa Hanover: University Press of New England. Hoehne, M.V. 2006. ‘Somalia: the changing spectrum of Islam and counterterrorism’, Horn of Africa Bulletin, 20–26. Höhne, M. V. 2014. Continuities and changes regarding minorities in Somalia, Ethnic and Racial Studies, 38:5, 792-807. Hoehne, M.V., 2015. Between Somaliland and Puntland, Rift Valley Institute, Nairobi. Hoehne, M.V., D.Feyssa, & M. Abdile. 2011. ‘Somali and Ethiopian diasporic engagement for peace in the Horn of Africa’, African Conflict & Peacebuilding Review, 1, 1: 71–99. Holt, L. 2008. ‘Embodied social capital and geographic perspectives: performing the habitus’, Progress in Human Geography, 32, 2: 227–246. Horst, C. 2002. Hawala: the importance of overseas connections in the livelihoods of Somali refugees in the Dadaab refugee camps of Kenya, Working Paper, University of Amsterdam. Horst, C. 2006a. ‘Buufis amongst Somalis in Dadaab’, The Journal of Refugee Studies 19, 2. Horst, C. 2006b. ‘Introduction: Refugee Livelihoods: Continuity and Transformations’, Refugee Survey Quarterly 25, 2: 6–22. 266 Horst C. 2008. Diaspora Engagements in Development Cooperation, Policy Brief 8/2008. Oslo: PRIO. Horst, C. & K. Grabska. 2015. ‘Flight and Exile—Uncertainty in the Context of Conflict-Induced Displacement’, Social Analysis 59, 1: 1–18. Horst, C. & N. Van Hear. 2002. ‘Counting the cost: refugees, remittances and the ’war against terrorism’’, Forced Migration Review 14. Horst, H.A. & D. Miller. 2006. The Cell Phone: An Anthropology of Communication. Oxford: Berg. Hovil, L. 2010. Hoping for Peace, Afraid of War: The Dilemmas of Repatriation and Belonging on the Borders of Uganda and South Sudan. New Issues in Refugee Research, Paper No. 196. Geneva: UNHCR,. Human Rights Watch. 2014. Kenya: End Abusive Round-Ups Detainees Describe Mistreatment, Lack of Access to UN Agency. 21 May. Available at https://www.hrw.org/news/2014/05/12/kenya-end-abusive-round-ups [accessed 12 February 2015] Hughes, C. 2015. ‘Half a million refugees gather in Libya to attempt perilous crossing to Europe’, The Guardian, 6 June. Available at http://www.theguardian.com/world/2015/jun/06/cameron-merkel-at-odds-resettle- refugees-europe-migration [accessed 16 August 2015] Hughes, N. & S. Lonie. 2007. ‘M-PESA: Mobile money for the “unbanked”: Turning cellphones into 24-hour tellers in Kenya’, Innovations: Technology, Governance, Globalization 2, 1–2: 63–81. Human Rights Watch. 2010. Welcome to Kenya: Police Abuse of Somali Refugees. Available at www.hrw.org/reports/2010/06/17/welcome-kenya-0 [accessed on 03 February. 2015] Human Rights Watch. 2013. Hostages of Gatekeepers: Abuses against Internally Displaced in Mogadishu, Somalia. March. Available at http://www.hrw.org/sites/default/files/reports/somalia0313_ForUpload.pdf [accessed 25 April 2015]. Human Rights Watch. 2014. The Power These Men Have Over Us” Sexual Exploitation and Abuse by African Union Forces in Somalia. September. Available at https://www.hrw.org/report/2014/09/08/power-these-men-have-over-us/sexual- exploitation-and-abuse-african-union-forces [accessed 21 April 2015]Humphreys, 267 L. 2010. ‘Mobile social networks and urban public space’, New Media & Society 12, 5: 763–778. Humphrey, J. & H.Schmitz. 1998. ‘Trust and inter-firm relations in the developing and transition economies’. The Journal of Development Studies, 34, 4: 32-61. Hyndman, J. 2000 Managing Displacement: Refugees and the Politics of Humanitarianism. Minneapolis: University of Minnesota Press. Hyndman, J. 2002. ‘Business and bludgeon at the border: A transnational political economy of human displacement in Thailand and Burma’, GeoJournal 56, 1: 39–46. Hyndman, J. 2011. ‘A refugee camp conundrum: Geopolitics, liberal democracy, and protracted refugee situations’, Refuge 28, 2: 7–15. Hyndman, J., 2012. ‘The Geopolitics of Migration and Mobility’, Geopolitics, 17, 2: 243–255. Ibrahim, G. & V. Galt. 2003. ‘Ethnic Business Development: Toward a Theoretical Synthesis and Policy Framework’, Journal of Economic Issues, 37,4: 1107–1119. Igarashi, H. & H.Saito. 2014. ‘Cosmopolitanism as Cultural Capital: Exploring the Intersection of Globalisation, Education and Stratification’, Cultural Sociology, 8, 3: 222–239. ILO. 1972. Employment, Incomes, and Equality: A Strategy for Increasing Productive Employment in Kenya. Geneva: International Labour Organization. Inda, J. X., & R. Rosaldo. 2002. ‘Introduction’, in J. Inda and R. Rosaldo (eds.), A World in Motion. The Anthropology of Globalisation Blackwell Publishing, 1-34. International Crisis Group, 2011. Somalia: The Transitional Government on Life Support, Africa Report 170, February. International Crisis Group, 2012. Assessing Turkey’s Role in Somalia. October. IRIN. 1999. Kenya: Police arrest 600 in security raids focused on illegal aliens. 3 September. Available at http://www.irinnews.org/report/8881/kenya- 268 police-arrest-600-in-security-raids-focused-on-illegal-aliens [accessed 01 April 2015]. IRIN. 2002. Kenya-Somalia: UNHCR Preparing to Move Somali Bantus to Kakuma, 9 May. Available at http://www.irinnews.org/report/31830/kenya- somalia-unhcrpreparing-to- move-somali-bantus-to-kakuma [accessed 11 May 2015]. Isotalo, R.2009. ‘Politicizing the transnational. On Implications for Migrants, Refugees, and Scholarship’, Social Analysis 53,3: 60–84. Issa-Salwe, A.. 1994. The Collapse of the Somali State – The Impact of the Colonial Legacy. London: Haan Associates. Ito, M. & D. Okabe. 2005. ‘Technosocial Situations: Emergent Structurings of Mobile Email Use’, in M. Ito, D, Okabe, and M. Matsuda, eds. Personal, Portable, Pedestrian: Mobile Phones in Japanese Life. Cambridge: MIT Press. Jaji, R., 2012. ‘Social technology and refugee encampment in Kenya’, Journal of Refugee Studies 25, 2: 221–238. Jaji, R., 2013. ‘Somali Asylum Seekers and Refoulement at the Kenya– Somalia Border’, Journal of Borderlands Studies, 28, 3: 355–368. Jamal, A. 2000. Minimum Standards and Essential Needs in a Protracted Refugee Situation: A Review of the UNHCR Programme in Kakuma, Kenya, edited by Evaluation and Policy Analysis Unit, United Nations High Commissioner for Refugees, Geneva. Jamal, S. & A. Sandor, 2010. Temporarily Permanent: Agamben and Palestinian Refugee Camps in Lebanon. Annual Conference of the Canadian Political Science Association. James, D. 2015. Money from Nothing. Indebtedness and Aspirations in South Africa. Stanford University Press. Jansen, B.J., 2008. Between vulnerability and assertiveness: Negotiating resettlement in Kakuma refugee camp, Kenya. African Affairs, 107(429), pp.569– 587. Jaramogi, P. 2008. ‘Uganda: Army Warns Somali on Terror”, The New Vision, 12 October. 269 Jaramogi, P. and Wagawulo, C. 2008. ‘Uganda: City Explosion Kills Two in Kisenyi’, The New Vision, 7 November. Jok, J.M. & S.E. Hutchinson, 1999. ‘Sudan’s prolonged second civil war and the militarization of Nuer and Dinka ethnic identities’, African Studies Review, 42, 2: 125–145. Johnson, S. & M. Nino-Zarazua, 2011. Financial access and exclusion in Kenya and Uganda. Journal of Development Studies, 47, 3: 475–496. Jonsson, G. 2009. Comparative report: African Migration Trends. African Perspectives on Human Mobility Programme. International Migration Institute, Oxford. Kaiser, T. 2006. ‘Between a Camp and a Hard Place: Rights, Livelihood and Experiences of the Local Settlement System for Long-term Refugees in Uganda’, Journal of Modern African Studies 44: 597-621. Kaiser, T., 2010. ‘Dispersal, division and diversification: durable solutions and Sudanese refugees in Uganda’, Journal of Eastern African Studies 4, 1: 44–61. Kajoba, N. 2009. ‘Uganda: Somalis Commend Peacekeepers’, The New Vision 29 July. Kantai, P. 2013. ‘The Rise Of Somali Capital’, The Chimurenga Chronic Vol. 16. Available at http://chimurengachronic.co.za/the-rise-of-somali-capital-2/ [accessed on 12 Feb. 2015]. Kaplan, R. 1994. ‘The Coming Anarchy’. The Atlantic February. Available at http://www.theatlantic.com/magazine/archive/1994/02/the-coming- anarchy/304670/ [accessed on 15 August 2015]. Kapteijns, L. 2013. Clan Cleansing in Somalia: The Ruinous Legacy of 1991. Philadelphia: University of Pennsylvania Press. Kassim, M.M. 1995. ‘Islam and Swahili Culture on the Banadir Coast’, Northeast African Studies 2, 3: 21-37. Kaufmann, V. & M.M. Bergman. 2004. ‘Motility : Mobility as Capital’, International Journal of Urban and Regional Research 28: 745–756. Kearney, M. 1991. Borders and boundaries of states and self at the end of empire. Journal of Historical Sociology 4: 52–74. 270 Kendall, J., B. Maurer, P.Machoka, & C.Veniard. 2012. An Emerging Platform: from Money Transfer System to Mobile Money Ecosystem Innovations. School of Law, University of California Irvine, Legal Studies Research Paper Series No. 2011-14, available at: http://www.gsma.com/ connectedwomen/wp- content/uploads/2012/06/ssrnid1830704.pdf (accessed 9 April, 2015). Kesserling, R. 2015. Moments of Dislocation: Why the Body Matters in Ethnographic Research, Working Paper 8, Institute of Social Anthropology University of Basel. Khorshid, A.A.R., 2001. Islamic Insurance. A modern approach with particular reference to Western and Islamic Banking. The University of Leeds. Unpublished PhD Thesis. Kibreab, G., 1991. ‘Stranded ’Birds of Passage'? Eritean and Ethiopian Refugees in Khartoum’, Refuge: Canada’s periodical on refugees, 10, 4. King, R. 2012. Theories and Typologies of migrations: An overview and a primer. Willy Brandt Series of Working Papers in International Migration and Ethnic Relations 3 King, R., & R. Skeldon. 2010. ‘“Mind the Gap!” Integrating Approaches to Internal and International Migration’, Journal of Ethnic and Migration Studies 36, 10: 1619–46. Kleist, N. 2004. Nomads, sailors and refugees. A century of Somali migration. No. 23, Sussex Migration Working Paper. Sussex: University of Sussex. Centre for Migration Research. Kloosterman, R., Van Der Leun, J. & J. Rath. 1999. Mixed embeddedness: (in) formal economic activities and immigrant businesses in the Netherlands. International journal of urban and regional research, 23, 2: 252–266. Kloosterman, R. & J. Rath. 2001. Immigrant entrepreneurs in advanced economies. Journal of Ethnic and Migration Studies, 27, 2: 189–201. Kloosterman, R.C., 2010. Matching opportunities with resources: A framework for analysing (migrant) entrepreneurship from a mixed embeddedness perspective. Entrepreneurship & Regional Development, 22,1: 25–45. Kolm, S.C. 2008. Reciprocity An economics of social relations. Cambridge University Press, Cambridge. 271 Korac, M. 2009. Remaking Home: Reconstructing Life, Place and Identity in Rome and Amsterdam. New York: Berghahn. Kumar, K., C. McKay & S. Rotman, 2010. Microfinance and Mobile Banking: The Story So Far. Focus Note 62. Washington, D.C.: CGAP, July. http://www.cgap.org/gm/ document-1.9.45546/FN62_rev2.pdf Kumsa, M.K. 2006. ‘“No! I”m not a refugee!’ The poetics of be-longing among young Oromos in Toronto’. Journal of Refugee Studies, 19, 2: 230–255. Kurtulus, E.N., 2012. ‘Exploring the Paradoxical Consequences of State Collapse: The Cases of Somalia 1991-2006 and Lebanon 1975-1982’, Third World Quarterly, 33, 7: 1285–1303. Kusimba, S. 2013. Social Networks of Mobile Money in Kenya. Institute for Money, Technology and Financial Inclusion, University of California Irvine. Kusow, A. M.(ed.). 2004. Putting the Cart Before the Horse: Contested Nationalism and the Crisis of the Nation-State in Somalia Trenton: Red Sea Press. Laitin, D., 1976. ‘The Political Economy of Military Rule in Somalia’, The Journal of Modern African Studies 14, 3: 449–468. Laitin, D., & S. Samatar. 1987. Somalia. Nation in Search of a State. Boulder, CO: Westview Press. Lambo, I. 2012. In the shelter of each other: notions of home and belonging amongst Somali refugees in Nairobi, New Issues in Refugee Research Working Paper No. 233. Geneva: UNHCR. Landau, L.B. & I. Freemantle. 2010. ‘Tactical Cosmopolitanism and Idioms of Belonging: Insertion and Self-Exclusion in Johannesburg’, Journal of Ethnic and Migration Studies, 36, 3: 375–390. Lareau, A. & E. B. Weininger .2003. Cultural capital in educational research: A critical assessment. Theory and Society 32: 567–606. Law, J. 1992. ‘Notes on the Theory of the Actor-Network: Ordering, Strategy and Heterogeneity’, Systems Practice, 5: 379-93. Lee, M. 2014. Africa’s World Trade: Informal Economies and Globalisation from Below. Zed Books: London. 272 Lessault, D. & M.L. Flahaux. 2013. ‘Regards statistiques sur l’histoire de l’émigration internationale au Sénégal’, Revue européenne des migrations internationales 29, 4: 59-88. Levitt, P. 1998. ‘Social remittances: migration-driven local-level forms of cultural diffusion’, International Migration Review 32, 4: 926-48. Levitt, P. & D. Lamba-Nieves. 2011. ‘Social Remittances Revisited’, Journal of Ethnic and Migration Studies 37, 1: 1–22. Lewis, I. M. 1972. “The Politics of the 1969 Somali Coup.” The Journal of Modern African Studies 10, 3: 383–408 Lewis, I. M. 1994. Blood and Bone: the call for kinship in Somali society. Lawrenceville, NJ: Red Sea Press. Lewis, I.M. 1999. A Pastoral Democracy. London, Oxford University Press. Lewis, I. M. 2002. A Modern History of the Somali: Nation and State in the Horn of Africa. 4th ed. Eastern African Studies. Oxford: James Currey. Lewis, I.M., 2010. Making and Breaking States in Africa: The Somali Experience. Trenton, New Jersey: Red Sea Press. Leyshon, A. 1995. ‘Geographies of money and finance I’, Progress in Human Geography 19: 531–543. Leyshon, A. 1997. ‘Progress reports. Geographies of money and finance II’. Progress in Human Geography, 21, 3: 381–392 Leyshon, A. 1998. ‘Geographies of money and finance III’. Progress in Human Geography 22, 3: 433–446. Linde, T., 2011. ‘Mixed migration - a humanitarian counterpoint’, Refugee Survey Quarterly 30, 1: 89–99. Lindell, I. 2010. ‘Introduction: the changing politics of informality – collective organizing, alliances and scales of engagement’, in I. Lindell (Ed) Africa’s informal workers: collective agency, alliances and transnational organizing Zed Books, London. Lindley, A. 2007. Protracted displacement and remittances: the view from Eastleigh, Nairobi. New Issues in Refugee Research Working Paper No. 143. Geneva: UNHCR. 273 Lindley, A. 2009. ‘Between “dirty money” and “development capital”: Somali money transfer infrastructure under global scrutiny’, African Affairs, 108, 433: 519–539. Lindley, A. 2010. The early morning phone call: Somali refugees' remittances Oxford: Berghahn. Lindley, A., 2011. Between a protracted and a crisis situation: Policy responses to somali refugees in Kenya. Refugee Survey Quarterly, 30(4), pp.14–49. Lindley, A. 2013. ‘Displacement in contested places: governance, movement and settlement in the Somali territories’, Journal of Eastern African Studies 7,2: 291–313. Ling, R. & Horst , H.A.. 2011. ‘Mobile communication in the global south.’ New Media Society 13: 363–374. Lins Ribeiro, G., 2009. ‘Non-hegemonic globalisations Alter-native transnational processes and agents’, Anthropological Theory 9, 3: 297–329. Little, P. 2003. Somalia: Economy without State. Bloomington, IN: Indiana University Press. Little, P. 2013. Economic and Political Reform in Africa: Anthropological Perspectives. Bloomington and Indianapolis: Indiana University Press. Lochery, E. 2012. ‘Rendering difference visible: The Kenyan state and its Somali citizens’, African Affairs 111, 445: 615–639. Lorch, D. 1995. ‘In Somalia, a Chameleon Thrives’, The New York Times, Monday July 31. Available at: http://www.nytimes.com/1995/07/31/world/in- somalia-a-chameleon-thrives.html [accessed 10 February 2015] Lourenço-Lindell, I. 2001. ‘Social Networks and Urban Vulnerability to Hunger’, in A. Tostensen, I. Tvedten & M. Vaa (eds.) Associational Life in African Cities: Popular Responses to the Urban Crisis, 30-45, Uppsala: Nordic Africa Institute. Lourenço-Lindell, I., 2002. Walking the Tight Rope: Informal Livelihoods and Social Networks in a West African City, Stockholm: Stockholm Studies in Human Geography 9, Almqvist and Wiksell International. 274 Lyons, T. & A.Samatar. 1996. ‘Somalia: State Collapse, Multilaterol lntervention and Strategies for Political Reconstruction’, Washington, DC: Brookings Institution. Luling, V. 1976. ‘Colonial and Postcolonial Influences on a South Somali Community’. Journal of African Studies 3.4 : 491. Luling, V. 1984. ‘The other Somali. Minority Groups in Traditional Somali Society’, in T. Labahn (ed.) Proceedings of the Second International Congress of Somali Studies Hamburg, Germany: Helmut Buske Verlag. 4: 39–56. Luling, V., 2006. ‘Genealogy as theory, genealogy as tool: aspects of Somali clanship’, Social Identities 12, 4: 471–485. Lund, C. 2006.‘Twilight institu- tions: public authority and local politics in Africa’, Development and Change 37, 4. McFarlane, C. 2009. ‘Translocal assemblages: Space, power and social movements’, Geoforum 40, 4: 561-567. MacDonald, J. S., & L. D. Macdonald. 1964. ‘Chain migration, ethnic neighborhood formation, and social networks’, Millbank Memorial Fund Quarterly 42. MacGaffey, J. & R.. Bazenguissa-Ganga. 2000. Congo-Paris: Transnational Traders on the Margins of the Law. Oxford and Bloomington IN: James Currey. Mahmoud, H.A. 2009. ‘Breaking the communication barrier using mobile phones in pastoral livestock marketing’, Haramata 54. Majid, N. & S. McDowell. 2012. ‘Hidden dimensions of the Somalia famine’, Global Food Security 1, 1: 36–42. Malkki, L. 1995 “Refugees and Exile: From ‘Refugee Studies’ to the National Order of Things”, Annual Review of Anthropology 24. Malkki, L. 2002. ‘News from nowhere Mass displacement and globalised “problems of organisation”, Ethnography 3, 3: 351–360. Mamdani M. 2001. When Victims Become Killers: Colonialism, Nativism, and the Genocide in Rwanda Princeton: Princeton University Press Ma Mung, E. 2004. ‘Dispersal as a Resource’,.Diaspora: A Journal of Transnational Studies 13, 2/3: 211-225 275 Mann, L. 2014. ‘Wasta! The long-term implications of education expansion and economic liberalisation on politics in Sudan’, Review of African Political Economy 41,142: 561-578 Manson, K., 2012. ‘Big money in little Mogadishu’, Financial Times. Available at: http://www.ft.com/cms/s/0/d4979b30-bf98-11e1-bb88- 00144feabdc0.html#axzz2L3p1DnZ5. [accessed 10 June 2015]. Mansur, A. 1995. ‘The nature of the Somali clan system’, in A. Ahmed (Ed.), The invention of Somalia Lawrenceville, NJ: Red Sea, 117–134. Mantzikos, I., 2010. An African version of the Taliban? The Islamic Courts Union in Somalia (2006) and the Taliban Afghanistan (1996). Comparative Islamic Studies, 4(4.1-4.2), pp.113–129. Marchal, R. 1993. ‘Les “mooryaan” de Mogadiscio. Formes de la violence dans un espace urbain en guerre’, Cahiers d’études africaines, 33, 130: 295–320. Marchal, R. 2000. Globalization and its impact on Somalia. Nairobi, UNDOS. Marchal, R. 2005. ‘Dubai: Global City and Transnational Hub’, in M. Al- Rasheed (ed.) Transnational Connections and the Arab Gulf, London: Routledge, 93-110. Marchal, R. 2011. The Rise of a Jihadi Movement in a Country at War harakat al-shabaad al mujaheddin. CNRS SciencesPo Report, March. Marcus, G. 1995. ‘Ethnography in/of the world system: The emergence of multi-sited ethnography’, Annual Review of Anthropology 24: 95–117. Marin, G. 1931. ‘Somali games’, Journal of the Royal Anthropological Institute of Great Britain and Ireland 61: 499–511. Markakis, J. 1987. National and Class Conflict in the Horn of Africa. Cambridge: Cambridge University Press. Martín I. & S. Bonfanti. 2015. Migration and Asylum Challenges in Eastern Africa: Mixed Migration Flows Require Dual Policy Approaches. European University Institute, Migration Policy Centre. Mas, I. & O. Morawczynski. 2009. ‘Designing mobile money services: lessons from M-PESA’, Innovations: Technology, Governance, Globalization 4, 2: 77–91. 276 Massey, D. 1991. ‘A global sense of place’, in S. Daniels & R. Lee (Eds), Exploring Human Geography London: Arnold, pp. 237–245. Massey, D., 1992. Politics and Space / Time. New Left Review, 65–84. Massey, D. 1993. ‘Power-geometry and progressive sense of place’, in J. Bird (Ed.), Mapping the Futures: Local Cultures, Global Change London and New York: Routledge, pp. 280–288. Massey, D.B. 1994. Space, Place, and Gender. Cambridge: Polity. Massey, D. S., J. Arango, G. Hugo, A. Kouaouci, A. Pellegrino, & J. E. Taylor. 1993. ‘Theories of International Migration: A Review and Appraisal’, Population and Development Review 19:431-466. Mathews, G., 2008. ‘Chungking mansions: a center of ’low-end globalisation'’, Ethnology 46, 2: 169–183. Mayer, R.C., J.H. Davis, & F.D. Schoorman. 1995. ‘An Integrative Model of Organisational Trust’, Academy of Management Review 20, 3: 709–734. Maurer, B. 2005. Mutual Life, Limited: Islamic Banking, Alternative Currencies, Lateral Reason, Princeton: Princeton University. Maurer, B. 2008. ‘Resocializing Finance? or Dressing It in Mufti? Journal of Cultural Economy 1, 1: 65–78. Maurer, B. 2010. ‘Finger counting money’, Anthropological Theory 10, 1-2: 179–185. Mauss, M. 1954. The gift: Forms and functions of exchange in archaic societies. Glencoe, IL: The Free Press. Mazzucato, V. et al., 2015. Transnational Families Between Africa and Europe. International Migration Review, 49, 1: 142–172. Meagher, K. 1990. ‘The hidden economy: informal and parallel trade in Northwestern Uganda’, Review of African Political Economy 17, 47: 64–83. Meagher, K., 1995. Crisis, Informalization and the Urban Informal Sector in Sub-Saharan Africa. Development and Change, 26: 259–284. Meagher, K. 2005. Social capital or analytical liability? Social networks and African informal economies. Global Networks, 5, 3: 217–238. 277 Meagher, K. 2006. ‘Social capital, social liabilities, and political capital: Social networks and informal manufacturing in Nigeria’, Africa. The Journal of the International African Institute 105, 421: 553–582. Meagher, K. 2009. Culture agency and power: theoretical reflections on informal economic networks and political process. DIIS Working Papers 27. Danish Institute for International Studies, Copenhagen, Denmark. Meagher, K. 2010. Identity economics: social networks and the informal economy in Nigeria African issues. James Currey, Suffolk, UK. Meagher, K., 2011. ‘Informal economies and urban governance in Nigeria: popular empowerment or political exclusion?’ African Studies Review, 54, 2: 47– 72. Meagher, K. 2013. ‘Unlocking the Informal Economy: A Literature Review on Linkages Between Formal and Informal Economies in Developing Countries’, WIEGO Working Paper No. 27, April. Medani, K. 2011. ‘Informal Networks, Economic Livelihoods, and the Politics of Social Welfare: Understanding the Political and Humanitarian Consequences of the War on Terrorist Finance’, UCLA Journal of Islamic and Near Eastern Law 52: 99-137. Menkhaus, K. 2000. “From Feast to Famine: Land and the State in Somalia’s Lower Jubba Valley”, in Besteman, Catherine and Cassanelli Lee V., (eds.), The struggle for Land in Southern Somalia: The War Behind the War. Haan Associates. Menkhaus, K. 2002. ‘Political Islam in Somalia’, Middle East Policy IX, 1: 109–123. Menkhaus, K., 2003a. ‘Bantu ethnic identities in Somalia’, Annales d’Ethiopie 19, 1: 323–339. Menkhaus, K. 2003b. ‘State Collapse in Somalia: Second Thoughts’, Review of African Political Economy 30, 97: 405-422. Menkhaus, K. 2005. Kenya–Somalia Border Conflict Analysis. Washington, DC: USAID. Menkhaus, K. 2006.‘Governance without government in Somalia: spoilers, state building, and the politics of coping’, International Security 31, 3. 278 Menkhaus, K. 2007. ‘The crisis in Somalia: Tragedy in five acts’, African Affairs 106, 424: 357–390. Menkhaus, K. 2010. ‘The question of ethnicity in Somali studies: the case of Somali Bantu identity’, in M. Hoehne & V. Luling. (Eds.) Peace and Milk, Drought and War: Somali Culture, Society and Politics: Essays in Honour of I.M. Lewis London: Hurst & Company, 87–104. Menkhaus, K. 2012. ‘No access: Critical bottlenecks in the 2011 Somali famine’, Global Food Security 1, 1: 29–35. Menkhaus, K. 2014. ‘Calm between the storms? Patterns of political violence in Somalia, 1950–1980’, Journal of Eastern African Studies 8, 4: 558–572. Menkhoff, T. & C. Labig. 1996. ‘Trading Networks of Chinese Entrepreneurs in Singapore, Sojourn’, Journal of Social Issues in Southeast Asia 11, 1: 130–54. Meyer, S. 2006. The “refugee aid and development” approach in Uganda: empowerment and self-reliance of refugees in practice, New Issues in Refugee Research, Working Paper No. 131, UNHCR Evaluation and Policy Analysis Unit: Geneva. Miller, D. 2006. The Cellphone an anthropology of communication, Oxford, New York: Berg. Milner, J. 2009. Refugees, the State, and the Politics of Asylum in Africa. Basingstoke: Palgrave Macmillan. Mishler, W. & R. Rose. 2001. ‘What are the origins of political trust? Testing institutional and cultural theories in post-Communist societies’, Comparative Political Studies 34, 1: 30-62. Mohamed, J. 2007. ‘Kinship and Contract in Somali Politics’, Africa 77, 02: 226–249. Mohammed, G. 2013. ‘Kenya: The Cost of Harassing Somalis Over Terror’, The Star, 9 Jan. Mohamoud, A. A. 2006. State collapse and post-conflict development in Africa: The case of Somalia (1960-2001). West Lafayette, Indiana: Purdue University Press. Mohamoud, A. O. 2012. Gobannimo: Charting a New Path Toward Conflict Transformation in Somalia Post London Conference, February. 279 Molony, T. 2008. ‘Running out of credit: the limitations of mobile telephony in a Tanzanian agricultural marketing system’, The Journal of Modern African Studies 46, 04: 637. Molony, T. 2012. ‘ICT and human mobility: cases from developing countries and beyond’, Information Technology for Development, 18, 2: 87–90. Monsutti, A. 2004. ‘Cooperation, Remittances, and Kinship Among the Hazaras’, Iranian Studies 37, 2: 219–240. Monsutti, A, 2013. Anthropologizing Afghanistan: Colonial and postcolonial encounters. Annual Review of Anthropology, 42(1), pp.269–285. Mukhtar, M. 1995. ‘Islam in Somali history: Fact and fiction’, in A. Ahmed (Ed.), The invention of Somalia. Lawrenceville, NJ: Red Sea, 1–27. Muñoz, J. M. 2010. ‘Business Visibility and Taxation in Northern Cameroon’, African Studies Review, 53, 2: 149–175. Musamali, G. 2007. ‘Uganda: Somalis Want UPDF Deployment’, The New Vision, 12 January. Musinga, M., D. Kimenye & P. Kivolonzi. 2006. The Camel Milk Industry in Kenya SNV Report. Available at http://www.ebpdn.org/download/download.php?table?resources&id?2337 [Accessed 8 March 2015]. Myers, G. A. 2011. African Cities: Alternative Visions of Urban Theory and Practice. London, New York: Zed Books. Nenova, T. & T. Harford. 2004. ‘Anarchy and Invention: How Does Somalia’s Private Sector Cope Without Government’, Public Policy for the Private Sector 280: 4. Newton, K. 1997. ‘Social capital and democracy’, American Behavioral Scientist 40, 5: 575. Nordstrom, C., 2000. ‘Shadows and Sovereigns’, Theory, Culture & Society 17, 4: 35–54. Nordstrom, C. 2004. Shadows of War: Violence, Power and International Profiteering in the Twenty-First Century. Berkeley: University of California Press. 280 North, D. 2016. ‘Institutions and economic theory’. American Economist, 61,1: 72–76. Nyamnjoh, F.B. & B. Page. 2002. ‘Whiteman Kontri and the enduring allure of modernity among Cameroonian youth’, African Affairs 101: 607-34. Nyberg-Sørensen, N. (Ed.) 2007. Living Across Worlds: Diaspora, Development and Transnational Engagement Geneva: International Organisation for Migration. Oakes, T. & L. Schein. 2006. ‘Preface’, in T. Oakes & L. Schein, L. (eds), Translocal China. Linkages, Identities, and the reimaging of space. London: Routledge, xii-xiii. Oka, R.C. 2014. ‘Coping with the Refugee Wait: The Role of Consumption, Normalcy, and Dignity in Refugee Lives at Kakuma Refugee Camp, Kenya’, American Anthropologist 116, 1: 23–37. Okungu, J. 2011. ‘Is Al Shabaab Head in Somalia or Nairobi?’ The Star, 28 October. Available at http://www.the-star.co.ke/opinions/jerry-okungu/46694-is-al- shabaab-head-in-somalia-or-nairobi [accessed 22 May 2015] Okurut, F. & T. Botlhole. 2009. ‘Informal financial markets in Botswana: a case study of Gaborone City’, Development Southern Africa 26, 2: 255 – 270 Omata, N. 2013. ‘The Complexity of Refugees’ Return Decision-Making in a Protracted Exile: Beyond the Home-Coming Model and Durable Solutions’, Journal of Ethnic and Migration Studies 39, 8: 1281–1297. Omata, N. & J. Kaplan.2013. ‘Refugee Livelihoods in Kampala, Nakivale and Kyangwali Refugee Settlements: Patterns of Engagement with the Private Sector’, Working Paper Series, No. 95, Refugee Studies Centre, University of Oxford. Available at http://www.rsc.ox.ac.uk/files/publications/ working-paper- series/wp95-refugee-livelihoods-kampala-nakivale-kyangwali-2013.pdf [accessed 21 April 2015]. Orozco, M. & J. Yansura. 2013. Keeping the lifeline open .Remittances and markets in Somalia. ADESO / OXFAM / IADB. Beechwood. Otha, I. 2005. ‘Coexisting with cultural ‘others’: social relationships between the Turkana and the refugees at Kakuma, Northwest Kenya’, in Ikeya, K. and E. Fratkin, (eds.): Pastoralists and Their Neighbors in Asia and Africa, Senri Ethnological Studies 69: 227–239, National Museum of Ethnology, Osaka. 281 Otunnu, O. 1992. ‘Factors Affecting the Treatment of Kenyan-Somalis and Somali Refugees in Kenya: A Historical Overview’, Refuge 12.5: 21–6 Oucho, J. 2008. ‘African Diaspora and Remittance Flows: Leveraging Poverty?’ African Migration Yearbook 2008 0–26. Overå, R. 2006. ‘Networks, distance, and trust: Telecommunications Development and changing trading practices in Ghana’, World Development 34,7: 1301–1315. Oxfeld, E. & D.L. Long. 2004. ‘Introduction: an ethnography of return’, in D.L. Long & E. Oxfeld (eds) Coming Home? Philadelphia: University of Pennsylvania Press, 1-15. Page, B. & C. Mercer. 2012. ‘Why do people do stuff?: reconceptualizing remittance behaviour in diaspora-development research and policy’, Progress in Development Studies 12, 1: 1–18. Page, J. & S.Plaza. 2006. ‘Migration remittances and development: A review of global evidence’, Journal of African Economies 15, SUPPL. 2: 245–336. Pahl, R. E. 1985. ‘The Politics of Work’, Political Quarterly 56, 4: 331-45. Palackal, A., P.N. Mbatia, D.B. Dzorgbo, R.B. Duque, M.A. Ynalvez & W.M. Shrum. 2011. ‘Are mobile phones changing social networks? A longitudinal study of core networks in Kerala’, New Media Society 13, 3: 391–410. Palmer, R. 2004. The informal economy in sub-Saharan Africa: Unresolved issues of concept, character and measurement. Occasional Paper No. 98. Ediburgh: Centre of African Studies, Edinburgh University. Panagakos, A.N. & H.A. Horst. 2006. ‘Return to Cyberia: technology and the social worlds of transnational migrants’, Global Governance, 6, 2: 109–124. Pankhurst, R.1974. ‘Indian Trade with Ethiopia, the Gulf of Aden, and the Horn of Africa in the Nineteenth and Early Twentieth Centuries’, Cahiers d'Etudes Africaines XIV, 3: 453-497. Parry P. & Bloch M. (eds).1989. Money and the Morality of Exchange. Cambridge: Cambridge University Press. Pascouau, Y., & H. Labayle. 2011 “Conditions for Family Reunification Under Strain. A Comparative Study in Nine EU Member States.” King Baudouin Foundation, European Policy Centre, Odysseus Network. 282 . Accessed on Febraury 23, 2016. Passas, N.1999. Informal Value Transfer Systems and Criminal Organizations; A Study into So-called Underground Banking Networks The Hague: Dutch Ministry of Justice. Available at http:// english.wodc.nl/onderzoeksdatabase/informal-value-transfer-systems-and-criminal- organizations .aspx. [accessed 12 Feb. 2015]. Passas, N. 2006. ‘Demystifying Hawala: A Look into its Social Organisation and Mechanics’, Journal of Scandinavian Studies in Criminology and Crime Prevention, 7, 1: 46–62. Pavanello, S., S. Elhawary & S. Pantuliano. 2010. Hidden and exposed: Urban refugees in Nairobi. Humanitarian Policy Group Working Papers. London: Overseas Development Institute. Peebles, G. 2010. ‘The Anthropology of Credit and Debt’, Annual Review of Anthropology 39, 1: 225–240. Perlman, J. 2010 Favela. Four Decades of Living on the Edge in Rio de Janeiro. New York: Oxford University Press. Pérouse de Montclos, M. A. 2008. ‘Humanitarian Aid, War, Exodus, and Reconstruction of Identities: A Case Study of Somali “Minority Refugees” in Kenya’, Nationalism and Ethnic Politics 142, 2: 289–321. Pérouse de Montclos, M. A. & P. Mwangi Kagwanja. 2000. ‘Refugee Camps or Cities? The Socio-economic Dynamics of the Dadaab and Kakuma Camps in Northern Kenya’, Journal of Refugee Studies, 13, 2: 205–222. Pieke, F., N.Van Hear & A. Lindley. 2007. ‘Beyond control? The mechanics and dynamics of “informal”remittances between Europe and Africa’, Global Networks 3, June: 348–366. Pieterse, E. 2009. ‘African Cities: Grasping the Unknowable’, Inaugural Lecture, African Centre for Cities, delivered at the University of Cape Town. Available at http://africancentreforcities.net/download/assets/grasping_the_unknowable_26_aug ust _09.pdf. [accessed 13 April 2015]. Pieterse, E. 2011. Introduction: rogue urbanisms. Social Dynamics, 37, 1: 1– 4. 283 Pirkkalainen, P. & A. Mahdi. 2009. The Diaspora – Conflict – Peace – Nexus: A Literature Review Jyväskyla: University of Jyväskylä. Porteous, D. 2006. ‘The Enabling Environment for Mobile Banking in Africa’, Bankable Frontiers Associates Boston USA April: 1–57. Porteous, D. 2006. The Enabling Environment for Mobile Banking in Africa, Bankable Frontiers Associates Boston. Available at: http://www.bankablefrontier.com/assets/ee.mobil.banking.report.v3.1.pdf. [accessed 18 March 2015] Porter, G., K.Hampshire, P. Kyei, M. Adjaldo, G. Rapoo & K. Kilpatrick. 2008. ‘Linkages between livelihood opportunities and refugee-host relations: Learning from the experiences of Liberian camp-based refugees in Ghana’, Journal of Refugee Studies 21, 2: 230- 252. Portes, A. 1997. ‘Globalisation from Below: The Rise of Transnational Communities’, Transnational Communities Programme Working Paper WPTC-98- 01. Swindon, UK: Economic and Social Research Council. Portes, A. 1998. ‘Social capital: Its origins and applications in modern sociology’, Annual Review of Sociology 24: 1–24.Portes, A., M. Castells & L. Benton (eds). 1989. Informal Economy: Studies in Advanced and Less Developed Countries Baltimore: Johns Hopkins University. Portes, A., and P. Landolt. 1996. ‘The downside of social capital’, The American Prospect 26: 18–22. Portes, A. & P. Landolt. 2000. Social Capital: Promise and Pitfalls of its Role in Development. Journal of Latin American Studies, 32, 2: 529–547. Portes, A. & R. Rumbaut. 1990 Immigrant America: A Portrait. Berkeley: University of California Press. Portes, A. & A. Stepick, , 1985. Unwelcome Immigrants : The Labor Market Experiences of 1980 ( Mariel ) Cuban and Haitian Refugees in South Florida. American Sociological Review, 50, 4: 493–514. Powell, B., R. Ford & A. Nowrasteh. 2008. ‘Somalia After State Collapse: Chaos or Improvement?, Journal of Economic Behavior and Organization 67, 3: 657-670. Prag, E. 2010. ‘Women leaders and the sense of power: clientelism and citizenship at the Dantokpa market in Cotonou, Benin’, in I. Lindell, ed., Africa’s 284 informal workers: collective agency, alliances and transnational organizing London: Zed Books. Prahalad, C.K. 2004. The Fortune at the Bottom of the Pyramid: Eradicating. Poverty Through Profit. Upper Saddle River, NJ: Wharton School Publishing. Putnam, R. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Putnam, R. 1995. ‘Tuning in, tuning out: The strange disappearance of social capital in America’, Political Science and Politics 28,4: 664-683. Putnam, R. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simson and Schuster. Qassim, A. 2004. ’The Collapse of the Somali Banking System’, in A. M. Kusow, ed., Putting the Cart Before the Horse: Contested Nationalism and the Crisis of the Nation-State in Somalia. Trenton: Red Sea Press, 185-186. Quirk, J. & D. Vigneswaran. 2015. Mobility Makes States: Migration and Power in Africa, Philadelphia: Pennsylvania University Press Rabinow, P. & N. Rose. 2006. ‘Biopower today’, Biosocieties 1:195–217 Ram, M. et al., 2003. Banking on “break-out”: Finance and the development of ethnic minority businesses. Journal of Ethnic and Migration Studies, 29(4), pp.663–681. Ramadan, A. 2013. Spatialising the refugee camp. Transactions of the Institute of British Geographers, 38(1), pp.65–77. Raphaeli, N. 2003. ‘Financing of Terrorism: Sources, Methods, and Channels’, Terrorism and Political Violence 15, 4: 59–82. Ratha, D. & S. Mohapatra. 2011. ‘Preliminary Estimates of Diaspora Savings’, Migration and Development Brief 14. Washington, DC: World Bank Rawlence, B. 2014. ‘Kenya’s anti-terror strategy begins to emerge’ African Arguments, 9 April. Available at http://africanarguments.org/2014/04/09/kenyas- anti-terror-strategy-begins-to-emerge-by-ben-rawlence/ [accessed 13 March 2015] Rawson, D. 1994. ‘Dealing with disintegration: U.S. assistance and the Somali state’, in A.I. Samatar, ed. The Somali challenge: from catastrophe to renewal? Boulder: Lynne Reinner, 147–87. 285 RCK. 2012. Asylum under Threat; Assessing the Protection of Somali Refugees in Dadaab Refugee Camps and Along the Migration Corridor, Nairobi: Refugee Consortium of Kenya. Read, D. M. & K.A.Parton. 2009. ‘Economic Deregulation and Trade Liberalization in Kenya, Tanzania and Uganda: Growth and Poverty’, Journal of Economic Issues. 43, 3: 567-586 Refugee Consortium of Kenya. 2003. ‘Refugee management in Kenya’, Forced Migration Review 16. Refugee Law Project, , (undated) ‘Critique of the Refugees Act 2006’, Kampala: RLP. Reno, W. 2011. Warfare in Independent Africa. Cambridge University Press. Reuters. 2012. ‘Bomb rips through Nairobi minibus, killing six’, 18 November. Reuters. 2015. ‘Somalia's planned 2016 elections will not be a popular vote’, 29 July. Rinelli, L. & S.O. Opondo. 2013. ‘Affective economies: Eastleigh’s metalogistics, urban anxieties and the mapping of diasporic city life’, African and Black Diaspora: An International Journal 6,2: 236–250. Robbins, J. & D.Akin. 1999. ‘An introduction to Melanesian currencies. Agency, Identity, and Social Reproduction’, in D. Akin & J. Robbins, eds. Money and Modernity. State and Local Currencies in Melanesia. University of Pittsburgh Press. Roberts, H. 2014. ‘600,000 migrants are lined up along North African coast and ready to enter Europe this summer warns Italy’, Mail Online 4 April http://www.dailymail.co.uk/news/article-2596907/Army-600-000-migrants-lined- North-African-coast-ready-invade-Europe-summer-warns-Italy.html [accessed 19 August 2015] Rockefeller, S.A. 2011 ‘‘Flow’’, Current Anthropology 52, 4: 557-578. Rutherford, S., New ways to strengthen old ways: M-PESA and informal finance | Financial Access Initiative. Available at: http://financialaccess.org/node/2988. Safran, W. 1999. ‘Comparing Diasporas: A Review Essay’, Diaspora, 8, 3. 286 Salter, M.B. 2008. ‘When the exception becomes the rule: borders, sovereignty, and citizenship’, Citizenship Studies 12, 4: 365–380. Samatar, A. I., 1988. Socialist Somalia – Rhetoric and Reality. London: Zed Books. Samatar, A. I. 1993. ‘Destruction of State and Society in Somalia: Beyond the Tribal Convention’, Journal of Modern African Studies 30, 4: 625–641. Sander, C. & S.M. Maimbo. 2003. ‘Migrant Labor Remittances in Africa: Reducing Obstacles to Development Contributions’, World Bank, November. Sassen, S. 1991. The Global City: New York, London, Tokyo Princeton: Princeton University Press. Saunders, D. 2011. Arrival City: How the Largest Migration in History is Reshaping our World, New York: Vintage Books. Schaeffer, E.C. 2008. ‘Remittances and Reputations in Hawala Money- Transfer Systems: Self-Enforcing Exchange on an International Scale’, The Journal of Private Enterprise 24, 1: 95–117. Schapendonk, J. 2013. ‘From Transit Migrants to Trading Migrants: Development Opportunities for Nigerians in the Transnational Trade Sector of Istanbul’, Sustainability 5, 7: 2856–2873. Schaub, M.L. 2012. ‘Lines across the desert: mobile phone use and mobility in the context of trans-Saharan migration’, Information Technology for Development 18, 2: 126–144. Schielke, S., & K. Graw. 2012. The Global Horizon: Migratory Expectations in Africa and the Middle East. Leuven: Leuven University Press. Schlee, G. 2013. ‘Customary law and the joys of statelessness: idealised traditions versus Somali realities’, Journal of Eastern African Studies 7,2: 258–271. Schmoll, C. 2012. ‘The making of a transnational marketplace. Naples and the impact of Mediterranean cross-border trade on regional economies’, Cambridge Journal of Regions, Economy and Society 5,2: 221–237. Schmoll, C. & G. Semi. 2013. ‘Shadow circuits: urban spaces and mobilities across the Mediterranean’, Identities: Global Studies in Culture and Power 20, 4: 377-392 287 Schoumaker, B., M.L. Flahaux, D. Schans, , C. Beauchemin, V. Mazzucato & P. Sakho. 2015. ‘Changing patterns of African Migration: A Comparative Analysis’, in C. Beauchemin. Migration between Africa and Europe : Trends, Factors and Effects, New-York: Springer-Verlag & INED, Population Studies series. Schwartz, S. J., Unger, J. B., Lorenzo-Blanco, E. I., Des Rosiers, S. E., Villamar, J. A., Soto, D. W., … Szapocznik, J. (2014). Perceived Context of Reception among Recent Hispanic Immigrants: Conceptualization, Instrument Development, and Preliminary Validation. Cultural Diversity & Ethnic Minority Psychology, 20(1), 1–15. Sedgwick, P. 2014. ‘Unit of observation versus unit of analysis’, BMJ (Clinical research ed.) 348: 3840. Sheller, M. & J. Urry. 2006. ‘The new mobilities paradigm’, Environment and Planning 38: 207–227. Sheekh, R. C., ed. 2009. War and Peace: an anthology of Somali literature. Translated by Martin Orwin. London: Ponte Invisibile and Progressio. Sheppard, E. 2002. ‘The Spaces and Times of Globalisation: Place, Scale, Networks, and Positionality’, Economic Geography, 78, 3: 307–330. Shipton, P. 1995. ‘Luo entrustment: foreign finance and the soil of spirit in Kenya’, Africa 65, 2: 165–196. Shipton, P., 2009. Mortgaging the Ancestors: ideologies of attachment in Africa New Haven CT: Yale University Press. Shipton, P. 2014. ‘Trusting and Transcending’, Current Anthropology, 55, S9: S51–S61. Shoumatoff, A. 1994. ‘The 'Warlord' Speaks’, The Nation 258, 4: 442–50 Simone, A. 2015. ‘Afterword: Come on out, you’re surrounded: The betweens of infrastructure’, City, 19, 2-3: 375–383. Simone, A. 2004. People as Infrastructure: Intersecting Fragments in Johannesburg, Public Culture, 16, 3: 407–429. Simpson, G., 2010. “Welcome to Kenya” Police abuse of Somali Refugees. Human Rights Watch. 288 Singerman, D. 1995. Avenues of Participation: Family, Politics, and Networks in Urban Quarters of Cairo. Princeton, N.J.: Princeton University Press. Sheikh, A. & I. Mohamed. 2009. ‘Somali mobile phone firms thrive despite chaos’, Reuters 3 November. Slater, D. & J. Tacchi. 2004. ICT innovations for poverty reduction New Delhi: UNESCO. Smith, M. 2004. Warehousing Refugees: A Denial of Rights, a Waste of Humanity U.S. Committee for Refugees and Immigrants. Smith, Y.J. 2013. ‘We All Bantu – We Have Each Other: Preservation of Social Capital Strengths during Forced Migration’, Journal of Occupational Science, 20, 2: 173–184. Sökefeld M. 2006. ‘Mobilizing in Transnational Space: A Social Movement approach to the Formation of Diaspora’, Global Networks 6.3: 265–84. SomConsult. 1985. Report on flow of remittances of Somali workers abroad Mogadishu: Somconsult. Sommer, J. 1994. Hope restored? Humanitarian Aid in Somalia 1990-1994. Refugee Policy Group, November. Sorensøn, N N & K.F. Olwig, eds. 2002. Work and Migration: Life and Livelihoods in a Globalizing World London: Routledge. Staples, J. 2007. Introduction: Livelihoods at the margins. in J. Staples, ed., Livelihoods at the margins. Surving the city Walnut, CA: Left Coast Press Stepick, A. 1989. Miami’s Two Informal Sectors, in A. Portes, M. Castells and L. Benton (eds) The Informal Economy. Baltimore: Johns Hopkins University Press. Stepick, A. & A. Portes, 1986. Flight into despair: A profile of recent Haitian refugees in South Florida. International Migration Review, 20,74: 329–350. Stevis, M. & M. Mesco. 2015. ‘How Money, Race and Religion Determine the Fate of Europe-Bound Migrants’, The Wall Street Journal 16 August. Strathern, M., 1996. Cutting the Network. The Journal of the Royal Anthropological Institute, 2,3: 517–535. 289 Sykes, K M. 2005 Arguing with Anthropology: an introduction to critical theories of the gift. London: Routledge,. Sudan Tribune. 2013. ‘Uganda police hunt German terrorist’, 4 October Susser, I. 1996. ‘The shaping of conflict in the space of flows’, Critique of Anthropology 16, 1: 39–47. Szreter, S. 2002. ‘The state of social capital: Bringing back in power, politics, and history’, Theory and Society 31, 5: 573–621. Szreter, S., & M. Woolcock,. 2004. ‘Health by associ ation? Social capital, social theory, and the political economy of public health’, International Journal of Epidemiology, 33: 1–18. Taussig, M., 1992. The Nervous System, London: Routledge. Teff, M. & Yarnell, M., 2012. Somali Refugees: Ongoing Crisis, New Realities,Field Report, Refugee Internationa, March. Tenhunen, S. 2008 ‘Mobile technology in the village: ICTs, culture, and social logistics in India’, Journal of the Royal Anthropological Institute 14: 515- 534. The Economist. 2010. ‘Banking on mobile phones: Out of thin air’, 10 Jun. Available at: http://www.economist.com/research/articlesBySubject/displayStory.cfm?story_id= 16319635&subjectID=894408&fsrc=nwl The Economist. 2011. ‘Mobile telecoms in Africa: Digital revolution’, 7 April Available at http://www.economist.com/node/18529875 [accessed 01 June 2012]. The Economist. 2013. ‘The Road is Long’, 17 August. Available at: http://www.economist.com/news/britain/21583710-somalis-fare-much-worse- other-immigrants-what-holds-them-back-road-long, [accessed 18 August 2015]. The New Vision. 2007. ‘Uganda: Somalis Demand Recognition’, 28 October. The Standard. 2012. ‘Eastleigh blast victim dies’, 6 December The Standard. 2012. ‘At least one injured in Eastleigh blast’, 17 December. Available at 290 http://www.standardmedia.co.ke/?articleID=2000073065&story_title=Kenya-At- least-one-injured-in-Eastleigh-blast [Accessed 18 Jan. 2015]. Thompson, E. A. 2008. ‘An Introduction to the Concept and Origins of Hawala’, Journal of the History of International Law 10, 1: 83–118. Thompson, E.A. 2011. Trust is the Coin of the Realm Oxford University Press. Thornton, P.H. & K.H., Flynn. 2003. Entrepreneurship, Networks, and Geographies. In Z. J. Acs & D. B. Audretsch, eds. Handbook of Entrepreneurship Research. International Handbook Series on Entrepreneurship. Springer US, 401– 433. Available at: http://link.springer.com/chapter/10.1007/0-387-24519-7_16 [Accessed 19 April 2016]. Thrift, N. 1996. Spatial Formations Thousand Oaks, CA: Sage. Tilley, H. 2011. Africa as a Living Laboratory: Empire, Development, and the Problem of Scientific Knowledge, 1870-1950. Chicago: University Of Chicago Press. Titeca, K. & T. Vervisch. 2008. ‘The Dynamics of Social Capital and Community Associations in Uganda: Linking Capital and its Consequences’, World Development, 36, 11: 2205–2222. Tölölyan, K. 1991. ‘The nation-state and its others: in lieu of a preface’, Diaspora 1: 3-7. Tran, M. 2013. ‘Somalis fear Barclays closure of remittance accounts will cut lifeline’, The Guardian, 24 June. Trefon, T. 2004. Reinventing Order in the Congo: How People respond to State Failure in Kinshasa. London: Zed Books. Tsing, A.L. 2005. Friction: An Ethnography of Global Connection. Princeton: Princeton University Press. Turton, E. 1972. ‘Somali resistance to colonial rule and the development of Somali political activity in Kenya, 1893-1960’, Journal of African History 13, 1: 119-43. Turton D. 2003. ‘Refugees, forced resettlers and ‘other forced migrants’: towards a unitary study of forced migration’, New Issues in Refugee Research, Geneva: UNHCR, Working Paper 94. 291 UNHCR. 2011. Statistical Yearbook 2012, Geneva: United Nations High Commissioner for Refugees. UNHCR. 2014. Statistical Yearbook 2013, Geneva: United Nations High Commissioner for Refugees. United Nations General Assembly. 2001. Investigation into Allegations of Refugees Smuggling at the Nairobi Branch Office of the United Nations High Commissioner for Refugees, A/56/733 Office of Internal Oversight Services. UNSC. 2012. Report of the Monitoring Group on Somalia and Eritrea pursuant to Security Council resolution 2002. Avaiable at http://www.securitycouncilreport.org/atf/cf/%7B65BFCF9B-6D27-4E9C-8CD3- CF6E4FF96FF9%7D/Somalia%20S%202012%20545.pdf [accessed 10 May 2015]. USCR. 2004. World Refugee Survey 2004 – Kenya, US Committee for Refugees and Immigrants. Available at http://www.refworld.org/cgibin/texis/vtx/rwmain?page=publisher&publisher=USC RI&type= ANNUALREPORT&coi=KEN&docid=40b4593e14&skip=0 [accessed 16 October 2013] Van Hear, N. 2014. ‘Reconsidering Migration and Class’, International Migration Review 48, Fall: S100–S121. Van Hear, N. 2010. ‘Theories of Migration and Social Change’, Journal of Ethnic and Migration Studies 36, 10: 1531–1536. Van Lehman, D., & O. Eno. 2003. The Somali Bantu: Their History and Culture. Washington: Center for Applied Linguistics. Van Notten, M. 2005. The Law of the Somalis. Trenton, NJ: Red Sea Press. Verdirame, G. 1998. Refugees in Kenya: Between a Rock and a Hard Place, Oxford: Refugee Studies Programme. Verdirame, G., & B.E. Harrell-Bond. 2005. Rights in Exile Janus-Faced Humanitarianism. New York: Berghahn Books. Vertovec, S. 1999. ‘Conceiving and Researching Transnationalism’, Ethnic and Racial Studies 22, 2: 447–462. Vigh, H., 2008. ‘Crisis and Chronicity: Anthropological Perspectives on Continuous Conflict and Decline’, Ethnos 73,1: 5–24. 292 Vigh, H., 2009. ‘Motion squared: A second look at the concept of social navigation’, Anthropological Theory 9,4: 419–438. Vlcek, W. 2010. ‘Alongside global political economy — a rhizome of informal finance’, Journal of International Relations and Development 13,4: 429– 451. Vuorela, U. 2002. ‘Transnational families: imagined and real communities’, in D. Bryceson, & U. Vuorela, eds. The Transnational Family. New European Frontiers and Global Networks Oxford and New York: Berg, 63-82. Wahlbeck, Ö. 2002. ‘The concept of diaspora as an analytical tool in the study of refugee communities’, Journal of ethnic and migration studies 28, 2: 221- 238. Walker, S. 2007. ‘Can Refugees Become Citizens of Uganda?’ Refugee Law Project, Briefing Paper July. Walters, W. 2002. ‘Deportation, Expulsion, and the International Police of Aliens’, Citizenship Studies, 6, 3: 265– 292. Wang, H. 2004. ‘Regulating Transnational Flows of People: An Institutional Analysis of Passports and Visas as a Regime of Mobility’, Identities: Global Studies in Culture and Power 11, 3: 351–376. Watkins, K. & M. Quattri. 2014. Lost in Intermediation: How excessive charges undermine the benefits of remittances for Africa, London: Overseas Development Institute. Webersik, C., 2004. Difference that Matter: The Struggle of the Marginalised in Somalia. Africa, 74(4), pp.516–533. Werbner, P. 2002. ‘The place which is diaspora: citizenship, religion and gender in the making of chaordic transnationalism’, Journal of Ethnic and Migration Studies, Vol. 28 (1): 119-133. Wesonga, N. 2011, ‘Uganda: Somalis Still Feel Safe in Country Even After Attacks’, The Monitor 11 July. Wheatley, J. 2005. ‘Ancient Banking, Modern Crimes: How Hawala Secretly Transfers the Finances of Criminals and Thwarts Existing Laws’, University of Pennsylvania Journal of International Law 26: 347. 293 Whitehouse, Bruce. 2012. Migrants and Strangers in an African City: Exile, Dignity, Belonging. Bloomington: Indiana University Press. Williams, B. 1988. ‘Formal structures and social reality’, in D. Gambetta, D., ed. Trust: Making and Breaking Cooperative Relations. Oxford, Blackwell. Williamson, O. E. 1985. The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting. London: Macmillan. Wimmer, A. & N. Glick Schiller. 2003. ‘Methodological Nationalism, the Social Sciences, and the Study of Migration: An Essay in Historical Epistemology’, International Migration Review 37, 3: 576–610. Woolcock, M. 1998. ‘Social capital and economic development: Toward a theoretical synthesis and policy framework’, Theory and Society 27, 2: 151–208. Wouterse, F., 2010. Remittances, Poverty, Inequality and Welfare: Evidence from the Central Plateau of Burkina Faso. Journal of Development Studies, 46(4), pp.771–789. Zelizer, V. 1994. The Social Meaning of Money, New York: BasicBooks. Zetter, R. 1991. ‘Labelling Refugees: Forming and Transforming a Bureaucratic Identity’, Journal of Refugee Studies 4, 1: 39–62. Zetter, R. 2007. ‘More Labels, Fewer Refugees: Remaking the Refugee Label in an Era of Globalisation’, Journal of Refugee Studies, 20, 2: 172–192. Zollman, J. 2014. Kenya Financial Diaries. Shilingi kwa Shilingi Kwa Shilingi. The Financial Lives of the Poor. Nairobi: FSD Kenya, August.