MIGRATION EN ROUTE AS A SITE AND METHODOLOGY

This study draws primarily on multi-sited ethnographic research with Oromo women migrating through Djibouti from eastern Ethiopia, and the policymakers, service providers, hired guides, smugglers, and other persons accompanying and attempting to assist them along the way. Data were collected during three joint fieldwork stints: in March 2017, in January–February 2019 and in May 2019. Our findings also draw on our combined 36 years of scholarly work in Ethiopia and Djibouti (Carruth Reference Carruth2016, Reference Carruth2018, Reference Carruth2021; Smith Reference Smith2013; Smith et al. Reference Smith, Howard, Giordano, Yossinger, Kinne and Martin2019). On our first joint research stint, in 2017, we conducted four focus group discussions with migrants and refugees, 18 open-ended private interviews, and participant-observation and strategic observation among women and men in coastal areas where migrants congregated and departed for Yemen, in two refugee camps in Djibouti housing Ethiopian and Yemeni refugees and asylum seekers, and with several stakeholders at governmental, non-governmental and United Nations agencies. We learned about migration from refugees travelling to and through the country in different directions; from men and women who were part of smuggling networks helping people and goods cross the Red Sea; from aid workers and police officers who spoke of female migrants’ experiences of and vulnerability to sexual exploitation, sexual violence and robbery; and from healthcare and other service providers who cared for women en route.

Our next two trips, in early and then in mid-2019, entailed participant observation, strategic observation stints, participatory mapping exercises, and semi-structured private interviews along migrant routes from eastern Ethiopia through Djibouti (see Figure 1). We conducted private interviews with female irregular migrants from Ethiopia (n = 10) as well as with delalawoch (an Amharic, Afaan Oromo and Somali language plural word for guides or brokers, commonly referred to by international aid agencies as smugglers), gendarme, physicians, nurses and policymakers along common migration routes (n = 26).

Figure 1.

Map of Djibouti indicating field sites. Source: University of Texas at Austin, Perry-Castañeda Library Map collection, <https://legacy.lib.utexas.edu/maps/djibouti.html>.

With all interview and focus group participants we collaboratively mapped the flows of women and the social contacts and networks they utilised, including Ethiopian citizens and women identifying with Oromo ethnic groups within Ethiopia, but also smaller groups of Amhara, Tigrayan and Somali women irregularly migrating and travelling through Djibouti to destinations in the Middle East. With all these different participants we also outlined the various logistics of women's journeys through eastern Ethiopia to ports in Djibouti and then to Yemen, such as their modes of transportation, methods of border crossing, expected and actual costs, the relevant safety and security apparatuses that migrants encountered, and the many persons that accompany, manage, support and surveille their journeys.

With each of the 10 female Ethiopian irregular migrants from the Oromia Region, we first collected basic demographic information and then conducted semi-structured, extended ethnographic interviews and participatory mapping exercises. We also spent time observing and speaking briefly with other migrants and delalawoch, or smugglers or guides, with whom these same women travelled. During the ethnographic interviews, we asked women about the circumstances in their hometowns during and before their decisions to migrate, the financial, social and logistical support and resources they received, lacked and/or needed during their journeys, and where resources are scarce or fall short. We observed and interviewed additional people in formal as well as informal spaces for migrant services and networking in Djibouti, with a focus on those people and spaces serving Ethiopian women – including roadside resting places, markets, stores, hotels, cafes, clinics, salons, restaurants, churches, mosques and/or other community spaces, and spoke with individuals in those places about their encounters with women and their knowledge of women's migrations back and forth to Ethiopia. Finally, during additional private interviews with clinical providers and policymakers, we learned about the health needs and health emergencies common during women's migrations, and how and from whom they obtain various medical services at different points in their journey.

During all three of our field work stints, our interviews with migrating women happened at different points along their routes: in remote locations during their migrations through Djibouti, upon their return from the Gulf en route back to Ethiopia, after they had given up on migrating and were staying temporarily in an IOM shelter awaiting repatriation to Ethiopia, or when they were residing for a time in Djibouti City, attempting to save money for onward journeys. Consequently, the women we interviewed had not yet completed their journeys or already worked stints in the Gulf, and so could not offer reflections on their challenges after crossing the Red Sea, traversing Yemen, working abroad and/or attempting to return home to Ethiopia again.

In general, most scholarly and programmatic research with migrants is retrospective and asks returning migrants about their motivations and fears when their perceived ‘success’ or ‘failure’ may shape memory and responses. By contrast, this methodology examines migrant women's complex and sometimes contradictory reasons for migrating out of Ethiopia, and their migration journeys, reflections and experiences en route. This strategy highlights the dynamic, emotional and social processes of migration and the way migrants reassess and re-narrate their motivations and strategies for leaving home and making decisions. In the following section, we distinguish among the various terms by which migrants are referred and we outline the context in which migration and violence against women are invoked, particularly within governmental and non-governmental international aid organisations.

THE ‘MIGRANT’ AND THE ‘HUMANITARIAN INTERNATIONAL’

In a world in which population displacement is too common, and scholarship and politics too often myopically focus on concerns in the Global North about ‘refugee crises’ and ‘immigration crises’, the continuous and multidirectional movement of women between the Horn of Africa and Gulf States receives relatively little attention. Further, aside from recent coverage of conflict and violence in the Tigray Region of Ethiopia (UNOCHA 2021), there remains relatively little news, policy or political attention to the conditions elsewhere in Ethiopia that motivate and shape displacement and irregular migrations abroad. Ethiopia continues to have some of the highest numbers of displaced persons in the world, including ~2.1 million internally displaced persons (UNHCR 2018; IDMC 2019) and 705,820 refugees during this research (UNHCR 2018). However, estimates of the numbers of displaced persons there vary dramatically – not just over time, but even between different organisations counting migrants and displaced persons. For example, prior to conflict in the Tigray Region in northern Ethiopia (UNOCHA 2021), in October 2019 the Government of Ethiopia claimed there were only 100,000 IDPs (Getachew Reference Getachew2019). Then a few days later, UNICEF reported Ethiopia had 3.19 million IDPs (UNICEF 2019). With the outbreak of the conflict in northern Ethiopia in November 2020, and given ongoing localised conflicts throughout the country, these numbers surely remain in flux. In the Tigray Region of Ethiopia alone, in 2021, there are believed to be 1.7 million IDPs (UNOCHA 2021).

Asylum seekers and refugees are legally recognised in international and national law as ‘displaced’ and typically qualify for protection and humanitarian services (UNHCR 2016). They are counted and surveilled by the international humanitarian system, with the help of NGOs, international donors and state governments, and they are traditionally rendered sympathetic subjects in need of humanitarian action (Fassin & Vasquez Reference Fassin and Vasquez2005). ‘Displacement’ is generally defined in international law as the forced movement of people from their homes, places of habitual residence, or environment and occupational activities due to conflict, violence and even potentially natural and human-made disasters (UNHCR 2016; IOM 2019a). However, this broad designation is often difficult to prove or determine, especially for people originating in or travelling through politically insecure settings, for people who move or seek asylum for several reasons, and for people who may not desire to cross borders or be able to seek asylum outside their home country. Internal displacement statistics also depend on governments who may not allow international aid agencies access to displaced persons, and who may ignore or elide the displacement of people from dissenting political parties or minority or marginalised ethnolinguistic groups.

For example, most migrants we spoke to left the Oromia Region of Ethiopia for work in the Middle East because of a combination of protracted violence in their communities, threats or harassment they or their family members endured, intimate partner violence, and economic insecurity. While migrants fleeing politically insecure eastern and southern parts of Ethiopia ‘choose’ to leave (e.g. Zeleke Reference Zeleke2019), others fleeing these same circumstances and communities in many cases ended up living in a refugee camp in Djibouti and/or seeking asylum in the Middle East or Europe. We met several Oromo (and other Ethiopian) women in Ali Addeh refugee camp in Djibouti who had fled political violence in 2018 and 2019, and were awaiting resettlement or political change in Ethiopia before they returned home. Additionally, even though forced migration and political violence do not figure prominently in the IOM's conception of Ethiopian migrants’ motivations to leave home (e.g. IOM 2020b), in fact most migrant Oromo women we spoke to for this research described vividly incidents of land confiscation and the arrest and detention of local activists in their hometowns as motivations for their own and others’ migrations to the Gulf. And in general, all male and female Oromo Ethiopian refugees and migrants we met in Djibouti testified they faced political insecurity and political violence at home. However, most did not choose or want to seek asylum in Djibouti, due to the limited economic opportunities and poor housing conditions in remotely located refugee camps.

A ‘migrant’, in contrast to a refugee or asylum seeker, is an umbrella term not clearly defined under international law, describing a person who moves away from his or her place of usual residence (UNHCR 2016; IOM 2019a). ‘Regular’ migrants (as defined by IOM 2019a) who follow the legal pathways of transnational mobility, such as refugees, are counted and surveilled by numerous governmental organisations and mechanisms. While refugees and asylum seekers are types of migrants, the invocation of the word ‘migrant’ in popular parlance and the news media often more narrowly signifies persons who move voluntarily and often to seek economic opportunities, and not because of a direct threat of persecution or death (Castles et al. Reference Castles, de Haas and Miller2014). Alternatively, migrants may be classified according to the process through which they move – they may voluntarily travel for ‘economic’ reasons, but they may also be ‘trafficked’ or ‘smuggled’ against their will. Finally, migrants are classified according to their legal status: for example, ‘regular’ migration occurs in compliance with the laws of the country of origin, transit and destination, while ‘irregular’ migration (otherwise termed illegal, unauthorised or undocumented migration) is defined as the movement of persons that takes place outside the laws, regulations or international agreements governing the entry into or exit from the state of origin, transit or destination (IOM 2019a).

Contemporary migrations of Oromo Ethiopian women to the Gulf additionally trace routes hundreds if not thousands of years old – ancient population movements with a provenance that extends much longer than the humanitarian and policy communities’ focus on ‘migration crises’ and ‘vulnerable migrants’. For centuries, traders, travellers, pilgrims, scholars, refugees, and enslaved and trafficked persons have travelled back and forth between Ethiopia and the Gulf States. Historians have documented this as a crucial slave trading route, with Edwards claiming that ‘during the nineteenth century, slaves were probably the most important export item from Ethiopia’, either to Sudan or to Arabia via Massawa or the Gulf of Aden ports (Edwards Reference Edwards1982: 4). In fact, Hopper notes, ‘European demand was a significant factor in the growth of the East African slave trade in the eighteenth century’ (Hopper Reference Hopper2015: 33). Pankhurst also traced these routes, noting that the ‘main Gulf of Aden ports involved in the slave trade were Taj[o]ura and Zeila’ (Pankhurst Reference Pankhurst1964: 226). The trade in enslaved persons was not outlawed in Ethiopia until the mid-20th century (Aidid Reference Aidid2020). Ethnic Oromos in Ethiopia, in particular, were key actors of the transnational slave trade during this period (Hassen Reference Hassen1990).

Transnational historical migration, trade and slavery routes, we found, also involved a combination of both voluntary and involuntary movements, and these movements follow many of the contemporary pathways that thousands of Ethiopian men and women still use, journeying through Ethiopia and Djibouti, and to and from the Arabian Peninsula to seek asylum or refuge from political insecurity, political repression, and conflict in Yemen, Eritrea, Ethiopia and Somalia. Every year, thousands of Ethiopian migrants follow well-trodden routes out of Africa, walking unpaved caravan trails through remote desert expanses. Bezabeh (Reference Bezabeh2011: 599) describes a ‘nation with considerable gradation of citizenship … at the bottom of [which] we find recent migrants, particularly Ethiopian migrants who work in various low-paid jobs’ (also noted in Chire Reference Chire2012; Zeleke Reference Zeleke2018, Reference Zeleke2019).

The World Bank estimates ~850,000 Ethiopians legally migrated out of Ethiopia in 2017, and 20% of these migrants travelled to the Kingdom of Saudi Arabia (KSA) for work. The Ethiopian Ministry of Labour and Social Affairs (MoLSA) estimates far more than that actually departed Ethiopia for work in the Middle East, mostly through irregular means (IOM RDH 2020b: 7). The IOM Displacement Tracking Matrix (DTM) confirms a preponderance of movement among Ethiopian irregular migrants along the so-called ‘Eastern Route’ through Djibouti and Somalia, across the Red Sea and Yemen, and with the goal of KSA or other Gulf Cooperation Countries.

Studies conducted by various international aid organisations and news outlets have examined the risks and struggles of the Ethiopian men who migrate through Djibouti and work in the Gulf States (Gardner & Rosser Reference Gardner and Rosser2018; IOM 2020a), but women's migrations remain largely absent from media profiles and policy documents (one exception, discussed subsequently in this paper, includes IOM RDH 2020b). Staff at the IOM, UNHCR and the World Bank acknowledged to us multiple times a lack of data about women's diverse and multifaceted experiences, and a need for interventions in response. Djiboutian and Ethiopian national policymakers also acknowledged the challenges they face designing appropriate services for migrant women, without a systematic understanding of the gendered risks these women face at home and en route, and the networks and resources women depend on for their invisibility, survival and protection.

Because of disputed and narrow legal designations, some scholars draw attention to the ways in which different migrants are described as ‘deserving’, based on their objectives, motivations, and the conditions in their countries of origin (de Genova Reference de Genova2002; Willen Reference Willen2012; Yarris & Castañeda Reference Yarris and Castañeda2015). The perceived ‘deservingness’ of different migrants affects their access to assistance, visibility and public sympathy. These scholars argue that we should move past the formalism of legal entitlement and into the moral constructions operating underneath definitional rationalisms (de Genova Reference de Genova2002; Willen Reference Willen2012; Yarris & Castañeda Reference Yarris and Castañeda2015; Cook Heffron Reference Cook Heffron2018). Displaced and migrant persons’ own invocations and appropriations of legal designations are also central to the findings we present here. Accordingly, we reject conventional and exogenous distinctions between voluntary/economic and involuntary/forced migration and we instead seek to better understand how different labels are applied to persons who cross national borders due to ‘various combinations of violence, choice, circumstance, and agency’ (Yarris & Castañeda Reference Yarris and Castañeda2015: 64–9). As already stated, Ethiopian women we met during this research dreamed of economic opportunities as they migrated abroad, but they also left home to escape multiple forms of violence. Classing them either as ‘displaced’ or merely ‘migrants’ elides the complexities of their motivations and the various forms of violence they faced at different points. Similarly, some of the methodological limitations of research conducted by international organisations and the focus by international organisations on distinct economic motivations for migration may contribute to failures on their parts to elicit information either about the political violence, structural violence (as defined by Galtung Reference Galtung1969; Carruth et al. Reference Carruth, Martinez, Smith, Donato, Piñones-Rivera and Quesada2021), and domestic, intimate partner or gender-based violence that contribute to women's migration decisions, or the contingent forms of sociality, power and agency women express by undertaking dangerous transnational journeys.

While economic motivations figured into Ethiopian women's irregular migrations to the Gulf, most women we spoke with also expressed a desire to leave home secretly and journey invisibly for their own safety, and expressed desires for escape and autonomy, and not just financial gain. This research therefore attends not just to the production of problematic binary categories in relevant policy and scholarship – such as between refugees and migrants, forced migration and economic migration, or even victims and perpetrators – but additionally, attends to how these mutual exclusions frame policy and elide the complexities and contingencies of mixed migration, including the repeated and multidirectional flows of people over time, the multiple positionalities and subjectivities women inhabit during and after migration, and women's multivalent decisions and strategies on the move (Mahmood Reference Mahmood2001; Yarris & Castañeda Reference Yarris and Castañeda2015). Women's objectives for migration reflect the kind of expansive, yet incomplete practices of citizenship that Smith (Reference Smith2013) describes in her study of federalism and ethnicity in Ethiopia. In that work, and here, meaningful citizenship ‘refers to the way in which rights are exercised, or the effective practice of citizenship’, not just the legal conferral of status (Smith Reference Smith2013: 4). By extension, migration constitutes an exercise of citizenship.

BEYOND SECURITY: AFRICAN WOMEN'S MIGRATIONS IN INTERNATIONAL RELATIONS

Migration policies in the Global North often seek to keep migrants out (Sajjad Reference Sajjad2018; Mourad & Norman Reference Mourad and Norman2020); consequently much of the scholarship on migration has sought to understand this phenomenon and strategy. So-called ‘deterrence narratives’, particularly in Europe and the USA, present migration through a security lens, and the host country's purported inability to accommodate migrants (Steinhilper & Gruijters Reference Steinhilper and Gruijters2018; Kent et al. Reference Kent, Norman and Tennis2020). These ‘externalisation’ policies by relatively rich advanced democratic states attempt to not only prevent migrants from entering but also to move the control of migrants and borders farther and farther from their own borders. Zolberg (Reference Zolberg, Fahrmeir, Faron and Weil2003) referred to this as a regime of ‘remote control’ of passports, visas and other border control measures. FitzGerald (Reference FitzGerald2019) identifies other metaphorical and architectural tools that highlight ways migration policies cage, control, and restrict modern migrants. Multiple control mechanisms have been documented by many and defined by Sajjad as ‘combining sophisticated technology, agreements, deployment of semi-military and military forces, extensive intelligence networks and the crude construction of walls and fences to deter irregular migration both by land and sea’ (Sajjad Reference Sajjad2018: 42). In this context, the EU–Turkey deal and the EU Trust Fund for Africa constitute crucial examples of externalisation (Mourad & Norman Reference Mourad and Norman2020).

Additionally, the focus on policy and politics in the Global North fails to fully account for what happens in the Global South and in migrants' countries of origin. Drawing on Hollifield's influential concept of the ‘migration state’, Adamson & Tsourapas (Reference Adamson and Tsourapas2019: 6) highlight the ways in which this term ‘centers on state management of migrant flows into a destination country and, thus, examines policymaking around questions of entry, integration, citizenship and naturalization’ (emphasis in original). The focus on migrants’ destinations and their effects on host communities potentially detracts from the circumstances in migrants’ home countries that motivate and shape dangerous and illicit journeys.

Unlike most countries of the Global North, many countries in the Global South are more concerned with out-migration and remittances from migrants as strategies of economic development and poverty alleviation. In many Global South countries with low state capacity, migration policy exists but it is typically ad hoc and responsive to a combination of domestic demands for remittances and international donor pressure (Adamson & Tsourapas Reference Adamson and Tsourapas2019). Throughout East Africa, out-migration represents an economic necessity for many communities and individuals. Accordingly, Ethiopians have left in large numbers in recent years seeking employment in the Gulf States; men mostly seek work in construction, and women mostly perform domestic labour (Fernandez Reference Fernandez2020).

The focus on the political and economic security of destination countries also obscures ‘various forms of political and forced migration, including conflict-induced migration, ethnic cleansing, refugee crises, asylum-seeking, modern and historical forms of slavery, human trafficking and smuggling’ experienced by migrants (Adamson & Tsourapas Reference Adamson and Tsourapas2019: 7). People may migrate for several reasons, and people may only disclose select reasons in different data collection settings; these reasons may be simplified, summarised, or glossed over as ‘economic’ for persons not officially classified as ‘displaced’. Mainwaring & Brigden (Reference Mainwaring and Brigden2016: 245) note, these ‘journeys are both ambiguous and clandestine’. More broadly, irregular migrations around the world, as well, are never either completely ‘forced’ – as people usually do have a role in shaping their own migration experiences, even in limited ways – nor are they the result of freely made economic choices alone (Menjívar & Salcido Reference Menjívar and Salcido2002; Akpinar Reference Akpinar2003; Merry Reference Merry2006; Martin & Tirman Reference Martin and Tirman2009; Donato & Gabaccia Reference Donato and Gabaccia2015; Campbell & Mannell Reference Campbell and Mannell2016; Dobrowolsky Reference Dobrowolsky2016; Freedman Reference Freedman2016; Ghebrezghiabher & Motzafi-Haller Reference Ghebrezghiabher and Motzafi-Haller2016; Goździak Reference Goździak2016; Belloni Reference Belloni2019).

One consequence of the dual focus among policymakers and aid agencies on security and economic motivations is the way these actors attempt to govern international migration and deter migrants from leaving home (Kent et al. Reference Kent, Norman and Tennis2020). For example, so-called ‘migration sensitisation’ campaigns or ‘migration deterrence’ programmes run by organisations such as the IOM in Africa seek to educate international migrants about the dangers they face en route and at work in the Middle East. IOM staffers attempt to find and converse with migrants at various points along their routes – in Ethiopian communities from where migrants depart, along the Ethiopia–Djibouti border, in towns where migrants hide and sleep en route, and in port cities such as Obock, where they depart to Yemen and other Gulf States (e.g. see IOM 2008, 2012, 2015). When the staff of IOM encounter or approach migrants, they typically show them posters and photographs of emaciated, deceased and mangled bodies that have been found nearby, and sometimes ask migrants to view testimonies from others who have undertaken the journey without success. Staffers tell migrants stories about boats sinking, migrants dying in the desert, and about incidents of sexual violence. This programme is designed to deter migrants from continuing on their journey, and to convince them to accept help from the IOM and their ‘assisted voluntary return’ back home (IOM 2015, 2020b). These methods of ‘sensitisation’ and ‘deterrence’ – also called ‘information’, ‘education’, and ‘migration suppression’ in the grey literature – did not in fact deter migrants from seeking work in the Gulf so far as we could tell, but rather only confirmed what migrants already feared and accepted as risks in these perilous journeys. As Sabine Dini notes in her incisive critique of IOM, ‘the organisational logic of IOM politicises mobility as deviant and statism as a proof of national allegiance’ (Dini Reference Dini2018: 1694). Bradley (Reference Bradley2020), as well, finds that the IOM has been much more focused on policy and creating a set of ‘humanitarian entrepreneurship’ activities such as supporting evacuation, so-called ‘voluntary’ returns and developing extensive data collection tools than providing other forms of protection or humanitarian assistance (Bradley Reference Bradley2020). Accordingly, the migration deterrence we observed reinforced the border management and sovereignty practices of both the Djiboutian and Ethiopian states, and did not enhance the protection or care of migrants.

VIOLENCE AND OPPORTUNITY: GENDERED PERSPECTIVES ON MIGRATION

There is rising demand for female domestic labour in the Gulf (Grabska et al. Reference Grabska, de Regt and Del Franco2019; Kebede n.d.). Survey research with return migrants in Ethiopia found that 60% were women and that the decision to migrate was theirs alone (Kuschminder & Siegel Reference Kuschminder and Siegel2014). However, since none of these recent survey reports asked female respondents specifically about intimate partner, gender-based or domestic violence as a potential reason to leave home, it is impossible to know the full extent to which these experiences may influence migration decisions (Kuschminder & Siegel Reference Kuschminder and Siegel2014; IOM 2020a, 2020b). Ethnographic research has been more successful in uncovering the frequency and effects of multiple interacting forms of and exposures to violence. For example, in her ethnographic study of Ethiopian female migrants, Zeleke (Reference Zeleke2018) found that economic contributions to family were a motivating factor in her respondents’ decision-making but migration was also a ‘strategy to escape from different forms of GBV,’ including arranged marriage and marriage by abduction.

Existing survey research in the region also demonstrates high levels of interpersonal and social violence and low-levels of employment and social autonomy for women and girls in Ethiopia (Pankhurst et al. Reference Pankhurst, Negussie and Mulugeta2016; UNFPA 2017), including among ethnic Oromo women in Ethiopia (Abeya et al. Reference Abeya, Afework and Yalew2012; Alebel et al. Reference Alebel, Kibret, Wagnew, Tesema, Ferede, Petrucka, Bobo, Birhanu, Tadesse and Eshetie2018; Yohannes et al. Reference Yohannes, Abebe, Kisi, Demeke, Yimer, Feyiso and Ayano2019). Approximately 50–60% of all Ethiopian women report some type of interpersonal violence during their lifetime, mostly at the hands of intimate partners and family members (World Health Organization 2005; UN Women 2020). Ethiopian women also experience high levels of economic precarity: compared with men, they face lower educational attainment, lower wages, higher unemployment in formal sectors, and higher prevalence of low wages and informal contractual work in domestic and agricultural sectors (McKay Reference McKay1993; Schewel Reference Schewel2018; Grabska et al. Reference Grabska, de Regt and Del Franco2019). Remarkably, decades of rapid economic growth, educational expansion, and governments’ and aid agencies’ focus on women's empowerment in Ethiopia seem to have done little to change the circumstances under which women seek to migrate. Schewel found that, still, ‘migration to the Middle East is framed as an investment for themselves and their families’ (2018: 6). Grabska et al. (Reference Grabska, de Regt and Del Franco2019: 65) conclude that ‘adolescent girls also migrate in order to escape oppressive gender regimes … (and) the lack of educational and employment opportunities’ (see also Yilma Reference Yilma2013).

However, there remain gaps in research focused on the gendered nature of various forms of violence in migrant women's home communities, and these gaps have in turn hampered interventions aimed at ending or reducing the various and gendered forms of interpersonal violence, political violence and exploitation that propel and shape women's migrations out of sending countries (Schewel & Fransen Reference Schewel and Fransen2018; Fernandez Reference Fernandez2020). For example, IOM's (2020b) recent report on women's migrations presumably only asked respondents to choose from either economic motivations or political persecution as motivation, without either considerations of the interdependent nature of these two phenomena or the potentially added effects of intimate, domestic, gender-based, sexual and interpersonal forms of violence. While this may be a methodological limitation to the IOM's work, it is reflective of a need to better understand how female migrants may describe, link and balance economic objectives, political motivations and fears of violence.

In general, intimate partner, domestic and gender-based violence within Ethiopian families and communities is difficult to study and confirm. Social stigma, family pressure and a weak policing and judicial system in Ethiopia all contribute to under-reporting (Kedir & Admasachew Reference Kedir and Admasachew2010; Burgess Reference Burgess2012). Even so, some estimates put the rates of intimate partner violence as high as 50–70% (World Health Organization 2005; Semahegn & Mengistie Reference Semahegn and Mengistie2015; Kassa & Abajobir Reference Kassa and Abajobir2018). A study by UN Women found that 20% of Ethiopian women reported experiencing physical and/or sexual intimate partner violence in the last 12 months, and higher over a lifetime (UN Women 2016). Another study found that nearly half of Ethiopian women experienced violence in their lifetime (Kassa & Abajobir Reference Kassa and Abajobir2018).

Migration and violence against women are linked in Ethiopia and elsewhere. De Regt & Mihret's (Reference de Regt and Mihret2020) research with sex workers in Addis Ababa finds that, ‘only seven out of thirty girls mentioned poverty as the main reason for their migration … One of the recurrent themes in the interviews was the restriction on girls’ mobility, in particular when they were growing up; abuse and violence at the familial level was a significant factor in motivating young women to seek control over their lives by seeking domestic work in Addis Ababa’ (De Regt & Mihret Reference de Regt and Mihret2020: 517). Similarly, studies of domestic violence in Peru, Chile, Central America, the UK and India demonstrate the importance of multiple forms of violence in understanding women's migration strategies and motivations (Parson Reference Parson2010; Bowstead Reference Bowstead2015; Anitha et al. Reference Anitha, Roy and Yalamarty2017; Cook Heffron Reference Cook Heffron2018). While a few Ethiopian women in our study explicitly mentioned state and communal violence as reasons they left home, many women also hinted at leaving marriages and families that wanted to control them. In interviews, focus groups and conversations, women repeatedly linked domestic experiences of violence, lack of personal autonomy including lack of control over finances, lack of economic opportunities, and a desire to improve their families' lives and lack of economic opportunity. Migrating out of one's hometown or country, and dreaming of building one's own house, represented strategies for navigating the violence both of economic deprivation and physical and economic control over women's lives. While this research does not permit us to assess the outcomes of these migrations, we find that female migrants seek to use these journeys to transform their lives in complex ways that transcend either securitised or economic frameworks. Likewise, Oromo Ethiopian women's migrations should be understood as more than economic strategies and more than responses to opportunities for work abroad or political persecution at home. These migrations were also journeys in search of autonomy and emancipation.

However, women's journeys toward the Gulf often stalled. Some Ethiopian women ended up in the capital of Djibouti City to work and earn money for a journey onwards to the Middle East, but sometimes they stayed in Djibouti indefinitely, unable to move on. These women often performed domestic labour or worked in restaurants in the truck depots and cafes near Ethiopian transit hubs outside the huge port of Djibouti which supplies most of landlocked Ethiopia's goods. Two young Ethiopian women told us of their experiences working as domestic servants there, hoping to earn funds for the onward journey. For example, one young woman said with some ambiguity that she left home because of ethnic violence between ethnic Somalis and Oromos in Ethiopia, but also because, ‘there was no ability to choose as a young woman’. Officials in government offices in Djibouti City repeatedly in conversations and interviews noted the large numbers of female Ethiopians employed in Djibouti, including in informal, unregulated housekeeping work. Similarly, Zeleke (Reference Zeleke2019) found that Ethiopian female migrants in Djibouti City stayed there and worked only a few weeks to six years. Many of these women intended to move on to the Gulf, but were often temporarily waylaid for various reasons, including lack of funds (Zeleke Reference Zeleke2019).

Delalawoch, or brokers and guides who surreptitiously transport Ethiopian migrants across international borders and through remote landscapes, testified to the frequency with which Ethiopian women are delayed or even extorted and forced to stay in Djibouti City. Medical professionals in clinics in the capital also reported regularly treating Ethiopian women for health concerns related to the violence they had experienced at home, en route, and related to their work in Djibouti City (Carruth et al. Reference Carruth, Martinez, Smith, Donato, Piñones-Rivera and Quesada2021). One physician in a clinic in Djibouti City reported providing free or reduced cost family planning, HIV testing and other medical care for Ethiopian women. Some of these women worked in ‘nightclubs’, she said, or as domestic workers, while others continued travelling soon after their arrival in Djibouti City, headed for the Gulf. Multiple clinicians reported that migrant Ethiopian women most frequently presented in their clinics with malnutrition, tuberculosis and dehydration from diarrhoeal diseases. Some had clearly faced violence, these clinicians attested, but women, ‘would not talk about this violence’.

Similarly, medical professionals in Obock, a port city close to where migrants cross the Red Sea to Yemen by boat, reported health issues for women but also a desire to continue to Yemen. A clinician at the IOM's Migrant Resource Centre (MRC) in Obock noted that female migrants who had come to the MRC for medical care (most commonly, for severe diarrhoeal diseases and dehydration, occasionally the result of confirmed cholera infections) would sometimes change their mind about seeking care from the MRC and want to leave before completing their medical treatment. Sometimes the delalawoch even came to the MRC to seek out their clients, purporting to ‘care for the women more than we do’ because of the value of their labour on arrival in the Gulf. We also heard from migrants that, rather than being criminals or abusive, most delalawoch tried to protect and care for them. They facilitated their escapes and earned their trust.

THE ESCAPES

Oromo women leaving Ethiopia during this research travelled through both legal and, more commonly, irregular or extra-legal means to find employment in Saudi Arabia and the other Gulf countries. In the three ethnographic narratives to follow, women narrate their objectives, aspirations, and journeys, and in so doing illustrate how their complex and interrelated aspirations, agency and fears structure these irregular migrations and their attempts to realise both escape and emancipation.

CONCLUSION: POWER THROUGH MIGRATION

It is challenging to disentangle the objectives and motivations of migrants. Yet based on our analysis and policy work on migration in this region, we find it is along the way, rather than after the fact, that women's goals, strategies and dreams are best ascertained through field research. The Oromo Ethiopian women we met en route through Djibouti expressed desires for a complex mix of economic, political and personal independence. Women's own narratives of their migration journeys, at least at this point in their lives and journeys, revealed they understood the risks they faced, but sought to better their lives through extra-legal transnational migration anyway. All but one of the women we met and interviewed departed home secretly, knowing others would want to keep them home. And they survived the arduous journey and multiple crises along the way drawing strength from their aspirations to build their own metaphoric and material houses. In so doing, these women articulated a profound form of transnational, active and meaningful citizenship (Smith Reference Smith2013) potentially common among many irregular migrants facing violence at home and during migration journeys.

Much of the migration literature produced by scholars and policymakers in the Global North centres either on political security or on migrants’ economic motivations and objectives. The securitisation of migration can render individual experiences and motives invisible, reify the threat of a ‘migration crisis’ or ‘refugee crisis’, and thus also fail to recognise the effects of intersecting forms of violence, poverty and insecurity on decision-making both in migrants’ hometowns and along the route. Moreover, Oromo Ethiopians testified to us that internal displacement within Ethiopia and seeking asylum outside Ethiopia were both dangerous prospects and failed to offer hope of economic opportunity. Given their limited options and the dire circumstances for ethnic Oromos in Ethiopia and in remote refugee camps in Djibouti, and given the fact that legal forms of labour migration remained too expensive or impossible for most to attain, we posit that irregular transnational migration may often be worth the risk, at least among targeted and impoverished ethnic minorities.

Additionally, scholarship that accounts for violence against women and their relative lack of economic opportunities also tends only to see them as victims – either victims of the states and families that drive them out, or victims of male traffickers, smugglers or employers in destination countries – rather than agents of their own destiny, however limited by structural conditions. The women we met, however, had specific and detailed visions, strategies and networks of support to achieve their desired outcomes.

Women's irregular migrations are therefore neither simply ‘forced’ or incidents of exploitative smuggling alone, nor are they the result of freely made or safe economic choices. Oromo Ethiopian women described their migrations abroad for work as decisions, motivated by their poverty as well as their experiences and threats of political violence, structural violence and intimate partner or domestic violence at home. Women survived violence and dangerous journeys to the Gulf States by drawing on social, emotional and logistical support from other migrants as well as their delalawoch.

While we found some delalawoch did exploit or abuse migrants, and while this system of irregular migration places women in positions of less power and authority compared with delalawoch they depend on, delalawoch were also key actors in local and transnational migration economies and communities throughout the region. Women depended on and often trusted delalawoch for their escapes, even when assistance and protection from Djiboutians and humanitarian aid organisations were available. This suggests that different approaches to migrant assistance and protection are needed, as well as more nuanced perspectives on ‘smuggling’ within governmental and aid agencies such as the IOM.

The activities, strategies and networks of delalawoch also remained largely ignored by global humanitarian actors such as the IOM and to Ethiopians hoping to migrate. Delalawoch provided no safeguards or guarantees for migrants’ safe arrival in the Gulf, and we found no efforts by delalawoch to provide transparency or accountability for their services. At the same time, delalawoch lacked power and authority to address the structural violence women faced before, during and after their migration journeys (Carruth et al. Reference Carruth, Martinez, Smith, Donato, Piñones-Rivera and Quesada2021); they said they felt helpless at only being able to offer skeletal services at certain geographic points along the way; and they themselves felt vulnerable to sanction and arrest. Delalawoch were not part of a lucrative or extractive industry in Ethiopia and Djibouti (see for comparison, Tinti & Westcott Reference Tinti and Westcott2016; Raineri Reference Raineri2018). Despite risks of abandonment or exploitation, as Ayyantu and others described, the greatest threat to women's lives and wellbeing remained instead the violent political, economic, social and structural conditions in women's home countries and in their destinations. Furthermore, women's irregular migrations in this region, while dangerous, were at the same time key to their survival and independence. They repeatedly expressed gratitude for the guidance delalawoch and various service providers could provide.

Finally, as previously described, the IOM in Ethiopia and Djibouti currently use methods of ‘migrant deterrence’ and ‘sensitisation’ to frighten migrants both prior to undertaking their journeys or along the route (IOM 2015). Yet no evidence or research exists to prove these methods are effective. The potential ineffectiveness of these methods and their continuing popularity, however, suggest an institutional blindness to the multifaceted circumstances people are escaping and negotiating en route. Our research suggests these deterrence methods ultimately fail because they do not account for migrants’ motivations and objectives – namely, to seek better lives, to achieve personal independence, and simultaneously, to escape multiple forms of violence at home. Furthermore, interventions such as fear-based educational campaigns to deter migrants’ transit, the militarisation of international borders to deter migrants’ crossings, and the criminalisation of the delalawoch women depend on for safety, cannot alone improve migrants’ protection and wellbeing. These tactics may indeed make migration more dangerous and traumatising to migrants who have already faced multiple forms of violence and oppression.

In sum, Oromo Ethiopian women migrate out of Africa to the Gulf States for work to escape the many forms of insecurity and violence they face at home, and in so doing, seek economic as well as personal and political autonomy. Therefore only when countries offer women not simply jobs or humanitarian assistance, but human rights and meaningful citizenship – in their hometowns, during migration, and also in the countries where they work abroad – will women's aspirations finally be realised. In other words, only when women such as the ones we introduce here, and others, can ‘build their own house’, safely and wherever they see fit, will dangerous irregular migrations be prevented.