Skip to main content Accessibility help


  • Access


      • Send article to Kindle

        To send this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle. Find out more about sending to your Kindle.

        Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

        Find out more about the Kindle Personal Document Service.

        Available formats

        Send article to Dropbox

        To send this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Dropbox.

        Available formats

        Send article to Google Drive

        To send this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Google Drive.

        Available formats
Export citation


Drawing on ethnographic research conducted in Djibouti's Markazi camp for refugees from Yemen between 2016 and 2018, this article examines the complex motivating factors that drove a subset of Yemenis to seek refuge in the Horn of Africa. Although the primary reason for their flight to the Horn of Africa was the ongoing war, a secondary but not inconsequential driver of many of these Yemeni refugees’ current displacement was their family histories of transnational migrations and interethnic marriages. This article argues that, for this group, it was their “mixed” (muwallad) Arab and African parentage and resulting alienation in Yemen that made their flight imaginable—and, in their view, imperative. Although “mixed motive migration” is not unusual, this example underscores how spatial and social (im)mobilities in Yemen and the Horn of Africa region have been co-constituted across generations. More importantly, it has critical implications for the recently adopted Global Compact on Refugees, which promotes (among other solutions) the “local integration” of refugees in their proximate host societies.

If you asked me why I fled Yemen and I were to tell you that I fled from the war—I'd be a liar. No—even before the war began, I had always hoped I would leave Yemen. But if I had fled then, no country would have taken me as a refugee—the United Nations wouldn't have taken me. It would've said, “there's no war in your country.” And yet, war was inevitable. In Yemen, we were suffering. Many Yemenis, millions of them, including me, were suffering. Suffering! I swear to God, I swear. I felt as if I wasn't a human being, as if I wasn't even Yemeni. I was [like] a refugee in Yemen, but without international recognition. A virtual refugee. Now I am a real refugee.

–Ahmad (a refugee from Taiz)1

It is widely recognized that the current war in Yemen has precipitated the greatest humanitarian crisis of our time.2 In addition to having lived through more than four years of armed conflict, millions of Yemenis have been suffering from chronic hunger, large-scale cholera and diphtheria epidemics, and critical shortages of medicines, fuel, and water.3 Nearly five million Yemenis—more than 15 percent of the country's population—have been internally displaced. And approximately 190,000 individuals have officially sought refuge in neighboring countries. Given the duration and scale of this crisis, it is striking that the numbers of persons fleeing Yemen have not been larger. Some of the reasons for this relatively small—and yet, still significant—outmigration of refugees from Yemen include the closure of Yemen's geopolitical borders and airspace; the mobility restrictions, financial costs, and physical hazards of traveling through Yemen, and across the sea; and the reluctance of many Yemenis to leave their homeland.

What this article examines, however, are the complex (“mixed”) motivating factors that drove a subset of Yemenis and foreign nationals living in Yemen to seek refuge in the Horn of Africa—even as refugees and migrants from the Horn of Africa continued to seek asylum in or passage through Yemen. Certainly, the primary reason for the emergence of Yemeni refugees in the Horn of Africa must be attributed to “the war” itself—and, indeed, entire coastal villages crossed the Red Sea and the Gulf of Aden to escape bombs and missiles. However, a secondary but not inconsequential driver of many Yemenis’ current external displacement was—as Ahmad, quoted in the epigraph, suggests—their lifelong feelings of abjection and alienation, even within Yemen. Drawing on ethnographic fieldwork in the Markazi camp for Yemeni refugees in Djibouti, this article argues that, for many, it was their family histories of transpelagic migrations, transnational marriages, and hybrid (s./pl. muwallad/īn: “mixed” Arab and African) ethnicities that made their flight imaginable—and, in their view, imperative. This history does not simply complicate their own experience of refuge. It underscores how spatial mobility and social mobility are co-constituted and evolving; how, in this case, these refugees’ and their (fore)fathers’ prior corporeal migrations resulted in a social immobility, engendering new migratory movements and new forms of disrupted mobilities (spatial, physical, social, imagined, and virtual): a likely catalyst, once again, for future transnational migrations.4 Indeed, as the trajectories of these “mixed-blood” (muwallad) Yemenis demonstrate, many fled persecution, conflict, economic hardship, and social alienation in pursuit of dignity. In doing so, they were moving also “across the binary categorizations” of “forced migrant” and “social or economic migrant.”5 This has critical implications for the value that these Yemenis attribute to their hard-won refugee status and for the push by the United Nation High Commissioner for Refugees (UNHCR) for their local “integration,” a key part of its new Global Compact on Refugees that many of the individuals I interviewed reject.

In what follows, I provide a brief overview of Yemen's history as a migrant-producing and refugee-hosting nation prior to the emergence of its own refugee population in 2015. I then describe the establishment and conditions of Djibouti's Markazi camp, the only UNHCR-managed camp for refugees from Yemen, where between December 2016 and March 2018 I interviewed members of one hundred refugee households, compromising nearly one-third of the camp's population.6 Although Markazi is a relatively small camp, it hosts refugees from former North and South Yemen—areas that are positioned on opposing sides of the so-called civil war. My initial goal was to understand why its inhabitants had fled Yemen when, in many cases, their relatives and neighbors had fled internally, from urban centers to their ancestral villages, for example. I also aimed to understand their migratory “choices,” if any: whether (and why) some had sought refuge in Djibouti instead of in other countries in the region. It soon became clear that as, if not more, divisive than any putative North–South divisions are the distinctions drawn by camp residents between those who consider themselves “pure” Yemeni (aṣlī) and those with “mixed” (muwallad) binational, biracial identity. After discussing some of the migration experiences that influenced this particular population's decision to flee Yemen and seek asylum in Djibouti, I address the implications of current developments in international refugee policy.

Recognizing the knowledge produced from this inquiry to be partial, situated, and contradictory, I acknowledge that my interlocutors’ responses—and questions—may have been calculated to achieve certain outcomes. Specifically, the refugees in Markazi are likely to have shaped their narratives according to what they thought visiting officials, journalists, and researchers would want to hear and could do for them. This means that claims of lifelong discrimination may be exaggerated or strategic. Even so, as these refugees seek to convince others of their reasons for having fled Yemen, they also convince themselves. This is evident in Ahmad's steadfast conviction that he will “endure all of this suffering” until he “eventually finds success”: success, meaning for Ahmad, and others like him, “being resettled in a country that respects its citizens, a place where my children can have a proper future.”


The territory that is now the Republic of Yemen has long been known for producing diasporas. As early as the 13th century, Hadhrami seamen, merchants, and religious leaders migrated around the Indian Ocean basin, from the Horn of Africa and the Swahili coast to India and the Malay Archipelago.7 In the late 19th and early 20th centuries, Yemenis seeking to escape famine, poverty, high taxation, and political repression in both North and South Yemen migrated to the port city of Aden and, from there, to the Horn of Africa, where they found work as porters, clerks, or merchants and often married “local” women.8 Others established emigrant communities around and across the Atlantic: in the United Kingdom, where many settled as sailors and dockworkers in port cities and as steelworkers in the British Midlands9; and in the United States, where many settled as automobile workers in the Midwest and as agricultural laborers in New York and California.10 Moreover, by the 1970s and 1980s, a significant portion of the male labor force in both Yemens had migrated to the petroleum-rich cities of the Arab Gulf.11 This economically driven outmigration was so massive that the Yemens suffered their own labor shortages, resulting in high domestic wages and economic underdevelopment at home.12

One of the first notable reversals of these waves of emigration from North and South Yemen occurred also in the 1970s, when Yemeni labor migrants began returning from the Horn of Africa due to these newly independent states’ nationalization policies and rising anti-Arab sentiment.13 The return of Northern Yemenis was even courted by then President al-Hamdi (r. 1974–77), who sought to attract educated Yemenis into the labor force. However, despite Hamdi's promises of employment, free housing, and equal opportunities, these returning migrants were often blamed for Yemen's social, economic, and political problems.14 The second significant return migration occurred in 1990–91, when more than 800,000 Yemeni migrants were effectively expelled from Saudi Arabia and other Gulf Cooperation Council (hereafter, Gulf) countries as a result of the newly unified Republic of Yemen's support for Saddam Husayn during the 1990 Gulf War.15 This influx of largely unskilled returned migrants cost Yemen its remittances at the same time that it led to a 7 percent increase in Yemen's population and a 30 percent increase in its labor force.16 And, again, the migrants were blamed for the country's rising unemployment. Moreover, it was during this critical period that the Republic of Yemen began hosting large numbers of Iraqi and Somali refugees.17 Along with being the only country in the Arabian Peninsula that is signatory to the 1951 Refugee Convention and its 1967 Protocol, Yemen generously accepted Somali refugees on a prima facie basis, with few stipulations. It had also accepted Eritrean refugees and asylum seekers before that (many of whom fled to Yemen during Eritrea's war of secession from Ethiopia, 1961–91) and would begin hosting Ethiopian refugees in the mid 2000s. By 2006, Yemen had the world's tenth-highest refugee population per one USD gross domestic product (GDP) per capita (purchasing power parity).18 Meanwhile, tens of thousands of unabsorbed Yemeni returnees were still living in the makeshift settlements they had set up on the outskirts of al-Hudaydah and Aden more than a decade earlier. In some areas, such as the pointedly named Saddam (Husayn) Street in al-Hudaydah and in the Basateen neighborhood in Aden, Yemeni returnees and Somali refugees were living side by side. All this is to say that Yemenis have long conceived of themselves and their fellow compatriots as migrants. A Yemeni saying, “the Yemeni is an eternal migrant” (al-yamanī al-muhājir al-dā ʾim), reflects the sense in Yemen that “migration, a result of economic and political discrimination in the homeland, is a permanent, if unwelcome, fate.”19 But as several Yemenis living in Djibouti told me, they had never imagined themselves as “real” (legally recognized) refugees.

This migratory imaginary changed abruptly in 2014, when the outbreak of the civil war led to unprecedented displacement within the country and the first outflow of refugees from Yemen.20 Since March 2015, nearly 4,900,000 persons—more than 15 percent of the country's population—have been internally displaced.21 In addition, more than 190,000 refugees and asylum seekers have fled to neighboring countries: Oman, Saudi Arabia, Djibouti, Ethiopia, Somalia, and Sudan.22 (Meanwhile, around 280,000 UNHCR-assisted refugees and asylum, most of them from the Horn of Africa, remain in Yemen, unable or unwilling to return to their country of origin. And tens of thousands of Ethiopian migrants continue to enter Yemen irregularly each year hoping to reach and secure work in Saudi Arabia.23) Most UNHCR-registered arrivals in the Horn of Africa were nationals from Somalia and Ethiopia. But among the approximately 100,000 arrivals from Yemen in the Horn of Africa by October 2017, nearly 30,000 were Yemeni nationals.24 In Ethiopia, Somalia, Sudan, these Yemeni refugees reside primarily in urban centers where they seek work, rent accommodations, and enroll their children in local public or private schools, with limited UNCHR support in the form of subsistence allowances, health care assistance, and livelihood program enlistments.25 In Djibouti, Yemeni refugees are divided between Djibouti City and Obock, a tumbledown port that saw its glory as the initial capital (1884–94) of French Somaliland. Situated just hours by motorboat from the Red Sea coastal villages of Yemen, Obock runs—even now—on fresh produce, processed foods, and subsidized fuel imported from Yemen. Although several Yemenis have opened shops and other businesses in Obock, most of the refugees in this region live in “Markazi,” the gated UNHCR camp located five kilometers west of the town. The town is also the major gateway for Ethiopian migrants—many of them unaccompanied minors—seeking passage to Yemen and Saudi Arabia. Having crossed the desert on foot carrying little more than a water bottle, these boys rely on the scraps of the refugees living in Obock and in Markazi. Their relative privilege as (Arab) “refugees”—who receive aid from several Gulf-based humanitarian organizations—is thus abundantly clear to the Yemenis who pity the (African) “migrants” begging for food through their camp's chain-link fence.


A blue signboard in front of the camp's guard station welcomes visitors and donors to Markazi. Markazi camp was established in late April 2015 to accommodate the arrival of nearly 900 persons from Yemen in northern Djibouti following the escalation of the conflict in March 2015.26 Initially, the arrivals were housed in an orphanage run by the Kuwait's Al-Rahma International Charity, a sports stadium, or the nearby International Organization for Migration (IOM) Migrant Reception Center, separated according to their origins. Once registered by the UNCHR and Office National d'Assistance aux Réfugiés et Sinistrés (ONARS), the Djibouti government's refugee entity, the refugees were transferred to the newly erected camp where they were allocated tents and food rations. By mid-June 2015, of the 2,110 persons arriving from Yemen who had claimed refugee status, 1,575 were officially living in Markazi camp. (The rest were living with relatives or renting apartments in Djibouti city.) Although Markazi was operating at its then current capacity, the government expected it to host more than 5,000 refugees and prepared the camp for that eventuality.27

What officials—or, at least, the refugees—were less prepared for were the scorching desert (khamsīn) winds that darken the skies and scoured the UNCHR-issue tents to shreds. Refugees still living in Markazi today recall their first stifling summer and say that this is when many of the camp's residents decided to return to Yemen. Others sought shelter with family or friends in the city or depleted their scarce savings to rent rooms in Obock town. A headcount conducted by the UNHCR in mid-September found only 581 refugees living in Markazi.28 Meanwhile, Gulf-based charities had rushed to their fellow Arabs’ aid. Bahrain's Royal Charity Organization donated colorfully patterned “Sahara tents” to replace the standard-issue UNCHR tents. And the Qatari Red Crescent donated 300 Refugee Housing Units designed by the IKEA Foundation.29 Although these flat-pack, steel-framed “Better Shelter” units have won several accolades for their innovative design, their polypropylene siding makes them oppressively hot in Djibouti's tropical desert climate.30 Many families use these, what they call “caravans,” not for shelter, but as kitchens or storage units—and, in one case, as a chicken coop. Nevertheless, as a result of these various donations, most families have managed to cobble together a housing compound that accommodates their specific needs. In addition, the Markazi refugees have received relatively copious amounts of charitable aid and private donations—especially in contrast to the Somali and Ethiopian refugees languishing in Djibouti's older, underfunded, and overpopulated camps. “It's a five-star camp,” an IOM official told me, after describing Markazi's outsized funding.31 “I've never seen anything like it,” a UNHCR representative exclaimed.32

“There is no other camp like this!” Walid reiterated—in words, not meaning—as we sat under a plywood shelter waiting for the distribution line to thin out. “We call it the camp of the oppressed,” his friend Nabil explained: “we may have fled death, but now we are living in hell.” Those who cannot return to Yemen and are not used to the coastal climate describe feeling imprisoned by the camp's remote location and harsh environment. “It was like we were chickens in a microwave,” one woman said, describing the summer months in her caravan. Others showed me photographs they had taken of the scorpions and snakes that entered their tents during this period. “We used to take our children to parks, to go see animals in a zoo … Now it's like we are in the zoo, and the animals come to gaze at us!” Rashid continued, pointing to the troops of baboons stationed atop the perimeter fence. This sense of living in a prison or, even worse, a zoo—in the sense that one's world has been turned upside down—is compounded for many by their effective internment in Obock, in a country they view as even less developed than Yemen. Although officially the refugees have been granted freedom of movement and the right to work (along with access to education and social services), in practice they must obtain permission from the ONARS camp managers or the prefect's office in order to board the ferry to Djibouti City. (Some admitted that they bribe their way aboard.) Those lucky enough to secure coveted camp labor as a teacher, health assistant, NGO “volunteer,” or day laborer are paid a maximum of 225 USD per month33; to work legally in Djibouti City, one must obtain a work permit, which few can afford. As a result, many of the camp residents are unemployed, bored, anxious or depressed, and prone to conspiratorial thinking.

It is also clear that few had expected to languish in Markazi as long as they have. Four years after the camp's opening, only one family has been officially resettled to a third country (in Europe). A handful of families, anxious about the extended disruption of their university-aged children's education, moved to Egypt; others, tired of languishing in the camp, returned to Yemen. But many are now experiencing and suffering a halted mobility and increasing constraints on several fronts: physical and legal constraints on migrating onward; socioeconomic constraints on making a living, much less a livable life; and technological constraints on accessing legal counsel or even the Internet.34 Several complained less of stasis than of moving “backwards.” As Walid put it: “There are good things here in Obock, like security. We can sleep at night. But we have gone backwards. The Syrians, they ended up in a better place—in Turkey.” Describing Obock as a “Stone Age” town, the refugees are well aware of the foreign aid and employment that their presence has conjured. Some, feeling trapped, fear that their encampment is permanent—that they will be prevented from being resettled. “We're like the goose that lays the golden egg,” several people repeatedly told me. “We're developing the people of Obock,” Rashid said. “They're the Bedouin—they're the ones benefitting. We're like a football tossed between Yemen, Djibouti, and Saudi Arabia.”


At the time of writing (in spring 2019), Markazi hosts around 1,250 Yemeni refugees.35 When the camp was initially set up, it was organized into four sections, roughly peopled by those from the same region. UNCHR introduced this geographic organization, I was told, because it feared that the North–South hostility in Yemen would permeate the camp. What these divisions belie, however, are the tensions that cut across geographic regions: growing rifts between the urban and rural populations, the educated and the undereducated, and Yemen's rigid social classes. They also obscure the histories of transnational migrations and hybrid backgrounds that transect each region, such as the large number of muwalladīn families: in this case, Yemeni citizens with African roots. At present, most of the camp's residents originate from Dhubab, a small Red Sea town that has been hit numerous times by airstrikes, and from fishing villages in the Bab al-Mandab region overlooking the Bab al-Mandab strait, a key “chokepoint” between the Arabian Peninsula and the Horn of Africa. Another group comes from Wahija, a coastal village south of al-Mukha where more than forty civilians were killed by airstrikes on a wedding party on 28 September 2015.36 All three of these origin points—Wahija, Dhubab, and Bab al-Mandab—are located in Yemen's Taiz governorate, one of the main frontlines in the war between the Ansar Allah (Houthis) and the progovernment forces. A smaller percentage of the camp's residents come from Yemen's three largest urban centers: Taiz, Aden, and even Sanaa. Where the refugees originate from correlates, in large part—but not conclusively—with where they hope to end up. A thirty-three-year-old man from Wahija (whose mother's sleeping body was shredded by shrapnel on the night of the wedding bombing) told me that the camp's population is divided into three factions: the coastal inhabitants (aṣḥāb al-sāḥil), who wish to return to their fishing villages in Yemen; the people of Dhubab, who cannot return because their houses were destroyed and the area was transformed into a military zone; and the urbanites (ʿ ayāl al-mudun) who long for resettlement in a “first-world” country.37

While these three outcomes do not correlate exclusively with the desires of each person so categorized, they do correspond almost directly to the three “durable solutions” to refugee protection promoted by the UNHCR: voluntary repatriation, local integration, and third-country resettlement.38 Many of the camp's residents say that they had anticipated returning to their homeland within a few months, if not weeks. However, four years into the war, it has become increasingly evident that neither the repatriation nor the resettlement of Yemen's refugee population is viable: the conflict remains intractable and few third countries are accepting refugees from Yemen. This leaves local integration as the only “durable” solution. To this end, the Government of Djibouti—which has already extended prima facie recognition to refugees from Somalia and Yemen and currently hosts more than 29,000 refugees and asylum seekers (mainly from Somalia, Ethiopia, Eritrea, and Yemen) despite the country's endemic poverty and high unemployment—has taken progressive steps toward socio-economically “integrating” refugees into the surrounding communities.39 In January 2017, President Guelleh promulgated new refugee laws aiming to safeguard and strengthen the refugees’ access to education, healthcare, employment, and eventual naturalization.40 Most immediately, these laws pave the way for the inclusion of refugees in the national health and education systems—a move that, in the realm of education, will enable refugees following the Djiboutian curriculum to receive nationally recognized certificates to facilitate their employment. (Previously, camp schools were funded and managed by international nongovernmental organizations that had relied on the English-language Kenyan curriculum in camps for Somali, Ethiopian, and Eritrean refugees and the Arabic-language Yemeni curriculum in Markazi. Health care, too, had been the provenance of international NGOs.) In December 2017, the new laws came into effect by presidential decree.41 And, to much UN fanfare, the government announced that the country's camps would henceforth be considered villages.42

This semantic shift is being accompanied in Markazi camp—now village—by a structural transition from their temporary tents to more durable air-conditioned residential housing units. Funded by Saudi Arabia's King Salman Humanitarian Aid and Relief Center (KSRelief) and constructed by two Turkish companies, these 300 twenty-four–sq. meter composite container homes are presented as being “tailored to” and more “suitable for” refugees in such a harsh desert climate.43 Yet many of the residents of Markazi would disagree. In their view, the houses are too small and too closely packed together—like “biscuit boxes,” a man from Mawza told me—to make them an attractive alternative to their makeshift but roomy tent-and-caravan compounds. For large families and those from rural areas used to gender segregation, outdoor latrines, and ground ovens, these one-bedroom structures with their miniature indoor kitchens and lavatories are culturally, socially, and architecturally unsuited. Even urbanized couples with fewer children worry that these unmodifiable containers will increase their discomfort—especially were there to be a summer power failure. More importantly, even comfortable houses and increased government services cannot make up for the camp's barren location, the town's impoverishment, and the country's lack of economic and educational opportunities for refugees from Yemen. “If the Saudis gave us a villa, I would still refuse to move [out of this tent],” a woman from Sanaa insisted. For her and others who yearn to be resettled in Europe or North America, these houses represent a gilded cage. Many of Markazi's residents believe that if they move physically from their impermanent tents to these durable containers (now called “the Saudi Village”), they will be moving jurisdictionally from the shelter of international protection to the de facto prison of permanent displacement. This concern has been compounded by visiting Yemeni ministers’ insistence that the Markazi residents are merely “displaced”—a renunciation of their refugee status seemingly echoed in the Djibouti government's determination that the camp has become a village. “We're not refugees,” a man parroted sarcastically during a ministerial visit to the camp in January 2017; “we're displaced. Don't you know that Obock is a governorate of Yemen?!” “They say this as if we just moved from one city to another,” his friend replied. “But we crossed the sea! We fled from Yemen!”

Accordingly, for a small but vocal population in Markazi, neither a “five-star” camp nor an emerging “village” approximates a viable solution. Certainly, most of the population fled the acute brutalities of war: bombs dropping on their neighborhoods; missiles hitting their loved ones at distribution points; Houthi rebels threatening to conscript their young sons; corpses lying in the streets. Many individuals I interviewed lost their immediate family members and relatives to this war. Some lost them in the literal sense; separated from their spouses or children, they do not know whether they are still alive, much less their whereabouts. Several families fled in need of critical medical care; following the naval and air blockade of Yemen in March 2015, they left by any way possible to seek medical operations abroad. Some young men fled alone to avoid being conscripted by the Houthis or the army. Others are seeking political asylum. For these persons, improved housing may provide temporary relief. But of the one hundred households I interviewed in 2017, more than one third insist that they will never return to Yemen and reject the plans for their “integration” (indimāj) in Djibouti.44 What this population seeks—many of them educated and urban, but many also the so-called hybrid descendants (muwallad/īn) of both “Arab” and “African” parents—is refuge from generations of cross-sea migration. In the view of those born to Yemeni fathers and Eritrean, Ethiopian, or Somali mothers, neither Yemen nor the countries in the Horn of Africa had ever fully integrated them. This experience of socioeconomic marginalization—not the war, itself—is why many had traveled to Djibouti to become UNHCR-recognized refugees: refugees who, through the promise of third-country resettlement and access to meaningful citizenship, hoped to escape their (often involuntary) migration pasts.


The Yemeni men who emigrated in the late 19th and early to mid 20th century had been born into—and, in many cases, were seeking to escape—a rigid, hierarchical system of social stratification. Prior to the 1962 overthrow of the Imamate in North Yemen and the 1967 revolution against British colonial rule in the South, Yemen's traditional social order was comprised of several caste-like status groups based on descent and profession. Whereas the religious elites (sāda) and the land-owning tribes were distinguished by their Prophetic or otherwise “pure” (Arab) lineage, the lowest social classes—the ʿabīd (slaves) and the akhdām (servants)—were ostracized for their “presumed African ancestry.”45 Although the two republics to emerge following the 1962 and 1967 revolutions sought to eradicate these status differentials and replace them with a unified and homogenous citizenry, the regime of ʿAli ʿAbd Allah Salih exalted Yemen's tribal identity.46 At the same time, the akhdām—now referred to as the muhammashīn (“the marginalized ones”)—continued to be treated as social outcasts. Moreover, the migrants who later returned to Yemen with their Asian or African wives and their biracial children found that their wives were often marginalized and their muwallad children mistrusted.47 This was especially the case for the Yemeni migrants and their muwallad children returning from the Horn of Africa due to the stigma attached to Yemenis with African ancestry and to Yemenis with darker skin.48

So prevalent were these experiences—of dispossession and oppression in prerevolution Yemen; of long-term emigration as a short-lived escape; and of the marginalization of the muwallad children born of these migratory pathways—that they emerged as the major literary themes in Yemen's early (1970s) novels.49 Muhammad ʿAbd al-Wali (1940–73), one of the most prominent Yemeni authors of this pioneering “seventies generation,” dramatized his own experiences as the son of a Yemeni émigré father and an Ethiopian mother in his celebrated 1971 novella, They Die Strangers (Yamūtūn Ghuraba ʾ). Writing about the grandiose dreams but solitary death of ʿAbdu Saʿid, a destitute Yemeni storekeeper in Addis Ababa, ʿAbd al-Wali underscores the alienation of Yemeni migrants—and especially, their “half-breed” (muwallad) children—both in their homeland and abroad. As one of the novella's characters, himself the muwallad son of a Yemeni migrant and an Ethiopian mother, reflected:

[ʿAbdu Saʿid] died like a stranger, like hundreds of Yemenis die in other lands. As for his grave—it's not his. It's not his land; it belongs to other people, to the Ethiopians. We even occupy their graveyards. As if it's not enough to swallow the meager scraps meant for their mouths, we occupy their graves, too. Oh God, we are such strangers!50

The anguish of being “a stranger”—not just a temporarily displaced migrant, but the permanently rootless “half-breed”—clung to these muwallad children even following their eventual “return” to Yemen. It was this social alienation, compounded by racial discrimination, a patronage system favoring those with strong tribal connections, and the continued scapegoating of migrants and muwalladīn for Yemen's social and economic problems, that made people like Ahmad (who insists he is not muwallad) feel as if he were “not even Yemeni.” As a person with “dark” skin and a “weak” tribal affiliation, Ahmad claims to have been treated as if here were a khādim (“servant,” of akhdām status)—as if he were a “virtual refugee.” Thus, for Ahmad, and many other socially, economically, and politically marginalized Yemenis, the occasion to become a bona fide refugee with international protection and the prospects for third-country resettlement was an opportunity worth fleeing for.

To illustrate the mixed migration motives of Yemen's muwalladīn refugees, I will present three migration narratives, one involving Ethiopia, one Somalia, and one Eritrea—Djibouti's three neighboring countries to which these persons of “mixed blood” might have otherwise “returned.” Although my focus here on male trajectories is intentional in that, until recently, nationality in Yemen was passed primarily through patrilineal descent, there is much to be said about women's (im)mobilities and the ways in which migration is gendered.51 My aim here, however, is to underscore why camp residents (men and women) view their refugee status as a potential pathway toward self-realization—and not, as the term “refugee” commonly connotes, as a stamp of racialized abjection.52

Take the story of Yusuf, for example. Both of Yusuf's parents were born in Addis Ababa to Ethiopian mothers and to Yemeni fathers from Taiz who had migrated there in the early 20th century to escape “the poverty, chaos, illiteracy, and unemployment” under the Imamate.53 Yusuf's mother and father married in Ethiopia and had twelve children. Yusuf, the youngest, was born just following their return to Yemen in 1981. After initially settling in Taiz, the family moved to Sanaa in 1991 in search of economic opportunities. Yusuf completed secondary school, played football for several years, and began working for various companies. In 2008, he met Naima, an Ethiopian woman from Addis Ababa who had moved to Sanaa for work. They were living in the same apartment building and fell in love, he told me, beaming as he glanced her way. They wed in 2012, despite his parents not approving of the marriage. “They said you could have married better, why go back to Ethiopia,” he recalled. Not only is Naima a Christian, but Yusuf's parents did not want his children to face the same discrimination they had suffered, raising “half-breed” children. Eventually, Yusuf's Yemeni-Ethiopian mother warmed to Naima and, once their son Hamza was born in 2014, his Yemeni-Ethiopian father did, too.

Before his marriage, Yusuf had been active with the revolutionary youth. Two of his friends were among those lost their lives on the “Friday of Dignity” (18 March 2011); Yusuf, who had been nearby, narrowly escaped the massacre himself.54 Three years later, in August 2015, missiles struck their neighborhood in Sanaa and a projectile hurled into their bedroom. “It was like Judgment Day,” Yusuf said, describing the fire and smoke ravaging their apartment. They grabbed some clothes and Yusuf's treasured jambīya (dagger) and moved to Taiz to stay with relatives. But in Taiz, the fighting was worse. Yusuf had seen Al Jazeera's media coverage of the Markazi camp, which had opened five months earlier, and had even heard that it was a place “unfit for animals.” But Yusuf and Naima felt they had suffered enough discrimination in Yemen and they worried for Hamza's future. So, a few days later, they traveled to al-Mukha and boarded a boat to Djibouti—as if he was following in the steps of his grandfathers, he reflected.

But Yusuf does not intend to remain in Djibouti. Nor is he considering the idea of moving onward to Ethiopia, where he has never lived, and where it is similarly difficult to make a living. Instead, he views life in the camp as a necessary if painful step to a new beginning outside of the Yemen and Horn of Africa region. “Returning to Yemen is impossible for me, for us, because of what my wife and I have suffered—because of society's disdain for me because I married a Christian,” he said. “It is difficult for me to live in such a tribal, patriarchal, shaykh-dominated society. I want to live in a society comprised of all civilizations, all cultures, all types, all nationalities.” He then quoted the esteemed Yemeni poet ʿAbd Allah al-Baraduni (1929–99): “‘Yemenis’ they call us, when we are in exile. Exiled we are, when we are in Yemen (al-yamaniyyūn fī al-manfā wa-manfiyyūn fī al-yaman).”55 “What does this mean to you?” I asked Yusuf. “If a person is in Yemen, he is in exile, and if he is outside of Yemen, he is also in exile,” he replied. “The Yemeni has no value…. Anywhere he goes in the world, he is deemed worthless by others because his own state and president consider him worthless. This is very difficult. It is for this reason that I have decided never to return Yemen, ever.” Yusuf and Naima plan to wait as long as it takes to be resettled—“twenty years even, if necessary.” Meanwhile, their four-year old son has adopted a feral cat: its English-language name is “Travel” (sāfir, the Arabic word for travel, is also the term the Markazi refugees use for resettlement).

In contrast to Yusuf who was born in Yemen, Salim was born and raised in Mogadishu. Salim's grandfather had migrated from Yemen to Somalia in search of work and had married a Somali woman there. Salim's father was born in Mogadishu; he too married a Somali woman, with whom he had five children, including Salim, born in 1963. Salim's sister Fatima died in childbirth a few years prior to the start of the Somali civil war. Two years later, his sister Asma was assaulted and shot in the head by a gang; she was twenty years old. Shortly thereafter, the vehicle in which his sister Shirin was traveling from Mogadishu to Bosaso was attacked and crashed; Shirin died. Moreover, once the war started, Salim's surviving family no longer felt welcome in Mogadishu. “We had become strangers,” he told me in English, as if recalling the words of ʿAbd al-Wali.56 During the war, their family dispersed. Salim's mother and older brother made their way to Libya in the hope of reaching Italy, but were unsuccessful. Salim and his father left for Aden and moved into al-Basateen, a neighborhood populated by the Somali refugees who had fled to Yemen following the collapse of the Somali state in 1991 and the Yemeni labor migrants who had been forced to leave Saudi Arabia following Yemen's support for Saddam Husayn in 1990. There, Salim and his father continued to face discrimination based on their skin color and status:

We could not get a good life in Yemen. We faced a lot of discrimination. When you're black, nobody trusts you. You need to have a tribe, you know? In Yemen, if you have a strong tribe, you get anything. If you don't—even if you do have a big tribe, but if you're black, you will be treated like a Negro. No one will respect you. That's the truth. So, what can you do? You need to survive, you understand? […] When all the doors close in your face and you have family, you need to do anything you can. Not once—I never felt freedom in Yemen. I never had one year with work and without any problems. Not even a year! Every two months, three months, there was a new problem … Even when I'd get work, for example, they'd ask me why I am black; why you are from Somalia? So, I would become nervous a lot of time, you understand? Until right now even, I am working here—right now, we have discrimination here.57

When the civil war broke out in Yemen in 1994, Salim and his ailing father moved to Sanaa. A few years later, Salim heard that his mother and brother had returned from Libya and moved to Aden; there, his mother died, and then his brother died, before he was able to see them again. Meanwhile, in Sanaa, Salim's situation was no better, so he and his father returned to al-Basateen, where Salim met his wife, Hayat, a refugee from northern Somalia. In 2008, they married; a year later, she gave birth to twins: a boy and a girl. Salim moved his family, his father, and his wife's mother to Sanaa in search of work. In Sanaa, Salim taught English in a government school and their life was stable for a time.

Until this war began. That is when a neighbor came to his house and said: “you are from the south; you must leave, or you'll face danger.” Salim placed his father and five-year-old twins on a bus to Taiz; he and Hayat and her mother would follow later. A Houthi-fired missile struck the bus and killed all twenty-five passengers. Two days after Salim and Hayat visited the mass grave, they left Yemen. Salim, who now lives in Markazi camp with his wife and his mother-in-law, has lost every member of his birth family and his two children. He insists that he cannot return to Yemen, or to Mogadishu, or even to Hargeysa (Somaliland), where his mother-in-law comes from, because they no longer have family or other connections there and he is descended (maternally) from a southern clan. “In Somalia, they hate us and, in Yemen, they hate us. People in both countries discriminate against the muwalladīn, but it's not their fault they have two bloods. It's the fault of their grandfathers who migrated,” he had told me in January 2017. “My grandfather was a stranger and now we are strangers,” Salim reflected bitterly.58

Like Yusuf and Salim, forty-year-old Nuh is part-Yemeni and part-African; unlike Yusuf and Salim, he is not a Yemeni national, but an Eritrean citizen—and a former refugee in Yemen. Nuh's maternal grandfather was a migrant from Aden who had moved to Assab during the British colonial period and worked as a fisherman there. Nuh's mother, born to this Yemeni migrant and an Eritrean mother, married an Eritrean doctor: Nuh's father. Nuh attended school in Assab through sixth grade, after which he worked in the salt trade. He met his Yemeni wife, Jamila, who was born to a Yemeni father and an Ethiopian mother in al-Hudaydah (Yemen), when she was visiting her mother in Assab in 1990. They married and lived in Assab, where she gave birth to her two eldest of six children. During the Eritrean–Ethiopian war in 1998, Nuh was conscripted and shot in the arm by a fellow Eritrean soldier. After his wound healed, the government forced him to return to service, but he fled to Assab. There, he discovered that Jamila's father had come to fetch her, so Nuh traveled by boat to al-Hudaydah to reunite with his family. In 2000, they all moved to Sanaa, where eventually Nuh was recognized by the UNCHR as a political refugee. Once he obtained this status, his mother in Assab sold the land he had owned there and sent him the proceeds. With this, Nuh bought a car and worked as a taxi driver; later, he opened an office for domestic workers from Ethiopia. They were doing well; “we were better off than many Yemenis around us,” he said.59

Then the war started and his house shook from the bombs dropping around it. All their neighbors were fleeing, so they fled, too. Nuh thought that if he arrived in Djibouti as a (second-time) refugee from Yemen, a third country would surely accept him for resettlement. Nuh claims that he cannot return to Eritrea because of his status as a deserter; he cannot move to Addis Ababa, where his mother-in-law lives, because of his former activities as an Eritrean soldier; and he sees no future for his children in Yemen. But in Djibouti, he has little work and limited mobility. Nuh compares his current spatial and social immobility to his experience as a refugee in Yemen, where he could work easily, travel freely, and enroll his children in public schools. When I met him in early 2017, Nuh's mother and sister had just arrived on foot from Assab; they had paid smugglers to help them cross the border and were on their way to a refugee camp in Ethiopia.60 Since then, Nuh has been debating whether to join them. Perhaps being a triple refugee—this time as an Eritrean refugee encamped in Ethiopia—would finally open the door to his third-country resettlement in the West. Were he to succeed, Nuh would eventually send for Jamila and their children who plan to remain in Markazi until then.

These experiences of alienation—of a Yemeni-Ethiopian shunned for marrying “back into” Ethiopia; of a Yemeni-Somali fleeing war-torn Somalia for Yemen, where he was discriminated against for being black; of an Eritrean with Yemeni roots seeking asylum in Yemen—and scores like them illustrate how Yemenis’ migrations in the early 20th century are playing out a century later in the lives of families tossed around “like footballs” between Yemen and the Horn of Africa. Markazi is a center of such mixed migration histories: stories of Yemeni men migrating to East Africa to escape the desperate conditions in their country, whose Yemeni-African children returned to Yemen to flee warfare or poverty in Somalia, Ethiopia, and Eritrea, whose grandchildren are finding their way back, only to feel more immobilized than ever. In this sense, the war was, as Ahmad suggests, less the catalyst than the conduit for escaping a lifetime of forced movement, alienation, and loss. This does not make their claim to refugee status any less legitimate; rather, it points to the limitations of the UNCHR's new push for local integration—in another view, containment—in a region where ethnic, clan, and tribal-based politics have marginalized peoples for generations. And continue to do so: even in a so-called village comprised of persons forcibly displaced from Yemen, those who claim to be “real Yemenis” disparage those they consider muwallad(īn).

It is also why one of the greatest tensions in the camp today is not the conflict between so-called northerners and southerners. Rather, it is the struggle between those who accept the label “displaced”—as several visiting officials have described the refugees, suggesting that their condition is physical and temporary—and those who reject it. People like Yusuf, Salim, and Nuh can't return—or they won't. “Yemen took everything from me, and gave me nothing,” Ahmad told me. “It took my youth; it took my work; it took my house; it took my future!” All they have left, they say, is their status as refugees. The value and materiality of this status cannot be overstated; it is, literally, what they are holding onto—their refugee cards; some have two or three—as their only ticket “out.”61 (Moreover, their relative privilege as refugees is displayed to them daily by the steady march of young African “economic migrants” passing the camp en route to Yemen: migrants whose border crossings are considered “illegal.”) Thus, when visiting officials insist that the inhabitants of Markazi are merely displaced and when charities prepare to move them into fixed structures, many of the Markazi residents read this as a sign that the powers that be are colluding to keep them in their (abject) place.62 Whether these fears are well founded or the fancies of people living in a heightened state of precarity, what is real is that the Markazi refugees feel doubly incarcerated: as occupants of a securitized camp and as persons who, despite having crossed the sea, have not in fact escaped the failings of Yemen. Or the alleged faults of their grandfathers.


In this current era of unprecedented and protracted displacement, countries that have accepted and are willing to “integrate”—as opposed to encamp—refugees and asylum seekers are to be commended. (What this means essentially is that refugees are included in their host society's “existing economic and social structures rather than constructing parallel ones.”63) This is especially the case for the less affluent and smaller nations like Djibouti, which have had little choice but to take on more than their so-called share. For this reason, from a state-centric point of view, Djibouti's steps toward the refugees’ socio-economic integration are exemplary. They are also in keeping with the UN's New York Declaration for Refugees and Migrants, adopted on 3 October 2016, in which member states committed “to a more equitable sharing of the burden and responsibility for hosting and supporting the world's refugees.”64 As a result, Djibouti and eleven other member states (including Ethiopia and Somalia) agreed to begin applying a Comprehensive Refugee Response Framework (CRRF) aimed to benefit both the refugees and the communities that host them. It was, in part, their pilot experiences with the CRRF that helped guide the development of the Global Compact on Refugees affirmed by the United Nations General Assembly in December 2018.65

However, when viewed through the lens of Yemeni refugees in Djibouti, it becomes clear that this heralded development stands in tension with the experiences and aspirations of certain refugees themselves. First, these second- or third-generation children of migrants do not trust that they will ever be “integrated” or even partially included in countries or in a region from which they, their parents, or their grandparents once fled. For Yemen's socially marginalized muwalladīn and akhdām/muhammashīn, especially, displacement is neither temporal nor geographical; it is intergenerational and interiorized. For this reason, access to government health and educational services, a “village” veneer, and even promises of eventual naturalization are unlikely to resolve their feelings of life-long alienation: an “exile” and “suffering” that has only strengthened their determination to leave Yemen and the Horn of Africa altogether. Second, they see and experience this new global “burden-sharing” commitment as the extension of Fortress Europe and Fortress America: the propping open of some borders so that others can remain tightly shut. It is for this reason that many of these families and individuals crossed the sea to become “real refugees.” Having felt incarcerated by social and regional borders for generations, they have seized onto a legal category to leverage themselves out of their current regional circuits of migration and into a more open world. To say that their situation today is the “fault of their grandfathers” is not to exonerate the Salih regime, the Yemeni warlords, the foreign governments, or the international community responsible for this humanitarian catastrophe. Rather, it demonstrates these refugees’ painful awareness that their corporeal, physical, and social “displacement” is the result of problems that are structural, systemic, and deep rooted. The solution many of these refugees are holding onto, and out for, is third-country resettlement in a place—geographical, socioeconomic, and sociocultural—from where they no longer feel compelled to migrate.


Author's note: The research for this article was supported by NYU Abu Dhabi (NYUAD)’s Research Enhancement Fund and reviewed by NYUAD's Institutional Review Board. I am grateful to Djibouti's Office National d'Assistance aux Réfugiés et Sinistrés for permitting my entry to Markazi camp; to the representatives of the UNHCR, IOM, and various NGOs for meeting with me; and to the photographer Nadia Benchallal who accompanied me during much of this fieldwork. Versions of this article were presented, in 2017, at the Meeting of the Middle East Studies Association (American Institute for Yemeni Studies General Meeting) in Washington, D.C. and, in 2018, at the Arab Crossroads Studies Lecture Series at NYUAD, the NYUAD Institute in New York, the Council on Middle East Studies’ Arabian Peninsula Lecture Series at Yale University, and Yonsei University in South Korea. I thank these audiences and the anonymous reviewers and editors of IJMES for their constructive feedback. I am especially grateful to the refugees in Markazi who shared their experiences and extended their hospitality and trust. I sincerely hope they will achieve what they consider “a proper future” for themselves and their children.

1 Interview with the author (Arabic) in Markazi Camp, Obock, Djibouti, March 2018. All names are pseudonyms. Most interviews were conducted in Arabic. However, some interlocutors chose to speak English during parts of our interviews or conversations (as indicated in the respective footnotes).

2 See, for example, “Remarks by the Secretary-General to the Pledging Conference on Yemen,” 3 April 2018, accessed 1 June 2018,

3 United Nations Office for the Coordination of Humanitarian Affairs, “Humanitarian Needs Overview 2018: Yemen,” December 2017, accessed 1 June 2018,

4 Hannam, Kevin, Sheller, Mimi, and Urry, John, “Editorial: Mobilities, Immobilities, and Moorings,” Mobilities 1 (2006): 122.

5 Zimmermann, Susan E., “Reconsidering the Problem of ‘Bogus’ Refugees with ‘Socio-Economic Motivations’ for Seeking Asylum,” Mobilities 6 (2011): 341. For more on the problematic distinctions drawn between refugees, asylum seekers, and socioeconomic migrants (or between forced migration and “voluntary” migration), see Jureidini, Ray, “The Convergence of Migrants and Refugees: Western and Muslim Perspectives,” Sociology of Islam 5 (2017): 224–47.

6 During fieldwork conducted over six seasons (December 2016–January 2017, March 2017, October 2017, January 2018, March 2018), I conducted one hundred formal interviews, either through one-on-one semistructured interviews with the male or female heads of household or through group interviews with several household members at once. Although the camp's population is constantly fluctuating, it has been estimated that the camp hosts approximately 300 households at any given time (as suggested also by its recent acquisition of 300 housing units). If this is accurate, then I have interviewed close to one-third of the Markazi households—although, of course, my data comprises interviews with persons who have since left Djibouti as well as with newer arrivals. In addition to these interviews, I gathered qualitative data through group discussions with male and female coworkers and friendship groups and through participant observation in homes, shops, and public areas. Before each interview or new gathering, I explained that I had no connection to any government or relief agency and that I was therefore unable to assist anyone directly. It was especially important that I stress the academic nature of this project given that I was usually accompanied by the photographer Nadia Benchallal who was taking photographs of daily life and portraits of refugee households for an exhibit that we organized at New York University Abu Dhabi. We also gave cameras to nine refugees whom Benchallal guided through workshops and whose photographs we exhibited in Markazi and in Djibouti City. The photographic component of this research project—which deserves more analysis, one that is beyond the scope of this article—helped us to build relationships with refugees throughout the camp and to include a core group of them in this collaborative framing. See Rebecca Buxton, “How Yemeni Refugees Photographed Their World,” Refugees Deeply, 31 May 2018, accessed 22 March 2019,

7 See, e.g., Bang, Anna K., Sufis and Scholars of the Sea: Family Networks in East Africa, 1860–1925 (London: Routledge, 2003); Boxberger, Linda, On the Edge of Empire: Hadhramawt, Emigration, and the Indian Ocean, 1880s–1930s (Albany, N.Y.: State University of New York Press, 2002); Brehony, Noel, ed., Hadhramaut and Its Diaspora: Yemeni Politics, Identity and Migration (London: I.B.Tauris, 2017); Freitag, Ulrike, Indian Ocean Migrants and State Formation in Hadramout: Reforming the Homeland (Brill: Leiden, 2003); Ho, Engseng, The Graves of Tarim: Genealogy and Mobility across the Indian Ocean (Berkeley, Calif.: University of California Press, 2006); and Manger, Leif, The Hadrami Diaspora: Community-Building on the Indian Ocean Rim (Oxford: Berghahn Books, 2010).

8 Bezabeh, Samson A., Subjects of Empires/Citizens of States: Yemenis in Djibouti and Ethiopia (Cairo: American University in Cairo Press, 2016); Colton, Nora Ann, “Political and Economic Realities of Labor Migration in Yemen,” in Yemen into the 21st Century: Continuity and Change, ed. Mahdi, Kamil A., Wurth, Anna, and Lackner, Helen (Reading, UK: Ithaca Press, 2007), 5378; Miran, Jonathan, Red Sea Citizens: Cosmopolitan Society and Cultural Change in Massawa (Bloomington, Ind.: Indiana University Press, 2009); Regt, Marina de, “‘Close Ties’: Gender, Labor and Migration between Yemen and the Horn of Africa,” in Why Yemen Matters: A Society in Transition, ed. Lackner, Helen (London: Saqi Books, 2014), 287303.

9 See Halliday, Fred, Arabs in Exile: Yemeni Migrants in Urban Britain (London: I.B.Tauris, 1992); and Searle, Kevin, From Farms to Foundries: An Arab Community in Industrial Britain (Oxford: Peter Lang, 2010).

10 See Abraham, Nabeel and Shryock, Andrew, eds., Arab Detroit: From Margin to Mainstream (Detroit, Mich.: Wayne State University Press, 2000); and Friedlander, Jonathan, ed., Sojourner and Settlers: The Yemeni Immigrant Experience (Salt Lake City, Utah: University of Utah Press, 1988).

11 By 1975, as many as 1,230,000 Northern Yemenis—approximately 19 percent of its total population—had migrated to Saudi Arabia and other Gulf States. See el-Mallakh, Ragaei, The Economic Development of the Yemen Arab Republic, Reprint Edition (New York: Routledge, 2015 [1986]), 1527. Moreover, it is estimated that more than 18 percent—and perhaps as much as one-quarter—of Yemen's total labor force (male and female) was working as short-term migrant laborers in the Arab Gulf; Fergany, Nader, “Impact of Emigration on National Development in the Arab Region: The Case of the Yemen Arab Republic,” International Migration Review 16 (1982): 761.

12 Colton, “Political and Economic Realities”; Fergany, “Impact of Emigration on National Development.”

13 Bezabeh, Subjects of Empires/Citizens of States; Connie Carøe Christiansen, “Gender and Social Remittances: Return Migrants to Yemen from Eastern Africa,” Chroniques yéménites 17 (2012),; De Regt, Marina, “From Yemen to Eritrea and Back: A Twentieth-Century Family History,” Northeast African Studies 17 (2017): 2550; De Regt, “Close Ties”; Rouaud, Alain, “Pour une histoire des Arabes de Djibouti, 1896–1977,” Cahiers d’Études africaines 387 (1997): 319–48.

14 De Regt, “Close Ties”; De Regt, “From Yemen to Eritrea and Back.”

15 Colton, Nora Ann, “Homeward Bound: Yemeni Return Migration,” International Migration Review 27 (1993): 870–82; Van Hear, Nicholas, “The Socio-Economic Impact of the Involuntary Mass Return to Yemen in 1990,” Journal of Refugee Studies 7 (1994): 1838; Stevenson, Thomas B., “Yemeni Workers Come Home: Reabsorbing One Million Migrants,” Middle East Report 181 (1993): 1520.

16 Colton, “Political and Economic Realities,” 69.

17 Thiollet, Hélène, “From Migration Hub to Asylum Crisis: The Changing Dynamics of Contemporary Migration in Yemen,” in Why Yemen Matters: A Society in Transition, ed. Lackner, Helen (London: Saqi Books, 2014), 267–86.

18 The Republic of Yemen hosted seventy-nine refugees per one USD GDP (PPP) per capita. Although this was only one-tenth of what the Republic of Tanzania was hosting at the time (738 refugees per one USD GDP per capita), it was considerably more than France and Canada (four), the United Kingdom (eight), the United States (twelve), and Germany (twenty-six); UNHCR, Statistical Yearbook 2006, Chapter VII: Capacities and Contributions of Host Countries [2008], accessed 1 June 2018,

19 Halliday, Fred, preface to From Farms to Foundries: An Arab Community in Industrial Britain, by Searle, Kevin (Oxford: Peter Lang, 2010), xv.

20 Marina De Regt, “A Grim New Phase in Yemen's Migration History,” Middle East Report, 15 April 2015, accessed 22 March 2019,

21 This number, calculated in November 2018, includes 3.6 million internally displaced persons (IDPs) and the 1.28 million IDPs who have returned to their place of residence—not because conditions had improved but because they had become equally dire in the areas to which they had fled; IOM, “Yemen Displacement Report Round 37,” March 2019, accessed 8 May 2019,

22 UNHCR, “Yemen Situation Regional Refugee and Migrant Response Plan,” 31 October 2017, accessed 22 March 2019, These figures were last updated on 31 October 2017; newer figures are available from monthly reports by country but the data sets are not readily comparable.

23 In 2018, more than 160,000 East African migrants crossed the Red Sea and entered war-torn Yemen, exceeding the number of migrants (144,000) who entered Europe via the Mediterranean Sea the same year; Helen Lackner, “Migrant Crisis in Europe? Look at Yemen,” Open Democracy, 19 February 2019, accessed 22 March 2019, Most of these are Ethiopian migrants who walk from their villages to Djibouti's Red Sea port of Obock from where they are smuggled across the sea into Yemen, a crossing that took the lives of at least fifty-two migrants in early 2019 when two overcrowded boats capsized off Djibouti; Associated Press, “At Least 52 Dead After Boats Capsize Off Djibouti, U.N. Migration Agency Says,” The New York Times, 30 January 2019, accessed 22 March 2019,

24 UNHCR, “Yemen Situation Regional Refugee and Migrant Response Plan,” 31 October 2017, accessed 22 March 2019,

25 UNHCR, “Somalia: 1–31 December 2017,” 14 February 2018, accessed 1 June 2018,

26 By late May 2015, an estimated 13,000 people had arrived in Djibouti; just under half of these were Yemeni nationals (the rest were transiting third-country nationals and Djiboutian returnees). See International Federation of the Red Cross, “Emergency Plan of Action, Operation Update: Djibouti: Yemeni Refugees,” 2015, accessed 22 March 2019,; and UNHCR, “Djibouti Response Plan for Yemen Crisis, April–September 2015,” May 2015, accessed 22 March 2019,

27 UNCHR, “Yemen Situation Regional Refugee and Migrant Response Plan. January-December 2015,” December 2015, accessed 1 June 2018,

28 The same headcount found 2,768 Yemenis registered as refugees in Djibouti city; UNHCR, “Djibouti Fact Sheet,” September 2015, accessed 1 June 2018,

29 See “UNHCR/Qatar Red Crescent and Better Shelter,” n.d., accessed 1 June 2018,

30 See “MoMA Includes Better Shelter in Its Permanent Collection,” 13 October 2016, accessed 1 June 2018,; and “Beazley Designs of the Year 2016,” 24 November 2016, accessed 1 June 2018,

31 Interview with the author (English), Djibouti, 1 December 2016.

32 Interview with the author (English), Djibouti, 1 December 2016.

33 This “incentive”— of approximately ten USD per day—is the same rate received by the Ethiopian “illegal migrants” passing by the camp if they reach and work illegally in Saudi Arabia.

34 On the overlap between mobilities and forced migration—especially how “movement begets constraint, constraint begets movement, and movement occurs within constraints and constraints within movements”—see Gill, Nick, Caletrio, Javier, and Mason, Victoria, “Introduction: Mobilities and Forced Migration,” Mobilities 6 (2011): 301–16.

35 The total number of registered refugees from Yemen in Djibouti at the end of December 2018 was 4,916; UNHCR, “Djibouti Fact Sheet, January 2019,” accessed 22 March 2019, Between March 2015 and October 2017, an estimated 38,000 refugees entered Djibouti; of these approximately 20,000 were Yemeni nationals; UNHCR, “Yemen Situation Regional Refugee and Migrant Response Plan,” 31 October 2017, accessed 22 March 2019, Markazi's population rose to around 3,000 toward the end of 2015, but nearly half of its residents returned to Yemen in early 2016 following assurances by Gulf countries of assistance at home; UNHCR representative, interview with the author, Djibouti, 1 December 2016. See also UNCHR, “Yemen Situation Regional Refugee and Migrant Response Plan: January–December 2016,” December 2015, accessed 22 March 2019,

36 Shuaib Almosawa and Kareem Fahim, “Airstrikes in Yemen Hit Wedding Party, Killing Dozens,” New York Times, 28 September 2015, accessed 22 March 2019,

37 Interview with the author (Arabic), Markazi, 5 January 2017. In late 2018, refugees from Dhubab began returning to their village as the conflict had moved elsewhere.

38 See Chimni, B. S., “From Resettlement to Involuntary Repatriation: Towards a Critical History of Durable Solutions to Refugee Problems,” Refugee Survey Quarterly 23 (2004): 5573; and Long, Katy, “When Refugees Stopped Being Migrants: Movement, Labor and Humanitarian Protection,” Migration Studies 1 (2013): 426.

39 These 29,000 refugees and asylum seekers comprise nearly 3 percent of Djibouti's population (estimated just under 1,000,000) at a time that its unemployment hovers around 40 percent; “Djibouti Economic Outlook - Spring 2018” (English) (MENA Economic Outlook Brief, Washington, D.C., 2018), accessed 22 March 2019,; UNHCR, “Djibouti Fact Sheet, January 2019,” accessed 22 March 2019,

40 These actions followed the 2016 New York Declaration for Refugees and Migrants, which laid out a vision for a Comprehensive Refugee Response Framework (CRRF) to be applied in a dozen countries—Djibouti, Ethiopia, Kenya, Uganda, Somalia (and neighboring counties), Zambia, and six countries in Central America—and from Djibouti's embrace of the Nairobi Declaration on Durable Solutions for Somali Refugees in March 2017. See “CRRF Global Digital Portal: Djibouti,” n.d., accessed 1 June 2018, For an overview, see Hansen, Randall, “The Comprehensive Refugee Response Framework: A Commentary,” Journal of Refugee Studies 31 (2018): 131–51.

41 Republic of Djibouti, Décret fixant les modalités d'exercice des droits fondamentaux des réfugiés et demandeurs d'asile en République de Djibouti, 2017-410/PR/MI, 7 December 2017, accessed 1 June 2018,

42 Catherine Wachiaya, “UN Refugee Chief Praises Djibouti New Refugee Laws,” Relief Web, 15 December 2017, accessed 1 June 2018,

43 The total cost of the project (for 300 housing units, a 300-sq. meter school, and a 180-sq. meter mosque) was 2,500,000 USD; Nuryap Inşaat, “Nuryap and Nevka Companies Completed the Obock Camp Project in Djibuti,” n.d., accessed 1 June 2018,

44 Because of the dramatic turnover in the camp, an argument based solely on numbers would be misleading. Nevertheless, of the one hundred households I interviewed—fifty-three of which had come from the Taiz governorate (which includes Dhubab and Wahija), thirty from Aden, twenty-eight from Sanaa, six from al-Hudaydah, and one each from Ibb, Shabwa, and Lahij—at least twenty-nine households self-identified as muwallad and four households had been refugees in Yemen prior to the war (three from Eritrea, one from Somalia).

45 De Regt, “From Yemen to Eritrea and Back,” 30. Reputed to be the black descendants of Christian Ethiopian soldiers who invaded Yemen in the 6th century, the akhdām (who refer to themselves as al-muhammashīn, “the marginalized”) continue to be treated as social outcasts confined to jobs such as garbage collection and street cleaning. Comprising 5 to 10 percent of Yemen's population, the akhdām/muhammashīn are suffering additional vulnerabilities and discrimination due to the war; Rania el-Rajji, “‘Even War Discriminates’: Yemen's Minorities, Exiled at Home” (brief for the Minority Rights Group International, 2016), accessed 22 March 2019,, 12.

46 Gabriele vom Bruck, Islam, Memory, and Morality in Yemen (New York: Palgrave MacMillan, 2005), 8–9.

47 De Regt, “Close Ties”; Pandya, Sophia, “Yemenis and Muwalladīn in Addis Ababa: Blood Purity and the Opportunities of Hybridity,” Journal of Arabian Studies: Arabia, the Gulf and the Red Sea 4 (2014): 96114; Walker, Iain, “Hadramis, Shimalis and Muwalladin: Negotiating Cosmopolitan Identities between the Swahili Coast and Southern Yemen,” Journal of Eastern African Studies 2 (2008): 4459.

48 Ho discusses the alienation of muwalladīn of Hadhrami-Asian descent who felt out of place in their homeland but did not seem to have experienced as much overt racism, perhaps also due to their elite (sāda) social status. See Ho, Enseng, “Hadhramis Abroad in the Hadhramaut: The Muwalladīn,” in Hadrami Traders, Scholars and Statesmen in the Indian Ocean, 1750s–1960, ed. Freitag, Ulrike and Clarence-Smith, William G. (Leiden: Brill, 1997), 131–46; and Ho, Graves of Tarim.

49 Al-Jumly, Mohammed and Rollins, J. Barton, “Emigration and the Rise of the Novel in Yemen,” World Literature Today 71 (1997): 41.

50 Abdul-Wali, Mohammad, They Die Strangers: A Novella and Stories from Yemen, trans. Bagader, Abubaker and Akers, Deborah (Austin, Tex.: University of Texas Press, 2001 [1997]), 65.

51 Prior to 2003, only children born to a Yemeni father acquired Yemeni citizenship at birth (except for foundlings or children born to a Yemeni mother and a legally unknown father). Dual nationality and naturalization were possible only in exceptional cases. A 2003 amendment extended citizenship to children of Yemeni mothers who had become divorced from their foreign spouse. A 2010 amendment finally permitted full maternal jus sanguinis as well as dual citizenship. In Ethiopia and Somalia, citizenship was also generally conferred by male lineage, until recently; in Somaliland and Djibouti, citizenship is conferred by the father only, unless the child's father is unknown. Except in certain cases, Eritrea and Ethiopia prohibit dual citizenship. Somaliland's 2002 Citizenship Law and Djibouti's 2004 Nationality Law permit dual citizenship; as of 2012, Somalia's Provisional Constitution does, too. This means that, even now, children born of mixed marriages between Yemeni nationals and nationals of Horn of Africa countries will assume the single nationality of their father—with the exception of children born to Yemeni fathers and Somali mothers or children born to Yemeni mothers and Somali, Somalilander, or Djiboutian fathers who could, theoretically, acquire dual citizenship. See the Law on Yemeni Nationality (No. 6 of 1990), 26 August 1990,; Proclamation on Ethiopian Nationality (No. 378 of 2003), 23 December 2003,; Eritrean Nationality Proclamation (No. 21/1992), 6 April 1992,; The Provisional Constitution of the Federal Republic of Somalia, 1 August 2012,; Republic of Somaliland Citizenship Law (No. 22/2002), 3 June 2002,; Code de la Nationalité Djiboutienne (No. 79/AN/04/5ème L), 24 October 2004,, all accessed 22 March 2019. See also Laura Van Waas, “A Comparative Analysis of Nationality Laws in the MENA region,” 9 September 2014, accessed 1 June 2018,

52 See, e.g., Masquelier, Adeline, “Why Katrina's Victims Aren't Refugees: Musings on a ‘Dirty’ Word,” American Anthropologist 108 (2006): 735–43.

53 Interview with the author (Arabic), Markazi, 14 December 2016.

54 On this day, President Salih's supporters opened fire into a crowd of unarmed, antigovernment protestors gathered for prayer in Change Square, killing forty-five individuals and wounding hundreds, prompting many government officials to resign; Human Rights Watch, “Unpunished Massacre: Yemen's Failed Response to the ‘Friday of Dignity’ Killings,” 2013, accessed 25 March 2019,

55 This is a line from al-Baraduni's 1973 poem, “al-Ghazu min al-Dakhil” (The Attack from Within). See ʿAbd Allah al-Baraduni, Diwan ʿAbd Allah al-Baradduni, vol. 1 (Sanaa: al-Hayaʾ al-ʿAmma li-l-Kitab, 2002), 680–84. A more literal translation is “Yemenis are in exile, and others are exiled in Yemen”; see Al Areqi, Rashad Mohammad Moqbel, “Ideology of Exile and Problematic of Globalization in Al Baraduni's Poetry,” International Journal of Applied Linguistics and English Literature 5 (2016): 22. This translation, which comes closer to Yusuf's interpretation, is taken from Abdulsalam al-Rubaidi, “Imagining an Alternative Homeland: Humanism in Contemporary Yemeni Novels as a Vision for Social and Political Reform,” Center for Applied Research in Partnership with the Orient (CARPO) Study 6 (2018): 11, accessed 22 March 2018,

56 Interview with the author (English), Markazi, 28 December 2017.

57 Interview with the author (English), Markazi, 13 March 2018.

58 Interview with the author (English), Markazi, 28 December 2017.

59 Interview with the author (Arabic), Markazi, 30 December 2017.

60 For an overview of why Eritreans have been fleeing and of the camps that house them in Ethiopia and Djibouti (among other countries), see Connell, Dan, “Refugees, Migration, and Gated Nations: The Eritrean Experience,” African Studies Association 59 (2016): 217–25.

61 Theoretically, these individuals could be considered “flexible refugees” in the same way that Ong describes holders of multiple passports as “flexible citizens”: entrepreneurial and cosmopolitan migrants for whom refugee status is a necessary surrogate for weak citizenship from failing states; Ong, Aihwa, Flexible Citizenship: The Cultural Logics of Transnationality (Durham, N.C.: Duke University Press, 1999).

62 This need not be conspiratorial. For a review of the sedentarist (and paternalistic) assumptions guiding development interventions in Africa, see Oliver Bakewell, “‘Keeping Them in Their Place’: The Ambivalent Relationship between Development and Migration in Africa,” Third World Quarterly 29 (2008): 1341–58.

63 Hansen, Randall, “The Comprehensive Refugee Response Framework: A Commentary,” Journal of Refugee Studies 31 (2018): 141.

64 United Nations General Assembly, New York Declaration for Refugees and Migrants: resolution/adopted by the General Assembly A/RES/71/1, 3 October 2016, paragraph 68, accessed 1 June 2018,

65 Nathalie Riise, “UNGA Votes to Adopt Global Compact on Refugees,” 18 December 2018, accessed 23 March 2019, For the text, see United Nations General Assembly, Report of the United Nations High Commissioner for Refugees: Part II Global Compact on Refugees A/73/12 (Part II), 13 September 2018, accessed 22 March 2019,