UNHCR and the Algerian war of independence: postcolonial sovereignty and the globalization of the international refugee regime, 1954–63

Abstract

The Algerian war of independence (1954-62) was crucial to the extension of the modern international refugee regime beyond Europe. It is also the exemplar of how that regime became a site for the establishment of postcolonial sovereignty, globally. Tunisia and Morocco, newly independent, requested UNHCR’s help in assisting hundreds of thousands of Algerian refugees: interacting with the refugee regime allowed them to establish their credentials as independent states while asserting sovereignty over their own territories. In Algeria, the 1951 Refugee Convention applied before the war started, and UNHCR worked there to support ‘old’ refugees. During the war, the Front de Libération Nationale asserted itself as a state-in-waiting by engaging with UNHCR outside Algeria as the agency coordinated a vast relief operation. After the war, as refugees returned to a landscape riven by mass displacement, interacting with the refugee regime helped the new state assert sovereignty over Algeria’s territory, and Algerian bodies.

Introduction

The Algerian war of independence (1954–62) was crucial to the extension of the modern international refugee regime beyond Europe. ¹ The 1951 UN Convention Relating to the Status of Refugees had initially been intended to settle Europeans still out of place after the Second World War. The mandate of the Office of the United Nations High Commissioner for Refugees (UNHCR), the agency tasked with helping states to implement it, was set to run for only eight years. But two new population displacements in the 1950s brought about the expansion of its remit in both time and space. Within Europe, the arrival of tens of thousands of Hungarians fleeing Soviet repression in late autumn 1956 led Austria—newly independent after Nazi Anschluss and Allied occupation—to request UNHCR’s assistance: first in providing relief, then in coordinating a mass evacuation. And in north Africa, shortly afterwards, the arrival of tens of thousands of Algerians fleeing French repression led Morocco and Tunisia—newly independent from French colonial rule—to request its help too. They were among only five states outside Europe to adhere to the convention before 1960. ²

But the story of UNHCR’s involvement in the Algerian war is more complex than this, and less schematic. The Algerian war did not bring the agency to north Africa for the first time: the 1951 Convention already applied in the Maghrib before the war began, and UNHCR already operated there, because France had signed it on behalf of ‘all territories for the international relations of which France is responsible’. ³ That included Algeria, legally part of the French Republic, as well as Morocco and Tunisia, as protectorates. ⁴ The definition of ‘refugee’ in the convention’s text was not restricted to members of a specific group or groups, as earlier refugee law had been, but it was limited temporally to people displaced before 1 January 1951. It also offered contracting states two alternative versions: one limiting its application geographically to refugees displaced as a result of ‘events occurring in Europe’ and the other applying to those displaced by ‘events occurring in Europe or elsewhere’. ⁵ In 1967, a protocol was added to it removing both limitations, allowing the institutions of the refugee regime to respond to the ‘new refugees’ of the decolonizing 1960s. But foregrounding the perspectives and agency of actors in the newly decolonized world decentres the UN institutions and the major powers in our understanding of how this happened. Throughout the later 1950s and 1960s, former French colonies in north and then west Africa took advantage of their succession to the convention to push the refugee regime towards globalization: they saw possibilities in it far beyond what the old colonial power had intended, and beyond what legal scholars and historians have recognized. ⁶

Simply to state that Morocco and Tunisia invited UNHCR onto their territories as newly independent states, and that the agency accepted the invitation under its ‘good offices’ doctrine, gives us a thin understanding of what happened. The sovereignty of newly independent states—their ability to assert the primary claim to rule over a particular land and a particular group of people, and to have that claim recognized on that land, among those people, and beyond—was not made overnight at independence: it quickened and thickened over a much longer period. ⁷ Globally, throughout the twentieth century, this process was closely connected with population displacement. In eastern Europe, the emerging nation-states that replaced dynastic empires after the First World War staked their territorial claims, and defined their ‘national’ populations, partly through their efforts to manage the return and settlement of people displaced during the war. ⁸ The territories, institutions, and national identities of post-Ottoman nation-states in the Arab Middle East were formed around and against refugees, and in dialogue with an international regime of refugee protection, between the wars, when they were ruled—as nominally independent states—by Britain and France under League of Nations mandates. ⁹ Similarly, what sovereignty meant in south Asia, in terms of borders, citizenship, and property regimes, was articulated around the populations displaced at partition in 1947. ¹⁰ That displacement preceded the convention, and the UN refugee agencies played little role in responding to it. ¹¹ But the Algerian war showed that something had changed after 1951: the UN refugee regime itself had become a site for the articulation of sovereignty. When Tunisian prime minister Habib Bourguiba requested help from the High Commissioner for Refugees, on 31 May 1957, he was taking the opportunity to interpellate the UN institutions as leader of a newly but fully fledged member state. Morocco soon followed. As UNHCR coordinated a large relief operation for refugees from Algeria with its ‘implementing partner’ on the ground, the League of Red Cross Societies, the two states asserted their territorial sovereignty by deciding which foreign and international agencies would be allowed to participate in it. And interacting with the refugee regime also helped to define the meaning of sovereignty within those territories—notably in the question of citizenship and nationality, as decisions were made over who would be defined as a ‘refugee’ and therefore eligible for international assistance.

In the historiography of Algerian independence, meanwhile, it is now well established that the National Liberation Front (Front de Libération nationale, FLN) successfully used international forums to assert itself as the internationally recognized representative of the Algerian people. ¹² During the war, the FLN developed national institutions of health care and humanitarian assistance, in a struggle with the colonial state for sovereignty over Algerian bodies that was also a struggle for international legitimacy. ¹³ The main objects of this contest were the Algerian Muslims displaced during the war in their millions—a displacement whose astonishing scale and profound impact historians increasingly recognize. ¹⁴ Most were displaced within Algeria’s borders, but, as we will see, the refugees beyond them also figured in this history. The FLN took responsibility for them, directly or through affiliated agencies such as the Algerian Red Crescent, and interacted with host states and the refugee regime on their behalf. At the end of the war, the repatriation of hundreds of thousands of refugees across border zones militarized by the French army demonstrated how hard it would be for the new state to establish full territorial sovereignty. And the relief operation to assist the refugees, following them back into Algeria as the transition to independence began, discovered that many of the ruined country’s residents were in even greater humanitarian need. The shared experience of displacement was a powerful driver of national feeling among Algerian Muslims, but created a humanitarian emergency that would test the new state’s sovereignty over Algerian bodies to the limit. Finally, independence raised the question of UNHCR’s relationship with the new state of Algeria—not least because the agency remained responsible for several thousand ‘old’ refugees still living there. Despite the immense problems preoccupying the new government, it swiftly followed its neighbours in succeeding to the convention. Why?

This article makes two arguments at the intersection of refugee history and global history. First, that if the Algerian war globalized the international refugee regime, it did so because of actors outside Europe—a shift in the regime’s centre of gravity that would gain further momentum in the 1960s. Second, that this made the regime into a key site for the articulation of sovereignty in the decolonizing global south. The article draws on sources from the archives of UNHCR, the International Federation of Red Cross and Red Crescent Societies, and the International Committee of the Red Cross, supplemented where possible by narrative sources. It starts with UNHCR’s operations in Algeria before and during the war. It then looks at the experiences of the 300,000 or so refugees from Algeria who fled to Morocco and Tunisia during the war, and how these states interacted with the international refugee regime. Next it explores the much larger landscape of displacement that refugees returned to at the end of the war. It ends by briefly discussing independent Algeria’s interactions with the refugee regime immediately after 1962. Throughout, we will also stress global comparisons for the Algerian case.

UNHCR in Algeria

The Algerian war of independence began on 1 November 1954, with a series of bombings by the FLN, then a small nationalist organization. ¹⁵ Its aims, presented in the ‘November First Declaration’, were to end French colonization and achieve Algerian independence. This was unthinkable for France. Algeria was its only major colony of settlement, with around a million European settlers (originally from France, Spain, Italy, and elsewhere) among some nine million Algerian Muslims. Since the 1840s, it had been legally a part of France itself, though one where discriminatory legal codes and rigged electoral colleges kept the colonized population subjugated. The FLN’s terrorist actions provoked a disproportionate French response against the whole colonized population, as they were intended to. This intensified rather than dispersing support for the FLN and its armed wing, the National Liberation Army (Armée de Libération nationale, ALN). French ‘victories’ like the brutal eradication of the FLN’s organization in the capital Algiers only pushed the colonized population further into opposition. By the late 1950s, the war, correctly understood by Algerian Muslims as a revolutionary struggle, had mutated from an urban counterinsurgency campaign into a countrywide conflict marked by aerial bombardment and mass population displacement. France held the military advantage, but its geopolitical situation was weakened, and the FLN’s strengthened, by the diplomatic disaster of the ‘tripartite aggression’ against Egypt in November 1956—France’s participation was intended to destroy the Free Officers’ regime as a backer of the FLN. Meanwhile, the FLN gradually absorbed or eliminated the other nationalist organizations and asserted itself internationally as a state in waiting. In 1958, it established a provisional government of the Algerian Republic (Gouvernement provisoire de la République algérienne, GPRA), headed by Ferhat Abbas and based in Cairo then Tunis, which sent a delegation to the UN General Assembly in New York. And the French public gradually lost faith in a manifestly unwinnable war, fought by an army of mass conscription, that was triggering increasing political violence and instability in the metropole itself. A negotiated end to French rule in Algeria became inevitable. The Evian Accords of March 1962 ended the war, and formal independence followed that July. ¹⁶

By the time the war began, the 1951 UN Convention Relating to the Status of Refugees already applied there. But only just. The convention entered into force on 22 April 1954. ¹⁷ France, having signed the convention in 1952, ratified it on 23 June 1954, bringing the convention into force for France ninety days later, on 21 September. ¹⁸ And not only for France but for ‘All territories for the international relations of which France is responsible’, which at that date still included Algeria, Morocco, and Tunisia. Across all three territories, and many more, France was therefore committed to grant anyone who met the convention’s definition of refugee the rights and protections that it enshrined, and the Office of the UNHCR was mandated to assist it.

In the 1950s, UNHCR was primarily a legal agency. It helped states implement the convention in legislation and practice, and it helped refugees access the assistance they were entitled to under the convention. In Algeria, these were mostly Europeans who had been refugees for years, including Spanish Republicans exiled since 1939 and even some Russian refugees from the 1919–21 civil war. ¹⁹ UNHCR documents referred to these groups as ‘old refugees’ or ‘Convention refugees’. Many of them now had children who were French, having been born on French (Algerian) soil. UNHCR had responsibility for these ‘old refugees’ throughout the Algerian war, working with different partners. The French Red Cross provided humanitarian assistance, while the relevant French state agencies were the French Office for the Protection of Refugees and Stateless Persons (Office français de protection des réfugiés et apatrides, OFPRA) and the Social Service for Emigrant Assistance (Service social d’aide aux émigrants, SSAE). ²⁰

How many of these refugees were there? It was hard for UNHCR to know: most had been in Algeria since long before the agency was founded, and not all approached it for assistance. During the war, and in its chaotic aftermath, confirming their numbers was even harder. In December 1962, an employee of the SSAE who was still in Algeria, Mme Palix, informed the agency that in 1960 the authorities had estimated the total number of foreigners in Algeria, including refugees, at 30,000. But now she reckoned only 2,000–5,000 refugees remained, mostly Spanish refugees living in Oran. ²¹ This sharp decline, and the location of the remainder, highlights a key point about the ‘old’ refugees: to the extent that they were integrated within French Algeria, it was as a part of settler society. (Oran, though smaller than Algiers overall, had the colony’s largest ‘European’ population.) Out of a settler population of roughly one million, some 650,000 left Algeria in 1962, just before or soon after independence. Most of the refugees seemed to have simply left with them.

UNHCR’s operations in Algeria during the war, then, were an extension of its European mandate: working with European refugees who were living in a European settler society. As the colonial state dismantled itself in 1962, the OFPRA and the SSAE went with it. The French Red Cross also pulled out, replaced by the Algerian Red Crescent. France viewed the remaining ‘old’ refugees as, now, independent Algeria’s responsibility. But UNHCR recognized that its own role in assisting them must continue within the new state.

Algerian refugees and postcolonial sovereignty in Morocco and Tunisia

A second group that relied on support from UNHCR was composed of Algerian refugees who fled into Morocco and Tunisia during the war. After Algeria’s neighbours became independent in 1956, Algerians quickly started to seek refuge. Their numbers increased sharply after the French army began creating ‘forbidden zones’ (zones interdites) in 1958—some Algerian Muslims fled across the border to avoid being sent to regroupment camps (see below). Others were displaced from their homes by the militarization of the borders themselves, as the French army turned the borderlands into a terrain of minefields and electric fences up to 100 kilometres deep to prevent Algerian combatants from entering the country from bases in Tunisia and Morocco. ²² Like other nationalist fronts unable to exercise sovereignty in the occupied homeland, such as the Palestine Liberation Organization from the 1960s and the Sahrawi Polisario Front from the 1970s, the FLN tried to make itself the refugees’ representative government-in-exile in their interactions with the host states and international organizations. ²³ For Morocco and Tunisia, the arrival of the refugees triggered direct interactions with the refugee regime. They were the first postcolonial states to see that the regime offered them opportunities to assert their new-found sovereignty in the international sphere, and to learn that doing so would raise questions about the nature and limits of their sovereignty on the ground. They would not be the last.

Initial support for the Algerian refugees was provided by their ‘own’ authorities, the FLN and, after 1958, the GPRA. A 1957 report for the International Committee of the Red Cross (ICRC) by J-P. Schoenholzer described the assistance the FLN offered refugees in Morocco, where it had several bases. ²⁴ It provided consular support, a family allowance, legal protection, and, more importantly, help finding housing. The Amicale des Algériens musulmans (AAM), an FLN-affiliated organization tacitly recognized by the Moroccan authorities, gathered arriving refugees and found them places to settle, obtaining authorization to occupy empty or abandoned buildings or install tents or packed-earth huts on land near water sources. Some settled in caves. Refugees were also settled in camps near the borders, sometimes within earshot of the war. ²⁵ The AAM, Schoenholzer claimed, registered the refugees; recorded births, weddings, and deaths; and issued documents.

This is a striking instance of the FLN establishing itself as a ‘national counter-state’ for the colonized population—a strategy that was pursued outside Algeria’s borders as well as inside them. ²⁶ Outside Algeria, France’s ability to hinder the strategy through violent repression was limited. It could not prevent the FLN and its affiliates from providing social support to Algerian Muslim refugees, a ‘state-like’ activity that evidently produced statistical reports on them as a population too (Schoenholzer cited AAM figures of 47,500 Algerian refugees in Morocco, 20% men, 20% women, and 60% children under 12). It could only discourage other states from giving the GPRA diplomatic recognition and try to dissuade international agencies from working with the FLN. UNHCR was one such: Schoenholzer was in Morocco to observe the situation before the agency launched a vast ‘joint relief operation’ for Algerian refugees there and in Tunisia in 1959.

The relief operation was led by UNHCR and the League of Red Cross Societies. The legal basis for it was the Tunisian government’s request for help to August Lindt, the High Commissioner, on 31 May 1957. ²⁷ One condition of the League’s involvement was that there must exist national Red Crescent societies in Morocco and Tunisia: Tunisia already had one, established in autumn 1956, and the creation of a national Moroccan society in 1957 was also accelerated by the need to care for Algerian refugees. ²⁸ On the ground, the operation also collaborated with the FLN-affiliated Algerian Red Crescent, though the League did not formally recognize it until 1963. ²⁹ The joint relief operation benefited at its launch from the publicity around World Refugee Year (1959–60) and lasted beyond the end of the war. It was well timed: the number of refugees was spiking as the war intensified.

The question of numbers is revealing. ³⁰ It hints at the continuing exchanges between, on the one hand, UNHCR and its partners, and on the other, the GPRA and Algerian structures on the ground. Although they seem to be under-documented in the archives, such exchanges may have contributed to the confusion in numbers, with Algerians providing figures based on the number of individuals they registered while UNHCR and the League counted rations distributed. Better documented are the almost inevitable disagreements between host states and international agencies over refugee numbers: figures used in the joint relief operation in Tunisia, for example, were consistently lower than those given by the Tunisian authorities. ³¹ Host states often give higher estimates than humanitarian agencies for the size of their refugee populations. This may be for ‘bad’, or self-interested, reasons: to embarrass the government of the country of origin and perhaps channel support to its exiled opponents; to maximize incoming humanitarian assistance, as a source of foreign exchange and opportunities for employment (and embezzlement); and to polish the host state’s own humanitarian reputation. And it may be for ‘good’, or altruistic, reasons: because they recognize that registering refugees is difficult, so humanitarian agencies often undercount the population in need; because they know that aid often falls short, and more refugees may arrive before it comes; and to offset the losses incurred by their own initial absorption of the costs of assistance and the longer-term impact on their own populations (as, e.g., food prices rise). ³² Several of these factors were likely at play in Morocco and Tunisia. Certainly, both states were keen to see Algeria reach independence, without wanting to confront France directly themselves, and the international agencies feared that some beneficiaries of their assistance might be engaged in military action against France. ³³ The joint relief operation was also marked by the equally common lack of consensus over figures among international agencies. ³⁴

To manage these issues, in June 1960 UNHCR and the League drafted ‘Joint instructions on criteria for material assistance to Refugees from Algeria in Morocco and Tunisia and guiding principles on number of registrants’. They stressed that while the governments’ official figures should be used in planning the relief operation, these included many non-refugees: ‘it is therefore of the utmost importance that every effort be made not to increase this figure but rather to endeavour to reduce it and to apply the criteria for material assistance to all new registrations rigidly’. ³⁵ Assistance should only go to ‘Persons who had habitual residence in Algeria and who fled to Morocco and Tunisia from there since 1956 as a consequence of events in Algeria, and who are in need [emphasis in original].’ Numbering refugees, that is to say, also meant defining them. But this was less simple than it appeared. The instructions immediately excluded from any material assistance ‘(1) Nomadic tribes undisturbed by events in Algeria and who merely wish to take advantage of the possibility of free food; (2) Persons from Algeria who had habitual residence in Morocco or Tunisia (carte frontalière).’

Determining who qualified for international assistance highlighted the ‘nationalization’ of the previously fluid and intermingled populations of French north Africa. As the porous internal frontiers of the colonial period gave way to harder national borders, nomadic tribes needed to be allocated to one state or another, regardless of their customary migrations. Similarly, the relative freedom of movement between the three countries in the colonial period meant that there were many ‘Persons from Algeria who had habitual residence in Morocco or Tunisia’. There were also many Moroccans and Tunisians who had been resident in Algeria: as late as January 1962, it emerged that as many as 65,000 of the then total of 153,000 ‘Algerian’ refugees in Morocco were ‘in fact Moroccan subjects [ressortissants marocains], perhaps refugees from Algeria but incapable of proving it’. ³⁶ Such people might not only have ‘habitually resided’ in Algeria but lived their whole lives there. ³⁷ Independence for Morocco and Tunisia, and the war to achieve it in Algeria, sharply posed the question: how should state sovereignty extend over groups and individuals among the still partly colonized people of north Africa? The decisions taken between the host states and UNHCR over who qualified for international assistance as a refugee helped to answer this question.

International assistance took many forms. In February 1960, the High Commissioner reported on his agency’s recent work to the UN General Assembly (noting that the actual number of refugees was considerably higher than the 200,000 that had been budgeted for). ³⁸ The operation had established thirty-seven aid distribution points in Tunisia and twenty-three in Morocco, in close cooperation with both governments. Its primary aim was to distribute food, ‘since it is an obvious fact that before all else the refugees must be fed’: the report recognized that in the operation’s early months, short on funds and supplies, distributions ‘never reached a satisfactory calorie level’. The aim was to ensure rations of 1,540 calories per day, with children, the majority of the refugees, receiving an additional 135 calories per day in condensed milk. At this stage, there were no formal camps—some were created later—and the refugees were living in gourbis (shelters) or tents of their own construction, so next came provision of blankets and clothing, contributed by national Red Cross and Red Crescent societies: some 230,000 blankets had been issued. A small number of tents had been provided, and some shelters had been improved before the winter, but this was mainly by the refugees’ own initiative with assistance from the local authorities. Seventeen milk stations had been set up to provide a daily ration of reconstituted powdered milk to children and pregnant or nursing women, and the agency aimed to increase this number to forty-fifty. In Tunisia, multi-purpose centres were planned, incorporating milk stations, facilities for providing hot meals, and dispensaries. A medical survey conducted in August 1959 had indicated that there was no serious health emergency among the refugees, though it recommended improved nutrition. Both countries had opened their health services to the refugees, but the agency thought that additional services might prove necessary, and was investigating ways to improve the situation—though this would be costly. There were also pilot projects to provide support for refugee livelihoods. Almost inevitably, given that most of the working-age adults among the refugees were women, these consisted of sewing rooms to train women and in Morocco an experimental mat-weaving project. ³⁹

Funding, too, came from many sources. Between 1 February 1959 and 31 January 1960, UNHCR had received over a million dollars from governments and over $600,000 from other sources for the relief operation. The US was easily the largest state donor in this period, at $610,000 (part-pledged), but the second largest was France, which had contributed just over $250,000. ⁴⁰ Of the other sources, the most significant was Vrijzinning Protestante Radio Omroep (VPRO), a Dutch broadcaster which had made the joint relief operation the focus of its charitable appeal for World Refugee Year and contributed nearly $500,000. ⁴¹ And the relatively well-funded relief operation also raised questions for the host states. What were their responsibilities towards their own populations? ⁴² Though UNHCR and the League may have been right that ‘false refugees’ were not entitled to relief because they were not technically refugees, that did not mean that they were not in need. The international assistance that refugees received highlighted the humanitarian needs of the host population: a common phenomenon in refugee history, that often fuels host-society resentment. The Moroccan government appeared to have growing concerns about its own population, whose difficulties the joint relief operation made more apparent. UNHCR therefore found itself discussing the possibility of charitable organizations, such as the American Friends Service Committee, providing relief to destitute Moroccan nationals. ⁴³ But did allowing foreign agencies to operate mean relinquishing sovereignty? One such was CARE, an American charity that had originated in 1945 as a provider of food aid to Europe. ⁴⁴ In January 1961, its head Willard Johnson lamented that neither Tunisia nor Morocco would allow his organization to operate freely: ‘both nations will not deal directly with an American organization, only with international agencies’. ⁴⁵ Tunisia had an ‘umbrella’ agreement with UNHCR allowing foreign organizations to work under UNHCR authority. ⁴⁶ Morocco did not share Tunisia’s blanket suspicion of foreign national agencies, but mistrusted organizations from specific countries. UNHCR had only ‘a moral role in advising the government on what programmes to encourage’. ⁴⁷ As the months went by, and a cohort of different national and international organizations came to work under UNHCR’s coordination, this was another way in which the refugee regime provided a site for Morocco and Tunisia to articulate their postcolonial sovereignty.

At the end of the war, the joint relief operation rapidly gave way to repatriation (Figure 1). ⁴⁸ The outlines of the operation were set by the Evian Accords between France and the FLN, which requested UNHCR involvement—the first time that a refugee repatriation had been referred to the agency directly by the countries involved. ⁴⁹ UNHCR and the League, now headed by Félix Schnyder and Henrik Beer, agreed after some discussion to organize the repatriation. ⁵⁰ Tripartite commissions would be created in Algeria, with representatives of the French government, the Algerian Provisional Executive, and UNHCR and the League. ⁵¹ At a joint press conference, Beer stressed the repatriation’s ‘truly universal character’: over fifty national Red Cross and Red Crescent societies were involved. It was, Schnyder added, ‘a humanitarian action benefitting from the widest support that has ever been seen [une action humanitaire bénéficiant de l’appui le plus large qu’on ait jamais vu]’. ⁵²

Figure 1. ‘Ces réfugiés vont retrouver la vie’ – League of Red Cross Societies brochure, 1962.

The repatriation was launched in May 1962. It was an immense undertaking: in January 1962, the League of Red Cross Societies had estimated that it was distributing rations to over 300,000 people, 171,790 in Tunisia and 135,113 in Morocco. ⁵³ A UNHCR communiqué called the operation ‘the most important undertaken with the assistance of an international organization since Nansen’s time, with the exception of the return to their countries of persons displaced by the Second World War’. ⁵⁴ This awkward phrasing aimed to convey both its magnitude—considerably larger than the recent evacuation of Hungarian refugees from Austria—and its significance in the history of international organizations. Other participating agencies included the International Committee of the Red Cross and Rädda Barnen, the Swedish branch of Save the Children. ⁵⁵ On the Algerian side, the Algerian Red Crescent, the National Liberation Army (ALN), and the provisional government were all involved. Transit camps were set up in Morocco and Tunisia to shelter refugees overnight before their departure for Algeria by train or truck, or on foot. The refugees tended to be in better health than Algerians who had remained in the country throughout the war—some later felt guilty for having experienced less hardship—but the French required them to be vaccinated and medically examined, ostensibly to prevent diseases from entering a territory still under French sovereignty. ⁵⁶ Itineraries were drawn up and carefully verified by the ALN: the refugees were not just crossing a borderline but a wide and heavily mined border zone (Figure 2). Claudine Chaulet, a French-born GPRA agent who participated in the repatriation, travelled back with nomad women who, familiar with the area, began making their own way. Then a French helicopter dropped a note explaining that they were in a minefield. They didn’t move until ALN trucks came to fetch them. But in the end, she later recalled, ‘there wasn’t a single mine victim during the transfer of roughly 200,000 people’. ⁵⁷

Figure 2. Map of repatriation operation.

This rapid repatriation left the host countries, Tunisia and Morocco, suddenly under-supported: they argued that they still required assistance with any refugees that remained, but the relief operation swiftly wound down. And it left the returning refugees under-supported too. The humanitarians coordinated by UNHCR in Morocco and Tunisia provided the departing refugees with tents and rations—medical staff stressed the importance of feeding them up before they left—because there was great uncertainty about the situation they were returning to. ⁵⁸

Refugee return and landscapes of displacement

Algeria is part of a global twentieth century history of nation-states forged by mass displacement, and the returning refugees were a relatively small part of it. The ‘regroupment’ of much of the colonized population within Algeria had fostered the very national feeling it was intended to repress, but created a humanitarian crisis that the new government would struggle to address. The unanticipated departure of much of the settler population helped create the conditions for a national rather than a colonial economy. But before that happened, the repatriation and resettlement of the refugees showed how hard it would be for the new state to establish sovereignty over its territory.

Take the issue of minefields. These were not just an obstacle to the repatriation operation: they hindered refugee resettlement and lastingly compromised the new government’s ability to exercise territorial sovereignty. By the end of the war, the French army had laid between eleven and twelve million antipersonnel mines in Algeria, mostly near the border. ⁵⁹ Withdrawing over the course of 1962, it left behind bases, camps, and barracks, some of which were also mined. Most of the refugees were from these border regions, a rural peasant population whose lands were now unusable for farming. Demining the land was an urgent necessity, but in the borderlands French soldiers refused to carry out last-minute demining operations: as one fictionalized memoir put it, ‘It would be stupid to get yourself crippled when the war is over.’ ⁶⁰

What this meant for one group of returning refugees was described by the UNHCR delegate for Algeria, John D. Kelly, on 1 August 1962. ⁶¹ Visiting Tlemcen in the west of the country at the end of July, he had been informed of the ‘desperate situation’ of 5,000 members of the Beni Boussaid tribal group. Their repatriation had been delayed because their lands had been ‘saturated’ with minefields: ‘It appears that the maps and plans of the mines have been lost, which enormously increases the difficulty of disposing of the mines.’ And when they did return to their tribal lands on the Moroccan border, on 20–25 July, they found themselves ‘completely hemmed in by the minefields’. Unable to till their lands, or even fetch water, they were entirely reliant on external assistance. The Tlemcen prefecture asked UNHCR to put pressure on the French to demine the area as the ALN was not equipped to do so, and the Beni Boussaid could neither be kept as ‘permanent charges on charity’ nor transferred elsewhere in Algeria. But Kelly’s response to this request was revealing. He returned to Rocher Noir near Algiers, where both the Transitional Executive administration and the UNHCR offices were located. There, he liaised with M. Benzerfa of the executive on the subject, ‘since it is one entirely within the competence of the Algerian authorities, and UNHCR could act, if at all, only if so requested by them [emphasis added]’. The minefields were a practical constraint on the new state’s territorial sovereignty. But paradoxically, its diplomatic sovereignty meant that UNHCR could only intercede with France if the Algerian government asked it to do so.

Demining was a formidable challenge for the new state, requiring money, material, and technicians that it lacked. ⁶² Former ALN combatants with experience in crossing the minefields during the war began the task, and from 1963 Soviet demining teams worked with them. ⁶³ But one of the main obstructions to demining remained the absence of maps of the mined areas—not lost, as Kelly thought, but kept by the French, and not handed over until 2007. It took another decade after that for demining to be completed. ⁶⁴ In this very basic sense, for over half a century after independence, French landmines remained a deadly hindrance to Algeria’s exercise of full territorial sovereignty. Part of the Beni Boussaid land was returned to civilian authorities only in 2013: what happened to the Beni Boussaid in the meantime is unknown. ⁶⁵ For them as for many other returnees, it was impossible to remain on their lands. But where else would they go?

For the returning refugees were not the only ones struggling to find a place to settle. In January 1963, the secretary-general of the League of Red Cross Societies described how their humanitarian needs had immediately been folded into the much greater needs of the population at large: ‘It rapidly became apparent that their fate could not be separated from that of the two million displaced and regrouped people who likewise converged on their douars of origin, nor from the mass of the needy population whose standard of living had dropped considerably following this prolonged conflict.’ ⁶⁶ During the war, the French army had created ‘forbidden zones’ (zones interdites) to cut the FLN off from the popular logistical support it needed: large swathes of territory whose inhabitants were resettled in so-called regroupment camps (camps de regroupement). These were concentration camps like those used by Spain and Britain during their respective imperial wars in Cuba and South Africa around the turn of the century: an instance of ‘barbed-wire imperialism’. ⁶⁷ By the end of the war, 3,525,000 people, or some 41% of the colonized population, had been forcibly displaced. Of those, 2,350,000, roughly a quarter of the colonized population, had been forced into regroupment camps; around 175,000 others spontaneously left their villages to regroup around them. ⁶⁸ An International Committee of the Red Cross mission visited twenty-one camps in the centre and east of Algeria in January and February 1962. A few were of permanent construction (bâti en dur) and situated in regions where cultivable land would be available to the residents after fighting ceased. But the ‘great majority’ were not. They were comprised of gourbis: ‘Built of bricks—obtained by mixing clay and straw—covered with thatch or twigs, they shelter a population living directly on a beaten earth floor, alongside the domestic animals.’ ⁶⁹ Their inhabitants lacked the most basic necessities. Hygiene and nutrition were poor; child mortality was high; tuberculosis, rickets, and trachoma were common. The sanitary situation of the ‘regrouped’ populations was worse than that of the refugees living in Morocco and Tunisia. ⁷⁰ And it ran the risk of suddenly deteriorating at independence, as the French authorities and the French Red Cross were leaving before the Algerian authorities could take over.

Independence left these encamped populations facing an uncertain future. Only about 15% of the regroupés had left the camps by the time of the ceasefire in March 1962. A month later, they still numbered 1.8 million (900,000 in the east, 600,000 in the region of Algiers, and 300,000 in the west), 30% of whom were children under ten. ⁷¹ It bears repeating that French military strategy had deliberately cut this large population off from the means of sustaining itself. As a result, 500,000 people were entirely dependent on outside support; another 300,000 were partially dependent. The risk of famine was a major concern well beyond independence in July 1962. Some regroupés returned to their lands of origin, some migrated elsewhere (many to the cities), and some stayed where they were. Some left the camps only to return, having discovered that their homes had been destroyed. ⁷² Like Europe in 1945, on a smaller but still vast scale, the entire country was full of displaced people in urgent need of assistance. ⁷³

Caring for these people was a matter not only of humanitarian emergency but also of sovereignty. For the FLN and for many Algerians, one central dimension of their revolution was regaining the ability to care for their own—that is, sovereignty over Algerian bodies. The Algerian transitional authorities and several relief agencies, following the refugees back ‘in-country’, divided the territory of Algeria north of the desert (Figure 2): the League operated along both borders and in most of the western third of the country, including Oran; Catholic organizations (Caritas and the National Catholic Welfare Conference) operated in the central third, including the capital; Protestant organizations (Cimade and the World Council of Churches) operated in the eastern third, including Constantine but minus the border zone. A Quaker relief organization, Quaker Service, operated in a small coastal area north of Constantine. But international humanitarian assistance, however much it was needed, called into question the new state’s sovereignty over Algerian bodies. When relief organizations divided up the territory, it even raised questions about territorial sovereignty. There was a tension, for the Algerian authorities, between proving their own capacity to feed, protect, and house their own people (Figure 3) and having to appeal for and accept international relief. ⁷⁴ This explains the efforts of the Algerian provisional government and its agencies not just to participate in but also to supervise and coordinate the repatriation and the humanitarian relief operation within Algeria.

Figure 3. al-Mujāhid, 29 Nov 1962. “For a better life for the children of the Martyrs. Yesterday began the house-building operation for refugees in the Eastern region. And so, thanks to the mobilisation of youth and army, the situation of misery and homelessness that hundreds of thousands of our brothers are suffering will soon come to an end.”

While the colonized population moved slowly out of mass displacement, the bulk of the colonial population (10% of the total population of French Algeria) was displaced in a different way. Well over half of the settler population left Algeria in 1962 alone, with more soon following. ⁷⁵ Most of the country’s Jewish population, 110–120,000 people, also left: they were not settlers, but had been made fully French by the Crémieux decree of 1870 and slowly assimilated to the colonial population since then. ⁷⁶ The French government had not anticipated departures on such a large scale. ⁷⁷ Houses, apartments, and land were left empty. When French troops withdrew from the Rivail regroupment camp, in the Cherchel region east of Algiers, its inhabitants destroyed the barbed wire fences that surrounded (and symbolized) the camp. Mohamed Sari, a child there at the time, recalled that his grandfather would have preferred to go back to their land in the mountains. But his father hoped that land abandoned by settlers might be distributed to Algerians and decided to move to the plains to take up the opportunity if it arose. ⁷⁸ As in other settler colonies where an indigenous majority population remained, the expropriation of the country’s most productive agricultural land had been central to the violence of colonialism. ⁷⁹ The departure of the settler population offered the new state the possibility of bringing settler-owned agricultural land and the previously European quarters of the cities alike under Algerian Muslim ownership—an essential part of building a sovereign nation-state.

These complex landscapes of displacement in Algeria are far from unique: mass displacement forged nation-states around the world throughout the twentieth century. Inspired by anthropologist Liisa Malkki’s study of Burundian refugees in 1980s Tanzania, historians of Europe have explored how the shared experience of forced displacement acted as a driver of nationalist mobilization in the age of the World Wars. ⁸⁰ In Europe and the Middle East, the displacement and internment of suspect populations by dynastic empires during the First World War—from the Poles evacuated away from eastern front by Hohenzollern Germany to the Armenians targeted for genocidal deportation by the Ottoman empire—spread a nationalist political consciousness among previously dispersed populations far more effectively than pre-war nationalist parties had managed to. ⁸¹ Resettlement of displaced populations, often on land or in property from which non-‘national’ populations had been displaced, became a core mission of the new nation-states that took shape in the ruins of empire: Onur Yıldırım’s description of ‘a period of national reconstruction at the centre of which stood thousands of homeless, jobless and hungry refugees’ applies not just to Greece and Turkey but to a band of states stretching from the Black Sea to the Baltic. ⁸² Repeated in Europe in the 1940s, this phenomenon also became global. In China, population displacement during the Sino-Japanese war dwarfed that even in contemporary Europe, setting people of all social backgrounds together on the roads of exile and flinging them deep into the hinterland. It acted as a kind of pressure cooker for nationalism, and also for new conceptions of a social state with a duty to intervene to support its people in times of crisis. ⁸³ In south Asia, British imperial retreat triggered the displacements of partition and made displaced people a central focus of state-building efforts in both India and Pakistan; in Palestine, it intersected with the shockwaves of the war in Europe, resulting in the creation of a new state, Israel, for one group ‘nationalized’ through the trauma of genocide and displacement, and the displacement of Palestine’s Arab inhabitants—for whom encampment would also be a crucible for nationalism. ⁸⁴ The Algerian war was another episode in this history, as it spread to the other colonial empires. The shared experience of forced population displacement during the war, whether into exile in Morocco and Tunisia or into centres de regroupement in Algeria itself, spurred nationalist mobilization. Dispersed rural populations were literally concentrated, in a shared and consciousness-raising relationship to both the colonial military and to the FLN militants who operated clandestinely in the camps. ⁸⁵ And the question of who would provide for the displaced became a key issue for the nationalists, during and after the war. What was new in the Algerian case, but would become a common feature in the new states of the decolonizing global south for the rest of the twentieth century, was the involvement of the post-1951 UN refugee regime in responding to mass displacement.

UNHCR and the old and new refugees of independent Algeria

As Algeria moved towards independence, UNHCR staff began to plan for the agency’s future relationship with the new state. Would Algeria adhere to the convention? If so, how would practical responsibility for refugees shift from institutions of the French colonial state to those of the Algerian nation-state? The new state faced immense challenges in its early years, but in UNHCR archives we can see how Algerian state institutions nevertheless took over responsibility for ‘Convention’ refugees without much delay—a process which was also a gauge of their establishment, faltering at first, of sovereignty. Algeria began to receive new refugees, too, particularly from other African countries (and colonies). This made it one of the places where UNHCR was obliged to reconsider its own understanding of who should be defined as a refugee.

The transition from French to Algerian sovereignty raised, more pressingly for UNHCR than for either government, the question of what would happen to the ‘old’ or ‘Convention refugees’. Three days before the 1 July 1962 referendum on self-determination in Algeria, the High Commissioner wrote to his envoy in Algiers about their status. Algerian independence, he wrote, would ‘raise various legal problems, such as the applicability of the Refugee Convention, the recognition of the refugee status, the documentation of refugees and their legal status in Algeria’. The ceasefire agreement made no provision concerning Algeria’s future obligations in international relations. Recently independent countries, not just Tunisia and Morocco but also Niger, the Cameroons, and Ivory Coast, had ‘made declarations of continued obligation under the Convention’. ⁸⁶ But their transition to independence had been much less violent than Algeria’s, and the aftermath much less chaotic. In September, Kelly reported from Algiers that ‘there is no real Government in this country’. ⁸⁷ Legally, the Provisional Algerian Executive held sovereignty, but it had no function in foreign affairs. Independence had brought a political crisis verging on civil war between the military leadership that had been based within Algeria during the war and the returning political leadership that had been based abroad. When Kelly raised the issue of Algeria’s inherited international responsibilities with M. Guelal, chief of information in the foreign ministry, the latter confessed that it was ‘just one of those matters about which they had no time to deal’. Guelal asked what other newly independent countries had done about conventions ratified by the former imperial powers, but ‘was sure that there would be no difficulty at all about UN Conventions’.

UNHCR was concerned for the ‘old’ refugees because they risked getting caught in a no-man’s-land between French and Algerian sovereignty. OFPRA, the French refugee office, had informed the agency that at the end of 1962 it would stop renewing the ID cards of ‘old’ refugees still living in Algeria. ⁸⁸ Among this group, many needed support or demanded the agency’s help in getting to France. At independence, they found themselves as refugees within the formerly colonized population: a much poorer one than the disappearing settler population, which changed the level of support they could now expect from the state they were living in, even if it applied the convention and recognized them as refugees. But more immediately, and materially, they needed identity and travel documents. In the end, OFPRA continued to renew old documents into 1963, but could no longer issue documents to new applicants for fear of infringing Algerian sovereignty. ⁸⁹ Meanwhile, although the new Algerian government was too busy to create an administrative office for the old refugees at state level, its local authorities began to deal with them directly. In March 1963, the prefecture of Algiers, which held a stock of old OFPRA travel forms, began overwriting them to produce new Algerian documents. ⁹⁰ But this risked creating a confusion. At this stage, refugees were an exception to the rule that it was no longer possible to enter France from Algeria with expired travel documents: French consulates were issuing laissez-passer to refugees even if their existing OFPRA cards or passports were out of date. But if refugees were issued ‘Convention passports’ overwritten from French documents by the Algerian authorities, ‘the French authorities will be forced to consider refugees coming from Algeria as having found a second host country and being under the protection of the Algerian state’. ⁹¹

As UNHCR assisted the Algerian authorities in dealing with ‘old’ refugees, it was also faced with the issue of new refugees—and at first, the agency preferred not to consider them. In May 1963, Kelly’s replacement as delegate for Algeria, the Lebanese Michel Moussalli, sketched out recommendations for a refugee status determination procedure in Algeria. He noted that ‘A large number [un nombre élevé] of refugees not generally recognized as coming under the Convention live in Algeria. Thus, some Portuguese, Angolans, Tunisians, Moroccans, etc.’ ⁹² Algeria after independence became known as the ‘Mecca of revolution’: it attracted many activists, and several liberation movements set up offices in Algiers. ⁹³ This quotation refers to such groups. In September, when there was talk of UNHCR assisting the Algerian authorities in taking a census of ‘Convention refugees’, Moussalli warned again that ‘Algeria will be the refuge of all sorts of new refugees, and it will require a great deal of diplomacy not to be pulled in a direction which the [High Commissioner] would not wish to take.’ ⁹⁴ For Moussalli, in other words, Algeria had the potential to force UNHCR to change its definition of whom it considered to be a refugee. Only a few years later, the 1967 Protocol brought just such a change.

Through the 1960s and 70s, UNHCR in Algeria continued working with refugees both old and new. Helping the new government set up its own Algerian Office for the Protection of Refugees and Stateless Persons provided an opportunity to discreetly monitor the government’s assistance to returnees from Morocco and Tunisia (which UNHCR subsidized), and to reassure the remaining ‘old’ refugees, who were wary of the new government’s plans to count them. ⁹⁵ This was difficult anyway in a country emerging from the chaos of the war, though a communiqué was published in La République d’Oran in January 1964 inviting them to come forward. ⁹⁶ By spring 1964, the country was already hosting a new refugee population, of several hundred Malians: UNHCR was involved in helping them go home, though in the 1970s another wave fled a drought-induced famine. ⁹⁷ The agency also continued to assist Algeria’s ‘old’ refugees into the 1970s, as they literally grew old. In 1969, Spanish Caritas signalled the existence of about 400 Spanish refugees in Oran, about fifty of whom had families. ⁹⁸ Older and poorer refugees received support from religious communities such as the Petites Sœurs des Pauvres, if only on their deathbeds; others found it difficult to find jobs now that French companies were no longer operating in Algeria; others seemed again to be living normal lives and were known to the French authorities for obtaining visas for tourism or work to travel to France. ⁹⁹ After the death of Franco, several retired Spanish refugees wanted to return to their country of origin, though they would not be able to be paid their Algerian pension there. ¹⁰⁰ By that time, UNHCR was almost unrecognizable from the small technical office, focused on legal issues and the coordination of other agencies’ humanitarian action, that had hesitantly extended its operations beyond Europe in the 1950s. Enshrined by the 1967 Protocol, the globalization of the international refugee regime had become unstoppable, and UNHCR was becoming a major humanitarian actor in its own right in response to the refugee movements in Southeast Asia. ¹⁰¹

Conclusion

UNHCR’s involvement in the Algerian war deserves our attention, both as a turning point in the history of the modern international refugee regime and as an exemplary case for understanding how that regime became an arena for the establishment of postcolonial sovereignty.

Writing the history of the refugee regime from North Africa shows, in a way that writing it from Geneva cannot, how it was reshaped in the 1950s under the impetus of the newly independent states of Morocco and Tunisia, and the anticolonial nationalist movement of Algeria. Recognizing the role they played in drawing the refugee regime ‘beyond Europe’ offers us a new way of understanding its path from the 1951 Convention to the 1967 Protocol: newly independent states in the global south created much of the momentum for the globalization of the refugee regime. Former French colonies in Africa took the lead, because France’s (unusual) blanket application of the convention to its colonial empire meant that they succeeded to it at independence. ¹⁰² Morocco and Tunisia extended UNHCR’s operations beyond Europe in the late 1950s by requesting the agency’s assistance. In the 1960s, former French colonies in West Africa would create momentum for the revision of the text of the convention itself: several of them, on becoming independent and succeeding to the convention, notified the High Commission that they would depart from French precedent and apply the geographically unlimited interpretation of the text. ¹⁰³ This decisively shifted the balance among contracting states away from the limited version and towards the 1967 Protocol. Why they did this, and whether their action was coordinated, remains to be seen. But there is a global history to be written of how states that succeeded to the convention as former colonies of European powers interacted with, and remade, the refugee regime. ¹⁰⁴ Many other present-day UN member states first joined the regime in this way. ¹⁰⁵

For Morocco and Tunisia, interacting with the refugee regime helped to establish their credentials as independent states, and brought the meaning of their sovereignty into sharper definition—whether for the populations who now found themselves firmly categorized as Algerian, Moroccan, or Tunisian, or for the territories from which they could now decide to exclude certain foreign agencies. This subject too would merit further attention: how did the arrival and settlement of refugees stimulate processes of territorialization in each country, and how did political and societal responses to the refugees relate to the articulation of their national identities? ¹⁰⁶ For the Algerian provisional government, meanwhile, interacting with the refugee regime during the war was part of its effort to use international diplomatic and humanitarian forums to make itself the recognized representative of the Algerian people. ¹⁰⁷ Resettling the returning refugees in the midst of the much greater displacement crisis of the regroupés showed the transitional authorities at and beyond the limits not just of their ‘state capacity’ but of their sovereignty over Algerian bodies and the new national territory. As the multiple crises of 1962 slowly subsided and the new state took bureaucratic, political, and diplomatic shape, Algeria’s participation in the refugee regime gradually stabilized. Following the refugees allows us to see the texture of postcolonial sovereignty in formation, a decades-long process that took place everywhere from the minefields of the border zone to the filing cabinets of urban prefectures, from the government buildings of Algiers to the seat of UNHCR in Geneva. Algeria’s participation in the regime shows how deep was the French imprint on the new state’s understanding of government: the administrative office it set up to support the implementation of the convention in Algeria was, as its name suggests, calqued on the Office français de protection des réfugiés et apatrides, which offered to train its personnel. ¹⁰⁸ But this is only part of the story, because Algeria also asserted its independence by choosing when not to participate in the regime. When exiled activists from other independence movements in and beyond Africa arrived in the new state, it did not see them as refugees or seek to involve UNHCR in assisting them. It hosted them as part of a revolutionary responsibility to assist the anticolonial struggle, successful in Algeria, as it continued elsewhere.