States and refugee advocates often insist that ‘refugee’ and ‘migrant’ are separate distinct categories, despite ample evidence that these labels blur in practice. However, little attention is paid to the fact that in the past refugees were considered as migrants, with international attention focusing on securing their access to existing migration channels. This article traces this tangled history of refugee and migrant identities through the 1920s to the 1950s, when ‘refugee’ and ‘migrant’ categories were separated. The article argues that treating refugees as migrants in the 1920s and 1930s failed to ensure their protection from persecution because their admission was entirely dependent upon economic criteria. Separating refugees from migrants in the 1950s—by providing refugees with an exceptional right to cross borders and claim asylum—helped to address this protection gap. However, the article shows that in creating a special route for admission deliberately set apart from migration, the humanitarian discourse that protects refugees from harm actually prevents refugees from finding durable solutions, which depend upon securing an economic livelihood and not just receiving humanitarian assistance. The article concludes that, in the interests of refugee solutions, the extent of separation between refugee protection and access to migration should be reversed. Refugee advocates should reconsider the many innovative lessons both from the Nansen era and the decade of experimentation that preceded the establishment of today’s contemporary refugee protection framework in 1951. While asylum and the ‘refugee’ category perform essential roles in admitting those in need of international protection, asylum alone—unlike migration—cannot meet long-term needs. Reconfiguring understandings of on-going refugee protection to facilitate movement and prioritize the securing of sustainable livelihoods would both better reflect the reality of people’s movements in conflict and crisis and offer more opportunities for durable solutions to protracted crises.
This article traces the tangled history of refugee and migrant identities.1 It explores how, until the 1950s, refugees were essentially considered to be a subcategory of migrants. The international community was less focused on the need to ensure universal access to asylum, and more on the need to find solutions for existing European refugees. As a result, refugee and migrant were fluid identities, with both words used to describe a ‘surplus population’ that included not only refugees but also unemployed nationals. Their collective emigration would solve Europe’s economic and political problems.
However, the signing of the 1951 Convention saw the identity of refugee clearly separated from that of migrant, above all to protect refugees’ rights of claiming admission and asylum. This was a response to the evident failure of the 1930s migration regime to offer adequate protections to those attempting to leave Germany in that period. Through the 1950s, UNHCR and other refugee protection actors also increasingly promoted the humanitarian resettlement of vulnerable refugees who could not work, while states’ interests in restricting and ‘managing’ migration grew. The solution of refugee resettlement was increasingly placed apart from normal migration channels.
Since this time, NGOs and UNHCR have broadly followed the line that ‘refugees are not migrants’ as a means of protecting asylum space, despite a broad recognition that the line between a ‘refugee’ and a ‘migrant’ is often relatively arbitrary. However, the result is that a humanitarian discourse intended to protect refugees has in fact strengthened many states’ restrictionist migration agendas, and prevented refugees being included within migration-development discourses. While asylum protects an immediate right of entry, this failure to consider refugees’ economic needs has seriously impacted on long-term quality of protection and refugees’ ability to access sustainable and meaningful solutions (see DFID 2007; Van Hear and Nyberg-Sorensen 2003).
I therefore argue that the extent of separation between refugee protection and migration has in fact proved to be counter-productive in many instances and should be revised. The existing system of humanitarian aid suffers from a ‘sedentary bias’ that exacerbates refugees’ inequality and dependency and has failed to meet broader geo-political interests in stabilizing and developing refugee-producing regions (Bakewell 2008; Long 2010). Reconfiguring understandings of longer-term refugee protection to facilitate movement and prioritize the securing of sustainable livelihoods would both better protect refugees’ human dignity and better reflect the reality of people’s movements in conflict and crisis.
1. A historical approach
In the past decade, a significant body of research has been carried out investigating the ‘asylum-migration nexus’ (see e.g. Castles 2003: 17). These findings have shown persuasively that ‘refugee’ and ‘migrant’ flows are often interconnected, with communities, families and even individuals shifting between these different policy categories. This article builds on this analysis, but offers a new—and previously relatively neglected—historical account of how these institutional categories have emerged.
Studies of forced migration have often been accused of ahistoricism. The year 1951—the first year in which UNHCR was operational and the year in which the Convention on the Status of Refugees was agreed—often functions as a ‘year zero’, in part because these remain the twin pillars of the contemporary refugee protection regime (Haddad 2008; Hathaway 2007). While this approach has obvious advantages, particularly for those concerned with the impact of current policies on contemporary refugee populations, its limitations are significant.
The 1951 regime is a historical artefact: understanding how and why it was constructed is important, not least to avoid any assumption that it represents the ‘natural’ or ‘obvious’ conclusion of earlier debates. Archival material makes very clear that the distinct humanitarian character of the existing refugee protection regime was by no means an inevitable conclusion of negotiations through the 1940s, but a politically crafted construction of Western states, intended to respond to the specific dynamics of the post-1945 European refugee crisis and shaped by Cold War rivalries.
Most importantly for the purposes of this article, stripping away the 1951 regime makes far more visible the fluid and changing understandings of poverty, persecution, and protection that international policy-makers employed in developing regimes to govern both refugees and migrants during the inter-war period. Many scholars have discussed the artificial lines drawn between ‘migrant’ and ‘refugee’. However, by moving back behind the 1951 line, it becomes easier to discuss the processes of refugee migration and related humanitarian and developmental responses without becoming entangled in contemporary debates surrounding the political use of ‘refuge’ and ‘migrant’ labels (see e.g. Betts 2009; Zetter 2007).
This article uses the term ‘refugee’ as it was broadly used by the international community in the inter-war period: that is, to identify those individuals and groups forced to leave their country of origin because they were politically excluded (as opposed to simply impoverished). Migration is understood as a primarily socio-economic phenomenon: this article focuses on labour migration, but recognizes that migration may also take place for other reasons, such as access to education or family reunification. The term ‘humanitarian’—used particularly in this article in relation to refugees’ admission into a country of asylum—is used to describe pleas for states to offer neutral and impartial protection from persecution. This does not of course preclude the notion that the decision to recognise such claims may be in fact highly political, particularly in terms of determining what constitutes persecution.
The body of this article is divided into three parts. The first section outlines the contemporary contradictions in approaches to ‘refugees’ and ‘migrants’ that dominate not only state but also NGO approaches to refugee protection. It highlights the growing number of field-based studies that show uses of mobility and migration as key protection strategies employed by refugees and the displaced themselves, despite the legal obstacles intended to prevent this movement.
The second part of the article outlines how understandings of refugee migration influenced early approaches to refugee protection during the inter-war period. It also suggests when and why ‘refugee’ and ‘migrant’ began to be approached as definitively separate categories by international policy-makers, splitting humanitarian and development approaches to mass movement. This main body of the article draws on archival research carried out in the League of Nations Archive in Geneva (2008), the UK National Archives (2011), Churchill College Cambridge Archives (2011), and Hansard’s online parliamentary records (2011).
The final section of the paper offers a brief conclusion, arguing that early approaches to refugee protection in fact demonstrated a more holistic, development-focused approach to ‘solving’ refugee crises than many contemporary strategies do. Replacing refugees within migration and development discourses is an essential strategy with which to respond to the growing number of protracted displacement crises.
2. The devil in the definition
The relationship between the displaced and migration is a complex one. On the one hand, displaced persons—whether meeting the narrow ‘well-founded fear of persecution’ definition set out in the 1951 Refugee Convention, or more general criteria of flight from generalized violence, breakdown of public order, or other crises of governance—are indisputably involved in movement (UNGA 1951; OAU 1969). Before 1951, refugees were identified not as individuals, but as specific nationalities no longer physically present in or politically protected by their state of origin. Movement was thus an integral dimension of being characterized as a ‘refugee’, but more emphasis was placed on the need for protection than the distinction between forced and voluntary migration.
Yet on the other hand, international policy-making today—both on issues of migration and refugee protection—is closely tied to presenting refugees and migrants as categorically different groups. This practice of separating refugee from migrant is widespread in contemporary discourses, involving both actors who are concerned with the exclusion of migrants and those concerned with the inclusion of refugees.
It is well recognized, of course, that states, border agencies, and immigration departments feed a discourse in which a ‘genuine’ refugee claim is detected among those of ‘bogus’ asylum seekers by the refugee proving an absence of economic intent in arriving in a host country. Humanitarian interests are to be defended: as UK foreign secretary William Hague insisted in November 2011, ‘I simply want this country to be a safe haven for the genuine refugee’ (Daily Mail 2011). However, because—in the words of Phil Woolas, former UK Labour Minister for Immigration—‘most asylum seekers, it appears, are economic migrants’ (Tibbets 2008), a refugee is generally presented as a figure of humanitarian rescue, qualifying for protection only by virtue of the absence of any explicit economic aspirations.
Perhaps more surprising, is the extent to which this same division between refugee and migrant is asserted by many of those interested in further refugees’ protection. Ruud Lubbers, then United Nations High Commissioner for Refugees, insisted in 2004 that ‘refugees and migrants are fundamentally different’ (BBC 2004). Erika Feller, currently Assistant High Commissioner for Protection at UNHCR, wrote in 2005 that ‘refugees are not migrants … it is dangerous, and detrimental to refugee protection, to confuse the two groups, terminologically or otherwise’ (Feller 2005: 27). In a climate of public hostility towards migration in general, preserving space for refugee protection is thus seen by many NGOs and refugee advocates depend upon separating refugees from migrants.
These debates are often replicated in research (for a good discussion of these debates see Bakewell 2011). Zetter’s seminal work on the functions and processes of ‘labelling’ helps underline that of course this space between ‘refugee’ and ‘migrant’ reflects not empirical reality but policy intentions: ‘[L]abels do not exist in a vacuum. They are the tangible representation of policies and programmes … labels develop their own rationale and legitimacy and become a convenient and accepted shorthand’ (Zetter 2007).
Yet even while insisting on maintaining the ‘crucial’ distinction between ‘refugee’ and ‘migrant’, refugee advocates are also adept at presenting the displaced to a sceptical public as valuable economic assets: consider the British Refugee Council’s 2010 Refugee Week Campaign ‘As British as Fish and Chips’, which focused on highlighting the contributions of refugee entrepreneurs to British business (Refugee Council 2011).
Equally, there is a general—albeit contradictory—recognition that poverty and persecution at the source are often interconnected. This is particularly true in states where crises of governance have led to long-term under-development of infrastructure and a consequent lack of basic goods and services. The development of ‘mixed migration’ discourse is a partial recognition of this (although it still treats refugee and migrant as separate individuals, albeit travelling in the same flow) (UNHCR 2007). Even while insisting that refugees were not migrants in 2005, Feller acknowledged at the same time the role migration might play in ‘solving’ refugee crises, suggesting that UNHCR should begin to look at ‘using the resettlement of refugees also in a migration context to assist states to address such demographic challenges’ (Feller 2005: 32).
Yet despite this recognition that the policy distinction between refugee and migrant is in many ways an artificial construction, there has been remarkably little research carried out which looks at the historical relationship between the category of refugee and processes of migration (two important exceptions are Karatani 2005; Elie 2010). This is a significant research gap, not least because it is very clear that when an international refugee protection regime was first established in the 1920s, refugees were very much considered to be migrants, both in terms of admission (asylum) and solution. The so-called Nansen Passport—a refugee travel document first introduced in 1922—was central to the nascent refugee protection regime and was intended to facilitate refugees’ onward movement and through this their access to labour markets.
Refugees’ problems remain two-fold. They first require access to asylum as protection against persecution in their country of origin (now guaranteed under the 1951 Convention). They then require on-going protection, and finally a solution to exile. Yet solutions today are instead largely conceived of in terms of stopping—and preferably reversing movement. Even refugee resettlement—the movement of refugees in a first country of asylum to a (Western) third country, where they will be offered permanent residency and a pathway to naturalisation—is largely framed as a humanitarian process, with applicants judged on protection needs alone (UNHCR 2011).
The question therefore needs to be asked: when—and why—did refugees stop being considered migrants? What impact did this shift in identities have on refugee protection? In looking to respond to these issues, this article now turns to consider how refugee protection was constructed around the twin facets of migration—movement and employment—in the first decades of the international refugee regime.
3. A history of refugee migration
3.1 The early years: Nansen passports and the role of the International Labour Office
The first co-ordinated international efforts at refugee protection were a response to the Russian exodus that followed the 1917 Revolution, subsequent Civil War and resulting famine. By February 1921, Gustav Ador, the President of the International Committee of the Red Cross, had written to the League of Nations about the condition of 800,000 Russian refugees ‘without legal protection and living in desperate poverty’, urging the League to intervene as ‘the only supranational political authority capable of solving a problem which is beyond the power of an exclusively humanitarian organization’ (Skran 1995: 84–85). The League’s response was to appoint Fridtjof Nansen as High Commissioner for Russian Refugees in August 1921: the mandate was later extended to include Armenian (1924), Assyrian and Assyro-Chaldean (1928), and Saarland refugees (1935).2
Refugees were thus perceived as suffering from both political disenfranchisement (in the case of Russian refugees, as was confirmed by the Soviet Union’s Denationalization Decree issued in December 1921) and economic poverty. It was the economic destitution of the refugees—and the difficulty of continuing to finance a relief-based humanitarian effort—that prompted Ador to request League intervention.
Repatriation to Bolshevik Russia was politically problematic (see Long 2009). Yet the prevalent logic of nationalizing state territories in Eastern Europe prevented any possibility of local integration. The need, therefore, was to facilitate the Russian refugees’ employment through onward movement. This meant that the Russian refugees’ lack of legal protection was most problematic—for the international community at least—because it prevented their onward migration through a Europe newly divided by passport controls, and ‘without passports, they lacked freedom of movement’ (Skran 1995: 103).
It was for this reason that Nansen’s refugee protection regime was centred on the facilitation of refugees’ onward movement. Following a conference in July 1922, delegates from sixteen governments agreed to adopt travel documents that would allow refugees to travel across international borders, in order to join family elsewhere (and through family a means of financial support) or seek out employment. Nansen’s objective was repeatedly described as ‘the admission of refugees to countries where they would be able to support themselves’ (LNA 1922). In other words, refugees were to be helped to become migrants: exile and destitution could be solved through continued movement. As Torpey has argued, ‘the Nansen passport system represented the first step toward resolving at the supra-national level the internal contradictions of a system of movement controls rooted in national membership’ (Torpey 2000: 129).
Treating refugees as migrants was, however, a double-edged sword. On the one hand, access to a Nansen passport allowed mobility, and through mobility an opportunity for immigration and employment, as well as an ability for refugees to ‘exercise their own initiative’. Yet as Kaprielian-Churchill (1994) has shown in the case of Canada, disputes over refugee ‘returnability’ in effect became a means of continuing to screen out ‘non-preferred’ classes of East European migrants. Because refugees lacked a country of return (rather than a country of origin per se), by 1926 Nansen complained to an international conference on the Russian and Armenian refugee problem that refugees faced ‘serious discrimination’ in attempting to travel ‘in pursuit of their livelihood or for health or family reasons’ (LNA 1926). Few European states—the notable exception being France (because of population losses suffered in the Great War)—were prepared to accept significant numbers of refugees.
The refugee passport system established in the 1920s was nevertheless remarkable not just for its insistence that the focus of international efforts should be solving refugee exile—which would be found through freedom of movement and resulting employment and settlement—rather than simply providing relief. Frustrated by states' continued reluctance to provide a revolving fund through which to facilitate the migration of refugees, Nansen eventually financed migration through the issue of passport stamps, for which non-indigent refugees were charged five gold francs.
Moving beyond merely facilitating freedom of movement in a bureaucratic sense, by the 1930s the Nansen stamp fund was large enough to directly help refugees become self-supporting by offering what would effectively be recognized today as micro-credit loans. In 1934, for instance, the Nansen Office assisted 782 refugees with loans intended to help support businesses—twelve restaurants, five laundries, five tobacco shops, five rabbit farms, and a doll factory (Skran 1995: 193). Again, this suggests an approach to refugee protection centred on development and economic inclusion, rather than classic forms of humanitarian relief.
Nansen passports were eventually recognized by some fifty-two governments (Skran 1995: 108). They stand as evidence of an approach to refugee protection in the 1920s that was above all fixed around facilitating individual freedom of movement as a form of burden sharing. If refugees could become migrants, their ‘problem’—viewed overwhelmingly in terms of the need to ensure economic self-sufficiency—could be solved. This in part reflected a wider contemporary aspiration to liberalise movement controls, with states repeatedly stressing during the 1920s their eventual intention to remove all passport and visa controls (Salter 2003: 79).
Further confirmation of the extent to which refugees were viewed as a special category of economic migrant can be found in the fact that between 1925 and 1929, the International Labour Office (ILO) assumed operational responsibility for refugees. Nansen’s office continued to deal with the legal and political aspects of the refugee problem. This economic perspective on refugee exile, and the close connection made between protection and work saw the ILO operate as an effective refugee labour exchange, matching refugees with labour needs outside Europe. Refugee exile was thus one part of a broader concern to tackle the much greater project of global unemployment through targeted migration. This bought together humanitarian and longer-term development aims.
A centralized migration bureau matched receiving countries’ employment needs with refugee quotas, often broken down by occupation. In 1926, for example, having opened two new offices in South America, the ILO Refugee Service reached agreement with the state of Sao Paolo in Brazil to accept 265 refugees a month, a number which was to include fifteen locksmiths, five stucco plasterers, and five bootmakers (LNA 1926). Many more refugees found work as agricultural labourers: 10,000 Ukrainians were placed on Canadian farms, while 7,600 Russian refugees were employed in France, with smaller groups being sent to Belgium and Luxembourg. In August 1928, the director of the ILO reported that the number of unemployed refugees had dropped from 400,000 to 200,000 since the office had become involved in refugee work, with 50,000 direct placements found through ILO auspices (Holborn 1939: 130).
There was thus no clear distinction made by the international community in the 1920s between political refugee and impoverished migrant. Although the Russian, Armenian, and other minority refugees had suffered a political exclusion at the hands of the Soviet and Turkish states, their ‘problem’ was viewed above all as one of economic poverty, which was compounded by their difficulties in establishing a legal identity in the new world of nation-states. Legal protections mattered above all because they could help to facilitate onward movement and settlement.
This is not to suggest that, for many of the refugees involved, the problem of exile was far more complex, involving an acute loss which could only be met through a return to a lost ‘home’. However, for the international actors involved in responding to the post-war refugee crisis, the solution to this crisis was very clearly framed by an understanding of refugees as impoverished economic migrants. What might be identified in today’s terms as the ‘relief-development’ gap—the vexed question of how to move away from humanitarian aid and provide sustainable solutions to displacement—was to be filled through facilitating refugees’ independent migration. This promotion of a path to settlement and employment was an alternative to costly relief-driven humanitarian projects. While the focus on migration was to some extent dictated by the lack of prospects for return or local integration, it also reflected a broader cultural understanding that placed economic poverty at the centre of a European plan for ‘emigration as development’.
However, while treating refugees as migrants offered solutions to some, this strategy also excluded many other refugees from moving beyond asylum. The unravelling of European empire had not only created a refugee population, but exacerbated the drive towards restrictive immigration policies intended to preserve ‘national’ cultures not only in Europe, but also the Americas, Palestine, and Australasia. Qualifications, age, health and race could all dictate migration policy. In 1929, the High Commission took back technical responsibilities for refugee settlement and employment from the ILO. This reflected the worsening economic climate, in which labour permits were increasingly difficult to secure and migration routes increasingly circumscribed.
As a result, while migration offered a pragmatic strategy through which to approach refugees’ economic poverty, it was not an approach easily reconciled with an idea of universal humanitarian compassion. Yet in the 1920s, there was also no real concern that Russian or Armenian refugees would not be granted basic asylum—particularly given Western Europe’s political opposition to Bolshevism and Turkish expansion. The humanitarian crisis lay in dealing with the effects of exile, rather than preserving asylum space per se. Nansen was still optimistic in 1929 that with sufficient funding, the refugee problem might be entirely resolved within a decade (Holborn 1939: 131).
The real difficulty, however, lay in persuading states to adopt more flexible approaches to the migration of the poor, a group which included refugees, once economic conditions worsened after 1929. Domestic politics meant states chose to spend one hundred times more on refugee relief than they collectively contributed to the ILO migration and settlement programmes between 1925 and 1929. Humanitarian aid and maintenance for refugees thus emerged as a deliberate alternative to facilitating their inclusion—or tolerating the expansion of international migration flows which might offer a solution to refugees’ economic poverty, but which ran against the nationalisation of state space driving migration controls (see e.g. Torpey 2000: 6–8).
3.2 Asylum and migration paralysis in the 1930s
The 1930s saw the almost total breakdown of both migration and asylum frameworks (see e.g. Marrus 1985; Skran 1995). From 1933, European states used a number of strategies to avoid offering asylum to Germany’s Jewish population, including redoubling visa restrictions to prevent arrivals at the border and physical sealing of borders. These strategies formed part of a wider trend to limit immigration quotas in the face of the Great Depression, which made labour visas extremely difficult for Eastern Europeans (including German Jews) to obtain. US quotas for immigrants from Germany were not actually filled during the 1930s, largely due to a narrow interpretation of who was ‘likely to become a public charge’ during the Hoover administration (see e.g. Mitchell 1992). However available in theory, these restrictions on the issuing of visas to all who might become a public charge (until the public charge clause was renounced by President Roosevelt in March 1938) meant that many of the persecuted who applied for entry to the USA were refused entry because they were also poor (see Mitchell 1992: 940–942; Torpey 2000: 139–141).
This collapse of the humanitarian framework for asylum admission in the 1930s—although exacerbated by German aggression and latent anti-semitism among host communities—was essentially the result of the collapse in economic migration. Refugees were still primarily viewed as impoverished migrants: but global economic crisis now meant that there was little appetite to support large numbers of destitute exiles. The 1933 and 1938 Refugee Convention (more limited in scope)—which included a prohibition against the forcible return of refugees—attracted little support from states, who continued to frame Germany’s Jewish refugee exodus in migration terms, and to insist on the impossibility of admitting impoverished migrants in times of economic chaos. Writing in 1939, the General Secretary of the Irish Refugee relief committee argued that the refugee problem was ‘primarily an economic one … even in Germany it is not difficult to see a main structure of economic trouble behind a façade of “race purity”’ (Ditchburn 1949: 282).
Domestic political opinion in host communities confirmed that tolerance of refugee numbers was closely linked to attitudes towards immigration. A 1938 US opinion poll, for example, showed that 94 per cent of Americans disapproved of Kristallnacht, but 77 per cent thought immigration quotas should not be raised to allow additional Jewish migration. States refused to treat refugees as a separate category with an additional moral claim for admission because of their concerns with immigration—the Canadian delegation at the 1938 Evian Conference, for instance, insisted that ‘under no circumstances would it agree to a special admissions quota for refugees’ (Skran 1995: 213, 132).
Even those anxious to broker refugees’ access to asylum space understood this as a form of economic migration—albeit with added humanitarian imperative. In part, this reflected the political preoccupations of 1930s Europe. Leon Blum’s insistence (on taking office in France in 1936 and ratifying the 1933 Convention) that ‘the right to asylum cannot be separated from the right to work’ (Skran 1993: 132) reflected not only a recognition that economic poverty was the immediate threat facing Spanish Republican and German refugees, but also a broader commitment to the idea of access to work as an essential component of a ‘full life’.
Similarly, studying the records of UK parliamentary debates illustrate the extent to which refugee advocates in the UK understood the worsening persecution of German Jews in economic terms, and presented the solution to this political crisis in terms of migration. As Philip Noel-Barker—who had previously worked as an advisor to Nansen in Geneva, and played a crucial role in securing the Kindertransport of Jewish refugee children to Britain in 1938—argued in April 1939, those Jews who remained in Germany were facing ‘the destruction of the[ir] earning power … the moment when they must leave Germany or die comes nearer’. Humanitarian relief being offered in Vienna was critiqued as ‘simply keeping the refugees alive’ rather than offering any hope of a ‘final settlement’. Importantly, Noel-Baker insisted that ‘it is vital that the whole question should be treated no longer as a humanitarian matter but as one of urgent political and economic importance to Europe as a whole … since 1931 the refugee has always been the first victim of the economic crisis’ (Philip Noel-Baker in Hansard 1939).
Both the impossibility of admitting refugees in the 1930s and the possibility of doing so were debated and justified in economic terms. On the one hand, many UK parliamentarians were clear that UK population density meant that ‘it is impossible for us to absorb any large number of refugees here’. British unemployment was the priority, so there was simply no room for refugees. Yet Noel-Baker’s response was to argue that refugees were an economic asset ‘even in this country, 11,000 refugees who have come to stay have given new employment to 15,000 British subjects’. Rather than insisting on a programme of admission as a humanitarian good, refugees were essentially presented as productive economic migrants. In fact, the problems of poverty and unemployment were presented as fundamentally connected to the ‘immobility of labour’. The need was to ensure the possibility of refugee movement, which had broken down under the stress of economic protectionism. Migration—freedom of movement facilitated by the international community—offered the pathway to solution: ‘[I]t may seem a little thin, its validity limited in both time and space, but to a refugee, a Nansen passport is often the beginning of salvation’ (Churchill College Archives: NBKR 4.574/5, June 1938).
Yet trying to broker space for refugees to be treated as migrants failed to offer these groups either humanitarian sanctuary or economic livelihoods. The terrible consequences of the breakdown of 1930s international co-operation are obvious. The refusal of states to admit refugee-migrants led directly to internment and death (Ogilvie and Miller 2006: 174).
In seeking to understand the pre-1951 relationship between refugees and processes of migration, what is important is to recognize that it was the closure of migration channels that damned many would-be refugees because as these closed they were denied access not only to a solution, but to asylum. The blurring of refugee and migrant identities—and the lack of any robust international refugee protection framework guaranteeing refugees’ admission regardless of economic rational—meant that states did not have to compromise national political protection by accepting broader global responsibilities to admit those who could no longer remain in their country of origin.
3.3 A mixed migration regime: European voluntary workers
After 1945, the struggle to define ‘refugee’ and ‘migrant’ continued to exercise both states and humanitarian actors, culminating in the establishment of a mixed regime in 1951 that—although it offered special protection to refugees—continued in many other ways to elide refugee and migrant identities.
The size of Europe’s post-World War II refugee crisis was irrefutable and led to a decade of intense experimentation in relation to an international refugee protection regime. At the close of World War II, between twenty and thirty million displaced persons were scattered across Europe (Loescher 2001: 34). Many were suffering from ‘simple’ physical dislocation and were eager to return home. In the three years following its creation in 1943, the United Nations Relief and Rehabilitation Association (UNRRA) assisted some seven million displaced persons with their repatriation.
However by the late 1940s, mass repatriation of Eastern bloc refugees—those who refused to return—was no longer a political option. UNRRA's successor organization, the International Refugee Organization (IRO), repatriated only 73,000 displaced persons between 1946 and 1952, while helping one million to resettle (Loescher 2001: 41).
The IRO constitution repeatedly stressed its mandate was to care only for ‘genuine refugees’—a formula which excluded all ethnic Germans expelled under the Potsdam agreements, as well as Soviet ‘quislings and traitors’ (IRO 1946; UNGA 1946). Establishing a refugee claim to IRO assistance therefore rested not only on humanitarian need, but also on criteria that excluded those ‘undeserving’ by virtue of their association with the Axis powers. Yet if refugees were a group whose ‘unreturnability’ and ‘deserving victimhood’ were increasingly being established as the twin cores of their identity, the relationship between refugees and processes of onward migration was less certain. Those fitting a refugee profile were offered formal international solutions that depended on their ability to meet economic migration profiles.
One example of this blurring of refugee and migrant identities by states can be seen in the UKs European Voluntary Worker (EVW) programme that ran between 1947 and 1951. These programmes saw several thousand Central and Eastern European workers recruited from refugee camps to fill the gaps in key industries—in particular mining, agricultural, and domestic work—that had resulted from wartime demographic shifts within the UK population leaving key industries under-strength at the same time as Continental Europe suffered the burden of refugees’ ‘surplus population’ (see Kay and Miles 1988, 1992; Vo 2010). EVWs were explicitly recruited as labour migrants, and were required to sign twelve-month work contracts as a condition of their provisional stay in Britain. In effect, what was being offered was a conditional migration as a pathway to potential settlement: ‘[T]his period of permitted stay will be extended further if the individual complies with the employment conditions set out immediately above and behaves as a worthy member of the British community’ (United Kingdom National Archives (UKNA) LAB 17/187).
The Balts, Ukrainians, and Poles who arrived in Britain as part of the Baltic-Cygnet and Westward Ho! Labour recruitment programmes understood themselves to be refugees as well as migrants. The politics of Cold War Europe meant these individuals could not return ‘home’. Yet the EVW scheme was seen by the British state almost exclusively in labour terms. The workers were refused access to a London Travel Document (the successor to the Nansen Passport, established in 1946) during their first twelve months of residency, for fear that the best recruits might ‘travel to the land of their choice’ (Vo 2010: 88). The idea of freedom of movement that had underpinned Nansen’s earlier plans for refugee migration was replaced with centrally controlled, indentured labour.
This migration offered opportunity to leave the privations of the displacement camps in central Europe, but the bargain was explicit: refugees were to work for their eventual right to settlement. Despite the IROs distressed protests, in the most obvious example of the instrumentalism of the UKs approach to this refugee migration, EVWs with medical needs—including pregnant women—were returned to displaced persons' (DP) camps in Germany, where the IRO was winding up its relief operations (Vo 2010: 90).
Yet internal memos underline just how sensitive the British state was to accusations that the EVWs were being treated as ‘slave labour’ (UKNA LAB 8/1731). This was an accusation also frequently levelled by Soviet officials at Western governments (see below). It was in rebutting Soviet charges regarding the exploitative nature of IROs work as a cheap labour exchange that the British state proved most interested in labelling the EVWs as refugees, allowing a convenient defence against the USSR’s critiques of capitalist labour exploitation.
It is also noteworthy that in planning a publicity campaign in October 1947, the Ministry of Labour and National Service focused on connecting the ideas of work (explicit in the title ‘European Voluntary Workers’) with a historical narrative that focused on the economic benefits that offering Europeans refuge from persecution in past centuries brought (reception of ‘offer’), with pamphlets citing examples of fourteenth century Flemish weavers and seventeenth century French Huguenots (Vo 2010: 84). By presenting EVWs as both workers and refugees, combining productivity with need, it was hoped that domestic political tolerance could be negotiated.
As Kay and Miles have concluded, the EVW programmes—recruiting workers from eastern European refugee populations—thus deliberately shifted between presenting these migrants as refugees, and presenting them as workers: ‘[R]efugees were selected and landed in Britain largely according to an explicit criterion of economic utility and not, as is generally the case with refugees, according to an articulated humanitarian concern to harbour a refugee from persecution’ (Kay and Miles 1988: 231). Issue linkage—a phrase beloved of contemporary policy strategists—allowed some form of refugee protection to be offered to those refugees who also met the criteria for working as much-needed industrial migrants.
By 1950 the initial restrictions placed on EVWs’ freedom of labour and movement had been largely removed, and the EVWs were allowed to settle in accordance with normal migration law. This resolved the dilemma of treating EVWs as contract migrants and restricting their freedoms upon arrival. It did not address the question of how those refugees who failed to meet migration criteria, and remained in the camps, should be offered a solution to their displacement, rather than merely relief. This hard core of ‘unwanted’ refugees—numbering around 150,000 by 1950—became the concern of IRO's successor organizations (Jaeger 2003: 10). This problem would eventually transform refugee resettlement from a migration-focused activity into one centred around humanitarian need.
3.4 The return (and retreat) of the ILO
The IRO successfully resettled over one million refugees, but at enormous financial cost. By 1949 it was clear that the USA—who had provided around forty per cent of IROs operating budget—would not renew the IROs mandate. The question now was what shape the next iteration of the refugee protection regime would take.
What was clear was that there was a political willingness in the West to develop a legal framework that created a space of humanitarian exception around refugees’ rights to seek admission and prohibited forcible return (refoulement). These norms were eventually codified in 1951 as Article 33 of the Convention on the Status of Refugees. In essence, this provision ‘solved’ the asylum dilemma of the 1930s. Refugees could now claim admission on humanitarian grounds, without reference to the socio-economic criteria applied to migration channels. However, in separating the humanitarian from the economic drivers of migration, the question of how to solve many refugee crises now became less clear.
The first dimension of the post-IRO regime to take concrete shape—in December 1950—was the establishment of the United Nations High Commission for Refugees. UNHCR was to be ‘humanitarian and social’ and ‘of an entirely non-political character’. Although charged with both ‘providing international protection’ and ‘seeking permanent solutions’ for refugees, both these tasks were to be accomplished through assisting governments and private organisations, rather than through direct operational activity (UNGA 1950: Annex, para 1). A line can therefore be drawn between UNHCR as imagined in December 1950 by the international community, and Nansen’s Office between 1925 and 1929, when it was concerned with the legal rather than the ‘technical’ or ‘operative’ aspects of protection.
The question of the relationship between ‘refugee’ and ‘migrant’ identities was largely absent from the statute. However, the core competencies of the High Commissioner were listed to include ‘promoting the admission of refugees, not excluding those in the most destitute categories, to the territories of States’ (UNGA 1950: Annex). On the one hand, this phrase suggested a continued use of normal migration channels to move refugees out of the displacement camps. Yet on the other hand, the explicit focus on those ‘in the most destitute categories’ suggested the High Commissioner would be expected to appeal to states' humanitarian compassion to open up new refugee resettlement channels for those who could not hope to qualify as economic migrants.
The second pillar of the post-IRO regime, the Convention on the Status of Refugees agreed in July 1951, was primarily concerned with establishing the rights of refugees already in exile. As such, it continued to confirm the importance of securing economic rights for refugees. Article 17 emphasises the importance of refugees’ access to work in line with ‘the most favourable treatment accorded to nationals of a foreign country in the same circumstances’. Still more important in terms of understanding the perceived relationship between refugee and migrant, Article 17 makes explicit reference to the treatment of refugee-migrants, stipulating that:
The Contracting States shall give sympathetic consideration to assimilating the rights of all refugees with regard to wage-earning employment to those of nationals, and in particular of those refugees who have entered their territory pursuant to programmes of labour recruitment or under immigration schemes (UNGA 1951: Article 17, my italics).
This suggests not only that those drafting the 1951 Convention expected refugees to continue to move as migrants, but that by virtue of these refugee-migrants’ contribution as labourers, they should be able to access an accelerated pathway to integration.
Article 28 of the Convention also established states’ obligations to supply refugees with a travel document ‘unless compelling reasons of national security or public order otherwise require’ (UNGA 1951: 28). These Convention Travel Documents (CTDs) were to guarantee the re-admission of the refugee to the issuing state, thus avoiding the vexations over ‘returnability’ that had plagued the early Nansen passports. The aim, however, remained the same: to empower refugees to ‘go out and find [their] feet in another country’ (Ad-hoc committee of the Conference of Plenipotentiaries, cited in Hathaway 2005: 847). Facilitating onward movement would allow refugees to find their own solution to exile.
As the noted refugee law scholar James Hathaway has argued, the aim of Article 28 was ‘to establish a unified regime for international freedom of movement that exists in parallel to the more general passport-based system’ (Hatahway 2005: 850). Refugees were to be included within a liberal system of global governance that envisioned the circulation of labour as well as goods.
Soviet participation in the discussions that established this framework were limited, but offered a revealing insight into the tensions between the humanitarian claims of Western states and their use of refugee labour, which meant permanent settlement was often conditional on completing a labour contract. Angry at the concentration of the IRO on resettlement rather than repatriation, the USSR representative repeatedly railed that:
The [refugee] camps were being turned into a slave market where representatives of the United States, the United Kingdom, France, Australia and so forth came to recruit cheap labour … The IRO presided over this market, and had been reduced to a mere employment agency, acting in the interests of the capitalists, whose only idea was to obtain the labour that they needed at the lowest price. No humanitarian principles inspired the work of the IRO. It was merely a profit-making concern for the capitalists of the beneficiary States (United Nations 1949).
Although there was an element of the absurd in the USSR’s interest in the labour conditions faced by Russians seeking refuge from its own totalitarian politics, the comments arguably help to explain some of the sensitivity felt even today by humanitarian actors when considering the role of refugee labour. Western states continued to stick doggedly to incorporating refugees within migration quotas, so that although asylum and basic relief continued to be offered in camps, settlement and the possibility of ending exile was often effectively conditional on meeting economic criteria.
The Convention and UNHCR to some extent circumvented this dilemma by focusing on broad refugee protections. The implication was that migration could certainly offer one means of securing protection, and the international community’s role was to ensure no discrimination against refugee-migrants. However, neither the Convention nor UNHCR were concerned with operational activities which were designed not just to protect the right to move or to work, but were intended to actively solicit such opportunities. Debate over who was to take on IROs direct responsibilities for refugee transport placed the relationship between refugee and migrant under far more direct scrutiny.
4. From the ILO to PICMME
At the beginning of 1951, the idea of Europe’s ‘surplus population’ remained more important in operational terms than any distinction between migrant and refugee. In October 1951, the ILO held an inter-governmental conference in Naples at which it advanced its plan for a comprehensive agency that would incorporate refugees within a wider migration framework and effectively operate as a labour exchange. A total of 1.7 million refugees would be helped to migrate over five years through this ‘migration administration’. The ILO would thus be responsible for migrants (including refugees) as workers: the United Nations would deal with the legal protections necessary for migrants (including refugees) as aliens (Karatani 2005: 533).
However, the Naples conference was a total failure for the ILO, who saw their plan for a comprehensive agency to deal with surplus population ‘wholly rejected’, with the Americans ‘largely responsible for wrecking the conference’ (UKNA, CO 4919/13—22/10/1951). The ILO–UN ‘quest to build a single comprehensive regime for people on the move was quashed’ (Karatani 2005: 536). The US Congress was not prepared to release $10 million in funding to an organization whose members included the Soviet bloc states Czechoslovakia and Poland, and instead favoured the establishment of a much more tightly controlled inter-governmental—rather than international—organization which would focus solely on the transportation of migrants and refugees. At a conference held one month later in Brussels, the Provisional Inter-governmental Committee for the Movement of Migrants from Europe (PICMME, later the intergovernmental committee for European migration, ICEM, and today the International Organization for Migration, IOM) was established. As Astroush and Muntz (2011) have recently underlined—IOM continues today to tread an uneasy path between state’s interests in migration management and humanitarian objectives.
Karatani’s claim that the failure of the ILO plan was the moment at which refugee and migrant regimes were definitely separated, however, is a misreading of these events (Karatani 2005: 17). PICCME/ICEM's activities did not draw a distinct line between migrant and refugee—continuing to focus on the amorphous mixed ‘surplus population’, some 400,000 migrants and refugees were resettled during the 1950s under the committee’s auspices. However, what was at stake were competing political visions of migration, with important implications for those interested in refugee solutions.
The ILO—echoing the claims of Nansen and Johnson in the 1920s—argued that ‘orderly migration would not only solve the dislocation of population distribution, it would also contribute to the realisation of peace and social justice in the post war world … it was within their mandate to attack social justice and human misery’ (Karatani 2005: 517). Migration was presented as a fundamental component of basic development. Despite the traditional association of migration with liberal politics, this was freedom of movement embraced by post-war socialism. Elie’s work has suggested that PICMME/ICEM were also concerned primarily with the economic value of Europe’s surplus population, with the refugee issue of secondary importance (Elie 2010: 350; personal correspondence 2012). However, by removing control of transportation mechanisms from the ILO (with its Soviet bloc membership), the USA also gained a means of making an apparently powerful political commitment to protecting refugees—those who could never return—from Communism.
These political fault lines can also be traced in UK government correspondence, revealing how co-ordinating international migration shifted from being associated with questions around work and labour—and thus the primary responsibility of the Ministry of Labour—to being associated with inter-state politics—and thus claimed by the Foreign Office as their responsibility. In the aftermath of the Naples conference, Foreign Office staff were scathing in their attacks on the Labour ministry: ‘[T]he ministry of labour are … emotionally engaged in favour of the ILO which has been their protégé since its foundation … they instinctively seek to avoid any criticism of all ILO proposals … the whole problem is in any event primarily of political importance’ (UKNA 371/94365, 4921/1).
By 1951, the foundations of an international refugee regime had been established. Refugees were increasingly defined in political terms as anti-Soviet. However, if the ‘development’ view of refugee migration advanced by ILO had been effectively rejected, it had not yet been replaced by a purely ‘humanitarian’ claim for refugee resettlement. States remained reluctant to take refugees who were not also economically productive migrants, largely because this allowed them to retain tight control over admissions. Thus, even as refugees’ independent freedom of movement was curtailed by increasingly stringent criteria governing international movement, and even as refugees were increasingly defined in political terms rather than by measures of poverty, states continued to blur refugee and migrant identities to meet their own political interests.
4.1 Dividing refugees from migrants: Cold War politics and humanitarian action
The politics of refugee migration increasingly intersected with the politics of Cold War flight. Nowhere was this clearer than in the case of the US migration policy, where advocates increasingly urged that refugees be admitted outside immigration quotas on humanitarian (and political) grounds.
In 1948, the USA passed its first Displaced Persons Act despite considerable political controversy. Valid for three years, it ‘regarded refugees as objects of political concern, not simply as suffering humanity, for it focused on them as anti-communist migrants’ (Wyman 1994: 4). Although the DP Act authorized an additional 200,000 settlement places for displaced persons from Europe, it did so within a migration framework rather than applying any criteria of humanitarian need. The Act stipulated that some forty per cent of those migrating were to be selected from the Baltic population from territories annexed by the USSR. Thirty per cent were to have prior experience working in agriculture. Most seriously, fifty per cent of the numbers admitted under the Displaced Persons Act were to be ‘mortgaged’ against existing national migration quotas, tying refugee resettlement to longer-term national migration flows.
President Harry Truman had been a vocal advocate in arguing for a far more generous refugee resettlement provision, stating that ‘it is with very great reluctance that I have signed S. 2242, the Displaced Persons Act of 1948 … . It is a close question whether this bill is better or worse than no bill at all’. National legislators were reluctant to offer more than a minimal commitment to the expansion of European migration, despite the claimed humanitarian needs. The result was Truman’s assessment that the bill ‘form[s] a pattern of discrimination and intolerance wholly inconsistent with the American sense of justice’, deliberately intended to exclude Jewish and Catholic migrants (Truman 1948).
In making the case for a new act in 1952, President Harry Truman pointed out that ‘half of the Latvian quota has been mortgaged ahead three centuries to the year 2274, the Estonian quota through the year 2146, the Lithuanian quota through the year 2087, and the Polish quota through the year 2000’. Resettlement and migration had been presented as a zero-sum game, which Truman decried as ‘a most begrudging method of accepting useful and worthy people … [this] will necessarily deprive many other worthy people of an opportunity to come to the United States in future years’ (Truman 1952).
US refugee resettlement was still closely tied to broader demographic and economic concerns by both sides of the debate, with Truman arguing that a new Displaced Persons' bill was necessary to fight the serious crisis caused by ‘overpopulation in parts of Western Europe, aggravated by the flight and expulsion of people from the oppressed countries of Eastern Europe’ (Truman 1952). Poverty and displacement were still being presented as two interconnected problems, both of which could be fought through migration. Yet the difficulty was evident: there was little appetite for expanding general migration.
The result was an increasing use of politics—and by extension, the language of ‘humanitarianism’—to frame migration from Eastern Europe. Truman depicted the Eastern European displaced as the ‘innocent and unhappy victims of communist oppression’. The politics of humanitarian rescue were conflated with the geo-politics of the West’s struggle against the East:
The miserable conditions in which these fugitives from communism find themselves, and their present inability to emigrate to new homes and start new lives, lead inevitably to despair. Their disillusionment is being effectively exploited by communist propaganda. These men and women are friends of freedom (Truman 1952).
The marriage of humanitarian virtue and political self-interest meant that Eastern European refugees were no longer being presented simply as impoverished economic migrants. This was a successful strategy: the Displaced Persons Acts which followed the 1948 Act were considerably more expansive than the first. Yet it also signalled the beginning of a definitive shift away from viewing refugees as migrants. Now—above all in the US context—refugees were above all victims of communism, making a (politically acceptable) claim to entry based on humanitarian need rather than economic or demographic need.
Separating refugees from migrants was also a strategy pursued by the newly established UNHCR in the 1950s. This was motivated not so much by a desire to circumvent growing national resistance to general migration, but because it was increasingly evident that in focusing on resettling the economically able, other refugees—the old, sick, and disabled—were being left behind, trapped into displacement. As a consequence, UNHCR made several appeals for refugee resettlement to be approached as a humanitarian and not a migration or ‘development’ act. As August Lindt, High Commissioner 1956–60, argued:
A new residue is developing, because of a continuous skimming process. The most skilled and most employable are being taken first, while the less skilled and less able-bodied are being rejected by the selection missions of the immigration countries (Lindt 1959).
Considerable energy was invested by UNHCR into trying to encourage refugee resettlement to be seen not only in terms of migration and labour recruitment, but—in order to include and even prioritise the vulnerable and the ‘unemigrable’—to focus on refugees as objects of humanitarian concern first, and only then consider economic value.
There was thus no single moment at which the identity of ‘refugee’ and ‘migrant’ were finally and decisively separated into separate groups by international policy makers. By 1960, the refugee had been shaped into a figure whose admission to the West was increasingly framed by political-humanitarian interests. Although it remained important to stress the economic contributions refugees would make once resettled, the criteria for their admission was no longer economic (Lindt 1960). As ‘refugee’ became a bureaucratic category and humanitarian management expanded, this had a significant impact on refugees’ independent ability to migrate beyond camp settings, as had been actively encouraged in the 1920s.
5. Conclusions: reconnecting refugees with migration
Refugees are clearly not ‘just’ migrants. Though legal emphasis on persecution as the qualifying criteria for refugee status may obscure the role that poverty plays in prompting forced movement, a refugee’s claim to freedom of movement goes beyond development to encompass humanitarian motives. A refugee—in the broad terms that were used before the 1951 definition, and which arguably remain relevant today—can not return ‘home’ or access rights guaranteed by a meaningful state–citizen relationship. Their plea for asylum—admission across a border as protection against persecution—is distinctively a humanitarian claim and should be recognized as such. The 1930s—when refugees were treated as migrants during a period of economic and nationalist retrenchment—provides ample evidence of the catastrophic consequences such an approach. The achievement of the 1951 refugee protection framework was to provide a universal humanitarian basis upon which asylum—basic and immediate security and relief from persecution—was to be granted to many of the displaced.
Yet if a refugee's claim to asylum is a moral claim that is intended to separate refugees (who need admission) from migrants (who want admission) (Walzer 1983: 48), long-term protection and ‘solving’ refugee crises often depends on refugees securing access to existing migration channels. This was formally recognized by the international protection framework in place between 1920 and 1950. This also fits Soguk’s (1999) claim that passports for refugees both ensured that the displaced remained outside any specific normal citizenship (i.e. in a host country), while simultaneously providing sufficient inclusion for refugees within the broader national-states system to ensure the continued viability of this newly nationalized rights framework. In other words, although refugees were not ordinary migrants, they would find economic self-sufficiency and national protection by being assisted to join existing migration processes.
Contemporary approaches to formal refugee protection often see movement as fundamentally at odds with the provision of humanitarian assistance. This trend is evident in states’ continuing use of encampment to restrict freedom of movement: UNHCR staff have also at times suggested that aid is easier to deliver in a camp setting (see e.g. Harrell-Bond 1986, 2002). Convention Travel Documents—the successor to Nansen’s passports—are seriously underused (UNHCR sources 2011). Refugees' protection has become a sedentary pursuit, with refugees expected to remain in fixed locations—camps in countries of first asylum—pending an eventual repatriation. By 2009, 42 million uprooted people were ‘just waiting to go home’ (UNHCR 2009).
Yet what is very clear is that migration—movement and labour—remains a de facto choice many refugees make in seeking their own solutions to exile. There is a compelling argument that migration has a crucial role to play in development, particularly in fragile post-conflict states where infrastructure is weak and state capacity to provide adequate socio-economic services very limited. The problem therefore arguably lies in the fact that formal international interventions tend to fixate on humanitarian relief rather than development and movement-based solutions. In separating refugee and migrant identities, and assigning one a humanitarian and one a development moniker, displacement has been largely absent from the migration-development discourses of the past decade.
Looking back to the period between 1920 and 1950—when refugees were very much treated as a special category of migrant—offers us some important insights into the flawed construction of the contemporary refugee and migrant regimes as well as reminding us of the possibility of an alternative framework. Many of the challenges faced by refugee advocates in these early decades are the common complaints of those working in refugee protection today: the ‘returnability’ issue; the closing down of migrant and by extension asylum space; public hostility in host communities; and state reluctance to move from financing relief to finding solutions to displacement, despite the additional costs involved in care and maintenance. Yet in marrying displacement to development, in seeking to protect refugees’ freedom of movement and in acknowledging the links between resolving poverty and fighting persecution, the era of the Nansen passport arguably offers an alternative vision of longer-term refugee protection centred not on relief but on freedom and development that remains relevant today.
The fundamental obstacle to resolving displacement crises remains—as it was in the 1920–50 period—the concern of states to protect national interests and respond to their domestic constituents first. Just as exclusion from nation–states' membership creates refugees, nation–state strangleholds on migration and entry into membership play an important role in preventing displacement from being resolved. An artificial separation between ‘refugee’ and ‘migrant’ ostensibly removes the humanitarian imperative for states to admit the needy in all but a minority of cases, opening up opportunities to restrict migration. This has also created a climate in which although refugee asylum—at least in terms of non-forcible return—is broadly respected, refugee solutions cannot be secured.
This is why the fate of refugee protection has always been inextricably tied to migration policies. Today, 7.2 million refugees remain in protracted exile, defined as a crisis lasting more than five years, although the average displacement is now approaching twenty years—with ‘no solution in sight’ (UNHCR 2011). Although resettlement is lauded as a protection tool, only one of every ten refugees assessed by UNHCR as in need of resettlement is actually allocated a resettlement place. There are almost no alternative migration routes available for those refugees who might benefit in economic terms, but do not fulfil protection–resettlement criteria. For refugee advocates at least, this should surely provide compelling motivation to reconsider the mantra that ‘refugees are not migrants’.
The 1930s provide a powerful warning that a migration system alone cannot protect the persecuted. The 1951 refugee regime established a basic right to protection secured through a process of exceptional and non-socioeconomic admission. Yet this ability to cross-borders, secured by separating refugee and migrant categories, has left refugees frequently unable to move freely in order to secure the livelihoods that form an essential part of long-term protection and sustainable solutions.
Understanding that refugees were once considered as migrants—and that this had positive as well as negative impacts—and tracing the process by which these two identities began to separate provides an important foundation for this. This does not mean abandoning refugees’ access to an exceptional space for admission but it does mean recognizing the limits of a separate ‘refugee’ identity, particularly in terms of crafting pragmatic and accessible solutions in long-term exile. Far from a radical new departure, scholars, policy-makers, and practitioners would do well to look also to the past and consider how refugees—especially those trapped in protracted displacement—might be reconnected to broader migration processes. Studying the complex migration–refugee relationships that emerged during this period certainly suggests that facilitating refugees’ freedom of movement and prioritizing the economic, developmental, and demographic impacts of displacement are important avenues to (re)explore in the contemporary search for durable solutions to displacement.
My thanks to Jérôme Elie and Oliver Bakewell for their thoughtful comments on an earlier draft of this paper, and to three anonymous reviewers for their helpful suggestions for revision.
Conflict of interest statement. None declared.