Realising the Right of Return: Refugees’ Roles in Localising Norms and Socialising UNHCR

ABSTRACT

Refugee repatriation rests on a fundamental norm: the right of return. The Office of the UN High Commissioner for Refugees (UNHCR) styles itself as the chief defender of international norms on refugees, educating refugees about their rights and socialising states to respect the cardinal rules of the refugee regime. However, UNHCR initially devoted little attention to the right of return. When UNHCR began to engage more actively with repatriation in the 1980s, it had only a nascent institutional conceptualisation of the right of return: it framed the implementation of the right of return as a non-political, humanitarian undertaking, and assumed sovereign states had largely unfettered discretion in its implementation. Drawing on extensive material from the UNHCR archives on repatriation movements from Honduras to El Salvador in the 1980s, this article examines how refugees themselves have influenced the governance of return by serving as norm entrepreneurs, localising the right of return and socialising UNHCR to rethink and support broader interpretations of this principle. It analyzes how Salvadoran refugees envisioned the right of return as a collective and deeply political process of asserting citizenship claims, and took direct action to implement this right, compelling UNHCR and government actors to adjust to their vision. These experiences have important implications for understandings of the right of return as an international norm, and the roles of refugees themselves as actors in norm localisation and socialisation processes. Reinforcing and expanding on recent studies of how refugees actively shape aid efforts, peacebuilding and the resolution of displacement, this study highlights the significance of subaltern power in the refugee regime, showing how it can reverberate across different sites and scales to definitively influence not only the execution of the regime’s core functions but also the interpretation of the normative commitments underpinning it.

In the international refugee regime, repatriation is celebrated as the ‘preferred’ solution to displacement, and is in theory governed by a key set of principles: returns are to be voluntary, safe and dignified.¹ These norms rest on the cornerstone of the right of return. As the main international organisation charged with protecting refugees and promoting durable solutions to their displacement, the Office of the UN High Commissioner for Refugees (UNHCR) styles itself as the chief defender of international norms on refugees, and the primary authority on their interpretation and implementation (Orchard 2014, 173–202). In this capacity, UNHCR educates refugees about their rights and ‘socialises’ states to respect the cardinal rules of the refugee regime.

Although all refugees should in principle be able to exercise the right of return if they choose, UNHCR initially devoted little attention to this norm, focusing instead on non-refoulement – refugees’ essential right not to be returned to countries where they would likely encounter persecution, torture or other abuses (Goodwin-Gill and McAdam 2007: 201). In the 1980s, UNHCR began to engage more actively with repatriation, but had only a nascent institutional conceptualisation of the right of return and the obligations it entailed (Bradley 2013; Loescher 2001). At the time, UNHCR saw the right of return predominantly in terms of an individual’s right to re-enter and resume living in her country of origin, typically in her prior community. In UNHCR’s view, sovereign states had largely unfettered discretion in the implementation of this right, which the agency generally framed as a non-political, humanitarian undertaking (Long 2014, 54). Through the lens of the mobilisation of thousands of refugees who repatriated from Honduras to El Salvador in the mid-to-late 1980s, this article examines how refugees themselves have influenced the governance of return as norm entrepreneurs who socialised UNHCR to rethink and support broader interpretations of the right of return.² Salvadoran refugees ‘localised’ this norm by developing their own conceptualisation of the right of return as a collective and deeply political process of asserting citizenship rights, one that was not necessarily tied to particular places of origin. They strategically built on codified law and leveraged their experiences as refugees to lend political and moral weight to their interpretation. They also demonstrated their ability to take direct action to implement this right (even at great personal risk), notwithstanding government opposition, by planning massive repatriation movements that compelled UNHCR and the governments of El Salvador and Honduras to adjust, however grudgingly, to their vision.³

In developing this account, I draw on extensive research in the UNHCR archives, focusing on the period of 1983–1989. The archives represent a rich source for this analysis as they contain extensive material prepared by Salvadoran refugee groups, including declarations, manifestos, correspondence and publicity, which provide insight into how displaced Salvadorans interpreted and took steps to realise the right of return. Importantly, the archives also contain memos, reports, cables and correspondence prepared by UNHCR and its governmental interlocutors, which shed light on how these actors reacted to the refugees’ positions.⁴ Institutional archives provide only a partial view of any period; I therefore situate and read this archival material in light of ethnographic and geopolitical accounts of this movement (e.g., Cagan and Cagan 1991; Edwards and Siebentritt 1991; Silber 2011; Todd 2010; Wood 2004).

Through their refusal to accept the narrow, ‘apolitical’ and humanitarian vision of repatriation favoured by UNHCR and the government of El Salvador, and their willingness to take direct action despite extreme insecurity to realise their right of return, Salvadoran refugees socialised UNHCR, prompting the agency to adapt its interpretation of this norm and its approach to supporting its implementation. Their actions had significant consequences for the resolution of the displacement crises stemming from the civil wars in Central America. Their experiences also have important implications for understandings of the right of return as an international norm, and the role of international organisations and refugees themselves as actors in norm localisation and socialisation processes. Reinforcing and expanding on recent studies of how refugees actively shape relief efforts, peacebuilding and the resolution of displacement (Bradley, Milner, and Peruniak 2019; Pincock, Betts, and Easton-Calabria 2020; Roepstorff 2019), this study highlights the significance of subaltern power in the refugee regime, showing how it can reverberate across different sites and scales to influence not only the execution of the regime’s core functions but also the interpretation of the normative commitments underpinning it.

I first introduce the concept of the right of return, and briefly situate the study in relation to research on refugees, repatriation and international norms, broadly understood as ‘standards of behaviour defined in terms of rights and obligations’ (Krasner 1982, 186). Second, I discuss the Salvadoran displacement crisis and analyse the localised conception of the right of return articulated by refugee leaders, and their implementation efforts.⁵ I then trace how the refugees socialised UNHCR to support new interpretations of the right of return, and close by considering the theoretical implications. My intention is not to suggest that this case is archetypal, or that the Salvadoran refugees’ interpretation of the right of return can simply be transposed to other communities. Rather, this movement offers a window into how power asymmetries in the governance of return may be challenged, and points to the need for more careful attention to the ways in which refugees imagine and affect the enactment of international norms in the refugee regime.

International Norms and the Governance of Refugee Returns

The right to seek asylum (Orchard 2014) and non-refoulement (Pirjola 2007), as well as UNHCR’s authority and persuasive power in advancing these cardinal norms, have been explored in some detail in the expanding literature on the refugee regime (Barnett and Finnemore 2004; Betts 2009). Some attention has also been devoted to norms on repatriation as a durable solution to displacement, particularly the principle that returns must be voluntary (Chimni 1993; Gerver 2018; Hathaway 1997) and unfold in ‘conditions of safety and dignity’ (Bradley 2013; Englbrecht 2004). Explicitly or implicitly, the right of return (as a matter of law and more broadly as a political and moral claim) underpins all repatriation movements, but has attracted somewhat less attention. Discussions of the right of return are perhaps most closely associated with the Palestinian refugees (see e.g. Aruri 2002, Khalidi 1992), and the UN Human Rights Committee affirms that the ‘right to return is of the utmost importance for refugees seeking voluntary repatriation’ (UN Human Rights Committee 1999, para 21; Hathaway 2021, 1181–1183). However, under international human rights law the right of return is a general principle not specific to particular refugee groups. For example, the 1948 Universal Declaration of Human Rights provides that ‘Everyone has the right to leave any country, including his own, and to return to his country’ (Article 13.2). The International Covenant on Civil and Political Rights (ICCPR) (1966) articulates the right of return in a legally-binding manner, confirming in Article 12.4 that ‘No one shall be arbitrarily deprived of the right to enter his own country’. This language was chosen in order to protect people such as refugee children born in exile who may have a right to enter a particular country (such as their parents’ state of origin) although they have never been there and so cannot ‘return’ in a straight-forward sense (Hannum 1987, 56; UN Human Rights Committee 1999). According to the UN Human Rights Committee, Article 12.4 is to be interpreted expansively, such that deprivations of the right of return can only very rarely be justified (UN Human Rights Committee 1999, para 21; Hathaway 2021, 1181–1183). As Hannum (1987, 20) argues, from the perspective of international human rights law, ‘the point of departure [for determining the legal content of the right of return] must be the widest possible scope of free movement into one’s own country … ’ Several regional human rights instruments also include important provisions on the right of return, with the 1969 American Convention on Human Rights indicating in Article 22.5 that ‘No one can be expelled from the territory of the state of which he is a national or be deprived of the right to enter it’.

The right of return is additionally addressed in international humanitarian law, and in a wide range of UN General Assembly and Security Council resolutions, peace treaties, national laws, and soft law tools such as the UN Principles on Housing and Property Restitution for Refugees and Displaced Persons (Pinheiro Principles) (Leckie 2007). These standards vary in their scope, but in contrast to provisions on the right of return in international human rights law, they often specifically focus on the return of refugees and other forced migrants, and have shaped how the right of return has been interpreted and implemented in the context of the refugee regime. Some have argued that properly interpreted, the right of return entails not only the right to re-enter one’s country of origin, but more specifically the right of those who have been uprooted to return to and reclaim their original homes and lands (see e.g. Leckie 2005, 3). However, over time the right of return has been interpreted in a range of ways. In addition to being understood as an individual right to re-enter one’s country of origin and repossess lost homes and lands, it has also for example been understood as a dimension of freedom of movement, as a form of redress for the wrong of forced migration, and as a claim for recognition as a member of the political community from which one has been exiled (Bradley 2018).

“Under international law, a refugee has the presumptive right to return home”, with the right of return serving as the fundamental norm underpinning repatriation movements (Hathaway 2021, 1179). However, in its first decades of operation UNHCR’s conception of the right of return and other norms related to repatriation were relatively under-specified. This is perhaps unsurprising as the 1951 Refugee Convention does not explicitly address the right of return; the UNHCR Statute also does not directly address the right of return.⁶ However, UNHCR’s mandate under its Statute includes ‘Assisting governmental and private efforts to promote voluntary repatriation’ (Article 8.c). In its early years UNHCR focused primarily on protecting and resettling refugees fleeing the Eastern bloc and conflicts associated with Communism, with the right of return seen as largely ‘theoretical’ (Long 2014, 77). This began to change in the early 1980s when western support for largescale resettlement began to wane, and governments pushed UNHCR to expand its involvement in returns (Loescher 2001). In 1980, the UNHCR Executive Committee (ExCom) passed its first Conclusion on Voluntary Repatriation (No. 18); strikingly for a protection-focused organisation, the Conclusion bears no mention of rights, including the right of return (UNHCR Executive Committee, 1980). Five years later, ExCom passed its second Conclusion on Voluntary Repatriation (No. 40), which opens by affirming ‘the basic rights of persons to return voluntarily to their country of origin’ (UNHCR Executive Committee, 1985). Conclusion 40 crystallises some tenets of UNHCR’s emerging conception of refugees’ right of return, stressing the voluntary, ‘individual character of repatriation’ and the ‘need for it to be carried out under conditions of absolute safety, preferably to the place of residence of the refugee in his country of origin’. Despite UNHCR’s limited operational experience with repatriation (compared to other ‘durable solutions’ such as resettlement and local integration) and its lack of detailed institutional positions on refugees’ right of return, Conclusions 18 and 40 assertively position UNHCR as the central authority alongside states in governing return movements and interpreting international norms related to repatriation. For example, while Conclusion 40 recognises the importance of ‘spontaneous returns’ – that is, movements organised by refugees themselves – it centres UNHCR’s judgements on the appropriateness of these movements, indicating that the High Commissioner should actively promote repatriation and aid for spontaneous returnees ‘whenever he deems that the prevailing circumstances are appropriate’ and ‘in the interests of the refugees concerned’.⁷

Scholarship on repatriation norms typically probes the moral and legal contours of these commitments (Bradley 2013; Chimni 1993; Gerver 2018; Hathaway 1997, 2021, 1179–1183; Goodwin-Gill and McAdam 2007: 492–497), rather than engaging key concepts from the international relations (IR) literature such as socialisation and localisation to explain their evolution and influence. However, application of the notions of socialisation and localisation may illuminate dynamics surrounding these norms; at the same time, examination of repatriation norms may help refine understanding of socialisation and localisation processes. Socialisation involves the social processes that lead to the adoption and internalisation of norms, such that they assume a ‘taken for granted’ quality (Finnemore 1993; Risse 2000). Socialisation research in IR has generally ‘been unified by a fundamental common assumption: the actor being socialized is the state’ (Efrat 2015, 648; Risse, Ropp, and Sikkink 1999). International organisations and transnational non-state actors, particularly advocacy NGOs and networks, have attracted the lion’s share of attention in the state socialisation literature, which has identified micro-processes such as social influence and persuasion through which socialisation is achieved (Goodman and Jinks 2013; Keck and Sikkink 1998).

Norm entrepreneurs play pivotal roles in introducing and (re)framing norms and inducing socialisation. Finnemore and Sikkink (1998, 896) offer an influential conception of norm entrepreneurs as “agents having strong notions about appropriate or desirable behavior in their community”. Building on this perspective, Orchard (2013, 37) helpfully defines norm entrepreneurship as “as an activity or process through which actors are willing to devote considerable resources (material and/or ideational) in order to introduce, change, or replace international norms in their areas of interest”. Traditionally, UNHCR has been assumed to be the chief norm entrepreneur in the refugee regime, and to drive the socialisation of international norms on refugees; only limited attention has been devoted to how international organisations like UNHCR are themselves affected by other norm entrepreneurs and socialised in their conception of the norms they promote (Efrat 2015; Park 2006). In particular, the possibility that subaltern or marginalised groups such as refugees may serve as norm entrepreneurs and shape the process of ‘socialising the socialisers’ has been under-examined.⁸

This imbalance has persisted despite a surge of interest in the localisation of humanitarian response and refugee assistance (Pincock, Betts, and Easton-Calabria 2020, 2021; Roepstorff 2019), and in the localisation of international norms – that is, the dynamic processes through which international norms are interpreted, reconstructed and diffused in local contexts, often resulting in new interpretations that sit between outright rejection or acceptance of the norm as articulated and promulgated by its international proponents (Acharya 2004). Localisation risks becoming a sprawling, ‘catch-all category’, as the notion of the ‘local’ is endlessly mutable (Roepstorff 2019; Zimmerman 2016, 99). This dynamic is particularly apparent in relation to the ‘localization’ of international norms on refugee protection, a process that necessarily spans multiple geographic locales and sites of political action. Just as socialisation scholars have focused on powerful institutional actors such as international organisations, NGOs and transnational advocacy networks, norm localisation researchers have mostly focused on comparatively prominent ‘local’ actors such as national authorities, regional organisations and well-connected NGOs, tending less to the contributions of communities on the brunt end of conflict, displacement and human rights violations (Kneebone 2016; Nah 2016; Zimmerman 2016).⁹ Examination of Salvadoran refugees’ actions as norm entrepreneurs, their localisation of the right of return, and their socialising influence on UNHCR, provides an opportunity to help correct these imbalances.

From Exodus to Exercising the Right to Return to El Salvador

In the 1980s, interconnected conflicts fuelled massive displacement across Central America. In El Salvador’s civil war (1979–1992), leftist revolutionary groups united under the Frente Fabarbundo Martí para la Liberación Nacional (FMLN), fought the country’s right-wing, Washington-supported government. Leftist organising in El Salvador was in part a reaction to entrenched disparities and gruelling poverty: in 1980, 1% of the population controlled 70% of arable land, while the poorest 20% had only 2% of the national income (Silber 2011, 36). During the war the FMLN controlled significant territory in the north and east of El Salvador; the armed forces used scorched earth tactics to depopulate FMLN-held territory, in an attempt to undermine the guerillas’ base of support.¹⁰ Between 1979 and 1982, some 1.5 million Salvadorans – primarily campesinos from impoverished rural communities – were displaced; 500,000 were uprooted within the country and over a million fled to the US, Mexico and Honduras. Some 21,000 refugees lived in five official camps in Honduras, mere miles from the border (García 2006, 34–35; Todd 2010, 3).

While the Honduran and Salvadoran governments and armed forces perceived the camps to be influenced if not outright controlled by the FMLN, UNHCR publicly stressed the civilian nature of the refugee population and privately lamented the FMLN’s sway in the camps (see e.g. UNHCR 1984 hon/hcr/818, UNHCR 1987c hon/hcr/0560). Beginning in 1985 and accelerating in 1987, refugee-led repopulation committees began planning large-scale returns to areas under FMLN control – a surprising development in some senses as their intended destinations were in deeply insecure warzones.¹¹ Some refugees were hesitant to return due to the prevailing insecurity, but felt pressured to do so; others were eager to return for intertwined reasons including commitment to the creation of new communities reflecting values of social equality and justice, and the desire to escape the strictures of camp life (Cagan and Cagan 1991; Edwards and Siebentritt 1991; Silber 2011; Todd 2010, 190–220). Given Salvadoran officials’ opposition to repatriation, and the risk of violence against would-be returnees, the archives clearly show that UNHCR was hesitant to support the process. Nonetheless, the refugees came together to envision and advocate for their own repatriation, and made the right of return a normative cornerstone of these efforts. The majority of the ‘massive’ return movements unfolded from 1987–1989 and concluded by the early 1990s, and significantly influenced the treatment of voluntary repatriation as a durable solution to displacement in the International Conference on Central American Refugees (CIREFCA), an influential inter-governmental process guided by UNHCR and the UN Development Programme that unfolded from 1989–1994. By the time the Chapultepec Agreement was signed in 1992 to officially end the war, the majority of those encamped in Honduras had already returned (Bradley 2011).

This is not to say that the returns were a resounding success from the perspective of any of the actors involved: many returnees were intimidated, abused or even killed, and in the ensuing decades repopulated communities have struggled in the face of continued poverty, structural inequalities, and gang-related violence, prompting renewed displacements (Brigden 2018; Silber 2004, 2011; Wood 2004). It is nonetheless remarkable that in the midst of war and despite concerted government opposition and deep UNHCR ambivalence, the refugees managed to exercise their right of return. The following sections address two elements of the refugees’ strategy: their localisation and diffusion of their own interpretation of the right of return, and direct mobilisation to realise this right.

Localising the Right of Return

Organised under the umbrella of repopulation committees, Salvadoran refugees in camps such as Mesa Grande, Colomoncagua and San Antonio explicitly articulated and disseminated their own interpretation of the right of return, and the conditions essential to its realisation. In contrast to UNHCR’s conception of repatriation as an apolitical, individualised process to be guided by the agency’s expertise and subject to the discretion of the Salvadoran authorities, the refugees envisioned the right of return as an individual but also collective right to be realised first and foremost through the political mobilisation efforts of the refugees themselves, as citizens of El Salvador (Todd 2010, 190–220). Beyond simply re-entering the country, the refugees argued that exercise of the right of return required that they be able to choose whether to go back to their prior places of residence, or resettle elsewhere in the country with other returnees (Sollis 1992, 55–56).

The refugees concertedly linked their vision to provisions on the right of return established in international human rights law and in the 1983 Constitution of El Salvador, which indicates that ‘Every person has the liberty to enter, remain in, and leave the territory of the Republic, save the limitations that the law establishes … No Salvadoran shall be expatriated, nor his entry into the Republic prohibited, nor a passport or other documents of identification for his return be denied’. (Chapter 1, Article 5). For example, the Mesa Grande repopulation committee argued, ‘We are protected by international treaties that recognize the right to freedom of movement and to freely choose the place of residence within the territory of their country … As Salvadorans we have the right to return to live and work freely in our country … We demand that the free choice of all the people in their place of settlement be respected’ (Mesa Grande Repopulation Committee 1987b). The committee further asserted that ‘We are Salvadorans and we have the internationally recognized right to return to our country under the conditions set forth in this project’ (Mesa Grande Repopulation Committee 1987b). Such communications reflect how the repopulation committees understood the implications of international law for their own lives, and leveraged international law to bolster their own aspirations for return.

Manifestos and open letters (published in Salvadoran newspapers) to prominent officials such as the President of El Salvador, the UN Secretary-General and the High Commissioner for Refugees announced intentions to return en masse, and further detailed their vision of the right of return as a collective undertaking. The refugees rallied under slogans emphasising return as a process of communal hope and struggle, such as ‘Happiness is returning in community. Let’s struggle to achieve it’ and ‘We must construct the conditions to repatriate. Let’s work together to achieve them’ (Todd 2010, 203). Similarly, pro-return publicity called out, ‘Brother refugee: Organization guarantees our security’ (Todd 2010, 203). The repopulation committees favoured collective returns as a means of increasing their security, meeting socioeconomic needs, and preserving the community structures developed over years in exile. Maximising the contribution collective returns could make to the creation of viable communities in line with the refugees’ political aspirations required some flexibility in terms of the destinations to which they ‘returned’. While many did go back to the areas where they previously resided, the repopulation committees stressed that their demand to return to their ‘place of origin’ did ‘not necessarily mean places of previous residence’; rather, ‘Places of resettlement [return] should be determined according to the results of fact finding missions’ the refugees requested (UNHCR 1989a hon/hcr/0401; Sollis 1992, 55–56).

The different elements of the refugees’ localised interpretation of the right of return and the inter-related conditions necessary for its realisation were reflected, for example, in a 1987 open letter to Salvadoran President José Napoleón Duarte from the repopulation committee of Mesa Grande camp. In this letter, the refugees recognise that ‘the reasons that prompted our departure still exist’, but that having been refugees for six years, and having lived in conditions without freedom of movement, without the means to work freely for our self-subsistence … We would like [to] express to you our decision to return to our places of origin … The minimum conditions we demand are the following:

1. That the civil population be allowed to live in their places of origin;

2. To work freely in those places;

3. That our children not be forcibly recruited;

4. That there exist no military or civil defence posts in these settlement sites;

5. That there not be raids or bombings of these places;

6. Freedom for circulation by land and water;

7. That the access of international assistance be allowed as well as the personnel working for such agencies for the sake of the development of our new communities; That there be freedom of the press and public media. [sic, UNHCR translation]

In closing, the refugees stress that they ‘are a civil[ian] population anxious for peace’ (Mesa Grande Repopulation Committee 1987a). By concertedly presenting themselves in peaceful (but political), constructive terms the refugees legitimised their conception of the right of return, while utilising their experiences of dispossession and survival to give weight to their claims. Their strategy effectively characterised the right of return as a norm that both depended on and enabled respect for related human rights principles such as the rights to freedom of movement, work, and secure enjoyment of the home.

In UNHCR’s view, the refugees’ portrayal of their ‘movement in a political light’ threatened its legitimacy and risked provoking violence (UNHCR 1988g sal/hcr/0249). In contrast, for many refugees their political interpretation of the right of return, and its connections to campesino social mobilisation traditions and their identity claims as Salvadoran citizens, was an essential source of legitimacy and security. The refugee leadership reasoned that they could better protect themselves by banding together, with the support of international solidarity activists, than by attempting to appease national authorities who viewed them with unrelenting suspicion if not outright hostility (Sollis 1992, 51; Todd 2010, 190–220; Wood 2003, 2004). Reflecting this belief, returnees publicly demanded the ‘Right to be recognized as they are: CITIZENS’ (UNHCR 1988e sal/hcr/0214). Over the course of the late 1980s, the refugees regularly reiterated and widely diffused these demands in person and in print, reaching audiences throughout the region and internationally. While the details varied somewhat, depending on the particular circumstances of the camps and communities concerned, the approach remained broadly consistent: the refugees did not generally request permission but asserted that as citizens they had decided to exercise their right to return, challenging government attempts to wield veto power over repatriation.

Mobilisation for Direct Implementation of the Right of Return

The implementation of international norms typically requires state recognition of the norm and active support for its realisation, such as through drafting legislation, funding programmes, or ensuring state agents refrain from prohibited behaviour like torture or arbitrary executions (Betts and Orchard 2014). In this case, despite constitutional and international legal provisions backstopping the refugees’ right of return, state authorities were vociferously opposed to implementation of the norm, particularly the expansive, collective interpretation embraced by the refugees. However, a constellation of factors meant the refugees could take the implementation process into their own hands in significant ways. As the camps were mere miles from the border and their final destinations, the refugees could force the issue of implementing the right of return by simply walking to the border – as they vowed to do in numerous open letters to UNHCR and the President of El Salvador. Of course, the refugees needed state cooperation and wanted UNHCR support to fully realise their vision for the right of return. Yet by publicising their plans to repatriate (with or without UNHCR assistance), connecting with international solidarity activists, and drawing media attention, the refugees asserted some control over the process, pushing UNHCR to support their demands and making overt military action to prevent returns an unpalatable option.

Reflecting humanitarinism’s entrenched paternalism (Barnett 2017), UNHCR actively tried to dissuade self-styled returns and downplay the possibility of upstart refugees attempting to repatriate on their own terms. In its interactions with would-be returnees, UNHCR (understandably) stressed the dangerous conditions en route and in return communities, while in its internal documents and démarches with the Salvadoran government, UNHCR often reframed the repatriation committees’ detailed demands as the refugees merely ‘requesting to return to their home country under certain conditions’ (Franco 1987b), UNHCR 1987a. UNHCR’s position reflected its assumption that ultimately acceding to returns was the state’s choice, and that the refugee agency and its government counterparts were the key repatriation powerbrokers. For instance, in a 7 October 1987 letter to President Duarte, UN High Commissioner for Refugees Jean-Pierre Hocké (1987) reaffirms ‘on the one hand, the right of the refugees to return, as well as the prerogative of the country of origin to adopt all the necessary measures so that the repatriation is carried out within conditions that respect the internal order of the country and security of the returnees themselves’.

Notwithstanding governmental and UNHCR authorities’ clear preference for orderly (read: controlled), individual returns, the refugees bucked attempts to dictate the terms of their return by contining to promote their localised, collective interpretation of the right of return, and laying plans to implement it by relying on their own means and their strategic connections with international solidarity groups who provided some material resources and whose presence deterred, at least to some degree, direct military violence. By organising themselves directly, ignoring the Tripartite Commission’s decree that massive returns be limited to 500 people, and publicly announcing specific dates and timelines for collective returns, on foot if necessary, the refugees increased pressure on the Salvadoran government and UNHCR to adapt and respond to at least some of their demands. UNHCR certainly perceived the refugees’ strategy as a pressure tactic: for instance, in August 1987 the head of UNHCR’s office in Honduras informed Geneva that 900 families from Mesa Grande ‘could start [to] walk towards border on October 1^st 1987. Different sources seem to confirm that this refugee decision … [the possibility that the refugees have decided to] Threaten [‘spontaneous’ repatriation] in order to press salvadorean government and UNHCR is not ruled out’ (UNHCR 1987 hon/hcr/0560, sic). While declared repatriation dates sometimes passed without action, UNHCR staff recognised that refugee-led, largescale returns could not be delayed indefinitely and sought to clarify their ability and willingness to support returns opposed by the Salvadoran government. On 21 August 1987, for instance, the High Commissioner’s Representative in Honduras told would-be returnees in Mesa Grande that ‘HCR cannot force El Salvador government to accept repatriants, but will be ready to accompany, rpt [repeat] accompany refugees along their movement from Mesa Grande to the border … [but] once at the border, UNHCR would not force the entry to El Salvador against Salvadorean government’s will’ (UNHCR 1987d hon/hcr/0603).

The refugees’ mobilisation to implement the right of return was not simply a romantic story of morally unimpeachable underdogs standing up to unbending institutions. Refugee leaders sometimes coerced dissenters to accede to their views, pushing families to repatriate despite fears of violence, or impeding departures outside the repopulation committees’ timetables or to unsanctioned destinations. In June 1988 in San Antonio camp, for example, UNHCR reported that a refugee who had indicated he wished to return immediately and individually was assaulted and detained by other refugees, and forced to declare that ‘he was repentant of his earlier statements’ (UNHCR 1988b hon/hcr/0397). Some refugee leaders saw it as their right to govern returns, and insisted that they attend UNHCR interviews with families requesting return, making it difficult if not impossible for them to freely express their preferences (UNHCR 1988a hon/hcr/0384). In UNHCR’s assessment, the ‘coordinators’ opposition against [individual and immediate] volrep is their fear of a substantial decrease in the camp population,’ which remained an important base of FMLN support (UNHCR 1988b hon/hcr/0397). While UNHCR was accustomed and even sometimes inured to the government of El Salvador impeding the right of return, it was unwilling to brook barriers created by the refugees themselves. Senior officials in Geneva argued that ‘refugee co-ordinators’ attitude totally unacceptable to UNHCR and constitutes a violation of basic refugee rights,’ and advised staff in Honduras to ‘maintain firm position regarding implementation of refugees’ right to return’ on UNHCR’s terms, and

urge refugee co-ordinators to immediately hand over to UNHCR remaining volrep candidates … Should refugee co-ordinators resist, believe situation should be restored by any legitimate means including, if you deem appropriate, intervention by public order forces at the request of national civilian authorities … (UNHCR 1988i hcr/hon/0348)

The abuse of dissenters and refugee leaders’ intolerance of return plans at odds with their own clearly posed major protection concerns demanding UNHCR action; indeed, in attempting to control the return process in this way, some refugee leaders displayed the same domineering approach that they vociferiously opposed when it came from UNHCR. However, UNHCR’s willingness to contemplate dramatically escalating the situation by calling in Honduran authorities reflects the agency’s frustration with refugees occupying positions of power in governing returns, and their desire to re-establish their own authority.

Throughout this operation, UNHCR vacillated between recognising that the refugees were driving the return process; deriding their approach as unsuitable; attempting to reassert control; and congratulating itself for enabling return as the ‘preferred’ solution. For example, UNHCR staff from field officers to senior management recognised that the returns happened because, as UNHCR’s Regional Representative for Latin America and the Caribbean (LAC), Leonardo Franco, expressed it, the refugees ‘were unwilling to negotiate their desire to return en masse’ (Franco 1987c). Yet Franco (1987c) also believed that the mass returns of October 1987 unfolded because UNHCR démarches

culminated in the decision taken by President Duarte himself on 9 October not only to permit refugees to return to their villages of origin but also to return en masse under UNHCR protection and assistance. This decision was expressly premised on the authorities’ acknowledgement of the seriousness and capability of UNHCR and out of personal respect for the High Commissioner … The enormous risks inherent in this operation were minimized by the seriousness and self-sacrifice of UNHCR staff, but were also overcome by a large dose of luck!

UNCHR’s involvement was pivotal, but such communications overlook the refugees’ own drive and determination as decisive factors in enabling implementation of the right of return. UNHCR was also often taken aback by refugee-led efforts to establish the socio-economic conditions they called for as integral elements of the right of return: in November 1987, for example, UNHCR staff reported that ‘all present witnesses’ were ‘surprised’ by ‘the dynamic activity and disciplined organization displayed by repatriates’ who ‘have cleaned up settlement areas, have constructed their provisional houses, have started vegetable gardens, have repaired adjacent roads, have installed latrines, have designed communal areas, etc’ (UNHCR 1987g sal/hcr/087, sic).

At other times UNHCR was compelled to more fully acknowledge the refugees’ place in steering the process. Debriefing a particularly contentious movement in 1988, for example, senior staff recognised that

The refugees as a community rather than a group of individual refugees took the political decision to return on a given day to given places … the refugees imposed their conditions: date, place of return, number of trucks, schedules, etc. In itself this intransigent position implied a constant challenge to both governments … From the refugee committee’s point of review it was their operation and the outside support/assistance was there only to make it easier. (Blatter 1988)

UNHCR attempted to reassert control and depoliticise the process, indicating for example in a major Geneva meeting with NGO leaders that in future, ‘Practical modalities will be determined by the UNHCR and governmental authorities, and then will be shared with the refugees’ (Del Rio 1988). Challenging UNHCR’s opposition to the refugees’ ‘politicization’ of the process, NGO leaders suggested that because ‘there is a risk that the Salvadorian military might exercise a veto to massive voluntary repatriation to El Salvador … the political mobilization of refugees might be regarded as a necessary condition to ensure that massive repatriation actually takes place’ (Del Rio 1988). Accustomed to governing refugees and interpreting norms through the lens of its own mandate, UNHCR insisted that ‘while political agitation may be legitimately regarded by certain parties as a viable mechanism to achieve given objects, it is not compatible with the humanitarian character of UNHCR mandate. Consequently, if this course of action is going to be followed in a future massive repatriation movement, UNHCR definitely will not be associated with it’ (Del Rio 1988). In the field, however, UNHCR grudgingly acknowledged the wisdom in the refugees’ strategy, with staff reflecting that the ‘logical approach’ of ‘individual (small groups) repatriations’ does ‘not however coincide with the refugee committee. Firstly having gained so much public attention through the 1987 and 1988 movements, individual repatriations will certainly not be favoured by the refugee committee. Secondly and paradoxically salgovt would not allow small groups to return to places/areas of origin in conflictive zones, i.e. only through heavy pressure of massvolrep did salgovt. accept places chosen by refugees’ (Blatter 1988).

As these communications recognise, however grudgingly, despite significant power differentials the repopulation committees compelled UNHCR and the Salvadoran government to adapt to their strategies for implementing the right of return. As I discuss in the following section, the refugees’ efforts helped socialise UNHCR to accept and promote key elements of their normative interpretation.

Socialising UNHCR

What effects did Salvadoran refugees’ efforts to localise and implement an international norm – the right of return – have on UNHCR? Conceptions of the right of return and the roles of UNHCR and refugees themselves in implementing it were deeply fraught and remain contested. Yet elements of the refugees’ approach became more widely accepted by the agency and reflected in its operations and key frameworks such as the CIREFCA Declaration and UNHCR ExCom Conclusions.

Resistance

For the most part UNHCR was initially resistant to the Salvadoran refugees’ localisation of the right of return and their determination to play an active part in its realisation as an inevitably and explicitly political process. UNHCR initially perceived large-scale, refugee-led returns as a ‘threat’ (Franco 1987a) and a potential embarassment for the Salvadoran authorities (UNHCR 1988i sal/hcr/0348), with some staffers seeming to identify more closely with the government’s desire to control the repatriation process than with the refugees’ aspirations and rights claims. For instance, despite the abduction and torture of some returnees and the strafing of repopulated communities, UNHCR officials approvingly summarised the government’s bare-bones position on the right of return: ‘Principle of right of return accepted in general terms by sal gov including armed forces. Return must be organized, gradual and only to non-conflictive areas’ (UNHCR 1987f cos/hcr/0604). In defiance of its recognition of return as a right, UNHCR joined the governments of El Salvador and Honduras in signing a 1987 Tripartite Commission communique describing the acceptance of 4,000 returnees by their own state as a ‘humanitarian act’, and applauded the president’s ‘generosity’ in allowing the operation (Tripartite Commission 1987).

At some points, UNHCR characterised the refugees’ attempts to influence decisions on issues such as return routes as ‘blackmail’, especially after ‘several concessions had been made to the refugees; the site of repatriation, date, number of movement and accompaniment’ (UNHCR 1988h). UNHCR also questioned the refugees’ interpretation of their right of return and NGOs’ support for refugee leadership in governing returns, complaining that these groups failed to appreciate the

basic difference between the ‘Protection of the refugees’ and the ‘defense of the refugees’ interests.’ Some of these interests however genuine they may be, are not necessarily linked to the existing refugee law and could even be in contradiction with the humanitarian and apolitical nature of our protection work … problems also derive from indiscriminate support of agencies’ personnel on refugee standpoints independently of their impact on the refugees’ protection and welfare in both the long and short-term. (Hocké and Franco 1988)

UNHCR was keenly aware that the Salvadoran authorities perceived it as subservient to the refugees and by extension the FMLN, with the Chief of Staff of the Salvadoran armed forces having expressed the view that UNHCR and the FMLN were the ‘same thing’ (UNHCR 1987b misc 534–4). Following a particularly conflictual collective return in August 1988, UNHCR’s Representative in Honduras lamented that the movement had been ‘a logistical success and a political failure … the operation has been de-naturated by the attitude of the refugees and manipulated by interests which were not humanitarian’ (UNHCR 1988f hon/hcr/0669). Clearly uncomfortable with refugees deviating from what the agency saw as their natural position of compliance, UNHCR tried to set clear conditions for its involvement, insisting that it could only support returns that were non-political and approved by the sovereign authorities. To this end, High Commissioner Hocké (1888a) issued a direct appeal to actors across the region in 1988, clarifying UNHCR’s ‘requirements’ for facilitating voluntary repatriation:

1. All governments concerned … must agree with the modalities of the repatriation and of UNHCR participation;

2. The operation should be carried out in accordance with the UNHCR Mandate in a strictly humanitarian and apolitical spirit;

3. The UNHCR protection mandate should be respected by all concerned.

Through such appeals, UNHCR attempted to reassert control over unruly repatriants, and the interpretation and implementation of the right of return.

Influence and Persuasion

The socialisation of UNHCR in this case pertained to how the agency interpreted and promoted the right of return as an international norm, as well as its openness to the direct involvement of refugees in governing returns, including as norm entrepreneurs. UNHCR’s socialisation was not a linear process in which the agency flipped from the positions of resistance discussed above to full acceptance of broader conceptions of the right of return, and the active participation of refugees in leading repatriation processes; nor was this the only mission that shaped UNHCR’s perspectives on these issues. Rather, over the course of years spent working on the Salvadoran displacement crisis, UNHCR staff and the agency itself were influenced by the refugees and their supporters, and persuaded to revisit and adapt elements of its institutional position and approach. UNHCR’s archives are replete with communications (some of which are discussed above) from Salvadoran refugee organisations detailing their conditions for repatriation, and notes from meetings in which displaced leaders debated the terms of their return with UNHCR staff. Agency staff were forced to grapple with these demands, parse out their normative and practical implications, and rethink their own assumptions and institutional mandate in light of the refugees’ arguments. UNHCR’s paternalistic impulses and its self-conception as the leading authority on international norms on refugees have persisted. Yet through its involvement in this operation and its legacies, more complex conceptions of the right of return, and of refugees as political agents with active roles in governing returns, started to assume a more ‘taken for granted’ (if still fraught) quality within the refugee agency.

The refugees’ influence on UNHCR’s interpretation of the right of return and their preferred means of implementation unfolded gradually through processes of contestation. At early meetings of the Tripartite Commission in 1986, the governments and UNHCR alike stressed return as an individual process, dependent on government authoritization, and only to places of origin approved by officials. Through their unflagging insistance on collective returns to places chosen by the refugees themselves (often but not necessarily places of origin), the refugees gradually convinced UNHCR to shift gears and accept the legitimacy of their interpretation and appoach, even if logistically suboptimal from UNHCR’s perspective. By 1988, UNHCR staff reflected that ‘Ideally, one could envisage group of 150 to 200 repatriates in five buses acompanied by fifteen trucks which could be handled on a regular basis by existing staff at regular intervals’ but recognised that ‘This is probably not the way in which possible candidates to repatriation to El Salvador would consider returning to their country’ (UNHCR 1988d hon/hcr/0833). In persuading UNHCR to support them, the refugees concertedly appealed to the predominant rhetoric and logics of the refugee regime, pointing out that the ‘international community recommends volrep as best solution to refugee problems’ – effectively suggesting UNHCR would be hypocritical for failing to back their repatriation ambitions (UNHCR 1989a hon/hcr/0401).

The refugees’ efforts to influence and persuade UNHCR were backstopped by a large, international network of NGOs and solidarity activists who sent hundreds of messages to UNHCR defending the refugees’ demands, and interceded with UNHCR in support of the refugees’ views. For example, pushing back against UNHCR’s attempt to quash the refugees’ political approach to returns, the representativative of an influential Catholic NGO argued that ‘it should be taken into consideration that the repatriation of Salvadorian refugees takes place amidst war. Thus, every action becomes a political expression. Furthermore, the actions of refugees are a result of the years living in closed communities in Honduras where refugees realized the power of organization’ (Juarez 1988).

UNHCR’s records from this mission underscore the socialisation of international norms as a dynamic, contentious process of grappling with new ideas and negotiating new practices. At some points, the repopulation committees tempered their own demands and strategies in light of changing political circumstances, and to gain greater traction with UNHCR, prompting the UNHCR working group on repatriation to conclude that while the refugees’ terms ‘cannot be fully shared by UNHCR, UNHCR cannot be indifferent to the substance of the matter, given the desire of a significant number of refugees to return voluntarily’ (UNHCR 1987, Rapport de la reunion de coordination). State approval for returns nonetheless remained a major sticking point: as an international organisation created by and answerable to states, UNHCR insisted it could not support returns without government approval. But as the refugees reiterated their calls for return, and decried government opposition as incompatible with international human rights norms, UNHCR leaders became convinced that the agency needed to advocate more actively in support of the right of return, arguing that ‘UNHCR cannot remain indifferent to refugee volrep request and must support right to return’ (UNHCR 1987f cos/hcr/0604), and pointing out that the refugees had waited months if not years for the government to agree to their calls for return (UNHCR 1987e hon/hcr/0685). Pressured and to some extent persuaded by the refugee leadership’s active approach to interpreting and implementing the right of return, UNHCR tried to square the circle: Hocké began to intercede more actively with the Salvadoran president and senior officals to support exercise of the right of return, suggesting that somehow the parties could have their cake and eat it by finding ‘principles that should govern voluntary repatriation operations to guarantee the right of return to the places that the refugees wish, in full respect for the sovereignty of the country of origin’ (Hocké 1988b). UNHCR lobbied Salvadoran officials to accept the return of refugees to their chosen destinations, even in conflict zones, explaining that otherwise UNHCR could not be involved as aspects of the returns would be involuntary. Following a particuarly contentious movement in 1988, Hocké attempted to assuage the incensed Salvadoran Minister of Defence, thanking him and his colleagues for their ‘attachement to and respect for international humanitarian and refugee law, and particularly the right to repatriation’ (Hocké 1988c). While this is an almost farcical description of the Salvadoran armed forces’ position, such interventions were important in terms of enabling thousands of continued returns.

This position of more active support for the right of return as a matter of protection was particularly challenging for UNHCR to adopt in this case as the refugees wanted to return to war zones in which their safety was highly uncertain. In light of the refugees’ full-throated support for return, UNHCR concluded that it needed to show greater respect for refugees’ choices, despite their riskiness, an institutional shift that challenged the agency’s typical paternalist posture. As Franco expressed it,

in cases where the conditions of safety and rehabilitation of returnees are not entirely fulfilled, but when refugees insist on being repatriated … UNHCR’s responsibility is always to do the best for refugees wishing to repatriate, even under very difficult circumstances, provided: a) We are absolutely convinced that they return of their own free will; and b) they are well informed of the limits of UNHCR action on their behalf in the country of origin. (Franco 1987a)

UNHCR’s embrace of a more supportive if still guarded position was also motivated by recognition of the refugees’ determination and ability to take direct action to implement the right of return, as discussed above. As a senior official conceded, ‘whatever salgov’s final position,’ or UNHCR’s for that matter, the ‘refugees will return and will in no time organize/mobilize international solidarity groups/churches, NGOs, parliamentarians, mass media, etc’ (UNHCR 1988c sal/hcr/0162).

UNHCR’s support for the resolution of the Salvadoran displacement crisis, including through collective returns, grew stronger in the late 1980s as it sought out a ‘new and comprehensive strategy’ for the region (Franco 1988). Again trying to square the circle of providing non-political, humanitarian support for an undeniably political population and operation, UNHCR settled on the compromise view that ‘In protecting and assisting refugees UNHCR must be non-political; in other words be neutral and not take sides. Any bias in its work can and should only be in favour of the rights of refugees’ (Franco 1988). In its key strategy document for this new approach, UNHCR was clearly cognisant of and accepted key elements of the Salvadoran refugees’ views on return, and stressed the need to consult with them, especially in ‘respect of individual rights, particularly the right to return’ (Franco 1988). UNHCR’s strengthened support for the refugees’ repatriation was challenged by the 1989 election of the ultra-right wing ARENA party in El Salvador, but in its rapproachements with the new government UNHCR delicately supported the refugees’ desire for collective returns, including to communities in conflict-ridden areas (Bradley 2011, UNHCR 1989a hon/hcr/0401).

UNHCR’s efforts to advance a new regional strategy culminated in the much-lauded, UNHCR-spearheaded CIREFCA process, which ran from May 1989–1994 with the goal of supporting durable solutions to Central America’s displacement crisis as a critical element of the region’s nascent peace processes (Betts 2006, 2009). The CIREFCA Declaration and Plan of Action were signed by governments across the region and distilled protection principles for voluntary repatriation processes. This text, shepherded by UNHCR, reflects the influence the Salvadoran refugees had on ideas about repatriation, not only in their own case but on UNHCR and across the region. For instance, the Salvadoran refugees’ approach directly inspired the more well-known repatriation efforts of Guatemalan refugees encamped in Mexico, who emulated some of the Salvadoran refugees’ organising strategies and their conceputalization of the right of return as a collective, deeply political process of reasserting citizenship rights and reshaping the state (Krznaric 1997, 71; Bradley 2014, 121). The CIREFCA Declaration and Plan of Action also reflect the limits of the refugees’ influence: for example, the Declaration ‘recognize[s] the crucial role played by the tripartite commissions’, which excluded direct refugee participation, and preserves some original elements of UNHCR’s interpretation of the right of return, such as its inistance on support for returnees as a ‘non-political’ undertaking. However, the Declaration and Plan of Action nuance these provisions in ways that reflect the socialising effects of the refugees’ advocacy, and the influence UNHCR’s experiences with highly mobilised refugee groups in the region had on its interpretation of international norms on refugee returns. For instance, the document insists that solutions must be pursued ‘on a strictly humanitarian basis, with the agreement and respect of all parties concerned’ – a frame which reflects UNHCR’s discomfort with refugees’ political activities and seems to uphold the state’s prerogative to quash undesired repatriation efforts. Yet the agreement tempers the extent to which the state of origin can (in principle at least) deny the repatriation of exiled citizens by stressing the ‘obligation to respect, in the first place, the right of refugees to return voluntarily to their countries of origin in order to resume a normal life’.¹² The ‘Programmes in favour of returnees’ mapped out in the Plan of Action were to be based on respect for principles closely mirroring many of the demands made by the Repatriation Committees since the mid-1980s, including the need for voluntary and safe returns; freedom of movement and in determining where to settle; non-discrimination; equitable access to livelihoods and land; and information on conditions in return communities. The Plan of Action insists on the ‘voluntary and individually-manifested character of repatriation’, a provision that may seem to be at odds with the Salvadoran refugees’ vision for collective returns. However, as the document also insists on the refugees’ right to freely choose their destinations and places of residence, it in fact accommodates their localised vision for realising the right of return.

The repopulation committees objected that they were not represented at the CIREFCA negotiating table (UNHCR 1989b sal/cos/hcr/0131), yet their interpretation of the right of return and their demands as repatriating citizens, conveyed to UNHCR over years of advocacy, clearly influenced the CIREFCA process. To be sure, it is important not to overstate the case: the refugees’ influence was limited in important ways, and contestation over repatriation norms and the implementation of returns continued, both in the region and internationally. Socialisation does not necessarily imply an end to normative contestation; in this case, the refugees’ efforts did not end debates on the right of return. Rather, their socialising influence was evident in how they shifted the terms of the debate and the assumptions UNHCR carries into it. Many of these concepts and commitments were pathbreaking and highly contested when the refugees first raised them in the mid-1980s, yet came to be more widely expected and accepted by UNHCR staff involved in repatriation operations, particularly in Central America. While it is difficult to definitively trace the broader implications of this case across regions and over time, given the diverse influences on UNHCR policies and practices at the global level, it is striking that many of the perspectives advocated by the Salvadoran refugee leadership were integrated into subsequent UNHCR ExCom Conclusions. In particular, many Conclusions adopted after the CIREFCA process emphasise that refugees’ right of return entails the state of origin’s obligation to admit and reintegrate their repatriating nationals (e.g. Conclusion No. 74–1994, No. 77–1995, No. 79–1996, No. 81–1997, No. 85–1998, No. 96–2003, No. 101–2004). Acceptance of the state of origin’s obligation to accept returning citizens was hard won in the case of Salvadorans returning from Honduras in the 1980s. In a bitter irony, however, it is now applied not so much to support the claims of refugees eager to return but to enable and legitimise repatriations carried out under pressure as host states look to offload unwanted populations.¹³

A Pyrrhic Victory? Implications for the Governance of Return and Recognising Refugees as Norm Entrepreneurs

The collective repatriation of thousands of refugees from camps in Honduras to their homes in El Salvador was a remarkable but Pyrrhic victory. Through their localisation efforts, the refugees imbued the right of return with distinct and powerful meaning in their struggle for a more equitable and peaceful El Salvador. Through their implementation efforts, Salvadoran refugees in Honduran camps proved themselves to be key players in the governance of return, who challenged UNHCR’s assumed authority as the guardian of international norms in the refugee regime, and helped socialise the agency into new ways of thinking about the right of return and refugees as agents in repatriation movements. Yet these achievements came at great cost: returnees were subject to often extreme state violence, and despite their idealistic rhetoric the returnees were in many ways unable to upend the profound inequalities structuring Salvadoran society. In the face of grinding poverty, disillusionment and gang-related violence, many returnees and their descendants have endured repeated displacement (Brigden 2018; Silber 2011). In this sense the case is a testament to the power of refugees’ agency, but also to the profound structural barriers limiting it.

More than 30 years on, this operation may now seem almost anachronous as many contemporary refugee populations struggle to resist repatriation rather than actively claim the right of return. What then is the enduring significance of this experience, practically and theoretically? Unfolding just as UNHCR was amplifying its focus on repatriation, refugees involved in these movements had outsized influence on local as well as international and institutional imaginaries of the right of return as the cornerstone of repatriation movements, and the role of refugees and UNHCR alike in advancing this right. Through years-long processes of persuasion and pressure the refugee leadership socialised UNHCR to accept and defend key elements of their normative interpretation, and provide practical support for their implementation efforts. Salvadoran refugees in Honduras were not always unified on this issue, and were not the only actors involved in shaping UNHCR’s stance on the right of return in this period. Yet this movement inspired return claims made by other refugee populations, and influenced the approach to the right of return particularly in the CIREFCA process, which in turn shaped UNHCR’s efforts to support voluntary repatriation in other regions. Its implications thus reverberate across the daily lives of thousands of returnees and their descendants, and the evolution of UNHCR as the central international institution in the refugee regime.

Much of the recent critical geopolitics literature on mobility draws attention to the ways in which displaced populations subvert state efforts to manage migration, often through clandestine movements (Hyndman 2012; Mainwaring and Brigden 2016), social mobilisation and de-nationalised protest (Ataç, Rygiel, and Stierl 2016; Ataç and Schwenken 2021) and the destabilisation of states’ efforts to ‘make and unmake’ refugees through processes of labelling and categorisation (Myadar and Dempsey 2021), typically in the context of contemporary struggles to move from states in the Global South to the Global North. This historical examination complements such work by also accentuating the significance of overt, refugee-led efforts to overcome the refugee label and relegation to statelessness through repatriation and the reclamation of citizenship rights within the Global South. Theoretically, this case suggests that as the emerging body of research on localisation in the refugee regime continues to grow, it is essential to consider not only how refugees mobilise to address their own assistance and protection concerns, but also how they serve as norm entrepreneurs who shape the ways in which key principles are understood. While international organisations are often assumed to be the drivers of norm socialisation processes, this case underscores that international organisations’ identities and commitments are not static, and demonstrates the importance of considering how international organisations like UNHCR may also be on the receiving end of socialisation processes driven not only by commonly acknowledged brokers like NGOs and transnational advocacy networks (Park 2006), but also by ‘subaltern’ actors such as refugees. It also illuminates some of the tactics displaced groups and other marginalised actors may deploy in norm localisation, implementation and socialisation efforts to limit and even leverage their subaltern position, such as cultivating international networks, connecting localised norm interpretations to widely recognised principles of international law, and drawing on experiences of dispossession and violence to legitimate localised conceptions of international norms.¹⁴ In this way it partially decentres the state and UNHCR in accounts of the influence of international norms in the refugee regime and the governance of return.

Acknowledgement

I would like to thank Ava Occhialini for her excellent research assistance, and the UNHCR archivists for their help. I am also grateful to Zeynep Sahin, Derya Ozkul and Tamirace Fakhoury, Arddun Arwyn, Laura Madokoro and members of the Doing Refugee History seminar for their feedback.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Social Sciences and Humanities Research Council of Canada .

Notes

1. I use ‘repatriation’ and ‘return’ as synonyms.

2. In examining socialisation, I am interested broadly in social processes that shape the interpretation, adoption and internalisation of norms, which influence behaviour (Efrat 2015; Finnemore and Sikkink 1998).

3. My analysis focuses on the refugees’ engagements with UNHCR and the government of El Salvador as these institutions were the primary targets of the refugees’ mobilisation efforts regarding the right of return.

4. Spanish and French archival materials have been translated by the author. All archival material is drawn from UNHCR Archives Classified Subject Files, Fonds 11, Records of the Central Registry, Series 3, Box 1108, 610.HON/SAL-Special Protection Problems-Repatriation. Volume numbers are indicated in the reference list.

5. In analysing the roles of refugee leaders, I consider those in formal leadership positions in refugee-led organisations such as the repopulation committees, as well as those who exercised leadership more informally, such as in the context of discussions between UNHCR and camp residents.

6. The protections afforded under the 1951 Convention hinge on refugees’ right not to be returned (the right of non-refoulement, see Article 33), but the Convention also indirectly addresses repatriation in Article 1 c, which indicates that refugee status ceases if an individual has ‘voluntarily re-established himself in the country which he left or outside which he remained owing to fear of persecution.’ The 1969 Organization for African Unity (OAU) Convention Governing the Specific Aspects of Refugee Problems in Africa effectively recognises refugees’ right of return in Article V(2), which states that “The country of asylum, in collaboration with the country of origin, shall make adequate arrangements for the safe return of refugees who request repatriation.” However, there is little indication that this legal development significantly shifted or deepened UNHCR’s perspectives on or approach to refugees’ right of return, particularly outside of Africa.

7. As Goodwin-Gill and McAdam (2007, 495–6) write that it ‘often does not matter what UNHCR, NGOs, or even States want; if refugees themselves choose to return, so they will, even to situations that outsiders consider highly insecure and undesirable.’ However, UNHCR has often been loath to accept the centrality of refugees’ own decision-making in return processes, particularly when refugees’ assessments and decisions conflict with UNHCR’s views and interests – such as when refugees want to return against UNHCR advice, or when they oppose return despite UNHCR indications that conditions are suitable. Legal analyses of UNHCR’s engagement in voluntary repatriation have often focused on the conditions in which the agency should actively promote repatriation, or merely facilitate it. See e.g. Goodwin-Gill and McAdam (2007, 494–495).

8. Orchard (2019, 23; 2014, 35) argues against attempts to identify a ‘discrete set of actors’ who operate as norm entrepreneurs, contending instead that it is more fruitful to focus on norm entrepreneurship as an activity that a wide range of players may engage in, as long as they have the requisite resources, authority, legitimacy, connections or other means of attracting support. While a focus on norm entrepreneurship as a process is indeed useful, it remains important to critically examine the sets of actors who are assumed to occupy and be excluded from these roles, in order to counter the elite bias that is typically built into studies of international norms. To be sure, refugee populations are highly diverse and internally stratified, and include their own elites. In general, however, refugees occupy a liminal position in the state system and experience major rights violations; they may thus be broadly understood as subaltern actors in the socialisation and implementation of international norms.

9. For important exceptions, see e.g. Kauffman (2017), McConnachie (2014). Harley (2021) provides a significant discussion from a legal and historical perspective of how refugees from relatively elite backgrounds shaped the development of early international refugee law and policy.

10. El Salvador’s truth commission found that the FMLN committed 5% of abuses during the war, right-wing paramilitaries and death squads 25%, and the armed forces 60%.

11. On the organisation and dynamics of the repopulation committees and their roots in campesino movements, see e.g. Todd (2010).

12. In this way, the approach to the right of return adopted in the CIREFCA process and by UNHCR in many subsequent repatriation operations more squarely reflects international human rights law standards, such as ICCPR Article 12(4), which allow only very limited grounds for legitimate state deprivation of the right of return.

13. On coerced returns and repatriations that fall short of standards on voluntariness, see e.g. Webber (2011).

14. Refugees may apply such tactics to in serving as norm entrepreneurs in relation to other principles, such as the administration of justice in camps. See e.g. McConnachie (2014).