Part III - Public Territories and Private Borders: Tracing Transnational Power Relations

10 From the Colony to the Border The Lawful Lawlessness of Racial Violence

Ayten Gündoğdu

Border governance has come to operate as a regime of terror, death, and disappearance with the adoption of ever more restrictive policies and technologies of immigration control. Over 67,000 migrants have died or disappeared as they tried to cross borders since 2014, according to a conservative estimate, as of July 2024 (International Organization for Migration [IOM], 2024). Overwhelming evidence points to a close connection between migrant deaths and disappearances and border control policies that push migrants to ever more dangerous routes (Callamard, 2017; Shatz & Branco, 2019). Among the policies that systematically give rise to death and disappearance is what scholars have described as “extraterritorialization” of migration control, which includes a wide range of practices, including the removal of certain territories from jurisdiction for immigration purposes, interstate agreements to “outsource” border control to countries of origin and transit, and maritime interdiction of migrants to prevent their entry into territory.

As many critics have underscored, such policies aim to manipulate and dissolve the close connection between territory and jurisdiction that is central to modern conceptions of law and politics (Benhabib, 2020; Gammeltoft-Hansen, 2011; Gündoğdu, 2018; Shachar, 2020b). With its claim to exclusive jurisdiction within its territory, the modern state is supposed to extend its legal framework to all of its inhabitants (Arendt, [1951] 1968: 230; Brubaker, 1992: 25). In this regard, modern territorial jurisdiction stands in distinction from medieval common law that organized jurisdiction around status and allocated rights and obligations accordingly (Dorsett & McVeigh, 2012: 83). A corollary of territorial jurisdiction within the migration context is that a migrant’s entry into the territory of a state triggers state obligations under domestic and international laws, including protections against refoulement. However, states have strategically manipulated the principle of territorial jurisdiction to evade such obligations and adopted border control policies that amount to a politics of nonentrée (Hathaway, 1992; Hathaway & Gammeltoft-Hansen, 2015). Extraterritorialization allows the states to disavow responsibility for the consequences of their migration policies by “delinking … the bond, between territory, jurisdiction, and the public in whose name and with whose authorization law and coercion are presumably exercised” (Benhabib, 2020: 88). This delinking places migrants in what Hannah Arendt ([1951] 1968) called “a condition of rightlessness,” denying them access to territory, personhood before the law, and the right to have rights (Gündoğdu, 2015, 2018). Extraterritorialization is also problematic because it reintroduces status-based discriminations in conflict with the modern principle of equality before the law. As “the shifting border” is unmoored from the geographical frontiers of each state, it is ‘transported’ to the bodies of migrants to the effect of “barring certain bodies and populations from territorial arrival at the shores of well-off countries” (Shachar, 2020b: 37, 33).

In this chapter, I propose rethinking policies of extraterritorialization as one key element in a contemporary constellation that can be best understood in relation to other projects of racial domination such as colonial rule. In that, my analysis joins the recent efforts to rethink migration in relation to colonialism and empire (e.g., Achiume, 2019; El-Enany, 2020; Gündoğdu, 2022; Mawani, 2018; Mayblin, 2018; Mayblin & Turner, 2021; Mongia, 2018; Reynolds, 2021), but it makes a distinctive intervention by examining some of the striking, yet unexplored, resemblances between the juridical formulas and techniques employed in colonial and migratory contexts. These include, among other things, strategic manipulation of jurisdiction, a status-based legal system, racialization of status categories, normalization of a state of exception, and racialized determinations of culpability.¹

Law plays a crucial role in crystallizing these elements into racial domination – under colonial rule and within contemporary border governance. In fact, the juridical regime characteristic of both of these constellations can be best described as one of “lawful lawlessness,” to borrow a term introduced by Austin Sarat and Nasser Hussain (2004) in a different context. On the basis of a comparative analysis of colonial and migratory juridical regimes, I argue that policies that continuously manipulate jurisdiction to create a condition of rightlessness are neither exceptional nor unprecedented; they have long been a key weapon in the arsenal of racial domination. In making this argument, my goal is not to draw a direct, causal link from colonialism to contemporary migration control but rather to identify certain shared elements between these two contexts to understand how and why the lines between lawfulness and lawlessness are blurred in regimes where law is put in the service of race rule.

The chapter develops this argument as follows: First, I outline the peculiar juridical maneuvers colonialism introduces to reconcile racial domination with the principle of equality before the law championed by colonial powers such as France and Britain. In Section 2, I discuss the Spanish enclaves of Ceuta and Melilla, the EU’s only land borders with Africa, to point out the ways in which the colonial juridical techniques are reconfigured within the migration context. In Section 3, I examine how law becomes complicit in maintaining borders as regimes of racial domination by turning to the 2020 ruling of the European Court of Human Rights (ECtHR) in N.D. and N.T. v. Spain, which in effect condoned the Spanish pushback operations in Melilla. By way of a conclusion, to address the question of whether law can have any emancipatory possibilities within this context, I briefly point to a diverse range of struggles that seize hold of legal discourses in inventive ways – inside and outside courts – to resist the transformation of borders into zones of death and disappearance.

1 Lawful Lawlessness under Colonialism

Colonial orders could not have been established and maintained without the use of force, but neither could they have been established and maintained solely on the basis of force. Law was crucial to racial domination under colonialism, not only as a justificatory framework authorizing violent practices, but also as a productive mechanism that created the jurisdictions and statuses used for the differential allocation of rights, privileges, and penalties. That allocation was racial through and through – interweaving phenotypical and biological conceptions of race with sociocultural assumptions about “civilizational” differences (Saada, 2005). In the colonies, “the rule of law,” stripped off its normative connotations and mobilized in the service of racial domination, categorized humanity into different juridical statuses and replaced equality before the law with “the rule of colonial difference” (Chatterjee, 1993).

What arose as a result was a system that continuously blurred the line between lawfulness and lawlessness: The decrees and codes used to entrench racial domination under colonialism were so arbitrary, indeterminate, and unstable that some would hesitate to place them under the rubric of “the rule of law,” given the association of the term with normative expectations such as impartiality, clarity, and predictability (Mann, 2009). “The rule of law,” however, has remained an ambiguous term since its origins in nineteenth-century German jurisprudence, signifying not only a normative framework constraining state power, but also an instrument of state domination (Saada, 2002: 104). It was this latter meaning that prevailed in the colonies where a certain form of race rule “exploded … the alternative between lawful and lawless government, between arbitrary and legitimate power,” to adopt Arendt’s ([1951] 1968: 461) characterization of totalitarian regimes.

Such juridical orders embody a form of “lawful lawlessness,” to borrow a term introduced by Austin Sarat and Nasser Hussain (2004) in their analysis of executive clemency. My use of the term differs from theirs to the extent that Sarat and Hussain are primarily interested in the occasional discretionary exercise of sovereign power in a constitutional democracy that otherwise rests on a legal system of predictable rules. I use the term instead to describe a juridical order in its entirety – and not sporadic acts of discretion. “Lawful lawlessness” becomes the norm rather than the exception in systems where law is placed in the service of racial domination, as I discuss below by drawing on the literatures on French and British colonialism.

One key element of colonial juridical orders was the manipulation of jurisdiction, which allowed European powers to cover up the contradictions arising from their commitments to equality before the law and their routinized violation of this principle in their colonies – a strategy that bears striking resemblances to contemporary efforts to evade jurisdiction in migration control. Assertion of jurisdiction was essential to establishing the authority of colonial laws (Dorsett & McVeigh, 2007: 4–5), but that authority was maintained by deploying multiple, and often conflicting, concepts of jurisdiction to justify racial domination (Esmeir, 2012: 15; Mawani, 2015: 421, 2018: 24–25; Pasternak, 2014; Saada, 2002: 103). In the case of French colonialism, for example, the French state worked with a territorial conception of jurisdiction to assert its sovereignty over the colonial territory, but it also simultaneously marked that territory as an exceptional space where French laws did not apply. This territorial exceptionality went hand in hand with another notion of jurisdiction based on status (Blévis, 2014; Le Cour Grandmaison, 2006, 2010; Saada, 2002, 2005): While the French Constitution did not apply to French Algeria, the French “settlers” continued to maintain their rights under that Constitution. Those same rights could not be extended to the “natives” (indigènes), however, who were “French subjects, or peoples under French protection or administration, and not French citizens” (Henry Solus quoted in Le Cour Grandmaison, 2006: 39; emphasis in original).

Status distinctions within colonial juridical orders were based on race – a problem that reappears within contemporary immigration policies and laws. In the French context, le code de l’indigénat, the continuously changing ensemble of laws that subjected colonized populations to an arbitrary rule by decrees, established the status of indigènes on a racial basis, working with the assumption “to each race its law and to each law its race” (Saada, 2002: 110). Within the British context, similarly, race was “the most obvious marker of colonial difference,” which rendered it impossible to “[administer] an impersonal, nonarbitrary system of law” (Chatterjee, 1993: 20; see also Hussain, 2003; Kolsky, 2010). Even when there were attempts to codify a unified set of laws that would apply equally to all the inhabitants of the colony, the failure of those attempts was due to the tenaciousness of the idea of racial difference. In fact, such attempts, as illustrated by the Code of Criminal Procedure established by Britain in colonial India in 1861, “expanded legal distinctions, exceptions, and inequalities” and exacerbated the problem of racial inequality before the law (Kolsky, 2010: 73).

The construction of racial difference in and through law was central to instituting racial segregation under colonialism (Saada, 2002: 105), as illustrated by the subjection of the movement of the colonized to an elaborate system of controls, which, in many ways, is the precursor to contemporary migration controls. In the Belgian Congo, for example, local populations were required “to acquire an internal passport and to seek the authorization of the territorial administrator to be able to leave their home district” (Le Cour Grandmaison, 2006: 43). Similarly, with the Criminal Tribes Act of 1871, the British made it illegal for anyone affiliated with a “criminal tribe” to go ‘beyond the limits so prescribed for [their] residence’ without a pass (Hussain, 2007: 522). We should also recall within this context that the contemporary system of passport controls finds its origins within the British efforts to control the movement of unindentured Indians within the empire (Mongia, 2018).

The abandonment of the modern idea of equality before the law was justified within the colonial context on the basis of assumptions about the special conditions of the colony, which foreshadow the construction of borders as spaces of exception. This exceptionality was asserted by major colonial powers such as Britain, France, and Belgium even during the negotiations to create an International Bill of Human Rights after World War II, and it is enshrined in the Article 56 of the European Convention of Human Rights, infamously known as the “colonial clause,” which allows “due regard … to local requirements” in circumscribing the “territorial application” of the Convention (Mayblin, 2018; Spijkerboer et al., 2021).

As a result of this exceptionality, in the colonies law operated in a more or less permanent emergency, under which the generality and relative stability of laws gave way to ever-shifting decrees, arbitrarily laid down by the colonial administrators (Hussain, 2003; Reynolds, 2017). Punitive measures, normally deemed unacceptable under the rule of law, were permissible under colonial rule. Administrative internment, for example, allowed for the detention of colonial subjects with no right to appeal and often with no clearly specified time limit (Le Cour Grandmaison, 2006: 45). Even collective punishment was legally permissible; in French Algeria, for instance, the governor-general could impose a collective fine on an entire tribe or village if they did not cooperate with the authorities (Le Cour Grandmaison, 2006: 47). Political protests and local uprisings, even those involving unarmed, nonviolent protesters, were violently suppressed by invoking martial law and authorizing the use of lethal force (Hussain, 2003: 99–131). The state of emergency that characterized colonial regimes created a regime of impunity in which it was extremely difficult to uphold colonial officials (and nonofficial Europeans) accountable for the use of violence against the colonized (Kolsky, 2010).

While the colonial regime exempted the colonizers from any culpability, it rendered “the native” suspicious in the eyes of the law on the basis of a set of racist assumptions about the innate characteristics of the colonized, including aggressiveness, mendacity, irrationality, and criminality. Frantz Fanon’s scathing critique of the infamous Algiers School of Psychiatry, founded by Antoine Porot, brings to view how the pseudo-science of the time participated in the propagation of these claims. According to Porot and his followers, “the North African” was predisposed to criminality because of congenital defects tied to the underdevelopment of the cortex (Fanon, [1961] 2004: 227). Equally problematic were assumptions about “native mendacity,” which rendered the “native” an unreliable witness and made it very difficult to hold Europeans accountable for the crimes they committed against the colonized (Kolsky, 2010: 110). Similar assumptions about native culpability and untrustworthiness shape the ways in which states and courts interpret the words and deeds of migrants of color, as I discuss in Section 3, turning their nonviolent actions (e.g., climbing a fence) into belligerent acts and rendering their testimonies unreliable.

Colonialism, in short, turned the colonized into rightless entities who could be subject to arbitrary rule and violence with impunity. Doing away with the fundamental rights and basic protections associated with the rule of law, it elevated “lawlessness, inequality, and multiple daily murder of humanness” to the status of “legislative principles” (Fanon, 2018: 434). The lawlessness associated with colonial order is a problem arising not from the absence of law altogether but rather from the monstrous peculiarities of colonial law, which can shed critical light onto new forms of lawful lawlessness that arise in the migration context.

2 Lawful Lawlessness of Borders

On June 24, 2022, the Spanish border enclave of Melilla became the scene of a bloody massacre as the Spanish and Moroccan border forces violently thwarted the efforts of over 1,500 migrants, mostly from Chad and South Sudan, to cross the border to Spain. Described as the deadliest incident in the history of the Spanish–Moroccan border, the brutal operation led to the death of at least thirty-seven migrants. Some were killed in a crush, as they were trapped by the Spanish and Moroccan forces in riot gear; others fell from the fences due to the use of tear gas and rubber bullets (Gilmartin, 2022b).

The aftermath of the massacre was business as usual. In an interview with the Spanish daily newspaper La Vanguardia, Spain’s Socialist prime minister Pedro Sánchez briefly expressed his regret for the loss of lives, but without taking any responsibility; instead, thanking the Spanish and Moroccan forces for their work, he shifted the blame to the “mafias and criminals who organize violent actions against our border” (quoted in Hedgecoe, 2022). On September 13, 2022, his party joined the conservatives and the far right to quash a proposal for a parliamentary investigation into the events of June 24 (Gilmartin, 2022a). Meanwhile on the Moroccan side, a state-backed investigation claimed that the cause of death was “mechanical asphyxiation” due to stampede rather than the use of force (Sanderson, 2022). Moroccan authorities quickly started prosecuting sixty-five migrants who were among those who tried to cross the border on June 24, 2022 on numerous charges, ranging from illegal entry and violence against law enforcement officers to facilitation of illegal immigration (Le Monde, 2022).

The Melilla massacre brings to view the key elements that contemporary border governance shares with colonial regimes: strategic manipulation of jurisdiction to create lawless zones; differential allocation of rights on the basis of racialized status distinctions; justification of excessive force in a normalized state of exception; institution of a regime of impunity; and allocation of culpability to the victims of violence. In what follows, I examine the Spanish–Moroccan border regime in Ceuta and Melilla in order to understand how these elements are reconfigured within a new constellation of racial domination that redeploys lawful lawlessness.

Ceuta and Melilla, two Spanish enclaves bordered by Morocco, are the only land borders between the EU and the African continent. Spain conquered Melilla in 1497, and Ceuta, captured by Portugal in 1415, was transferred to Spain in 1668 (Saddiki, 2017). In an effort to control migration from the African continent, Spain started building fences in Ceuta and Melilla in 1993, and currently, there are three 6-meter-high, 12-kilometer-long fences. As “the only African territories of an EU member state (excluding the Canary Islands) since Algeria became independent,” Ceuta and Melilla remain an unresolved territorial dispute between Spain and Morocco (Aris Escarcena, 2022: 64). Despite its cooperation with Spain on migration control, Morocco has never recognized Spanish sovereignty over Ceuta and Melilla, and continues to characterize these enclaves as occupied territories (Boeyink, Sahraoui, & Tyszler, 2022; Saddiki, 2017). In fact, in a letter addressed to the United Nations Human Rights Council, denying any responsibility for the pushback operation on June 24, 2022, the Moroccan government contended that “Melilla is a prison occupied by Spain” and that “the kingdom of Morocco does not have any land borders with Spain” (quoted in Ramajo, 2022). Outraged by this statement, Spain’s prime minister forcefully reasserted Spanish sovereignty in the enclaves: “Ceuta and Melilla are Spain, full stop” (quoted in Ramajo, 2022).

The peculiar maneuvers that the Spanish government makes in dealing with questions of jurisdiction in these enclaves highlight the need to rethink contemporary border control policies in relation to colonialism. If we recall, a signature move of colonial powers was to manipulate different conceptions of jurisdiction to mark their colonies simultaneously as their sovereign possessions and as exceptional spaces where their laws did not apply. This colonial technique sheds critical light on Spain’s concept of “operational border,” a legal fiction that turns the border from a fixed line into a continuously fluctuating one. While Spain asserts sovereignty within Ceuta and Melilla, which should mean that the three fences are within its territorial jurisdiction, it also argues that migrants enter Spanish jurisdiction only after climbing over the third fence and crossing the last line of Spanish security forces. With this maneuver, the border is no longer the legal one defined by international treaties but rather an operational one that can be arbitrarily moved at the Spanish government’s whim (González García, 2019: 217). This is why, despite its claims to sovereignty in Ceuta and Melilla, Spain tried to evade any legal accountability for the Melilla massacre of June 24 by declaring that the events did not take place within Spanish territory (Sáiz-Pardo, 2022).

As exemplified by Spain’s jurisdictional subterfuge, extraterritorialization, similar to the colonial tactic of manipulating jurisdiction, aims to create zones of exception where rights and legal protections can be evaded. Locating Ceuta and Melilla within the longue durée of colonialism helps us understand how and why such policies target racialized subjects. Before becoming an “open-air prison” for migrants from Africa, Ceuta used to be a “penal colony” for “undesirable” subjects, including dissidents from “seditious America” – that is, free Blacks and Creoles engaged in abolitionist and separatist struggles in Cuba (Sánchez, 2018: 338, 333–334). That history urges us to reconsider extraterritorialization alongside the colonial orders that denied equality before the law on the basis of racialized status distinctions.

On its surface, the legal regime at the Spanish–Moroccan border seems to have nothing to do with colonial juridical orders, as it purportedly targets conduct – that is, illegal entry – and not a racially determined status (e.g., indigène or “native” under French colonialism). The move away from explicitly racial status categories, however, points to a radical reconfiguration, rather than complete abolition, of race rule within the migration context. To clarify, the distinction between status and conduct has never been clear-cut, despite the conventional assumption that modern law leaves behind status distinctions and concerns itself only with voluntary conduct. “Status” can appear in new forms in modern law when it gets disaggregated into legally “regulable component ‘acts,’” as illustrated by the antihomeless legislation that criminalizes activities associated with unhoused persons (Feldman, 2004: 50). Similarly, migration laws that target illegal or unauthorized entry amount to a de facto criminalization of status to the extent that one’s mobility, including eligibility for visas or visa-free travel, is significantly shaped by factors such as country of origin. Within the context of the Spanish–Moroccan border, due to cooperation schemes that the EU devised in order to outsource migration control, “illegality” becomes a racialized status. Particularly since the 2015 Valetta Summit, which established the EU Trust Fund for Africa, the EU has been providing financial assistance to African countries in return for their cooperation in migration control. The EU and its member states have also signed bilateral agreements with countries such as Morocco, Libya, Mauritania, Nigeria, Senegal, and Ethiopia. These schemes, euphemistically defined as “mobility partnerships,” aim to immobilize migrants from “sub-Saharan Africa” in particular. This term, normalized within the European debates on migration as if it were simply a geographical designation, is racially charged to the extent that it lumps together most of the countries in the African continent, despite their differences, and excludes only the countries in North Africa whose inhabitants are perceived to be closer to “whiteness” (Ball, Lefait, & Maguire, 2021). The coordinated efforts to illegalize “sub-Saharan” mobility have led to systematic violence against Black African migrants (Council of Europe: Committee for the Prevention of Torture [CPT], 2015; European Center for Constitutional and Human Rights [ECCHR], 2020), as illustrated by the massacre of June 24.

Such violence has become routinized in a normalized state of emergency, which, in ways reminiscent of “the rule of colonial difference” (Chatterjee, 1993), has been justified on the basis of assumptions about the peculiar characteristics of borders as spaces of exception (Reynolds, 2021; Spijkerboer et al., 2021). This emergency logic is exemplified by the amendment that the Spanish government introduced to its Aliens Act in March 2015, during the adoption of the Law “on the protection of citizen security” (Asylum Information Database [AIDA], 2021: 25). Working in the spirit of the emergency decrees characteristic of colonial rule, this amendment establishes a “special regime” in Ceuta and Melilla, as it authorizes the Spanish authorities to expel migrants suspected of “illegal entry” back to Morocco, without any inspection of their individual circumstances. Efforts to render the amendment unconstitutional has come to an end with the Spanish Constitutional Court’s 2020 ruling that the “special regime” in Ceuta and Melilla is in accordance with the Spanish Constitution and the ECtHR’s jurisprudence (AIDA, 2021: 27–28).

This special regime is presented by Spain as a justified response to an “invasion” by migrants who are depicted as “enemies” acting belligerently (Garzon, 2015). Migrants’ efforts to cross the border collectively are described as “quasi-military” attacks aided by the mafia, justifying the use of anti-riot gear and tactics (Groupe Antiraciste de Défense et d’accompagnement des Étrangers et Migrants [GADEM], Asociación Pro Derechos Humanos de Andalucía [PDHA], La Cimade & Migreurop, 2015: 46). This militarized discourse of “invasion” and “enemy” is tied to a racialized imagery of “Africa” and “Africans” within the European immigration context, as can be seen in the following statement of a colonel of the Guardia Civil: “It is Europe that must go to Africa, not Africa to Europe … They must be taught democracy, education, and almost by force if necessary” (quoted in Boeyink et al., 2022: 63). Within this Eurocentric imagery, “Africa” is represented as a place of absolute deprivation, devastation, and precariousness (Mbembe, 2017: 48–53) and “Africans” as destitute “savages” to be saved and/or disciplined by the “civilized” Europeans. Racialized migrants, especially those from “sub-Saharan Africa,” are held captive to representations that associate them with aggressiveness, criminality, mendacity, and irrationality, which were central to the construction of “the native” as a suspicious subject within colonial juridical orders. As a result, under the normalized state of exception at the borders, they stand before the law as always already culpable even for the violence they face regularly at the hands of border agents, as illustrated by the 2020 ruling of the ECtHR in N.D. and N.T. v Spain.

3 Always Already Culpable Subjects

N.D. and N.T. v Spain (2020) concerns the “pushback” of two migrants, N.D. from Mali and N.T. from Ivory Coast, by Spain to Morocco. N.D. and N.T. were part of a group of around 600 migrants who tried to cross the Spanish–Moroccan border through the border fence in Melilla in August 2014. They succeeded in doing so, but the Spanish authorities immediately returned them to Morocco, without assessing their individual circumstances. The question before the Court was whether Spain violated the prohibition against collective expulsion and the right to effective remedy by depriving these migrants of the means to challenge their immediate return. In 2017, a Chamber of the ECtHR found violations on both counts, but the case was then referred by Spain to the Grand Chamber, which ruled in February 2020 that the Spanish pushback did not amount to a violation because it was the applicants themselves who “placed themselves in an unlawful situation” as they “chose not to use the legal procedures … to enter Spanish territory lawfully” (N.D. and N.T. v. Spain, 2020: § 242).

N.D. and N.T. v Spain highlights the need to examine the shifting constellation of techniques mobilized for the racial governance of borders rather than focusing solely on extraterritorialization. While the Court rejected the arguments the Spanish government invoked to evade jurisdiction, it ended up upholding various other techniques associated with race rule – including an emergency logic that justifies exceptional forms of punishment, a status-based distribution of rights, and a racial allocation of culpability.

The Spanish government’s efforts to evade jurisdiction by invoking the concept of “operational border” were ultimately rejected by the Court, though not without some serious problems in reasoning. Seeing through the government’s intention to create a lawless zone, the Court underscored that “the special nature of the context as regards migration cannot justify an area outside the law where individuals are covered by no legal system” (N.D. and N.T. v. Spain, 2020: § 110). In its consideration of the Spanish government’s invocation of an “exception to jurisdiction,” however, the Court also entertained a possible scenario in which a state might be significantly constrained in its “effective exercise” of authority over its territory (§105). In making this concession, the Court was responding to Spanish government’s claim that it was the applicants’ “illegal storming of the border fences” (§128) that justified its action in accordance with Article 51 of the UN Charter, which justifies the states’ right to self-defense against ‘an armed attack’” (§126). While the Court did not take the argument of self-defense seriously in its literal sense as laid out in the UN Charter, it took on board the Spanish government’s image of an invasion by immigrants (“an attempt by a large number of migrants to cross that border in an unauthorised manner and en masse”) as it underlined the challenges confronting European states as a result of the political and economic crises in “certain regions of Africa and the Middle East” (§§166, 169).

This logic of emergency was invoked by the Spanish government to shift culpability for collective expulsion to the applicants themselves. During the Grand Chamber hearing of the case in 2018, the lawyer for Spain justified the pushback operation by inviting the judges to imagine themselves sitting at their homes with their families and suddenly seeing “600 strong men coming, menacing to break your windows”: “And then, your reaction is to close the windows and secure shutters and blinds. … Who is being violent and unlawful?” (N.D. and N.T. v. Spain, 2018). The Court effectively agreed with that reasoning, which presented the state itself as the “victim” rightfully engaging in self-defense. The Spanish government asserted, citing Hirsi Jamaa and Others v Italy – a 2012 ruling of the ECtHR that is often read as a historic victory for migrants’ rights against extraterritorial migration policies – that a state cannot be held accountable for “collective expulsion” if “the lack of an individual removal decision could be attributed to the culpable conduct of the person concerned” (N.D. and N.T. v. Spain, 2020: § 133). It was precisely this reasoning that the Court adopted, as it declared that such culpability exists when migrants “deliberately take advantage of their large numbers and use force … to create a clearly disruptive situation” (§201). The Court reinvoked this reasoning in its later ruling in Shahzad v. Hungary (2021), when it drew a sharp contrast between the irregular but undisruptive entry of Shahzad and the eleven other migrants he traveled with and the “clearly disruptive” acts of N.D., N.T., and their companions, whose “storming” of the border fences created a “situation which was difficult to control and endangered public safety” (Shahzad v. Hungary, 2021: §§ 59).

Such reasoning renders human rights conditional on arbitrary interpretations about the “culpable conduct” of migrants, but it is even more sinister in that it turns unauthorized border crossing into a de facto status crime when undertaken collectively by subjects who are perceived to be threatening. Determinations of culpability assume racialized overtones within the context of the Spanish–Moroccan border control, as discussed in Section 2, since this kind of “culpable conduct” is most likely to be undertaken by migrants from “sub-Saharan Africa” who are systematically denied access to legal entry procedures.

It was precisely this racial dimension of the problem that was disavowed in the Court’s ruling. I use the term “disavowal” in its Freudian sense to refer to a double gesture that consists of “a ‘refusal to recognize the reality of a traumatic perception’ combined with an oblique acknowledgment of that disturbing state of affairs” (Steinmetz, 2006: 448–449). We see such a gesture in the Court’s response to numerous forms of evidence submitted by third parties, including the United Nations High Commissioner for Refugees, which unanimously highlighted that “persons from sub-Saharan Africa” are systematically denied access to regular border crossing points and asylum offices and had no option but to enter Spain by climbing the fences (N.D. and N.T. v. Spain, 2020: §§ 58, 143, 155, 163). While the Court noted this situation, it also suggested that the reasons for it were not clearly established (§218). Obliquely acknowledging but also denying racism, it chose to assign culpability not to the Spanish state but rather to the migrant applicants, presenting the latter in effect as irrational subjects who could not produce “cogent reasons … based on objective facts” for not following the legal procedures for entering Spain (§229).

The disavowal of racism in N.D. and N.T. v Spain (2020) illustrates the Court’s reluctance to recognize what scholars have called “structural” racism, which takes us beyond individual prejudices and centers on the racial disparities arising in the implementation of seemingly colorblind policies (Armenta, 2017). A similar disavowal can be seen in Memedov v North Macedonia (2021), in which the Court acknowledged the overwhelming third-party evidence demonstrating the systematic police abuse that the Roma faced in North Macedonia, while refusing to read that evidence as proof of “institutionalised racism” or as evidence that racism was “a causal factor in the impugned conduct of the police officers” (Memedov v North Macedonia, 2021: §§ 55, 54).

In N.D. and N.T. v Spain, the Court’s reluctance to recognize structural racism is combined with racialized assumptions about disposition to law-breaking – not unlike the assumptions that underlie “the North African criminality” thesis criticized by Fanon, albeit this time attributing culpability to “sub-Saharan Africans” who were similarly declared to be “impervious to ethics” owing to their unlawful and violent conduct (Fanon, [1961] 2004: 6). In shifting culpability to migrants and suggesting that they could not produce compelling reasons for their “illegal” entry, the Court’s ruling effectively paints the migrants as irrational, impulsive, and aggressive subjects who would rather climb three fences, risk being injured by razor wire, and expose themselves to violence by border guards instead of following legal routes.

N.D. and N.T. v Spain highlights the need to understand extraterritorialization as one key element in a constellation of techniques put in the service of racial governance of borders. While the Court read the case as a clear-cut instance of territorial jurisdiction, its reasoning highlights how this territorial conception coexists with various competing conceptions in the legal domain, including a functional one (often applauded by migrants’ advocates) that can devolve into a dangerous logic of emergency during which a state, facing a violent attack by migrants, might be said to be constrained in exercising “effective authority” in its territory. In accordance with this logic, the racialized image of a violent horde (“taking advantage of their large numbers”) acting stealthily (“they frequently operate at night in order to produce a surprise effect”) and attacking the borders of Europe (“storming the fences”) is invoked to justify state violence. Colonial orders frequently invoked emergency in order to allow punitive measures that would not normally be allowed under the rule of law, including collective punishment. Similarly, in N.D. and N.T. v Spain, when the Court holds the migrants “culpable” (in ways similar to the racialized assumptions of “native” guilt under colonialism), it ends up justifying a collective punishment that would have been normally disallowed by the prohibition against collective expulsion.

Conclusion: Jurisgenerative Politics against Lawful Lawlessness?

N.D. and N.T. v Spain highlights the crucial role that law plays in transforming borders into “death-worlds” where racialized subjects become vulnerable to a form of necropolitics that either literally kills them or “confer[s] upon them the status of the living dead” (Mbembe, 2017: 92). But this process been contested by a wide range of struggles, including the numerous protests and uprisings of the migrants themselves, as exemplified by the March of Hope that shook Europe from mid-2015 to mid-2016 as thousands of migrants organized collectively and marched hundreds of kilometers chanting their demands for freedom and dignity. As many scholars have noted, such struggles highlight how migrants’ agency has transformed borders into “spaces of constant tension, conflict, and contestation” (Hess & Karakayali, 2018: 418); (see also Celikates, 2022; Mezzadra, 2018; Mezzadra & Neilson, 2013).

To what extent can law play a constructive role in such struggles? It has become increasingly difficult for law to assume such a role within official juridical spaces in which a statist understanding of immigration control has come to prevail. Within these spaces, law has increasingly functioned in a “jurispathic” way as a mechanism that endows state violence with an aura of legitimacy (Cover, 1983). For some, this problem reflects the “built-in” limitations of international human rights law, arising from “the asymmetry … between the right to leave any country and the right to enter” (Costello & Mann, 2020: 313). Given the racial inequalities at the heart of access to mobility, this asymmetry renders it especially difficult for migrants of color to seek redress within courts, as my analysis of the ECtHR highlights.

Not all is lost, however, and there are possibilities for pushing law in an egalitarian direction even in these more official sites, as illustrated by numerous strategic litigation efforts that have turned to international criminal law, maritime law, and tort law to tackle the limitations of human rights law (Costello & Mann, 2020). Such efforts to mobilize different kinds of law to contest the regime of impunity surrounding border violence are not unlike the struggles that turned the pluralistic colonial juridical regimes into sites of contestation – an “unequal contest,” to be sure, but one that involved “non-compliance and appropriation by the subjugated” (Merry, Sally, Engle, 1991: 891).

Moreover, when we approach such juridical efforts with a shift of perspective and move from “seeing like a state” to “seeing like a migrant” (Mezzadra & Neilson, 2013: 166), even cases such as N.D. and N.T. v Spain appear in a new guise as acts of claims-making that strive to transform borders into sites of contestation. Such a shift suggests that law cannot be reduced to the ultimate ruling issued by the judges, as it also involves an alternative archive consisting of the testimonies of survivors, third party reports, dissenting opinions, and petitions ruled inadmissible, among other things. Even a case that appears to be a simple “failure” from a statist perspective can usher in new lexicons and strategies of contesting violent borders.

Finally, law cannot be just reduced to official juridical sites and decisions. As Robert Cover famously reminded us, “the formal institutions of the law” do not exhaust the meaning of law as “nomos – a normative universe” or “a world of right and wrong, of lawful and unlawful, of valid and void” (Cover, 1983: 4). Law in this sense is “a system of tension or a bridge linking a concept of a reality to an imagined alternative,” one that relates “is” not just to “ought” but also to “what might be” (Cover, 1983: 9–10). Law as nomos, practiced beyond the official juridical sites, introduces us to the idea of “jurisgenesis,” which involves creative acts of legal meaning-making that contest the hegemonic interpretations of law (Cover, 1983: 11). Such jurisgenerative practices can be seen, for example, in the various political mobilizations around migrant deaths and disappearances, which have appropriated legal idioms such as “human rights” and pushed them into a more egalitarian trajectory for holding states accountable (Gündoğdu, 2021: 580–582; Rygiel, 2016). Such mobilizations highlight that law, understood as a normative universe that cannot be limited to state courts, can play a crucial part in the struggles that contest border violence and call for the universalization of a right to free movement.

11 Private Borders, Hidden Territories

Anna Jurkevics

The premise of this volume is that borders are shifting, and that as borders shift, rights and democratic legitimacy ought to shift with them. For good reasons, the focus of our discussion has been on sovereign states and their borders. However, sovereign borders do not exhaust the types of boundaries that shape and circumscribe human freedom. In this chapter, I discuss private economic geographies. Private geographies – home ownership, land monopolies for agribusiness and extraction, special jurisdictional zones for export and trade – often help funnel capital around the bottlenecks of state regulation, and they can act as a shield against the redistributive demands of democratic constituencies. In particular, I am interested in special economic zones (SEZs), which are “industrial areas governed by more liberal economic policies than those than apply to the rest of the country” (Levien, 2013: 394). The expansion of these liberalized economic enclaves has bolstered the power of states by shoring up gross domestic product (GDP), job growth, and the wealth of key constituencies. Politicians point to the concentrated economic gains that accumulate in SEZs as proof of their commitment to the good of the people, and in so doing, bestow legitimacy on a set of political practices that remove land, wealth, and labor from the jurisdictional reach of democratic regulation. Private geographies are of special interest because they reveal how states and capital cooperate in monopolizing and bounding land, acting in ways that consolidate and increase each other’s power. In this sense, private borders and properties call into question the public–private divide and reveal how power over land is determined in the global age, often in ways that evade democratic control.

Any inquiry into shifting borders, therefore, must account for the shifting map of the private realm. This realm is shifting both domestically and globally. Attention to the way that power is deployed in private geographies can help us understand the fraying of the social contract described by Seyla Benhabib as she considers the consequences of economic globalization: “State jurisdiction and territoriality are uncoupled as new agents of jurisdiction in the form of multinational corporations (MNCs) emerge. In some cases, the state disburses its own jurisdiction to private agencies in order to escape the territorial control of popular legislators. The social contract is increasingly frayed” (Benhabib, 2007: 28). This description of the interplay between the state and MNCs is apt. Current transformations in sovereignty do not dissolve the power of the territorial state, but instead rechannel it, often into the private realm. Therefore, instead of thinking about globalization in terms of de-territorialization (e.g., Shachar, 2020b), it makes more sense to think of it as a territorial reconfiguration, one which includes the “borders” of private enterprise.

Looking forward, then, it does not make sense to talk about globalization as bringing about a “post-territorial mode of political differentiation” (Fraser, 2009: 24). We are territorial creatures. Our global economy is a complex map of nodes and passageways, plantations, extraction sites, highways and rails, ports and seaways, markets, industrial parks, export processing zones (EPZs), trade zones, and SEZs. These networks of production and trade have always exceeded the bounds of the Westphalian territorial state, but they are territorial nonetheless (Benton, 2009). Therefore, as we think about the ways that globalization detaches certain processes from the territorial state, we must also consider how and where these processes reattach to the land, for example through offshoring. A better grasp of private territorial reconfiguration will allow us to consider how (or whether) the power exercised in economic enclaves can be democratically legitimated.

This chapter wades into the complexities of the global economy by considering the special economic zone (SEZ) as a form of contemporary territoriality. SEZs appear in many forms: the industrial park, the free port, the export processing zone, and the liberalized planned city. How, exactly, does this jurisdictional innovation for market liberalization allow private citizens and corporations to amass power that evades the grasp of democratic constituencies? And what is it like to live and work in an SEZ? SEZs are not abstract processes; they are real places with real fences, workplaces, dormitories, and planned neighborhoods. Workers’ lives are sometimes entirely circumscribed within these zones; others enter only to labor. These shifting private geographies, and the way they are navigated by those who traverse them, may hold the key to legitimation crises that emerge when the complexity of the global supply chain obscures the exercise of power.

I begin with a review of the direct precedent for SEZs, which is the export processing zone, or EPZ. I then detail the development of SEZs in China, Africa, and India. I pay particular attention to Shenzhen, China, a planned SEZ-turned-city that morphed from a fishing village to a megalopolis of 12 million within a generation. I conclude by drawing out a number of theoretical insights about private territories. These enclaves are economic spaces of exception, or better yet spaces of exploitation, and they are governed by their own shadowy set of private–public liberal economic norms. Because of their nature as enclaves, they are also largely invisible to the outside world, which has its own repercussions for democratic legitimacy. Finally, SEZs and other forms of economic enclave reveal the logic of public–private cooperation in the monopolization of land in the global age. To conclude, I suggest that the only way to reign in illegitimate private geographies is through overlapping, nonsovereign forms of jurisdiction.

1 Spaces of Exception between the Public and Private

Before I dive into the details of EPZs and SEZs, it will be helpful to review some of the theoretical tools I am using to analyze these phenomena. My focus will be on geographical-legal enclaves that are designed to protect economic activities, and to shield these activities so that they can be carried out without the burdens of state regulation. Where these enclaves are subject to special jurisdictions designed to shield owners from the law of the land, it is possible to think of them as economic spaces of exception. I use this term with caution, as it draws on Giorgio Agamben’s (2005) theory of the “state of exception,” which I do not endorse in its entirety. Agamben draws his theory from Carl Schmitt’s controversial declaration that “sovereign is he who decides on the exception” (Schmitt, [1922] 1985: 5). The state of exception, for Schmitt, is a situation in which the law has reached its limit, and the sovereign must rise above it (or suspend it) and take the matter into his own hands. For the many critics of Schmitt, this Hobbesian move to place the sovereign above the law is merely a justification for the abuse of power. I accept this critique. Yet I find Agamben’s elaboration of Schmitt useful because it acknowledges the capacity of unrestrained sovereigns to extend their spatial power beyond constitutional and jurisdictional limits. Offshore military detention sites such as Guantanamo Bay are the prime example of a space of exception. In these spaces, the sovereign disappears its delegation of untrammeled domination from public view, thus shielding citizens from knowledge of constitutional and jurisdictional breaches.

An economic space of exception, as I am formulating it here, operates differently from a military detention site because its purpose is to protect private economic power. These spaces emerge from public–private cooperation in shielding economic activities from the law of the land. In economic spaces of exception – such as the EPZ or an SEZ, though we could also include here contemporary land grabs or ante-bellum American plantations – the law of the land is suspended in favor of the law of the enclave (see Jurkevics, 2021). The law of the EPZ, for example, is usually a combination of policy set by the investor, host government incentives, and trade agreements. In this sense, it is neither public nor completely private. Importantly, one of the main activities that economic spaces of exception hide is exploitation of workers. Special jurisdictional rules allow corporations to underpay and overwork their employees, who often also carry the burden of dangerous or inhumane working conditions. EPZs and SEZs, then, are not only economic spaces of exception, but also often spaces of exploitation.

Private-public cooperation in removing jurisdictions from the law of the land is a phenomenon that deserves more attention because it overruns important distinctions in legal and political theory, and so can appear perplexing if we try to use traditional categories to understand what is going on. For example, in Chapter 7, “The Materiality of Territory,” Nishin Nathwani illuminates the long-standing distinction in Roman law, later taken up by early modern European theorists, between dominium and imperium rights. Traditionally, the former is a proprietarian right to own and dispose of objects and land privately according to one’s natural right. The latter is the public right to rule over either persons or territory, and is intimately tied to the notion of jurisdiction. I do not deny the importance of this distinction, but in the public–private partnerships I review here, it is often blurred beyond recognition. If states alienate their jurisdictional authorities to private actors, in what sense is imperium still functional? And if a private owner rules their property as quasisovereign, protected by jurisdictional shields, then in what sense is dominium really different from imperium? The concern thrown to light in this chapter is that the blurring of these rights is a significant problem for democratic legitimacy in the global age.

2 Export Processing Zones

In this section, I consider an important precedent for the contemporary SEZ, which is the EPZ. However, it is important to point out that economic enclaves have a long history prior to EPZs and SEZs. The cordoning off of land for special economic purposes, where life inside is vastly different from life without, is nothing new. Modern economies have emerged from a whole spate of segregated economic sites: colonial trade enclaves, plantations, mines, ports, workhouses, company towns, industrial parks, and, of course, the factory (Benton, 2009; Guimarães Pinheiro, 2019).

EPZs were pioneered as part of American economic policy in Latin America during the Cold War. Michael Levien explains, “EPZs evolved as a way to enable foreign companies to exploit cheap domestic labor in ‘third world’ countries by creating islands of political stability, good infrastructure and ‘favorable’ economic policies” (Levien, 2013: 394). The World Bank defines EPZs as “fenced-in industrial estates specializing in manufacturing for exports that offer firms free trade conditions and a liberal regulatory environment” (Madani, 2003: 5). The first EPZs were established in Latin America in the 1950s, and the strategy proliferated in response to structural adjustment programs in the 1980s. During this latter period, EPZs spread across Asia and Africa, where experimentation with the form gave rise to the SEZ.

Structural adjustment programs are an important part of this story. In the spirit of the Washington Consensus, EPZs were often initiated in small developing economies in order to secure International Monetary Fund (IMF) and World Bank loans, and to establish trade with countries such as the US, and later, China. The embrace of EPZs was also a decisive move away from import-substitution policies that had prevailed in many Latin American economies before liberalization. EPZs allow small economies to capitalize on their competitive advantage in abundant unskilled labor, and can deliver a remedy for large-scale unemployment. EPZs are often considered success stories because they provide jobs, but these numbers obscure wage depression. As EPZs proliferate, corporations seeking the lowest possible wages can play zones off against each other in a wage race-to-the-bottom, resulting in what Raphael Kaplinsky calls “immiserizing employment growth” (Kaplinsky, 1993: 1851). Additionally, exploitation and dangerous working conditions are widespread in these enclaves (Kaplinsky, 1993: 1856–1857).

There are a number of striking aspects of EPZs in the context of this discussion. First, EPZs are an economic space of exception par excellence. The zones are granted special jurisdiction by host states, which “remove almost all vestiges of state intervention, including taxes … allow[ing] ‘free market’ determination of wages, thereby enabling firms to make use of the developing country’s prime resources: abundant (and hence relatively cheap) unskilled labor” (Kaplinsky, 1993: 1851). The EPZ is a laissez-faire jurisdictional island within a state’s territory, where the applicability of laws protecting workers or the environment stops where the fence begins. The soft law that governs EPZs, moreover, is heavily dependent on the will of foreign trade partners. In Latin America, for example, EPZs were designed for exclusive trade with corporations in the US. The US uses bilateral agreements with such states to limit how much value can be added to products made in the EPZs, thus making sure that the labor remains unskilled. The US also enforces monopolistic rights to provide the raw materials that are processed in these sites. In this way, the EPZ becomes a satellite territory of the US economy, one which escapes US labor and environmental regulations as well as the laws of the host state. The host country, which relies on the EPZ to guarantee loans from the World Bank and IMF, becomes completely dependent on its overpowering trade partner. Any move that imperils these relationships, for example pushing back on wage depression, is avoided to secure loans.

Another striking aspect of the EPZ, which carries over into contemporary SEZs, is the lack of linkages with local economies. One conundrum economists have faced is the complete disconnection of the activities that take place in these enclaves with local economies. The materials processed in EPZs are not drawn from host states, nor are the products sold in host states. The only direct link to the local economy is the wages that are paid out to the workers. It is unsurprising, then, that EPZs have come under fire by citizens who were sold these projects by governments claiming that they would lift employment and bring wealth to the people. The wealth rarely makes it outside the border-fences of the zone (Kaplinsky, 1993). It stays in the hands of the factory owners and managers who work directly with foreign investors.

3 Special Economic Zones

Contemporary SEZs emerged from experimentation with EPZs, applying the economic space of exception to a broader range of activities that can take place in an enclave. The World Bank defines SEZs as “demarcated geographic areas contained within a country’s national boundaries where the rules of business are different from those that prevail in the national territory … the zone is given a business environment that is intended to be more liberal from a policy perspective and more effective from an administrative perspective than that of the national territory” (Akinci & Farole, 2011: 3).

The definition is vague because, as the authors admit, there are all sorts of contemporary experiments in liberalized economic zones, from traditional EPZs, to free ports and industrial parks, to cities such as Shenzhen, Singapore, and Dubai. What is essential for SEZs is that they procure exceptions from national and local law, implement liberal economic norms, and attract foreign direct investment from multinational corporations (MNCs) seeking cheap labor and tax breaks. Broadly speaking, SEZs are geographical experiments in the privatization of land for industry and economic liberalization.

The growth of SEZs globally over the last forty years has been explosive. According to the World Bank, the International Labour Organization’s database of SEZs reported 176 zones in 47 countries [in 1986]; by 2006, this number rose to 3,500 zones in 130 countries. SEZs now are estimated to account for more than US$200 billion in global exports and employ directly at least 40 million workers (Akinci & Farole, 2011: 5). These numbers are from 2011. Since then SEZs have only expanded, though it is notable that many have also failed and been abandoned. Plots of half-cleared land, long since expropriated from peasants and farmers, sit unused and decrepit, a haunting reminder of overzealous land accumulation (Levien, 2013: 388).

While traditional EPZs are public–private ventures, SEZs have become increasingly privatized, where the land itself is owned by an investor or MNC (Akinci & Farole, 2011: 7). Therefore, SEZs must be put in the broader context of “neoliberalism” and the dynamic uncoupling of economic power from state control (Brown, 2015; Honig, 2017; McKean, 2020). These enclaves are designed to evade democratic constituencies, and are purposefully insulated. Their geographical isolation and boundedness only reinforce their systematic legal untethering from sources of democratic legitimacy, such as state or local legislatures, or even the public sphere. SEZs are notoriously difficult to gain access to, if we are to believe the reporters who attempt entry (Merchant, 2017).

China has been the most enthusiastic implementer of SEZs, extending liberalized enclaves beyond industrial parks and ports to entire cities. SEZs have played a key role in China’s incredible success in developing a manufacturing export-based economy, and in the rise of the “Asian tigers” more generally (Akinci & Farole, 2011). China uses SEZs to experiment with economic liberalization. The SEZ model has been applied to planned cities such as Shenzhen, which I explore later, and the policy architects behind Chinese SEZs have gradually moved to shift these zones offshore into Africa in order to consolidate other geopolitical gains along with economic ones.

The first African SEZ was established in 1971 in Mauritius, whose island position to the east of Madagascar situates it well as a free trade center linking Africa to South and Southeast Asia (Bräutigam & Xiaoyang, 2011: 29). However, the real boom in SEZs in Africa has come on the heels of China’s plan to internationalize its manufacturing power, starting in the early 2000s. By building SEZs, China has gained a foothold in Africa, which holds attractive resources and land for investors. Additionally, many African countries hold favorable international trade agreements that the Chinese can take advantage of when they funnel production through proxy states (Bräutigam & Xiaoyang, 2011: 37). Most of the zones in Africa are being developed by private companies, though they receive support from the Chinese government (e.g., diplomatic support, start-up funds, and loans). In return, the African host governments give standard SEZ packages to the investors: “tax holidays, and waivers on import tariffs for raw materials and inputs, along with restrictions on strike activity” (Bräutigam & Xiaoyang, 2011). African partners in host countries are expected to play a background role in supplying land and labor, and if they have a stake in the SEZ, it is generally under 20 percent (Bräutigam & Xiaoyang, 2011). There are also ongoing conflicts over the percentage of Chinese workers in these zones, which can be quite high (Bräutigam & Xiaoyang, 2011: 44–45). Where SEZs do not even relieve joblessness in their vicinity, their real purpose as an economic space of exception and exploitation becomes more obvious.

What does the proliferation of Chinese SEZs in Africa mean in territorial terms? Put plainly, SEZs are geographical enclaves of quasiprivate Chinese territorial rule within African countries. Offshore SEZs are reminiscent of the colonial trade enclaves and extraction sites of the great European powers during the age of empire (Benton, 2009). This is an old territorial strategy with a new name, an economic space of exception and private jurisdictional dead-zone immune to popular self-government in the host state. In this sense, the SEZ shares much with the privatization and foreignization of land that has taken place through the phenomenon of land grabbing, wherein huge swaths of land in Africa (among other places) have been transferred to private MNCs for agribusiness. African land is once again, as it has been so many times before, foreignized (Zoomers, 2010).

The incredible success of Chinese SEZs has encouraged other large economies to get into the game, with sometimes disastrous results. In 2005, the Indian government passed the SEZ Act, which encourages private companies “to develop, maintain and effectively govern economic zones on a for-profit basis” (Levien, 2013: 394). The Act has roused backlash for helping developers and investors expropriate land from farmers and raise rents.

The postcolonial history of economic enclaves in India can be traced back to Nehru’s nationalization of the steel industry, which saw the creation of company towns and early industrial parks (Levien, 2013: 385). Between 1947 and 2004, the Indian government displaced an estimated 60 million people to make way for national development projects (Levien, 2013: 403). The “regime of dispossession” (Levien, 2013: 381) of this era was one of government expropriation of land for development. Manufacturing and infrastructure were the main goals of these projects. While not without conflict and pushback, these projects were nationalized and therefore had a more solid claim to bolster “the good of the people” than privatized SEZ projects. Since then, however, the global trend of privatization has entrenched itself in Indian policy.

Nationalized development projects set the stage for SEZs in India insofar as they taught the government how to expropriate land. But SEZs are a distinct animal. The SEZ Act lays out a policy wherein the investor must only prove that their activity on the land parcel will produce more GDP than the current activity, which is almost always farming. This is not hard to prove, especially if the investor builds housing on the SEZ. Rental housing on Indian SEZs generates far more wealth than other industrial activities, for example, tech centers, on the same lands. As one might imagine, investors are quick to see the terms of a great deal: The government expropriates and clears the land, provides infrastructure, and the investor comes in, builds apartments alongside some nominally industrial activity, and collects rents. SEZs have thus popped up on the outskirts of Indian cities, where housing needs are concentrated. As housing and land values increase – a broader global trend – SEZ owners accumulate ever more wealth, which they can use to invest in further land speculation (Levien, 2013: 395). Thus in India, the SEZ has morphed from a jurisdictional tool that encourages production, export, and trade into a prime opportunity for rent-seeking (Levien, 2013: 384). The Indian approach to SEZs since 2005 must be counted in the shift towards privatization and land speculation that has exploded since the 1980s (Byerlee et al., 2011).

For Levien, “SEZs represent the shift to a regime of dispossession in which the state has become a mere land broker for capital, expropriating land for any private use … no matter how immaterial, speculative, or consumptive” (Levien, 2013: 384). Unsurprisingly, land conflicts have intensified in India. The recent protests and uprisings of Indian farmers are a response to market liberalizations and reflect frustrations that have come to a head after a new wave of dispossessions. Increasingly, such policies are coming under fire because Indians can see these projects for what they are, namely, rent-seeking, unfair land monopolization, and real estate scams.

The development of the SEZ as a private, bordered territory is likely to have political consequences for many years to come. Even where democratic constituencies are successful in legislating redistributive economic policy, environmental protections, or labor protections, these policies cannot touch the SEZ. SEZs are so thoroughly cordoned off from national territory – both jurisdictionally and sometimes literally – that they may as well be another country. Some of these zones, such as the Foxconn manufacturing plants in Shenzhen, are carefully guarded to keep journalists and outsiders out.

What is to be done? Can democratic politics, as it is practiced today, breach the walls of the SEZ? Unfortunately, most of the examples to be found of improvement in conditions in SEZs have come on the heels of disaster and scandal. For example, in 2005, an explosion in a Chinese dynamite factory in Zambia – the result of unsafe labor conditions – killed fifty workers and precipitated reforms. The Zambian workers in the China Non Ferrous Metal Company mines were granted the right to unionize. But is scandal the only way to tame the SEZ? I come back to the question in the conclusion.

4 Shenzhen

In this section, we enter Shenzhen, an SEZ and planned city whose recent meteoric rise has already been thoroughly mythologized. Shenzhen grew from a mere fishing village into a megacity city of more than 12 million in only forty years. The legend is captured well in this nationalist propaganda song “The Story of Spring”:

The Year of 1979

That was a spring

There was an old man

Drawing a circle by the South China Sea

Mythically building a great city

Miraculously forming a mountain of gold

Shenzhen! Shenzhen!

The Test Pilot of China’s Reform and Opening¹

The old man here is Deng Xiaoping, who succeeded Mao and is “credited with having single-handedly pivoted China toward economic reforms” during his period of rule from 1976–1989 (Du, 2019: 1–2).

Shenzhen was established in 1979 as an SEZ. The location was chosen for its proximity to Hong Kong. The city opens out from China’s southern coast onto the South China Sea, with its concentrated shipping and trade routes. Shenzhen has been championed as a miracle of modern centralized state planning. It is an economically diverse city of hopeful workers, many of whom live in the varying conditions of intracity “urban villages.” Some of these villages are slums, but there are gradations in conditions that reflect one of Shenzhen’s aims: to pull the rural Chinese, who have migrated to the city en masse, out of poverty.

Inequality in Shenzhen is jarring. Baishizhou is the name of the poorest urban village, a slum of newly urbanized peasants, a place of mythic violence, the mob, and winding alleys. To its west is Shenzhen University and Technology Park. To the east of Baishizhou is the “Overseas Chinese Town” (OCT), a planned neighborhood filled with theme parks and luxury residences. Tourists can find everything they need in the OCT, which is separated from Baishizhou by guarded wall, tall enough to block the neighborhood entirely from the eyes of those on the other side. Du paints a striking picture of the separation of the OCT and the slum: “One portion of the separation wall that lines the Window of the World theme park is disguised by a landscaped hillside, on top of which stands a five-meter-tall statue of Christ the Redeemer with … arms open toward the theme park, while his back is turned to the masses of Baishizhou” (Du, 2019: 268).

The artificial world of the OCT is capped by the Portofino neighborhood in the north, a luxury gated community modeled on the Italian town. “The pastel-colored villas, high-end restaurants and bars, cobbled piazza, and medieval bell tower are all scattered around a natural spring-fed lagoon named Swan Lake” (Du, 2019: 268). The wall that separates “Portofino” from Baishizhou is covered in barbed wire.

Life and urban form in Shenzhen follow the dictates of capitalist growth, indeed only exist for those dictates. When an SEZ is diffused into a megacity, the separation of labor and capital take the shape of the city. Neighborhoods are bordered. The wealthy are never to see those whose hands produce. The constraints of class are written purposefully into the city map, where what used to be outer SEZ walls penetrate the city in varied forms of segregation. The playgrounds of the global elite sit happily on top of the slums, where the only thing they appear to share is the polluted air of a manufacturing city. Is this any different from other centers of global capital: New York, London, Singapore, Shanghai? In many ways, the answer is yes. Those cities were not planned like Shenzhen. There is nevertheless an uncanny convergent evolution of global cities, where the unplanned cities are becoming increasingly hierarchical, spatially segregated playgrounds of the super-rich, while the planned city of Shenzhen labors to imitate that world, and at the same time makes it possible. London is not a manufacturing city; it needs no walls to keep workers out of its rich neighborhoods. It has, however, subtly pushed its underclass further and further to the periphery with each new wave of rent-increases, increases that follow the logic of global speculative markets.

The composition of Shenzhen is strange because while the whole city is an SEZ, governed by liberalized economic norms, there are also industrial parks and manufacturing campuses within the city that have the fenced-off characteristics of traditional EPZs. Shenzhen is a city of SEZs-within-SEZs, of private borders within private jurisdictions. The most famous industrial park in Shenzhen is Foxconn City, where major tech companies such as Apple produce devices. The iPhone is still produced today at the Longhua complex in Foxconn City. It was this place that gained international attention for a wave of worker suicides in 2009–2010. The haunting image of workers hurling themselves off the top of dormitories in Foxconn City has lingered in the global consciousness and sparked interest from journalists, who have worked hard to breach the increasingly secure complex. To use the concepts I introduced earlier, Foxconn City is a space of exploitation within a space of exception.

In the words of a journalist who was able to make his way into Longhua, “it’s a fairly aggressively shitty place to spend long days … Longhua [feels] like the dull middle of a dystopian novel, where the dread sustains but the plot doesn’t” (Merchant, 2017). In 2010 alone, there were at least eighteen suicide attempts and fourteen confirmed successes. Many others had to be talked down from roofs. In response, “Foxconn CEO, Terry Gou, had large nets installed outside many of the buildings to catch falling bodies. The company hired counsellors and workers were made to sign pledges stating they would not attempt to kill themselves” (Merchant, 2017). For the estimated quarter of the workers who live in the dormitories onsite, the legacy of the suicides is to be found in the “cage-like fences built out over the roof and the windows” (Merchant, 2017) of their living quarters.

Attempts at PR control from Foxconn and Apple seemed to miss the point. They argued that Foxconn work conditions are on a par with other manufacturers in China, and that the suicide rate at their factories is not actually higher than the population at large (Merchant, 2017). But the image of these striking deaths, of their frantic flights and suicide notes detailing their unlivable lives, of the net that now turns desperate escapes into circus acts – these images cannot be unpainted by statistics about the generality of unbearable lives in manufacturing economies. The Foxconn suicides are a jolt; they reveal the depth of life and death that makes up our supply chains, keeping our iPhones in our pockets to provide all of our on-demand digital needs.

In a dark turn, the attention paid by the international press to the 2010 suicides has sparked a new strategy for workers:

In 2012, 150 workers gathered on a rooftop and threatened to jump. They were promised improvements and talked down by management; they had, essentially, wielded the threat of killing themselves as a bargaining tool. In 2016, a smaller group did it again. Just a month before we spoke, Xu says, seven or eight workers gathered on a rooftop and threatened to jump unless they were paid the wages they were due, which had apparently been withheld. Eventually, Xu says, Foxconn agreed to pay the wages and the workers were talked down.

(Merchant, 2017)

Once again, I ask, is scandal the only way to achieve reform within an SEZ?

In the Shenzhen model, the sharply fenced-off world of the SEZ is diffused into the fabric of a city designed to meet the needs of production. The zone becomes much more complex than a single fenced-in site, though it contains those fences as well. The city is divided and proliferated into overlapping bordered enclaves: manufacturing spaces with their special jurisdictions, workers dormitories, urban villages and slums, high-end gated communities, themes parks, luxury hotels, and the business district. Each space serves a function in the Shenzhen model, and each space is heavily insulated from the others so that people only see the spaces relevant to their status. The stratifications and internal borders of the city make the structure of the SEZ largely invisible to those who live and work in it, because it simply becomes the fabric of their lives.

Conclusion

We are living in a moment of democratic backsliding. Ours is an age of legitimation crisis, in which citizens wade perilously through untrustworthy information to figure out how power is deployed over their lives. But if we do not even know where power resides – and when I say where, I mean it literally – then we cannot govern ourselves. It is for this reason that I think it is important to incorporate the study of private geographies into scholarship about territoriality. Borders and boundaries contour our lives, mobility, and freedoms. The salient borders that we encounter are not all public, and for many people across the globe, especially laborers, private borders may be the most salient ones in their lives. In this chapter, I have argued that the invisibility and insularity of economic enclaves such as SEZs make them difficult to govern. These private territories must be brought into the light so that everyday workers and citizens can demand legitimation of the activities that go on within.

Another theoretical issue that has emerged in this study is the entanglement of public and private forms of power when it comes to producing territoriality. Territorial innovations such as SEZs are the product of cooperation between public powers and private investors, who make mutually beneficial agreements over the heads of citizens. Often, these agreements are posed as projects that are for the good of the people. With the privatization of economic enclaves, which also increasingly follow the logic of speculative land markets, these claims are disingenuous. The truth is that the private transformation of land has long been entangled in advances of state power. For those of us concerned about the effects of this cooperation for democracy, we ought to be suspicious of laws that protect private venture from regulation, and we ought to be suspicious of special jurisdictions – these can be the first step in the foreignization of territory and the creation of jurisdictional dead-zones. Ultimately, sovereign states and private investors can cooperate on these ventures because they have a shared goal: monopolization of land. Investors want to “see like a state” so that they can extract as much labor and resources as possible from their ventures. Perhaps Marx goes too far when he accuses the state of being a steering committee of the bourgeoisie,(Marx & Engels, 1978: 475), but where SEZs are concerned, he hits the mark.

The cooperation of the state and private enterprise in monopolizing land and shielding it from democratic regulation is bolstered by jurisdictional tools such as SEZs that help the state create an economic space of exception. These spaces – and I’ve argued that EPZs, SEZs, plantations, and land grabs can become spaces of exception – create territorial gaps in which power and money can be pooled, and removed from the reach of democratic constituencies. SEZs are economic pressure-release-valves within the sovereign states system: They create space for states to move outside their borders without overtly undermining the international principles of sovereign territoriality and nonintervention. They also create opportunities to employ “floating populations” regarded by host states and investors as superfluous, to use Sibel Karadağ’s term from Chapter 12, “floating populations.” To fulfill these purposes, the innovators and governors of SEZs have developed norms unique to their use. It is, therefore, somewhat deceptive to call SEZs spaces of exception because they are brimming with law, treaties, contracts, and norms. They are excepted from national law, but they still follow a logic based on the principles of profit-maximization and economic liberalization. To paint these enclaves as spaces without law would be to mystify them. SEZs and other economic enclaves are full worlds of capitalist modernity. While exceptional in some ways, they also bring the contours of our global systems into sharp relief, and so could be characterized as a “constitutive exterior” (Derrida) of a sovereign states system committed to the principles of capitalist growth.

The only successful attempts to harness power within SEZs that I have mentioned in this chapter came on the heels of scandals – explosions in mines, mass worker suicides. But what should the democratic response to SEZs be? Ultimately, to handle these shifting lines of territorial power – both public and private alike – we must shift the lines of democracy. The spaces of power that determine our lives do not just correspond to the borders of the sovereign nation-state. They do not now, and they never have. We must acknowledge the fundamental pluralism of territorial powers. Moreover, we cannot always rely on the sovereign state to be the source of democracy, and so there must be more fora beyond the national stage for democratic decision-making, and these must correspond to the places where power is geographically broadcasted. Therefore, I believe it is time for democratic theorists to let go of our obsession with sovereignty and theorize democracy as a set of dispersed sites of inclusive political action in a configuration of overlapping, nonsovereign democracies (Jurkevics, 2019). Jurisdictions, for example regional organizations, that overlap with SEZs ought to support attempts at workplace democracy and unionization that emerge within these zones. As new privatizations of land reconfigure our global experiences of “home,” we must encourage the rooted democratic politics that emerge in resistance and response. The sites of democracy already exist. What they need is support and the sheen of legitimacy that theory is specially positioned to bestow upon them.

12 Cycles of (Im)mobility: Floating Populations in the Case of Turkey

Sİbel Karadağ

Introduction

In the summer of 2015, during which approximately 1 million people arrived in European territories from Turkey and an estimated number of 799 migrants died,¹ an immediate response was formulated under the 2016 EU–Turkey Statement (the so-called Refugee Deal),² with the intention to construct and intensify the transnational infrastructure of externalized border governance not only across the Aegean Sea and within Turkey, but also in the countries of origin.³ With the Statement, the EU has presumed Turkey to be a “safe third country” despite the fact that it does not provide refugee status to people coming from a non-European country owing to the principle of geographical limitation in the 1951 Convention.⁴ Since then, Turkey has been hosting the largest displaced population in the world. What was initially framed as the return of those who “irregularly” crossed from Turkey to the Greek islands in 2016 has culminated in the return of Syrians from Turkey being at the center of political debates, with society on the brink of social turmoil and escalating antirefugee discourse and hyperimpoverishment in recent years.

In 2021, a newly established far-right party whose main platform is strident antirefugee rhetoric came to prominence. Under its sponsorship, a nine-minute video, entitled “The Silent Occupation,” was posted on YouTube and widely circulated on social media. Portraying a dystopian future in which the Turkish republic had collapsed, the video shows Turks as second-class citizens who are being discriminated against and oppressed by Arab migrants. Following a social media campaign dubbed “send Syrians home” and the general discourse of the main opposition, the issue of migration became one of the most significant aspects of the national election process, and on May 4, 2022 the Turkish president announced the commencement of a return plan for 1 million Syrians to prefabricated houses, schools, and hospitals being constructed in northern Syria under the control of the Turkish army.⁵ Reconnect, a project for training Turkey’s first national voluntary return counsellors, was launched in October 2022, co-implemented by the International Centre for Migration Policy Development (ICMPD) and the Presidency of Migration Management of Turkey (PMM).⁶

This chapter turns the spotlight onto Turkey, a country at the EU’s frontier and a candidate country that has a frozen accession history and is currently hosting the world’s largest displaced population (around 4 million as of spring 2024, including registered Syrians and non-Syrians),⁷ Turkey itself is witnessing economic downturn, hyperinflation, extreme impoverishment coupled with authoritarian rule, and deterioration of the rule of law. The EU’s decades-long border policies, producing new spaces of intervention exercised with its neighbors and along migratory routes, have sparked considerable scholarly interest and criticism. This has led to an examination of the reconfiguration of border mechanisms as outsourced, offshored, shifting, mobile, itinerant, and dispersed, operating within and beyond the territorial limits of sovereign states (Balibar, 2004; Bialasiewicz, 2008, 2012; Casas, Cobarrubias, & Pickles, 2011; Casas-Cortes, Cobarrubias, & Pickles, 2013, 2016; Coleman, 2007; Newman, 2006; Rumford, 2006, 2008; Shachar, 2020b; Walters, 2004, 2006; Weizman, 2007). These processes have not only turning the Mediterranean Sea into a “death world” (Mbembe, 2003), but have also created various forms of “deterritorialized zones of lawlessness” (Benhabib, 2020: 96).

For historians, the assumption that the regulation of movement is a constitutive aspect of any political order is nothing new. Throughout history, apparatuses of enclosure, forced displacement, and free circulation of movement have been exercised in various formulations and allocations (Kotef, 2015; Nail, 2016). In today’s world, we witness borders and legal barriers to be increasingly in flux; as Ayelet Shachar nicely puts it, operating “in a quantum-like fashion, simultaneously fixed and fluid, stationary and portable, exerting influence over those coming under its kaleidoscopic dominion” (Shachar, 2020b: 9). Besides the spatial mobility of today’s borders, Elizabeth F. Cohen (2018) reminds us how time as well is assigned as a form of political value and is frequently inserted into political procedures for conferring and denying citizenship rights. Many others are analyzing the complex temporalities at stake in contemporary modes of governing migration (Jacobsen et al. 2021). Yet concerning the overlapping of time–space coordinates at borders, the Western gaze has to be transcended in order to go beyond current concepts and to grasp the level of “liquidity” in the Global South as opposed to that in the Global North (see Bauman, 2000).

As the largest refugee host country in the world, Turkey, I argue, manifests a varied set of legal and governing techniques that are designed to monitor millions of displaced people within a broad design of temporality and spatiality. When the country’s contested gatekeeping function for Europe coincides with a shrinking economy, an authoritarian rule seeking to hold on to its political power, and the decades-long ideological aspiration to be a regional actor in the Middle East, millions of displaced bodies are turned into an instrument of demographic engineering, at a time when this is strikingly in flux (İşleyen & Karadağ, 2023).

In the chapter, I first focus on the legal and spatial/temporal architecture of Turkey’s migration regime, one that produces layered levels of legal precarity and temporal gaps that are intertwined with demarcated spatial boundaries. Second, I explore how this legal and spatial/temporal architecture is daily violated and self-sabotaging in line with the needs of the informal labor market, thus continuously generating irregularity and arbitrariness. Finally, by examining events, I explain a set of governing technologies, at times paradoxical, that transform these irregularized bodies into floating populations engaged in cycles of (forced) movement. The study not only examines the role of legal design in the country, but also analyzes the daily implementation of legal texts, aiming to capture the level of arbitrariness that is present. To that end, I conducted interviews with five lawyers who specialize in Turkish refugee law.

1 Legal Architecture of Protection: Degrees of Temporality and Spatial Organization of Movement

The basis of Turkey’s international protection regime dates to the 1951 Geneva Convention and thereafter the 1967 Protocol, with Turkey still holding to the principle of geographical limitation in its implementation. What this entails is that refugee status is granted only to persons originating from European countries, while non-European asylum seekers have to make an individual asylum application and undergo “refugee status determination” procedures before being resettled in a third country. That is to say, non-European asylum seekers can only stay in Turkey temporarily until they are resettled in a third country. Despite receiving global migration flows over the last few decades, Turkey lacked a fully-fledged legal architecture of international protection until 2013, instead pursuing ad hoc reception and settlement policies that could be linked to her nation-building process (Table 12.1). Accordingly, the arrival of Turkic and Muslim populations was chiefly encouraged in cases of mass displacements,⁸ based on the status of “Turkish descent and culture,” while the emphasis on Turkish ethnicity has continued to be a significant factor (Karadağ, 2021; Karadağ & Sert, 2023; Kasli & Parla, 2009; Parla, 2011).

Table 12.1 Critical junctures in Turkish migration policy since the early twentieth century

Homogenization of population during the nation-building process, 1920s–1930s Convention on the Exchange of Greek and Turkish/Muslim Populations (1923) The early years of the Republican period witnessed a bilateral agreement with Greece aiming at population exchange, namely, the emigration of non-Muslim populations from Anatolia and the immigration of Turkish/Muslim populations. The estimated number of people living under the Ottoman period during World War I was approximately 16 million, including 13 million Muslims and 3 million non-Muslims. As the result of the population exchange and forced displacements, the percentage of the non-Muslim population in Turkish Republic dropped from 19 percent to 3 percent in 1927.9 Law 2510/1934 Settlement Act (1934) The Turkification and Islamization of the population became more pronounced with the 1934 Law that facilitated immigration and integration of those of “Turkish descent and culture” while preventing the entry of those who did not meet this criterion. After the 1950s Geneva Convention relating to the Status of Refugees (1951) Turkey became a signatory to the 1951 Geneva Convention while reserving temporal and geographical limitations. Accordingly, Turkey granted refugee status only to those originating from Europe. Meanwhile, non-European asylum seekers were given temporary residence in the country. Further emigration of non-Muslim populations (1955) Turkey–West Germany labor recruitment agreement (1961) Based on the official records, nearly 800,000 workers went to Europe (81 percent to Germany) on the basis of a plan aiming to boost economic growth in Europe. UN Protocol relating to the Status of Refugees (1967) Turkey kept the principle of geographical limitation from the 1951 Convention. New migration patterns in the 1980s–1990s USSR’s invasion in Afghanistan and 1982–1983 Afghan migratory movement to Turkey The official policies concerning “Turkish descent and culture” in the access to citizenship continued in this period. Mainly Uzbek and Turkmen Afghan families (4,163) were accepted, based on their ethnic background, and they were settled in the six designated provinces of Turkey. Gulf War and mass immigration of Iraqis, Iranians, and Kurds Ad hoc reception and settlement policies continued in the cases of Iraqis, Iranians, Kurds, and others, but there was not a comprehensive asylum and protection law. During these settlement policies, the status of “Turkish descent and culture” remained a determining criterion. 1989 reception of Bulgarian Turks fleeing the Zhivkov regime Regulation No. 6169/1994 on the Procedures and Principles related to Possible Population Movements and Aliens Arriving in Turkey either as Individuals or in Groups Wishing to Seek Asylum either from Turkey or Requesting Residence Permission in order to Seek Asylum from Another Country (1994) The 1994 regulation laid out the procedure for those seeking international protection for the first time. However, it was not a fully-fledged legal document and owing to the principle of geographical limitation, Turkey granted only temporary stay to non-European asylum seekers until their resettlement in a third country. Europeanization of migration policy, Syrian migratory movement, and First Law on Protection, 2000–2023 Turkish National Action Plan for Asylum and Migration (2005) The 2000s witnessed new modes of mobility regarding transit and undocumented migrants as well as asylum seekers from Middle Eastern, Asian, and African countries. Simultaneously, Turkey was further influenced by her EU candidate status. The Action Plan of 2005 was a result of Turkey’s harmonization process with the EU on asylum and border governance. This aimed for the lifting of the geographical limitation, the establishment of an agency for migration, a fully-fledged international protection system, and other measures. Syrian migratory movement and Turkey’s “open door policy” Law 6458/2013 on Foreigners and International Protection (LFIP) Turkey’s first comprehensive law on international protection. The law defines the preconditions for each legal category under international protection by maintaining the principle of geographical limitation. 2015 migratory movement from Turkey to Europe 2016 EU–Turkey Statement

In 2013, the legal construct of migration regulation finally came into force, with the adoption of the Law on Foreigners and International Protection (LFIP), a long-awaited product of the harmonization process with the EU border regime that had developed in tandem with EU membership negotiations since the 2000s. The LFIP contains its own peculiar technologies of governance which render it a sui-generis case of temporal and spatial architecture. First, it still applies geographical limitation and thereby provides refugee status only to those from European countries. Considering that European applicants are nonexistent in practice,¹⁰ the law mainly lays out the temporal hierarchies according to specific legal categorizations for millions of displaced populations from non-European geographies, in a way echoing what Tize (2020) calls “permanent temporariness” to varying degrees. It also delivers legal spatiality as a settlement policy in which legal access and accordingly entitlements are available for those who reside in their provinces of registration.

The law rests on two main legal pillars: temporary protection status (TPS) for Syrians (3,763,565 persons as of March 2022) and international protection status (IPS) for non-Syrians (330,000 persons as of December 2021).¹¹ The latter category involves displaced persons predominantly from Afghanistan, Iraq, and Iran among others, seeking third-country resettlement. A third category, much broader and more contentious than the other two, includes a large spectrum of irregularized migrants, such as apprehended migrants, those kept in removal centers, rejected IPS applicants, those pushed back by the Greek authorities, transit migrants, and those who have never registered. The scope of the third category, which concerns those without a legal status, is totally unknown and still expanding given the cracks in the registration system; this is explored further in Section 2. To illustrate the magnitude of this, the Migration Board Meeting of the Ministry of the Interior announced on September 15, 2021 that 1,293,662 irregular migrants had been “apprehended” in Turkey between 2016 and September 2021.¹² According to the same statement, Turkey deported a total of 283,790 of these “apprehended” migrants back to their countries of origin during the same period. The number of people deported in the first eleven months of 2022 was 101,574, according to official figures.¹³

According to this condition of layered legal precarity that regulates millions of displaced persons, there is no right of permanent stay but only temporally and spatially restricted movement. Syrians under TPS constitute the relatively privileged group within this period of prolonged uncertainty. Temporary protection policies are not new in refugee law, especially when a large number of people are fleeing an armed conflict or a disaster – given the limited scope of the 1951 Refugee Convention.¹⁴ In such cases, temporary protection policies designate complementary or exceptional forms of protection so that states are able to provide an immediate response to a humanitarian crisis. Yet in the Turkish case, almost a decade after its initial implementation, TPS is still used as a legal status that designates up to 3.2 million Syrians (as of March 2024) but does not provide any specified time frame. A person from Syria is required to approach the provincial directorates of the PMM, in a province of her own choice, to receive a temporary protection identity card. Once provided with an ID, she can benefit from the right to access to healthcare and education, but does not yet receive a work permit. The 2016 “Regulation on Work Permits of Foreigners Under Temporary Protection” gives Syrians access to work permits under TPS, while the regulation only allows employers to initiate the procedure, thus leading to the minimal numbers (63,789 by 2019) of Syrians who have obtained a work permit.¹⁵ As a result, the fate of the vast majority is in the cruel hands of the informal labor market.

Syrians under TPS can choose their province of registration provided that it is open to new registrants. However, once they are registered, they are bound to that province and require official permission from provincial authorities to be able to move between cities. Concurrently, their access to healthcare, education, and any other public services is restricted to their province of registration. This regulation vividly illustrates how the spatial configuration of movement defines the core of migration governance, in tandem with the legal production of temporality. In her influential book, Hagar Kotef (2015: 5) notes that neither subject positions nor political orders can be divorced from movement regimes, which have been differently configured and materialized within sets of “material, racial, geographic and gendered conditions” of those subjects. Consequently, regimes of movement are never simply a form of controlling or regulating mobility, but rather produce different categories of subjectivity, that is, ways of being, defining the contours and limits of subjects. The mobility gap between citizens and displaced persons, marked by the city-based registration system, demarcates the spatial boundaries of legality as well as the spaces in which they move (Shamir, 2005). Therefore, city-based registration corresponds to, in Torpey’s words (Torpey, 2018), the “legitimate means of movement” in the Turkish legal architecture.

A similar apparatus of closure or inward borders applies to non-Syrians under IPS, though in a stricter and more precarious manner. As defined in the LFIP, the PMM assigns those seeking international protection from non-European geographies to satellite cities – from which large metropolises and provinces located on the European borders are omitted.¹⁶ As long as they reside in their assigned cities, they can access public services and healthcare, but this is limited to one year following their registration. Again unlike Syrians under TPS, they have to register and reside in those designated satellite cities without having any prior knowledge of them or of relatives there. It should be noted that these satellite cities are not camp-like spaces, but rather small open cities that accept applications for international protection. Applicants have to stand on their own feet, by seeking a job in the informal sector. If an application is considered positively, the applicant is granted “conditional refugee status” (eligible for third country resettlement) or “subsidiary refugee status” (not eligible for resettlement but with notable risks of persecution in case of return). Another prominent discrepancy between Syrians under TPS and non-Syrians under IPS is that the latter have to provide proof of their presence to the provincial authorities on a weekly basis while the former do not.

Following this illustration of the spatial organization of the migration and border regime in Turkey, with its legal basis embedded in a temporal architecture, Section 2 discusses the ways in which this legal infrastructure becomes a self-undoing mechanism.

2 Demarcations Overstepped: Productions of Irregularities

“What we are facing is not only irregularity of thousands of displaced persons, but also irregularity of regulars,”¹⁷ said one of my interviewees, a lawyer, who pointed out the striking discrepancy between the legal architecture and daily transgressions. Despite the legal design that is aimed to limit both Syrians under TPS and non-Syrians under IPS to their provinces of registration, they tend to move to other cities, without travel permits, in order to find jobs in the informal sector. Because their legal statuses do not grant work permits, migrants face two options: either a life with legal status or a job.

The lawyers I interviewed frequently drew attention to the institutional changes that occurred in the aftermath of 2018. As of September 2018, the United Nations High Commissioner for Refugees (UNHCR) terminated its operations, and subsequently, the processes of first registration, refugee status determination, and resettlement have been entirely under the mandate of the PMM. Following this institutional shift, new registrations were halted in metropolitan cities owing to the overwhelming burden of internal mobility (European Council on Refugees and Exiles [ECRE], 2020). As a result, Syrians under TPS were deprived of the opportunity to transfer their registration to the larger cities where they informally reside and are employed. For non-Syrians under IPS, larger cities have never been considered to be satellite cities, so all applicants have been guided to the assigned small cities. A similar scenario is valid for them. Moving to bigger cities, even though it is prohibited, has become the only way in which many can survive, given the one-year limit on health services and the shortage of jobs.

The lawyers I interviewed underlined the devastating implications of closed registrations in large cities for Syrians, which had been previously tolerated to a large extent. As noted, until 2018, Syrians under TPS easily accessed healthcare and education in Istanbul even if their province of registration was different. With the hardening of migration policies in the metropolitan areas, “a great crisis has appeared at the door,” as an interviewee noted:

There has been a huge problem regarding the joint children of the spouses in different provinces. Marriage and family reunification is close to impossible in Istanbul […] The parents are registered in Istanbul, there is a minor mistake of a letter in the identity information, then the registration process is reset […] There are also problems with child registration. If one of the parents is registered outside of Istanbul, the approach of the PMM is to refer them outside of Istanbul, that is, to make the family reunification in another province. In the cases of severe illness requiring treatment in a university hospital, again other provinces are prioritized. Since about 2018, Istanbul has been following a very strict registration policy.¹⁸

Strict registration policies in metropolitan cities result in Syrians under TPS falling into de facto irregularity owing to their inability to change their city of registration. Though the exact number is not clear, a recent field study conducted by the PMM and IOM in 2018–2019 estimated that approximately 1 million Syrians were residing in Istanbul with only 496,000 of them registered in the city.¹⁹ Hence, nearly half a million Syrians under TPS became irregularized in 2019.

The satellite city system for non-Syrian IPS applicants gives us an even more precarious picture. Testimonies of lawyers demonstrate that the vast majority of cities have been closed to registration; applicants have to go continuously from one city to another to access registration. It is officially uncertain which satellite cities are available for new registrations because of the lack of coordination between provincial authorities. “We, as lawyers, put an Excel tab on the Google Drive, and change the color of provinces as yellow or green to mark whether they are open or not. This can change overnight,” noted my interlocuter.²⁰

The obstacles are far from over when registration finally takes place. The applicant finds herself in a small city to which she has been assigned without any guidance regarding accommodation, job, or aid. Hence, she becomes trapped in “chronic waiting,” as Khosravi (2014) puts it, for resettlement. This is often a remote possibility in the distant future owing to the astonishingly small quotas for the countries in the Global North:

When the UNHCR suspended all operations, the fate of IPS was plunged into complete uncertainty. Some cities are closed, some receive, and others do not, it is totally unclear. Third country resettlement, despite obstacles, was at least pursued under the UNHCR until 2018, but when the mandate changed, it was completely turned into a lottery; who gets the resettlement message no one knows […] I applied for IPS last year in the city of Uşak, for my LGBTIQ+ Iranian client. It was said that they would call for an interview in two weeks [refugee determination status interview]. No call. We approached again after two months. We trusted them that they would call at some point. It has been two years now.²¹

Another lawyer points to the blindness of the international community and the UNHCR who have entirely abandoned the persons seeking IPS:

The purpose of the satellite city was related to conditional refugee status and resettlement in a third country. People were kept under control with the aim that this would increase their chances of being resettled. However, as the numbers increased, the resettlement decreased, and now it no longer works. It is almost nonexistent. The UNHCR proudly announced that it had resettled 5,000 families in 2021. Really? Five thousand among hundreds of thousands of persons? After seeing the news, I was deeply embarrassed on behalf of them […] There is no international protection in operation here […] As a lawyer, I just laugh at the social integration projects which were pursued for ten years. If there is no registration, then you cannot talk about integration. This is not an issue of humanitarian or education aid; they [the UNHCR] see it is not working. Maybe they saw that it would not work, and they terminated their operations.²²

Closed registrations in satellite cities constantly produce irregularity: Job shortages, lack of access to healthcare (limited to one year), and protracted waiting push registered asylum seekers, who are required to sign in every week, to move to metropolises at the expense of their temporary legality.

Consequently, both for Syrians by the million under TPS and non-Syrians by the thousand under IPS, the city-based legal system has become a self-sabotaging mechanism, that gives rise to the arbitrary exercise of power and triggers considerable internal movement. My interviewees pointed out that the PMM, as the sole authority in migration governance since 2018, has rationalized that irregularity and arbitrariness are the new rules, both in the sense of legal protection and the mode of governance. One lawyer referred to Franz Kafka’s novel The Castle when he illustrated the modus operandi, a view shared by my other interviewees:

PMM is not a state institution that is legally ascribed, impersonal, has a corporate culture, does not hinge on personal decisions, and is bound by the legislation. There is no corporate precedent. It is completely personal and arbitrary, whose procedures and decisions change from day to night. It is a group of people that we, as lawyers, cannot deal with. We cannot even make them implement court decisions.²³

3 Cycles of (Im)mobility: Floating Populations

In explaining the conditions of deportees in France, whose lives are reduced to a forced circular mobility through prisons, detention centers, and public space, Carolina Sanchez Boe (2017) uses the term of floating populations. Nicholas De Genova (2021) furthers this discussion by referring to “the carceral circle,” alluding to Foucault (1995), and delves into how rejected asylum seekers or illegalized migrants become susceptible to “a repetitive cycle of rejection and detention, expulsion and capture” (Genova, 2021: 192). While these theoretical attempts aptly characterize the disposability of migrants’ lives, they all focus on countries on the Global North, in which the number of cases stands dwarfed in the face of the millions of displaced persons under the temporal lacuna in the Global South, particularly in Turkey.

Here, I seek to extend the scope of the term “floating populations,” by drawing on Turkey. I argue that this showcases a stretching of the term in terms of speed and magnitude by means of numerous modes of governance that at times seem paradoxical. The turning points to be explored in this section, though taking place within a very short period since the 2016 EU–Turkey Deal, correspond to episodes of tightening and loosening of borders in line with the changing political conjunctures and priorities of labor markets.

Following the closure of new registrations in metropolitan cities in 2018, the provincial governorate of Istanbul announced another major policy turn in July 2019. This corresponded to the defeat of the AKP government in the municipal election in Istanbul for the first time since 1994, one of the principal reasons being growing public uneasiness about the overwhelming mobility of Syrians and non-Syrians toward the city. The previous de facto policies of toleration were terminated overnight and the number of police checkpoints dramatically increased across the city. Reports indicate that even though the law (LFIP) does not mandate it, Syrians under TPS and non-Syrians under IPS were detained in large numbers because of their violation of residence restrictions; they were registered in other cities (ECRE, 2020). Official announcements indicated that some were sent back to their provinces of registration and 315,000 Syrians “voluntarily” returned to their country.²⁴ Published reports question the element of voluntariness in these return procedures (Amnesty International, 2020).

According to one of my interviewees, “that was the period when the policy of dilution and sweeping, as they call it, came into force.” My lawyer respondent continued:

Persons were intercepted from various parts of Istanbul; targets were given to the police such as “you will reach thirty to forty persons a day or a target figure per week.” There is long list of those apprehended: those who were not registered in Istanbul but in a different province, those who were totally unregistered or undocumented; those who were involved in a crime, those who had problems with their identity documents, and those who had been found to have false statements on their IDs. This is done by internal circulars mostly, and we do not even see some of these circulars.²⁵

The process lost momentum over six months and restrictive policies were slightly weakened, as lawyers testify. People who were sent back to their provinces of registration or even to Syria often sought to find ways to return. Resonating with how Franz Fanon (2004) describes the rationalities of power exercised upon colonized subjects, in which “confused by the myriad signs of the colonial world, he never knows whether he is out of line,” lines are constantly shifting in the Turkish migration regime, and people suddenly find themselves to be violating the law.

The next episode opened with an unprecedented moment that implied de facto infringement of the EU’s outsourced border policies and engineered migration at the Greek-Turkish border (İşleyen & Karadağ, 2023; Karadağ & Bahar, 2022). On February 27, 2020, another announcement came from the government, stating that Turkey had not received adequate financial support from the EU and would no longer stop displaced persons aiming to reach Europe (Human Rights Watch [HRW], 2020). Right after the announcement, during a single night, the Turkish police, gendarmerie, and border guards were ordered to stand down and let thousands of people rush to the Pazarkule border gate in Edirne at the Greek–Turkish land border. As cited in the reports, the testimonies of nongovernmental organizations who were present in the area hint that removal centers were emptied across the country, with undocumented persons without legal status constituting the vast majority, approximately 13,000 people, at the border gate (International Organization for Migration [IOM], 2020; Karadağ & Üstübici, 2021). However, there were also Syrians under TPS and non-Syrians under IPS, albeit the minority. People stayed in the buffer zone for a month while they were brutally beaten, assaulted, robbed, and pushed back by the Greek security forces (HRW, 2020). By the end of March 2020, Turkish authorities decided to clear the zone, this time by evacuating the tents and transferring people back to removal centers in random cities. Since these events coincided with the COVID-19 outbreak, people were quarantined in removal centers and dormitories for fourteen days. Strikingly enough, testimonies reveal that they were released from the removal centers in cities in which they did not reside or were registered after quarantine in the midst of total lockdown. None of my respondents had any information about what had happened to those under TPS or IPS, that is, whether they lost their existing status or not. The Pazarkule events, symbolizing an orchestrated mobility at the borders, exacerbated the already existing precarity of displaced populations. Since then, the city of Edirne at the Greek–Turkish border has been witnessing hypermobility that circulates around Istanbul, border villages in Edirne, and the Evros river, with many eventually being pushed back by Greece²⁶.

When it comes to people without any legal status, production of a floating subject materializes in microepisodes every day of their lives. As subjects on the move and deprived of any rights, they constantly rotate from one city to another, despite intercity mobility restrictions, going wherever jobs are available and constituting the cheapest labor force in the country. Some seek to reach Europe but then get pushed back, continuing their journey across the country and taking part in the informal job market until their next attempt to enter Europe or their detention or deportation. These two latter acts might come at any time in their daily lives, with indeterminant periods of containment following.

There are currently twenty-eight removal centers in Turkey, operating in different twenty-four cities and holding approximately 20,000 persons.²⁷ The construction of new centers and refurbishment of reception centers as removal centers have been undertaken using the EU budget as part of the EU–Turkey Deal.²⁸ From a legal perspective, when an undocumented person is intercepted by the police on the street, they are first taken to the police station. After assessing fingerprints and an identification process, the police transfer files to the PMM as the institutional body that determines whether administrative detention (six months maximum to be extended for an additional six months) or deportation is called for, based on the scope of Articles 54 and 57 of the LFIP. According to the lawyers, this procedure, aside from its arbitrariness, contrasts with the state of administrative law from the very beginning:

[T]he first problem in terms of legality is that the decisions of administrative detention are taken by an unauthorized institution, which is the PMM […] Lawlessness starts from here: The law says that the decision of an administrative detention shall be taken by the Governor’s Office, but in effect it is the PMM that takes the decision. It is based on the transfer of authority protocol made between the two offices. I constantly voice the unlawfulness of this practice and make legal objections, but they are ignored by the courts.²⁹

Once someone is put under administrative detention in a removal center, nobody knows what will happen next. The duration of stay and whether they will be released or deported are unknown and subject to numerous dynamics: the ability to communicate with officers in Turkish, whether or not they are pressured into signing the voluntary return form, the financial arrangement of the charter flights (supported by the IOM budget), the occupancy rate of removal centers, the number of available seats in the buses taking people to another removal center, and so on.

It needs to be underlined that the very act of detention is not exercised for its actual outcomes, but rather it is the potential for detainability, as De Genova (2007) aptly puts it, that operates as a disciplinary form of power. The potential constantly propagates fear, which is not equally distributed. For instance, the testimonies of undocumented Afghans indicate that “proper” looks, proficiency in the Turkish language, and an ability to keep away from trouble might increase the chances of avoiding detention on the street (Karadağ, 2021). Police checkpoints do not often include workplaces involving construction, manufacturing, restaurants, and sheepherding areas, where displaced persons are generally hosted. And once they are put in detention, some people are released in weeks or months while others are deported. On the official website of the Ministry of Interior, it is stated that 320,172 foreigners have been deported (as of April 2022) “within the scope of the fight against irregular migration” since the EU–Turkey Deal was signed in 2016.³⁰ For the lawyers I interviewed, the final verdict is predicated upon the signing of the voluntary return form:

I had two Afghan clients aged eighteen to nineteen; they stayed in Tuzla removal center (in Istanbul) for about two and a half months. The authorities interviewed them regularly every week, tried to persuade them that as there was a search warrant against them, they could not live in Turkey. The clients were patient for two months and did not sign. They could not deport them, then they got them on buses with other 150 people and dropped them off in Edirne [a city at the Greek–Turkish border]. Nobody crossed the border, everybody came back.³¹

For those who have been given a deportation decision, another type of forced mobility can occur: movement between removal centers across the country:

[Though they are detained in Istanbul] there is no official record in their names in Istanbul. They are sent to other cities. Deportation orders for them are made in those cities. After a person is caught, the family usually receives the news when they are taken to another city. Tuzla removal center is also used for the purpose of transfer. There is no payphone, no lawyer, no notary public, no file. Because the file will be created in the province where they are taken, you can hear from the person a week later. You receive a call from the other side of the country, a couple of hours before the deportation.³²

Since January 2022, especially after the general elections in May 2023, Turkey has entered another episode, which is to a large extent the institutionalization of the return policy. Since the beginning of a series of campaigns on social media launched in 2021, there has been use of phrases such as “there is a deadline for hospitality” and “this is an invasion.” Following the surge of public reaction against refugees in the summer of 2021, the process of dilution and sweeping is now at its peak. Research dated 2021 shows that the most pressing issue for the public is the economy, while the question of refugees comes second (rising to 17.9 percent from 6 percent in 2020).³³

The current political atmosphere in Turkey has a direct effect on evermore intensifying the speed and magnitude of forced mobility of those holding various legal statuses. As part of the dilution and sweeping policy, numerous provinces and districts were completely closed for registration renewal beginning in February 2022.³⁴ Furthermore, Syrians under TPS who do not reside in their province of registration (they are in the thousands, as previously noted), were given forty-five days to reregister in their provinces and notified that their IDs would otherwise be cancelled. “We have entered a period in which temporary protection status has been cancelled very intensely,” noted my lawyer interlocuter. Concomitantly, for undocumented persons (non-Syrians), the PMM announced that deportations were being radically stepped up. They added that the rate of deportation had gone up to 50 percent of “irregular migrants apprehended,” compared with 18 percent in Europe.³⁵

The repetitive cycles of im/mobility of displaced populations in Turkey demonstrate the inconceivable number of moving subjects, in the millions, exist in perpetually shifting spatialities and temporalities. The episodic policies turn into a self-sabotaging mechanism whose modes of operation give rise to the materialization of arbitrary power. The borders are changing nonstop while disruptive uncertainty and arbitrariness become the rules.

Concluding Remarks

In his book, Pascalian Meditations, Pierre Bourdieu (2000) notes that “absolute power is the power to make oneself unpredictable and deny other people any reasonable anticipation and to place them in total uncertainty.” Bourdieu’s description of “absolute power” here does not necessarily account for the absolute coercion enforced by the sovereign, but the powers that create a wide range of governing techniques that constantly produce precarious subjects. The case of Turkey epitomizes the level of instrumentalization of millions of displaced bodies that are trapped within a temporal lacuna. As their mobility is blocked by European countries, they are sentenced to waiting for indeterminate periods in Turkey, periods that might even exceed their lifetimes.

By turning the spotlight on the EU’s contentious frontier, onto the world’s largest refugee hosting country, this chapter not only shows the devastating fallouts of Europe’s externalization policies that lead to neighboring countries bearing the burden of refugee movements, but also describes the ways in which those countries in turn capitalize on displaced bodies and turn them into a political weapon.

As the political process of disenfranchisement continues, Turkey is sliding ever deeper into realms of lawlessness, and even her own citizens verge on becoming “rightless” subjects in terms of freedom of speech, human rights, and justice. Transforming millions of displaced people into irregularized and floating subjects is accompanied by the native population’s feeling that it is being left to its own fate, while the long-brewing collective anger towards noncitizens is about to boil over. In a context within which civil, political, and social rights are seriously hampered even for its own citizens, we witness a growing affinity between citizens and noncitizens in terms of economic deprivation, civil and political freedoms, and legal precarity. The situation does not result in the total erosion of citizen/noncitizen distinction, but the conditions of lawlessness and hence the undermining of rights lead to the subjection of both the former and the latter to the ad hoc aspirations of decision-makers in Turkey.

13 UNHCR and Biometrics Refugees’ Rights in a Legal No-Man’s Land?

Marie-Eve Loiselle

Introduction

For a long time, the border was the site where states have controlled foreigners’ access to their territory. In the last few decades, however, governments have increasingly sought to relocate away from their territorial limits border practices that serve to identify, filter, and, if necessary, prevent crossings into their national space. Digital innovations, including biometric technologies, are now offering new opportunities for making migrants legible from afar, miles away from “the legal gates of admission” (Shachar, 2020b: 9). For instance, it is common for states to collect travelers’ biometric data at the time of application for a visa in the country of origin. Yet the turn to emerging technologies is not the preserve of governments. The United Nations Refugee Agency (UNHCR) has been using digital technologies, including biometrics, for several decades. For UNHCR, biometrics, first deployed in the Middle East, is becoming an essential tool in accounting for the populations under its protection. This chapter explores this development and the associated risks for refugees. The first part surveys the emergence of states’ practices toward the deterritorialization of border controls with considerations for the enabling role of biometrics and digitization in that process. Then it demonstrates how the use of biometrics by UNHCR maps onto these states’ practices. The second part considers the consequences of UNHCR’s practices surrounding the biometric registration of refugees. It discusses the risks posed by the collection of biometric data in the refugee context before assessing how the institutional and structural conditions in which UNHCR operates, especially with regard to consent, accountability mechanisms, and legal safeguards, may undermine refugees’ control over their data.

1 Border Control, Identification, and Digital Technologies

The assumption that the border marks the beginning and the end of a state’s effective sovereignty over its territory is deeply entrenched in the Westphalian sovereignty model. This is why governments worldwide have identified borders as the locus of control over movement in and out of their respective space of authority. In that respect, the border and its associated administrative practices of admission and exclusion work like a sieve to filter beings, matters, even ideas, letting in what is good and rejecting the rest following a subjective assessment about national interests. Yet the border–control nexus is only one aspect of a larger paradigm of sovereign authority. Indeed, states increasingly opt to multiply the locations of (border) control practices at various sites beyond and within their territories. Scholars of border studies have articulated the modalities of an emerging mobility regime that aims to screen people on the move as early as possible into their journey. Ayelet Shachar explains how states, relying on legal fictions, are redrawing the spatial limits of their authority to control migration and movement through the ‘shifting border’ (Shachar, 2020b). As she describes, when screening migrants, states project their border practices as far outward as possible from the state territory. By contrast, when assessing the validity of rights claims, that same fictional legal border retracts back within the state territory.

Practices of “remote border control” are not new. The transatlantic visa system developed in the twentieth century was an early form of migration control at a distance (Zolberg, 1997: 308–309). What is changing is the sweeping scale at which external controls are now deployed and the granular nature of the information collected about people afforded by technological innovations. Biometric technologies are becoming the new norm for identifying and screening people. Instead of using biographic data, such as name, nationality, and gender, biometrics uses the body (e.g., fingerprints, iris scan, facial recognition, gait, DNA) as a source of information and token of identification. The data produced in such a way are distinct from other forms of personal information. They are not information about a person but from a person (Smyth, 2019: 50). They constitute the behavioral and biological material or characteristics of the individual from which they are collected.

Despite being part of someone’s make-up, biometric data are severed from the body. Through its datafication, the subject is disembodied. Parts of its unique and unalterable composite travel across space unencumbered by a constructed cartography of sovereign powers. Yet if data’s potential for mobility knows few limits, the “unterritoriality” of data (Daskal, 2015) does not signify the end of territoriality. On the contrary, datafication serves territorial practices of control by making possible what was before inconceivable: the assessment of someone’s identity when she is at a distance, far from the state’s territory. Rather than the demise of the border widely forecasted in the 1990s, data and digital technologies enabled the erection of a more insidious digital border, where only the elements of the person’s identity necessary for assessing the validity of a claimed right to enter are allowed to travel.¹ These data include immutable biometric features nearly universal to all human beings yet unique to each individual, which makes them highly accurate and reliable for identification purposes. They are easily deployable, transferable, and duplicable to different contexts and localities, such as airports, land borders, and refugees’ camps when needed (Smyth, 2019: 51). What’s more, when combined with other forms of data, the breadth of knowledge they may impart about a person’s past and present – perhaps even her future – multiply.

Hence, a key attribute of the digital border is that it makes proximity expandable without undermining legibility. The physical presence of the person seeking entry into foreign territory – a requirement of border control in an analogue world – becomes superfluous. Instead, it is the markers of one identity that travel through servers and databases to reach the authorities of the potential host country to offer a reading of the risks and benefits of her entry. This physical distance, in turn, appeases acute fears underpinned by the perceived threats posed by immigration and transnational violence – what Ronen Shamir (2005) coined the “paradigm of suspicion” – by maintaining suspect individuals at bay. Distance is compelling for another reason. Indeed, it shields a state from its obligation of nonrefoulement provided for in the 1951 Refugee Convention and other protections found in international human rights law instruments, which, as Dana Schmalz explains in Chapter 4, require the refugee’s physical presence or proximity to the state territory. The digital border is therefore key to foreclosing the required proximity for right’s claims and becomes increasingly efficient at doing so as the number of foreigners’ databases, identity management systems, and biometric identity documents increases.

For these reasons, states employ biometrics for identification purposes in ePassports. These collect the bodily attributes of foreigners in databases such as Eurodac in Europe and US-VISIT (now the Office of Biometric Identity Management) in the United States. National identity management systems using biometric and biographical data are also growing in popularity. States increasingly rely on them to ascertain welfare assistance and public service rights. When linked to travel documents, these databases have the potential to deliver on both fronts of the shifting border described by Shachar, screening those standing outside the gates of admission as well as those already in. As such, biometric technologies are allowing states to expand and rearticulate their political authority over their territory, further reinforcing their monopoly on the “legitimate means of movement” (Torpey, 2018). Yet states are not the only proponents of identity management systems. A host of international bodies have developed their own identity management architectures supported by biometrics data – or promoted their deployment – including the UNHCR.

2 UNHCR’s Turn to Biometric Technologies

UNHCR was an early adopter of biometric technologies when it undertook in 2002 to collect the iris scans of all Afghan refugees returning to Afghanistan from Pakistan with its assistance to prevent multiple claims by the same person (Kessler, 2003). The Afghan experience was a precursor to UNHCR’s Population Registration and Identity Management EcoSystem (PRIMES), launched in 2015. PRIMES is a digital platform that aims to bring together the host state’s civil registry, UNHCR’s registration identity management databases, as well as several digital tools, available or planned, to connect UNHCR, its partners, and refugees. A key component of PRIMES is its Biometric Identity Management System. This biometric database contained the fingerprints and iris scans of 15.7 million individuals at the end of 2023, up from 8.8 million in 2019 (United Nations High Commissioner for Refugees [UNHCR], 2020: 9; 2023). As the Agency explains, the biometric registration of refugees has several benefits. It facilitates UNHCR’s efforts to address the needs of displaced populations worldwide. Biometrics enrolment can reduce fraud, cut down processing time, and provide a more accurate overview of the refugee populations. It also provides proof of identity for affected populations in the absence of formal identification documents and gives them timely access to services and resources in innovative ways.

Broader political and structural factors have contributed to UNHCR’s turn to biometrics. As legal anthropologist Sally Engle Merry explains, the audit and performance culture typical of the corporate world has spread to global governance (Merry, 2011: S84). Since the late 1980s, a result-based management paradigm has forced organizations to demonstrate the “value for money” of their programs (Sandvik, 2016: 138). Agencies such as UNHCR must now quantify their achievements and demonstrate their efficiency to secure evidence-based funding. In that context, biometrics provides the raw material for data production and the development of statistical measures about beneficiaries, aid distribution, and other indicators of performance used to allocate funding (Madianou, 2019: 586). The importance of biometrics as a tool to measure the Agency’s effective use of its budget is reflected in UNHCR’s Grand Bargain commitment to “expand the use of biometrics for refugee registration to a total of 75 country operations by 2020” in order to reduce duplication and management costs. The influence of donors in the Agency’s adoption of biometric technologies for registration is also evident in its statement that iris scanning was “important in securing the support and confidence of donors” (Troger & Tennant, 2008: 3).

Another factor that helps explain the development of UNHCR’s digital identity architecture is the trend toward the securitization of migration, which intensified at the turn of the twenty-first century, along with the growing reliance on biometric data for filtering the movement of suspicious populations. Following September 11, 2001, suspicion and hostility against refugees rose to unprecedented levels (Betts, Loescher, & Milner, 2011: 62). This prompted northern countries to adopt a wave of measures to curb arrivals, cut down resettlement programs, and contain refugees in their region of origin. Host countries in the Global South grew equally concerned about the perceived security threats posed by large influxes of refugees (Betts et al., 2011: 60). In this context, UNHCR has had to balance its responsibility for protecting refugees with pressure from host and resettlement countries to enhance the quality of the data it holds about refugees and share that data with them. On the one hand, UNHCR, as an “invited guest,” must try to accommodate requests from host governments authorizing it to operate within their territory (Wilde, 1998: 113). On the other, the Agency cannot ignore donor states’ demands for better data on refugees and spending, with voluntary contributions representing 85 percent of its budget (UNHCR, 2022a).

UNHCR’s response to states’ concerns has been to step up the biometric registration of refugees. It also answered the call of northern countries for improved biometric data-sharing in the resettlement process. For instance, in 2004, the US Department of State noted in a report that “assurance of positive identification via biometrics throughout the refugee assistance process and especially the resettlement process would carry enormous advantages in the post-September 11 climate” (United States Department of State, 2004). The process was officialized in a 2019 Memorandum of Understanding (MoU) between UNHCR and the US Department of Homeland Security (DHS) for the sharing of refugees’ biometric data. Under the MoU, UNHCR agrees to transfer directly into the DHS Automated Biometric Identification System the biometric and associated biographic data of refugees it refers for resettlement in the United States. As the DHS recognizes, under the MoU, the United States could come into possession of data from individuals who will never set foot in the United States for various reasons (e.g., rejection or withdrawal from the resettlement process).

Finally, the adoption in 2015 of the Sustainable Development Goals and its target 16.9 calling for a “legal identity for all by 2030” has contributed to the expansion of digital identity management systems into the humanitarian space. The appeal of legal identity is premised on the assumption that the state’s recognition of an individual’s existence – and evidence of such recognition – is necessary to access social services and the enjoyment of rights. Since its adoption, target 16.9 has been a catalyst for technology-based solutions to legal recognition that use biometric technologies for proof of identity. This is reflected in UNHCR’s Strategy on Digital Identity and Inclusion, which sets the objective of “achieving the digital inclusion and digital identity of refugees, stateless persons, and other forcibly displaced persons.” With its “state of the art biometrics,” the Agency explains, PRIMES can make a meaningful contribution to target 16.9, noting that “‘[l]eaving nobody behind’ applies to all countries and sectors of society. Even to the digital space.” The Global Compact on Refugees reflects this commitment when it provides that UNHCR “will contribute resources and expertise to strengthen national capacity for individual registration and documentation” (Sköld, 2021), including digitalization, biometrics, and the collection and sharing of quality registration data. Narratives describing these initiatives focused on digital inclusion and better services to refugees. Yet underneath these lofty goals another logic appears to be at play, which speaks to the aspiration of certain states to create a spatial distance between their shores and individuals seeking refuge without giving up on their capacities to scrutinize and screen them.

Hence the motivations – and justifications – for the use of biometric technologies are twofold. On the one hand, and perhaps most evidently, the datafication of refugees is seen by state actors as the most reliable tool for the identification of people crossing their borders. As UNHCR’s identification management system evolves, it will likely become more efficient in tracing refugees’ journey from registration onward. This potentiality explains why such a form of indelible identification has been compared with the tagging or tattooing of populations (Agamben, 2004). On the other hand, proponents of biometric registration are discursively constructing it as necessary for someone’s recognition as a person before the law through a legal identity agenda, emphasizing values such as justice, equality, and dignity. Legal identity, supported by biometrics and digital platforms, is cast as an empowering instrument that holds the potential for opening access to humanitarian aid and services. Yet it is important not to fall prey to an instrumental use of the language of human rights. As Martin Krygier (2006: 136) warned, “hurrah words,” such as democracy, the rule of law and human rights, are endowed with virtue making it difficult to challenge any proposals pertaining to secure these goals. Nonetheless, we must examine what may hide underneath the rhetoric and consider how the means proposed to achieve these goods can deliver on their promise of empowerment.

3 Challenges Posed by UNHCR’s Biometric Practices

Most citizens and travelers perceive the collection of their personal data with unease about being subjected to identification practices long associated with crime control (Gilman, 2012: 1394). This said, apprehension about data collection is soothed by the apparition of smart travelers’ programs that associate biometric with privileged status. However, for refugees and other groups with precarious legal status, the collection of biometrics holds no such benefits and may result in harms that have no equivalent in different contexts. One risk refugees confront is that their country of origin may access UNHCR registration data without their knowledge or consent. The risk is not theoretical. Human Rights Watch (HRW) notes that between 2018 and 2021, the Bangladeshi government transferred the biographical and biometric data of 830,000 Rohingya refugees to the Myanmar government to assess their eligibility for repatriation. The data transferred had been collected during a joint UNHCR–Bangladeshi registration exercise. Testimonies by Rohingya refugees have since been collected to the effect that they were unaware their data could be shared with Myanmar. One individual explained to HRW that although he was informed of the potential transfer, he felt obliged to agree in order to access aid: “I could not say no because I needed the Smart Card and I did not think that I could say no to the data-sharing question and still get the card” (Human Rights Watch [HRW], 2021). Although UNHCR insists it did not violate its policy on personal data, the incident raises doubt about the feasibility of obtaining free and informed consent in humanitarian contexts.

Another risk is that biometric data be used for initially unforeseen purposes or have unexpected effects on individuals, sometimes even beyond the refugee population. In the late 2000s, UNHCR introduced the biometric registration of refugees in Kenya. In 2016, the Agency handed over the responsibility for the registration of refugees to the Kenyan government, and in the process, data from UNHCR’s biometric database were integrated into the national register of persons. The operation revealed that an estimated 40,000 individuals registered with UNHCR were, in fact, Kenyan citizens. Among them, several minors at the time of registration were refused a Kenyan national identity card when reaching eighteen years old, the state claiming that they were aliens because their fingerprints were stored in the refugee database (Haki na Sheria Initiative, 2021: 31). Deprived of this document, Kenyan nationals are denied citizenship rights and privileges such as employment, freedom of movement, and education. For them, it is not the border that stands between them and the privileges of citizenship but their own biological markers.

These two examples illustrate the profound implications that may arise from UNHCR biometric registration practices. They also emphasize the importance of free and informed consent and redress mechanisms for preventing and responding to harm resulting from data processing. To be sure, this is not only true in humanitarian contexts. The collection of biometric data by states for identification purposes presents intricate legal and social challenges that are increasingly being questioned in public fora and national courts. Yet the legal environment in which UNHCR operates, as described earlier, bars most pathways for debate and contestation that would otherwise be available. The growing reliance on UNHCR to provide digital identities to refugees through biometric registration relocates data extraction and processing in an institutional and geographical space that offers an exceptionally enabling environment for such practices.

Legal scholar Julie Cohen (2019) notes that the emerging informational economy model of data extraction challenges the law and existing legal institutions. Increasingly shaped by private interests, digital practices thrive in regulatory settings that favor opacity, informality, and standard-based governance rather than legal rules. The subjugation of consent for collecting digital data sustains this model further. Considering UNHCR’s accountability regime, one cannot help but notice the striking commonalities it shares with the regulatory framework idealized by digital actors in the market economy. Governance by standards rather than laws, self-regulation through codes of conduct and guidelines, and, significantly, immunity from prosecution before national courts are inherent to the Agency’s regulatory environment. In short, delegating the biometric registration of refugees to UNHCR’s transfer datafication practices into a jurisdictional setting where safeguard and accountability against breach to data protection are weak.

a Doubtful Consent

The issue of consent is central to discussions about new technologies and the processing of personal data in digital forms. Biometrics is data about a person’s physical, physiological, or behavioral characteristics. These characteristics are unique to an individual and cannot be modified. Because of the sensitivity of these data, acquiescence or consent to the collection of biometrics is usually seen as essential to validate their processing. Individual consent gives moral legitimacy to the collection and use of data. It reflects the agency and autonomy of the person over how their data are collected, used, and shared. From a legal standpoint, a person’s consent transforms into a lawful act what would otherwise be the wrongful action of another. Indeed, collecting and processing personal data is considered an intrusion into someone’s private sphere that requires her “transformative act of consent” to be lawful (Schermer, Custers, & van der Hof, 2014: 174). Consent is valid when given freely, that is, voluntarily without undue pressure or influence. The European General Data Protection Regulation, for instance, specifies that the individual consenting must have a real choice to accept or refuse the processing of her data. Consent must also be informed; to that end, the entity collecting and processing personal data must disclose who is collecting which data, for what purpose, and with whom it will be shared: “Anything less than this requires a leap of faith” (Barocas & Nissenbaum, 2014: 58).

While the legal and ethical foundations of consent remain the same in the humanitarian context, considerations specific to refugee populations complicate the issue, especially with regard to consent’s voluntary element. In principle, UNHCR relies on consent as the legitimating basis for processing personal data. Yet scholars and refugee advocates have questioned whether refugees seeking UNHCR assistance can truly give their free consent when humanitarian aid is conditional on the collection of biometric data. Indeed, free consent may require not only the absence of constraint, but also the presence of enabling conditions, including basic material support for subsistence (Gould, 2019: 182).

On the issue of consent, UNHCR’s policies lacks clarity. The general rule is that everywhere it is possible for the individual to exercise free and informed consent, it should be the legal basis for collecting and sharing biometrics. UNHCR’s Guidance on Registration and Identity Management (2018) provides that individuals can refuse the collection of their biometrics on legitimate grounds and maintain their right to international protection using alternative methods for identification and registration. At the same time, UNHCR’s 2015 Policy on the Protection of Personal Data of Persons of Concern to UNHCR (hereinafter UNHCR Policy) and its attendant 2018 Guidance on the Protection of Personal Data of Persons of Concern to UNHCR recognizes that it is not always possible to obtain refugees’ free and informed consent, for instance, when a refugee seeking food or cash assistance has no other source of income. In these situations, the UNHCR Policy provides no details about alternative modes of identification. Instead, it suggests that data processing (e.g., collection and sharing) remains valid under other legal bases such as the vital or best interests of the data subject, UNHCR’s mandate, or the safety and security of refugees. On that analysis, reference to consent sounds hollow if it can be bypassed when the context makes it inconvenient to obtain.

Beyond the question of free consent, doubts remain about whether we can truly achieve informed consent in the context of new technologies. The technological complexities surrounding biometrics processing, its entanglement with big data, and unforeseen future applications make it nearly impossible to provide the information necessary to obtain informed consent – let alone in plain language. When processing someone’s data involves complex data flows between diverse institutions with distinct interests, as can easily be the case when refugees’ data are shared with states and implementing partners, we face what Solon Barocas and Helen Nissenbaum (2014: 58) call “the transparency paradox.” That is, trying to impart the necessary information to obtain informed consent (i.e., data collected, for what purpose, with whom it is shared, following which modalities) in simple and clear terms cannot faithfully capture how data will be used. Hence, consent alone may be an inadequate basis for legitimating data use. For this reason, Carol Gould (2019) calls for the introduction of perspectives from democratic theory, bringing together principles of participation, deliberation, and representations to develop a form of “collective consent” that would apply to decision-making concerning new technologies and big data. Achieving that would be complex enough in liberal democracies. In the refugee context, this proposal will collide with the democratic boundary problem discussed by Eva-Maria Schäfferle in Chapter 16, whereby participation and representation is tightly knit to citizenship. Crucially, in the international humanitarian environment, it would require a perspective shift from seeing refugees as aid recipients devoid of agency (see Frédéric Mégret, Chapter 5), to individuals entitled to their say in decisions that affect them.

b Non-binding Standards

UNHCR’s participation in collecting refugees’ biometric data and the provision of digital identity is the extension of a paradigm that sees the expansion of nonstate actors’ regulatory role in spheres that used to be the preserve of states. Particularly acute since the 1990s, the phenomenon has directed attention toward how international organizations can be held accountable when their decisions or actions affect the rights or wellbeing of people under their protection or control (Benvenisti, 2018: 30). Hence, the question of accountability raised by UNHCR’s biometrics practices is closely entwined with unresolved debates that emerged in the early 2000s about the responsibility of international organizations. While a number of theories are proposed to address these shortcomings, none offers a substitute for the protection found in national contexts. Several institutional characteristics undermine the Agency’s answerability towards refugees and control over their own data. One is the nonbinding character of rules governing the management of refugees’ data by UNHCR. The Agency’s practices surrounding the processing of biometric data are primarily constituted of self-imposed policies and guidelines that inform the work of its staff. The most relevant instruments include the UNHCR Policy and two sets of guidance, mentioned earlier, for applying the Policy and guiding the registration of refugees, both adopted in 2018. Similar nonbinding standards developed by the United Nations inform UNHCR’s practices.

Given the ongoing evolution and unsettled state of the law on data protection and new technologies, the adoption of standards by UN agencies is a welcome development. These policies and guidelines around basic data processing principles may promote a more uniform and right-enhancing approach to the treatment of biometric data within UNHCR. The standards also fill a gap in international law. While the right to privacy is recognized in international instruments such as the Universal Declaration of Human Rights and the International Covenant on Civil and Political Rights, no international instrument is dedicated to data protection. The prospect for the conclusion of an international treaty on this subject is dim. The ideological divide between the European and American approaches to data protection and the pace with which the field itself is moving makes such an agreement unlikely in the short term (Kittichaisaree & Kuner, 2015). In any event, even if adopted, an international treaty concerning data protection would be of little use to refugees owing to the difficulty in practice to apply human rights instruments to international organizations.

Still, the effectiveness of UN standards depends on implementation. An audit of the UNHCR Biometric Identity Management System by the Office of Internal Oversight Services (OIOS) in 2016 revealed shortcomings in the application of the Policy. The OIOS found that the information communicated to refugees during biometric registration was below the standards provided for by the Policy, especially regarding the conditions of access to data by third parties and the rights and obligations of refugees. It also found that in all five country operations under review, staff knowledge of the Policy was limited. The Rohingya and Kenya cases discussed previously suggest flaws remain in applying the principles identified in UN standards in the area of concerns to the OIOS. The transfer of refugees’ biometric data to the United States as per the MoU concluded between UNHCR and DHS in 2019 raises further questions about respect for UNHCR standards, especially with regard to the retention of data for longer than necessary – seventy-five years – and the use of the data for a purpose incompatible with the original reason for collection: humanitarian aid versus crime prevention (Department of Homeland Security [DHS], 2019).

c Legal No-Man’s Land

With the risk of misapplying the Policy, another concern is the absence of an independent mechanism to hold the agency accountable when processing refugees’ data. UNHCR’s Policy on data protection establishes three supervisory roles with regard to data management: (1) a data protection officer, located at UNHCR’s headquarters, with overarching responsibilities for monitoring, advising, supporting, and training personnel involved in data protection within UNHCR; (2) data controllers in each country office or operation, usually the Representative in a UNHCR country office, responsible for developing procedures for the handling of data that comply with the Policy; and (3) data protection focal points, designated by data controllers to assist them in their role. The focal point is usually the most senior UNHCR protection staff member in a country office or operation. These bodies are primarily responsible for defining institutional and country-wide procedures regarding data processing. As such, they may provide a level of oversight over decisions involving the processing of personal data. Still, (at least until recently) they were not responsible for handling individual complaints concerning breaches of the Policy. The main pathway available to refugees seeking to challenge the processing of their data is the UNHCR Inspector General’s Office (IGO), an oversight mechanism for investigating claims of misconduct by UNHCR personnel and staff. However, a number of structural deficits may prove to be severe obstacles to accountability (Johansen, 2020). This includes IGO’s limited jurisdiction ratione personae to UNHCR staff members, not the Agency itself. Hence, the outcomes of investigations are disciplinary measures rather than remedial obligations. There are also shortcomings regarding the impartiality of the IGO owing to staffing procedures. This said, the creation of a Personal Data Protection Review Committee, envisaged in the 2022 General Policy on Personal Data Protection and Privacy to review decisions by local officials, could provide an additional and independent layer of redress once established (UNHCR, 2022).

Closely related to the latter point, the absence of an independent system of judicial review at the international level further undermined the scope for accountability within UNHCR. In a national context, judicial review has proven an efficient avenue for citizens to circumscribe the deployment of digital identity systems and population registries. In India, Mauritius, Kenya, and Jamaica, citizens and civil society have challenged the constitutionality of digital identity systems using biometrics. In these instances, judicial review afforded a space for negotiating the modalities under which governments may collect and process biometric data.² In three of the four cases, courts have upheld the systems but imposed limits on the type of data that can be collected and for which purpose, on who can access data, and for how long data can be retained. The Jamaican Supreme Court in Julian J. Robinson v. The Attorney General of Jamaica (2019) went further and struck down the National Identification and Registration Act and the National Identification and Registration System (NIDS) altogether. The Court concluded that the NIDS, which would have made it mandatory for all Jamaicans and certain residents to register their biometric data and obtain a NIDS identity card, would violate their constitutional right to privacy.

Even if judicial review operates ex post to the political decision, it nonetheless offers the potential for a constituency to agree or challenge governmental decisions through judicial bodies. In these four cases, national courts afforded citizens a pathway for influencing the structure and scope of national identity management systems. The binding nature of these decisions helped redress structural power inequalities between members of the public and the state. However, this crucial check designed to reign in the legislative and executive powers is lacking when UNHCR assumes responsibilities for the development of identity management systems. When conducting their mandate and functions, the immunities and privileges afforded to UNHCR and its officials shield the Agency from judicial proceedings. Indeed, Article 105 of the UN Charter and the 1946 Convention on the Privileges and Immunities of the United Nations are interpreted as providing absolute immunity to the Organization, its representatives, and officials from the jurisdiction of national courts (Rosa & Lemay-Hébert, 2019). In this context, the agency of refugees in defining the conditions under which UNHCR can collect and process their data is restricted.

Scholars have argued that the absence of checks and balances and a weak legal framework in the international migration space explains the proliferation of biometric technologies for governing the movement of refugees (Jacobsen, 2015; Molnar, 2019). On that analysis, international organizations such as UNHCR are a conduit for states to circumvent their legal responsibilities to protect human rights. Eyal Benvenisti (2018) notes that global institutions may pose a greater regulatory challenge than state institutions. Indeed, the transfer of responsibilities from states to international organizations allows powerful countries to evade democratic deliberations and respect for individual rights. This logic underpins the extraterritorialization of migration control, which, as noted by Ayten Gündoğdu in Chapter 10, often places the refugee in a “condition of rightlessness.” In the present case, it is doubtful that the delegation of responsibility, if only partial, for the collection of refugees’ biometric data to UNHCR is the sole product of a conscious reflection by states about evading their legal responsibilities. Foreigners’ rights to data protection are generally already restricted when compared with the rights afforded by states to their own citizens (Guild, 2018). In parallel, we see a growing number of countries, especially in the Global South, adopting digital identity management systems and population registries. Most likely, the demands on UNHCR for producing ever more granular data about refugees is a consequence of the factors identified at the beginning of this chapter. The emergence of an audit culture at the global level, the securitization of migration, and the promotion of a legal identity agenda in the UN Sustainable Development Goals all contribute to this development. But whether or not the growing pressure on UNHCR to collect and share refugees’ biometric data is underpinned by a desire to evade legal obligations, this development produces a de facto transfer of biometric practices in a space of legal paucity that undermines refugees’ agency over their data.

Conclusion

In 2015, UNHCR quoted Olivier Mzaliwa, a Congolese refugee, when promoting its biometric system: “I can be someone now. I am registered globally with the UN and you’ll always know who I am.” Perhaps Olivier is right, and we will always know who (and where) he is, but is that an inherently positive outcome? This line reflects the tension between the imperative of identification for access to aid and territory on the one hand and the right to integrity, privacy and control over one’s data on the other. Kevin Haggerty and Richard Ericson (2000: 611) wrote that under a logic of technological surveillance, human bodies are abstracted from their territorial location and are reassembled into virtual “data doubles.” These doubles then circulate in space and between actors in ways that condition the allocation of resources and power of (and over) those they digitally represent, without them being aware. The biometric data of refugees collected and processed by UNHCR follow a similar trajectory and purpose. While Olivier’s body might well be trapped in a camp for a protracted period or stopped en route, his data, by contrast, are not only condoned but encouraged to cross national borders. In the legal no-man’s land where UNHCR’s biometrics practices are taking place, the risk of disenfranchising refugees, by undermining their ability to control the modalities under which their data are collected, processed, and shared, is real. It leaves them with little to no ascendance over the paths their data doubles will take, which borders they will cross, and who they will encounter along the way.