I. INTRODUCTION
The human right to leave any country is being quietly and selectively eroded. It is not a widespread phenomenon, but it warrants inquiry. Why is it happening, and why does it matter? This article answers these questions by documenting the subversion of the right to leave any country, the practical, legal and discursive techniques that have enabled its decline to pass largely unremarked, and what this might signify for the regulation of mobility.
The right to leave any State, including one's own, is enshrined in international human rights instruments, including Article 13(2) of the Universal Declaration of Human Rights and Article 12(2) of the International Covenant on Civil and Political Rights (ICCPR).Footnote 1 It also appears in regional human rights instruments in Europe, Africa, the Americas, Arab States and various specialized human rights instruments.Footnote 2 The right to exit has intrinsic value as an expression of liberty.Footnote 3 It is also a vital complement to refugee protection, because the refugee definition stipulates that a person must be outside their country of nationality to qualify for asylum. Absent a right to exit, a person would be deprived of the capacity to seek asylum. In 1968, political philosopher Hannah Arendt asserted that ‘to depart for where we will is the prototypical gesture of being free’.Footnote 4 Banning departure from a State is itself a repressive measure targeting those who, in the words of AO Hirschman, wish to express their disaffection through exit rather than voice.Footnote 5 In Arendt and Hirschman's time, violation of the right to leave was typified by the Cold War exit restrictions imposed by communist regimes.
From 1990 onwards, exit restrictions gradually ended along with the communist regimes that imposed them. Within a couple of decades, however, a shift began among States and institutions of the Global North away from taking exit for granted as a fundamental right and toward regarding it as an inconvenient impediment to advancing State interests in a revised world order. The right to leave emerged in international legal instruments post-World War II in a period where it performed two conjoined purposes. Legally, a right to exit one State was linked to (and was a precondition for) the exercise of the right to seek asylum in another State under international refugee law.Footnote 6 Politically, the Cold War narrative portrayed exit restrictions as imprisoning citizens in totalitarian regimes, and gave the West another cudgel to wield in its ideological battle against communism.Footnote 7As long as the paradigmatic refugee was the Soviet dissident defector, politics and law pointed in the same direction.
From a twenty-first-century securitization perspective, exit restrictions occupy a different political valence.Footnote 8 While the right to leave still appears in the text of various human rights instruments, its twentieth-century proponents have quietly switched teams. The States that championed exit rights now actively subvert them, seeking to contain migrants and asylum seekers as far away as possible from the European Union (EU), North America and Australia.
Enlisting States of origin or transit (‘departure States’) to prevent exit from their own territory has emerged as a tool for preventing entry into the territory of ‘destination States’.Footnote 9 Violeta Moreno-Lax and Mariagiulia Giuffré dub this trend ‘consensual containment’.Footnote 10 They join a handful of other scholars who provide incisive and critical analyses of the history and doctrine of exit rights under international law, especially in relation to refugees.Footnote 11 Scholarship under the rubric of ‘crimmigration’ or ‘border criminology’ interrogates the criminalization of migration and the carceral character of migrant detention. This article complements and extends earlier work by exploring how criminalizing exit effectively abrogates the right to leave to create an uninterrupted regulatory apparatus directed at the prevention of entry. This is labelled the ‘seamless border’. The concept of the ‘carceral State’ links the function of exit restrictions in confining actual or prospective migrants on the territory of a State, and the function of detention in confining migrants to dedicated carceral spaces within that territory.
The argument proceeds as follows: Section II explores the practical erosion of the right of exit as an element of externalization of border control, providing a range of examples and mechanisms through which departure States restrict exit in the service of preventing entry to destination States. The illustrations focus on cooperation between Member States of the EU and relatively proximate third States that are viewed as departure States for migrants and refugees,Footnote 12 and explain the legal, political and operational mechanisms through which the effacement of the right to leave is achieved and implemented. These include the criminalization of exit, the ‘crime’ of irregular departure, pull-backs at sea, and subsequent arbitrary detention, cruel, inhuman and degrading treatment, and refoulement.
Section III introduces the concept of ‘seamless borders’ to explain how exit restrictions by a departure State fuse with a destination State's entry restrictions to squeeze out the normative and physical space between the right to leave one country and entry into another. A seamless border thus creates an integrated apparatus for controlling movement. This is followed by an account of how the practical enjoyment of the right to leave is constrained by the architecture of international human rights provisions regarding mobility, legally occluded by the anti-smuggling regime, and replaced by the humanitarian mechanism of search and rescue (SAR).Footnote 13 The article turns next to addressing the impediments that regulating through exit restrictions creates for attributing legal accountability to the departure States that sponsor and support them.
Section IV draws on crimmigration and border criminology literature in revealing conceptual links between exit restrictions and migrant detention. Scholars focusing on migrant detention document the transition from a model of administrative detention to its contemporary character as a carceral space. It is argued that exit restrictions adopted to prevent entry to destination States render States of departure as the new ‘carceral State’, in which the State itself operates as a punitive space of confinement for detaining actual or prospective migrants who are unwanted by destination States.
Finally, the Conclusion queries whether current developments reveal that the commitment to a right of exit was always more fragile than supposed. Once shorn of its Cold War utility, the right to exit is exposed as politically dispensable. It may be that a paradigm of mobility organized around entry and exit is veering toward obsolescence and being replaced by a logic of governance over movement itself.
II. PREVENTING ENTRY BY PREVENTING EXIT
A. Externalization
Even before the Cold War ended, the vector of migration was changing. Post-war decolonization reversed the historic direction of flows from metropole to periphery among former colonial powers. From the 1980s onwards, asylum seekers from the Global South grew in absolute numbers and relative to those from Eastern Europe. The externalization of borders by States of the Global North is not new, and States often swap, emulate and adopt one another's policies.Footnote 14 Nor is it new that the targets of externalization are predominantly people who are racialized as Black or Brown. Recent scholarship and advocacy address contemporary bordering as a legacy of empire, the role of borders in maintaining and enforcing global racial hierarchies, and border violence as an extension of racial domination.Footnote 15 Simply put, race is everywhere in bordering.
An array of mechanisms is deployed to prevent unwanted people from reaching States where they may enter irregularly and/or seek refugee protection.Footnote 16 States have progressed from requiring passports for international travel, to selective demands to obtain visas in advance at a consulate or embassy abroad and, most recently, to online electronic travel authorization that functions like a visa-lite for travellers from visa-exempt States. Excisions of territory and airport departure lounges create the legal fiction that State territory is not State territory for the purposes of enforcing migration and refugee law. Immigration officials from destination States are dispatched to airports abroad to vet passengers before they board flights. Carrier sanctions discipline private airlines and shipping lines for transporting improperly documented passengers and deputize them to scrutinize the validity of travel documents. Some States send deterrent advertisements directly or through international bodies like the International Organization for Migration (IOM).
Safe third-country designations require asylum seekers to submit their refugee claims in the first ‘safe’ country of arrival and authorize transfer of asylum seekers back to those States. They enforce a norm of responsibility sharing that distributes the responsibility to protect refugees according to a State's geographic proximity to the refugee-producing State because the closer the third State to the country of origin, the more likely it is that a migrant will pass through it en route to the ultimate destination State. The endpoint of these ‘safe third country’ regimes is to augment the responsibility of States of the Global South, who already host 75 per cent of the global refugee population.Footnote 17 Maritime interdictions in the form of push-backs are straightforward exercises of force whereby the coast guard of a destination State (eg Italy, Malta or Greece) intercepts vessels with migrants aboard and repels them to the territory of another State. Sometimes passengers are transferred back to authorities in the State of departure (eg Libya or Tunisia), and, sometimes, the vessels sink or capsize and people drown.Footnote 18 The Mediterranean Sea is the world's deadliest maritime crossing, with almost 30,000 recorded dead or missing between 2014 and 2023.Footnote 19 Many more deaths are likely to remain unrecorded.
B. Making Exit Illegal
As an operational matter, States can externalize their entry controls and rely on the passive acquiescence of affected States. The last two decades, however, bear witness to an acceleration of extraterritorial tactics that engage other States in cooperative or coordinated action to prevent unwanted migrants or asylum seekers from reaching destination States.
The spectre of ‘offshore disembarkation platforms’ for asylum determination recedes and reappears with different State proponents. From 2022 to 2024, Britain doggedly pursued the UK and Rwanda Migration and Economic Development Partnership, whereby the United Kingdom (UK) would pay Rwanda to accept asylum seekers forcibly and permanently removed from the UK for asylum determination, settlement or repatriation. The plan was abandoned in mid-2024 by the newly elected Labour government. Italy's agreement with Albania to detain and deport asylum seekers interdicted at sea by Italian authorities was swiftly ruled unlawful by Italian courts in October 2024, but it is likely to be only a matter of time before another State revives and champions a similar scheme.Footnote 20
Readmission agreements between States of departure and destination States facilitate deportation by the latter to the former. United States’ (US) pre-clearance zones in Canadian and Irish airports enable the US to conduct border inspections in foreign airports, and ‘juxtaposed’ border controls allow the UK, France, Belgium and the Netherlands to do the same at certain cross-Channel routes. The efficacy of unilateral, bilateral and multilateral informal agreements, as well as privatized techniques of deterrence and deflection, are enhanced by increasingly sophisticated methods of surveillance, data collection, integration and sharing, including (but not limited to) biometric and artificial intelligence (AI) technology.Footnote 21
In all instances, extraterritoriality complicates legal accountability for the direct or delegated exercise of legal authority. The European Court of Human Rights’ (ECtHR) judgment in Hirsi Jamaa v Italy Footnote 22 addressed Italy's push-backs on the Mediterranean high seas. Italy intercepted a boat carrying 200 asylum seekers, forcibly returned them to Libya without allowing them to assert a refugee claim, and then transferred them to the Libyan authorities, who detained and subjected them to torture and/or cruel, inhuman and degrading treatment. Rejecting US Supreme Court precedent, the ECtHR ruled that Italy exerted extraterritorial jurisdiction over the applicants for the purposes of Article 1 of the European Convention on Human Rights (ECHR).Footnote 23 The Court found that the Italian push-backs violated Article 3 (torture and/or cruel, inhuman and degrading treatment) and Article 4 of Protocol 4 (collective expulsion) of the ECHR.Footnote 24
Despite this ruling by the ECtHR, push-backs reportedly continue apace more than a dozen years later. Refugee advocates have found themselves in a cat-and-mouse game concerning jurisdiction. As Itamar Mann explains, State authorities read adverse decisions not as a lesson about respecting human rights, but rather as advice on how to relocate the locus of the human rights violation beyond the ECtHR's jurisdiction to adjudicate.Footnote 25 One tactic is to supplement or substitute push-backs by destination States with pull-backs by departure States.Footnote 26 The logic of a push-back is prevention of entry; the logic of a pull-back is prevention of exit. The context in which these practices arise makes it indisputable that these interceptions are conducted in the interests of, and at the behest of, destination States. The critical point is that destination States do not exert direct, physical control over the people who are immobilized by these practices, leading Giuffré and Moreno-Lax to label them ‘contactless control’.Footnote 27
Desultory diplomatic efforts by destination States to persuade departure States to impede exit are not new, but anecdotal reports suggest that the practice gained momentum early in the new millennium. In 2000, Australia supplemented its maritime ‘push-back’ apparatus by entering into a Regional Cooperation Agreement with Indonesia and the IOM. Indonesia (with Australian financial, logistical and technical support) intercepts putatively Australian-bound third-country migrants and asylum seekers, and they are ‘accommodated’ in Indonesia while the IOM provides ‘repatriation assistance’ to return them to their countries of origin.Footnote 28 Indonesia is not a party to the Convention Relating to the Status of Refugees, and does not extend effective protection to refugees.Footnote 29 In 2006, the Conference of Ministers of the Western Mediterranean ‘welcomed the efforts of the countries of the southern Mediterranean to contain illegal emigration to Europe’.Footnote 30 Indeed, Senegal boasted of arresting more than 1,500 ‘potential illegal emigrants’ attempting to sail from Senegal to the Canary Islands.Footnote 31 Before Israel built a fence in 2012 to prevent Eritrean and Sudanese asylum seekers reaching Israel via the Sinai, Egypt acceded to Israel's request to ‘work to prevent future infiltrations [into Israel] from its territory’.Footnote 32 The Egyptian military thereafter (and unilaterally) adopted a practice of shooting at asylum seekers—killing and wounding many—as they attempted to cross into Israeli territory via the Sinai.Footnote 33
Although exit restrictions are pursued by various destination States, this article focuses on agreements by EU destination States with select departure States. A thread that runs through all of these practices is the creation or revival of laws in departure States that turn the exercise of a right to leave into an offence or crime of ‘irregular departure’.
1. Libya and Tunisia
Italy and Libya concluded their first Treaty of Friendship, Partnership and Cooperation in 2008, while Muammar Gaddafi still ruled Libya.Footnote 34 A post-revolution Memorandum of Understanding (MOU) providing for readmission of expelled migrants followed in 2012, which was in force when the Hirsi Jamaa case was litigated.Footnote 35 Five years later, in 2017, Italy and Libya's faltering Government of National Accord concluded a ‘Memorandum of Understanding on Cooperation in the Fields of Development, the Fight against Illegal Immigration, Human Trafficking and Fuel Smuggling and on Reinforcing the Security of Borders between the State of Libya and the Italian Republic’.Footnote 36 The stated objective of the Italy–Libya MOU is to halt the flow of migrants through Libya to Italy across the Central Mediterranean route.Footnote 37 Under the MOU, Italy provides financial, technical, military, logistical and infrastructure support to Libya in exchange for blocking migrants’ departure or intercepting them on boats, pulling them back to Libya, and detaining them in so-called ‘reception centres’.Footnote 38 In this model, pull-backs by Libya replace push-backs by Italy. The European Council's Malta Declaration of 2017 endorsed the Italy–Libya MOU and pledged EU support.Footnote 39
Migrants in Libya remain vulnerable to predation inside and outside detention. The Libyan Coast Guard and, more recently, the Tarek Bin Zayed (TBZ) militia, interfere with SAR efforts by non-governmental organizations (NGOs), reportedly let migrants drown, and brutalize those rescued by them.Footnote 40 Once pulled back to Libya and detained in ‘reception centres’, they are exposed to starvation, extortion, forced labour, enslavement, trafficking, torture and cruel and inhuman or degrading treatment, murder and refoulement.Footnote 41
The United Nations (UN) Support Mission in Libya notes that ‘the overwhelming majority of migrants and refugees are placed in indefinite detention pending deportation without being charged, tried or sentenced under applicable Libyan laws’.Footnote 42 Although Libya lacks a functioning government or justice system, it has enacted and revised laws from 1987 onwards to criminalize irregular migration. Libyan law allows for the automatic and indefinite detention, forced labour and deportation of those who commit the ‘crime’ of irregular departure, and provides the legal authority for enforcement via pull-backs.Footnote 43
In July 2023, the EU and Tunisia entered into an MOUFootnote 44 which promotes various economic and trade initiatives, as well as measures to enhance ‘cooperation in combating and reducing irregular migration flows’, in the context of a ‘migration/development nexus’.Footnote 45 Indeed, this relatively recent MOU can be viewed as extending the earlier pattern of bilateral or multilateral mobility partnership agreements dating from the 1990s and early 2000s. In exchange for financial, technical and operational support, the prospect of enhanced ‘legal mobility’ and ‘legal pathways’ for Tunisians seeking access to the EU, Tunisia undertakes to facilitate repatriation of irregular Tunisian migrants from the EU, and to improve ‘coordination of search and rescue operations at sea’.Footnote 46 This coordination translates into official affirmation of Tunisian pull-backs, a practice that predates the 2023 MOU.Footnote 47
As with Libya, Tunisian pull-backs are authorized by domestic laws prohibiting exit through ‘illicit’ means, which includes departure at an unofficial border crossing, or through fraud, impersonation or use of false documents. Unlike Libya, Tunisian law also prohibits Tunisian citizens from exiting without prior authorization. Similar laws exist in Morocco and Algeria. Maximum imprisonment terms range from months (Algeria and Morocco), to 20 years (Tunisia).Footnote 48 The Tunisian law criminalizing ‘irregular entry and exit’ and ‘any act of assistance to irregular migrants and asylum seekers’ dates from 2004, a year after Tunisia entered into an MOU with Italy.Footnote 49 Human rights organizations report that migrants intercepted by Tunisia are subject to beatings and robbery during apprehension and after return to Tunisia, and then expelled en masse across the land borders in remote and dangerous desert locations.Footnote 50
Noting the decline in maritime migrant arrivals in Italy after 2017, the IOM observed the rise in the number of interceptions conducted by Tunisia and Libya:
The … number of people being returned to North African shores has increased in recent years. Interceptions by the Tunisian and Libyan coast guards accounted for 8 percent of all search and rescue operations in the Central Mediterranean in 2016, but by 2018, 49 percent of the total number people recorded attempting to cross were brought back to Tunisia or Libya. This shift can be attributed to several factors, including the decreased maritime patrol area of Italian authorities and the shift of EU/Frontex assets from maritime vessels to drones incapable of conducting rescue at sea.Footnote 51
In 2022, the IOM reported that 58,900 migrants departed Tunisia via the Central Mediterranean, of which 26,500 (45 per cent) were intercepted and pulled back to Tunisia.Footnote 52 Of those returned, 29 per cent were Tunisian. The same IOM report notes that during the same year 24,738 migrants were intercepted by the Libyan Coast Guard and returned to Libya (31 per cent of those departing Libya).
2. The Balkans and Turkey
In 2013, the Council of Europe's Commissioner for Human Rights reported that Central and Eastern European States selectively restricted the exit of Roma in order to avoid jeopardizing visa-free access to EU Member States:
What appears to be happening in the Western Balkans is that as EU member states increase pressure on these states to the effect that if the numbers of their nationals applying for asylum in the EU does not decrease, then all nationals of the state will be subjected to a mandatory visa requirement (again), the authorities of these states are seeking to restrict the departure of individuals who they consider at risk of applying for asylum, that is, the Roma.Footnote 53
Western Balkan States are not unique in their willingness to trade the mobility of marginalized citizens for the promise of enhanced access to the EU for preferred citizens. The 2016 EU–Turkey Statement committed Turkey to readmitting asylum seekers, including Syrians, who had entered Greece from Turkey, in exchange for the EU resettling Syrian refugees via legal pathways in the same proportion.Footnote 54 Even prior to the EU–Turkey Statement, Turkey required Syrian refugees with temporary protected status in Turkey to obtain ‘the permission of the Directorate General [of Migration Management] as a condition of lawful exit from Turkey’—a form of exit law.Footnote 55 In the Statement, Turkey also pledged to ‘take any necessary measures to prevent new sea or land routes for illegal migration opening from Turkey to the EU, and [to] cooperate with neighbouring states as well as the EU to this effect’.Footnote 56 In practice, this cooperation appears to involve a combination of push-backs by Greece and pull-backs by Turkey of migrants suspected of using Turkey to transit to European destinations. For example, in 2022 to 2023, the Turkish Coast Guard/Police reported that it had prevented or intercepted 100,000 asylum seekers attempting to cross the Aegean Sea to Greece. About 60 per cent of the boats pulled back by Turkey had also been pushed back by Greece.Footnote 57 Turkey's Deputy Minister of the Interior also reported that in March 2022, ‘Turkish authorities stopped 17,587 irregular migrants in the country who were trying to make it to Europe “illegally”.’Footnote 58 Turkey currently hosts the largest number of refugees (mostly Syrian) in the world.
3. Cameroon and Niger
Whereas the cooperative regimes described above rely on bilateral agreements, Cameroon is unique because it incorporates into domestic law a pre-emptive restriction on exit for the purpose of ‘illegal’ entry elsewhere. With French technical, logistical and enforcement support, Cameroon has formally created the crime of exiting Cameroon with the intention of entering another State unlawfully.Footnote 59
Under the aegis of an EU-supported anti-smuggling campaign, Niger swept facilitation of ‘illegal exit’ into its definition of smuggling in 2015, and introduced carrier sanctions for, inter alia, Nigerien bus companies transporting people within Nigerien borders who did not possess documentation proving Nigerien citizenship or authorized presence in Niger.Footnote 60 This extensive regulation of mobility both within and at Niger's borders provoked considerable backlash. The domestic unpopularity of these migration regulations ‘from beyond’ is evinced in the fact that within months of a military coup in Niger in July 2023, the new regime repealed the 2015 anti-migration law that criminalized transporting non-citizens within Nigerien territory.
Anecdotally, Amanda Bisong reports that a Nigerian court convicted a dozen Nigerians deported from Niger for ‘attempting to irregularly migrate to Europe through Niger’, even though no Nigerian law criminalizes ‘irregular emigration’.Footnote 61 Niger and Cameroon (and at least one Nigerian court) formally turned their own law into an extension of other States’ extraterritorial migration enforcement apparatus, capturing not only migrants transiting through their territory, but their own nationals.Footnote 62 Among other impacts, these EU-incentivized exit restrictions, in tandem with bilateral arrangements with individual African States, undermine and thwart regional free movement initiatives of the Economic Community of West African States (ECOWAS).Footnote 63
Legacies of colonialism and race shape the dynamic between States of departure and destination—with one notable exception that proves the rule. Under the terms of the 2018 Anglo-French Sandhurst Agreement (renewed in 2023),Footnote 64 the UK agreed to pay France to patrol French beaches to prevent irregular small-boat departures across the English Channel. While the Agreement manifestly failed to stem the movement of people crossing the Channel in small boats, it appears that the French Gendarmerie, acting as border guards, have resorted to violence to prevent departures.Footnote 65
C. Coordination and Cooperation
The exchange of material benefits for preventing exit illustrates the use of conditionality to persuade departure States to advance EU destination States’ migration policy goals. Today, States of departure anticipate substantial remuneration in money, military equipment and infrastructure from destination States who contract with them to intercept migrants on their territory or in territorial waters. An obvious asymmetry of power structures these arrangements in a manner that subordinates the interests of departure States in facilitating regional and transnational mobility. Many States of departure rely on remittances from nationals working abroad and have little domestic motive to discourage emigration. They may also wish to reap the benefits of regional free movement regimes. Of course, some States hope that restraining the departure of migrants who are unwanted by the destination States will open up lawful channels for the admission of their own nationals. So, halting the migration of some is done with the expectation of facilitating the emigration of others, to the economic, diplomatic and reputational benefit of the State of departure.Footnote 66 However, improved access for nationals has not yet materialized for the origin States.
Collusion between States of departure and destination raises critical questions regarding the complicity and responsibility of EU destination States in human rights violations ensuing from prevention of exit, especially maritime pull-backs.Footnote 67 At the operational level, national coast guards and border enforcement officials in Italy, Greece and Malta coordinate and cooperate with their counterparts from countries of origin or transit in intercepting migrants on dinghies and boats, in obstructing maritime rescue by civil society actors, and in facilitating pull-backs. The European Border and Coast Guard Agency (Frontex) is an important team member. Frontex was created by regulation in 2004 with a mandate of integrating management of the surveillance and control of external borders by individual States. Initially, it played a coordinating and technical support role, but its mandate, budget and personnel expanded significantly over the course of four major amendments to its founding regulation between 2007 and 2019, especially after the Syrian crisis in 2015.Footnote 68
Frontex now boasts the largest budget of any EU agency. The EU Commission's 2023 European Integrated Border Management strategy conjoins national authorities and Frontex into the European Border and Coast Guard.Footnote 69 Frontex is empowered to exercise enforcement powers akin to national border/coast guards, and to monitor, surveille and conduct risk analysis of irregular migration at the EU external borders. It can develop and participate in joint operations with Member States and with third countries, gather and share data, and execute returns to third countries. One commentator describes EU external border management as ‘increasingly hybrid in nature featuring a “multi-actor” European Border and Coast Guard, comprised of Frontex and national competent authorities often interacting in convoluted ways, subject to heterogenous legal frameworks combining EU and national law’.Footnote 70
Frontex is widely reported to cooperate in push-backs, pull-backs and disruptions of maritime rescue.Footnote 71 The role of Frontex in maritime operations in the Mediterranean has instigated allegations of human rights abuses and an absence of accountability mechanisms, culminating in media disclosure of a damning investigatory report by the EU anti-fraud watchdog.Footnote 72 Frontex Executive Director Fabrice Leggeri resigned in 2022. Frontex's specific role in Libyan and Tunisian pull-backs was documented by Border Forensics and Human Rights Watch in 2022 and 2023.Footnote 73 In 2024, a Swedish coast-guard officer confirmed that Frontex continues to collaborate in Libyan pull-backs: ‘As soon as we see a migrant boat leave Libya, for example, we call [Libyan authorities] and try to persuade them to take them back. And a lot of the time we succeed.’Footnote 74
An obvious and critical question is whether the EU role in pull-backs brings their actions within the jurisdiction of any court or tribunal for the purpose of attributing liability. Before turning to this issue, however, it is important to assess critically the significance of restricting exit from one State in order to prevent entry to another in relation to the asymmetry between rights of exit and entry under international human rights law. Maritime pull-backs present the issue starkly.
III. SEAMLESS BORDERS
The high seas represent a rupture in the contiguous patchwork quilt of bordered territory over the surface of the earth; in practical terms, the high seas are the only place where an individual right to leave can be exercised without simultaneously confronting a concomitant State prerogative to exclude. This absence of a sovereign claiming jurisdiction over the high seas helps explain the coordination and cooperation among State actors to control human mobility in nautical space, over which no single State can claim sovereignty.
The bilateral and multilateral arrangements whereby destination States proffer funds, resources, technology, resources and infrastructure in exchange for departure States’ restrictions on exit by sea go beyond externalization of migration control by destination States and point toward the paradox of seamless borders. Where prevention of exit from one State conjoins politically, legally and operationally with prevention of entry to another State (or States), bordering fuses rather than separates the authority of national sovereigns. This functional articulation of exit and entry policies of different States seals the border from both sides. In the case of the high seas, it transforms a borderless space into a nautical border zone, where the border happens wherever a State actor confronts a migrant. It represents the terminus of strategies to confine migrants, potential migrants, and refugees in departure States unless and until their movement is authorized by destination States.
The resort to exit restrictions as a means of preventing entry is not concealed. The task is less to expose the phenomenon than to explain how it can hide in plain sight. This section opens with an account of the asymmetry of exit and entry and the difference between the right of an individual, the power of a State, and the right of a State in relation to mobility (Section III.A). It then explains how the conjunction of anti-smuggling and SAR represents States as victims of rights violations by smugglers and as humanitarian actors when assisting in pull-backs (Section III.B). Finally, this section identifies the hurdles of State responsibility and jurisdiction that hamper efforts to pursue legal liability for violations of the right to leave against both departure and destination States (Section III.C).
A. Asymmetry Between Entry and Exit
Under international law, individuals possess a right to exit any State, but no right to enter another. This asymmetry creates the obvious dilemma that the right of an individual to leave any State is hollow if no other State has a duty to admit them. The formal and evasive answer offered by international law is that the right to leave may be frustrated but is not abrogated by the absence of a right to enter any other State. The answer from liberal political theory is that sovereignty, self-determination, and stability as a self-governing political community, require that States control entry and membership, but the same principles do not justify restricting exit.Footnote 75
The ICCPR expresses the asymmetry by defining the scope of the protected right (or freedom) and its limitations in different ways. Article 12 articulates three general norms regarding freedom of movement. First, everyone ‘lawfully within the territory of a State’ enjoys freedom of movement within that State. Second, everyone has the right to exit any country, including their own. Third, everyone has the right to enter and remain in their ‘own country’.Footnote 76
The right to exit any State, including one's own, seems to require little explanation or justification within a legal framework that centres freedom as the absence of State coercion.Footnote 77 Indeed, the 1868 Burlingame Treaty between China and the US disclaimed Imperial China's prohibition on emigration of Chinese subjects, affirming ‘the inherent and inalienable right of man to change his home and allegiance, and also the mutual advantage of the free migration and emigration of their citizens and subjects …’.Footnote 78 Andrew Wolman notes that various attempts from the late nineteenth century to the interwar years to produce international agreements to combat ‘irregular migration’ through exit restrictions ultimately failed to win sufficient support.Footnote 79 Vincent Chétail argues that the right to leave any State has attained the status of customary international law.Footnote 80 As noted earlier, the right to leave is also a vital complement to the right to seek asylum.
The right to leave is the broadest of the mobility rights under the ICCPR because, unlike free movement within a State or the right to enter one's ‘own’ country, the right of exit does not depend on lawful presence or membership in the State. Nevertheless, exit may be limited for reasons of national security, public order (ordre public), public health or morals.Footnote 81 For example, the legality of the various COVID-19 pandemic restrictions on entry and exit imposed by States depends on the interpretation of ‘public health’ and non-arbitrary limitations on Article 12 rights. Commonly accepted limitations on the right to exit concern persons subject to conscription and the criminally accused facing trial.
The right to enter one's ‘own country’ is narrowest in scope because it protects only citizens and those who might be functionally equivalent to citizens, but once established, it permits few exceptions. A restriction on the right of entry must be non-arbitrary, and the Human Rights Committee (HRC) has declared repeatedly that ‘there are few, if any, circumstances in which deprivation of the right to enter one's own country could be reasonable’.Footnote 82
The obverse of this narrow but virtually unqualified right of entry into one's ‘own country’ is that non-nationals have no right to enter a country that is not their ‘own’. This, in turn, is frequently recast into the misleading proposition that, as expressed by the ECtHR in Saadi v Italy, ‘… as a matter of well-established international law, and subject to their treaty obligations, including those arising from the Convention, Contracting States have the right to control the entry, residence and removal of aliens’.Footnote 83
The author has argued elsewhere that the right of a State under international law to exclude non-citizens is a right against the State of nationality (which owes a reciprocal duty to readmit its citizen), not against non-citizens as individuals. International law recognizes the rights of States against other States, not against individuals, and under international human rights law, individuals have rights against States, but not vice versa. Thus, non-citizens do not possess a human right to enter a State, nor do States have a right to exclude non-citizens. Rather, States exercise a sovereign power to exclude non-citizens, but this power, like other State powers, is subject to the constraints of international human rights. Confusing the power of a State to exclude with a right to exclude is a formal error with profound normative repercussions.Footnote 84 Yet it is undeniable that border control is commonly represented as a right of States against non-citizens that flows from the principle of sovereignty. Casting State power to exclude as a State ‘right’ matters because the exercise of a right requires no justification beyond itself: ‘I can do X because I have a right to do X.’ It also facilitates the depiction of States as victims of rights violations when unwanted migrants (including refugees) attempt entry, instead of orienting the inquiry to whether and how the exercise of State power to exclude might violate an individual right to leave or to seek asylum. The next section explains how the discursive disappearance of the migrant as a rights-bearing legal subject is enabled by the coupling of the anti-smuggling regime with the practice of maritime SAR.
B. States as Victims and Heroes
Until the end of the twentieth century, international law did not regulate mobility through criminalization. This changed with the adoption of the UN Convention against Transnational Organized Crime,Footnote 85 supplemented by the Protocol against the Smuggling of Migrants by Land, Sea and Air (Smuggling Protocol),Footnote 86 the Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children (Trafficking Protocol),Footnote 87 and the Protocol against the Illicit Manufacturing of and Trafficking in Firearms, their Parts and Components and Ammunition (Firearms Protocol).Footnote 88
The Smuggling Protocol prohibits profit-driven facilitation of ‘illegal entry’. The Trafficking Protocol is directed at the coerced movement of people within and across borders for purposes of exploitation.Footnote 89 It matters that trafficking and smuggling of people are located in an instrument dedicated to suppression of transnational organized crime rather than, for example, the protection of human or labour rights. It is also significant that the movement of people is placed alongside the trade in firearms. Like illicit guns, illicit people figure as dangerous contraband commodities that organized crime actors transport across borders, inflicting harm on the destination State.Footnote 90
Even though trafficked persons (unlike smuggled persons) are depicted as victims of the private actors who exploit them, their vulnerability is subordinate to the victimization of the State by smugglers and traffickers. It is first and foremost the State, not smuggled or trafficked people, that is victimized by transnational organized criminals. That the State is the primary victim of a criminal offence is consistent with the structure of domestic criminal law where the State (and not the individual victim) prosecutes perpetrators in the name of the sovereign.Footnote 91
The EU anti-smuggling framework diverges in certain respects from the international approach. It embeds smuggling into a broader framework that is explicitly aimed at criminalizing irregular migration. The legal framework was established in 2002 with the ‘facilitators package’, consisting of a Council Directive and Council Framework Decision directed against the facilitation of unauthorized entry, transit and residence. The Decision defines smuggling and promotes the harmonization of minimum penalties and escalating sanctions among Member States.Footnote 92 Notably, the definition of smuggling does not require that smugglers gain a financial or material benefit. It permits, but does not require, exemption for those who act on humanitarian grounds. Domestic prosecution of humanitarian actors, SAR NGOs, family members and even smuggled persons themselves reveals that suppressing organized crime does not drive policy or enforcement; indeed, it is widely recognized that few smugglers belong to any ‘organized crime’ network anyway. Rather, anti-smuggling is simply a tool for criminalizing irregular entry as part of the EU's larger project of ‘migration management’.Footnote 93
In 2023, an Italian court referred a case to the Court of Justice of the EU on the question of whether the prosecution of a Congolese mother for smuggling her daughter and niece breached the EU Charter of Fundamental Rights.Footnote 94 A few months later, the European Commission proposed a Directive on smuggling to replace the facilitators package with minimum rules that would, inter alia, specify that a smuggling offence requires facilitation of entry, transit or stay where the accused receives a financial or material benefit or creates a ‘high likelihood of causing serious harm to a person’.Footnote 95 It does not, however, explicitly exempt familial or humanitarian assistance.
The logic of anti-smuggling, with its focus on how people cross borders, contrasts with the logic of the Refugee Convention, which constrains State power based on why people cross borders. Note that the means of movement (such as being smuggled) and the motives for movement (such as fleeing persecution) are fundamentally different. That someone is smuggled discloses nothing about whether they are a refugee, or the human rights consequences of excluding them. Article 31 of the Refugee Convention provides that refugees should not be penalized for entering or remaining irregularly, recognizing that desperate people will take desperate measures in order to obtain refuge.Footnote 96 In other words, the means of entry cannot be held against someone whose reasons for entry make them a refugee.
As a matter of law, anti-smuggling regimes cannot derogate from human rights obligations contained in other international instruments, including non-refoulement, non-penalization of irregular entry under the Refugee Convention, or the right to leave under the ICCPR, the ECHR and other instruments. As a matter of fact, they do. Most people who are apprehended and repelled at the EU's external borders are nationals of conflict-ridden and/or repressive regimes. Whether they would qualify for asylum is unknowable in the absence of the opportunity to seek it. There are no lawful, regular migration routes for asylum seekers, or for the many migrants whose desperation does not register in the refugee definition. It is incontrovertible that restrictive migration laws create the market for smugglers, and that the conditions that impel people to move will not improve through policy interventions designed to deter movement. Notwithstanding the structural factors that produce and perpetuate irregular migration, anti-smuggling provides States with an attractive explanation rooted in international and EU law that foregrounds rights (of States), crime (against States) and victimization (of States first, and migrants, second). The problem is smugglers, migrants are their contraband cargo, and the solution is to break the smugglers’ business model to prevent the importation of contraband.Footnote 97 Within this framework, migrants lose their status as rights-bearing legal subjects; rather, they are dangerous commodities—much like the weapons that are the subject of the Firearms Protocol—contraband humans to be managed and eliminated in the name of vindicating the State's ‘right’ to exclude.
This presentation of States as the primary victims of smugglers is incomplete, as the tragedy of people drowning in sinking boats in the Mediterranean creates overwhelming political pressure on States for a humanitarian response. Operationally, this response of destination States is not grounded in the legal obligations imposed by the Refugee Convention or human rights law. Instead, it is located within the humanitarian apparatus of maritime SAR.Footnote 98 Smugglers exploit and endanger migrants by setting them adrift in overcrowded, unseaworthy boats, callously abandoning them unless States intervene and rescue them. This enables interdiction and pull-backs to be depicted as fundamentally benevolent SAR operations in which State actors heroically rescue migrant victims from evil smugglers, rather than as forcible returns to departure States (who perpetrate grave human rights abuses, including refoulement).Footnote 99
While EU leaders acknowledge SAR as a legal duty, the practice is imbued with a humanitarian and discretionary character that masks its role in the larger framework of halting migration. SAR is organized around the urgent and immediate need to save lives that are episodically endangered at sea. The structural causes that produce maritime emergencies, and the predictable consequences of interception by brutal and corrupt State actors, lie outside the borders of the humanitarian snapshot that focuses on saving lives in a moment of crisis. The logic of anti-smuggling produces a narrative where the suffering and death of migrants in transit is tragic, but attributable to bad choices by individuals who attempt to migrate through irregular means, and to the smugglers they hire. States of departure and/or destination who engage in SAR act seemingly heroically from a perch of innocence and humanitarian duty.
As a humanitarian undertaking, SAR will always be reactive, and always fall short of saving the lives of everyone at risk, especially as EU actors progressively disengage from direct intervention and outsource rescue to departure States.Footnote 100 Within a SAR model, it matters little whether the life is saved by an NGO vessel or the Italian or Libyan coast guard, or where the rescued people disembark. But within a model of combatting irregular migration to the EU, it matters that State authorities practically and legally obstruct NGO vessels in their rescue operations, and that brutality and refoulement await migrants intercepted by Libya and Tunisia. And, of course, migrants inevitably disappear and drown in these incidents, serving as a cautionary tale to others who might follow. The notion that possible rescue from drowning operates as a ‘pull factor’ drawing migrants to the EU can be used to rationalize exposing some or many migrants to death by drowning and signals the subordination of SAR to the imperative of combatting irregular migration. The resort to exit restrictions is not denied in this narrative. Rather, it can be ‘laundered into an ethically sustainable strategy of border governance’Footnote 101 where preventing exit protects migrants from smugglers and pull-backs rescue them from smugglers.
The role of departure States in this regime is to immobilize migrants unwanted by Europe by preventing departure and/or by pulling them back if they do set sail. The Italy–Libya MOU usefully illustrates the operation of this jurispathic tactic of eradicating refugee and human rights law and supplanting it with anti-smuggling/trafficking law and SAR. The preamble of the Italy–Libya MOU declares the parties’ ‘resolute determination to cooperate in identifying urgent solutions to the issue of clandestine migrants crossing Libya to reach Europe by sea’, and sets out eight substantive articles in furtherance of suppressing irregular migration.Footnote 102 These include Italian financial and technical support for Libyan border guards and coast guards, the creation of ‘reception centres’ to detain irregular migrants with a view to their voluntary or forced removal to countries of origin, and cooperation with international organizations that facilitate removal to countries of origin.Footnote 103 The MOU also commits to job creation initiatives to enable Libyans to replace income generated by migrant smuggling and trafficking.Footnote 104 The MOU appears to have enjoyed some success on this front: the same criminal organizations and militia that previously profited from the lucrative business of smuggling and trafficking in migrants have, thanks to Italy, transitioned into the more profitable occupation of policing the Libyan coast to prevent departure. As one Libyan militia leader put it, ‘Right now, in Libya, you are either in the smuggling business or the anti-smuggling business.’Footnote 105 These militia perform both tasks with exceptional violence, cruelty and lethality.Footnote 106 Nowhere in the Italy–Libya MOU do the words ‘refugee’ or ‘asylum’ appear, except by implication in the oblique undertaking to ‘interpret and apply the present Memorandum in respect of the international obligations and the human rights agreements to which the two Countries are parties’.Footnote 107 The right to leave any country and the right to seek asylum sink from view.
The Italy–Libya MOU, like the EU–Tunisia MOU, demonstrates how the asymmetry between entry and exit rights can be resolved by flipping the problem. Instead of rationalizing how the right to leave a State can be preserved despite an inability to enter another, the right to leave is simply discarded in order not to frustrate the right of another State to exclude. The new offence of ‘irregular departure’ does the necessary semantic work. Preventing ‘irregular departures’ is an explicit objective of the EU Anti-Smuggling Operational Partnerships with States of departure.Footnote 108 Preventing irregular departure articulates perfectly with preventing irregular entry. No asymmetry remains. The alignment of exit with entry is seamless.
This brings the discussion full circle: a right that Chétail describes as ‘fundamentally at the heart of the theory of human rights’,Footnote 109 a right whose breach was once regarded as a paradigmatic form of repression by communist States, turns out to be a right that is expendable. Today, of course, the threat to exit rights comes from destination StatesFootnote 110 who staunchly defended the right to leave, yet now regard exit restrictions as an attractive (or at least unobjectionable) mechanism of migration control.
C. Legal Accountability
Legal challenges to exit restrictions against both departure States and their destination State sponsors have recently gathered momentum in various forums. The prospects for accountability are dimmed by an array of practical, procedural and theoretical challenges that complicate the attribution of State responsibility and jurisdiction to adjudicate extraterritorial conduct under international and regional instruments. Indeed, one attraction of pull-backs from an EU perspective is that they distance destination States from the direct, physical contact that triggered liability for extraterritorial push-backs in the Hirsi Jamaa case. In other words, evading jurisdiction under human rights instruments is the whole point of destination State-sponsored pull-backs. This section offers a survey of current litigation against exit restrictions in regional and international forums. No case has proceeded to hearing, been adjudicated, or otherwise reached a resolution.
Before it was repealed, the pre-2023 Nigerien law was subject to a complaint to the ECOWAS Court of JusticeFootnote 111 on the basis that it impeded the free movement guarantees under ECOWAS and under Article 12 (right to leave) of the African Charter on Human and Peoples’ Rights.Footnote 112 Otherwise, most of the litigation to date targets Libya as departure State and Italy, Malta and/or Greece as destination States, as well as the EU Agency, Frontex.Footnote 113
A coalition of NGOs submitted a Communication to the UN HRC in 2020 regarding the role of Libya, Italy and Malta in violating, inter alia, the right to leave any country under Article 12(2) of the ICCPR.Footnote 114 The same coalition also initiated a request with the African Commission on Human Rights to investigate Libyan atrocities against migrants.Footnote 115
In an Article 15 Communication to the Office of the Prosecutor (OTP) of the International Criminal Court launched in 2022, the European Center for Constitutional and Human Rights, the International Federation for Human Rights, and Lawyers for Justice in Libya alleged violations of international criminal law in respect of Libyan pull-backs.Footnote 116 The Communication called on the OTP to investigate 24 named officials in Libyan, Maltese, Italian and EU agencies (including Frontex) concerning the coordinated EU–Libya interdictions of migrants in the Central Mediterranean by Libya. It argued that pull-backs constitute crimes against humanity, and identified the officials as co-perpetrators who formulated a ‘common plan’ to intercept and return migrants to Libya, making ‘essential contributions’ to operationalizing the plan, with knowledge of the exploitation and abuse (amounting to a crime against humanity) that Libya perpetrated on migrants and refugees.
Since destination States deliberately refrain from direct physical participation in pull-backs, or the enforcement of exit restrictions, attribution of State responsibility before international or regional institutions is particularly contentious. As previously noted, evading judicial scrutiny is widely assumed to be one of the objectives of the multi-actor, informalized, dispersed and externalized governance of European border control, including the measures directed at preventing ‘irregular departures’.Footnote 117
The International Law Commission (ILC) Draft Articles on State ResponsibilityFootnote 118 address customary law principles of attribution to States for acts or omissions that breach international legal obligations. Only States can formally invoke the responsibility of another State for its wrongful acts and pursue remedies of cessation and reparation before the ICJ. However, national, regional and international tribunals rely on the ILC Draft Articles for interpretive guidance in attributing State responsibility across a range of legal forums and contexts. Chapter IV of the Draft Articles addresses collaboration among States in the commission of wrongful acts, and Article 16 deals specifically with complicity in the form of one State aiding and abetting the commission of an internationally wrongful act by another State.Footnote 119 Establishing a destination State's complicity in a departure State's breach of the right to leave, along with the human rights violations entailed by pull-backs, detention and refoulement, would require proof that Libya's conduct constituted a wrongful act or acts and that Italy substantially contributed to the commission of the wrongful acts. Beyond the causal nexus, Italy's assistance would require knowledge of the circumstances that made Libya's actions wrongful and awareness of the serious risk that Libya would act in breach its international obligations.Footnote 120
Turning from general principles of international legal responsibility to the domain of international human rights law, the most significant procedural hurdle is jurisdiction. The attribution of wrongful conduct to a State is analytically distinct from the existence of a jurisdictional link between the State and the individual suffering harm under human rights law. The key question in respect of EU States is whether jurisdiction under the ECHR or the ICCPR extends to extraterritorial conduct where the State does not exercise physical control over the victims of alleged human rights violations. The answer remains unclear.
Among States that apply exit restrictions directly, Turkey, Libya, Tunisia, Cameroon and Niger are parties to the First Optional Protocol of the ICCPR.Footnote 121 Among destination States implicated in pull-backs, Greece, Italy and Malta are also parties. In early 2021, the HRC issued a decision under the ICCPR arising out of communications against Italy for its failure to rescue a ship in distress in 2013, which resulted in the drowning of more than 200 migrants, including at least 60 children.Footnote 122 The HRC determined that while the ship was located beyond the territory of both Malta and Italy, the requirement of a jurisdictional link between Italy and the victims was satisfied because the victims ‘were directly affected by decisions taken by the Italian authorities in a manner that was reasonably foreseeable’.Footnote 123 This approach to jurisdiction combines both causality (the effect of Italy's decisions on victims) and objective foreseeability of harm flowing from those decisions.Footnote 124
The Hirsi Jamaa case involved a push-back by Italy to Libya, where Italy's de jure and de facto control over the passengers on board Italian military ships—and thus extraterritorial jurisdiction under the ECHR—was relatively straightforward.Footnote 125 However, a case raising the issue of jurisdiction in the absence of physical control has been pending before the ECtHR since 2018. In SS and Others v Italy, the Global Legal Action Network and 17 survivors have alleged that Italy, pursuant to the Italy–Libya MOU of 2017, coordinated, facilitated and supported the Libyan Coast Guard's interference with the NGO Sea Watch's rescue of a vessel in distress and the subsequent interception of the vessel by the Libyan Coast Guard. Sea Watch rescued 59 passengers but at least 20 people drowned, and 47 were pulled back to Libya and into detention.Footnote 126
The ECtHR's criteria for extraterritorial jurisdiction, set out in Al-Skeini v United Kingdom and applied subsequently in Hirsi Jamaa, enumerate three bases for personal jurisdiction. The first emerges from acts of diplomatic or consular officials, the second arises from the State's exercise of public powers, and the third occurs where the State exerts actual authority and physical control.Footnote 127 None apply neatly to ‘contactless control’, where the identity of the State exercising power in a collaborative multi-State enterprise is not the State exercising physical control. In extraterritorial scenarios, the current interpretation of the three bases of jurisdiction risks incentivizing and rewarding States’ exploitation of this lacuna by developing governance models that ‘game the system’ to evade jurisdiction. At the same time, attenuating the nexus required between State and individual in order to establish jurisdiction risks creating a theoretically and practically unstable universalization of human rights responsibility.
Violeta Moreno-Lax responds to the risk of exploitation by States by proposing a ‘functional’ approach to jurisdiction that does not focus narrowly on the exercise of control at the precise moment of rights violation. Rather, her approach enquires whether the State asserts effective (but not necessarily exclusive) control in the creation and operation of a system that imposes material effects on individuals. Her reinterpretation of the ‘public powers’ branch of jurisdiction develops a conception of effective control that aggregates policy-making and operational implementation, as well as legislative, executive and adjudicative actions. She measures effectiveness by outcomes that determine a course of events culminating in interception and pull-back, rather than by insisting on physical force or presence at the moment of interception.Footnote 128 This revision attempts to confront and resist the tactical dispersal of control among States by refashioning the indicia of control to respond to models of governance that are informalized, complex, multi-actor and collaborative. One consequence is the possibility of concurrent jurisdiction of more than one State under international human rights law.Footnote 129
Vladislava Stoyanova is alert to the risk of universalizing human rights responsibility and cautions against expansive approaches to human rights jurisdiction that attenuate the required nexus between State and individual. She contends that the normative predicate of jurisdiction in human rights law (unlike general international law) presupposes an exclusive relationship between the legal subject and the sovereign that transpires within a bounded political community of interdependent political equals. Decisions and actions by the State distribute burdens and benefits within that community, in which members of the community participate and the State ostensibly governs to advance the common interest. Calculating and incorporating the interests of ‘outsiders’ to this relationship cannot be inserted easily into this model. Nor can the claim to an exclusive relationship of political authority accommodate concurrent sovereignty by more than one State even if, for example, Italy actively participates in the extraterritorial regulation of migrants. Stoyanova reaches the grim conclusion that existing jurisdictional principles ensure that the ‘interests of the individuals as protected by human rights law cannot be opposed to the interests of the States that actually exercise powers in ways that seriously harm these individuals’.Footnote 130
There is force to these arguments, although one may query whether Stoyanova takes sufficient account of the normativity of the relationship created between States and non-citizens through the principle of non-refoulement. The eventual ECtHR judgment in SS and Others v Italy may reveal whether a governance model that disperses authority and control across different actors, entities, instruments and operations can successfully ‘game’ the system of jurisdiction. If so, perhaps the law will permit Italy to do indirectly to migrants what Hirsii Jamaa says they cannot do directly, which is ensure the delivery of Italy-bound migrants and asylum seekers to Libya.
Even assuming that the jurisdictional hurdles in SS and Others v Italy can be surmounted, adjudication of the merits requires a determination of whether Libya's exit restrictions in law and practice violate the right to leave any country, such that Italy breaches its own obligations under the ECHR through its cooperation and assistance.Footnote 131 Since Libyan law imposes a general prohibition on exit by a non-citizen without a visa, the conclusion that Libya breaches its own legal obligation to respect the right to leave seems compelling.Footnote 132
A 2024 decision by a first-level Italian Civil Court illustrates how States can try to ‘change the channel’ on a human rights narrative and try to replace it with a SAR script. In March 2024, a SAR NGO (SOS Humanity) was conducting a rescue operation when the Libyan Coast Guard intervened and directed it to cease and depart. SOS Humanity refused, and was subsequently detained in Italy after disembarking survivors, on the grounds that it had ignored orders from Libyan authorities and thereby endangered the lives of the migrants. In other words, SOS Humanity unlawfully thwarted a Libyan Coast Guard SAR operation. The Civil Court of Crotone resisted the claim that the Libyan Coast Guard was, in fact, conducting a rescue operation, despite the narrative constructed by the Italy–Libya MOU. Instead, it found that the Libyan Coast Guard was armed and violent, firing shots at both the migrants’ and the SOS Humanity vessels, and subjected migrants to extreme abuse during and after their return to Libya. After reviewing the evidence, the court concluded that:
As things stand, Libya cannot be considered a safe place within the meaning of the [1979] Hamburg [SAR] Convention, as the Libyan context is characterised by gross and systematic human rights violations and Libya has never ratified the 1951 Geneva Refugee Convention. … All these elements are sufficient to exclude the existence of any qualification of the operations carried out by the Libyan coastguard … as rescue operations, in the sense recognised by the multiple international sources.Footnote 133
Assuming that higher courts are similarly unconvinced by the re-casting of Libyan pull-backs as benevolent SAR operations, the legal analysis will focus on whether Italy's complicity in Libya's breach of exit rights (among other violations) is justified by reference to the permissible justifications for restrictions on the right to leave under Article 2(3) of Protocol 4 to the ECHR. Limitations on the exercise of the right must be ‘in accordance with law and … necessary in a democratic society in the interests of national security or public safety, for the maintenance of ordre public, for the prevention of crime, for the protection of health or morals, or for the protection of the rights and freedoms of others’.Footnote 134
Halting ‘irregular migration’ appears to be the objective motivating Italy to collaborate in obstructing the right to leave Libya. The ECtHR considered the legitimacy of prohibiting exit from one State to prevent entry to another in Stamose v Bulgaria.Footnote 135 Bulgaria denied Stamose a passport for two years at the behest of the US, which had deported Stamose. Bulgaria enacted a law authorizing a travel ban to deter its nationals from violating the immigration laws of other States, in the hope that this would reduce the risk of other States—especially EU States—excluding Bulgarian nationals and/or imposing restrictive visa requirements. In its judgment, the ECtHR equivocated on the legitimacy of banning exit in the service of preventing entry:
Although the Court might be prepared to accept that a prohibition to leave one's own country in relation to breaches of the immigration laws of other States may in certain compelling situations be regarded as justified, it does not consider that the automatic imposition of such a measure without any regard to the individual circumstances of the person concerned may be characterized as necessary in a democratic society.Footnote 136
The ECtHR did not elaborate on what might constitute a compelling circumstance. In 2016, Macedonia admitted to obstructing the exit of Roma nationals in the interests of advancing visa liberalization with the EU, and conceded before the ECtHR that it breached the freedom of movement, right of exit and non-discrimination of Roma applicants under the ECHR.Footnote 137 Although the jurisprudence is tentative, it does not sanction a generalized prohibition on exit in the service of enforcing other States’ immigration laws, especially in the case of Libya (and Tunisia), where all departures require an exit visaFootnote 138 and the restriction is automatic, has no regard to individual circumstances, and precludes access to any asylum process.
To the extent that Italy can re-cast its objectives as the protection of migrants from deadly voyages and/or combatting human smuggling, analysis shifts to the means deployed to attain the objectives. It might be inferred that evidence of Libya's conduct required to establish jurisdiction under the ECHR would undermine the argument that facilitating interceptions and pull-backs to Libya is an acceptable means of protecting the lives of migrants or combatting smuggling. Less restrictive alternatives that could better protect human rights and/or reduce smuggling would include resumption of destination State SAR operations, refugee resettlement, asylum/humanitarian visas, and other regular pathways to protection. This would require Italy and other EU States redirecting the collaboration, cooperation and coordination currently expended on repelling migrants and refugees into the creation of a regime that would necessarily result in the lawful admission of more refugees (and possibly other migrants) to European States.
If these alternatives seem unrealistic and infeasible in the present or foreseeable future, perhaps it is because they clash with the actual objective of Italy and other States, which is simply to prevent irregular migration, whether or not smugglers are involved, and whether or not the migrants are refugees. This acknowledgement flows downstream into the final stage of a proportionality analysis, namely the balancing of the interests of those subject to interception, pull-back and detention in Libya, and Italy's interests. The right to leave is subject to various potential justifications that would ordinarily permit a balancing of interests but recalling that a right to leave is indispensable to the realization of a right to seek asylum, this justification for upholding a limitation on exit rights seems especially weak where the people exercising a right to leave may be refugees.Footnote 139
The dilemma of balancing extends beyond the formal question of whether the ECtHR will explicitly engage in a proportionality analysis if SS and Others v Italy reaches the adjudication phase. The claim that Italy violates the right to exit by orchestrating and assisting Libyan interdiction and pull-backs necessarily confronts the competing claim that a State must be able to exercise migration control as a condition of self-determination and sovereignty. A confounding aspect of this migration control argument is that it does not actually rely on the benefits of a given law, policy or practice beyond the fact that it instantiates control. Similarly, the harm averted by enforcement may be unrelated to a specific risk posed by any particular migrant. Instead, the harm consists of the diminution of control over admission that results from the imposition of human rights and rule of law constraints. This ostensible justification for migration control is at once highly abstract and existential, in which the admission of people as a matter of legal obligation rather than sovereign prerogative is the harm; every gate is a floodgate on this view. This is erroneous. Requiring more from the State in terms of justification does not entail the repudiation of State power to regulate migration (open borders), but it does impose a duty to defend limitations on the right of exit with more than a generic assertion that it advances border control.
IV. MOBILITY AND THE NEW CARCERAL STATE
Formal exit restrictions remain exceptional. The offence of ‘illegal departure’ seen in Libya, Tunisia, Niger, Cameroon and Turkey over the last decade serves a different purpose from its Cold War-era predecessors, or even current restrictions in Eritrea or North Korea.Footnote 140 Yet, the novelty of criminalizing exit recedes when set against the criminalization of migration more broadly. The rich legal and interdisciplinary scholarship on the theme of crimmigration and border criminology documents the nexus between criminal and immigration law, policy, practice and discourse.Footnote 141 A prominent illustration of the ‘crimmigration’ phenomenon is the transformation of unauthorized entry into the crime of ‘illegal entry’, a matter previously regulated through administrative law as breach of an immigration statute.Footnote 142 Criminalizing exit extends the logic of criminalizing entry, and takes it further. Whereas criminalizing entry turns the wrong of an administrative offence into a crime, the criminalization of exit has the added feature of transforming the exercise of a right (to leave) into a wrong.
The evolution of migrant detention also lies at the nexus of criminal and immigration fields. Migrant detention is formally classified as administrative detention, but it has become harsher, more ‘prison like’ and more punitive as the migrants subjected to it have become increasingly criminalized as ‘illegals’ in public discourse and in law. Although the legitimate uses of migrant detention must relate to the administration of immigration law,Footnote 143 detention is frequently an instrument of specific and general deterrence—a tool for inflicting suffering on detained migrants so they give up and depart, and for warning prospective migrants of the suffering that awaits them if they attempt the journey. The experience of detention as punishment means that migrant detention operates not merely as a space of containment, but as a specifically carceral space.
Exit restrictions also matter to our understanding of migrant detention and carcerality. When a departure State criminalizes exit, enforcement amounts to confinement on the territory of the departure State on behalf of the destination State. This invites a reimagining of the metaphor of the ‘carceral State’. Criminology scholars invoke the term to identify the punitive ideologies, institutions and practices of confinement, containment, surveillance and incarceration that extend beyond the traditional radius of the penal system within a State. The concept of the new carceral State is invoked here to convey how the territory of a State can function in toto as a punitive space of confinement in relation to other States. Libya is Italy's carceral State, Cameroon is a carceral State for France. Turkey, Niger and Tunisia are carceral States for Europe's unwanted migrants. When people try to escape, they may be apprehended, returned and punished by confinement within those States.
Thinking about the ‘carceral State’ in this way also invites attention to the links between the way exit restrictions confine people to a State, and the way migrant detention confines them within a State. In the recently overturned Australian case of Al Kateb v Godwin, the Australian High Court offered a rationale for indefinite migrant detention where deportation was not foreseeably possible. The court reasoned that if the State has the power to exclude non-citizens from the territory, it must also possess the power to ‘permit … exclusion from the Australian community’—by indeterminate separation from the Australian community via detention if an inadmissible non-citizen managed to enter the territory.Footnote 144
Although the High Court subsequently repudiated ‘separation from the community’ as a legally cognizable purpose of migrant detention, the political force of this logic should not be underestimated.Footnote 145 On this view, detention is not merely a prelude to exclusion from the State, or a response to a specific risk posed by the individual, but a third-best mechanism of exclusion, ranking somewhere below prevention of entry and post-entry expulsion. Segregation via detention prevents entry into the socio-political space of the State following the failure to prevent entry to the territorial space. Indeed, this resembles the function of refugee camps in the Global South, as well as Europe's refugee detention camps on Greek islands or Lampedusa in circumstances where States lack either capacity or will to prevent entry physically or effect the expulsion of refugees en masse.
It is important to situate detention within the larger trajectory of exclusion measures that comprise bordering regimes. It begins in countries of origin, continues through transit countries, and stretches into detention centres, segregating non-citizens within the territory of a destination or transit State where entry has not been prevented. Migrant detention resembles imprisonment in its punitive aspect, but it differs from penal incarceration because the segregation from community that detention imposes is not ancillary to the project of controlling human movement, it is an instantiation of it.
In most cases, migrant detention is a rear-guard action by destination States once the attempt to deter entry has failed, or where non-citizens have lost their status and been marked for deportation. But exit restrictions insert detention into the migration trajectory as a pre-emptive tactic. In countries of departure like Libya, detention aids in precluding migrants’ access to Italy. Alison Mountz's concept of the detention archipelago graphically illustrates how islands like Manus Island, Lampedusa and Guantanamo Bay operate as containers for unwanted migrants.Footnote 146 Susan Coutin recounts how Central Americans detained and deported from the US experience their country-of-origin post-return as a ‘zone of confinement’ that, in practical terms, they cannot leave.Footnote 147 She observes that:
The fact that national territories in some ways resemble detention centers—both of these confine, both restrict movement—challenges liberal notions of nations as entities through which individuals can realize their capacities.Footnote 148
Libya's role as the EU's extraterritorial prisoner-of-migration camp makes the confinement Coutin describes both literal and legal: the Libyan State apparatus (such as it is) deliberately incarcerates inhabitants on its territory for the ‘crime’ of irregular departure. Libya not only deflects migrants from the EU, but it also holds captive those whom the EU imagines as prospective migrants. Libyan ‘reception centres’ are smaller, grislier containers nested within a larger one. In Libya, migrant detention is to territorial containment what solitary confinement is to imprisonment—a more concentrated version of the same technology. Foregrounding the carceral dimension of exit restrictions reveals that what is ultimately at stake is the governance of mobility through immobilization.
V. CONCLUSION: TOWARD THE GOVERNANCE OF MOVEMENT
The claim that a State is sovereign to the extent that it controls its borders is a truism of conventional political discourse. Yet, if exclusion enacts State sovereignty, then the exit restrictions described in this article arguably undermine it. The European project of financing, equipping and training State and non-State actors to police the movement of people within and across African States’ borders gives the appearance of enhancing the border governance capacity of those African States, and thereby their sovereignty. And yet, the very act of subordinating their own migration governance to the interests and will of another sovereign, often at the expense of their own interests, only diminishes the sovereignty of these departure States.Footnote 149
One may detect whiffs of imperial arrogance in the ever more expansive reach of European governance into the domestic policies of foreign States. After all, controlling the mobility of the ‘natives’ in the colonies and overseas territories was a preoccupation of imperial powers.Footnote 150 Indeed, the merits of restricting the emigration of Brown and Black citizens of Global South countries in the name of averting a ‘brain drain’ was—and remains—a debatable topic of academic inquiry. Sara Dehm reviews the historical record within the UN concerning the denunciation of exit restrictions imposed by Communist States versus the ambivalence around ‘brain drain’ restrictions. She concludes that the divergence is revealing of the compromised status of citizens of the Global South as subjects of international human rights law:
While the people of the ‘highly industrialized’ world are placed in the position of the universal subject of international human rights law (and are granted a largely-unencumbered ‘right to leave’), the people of the Third World are instead enrolled within the enterprise of Third World development through the very interpretation of this ‘right to leave’.Footnote 151
But now, when Thomas Spijkerboer observes that ‘[p]opulation management in the Middle East and Africa is, once again, a legitimate subject of European policy’,Footnote 152 the important difference is that the populations are today being managed by Europeans for the ostensible benefit of Europeans.
That said, perhaps what is detected are also whiffs of desperation. Europe's extraterritorial forays into migration governance of departure States can be read as a tacit admission of depleted sovereignty, at least as measured in terms of effective border control. Destination States perpetually and inexorably fail to ‘manage’ migration within the margins of domestic political acceptability and/or legal legitimacy.Footnote 153 It seems that no political leader can afford to admit the limited capacity of their State—and all States—to fully control human movement. Instead, leaders continue to propound that perfecting border control is within reach, if only the right kind, combination and intensity of technology and force are expended. Even where departure States are recruited into aiding destination States in the task, leaders will still, inevitably, fail to satisfy populist demands to fortify the border ever more intensively in the name of sovereignty.
At the same time, the articulation of exit restrictions with entry restrictions—the seamless border—opens a window onto what a networked regime of circulation could look like. If it is accepted that realizing the goal of ‘managed migration’ legitimates regulating the movement of people to, from and inside Niger, for example, then controlling migration begins to converge with controlling movement. Where mobility itself is the target, borders and nationality become technologies to be activated in order to pool, distribute, and aggrandize States’ capacity to surveille, direct and contain movement as needed and as desired.
One can lament States’ defection from their international human rights obligation to respect and protect the right of exit, but the recrudescence of exit restrictions does more than demonstrate that the line between frustration and abrogation of exit rights is decomposing. It also suggests that a liberal model that proceeds from a premise of free movement, and then interposes episodic State-imposed obstacles to transnational movement, decreasingly fits with reality. States and regional units are expanding and coordinating their techniques of governance to subject human movement to conditions where it is knowable and controllable. The shift might be described as a departure from State-driven governance of borders (wherever they are deemed to arise), and toward the governance of movement.Footnote 154 The anodyne objective of ‘managed migration’ promoted in global State-led conversations cannot but generate a momentum to govern human circulation even before it begins.
The endurance of securitization following 11 September 2001 (9/11), accompanied by enhanced techniques of AI, surveillance, data-gathering, information-sharing, cooperation and coordination, is nudging governance away from simply the management of exit or entry per se, toward the assumption of control over movement as such. Although specific attention to COVID-19 pandemic restrictions lies beyond the scope of this article, the activation of surveillance and policing of movement inside State territory to contain the spread of the virus demonstrated both the will to govern movement of all people within State territory as well as across borders, and the rapid advancement of technologies to enhance the capacity to do so.Footnote 155
The investment by States in their capacity to govern movement invites us to reconsider the premise of free movement that is interrupted by intermittent State impediments erected at or near borders, or embassies, or airports. One might flip the picture and begin by identifying multiple, proliferating, dispersed and networked technologies that anticipate, detect, constrain, arrest or compel movement. Humanity is already subject to this regime of governing circulation. Privileged citizens of wealthy States do not notice the digital walls, the biometric gates and the remote-control portals because they are less visible and violent than their material analogues and because one passes through them unhindered, but they are no less operational for that reason. Individuals, especially those regarded as highly mobile, might subjectively experience their movement as free, unimpeded and unrestrained, but what is perceived today as free movement may be drifting toward permitted movement. One might not realize it until permission is denied. This is one reason to pay closer attention to the quiet erosion of the right to leave.
ACKNOWLEDGEMENTS
The author especially thanks Jaya Ramji Nogales and Thomas Spijkerboer for incisive and constructive criticism, and participants in the Global Migration Law workshop and the MOBILE Centre workshop at the University of Copenhagen for excellent feedback and suggestions. Gratitude is also extended to Tommy Pham and Lisa Ariemma for superb research and editorial assistance.
Open access
