The Borders of the Law: Legal Fictions, Elusive Borders, Migrants’ Rights

Bordering processes take place through different means and are carried out by different actors. Laws and regulatory activ‐ ities have a prominent place among border‐drawing instruments: Their capacity to mobilise actors, allocate funds, and determine procedures and remedies make them a formidable and multifaceted bordering tool. It is therefore not sur‐ prising to notice that EU institutions have heavily relied on regulatory tools when the need to resort to new bordering processes emerged in the aftermath of the so‐called migration crisis. This article delves into a particular (re‐)bordering process emerging from the legislative proposals attached to the Commission’s 2020 New Pact on Migration and Asylum: the attempt to uncouple the duty to fully respect and protect fundamental rights from the reality of migrants’ presence on national territory. This objective is pursued by the proposed legislative package through non‐entry fictions, capable of untangling the legal notion of “border” from its physical reality for the purpose of immigration law (only). The analysis of the relevant provisions provides the reader with a number of insights into the transformation of EU borders. First, borders (as defined by the law) are subject to a peculiar legal regime. Secondly, the legal notion of borders is increasingly indepen‐ dent of its physical/geographical correspondence. Thirdly, legal border lines are not linked to any place on the ground, but rather follow irregular migrants as they move, confining them to areas of less law, no matter their location.


Introduction
State borders are physical places designated to delineate national territory. They have been traditionally understood as defining the boundaries of a state's sovereignty and jurisdiction (Ryngaert, 2017, p. 53). The anchoring of sovereignty/jurisdiction in the real-world element of space has been defined as "legal spatiality" (Raustiala, 2006, p. 219). It underpins not only theoretical approaches to borders but also state practice and case-law (Al Skeini v. UK, 2011, para. 131). Although contested as to its relevance in a world with fluid and flexible borders (see, among others, Appadurai, 1996), the possibility of geographically determining the boundaries of states remains foundational for national legal systems. These systems encompass a world of institutional actors, principles, rules of conduct, and enforceability mech-anisms that have the national territory as their stage. Every person on a state's territory must respect its laws and is subject to its enforcement powers. In the large majority of cases, from the individual perspective, being subject to a legal system depends on the objective factor of physical presence in a certain place. The link between the physical reality of territory and the social construction of the legal system is expressed by the concept of territorial jurisdiction.
Territorial jurisdiction has long been the dominant lens to assess the reach of a state's powers and the extent of its responsibility to protect fundamental rights. "Classic" instances of extraterritorial jurisdiction, for example, that of the flag state over ships in the high seas (United Nations Convention on the Law of the Sea, 1982, Art. 92), have often been characterised as exceptional (Bankovic and Others v. Belgium and Others, 2001, para. 61). More importantly, they have themselves been grounded on an objective "spatial" element (i.e., presence on board of a specific ship). Be that as it may, the territoriality of the legal order has evolved in connection with what is broadly referred to as globalisation. The quasi-coincidence between the geographical borders of the state and the reach of its laws and enforcement powers has been supplanted in certain cases by new models of jurisdiction (Raustiala, 2006, p. 220). The EU legal order makes no exception: In recent years, migration and border control laws and practices of the EU and its member states have weakened the physical hook on which the legal construction of jurisdiction relies. The so-called migration crisis and its aftermath have been particularly effective in pushing EU policy-makers to rethink territoriality and resort to new bordering processes.
A first example of (de-)bordering process relies on a narrow interpretation of extraterritorial jurisdiction. Instrumentalising the traditional link between territory and jurisdiction, EU member states such as Italy artificially and intentionally withdraw their ships from the Mediterranean and sanction rescuing efforts (Basaran, 2014, p. 374). In so doing, they avoid the on-boarding of migrants, which would trigger the responsibility to protect their rights. The avoidance of direct contact with migrants is coupled with the continued presence of equipment capable of detecting boats in distress in the Mediterranean and communicating their location to the authorities of third states (so-called "contactless control" or control "by-proxy"; see Moreno-Lax, 2020, p. 387). The legal literature has responded to this trend by affirming the need to resort to a "functional" model of jurisdiction, "predicated on the exercise of public powers, such as those ordinarily assumed by a territorial sovereign" (Moreno-Lax, 2020, pp. 386-387). According to this model, whenever public powers are exercised by a state, its jurisdiction (including its obligation to fully respect and protect fundamental rights) should be triggered. The concept of functional jurisdiction echoes that of functional borders, developed from the perspective of territoriality and its evolution and applied to similar cases (Riccardi & Natoli, 2019, p. 9).
A reflection from the perspective of territoriality is still largely missing concerning a second kind of (re-)bordering process: The mandatory application of non-entry fictions proposed by the New Pact on Migration and Asylum (European Commission, 2020a). Non-entry fictions untangle the legal notion of "border" from its physical reality for the purpose of immigration law (only). In essence, when migrants cannot be physically "exclud[ed] from territory," the Commission proposes to modify the legal implications of their presence on the territory, excluding them from "rights" (Moreno-Lax, 2018, p. 120). The result is a set of proposed norms whose effect is to keep certain categories of persons from ever being able to access the full protection granted by a certain legal system, regardless of their physical location. The non-entry fiction is a process of exclusion, but also a process of illegality and invisibility creation. By denying entry to migrants already present on their territory, EU member states make them illegal (see De Genova, 2002, p. 432). The non-entry fiction renders exclusion dynamics invisible by relying on the language of illegality (see Sati, 2020, p. 23). At the same time, as discussed below, it justifies quasi-systematic detention, rendering irregular migrants invisible to the rest of society (Van Houtum & Bueno Lacy, 2020, p. 721). As this is done through a fictional exercise, the borders of legality can be pushed indefinitely inward to prevent migrants from attaining them. This results in an uncomfortable uncoupling of reality from the legal system, which permanently fails to "see" and "be seen" by certain categories of persons. This uncoupling is not entirely new: A number of EU member states already apply non-entry fictions to international airports (ECRE, 2021, p. 25). Moreover, non-entry fictions resulting in the application of lesser procedural standards in certain parts of the borders in the context of so-called "border procedures" (a) have been accepted in principle by the ECtHR (e.g., Saadi v. UK, 2008, para. 65) and (b) are already allowed under EU law (Rasche & Walter-Franke, 2020, p. 4). However, the New Pact plans to turn this option into an obligation and to extend the reach of border procedures, so that they can take place to an unprecedented extent within member states' territories.

Let's Pretend They Are Not Here: The Elusive Nature of Shifting Borders
Non-entry fictions (entailing a distinction between the "physical" entry on the territory and a "legally recognised" entry on the territory) are not an invention of the European Commission. More than two decades ago, US law (Illegal Immigration Reform and Immigrant Responsibility Act, 1996) abandoned its earlier distinction between migrants who have (physically) entered the territory and those who have not to determine the level of procedural guarantees owed to them during removal procedures. The physical entry/non-entry divide was set aside in favour of a distinction based on admission. Migrants who have entered the territory, but who have not been admitted to it by competent authorities, have since been placed in the same position as those who have never physically entered the territory. As noticed by Bosniak (2002), this has created a hard-to-justify divide between those migrants who have overstayed their visa or visa-free period, and those who have never obtained a visa. Given the cost of a visa and the nationality-based criterion to determine who can enter the USA without one, this distinction has inevitably tended to run along lines of nationality and social class. As mentioned, the non-entry fiction is not extraneous to the EU legal regime itself. Currently, Article 29(2) of the Convention Implementing the Schengen Agreement (2000) specifies that member states, when complying with their obligation to process asylum applications lodged within their territory, maintain the right to refuse entry to the asylum seekers concerned. Similarly, in certain cases and for a period of up to four weeks, Article 43 of the Asylum Procedures Directive (European Parliament and Council Directive of 26 June 2013, 2013) allows member states to examine asylum applications while refusing to access their territory. However, at the moment, EU law does not require member state to apply non-entry fictions.
In its new Pact on Migration and Asylum, the Commission proposes precisely to mandate the large scale application of the non-entry fiction throughout the EU. A discussion of the proposed reform cannot but start with a short overview of the relevant provisions.
According to Article 8 of the Proposed Asylum and Migration Management Regulation, "member states shall examine any application for international protection by a third-country national or a stateless person who applies on the territory of any of them, including at the border or in transit zones" (European Commission, 2020b, Art. 8, emphasis added). This provision is coherent with the "geography" of a state's territory, which includes its borders and transit zones. In line with this approach, Article 21, entitled "Entry," affirms that: Where it is established…that an applicant has irregularly crossed the border into a member state by land, sea or air having come from a third country, the first member state thus entered shall be responsible for examining the application for international protec-tion….Th[is] rule…shall also apply where the applicant was disembarked on the territory following a search and rescue operation. (European Commission, 2020b, Art. 21, emphasis added) In other words, for the purpose of identifying the member state responsible for an international protection application, entry means entry(!). Geographical presence on the territory of a member state triggers its jurisdiction, including its power to apprehend the migrant, subject it to administrative and judicial proceedings, restrict his or her freedom and even detain him or her.
We will see that, when it comes to the parallel triggering of migrants' rights, the legal definition of entry is much more restrictive than physical entry.
To start, Article 3 of the Proposed Screening Regulation (European Commission, 2020c) requires member states to apply screening "at the external border," among others, to "all third-country nationals who apply for international protection at external border crossing points or in transit zones and who do not fulfil the entry conditions." Articles 4 and 6(1) specify that those third-country nationals "shall not be authorised to enter the territory of a member state" (European Commission, 2020c, Art. 4, emphasis added). The non-entry fiction applies even if the screening is carried out in the territory and, more precisely, "at locations situated at or in proximity to the external borders" (European Commission, 2020c, Art. 6(1)). To summarise, according to the Proposed Asylum and Migration Management Regulation, migrants at borders and transit zones are considered to have entered the territory of a member state. At the same time, according to the Proposed Screening Regulation, they are considered not to have entered it.
The fiction steps further away from reality in Articles 5 and 6 of the Screening Regulation. These provisions impose the application of the screening procedure at an appropriate location within the territory of a member state (for "third-country nationals found within the…territory [of a member state] where there is no indication that they have crossed an external border to enter the territory of the member states in an authorised manner" [European Commission, 2020c, Art. 5, emphasis added]). In itself, the idea of screening someone who has managed to enter the territory undetected is logical. The need to identify those who are on the territory of the member state does not depend on the location where they are first confronted with the authorities. What is problematic is the link between screening, non-entry fiction, and border procedures. Article 41(1) of the Proposed Amended Common Procedure (ACP) Regulation (European Commission, 2020d) establishes that "following the screening procedure…and provided that the applicant has not yet been authorised to enter member states' territory, a member state may examine an application in a border procedure." What matters for the border procedure is not whether the person has actually entered the territory, but rather whether he or she has done so in an authorised manner. If this is not the case, an asylum border procedure applies, entailing a limitation of the applicant's procedural rights so significant that it will almost inevitably affect the protection of his or her substantive rights (for example the right to non-refoulement; see Moreno-Lax, 2017, p. 459). The start of the border procedure is only the beginning of a non-entry fiction that can be protracted for several months. Member states are required to exercise their authority upon international protection applicants and even detain them for several months without authorising them to enter the territory and enjoy the full procedural protections that would apply there. The non-entry fiction remains intact even in case of relocation: Member states' authorities can transfer applicants to another member state without the need to previously acknowledge their presence on EU soil.
According to Article 41(5) of the Proposed ACP Regulation, applicants "shall not be authorised to enter the territory of the member state" throughout the asylum border procedure, which can last up to 12 weeks (European Commission, 2020d). "Following that period, the applicant shall be authorised to enter the member state's territory," safe when one of the numerous exceptions apply. The first one, provided by Article 41(11), read in combination with Article 41(a), provides that international protection applicants whose application has been rejected in the context of the asylum border procedure "shall not be authorised to enter the territory of the member state" (European Commission, 2020d). Thus, at a closer look, only international protection applicants whose asylum procedure is taking longer than 12 weeks can, in principle, be authorised to enter the territory of the member state under Article 41(5). Even after the expiry of the 12 weeks deadline though, the authorisation to enter must be denied to (a) applicants whose first asylum application has already been rejected and (b) applicants who have not requested or obtained the right to remain pending an appeal. As a rule, the asylum border procedure is carried out "at or in proximity to the external border or transit zones" (European Commission, 2020d). Nonetheless, it can be extended to "other locations within the territory" of the member state on a temporary basis, when the capacity at the borders and transit zones is insufficient. In this case, and already during asylum processing, the disconnect between the lack of authorisation to enter the territory and the actual transfer of applicants within the territory is particularly evident. While this disconnect is framed as exceptional for asylum border procedures, it constitutes the norm for subsequent return border procedures. Article 41a of the Proposed ACP Regulation affirms that: Third-country nationals and stateless persons whose application is rejected in the context of the procedure referred to in Article 41 shall not be authorised to enter the territory of the member state.…[They] should be kept for a period not exceeding 12 weeks in locations at or in proximity to the external border or transit zones; where a member state cannot accommodate them in those locations, it can resort to the use of other locations within its territory. (European Commission, 2020d, Art. 41a, emphasis added) As a result of this provision, border return procedures are characterised by a fictional refusal of access to the territory, coupled with a physical and prolonged detention of the person concerned within such a territory.

Why Pretend? The Implications of the Non-Entry Fiction
The analysis conducted so far has laid bare the legislative uncoupling of law from reality, but it has not yet delved into its effects. Legislatively implying that migrants who are on the territory should not be considered to be there does not alter the reality of their presence, either for them or for national authorities interacting with them and bearing the costs of this presence (Rasche & Walter-Franke, 2020, p. 5). However, it allows for a legal construction whereby migrants not authorised to enter the territory do not have access to the full set of rights that their presence in the member state would otherwise entail. Migrants in border procedures are subject to the power of enforcement of the competent member state and have a duty to cooperate with national author-ities, much like migrants in "regular" asylum (European Council Directive of 26 June 2013, 2013) and return procedures (K. A. and Others v. Belgische Staat, 2018, para. 105). Nonetheless, the fundamental rights restrictions mandated for them by the legislative framework are much more significant.
To start, in the context of asylum border procedures on the merits, an international protection application must be examined in an "accelerated" manner. This acceleration constitutes a "b/ordering practice" of the kind that "tells [EU citizens] that the fundamental rights to which the EU adheres do not fully apply to undocumented migrants" (Van Houtum & Bueno Lacy, 2020, p. 722). It reduces the time granted to applicants to prepare their claims and to adjudicators to examine all relevant elements. According to Article 41(10) of the Proposed ACP Regulation, the applicant has five days from registration or relocation to apply for international protection and the whole procedure is in principle to last no longer than 12 weeks, including the appeal stage. It will be for EU courts to establish whether this period is "sufficient in practical terms to enable the applicant to prepare and bring an effective action" (Samba Diouf v. Ministre du Travail, 2011, para. 66), as required by Article 47 of the EU Charter of Fundamental Rights (EU Charter). The accelerated border procedure applies to the assessment of the merits of asylum applications not only when applicants have been uncooperative with the authorities or are considered to be a danger to national security and public order, but also when they come from a third country for which the average recognition rate in the Union is 20% or lower. Such a percentage raises to 75% in situations of crisis, as defined-quite broadly-by Article 1 of the Proposed Crisis Regulation (European Commission, 2020e). This nationality-based criterion to determine the extent of the applicants' procedural rights is problematic for several reasons. First, the very idea of a nationality-based criterion to determine the extent of one's procedural right appears to contradict the non-discrimination principle enshrined in both Article 3 of the Convention Relating to the Status of Refugees (1951; see Mouzourakis, 2020, p. 175) and Article 21 of the EU Charter (Slovak Republic and Hungary v. Council, 2017, para. 305; see also Carrera et al., 2019, p. 31). Secondly, this criterion does not sit easily with the "individual nature of any application for international protection" (Carrera, 2021, p. 8). Thirdly, recognition rates vary widely among member states (European Commission, 2016, point 5(2)). Relying on a Union average without having previously taken measures to ensure a certain homogeneity of recognition rates risks leading to arbitrary results. Fourthly, when the Proposed Crisis Regulation is triggered, the threshold set for a certain nationality to be admitted to the regular procedure border procedure becomes so high (75% average recognition rate) that the border procedure is transformed into the standard way of examining the merits of international protection applications (Mouzourakis, 2020, p. 175).
Besides the curtailing of the time granted to each applicant in the context of border procedures, the latter also entail systematic detention. Van Houtum and Bueno Lacy (2020, p. 721) define the border camp as one of the pillars of the Union's bordering practices, for its role in the singling out of irregular migrants as "different" from the rest of society, while at the same time hiding them from sight. International protection applicants in border procedures must "be kept" at borders, in transit zones, or at specific locations within the territory. In practice, this means that they will "be isolated from the rest of the population" and obliged to "remain permanently in a…zone the perimeter of which is restricted and closed, within which [their] movements are limited and monitored, and which [they] cannot legally leave voluntarily, in any direction whatsoever" (FMS and Others v. Országos, 2020, paras. 217 and 231). This condition amounts to detention according to the recent case-law of the Court of Justice of the EU (CJEU) or, at least, to a significant limitation of liberty according to the less protective case-law of the ECtHR. The latter's judgment in Ilias and Ahmed v. Hungary (2019) is puzzling in several respects, from the confusion between the definition of detention and the determination of its necessary character (paras. 232-233) to the consideration of an illegal departure towards Serbia as a relevant option for the applicants (paras. 237-238). Be that as it may, based on Article 53 of the EU Charter, the CJEU is entitled to go further than the ECtHR in protecting fundamental rights, and member states are bound to comply with such a higher level of protection when they are acting within the scope of EU law, as per Article 51(1) of the EU Charter. With this in mind, one can conclude that, at least under EU law, border procedures will result in de facto systematic deprivation (rather than limitation) of liberty (Cornelisse, 2021). This is problematic, as it deprives the principle according to which detention should be a measure of last resort (see, among others, El Dridi, 2011, para. 39;K. v. Staatssecretaris, 2017, paras. 46-48) of any practical meaning. The detention of asylum applicants for up to 12 weeks (20 in case of crisis) can be followed by another equivalent period of detention in the context of the return border procedures. Not only are the latter accelerated and accompanied by systematic detention. Border return procedures might also fall outside the scope of application of most minimum procedural guarantees enshrined in the Return Directive (European Council Directive of 16 December 2008, 2008). According to Article 2(2)(a), member states are allowed not to apply the Return Directive to third country nationals "apprehended or intercepted by the competent authorities in connection with the irregular cross-ing…of the external border of a Member state and who have not subsequently obtained an authorisation or a right to stay in that member state." The right of appeal of migrants in border procedures is limited not only in terms of the deadline to challenge a first instance decision but also in terms of the number of appeals: According to Article 53(9) of the Proposed ACP Regulation, "member states shall provide for only one level of appeal" in these cases (European Commission, 2020d). This limitation complies with the right to an effective remedy as interpreted by the CJEU (Gnandi v. État belge, 2018, para. 57;Samba Diouf v. Ministre du Travail, 2011, para. 69), but it might create problems of compatibility with certain national constitutional orders which do not admit curtailing of the number of degrees of appeals available for particular sets of proceedings (Muir & Molinari, 2019, p. 56).
Under Article 5(1)(c) of the Proposed Crisis Regulation, the rights of migrants in return border procedures are further curtailed by means of the introduction of a presumption of risk of absconding in most cases. Such a presumption does not only entail the automatic deprivation of liberty. If the recast Return Directive was adopted as proposed by the Commission (European Commission, 2018), it would also compromise the availability of voluntary departure options and lead to the imposition of re-entry bans.
Finally, the strong limitations applied in the context of border procedures to the suspensory effect of appeals run the risk of violating the principle of non-refoulement, as guaranteed by the EU Charter and interpreted by the CJEU. According to Article 54(2)(a) Proposed ACP Regulation, when border procedures are applied, the lodging of an appeal against a first instance decision cannot be automatically suspensive. This means that the person concerned might be returned to a third country before a final appellate judgment is rendered. Besides the problematic nature of this and other EU-level norms aiming at determining a maximum, rather than minimum, level of fundamental rights protection (Muir & Molinari, 2020), it is not difficult to see that this provision compromises the right to a judicial remedy and, as a consequence, places concerned migrants at risk of been returned to a place where they will be subject to inhuman or degrading treatment. The CJEU has itself recognised in Centre public d'action sociale v. Abdida (2014, para. 46) that EU secondary law, read in light of Articles 19 and 47 of the EU Charter, "must be interpreted as precluding national legislation which does not make provision for a remedy with suspensive effect in respect of a return decision whose enforcement may expose the third-country national concerned to a serious risk" of inhuman or degrading treatment.
The considerations developed above are all the more worrying if we consider that border procedures are meant to apply not only to adults, but also to minors (European Commission, 2020d, Arts. 41(5), 40(5)(b)). Quasi-systematic detention of minors and severe limitations of their procedural rights are especially problematic in view of their particular vulnerability, as recognised even by a court as sensitive to border-control arguments as the ECtHR (see, among others, Kanagaratnam and Others v. Belgium, 2011;Mahmundi and Others v. Greece, 2012)

Conclusion
This short analysis of the non-entry fiction envisaged in the New Pact on Migration and Asylum from the perspective of territoriality paves the way for a few conclusions. First, physical borders are increasingly subject to a peculiar regulatory regime, characterised by the full application of the member state's enforcement and regulatory powers, on the one hand, and by a limited application of standard procedural guarantees directed at protecting migrants' rights, on the other. The transformation of external borders into "anomalous zones" (Campesi, 2021), which has so far been operated only by some member states or in specific areas of the borders (so-called hotspots), is now becoming generalised. This results in the de facto transformation of borders into something more similar to "frontiers," namely "peripheral spaces that are managed, where citizens do not live" (Linden-Retek, 2020, p. 41). Secondly, through the EU-wide application of non-entry fictions, the legal notion of borders is destined to set itself almost completely free from its physical/geographical correspondence. The specific legal regime reserved to borders can be applied also within the territory, in any designated location, by virtue of the characteristics of the migrants concerned (e.g., their nationality and past behaviour). As discussed, the disconnect between geography and legal construction in no way affects the state authorities' enforcement and regulatory powers, but it does limit the array of rights and remedies available to migrants who are identified as falling within real or imagined border zones. In this context, the call for a "functional" model of jurisdiction equating the exercise of public powers with the need to fully respect and protect the fundamental rights of those subject to such powers resonates (perhaps unexpectedly) even within the bounds of national territory.

Conflict of Interests
When the manuscript was submitted, the author was not yet employed by the European Commission. The article presents the view of its author only and is published under her sole responsibility. The author declares no conflict of interests.