The mission of this blog is to provide a critical analysis of recent developments in the immigration and asylum law and policy of the European Union. The articles published on the blog are written by university academics and other experts from across the European Union who specialise in immigration and asylum law.
Since the past few decades, legal scholars and practitioners have been progressively acquainted with a series of abbreviations: SIS II, VIS, Eurodac, EES, ETIAS, ECRIS-TCN and the concept of interoperability has arisen prominently in the EU agenda. Behind the abbreviations lies an elaborated and complex legal framework of Europe-wide databases, whereby millions of personal data collected by different groups of third-country nationals are stored and further processed for a variety of purposes. For the sake of convenience and to inform the subsequent analysis, find here a table summarizing the key characteristics of each system.
This table is illustrative of the different logics and objectives underpinning each database, but also of an array of common characteristics.
First, their scope ratione personae is primarily different categories of third-country nationals, with EU citizens personal data only processed in an incremental manner (by the law enforcement branch of the SIS II; by the VIS, as regards sponsors or family members of visa applicants, or in the future ECRIS-TCN in relation to dual nationals);
Second, those databases process a wide range of personal data, including biometrics (photographs and fingerprints, except for the ETIAS), which constitute a special category of personal data under Article 9 of the General Data Protection Regulation;
Third, law enforcement authorities and Europol are allowed to have access to the records, either under specific conditions (VIS, Eurodac, EES, ETIAS) or because of the law enforcement (security) mandate (SIS II and ECRIS-TCN);
Fourth, personal data are retained for a significant period of time;
Fifth, databases are multi-purpose, dynamic and flexible in nature, as evidenced by the increasing objectives attached to each system. These purposes range from modernising immigration control to (disturbingly) law enforcement, thus blurring the boundaries between immigration and criminal law.
Crucially, databases are compartmentalised; even though in the future all third-country nationals will be effectively captured by at least one database, the data pots are separate from each other. This will soon change; the final step towards an EU ‘Big Brother’ is the interconnection of the different ‘data pots’ under the umbrella term of interoperability. Against this background, this blog post aims at critically evaluating this important legal development from a privacy and data protection standpoint.
2.Compartmentalisation Is Dead: Long Live Interoperability
The story behind the emergence of interoperability has been discussed in detail elsewhere. For the purposes of the present post, it suffices to recall that the interoperability debates first started in the aftermath of 9/11. Then, in its 2005 Communication, the Commission defined interoperability as the ‘ability of IT systems and of the business processes they support to exchange data and to enable the sharing of information and knowledge’. Details on the legal aspect of interoperability were spared, as the concept was reduced to a technical matter. Since the Paris attacks of November 13, 2015, the connection of the ‘data jars’ has gained fresh impetus. A High Level Expert Group on Information Systems and Interoperability was set up and gave the green light for implementing this initiative. In the meantime, interoperability was already foreseen in the proposals for the EES, recast Eurodac (still under negotiations) and ETIAS, thus pre-empting its insertion through the back door without an assessment on its necessity. Following speedy and rather limited negotiations, Regulations (EU) 2019/817 and 2019/818 were published in the Official Journal on the 22nd May 2019.
Interoperability is understood as information systems ‘speaking to each other’ and as an evolutionary tool that will enable further uses through the aggregation of data from different sources. Its four main components are:
The ESP (European Search Portal) will enable competent authorities to simultaneous query the underlying systems and the combined results will be displayed on one single screen. Even though the screen will indicate in which databases the information is held, access rights will remain unaltered and will proceed following the rules of each database.
The Biometric Matching Service (BMS) will generate and store templates from all biometric data recorded in the underlying systems, thus effectively becoming a new database that compiles biometrics from the SIS II, VIS, Eurodac, EES and ECRIS-TCN and that will replace separate searches in the other databases.
The Common identity repository (CIR) at the core of interoperability which will store an individual file for each person registered in the systems, containing both biometric and biographical data as well as a reference indicating the system from which the data were retrieved. CIR’s main objectives are to facilitate identity checks of third-country nationals, assist in the detection of individuals with multiple identities and streamline law enforcement access. With respect to law enforcement, a two-step process is foreseen whereby law enforcement authorities will first consult all databases to check whether records on an individual exist in any of the databases without obtaining prior authorisation by a verifying authority. In the event of a ‘hit,” the second step is to obtain access to each individual system that contains the matching data through the procedure prescribed in the legal basis of each database.
Finally, the Multiple Identity Detector (MID) will use the alphanumeric data stored in the CIR and the SIS II to detect multiple identities; it will create links between identical data to indicate whether the individual is lawfully registered in more than one systems or whether identity fraud is suspected.
3.From a ‘Panopticon’ to a ‘Pangnosticon’ in Violation of Privacy?
With interoperability as the ‘cherry on top’ of a multi-layered cake of databases, the landscape of information processing through centralised databases will be forever changed. Whereas interoperability will not frustrate existing limits on access rights of national authorities, the use of personal data will be attached to new purposes, which are not to be found in the respective legal instruments. For instance, Eurodac data will be used to detect persons with multiple identities even though Eurodac’s mandate does not specify such use. Furthermore, a particularly worrisome change involves Article 20 of the Regulations on interoperability, according to which a Member State police authority is authorised to query the CIR with the biometric data of a person taken during an identity check for the sole purpose of identifying that person. This function of the CIR was not supported by the existing legal framework and the addition of specific circumstances under which police authorities are authorised for identification checks in the adopted text does not compensate for the lack of clarity. The proposals stipulated that under Article 20 the identification of the person must contribute to preventing and combating irregular migration or contributing to a high level of security. As the EDPS has correctly pointed out these objectives are unduly vague and do not explain whether these police checks will take place under immigration control or law enforcement procedures. Identity checks by police authorities may fuel discriminatory practices which may proceed to identification checks to third-country nationals on the spot solely on the basis of extensive profiling, rendering their status on the territory particularly precarious.
In reality, interoperability negates the relevance of the purpose limitation principle by essentially enabling databases to be used for almost any purpose as long as this is not incompatible with the original purpose for which the data have been originally collected. The multiple reconfigurations of the systems over time denote that the threshold for such ‘incompatibility’ is unreachable and the limits of these systems are far from being reached. The fact that the CIR will include both personal data collected and further processed for traditional law enforcement purposes (ECRIS-TCN) and personal data of immigration nature that will be used for the identification of individuals in multiple fora is probably the deathblow for the purpose limitation principle and privacy as the key value behind data protection law. Interoperability should not be seen as a way of altering the nature and purposes of existing databases, in other words interoperability should not become the end in itself. Compartmentalisation, which was heralded as a way of ensuring the protection of privacy and personal data protection, has become a problematic feature, allowing for ‘blind spots’ that hinder the work of national authorities. This logic does not correspond to the traditional understanding of migration control, but rather exemplifies and validates a growing understanding of databases for third-country nationals as ‘security systems’ and interoperability essentially reconceptualises all centralised databases as quasi-intelligence tools.
Perhaps the elephant in the room as regards the operationalisation of interoperability involves the masked setting up of new databases —the BMS, the CIR and the MID- based on combining data from different sources (albeit the latter will not hold personal data). The fancy wording that is used (‘component’ and ‘repository’) should not distract from the dangerous reality of massive catalogues of third-country nationals at EU level who are either administratively or criminally linked to the EU over a significant period of time. The aggregation of data through databases signifies a new information-processing paradigm of mass and indiscriminate surveillance. By combining information from different systems, authorities are empowered to draw more precise conclusions on the private lives of individuals, and data subjects will be unable to foresee how their collected information will be used. As Bunyan has noted, it is not far-fetched to characterise interoperability as a decisive step towards a single EU information system at the service of an EU Big Brother.
One could argue that the BMS does not process personal data as it merely stored templates of biometrics. Whilst this is an area that merits further attention, it has also been convincingly argued that templates constitute personal data. Importantly, the case of the CIR in particular reminds Foucault’s ‘panopticon’, whereby domestic authorities shall be able to see all different groups of third-country nationals through a digital catalogue for foreign population with an administrative or criminal law link to the EU. Repetitive references in EU documents to ‘blind spots’ that need to be covered so that everyone could be seen fits well with the analogy. Moving beyond, it is hereby argued that interoperability will enable domestic authorities to know third-country nationals better, by assembling records from the different systems and combine the different personal data to create richer profiles regarding their movement and administrative or criminal procedures that they have undergone. The ‘pan-opticon’ (coming from the ancient Greek ‘πάν’ (all) + ‘οπτικόν’ (of sight)) is thus progressively replaced by the ‘pan-gnosticon’ (‘πάν’ (all) + ‘γνωστικόν’ (of knowledge), an emerging know-it-all surveillance system, whereby authorities would be able to achieve total awareness of the identities of the individuals, with the ultimate aim of preventing, deterring, controlling, or in more neutral words ‘managing’ people.
In a series of judgments, the EU Court of Justice has placed important limits to Member States’ surveillance powers. In Opinion 1/15, concerning the transfer of PNR data from the EU to Canada, the Grand Chamber found that such transfer would not amount to a system of unlawful generalised surveillance, given that the personal scope involved merely those travellers from the EU to Canada. Conversely, in Digital Rights Ireland and Tele2, the Grand Chamber was adamant in proscribing mass surveillance, since it involved ‘practically the entire EU population’. It is hereby argued that whilst each database on its own may not qualify as establishing generalised and indiscriminate surveillance of movement following Opinion 1/15 because it involves only a fraction of third-country nationals, the CIR as a new database combining materials from the underlying systems ticks all the boxes to be considered as unlawful mass surveillance of movement.
Another key change brought about by interoperability involves law enforcement access to third-country nationals’ data. Access is currently subject to tailored conditions of access and verification by a verifying authority. In Digital Rights Ireland and Tele2, the CJEU made clear that such access should be subject to strict conditions and prior verification that those conditions have been met by a verifying authority, either a judicial or independent administrative one. Interoperability will not only retain the problematic modalities of access to the respective systems (for example see my chapter in this volume), but will further progressively lead to routine access. As noted by the EDPS, a ‘hit’ is significant since it reveals elements of an individual’s personal life, for instance that they are visa free travellers or asylum seekers, and, therefore, this first step of checking whether there is personal data in any of the underlying systems should also take place after fulfilling the specific conditions of access prescribed in the legal basis of each database. Importantly, it is hard to believe that upon finding that a database holds information on a person, the verifying authority ensuring the conditions for access have been met will not allow such access. In other words, not only the independence and objectivity, but also the very existence of a verifying authority may be biased by the two-step approach. Arguably, this new function may enable national authorities to engage in ‘fishing expeditions’. Therefore, more prosecutions and/or convictions of third-country nationals may take place, merely because a pool of information exists, since no equivalent EU-wide catalogue of records on EU citizens exists. This may further sustain a divide between the EU citizens and the foreigner and raise serious non-discrimination concerns.
In addition, the operationalisation of interoperability raises further concerns as regards individual rights (rights of information as regards the processing of the personal data, right of access, rectification and deletion). In view of the processing of personal data in a multiplicity of context, the right to information may be all the more difficult to be exercised. Finally, as Evelien Brouwer eloquently points out in her contribution to the blog, it must be recalled that the personal data stored have been long suffering in terms of quality, including fingerprints. However, if the personal data stored are not of sufficient quality, any aggregation of this data through interoperability may have lead to incorrect processing, with significant repercussions for non-EU nationals, particularly in the case of the MID.
First, interoperability is much more than a buzz word and a panacea for security and migration concerns; it has become the ‘Trojan Horse’ towards the silent disappearance of the boundaries between law enforcement and immigration control and the radical intensification of surveillance of all mobile non-EU nationals. With that step completed, it will not come as a surprise if PNR data or even national identity cards or passports of EU nationals will also make their way into interoperable centralised databases, so as to ‘rectify’ the imbalance between the treatment of third-country nationals and EU citizens in terms of surveillance of movement through personal data processing. Secondly, interoperability is the latest nail in the coffin of third-country nationals’ privacy; databases have progressively proliferated and their functions expanded without having been litigated in terms of fundamental rights compliance before the European Courts. In an era where strategic litigation seems to be the way forward, is it possible for centralised databases to find their way into courts, or will we have to wait until data surveillance hits our own door?
This blog has already been published on 25 May 2019 on Verfassungsblog and we thank the editors of Verfassungsblog for their kind permission to re-publish this contribution.
On 14 May 2019, the Council adopted two regulations, Regulation 2019/817 and Regulation 2019/818, establishing a framework for the interoperability between EU information systems in the Area of Freedom, Security, and Justice. The new rules on interoperability, upon which the European Parliament agreed in April 2019, will allegedly provide for easier information sharing and ‘considerably improve security in the EU, allow for more efficient checks at external borders, improve detection of multiple identities and help prevent and combat illegal migration’. All this, according to the press release of the Council, ‘while safeguarding fundamental rights’.
It is questionable whether this commitment made by the EU legislator is justified. Interoperability only ‘works’ as long as the reliability and trustworthiness of data in the databases involved are sufficiently guaranteed. Considering the number of states and the large-scale databases involved, the consequences of decision-making based incorrect or even unlawful data will be detrimental not only for the protection of fundamental rights but also for the effectiveness of interoperability as a data and border surveillance tool as such.
‘Interoperability’ and ‘interstate trust’ is a perilous combination if this would allow national authorities to rely on data stored in EU data systems, instead of making a careful examination of each individual case. Despite formal safeguards in the regulations and the applicable data protection standards, it will be hard for data subjects to oppose decision-making based on incorrect data, when the source or the author of that information is unknown or when states can ‘hide’ behind the back of interstate trust without providing access to effective legal remedies. The problems which may arise for individuals through the ‘blind’ use of large-scale databases will be illustrated with two recent case-studies:
the first concerns the entry of a human rights activist into SIS II, and the second one the reliance on Eurodac/DubliNet to determine the age of a minor asylum seeker. Considering the extensive scale of data processing, the fact that interoperability affects mainly third-country nationals, and because of the complexity of rules, the adopted instruments fail to meet the standards defined by the CJEU and the ECtHR on the basis of the rights to privacy and data protection.
Interoperability and multipurpose use of EU large-scale databases
Both Regulation 2019/817 which applies to information systems in the field of borders and visa and Regulation 2019/818 on systems in the field of police and judicial cooperation, asylum and migration, build on the same interoperability components, which include:
A European search portal, allowing competent authorities to search multiple information systems simultaneously, using both biographical and biometric data.
A shared biometric matching service, enabling the searching and comparing of biometric data (fingerprints and facial images) from several systems.
A common identity repository, which contains biographical and biometric data of third-country nationals available in several EU information systems.
A multiple identity detector, to check whether the biographical identity data contained in the search exists in other systems covered, to enable the detection of multiple identities linked to the same set of biometric data.
The interoperability structure is thus based on a network or a mechanism through which different authorities can check whether information on a particular individual is available in one of the current (and future) EU databases. This search will be facilitated by the use and storage of biometrics in the different databases. The databases which can be connected via interoperability are numerous. They include the existing large-scale databases Schengen Information System or SIS II, Eurodac, and Visa Information System (VIS), and the Europol and Interpol databases. Furthermore, the interoperability scheme applies to the following systems when operational: the Entry Exit System (EES), the European Travel and Authorisation System (ETIAS), the European Criminal Record Information System on third-country nationals (ECRIS-TCN).
With the aforementioned common identity repository including biometric and ‘biographical’ data, the interoperability rules only involve a centralized database with regard to third-country nationals. According to the European Commission in the explanatory memorandum to the proposal, this differentiation between EU citizens and third-country nationals is justified by the aim of preserving security in the EU: ‘Whilst not directly affecting EU nationals […], the proposals are expected to generate increased public trust by ensuring that their design and use increases the security of EU citizens.’
Multipurpose access to centralised databases
According to the press release, the interoperability systems will not ‘modify the rights of access as set out in the legal basis relevant for each European information system, but will ease and improve information sharing’. Here the legislator has a point as it was not even necessary to extend access to the different databases involved for further purposes: this was already provided for in the various laws adopted in the last few years. In 2015, Eurodac, an administrative system with data on asylum seekers in the EU set up for the implementation of the Dublin system, has been extended for law enforcement purposes on the basis of Regulation 603/2013. VIS, including data on all applicants for short-term visas, has been developed as a multipurpose tool from the start, and also the new border systems ETIAS and EES will be accessible for law enforcement and security purposes. And the inclusion of biometric data in SIS III on the basis of the Regulation 2018/1861 for the purpose of border control and the Regulation 2018/1862 in the field of police and judicial cooperation already changed the structure of SIS II into a more general investigation tool.
Vice versa, the ECRIS-TCN, a database which was presented as necessary for the purpose of judicial cooperation, will be accessible for immigration control purposes. ECRIS-TCN provides for a centralised system with data on all third-country nationals with criminal records in the EU. In accordance with Article 7 of the ECRIS-TCN Regulation 2019/816 the information stored on previous criminal convictions on third-country nationals can be requested for criminal proceedings ‘or for any other of the following purposes, if provided under and in accordance with national law.’ Despite early and serious concerns of the Fundamental Rights Agency (FRA) against the use ECRIS-TCN for national immigration law purposes, the adopted Regulation now explicitly defines as one of the ‘following purposes’: ‘visa, acquisition of citizenship and migration procedures, including asylum procedures’.
This implies that ECRIS-TCN data can be used not only for rejecting short-term visa, but also for the refusal or withdrawal of residence permits of third-country nationals. The FRA warned against these secondary effects from national convictions based on previous irregular entry or stay, specifically for refugees and children and referred to the differentiated practices in the Member States with regard to criminalization of irregular stay and entry. According to the Commission in the explanatory memorandum to the 2016 proposal, the extent to which criminal record information is processed for other purposes would be ‘a matter of national law’. Therefore, limitations to this further use would not be possible in the ECRIS-TCN proposal. The adopted rule in the interoperability Regulation shows that the legislator is not only aware, but also approves that Member States will use information on criminal records in the ECRIS-TCN for immigration law purposes, even if these decisions are based on the (possibly very different) criminal law systems of other states.
The case of a human rights activist in SIS II
On 13 August 2018, Lyudmyla Kozlovska, Ukrainian national, President of the Open Dialog Foundation (ODF) and married to a Polish citizen, was detained by the Belgian authorities following a passport control at the Brussels airport on the basis of a Polish alert in the SIS II for the purpose of refusal of entry or stay. The next day, the Belgian border authorities deported her to Kiev, Ukraine. Prior to the issuing of the SIS alert, Kozlovska was legally resident in Poland and had applied for an EU long term residence permit. During that procedure, she was informed by the Polish authorities that she was not listed in the SIS II. In the period following her expulsion from Belgium, Mrs Kozlovska was allowed to visit several Schengen states (including Belgium and Germany) for several days to talk with members of the European Parliament and national parliaments, despite the fact that the Polish alert was not withdrawn. In March 2019, the Belgian authorities provided her with a five year residence permit, after which decision Poland is obliged to withdraw the SIS alert in accordance with the Schengen rules.
In E., addressing the obligation of consultation between Member States in these matters related to SIS, the CJEU confirmed that an individual has the right to rely before courts on the obligation to withdraw an alert from SIS if that is the outcome of this consultation. Access to legal remedies against the Polish alert is however difficult, also because she has never been informed about the precise reasons for the SIS alert. Furthermore, until the SIS alert is deleted, Mrs Kozlovska risks to be denied entrance or expelled by other Schengen states. Her case illustrates what happens if national authorities ‘blindly’ rely on data in EU databases and the effect of these decisions for human rights, including the right of freedom of expression, freedom of movement and effective judicial protection. Even if in the end she obtained a residence permit by one state, it will remain difficult to get rid of this ‘digital entry ban’. It is to be feared that interoperability will only enhance this informational bureaucracy, affecting the effective protection of fundamental rights.
The case of a minor unaccompanied asylum seeker in Eurodac
In May 2019, the Dutch radio-programme (VPRO) broadcasted a documentary on a case of a minor asylum seeker who, when entering EU territory via Italy, was registered into Eurodac as being an adult by the Italian authorities. Traveling further north, she applied for asylum in the Netherlands where she informed the immigration authorities (IND) she was 15. The IND however refused to treat her as a minor relying on the data submitted by the Italian authorities via DubliNet after consultation of Eurodac on the basis of the principle of interstate trust. Because of this refusal the minor did not receive further protection even though during the procedure she substantiated her age with documents and, as her lawyer submits in this documentary, from sight and behaviour, she clearly is a minor.
This Dutch policy of relying on data as registered by another Member State on the basis of the principle of interstate trust without any further investigation in to the age of the minor, has been approved by the Dutch highest administrative court (ABRvS). In aforementioned case, the minor herself had lied about her age to the Italian authorities in fear of being separated from the group with which she entered Italy. But the documentary officials of IOM and UNHCR confirm that incorrect registration of age of minors is common practice. A wrong birth date can be registered in periods of chaos when large numbers of asylum seekers arrive, but sometimes also on the basis of miscommunication or lack of information provided to minor asylum seekers.
As the report Digital Identity in the Migration and Refugee Context shows, during the identification process of migrants and refugees, the protection of privacy, informed consent and data protection are often compromised, not only for minors. The specific problems of children in registration procedures for immigration purposes were also established in the report Under watchful eyes of the FRA. Dealing with the collection of data during visa applications or for the purpose of the Dublin system, the FRA found that rights of children were affected in different ways, dealing with child-unfriendly treatment, doubts with regard to the quality and reliability of fingerprints, and risk of re-traumatization. Where interoperability might be a tool to trace missing children, the FRA in a report of 2017 underlined rightfully that this is only the case if Member States will make more use of the existing possibility to report missing children into SIS II and improve cooperation between their national authorities.
The first reason why interoperability of databases combined with interstate trust may have serious human rights and accuracy implications is the wide extent of the use of the stored into information. Large-scale databases are large-scale databases: they contain millions of data sets on individuals and they are used by a high number of national authorities, with different tasks and powers. This means that incorrect or outdated information in one of the aforementioned databases has a high risk of being multiplied in other EU databases and at the national level.
For example, Eurodac, used by 32 EU and non-EU states, at the end of 2017 included more than 5 million ‘fingerprint datasets’. More than one million transactions to Eurodac took place in 2017, with a peak in February 2017 of 6.000 transactions a day. At this moment data on age of persons registered in Eurodac are exchanged via DubliNet, but the Commission proposal (COM (2016) 272 final) for a recast-Eurodac Regulation provides for the registration of not only age, but also name, place of birth, citizenship and facial image. For the purpose of law enforcement, 550 Eurodac searches were performed by the ‘designated authorities’ of Member States and 114 by Europol.
At the end of 2018, SIS II contained more than 82 million alerts of which almost a million alerts were on persons. SIS II in 2018 was accessed more than 6 billion times by Member States and Associated Countries, representing an increase of 20% compared to 2017. In 2018, 267.239 foreign hits were reported of which 77% were triggered by alerts on persons. And finally VIS, accessible by 26 states (EU and non-EU Member States), contained at the end of 2017 about 49 million visa applications and 42 million fingerprints sets.
Second, the effects of interoperability will in the first place affect third-country nationals. As pointed out by the EDPS in Opinion 4/2018, the interoperability regulation in itself create a new centralised database containing information about millions of third-country nationals, including their biometric data. The consequences of any data breach could seriously harm a potentially very large number of individuals and, according to the EDPS, if ‘such information ever falls into the wrong hands, the database could become a dangerous tool against fundamental rights’. The risk of misuse of data is a risk which addresses the heart or ‘essence’ of the right to private life and data protection, protected in Article 8 ECHR and 7, respectively 8 of the Charter on Fundamental Rights.
Both the CJEU and the ECtHR have repeatedly warned in their case-law against the unrestricted or disproportional processing and use of personal information. In S. and Marper v. UK, the EctHR warned against the risk of stigmatisation where information on large groups of unsuspected citizens is stored into centralised databases for law enforcement purposes. One of the criteria in Digital Rights Ireland for the CJEU to find the Data Retention Directive in violation of the right to privacy and data protection, concerned the fact that its implementation would entail the data processing of ‘practically the entire European population’, also involving persons without any link to criminal prosecution. And in Schwarz v. Bochum, the CJEU found that the Passport Regulation 2252/2004 did not amount to violation of the rights protected in 7 and 8 of the Charter, because it only provided for the recording of two fingerprints and facial image on the passport, it did not prescribe a central storage of the data of passport holders, and the purpose of the use of the data was limited to identification of the owner and verification of the authenticity of the document and would not be used for other purposes.
These criteria are relevant when assessing the use of large-scale databases and the effects of interoperability, especially where these measures primarily address third-country nationals. Discretionary national powers and the availability of mobile devices will make it possible to check individuals not only at the borders, but also within the national territory. The risk of discriminatory checks within the EU borders is enhanced, as pointed out by Vavoula in her eumigrationlawblog, by Article 20 of the Regulation 2019/818 according to which police authorities can check the aforementioned CIR or Common Identity Repository solely for the purpose of identification of a person, for example if ‘a person is unable or refuses to cooperate’. This entails the risk of extensive use of databases, in violation of the strictly necessity test as defined by the CJEU in Digital Rights Ireland and Schrems and emphasised in the EDPS toolkit on assessing the necessity of data processing. The fact that interoperability and the centralization of data bases mainly affect third-county nationals (or EU citizens with a third-country nationality as in ECRIS-TCN) fails to respect the fundamental right to..
By Kris Pollet, Head Legal and Policy Research, European Council on Refugees and Exiles (ECRE)
This blog is written in the personal capacity of the author and does not necessarily represent the views of ECRE.
The reform of the Common European Asylum System (CEAS) through a third phase of harmonisation of the EU asylum acquis, launched by the Commission in 2016, proved too ambitious to be concluded under the 2014-2019 parliamentary term. While being blocked because of disagreements between Council and Parliament especially about solidarity between Member States, the legislative work done risks to be annihilated by the European elections, knowing that appetite of Member States for legislative harmonisation may diminish in the future.
Contrary to popular perception, from an institutional perspective considerable progress has already been made on parts of the asylum package. As discussed in more detail below, with five of the seven proposals having reached the stage of so-called trilogue negotiations and short of an inter-institutional agreement within less than three years, co-legislators have certainly moved at reasonable pace. Keeping in mind the fact that both the first and second phase of legislative harmonisation in the asylum field took five years and various amended proposals to conclude, the third phase of harmonisation is certainly still on schedule.
However, discussions on the Commission’s Dublin IV proposal in particular and the Asylum Procedures Regulation have triggered a range of fundamental legal and political questions among the Member States preventing the Council from adopting a negotiating mandate so far. A final push by the Commission and some Member States to have at least some of the least controversial proposals adopted before the European elections appeared to be in vain as there was eventually too little appetite to split the package. The uncertain legal and political ramifications of a partial deal finally prevented both the European Parliament and the Council from moving ahead. Faithful to the package-doctrine, the negotiations on the asylum package were finally de facto referred to the next parliamentary term.
Failure to complete negotiations on legislative proposals within one Parliamentary term triggers a number of institutional and political issues, in particular from the perspective of the European Parliament. Although the European Parliament has always been one step ahead of the Council in the legislative process since the Commission proposals were submitted, the installment of a new European Parliament may have important consequences for the negotiating mandates that have been adopted on all asylum proposals under the resigning Parliament.
This contribution discusses the possible institutional implications of the establishment of a new parliament after the elections for the negotiating mandates adopted by its Libe Committee on Civil Liberties under this parliamentary term. Whether and to what extent the new Parliament is bound by the commitments made by the previous assembly has obviously a huge impact on the continuation, if any, of the legislative process after the elections. In a second part, the state of play of negotiations on the asylum package will be briefly discussed with a particular focus on three key features of the Commission proposals relating to the Qualification Regulation and the recast Reception Conditions, illustrating the impact of positions taken by the LIBE Committee on the provisional compromise agreements that had been reached with the Bulgarian Presidency in June 2018.
1. New Parliament, new beginning?
Negotiations on the seven files constituting the asylum package are at different stages of the ordinary legislative procedure in accordance with Article 294 of the Treaty on the Functioning of the European Union. As none of the proposals have been adopted before the elections, negotiations with the Council will have to be taken forward by the next European Parliament, unless the current or future Commission decides to withdraw them. Assuming the objective of both last institutions is to achieve a first reading agreement, this triggers specific challenges from an institutional and political perspective for the next European Parliament.
From an institutional perspective, the failure of the co-legislators to conclude negotiations on (parts of) the asylum package before the elections could have important consequences for the positions which have been negotiated within the European Parliament under the current parliamentary term. In theory, all the work invested by the previous European Parliament in the asylum package could be lost. This is because of the so-called “unfinished business” provision laid down in Rule 229 of the Rules of Procedure of the European Parliament’s 8th parliamentary term. This provision states the principle that “at the end of the last part-session before the elections, all Parliament’s unfinished business shall be deemed to have lapsed”.
This would mean that all negotiating mandates adopted by the LIBE Committee on the asylum package, which all have fundamentally amended the initial Commission proposals and have included important human rights safeguards without exception, would no longer bind the new European Parliament. As a result, the process within the European Parliament would have to start all over again, with the appointment of rapporteurs and shadow rapporteurs, a first draft by the rapporteur, amendments by LIBE Committee members, a vote on the draft report and negotiating mandate in the LIBE Committee and endorsement of the negotiating mandate by the plenary assembly.
However, the abovementioned principle is subject to the provisions of the second paragraph of Rule 229 which states the following: “[A]t the beginning of each parliamentary term, the Conference of Presidents shall take a decision on reasoned requests from parliamentary committees and other institutions to resume or continue the consideration of unfinished business”. The Conference of presidents of the new Parliament could therefore decide to continue or resume the consideration of such unfinished business, upon request of a parliamentary committee or other institutions. The latter may include any other EU institution, including the Commission or the Council, should no such request be made by the parliamentary committee concerned. This is to be considered an institutional measure of last resort to address inaction by the parliamentary committee concerned and where it is considered politically desirable to continue the discussions in order to avoid additional delay from having to start the procedure from scratch again in the responsible committee.
Given the sensitive nature of the asylum package, the amount of work invested in the asylum proposals by the LIBE Committee and the advanced stage of negotiations with the Council on five of the seven asylum proposals, it is beyond doubt that such a request would be considered as “reasonable”. For the same reasons, it seems inconceivable for the next LIBE Committee not to submit such a request to the Conference of Presidents. It seems unlikely that the Conference of Presidents would refuse such request, but the fact remains that the decision to continue or resume unfinished business under the previous Parliament is at the discretion of the Conference of Presidents. The European Parliament’s rules of procedure do not provide for a specific definition of “unfinished business”. However, it must be assumed that it includes any type of activity within the European Parliament’s mandate which requires a decision, including the negotiation of legally binding acts. This follows a contrario from paragraph 3 of Rule 229, according to which the provisions on unfinished business “shall not apply to petitions and communications that do not require a decision”.
It should be noted that the basic rule according to which any unfinished business lapses, does not apply to legislative files on which the Parliament has voted in plenary. All votes taken in plenary by Parliament before the elections remain legally valid for the next Parliament. This means that, after the elections, the new Parliament will pick up the files where the previous left them and will continue with the next stage of the decision-making procedure. However, none of the LIBE reports on the asylum proposals have been subject to a vote in plenary before the last part-session of the European Parliament sealing the Parliament’s first reading position, except for the endorsement of the respective negotiating mandates. Therefore, the EP reports voted in LIBE could not benefit from this institutional practice and be saved from oblivion without a decision of the new Parliament’s Conference of Presidents.
Certain forecasts of the outcome of the European elections predict important changes in the composition of the European Parliament and the balance of powers between the traditional political families. While the European People’s Party and the Socialists and Democrats (S&D) may remain the two biggest political families, they may face a considerable loss, while the success of eurosceptic parties in the slipstream of the Legal Nord in Italy remains the big unknown. The effect of such changed composition on the European Parliament’s position in negotiations with the Council on the asylum package after the elections is hard to predict and will also depend on whether the current allocation of asylum files to the various political groups will be maintained. Traditionally, reports assigned in the preceding parliamentary term are not claimed by another political group after the elections. However, this is not set in stone and could be questioned at any time. As some of the appointed rapporteurs on the asylum files are not running for re-election and others may not be re-elected, new rapporteurs will have to be appointed. Whether this opportunity will be taken by certain groups to claim the rapporteurship from other groups remains to be seen, but even if all files remain within the same political group, a new rapporteur will inevitably bring a new dynamic and may set different priorities in the negotiations with the Council, or in case new reports need to be adopted, within the European Parliament.
2. The state of play of negotiations on the asylum reform
A partial inter-institutional agreement on the proposed EU Asylum Agency was even reached between the Council and the European Parliament under the Maltese Presidency in June 2017. However, this was never formally adopted due to the abovementioned package approach, while an amended Commission proposal submitted in September 2018 reopened discussions on the one asylum file co-legislators were in principle in agreement on. The Commission’s move provoked a fierce reaction from the LIBE Committee, which rejected all proposed amendments and in doing so made clear that the text as agreed with the Council in June 2017 was to be considered as the final deal ready for formal adoption. Largely inspired by activities carried out by EASO in accordance with the operating plan 2018 and 2019 for Greece, the amended proposal includes a further expansion of the EU Asylum Agency’s role in national decision-making by entrusting it with the “preparation of decisions on applications for international protection” to be consequently provided to the national asylum authorities. It also entrusts the Agency with a role in assisting Member States in “the follow-up of asylum procedures with return procedures in case of a negative decision”, thereby entering European Border and Coast Guard’s territory. Particularly controversial is the proposed role of the asylum support teams in assisting the Courts and tribunals with handling appeals by performing legal research, analysis and other legal support. In particular, as the role of the Agency in first instance decision-making increases, any interference with the appeal process is highly problematic as it undermines the independent and impartial nature of appeals and inevitably creates a conflict of interest. In the name of harmonisation and quick procedures, key features of the right to an effective remedy are apparently easily overlooked.
Furthermore, trilogue discussions have resulted in provisional compromise texts between the Bulgarian Presidency and the European Parliament on the Commission proposals for a Qualification Regulation and a Union Resettlement Framework, as well as the recast Reception Conditions Directive. However, these provisional agreements were eventually not endorsed by the Council, which further amended the provisional agreements and updated its negotiating mandates inter alia with respect to the detention of children, exclusion clauses and revocation of international protection status. Unacceptable to the European Parliament, trilogue discussions have been on hold since. In all three files, so far trilogue negotiations have altered key aspects of the initial Commission proposals. Some of these changes have been triggered by amendments proposed by the European Parliament’s negotiating mandate, others have been proposed by the Council as well.
The latter was the case, for instance, for the Commission’s proposal for a mandatory review of the continued need for international protection of status holders at the occasion of the renewal of their residence permit. Both institutions finally objected to the inclusion of such an obligation in the Regulation but for different reasons. For a majority of Member States, the obvious additional administrative burden on asylum authorities outweighed the potential deterring effect of a more temporary protection status. Adverse effects on integration of beneficiaries of international protection stemming from increased uncertainty over the length of their stay on the territory were used as justification for the rejection of the proposed provision by various LIBE Committee members. Although the rejection of the mandatory review provision is an important improvement of the Commission proposal, it would not prohibit Member States wanting to systematically re-assess whether beneficiaries continue to require international protection from doing so as a matter of national law.
Mandatory application of the internal protection alternative(IPA), optional for Member States under the current recast Qualification Directive is another key feature of the Commission proposal for a Qualification Regulation. Here, the Commission proposal is further nuanced in the provisional compromise text. Contrary to the European Parliament’s negotiating mandate, the latter maintains the mandatory assessment of an internal protection alternative, where persecution emanates from the State but provides for a presumption against the existence of an internal protection alternative where persecution emanates from the State, except where this is limited to a specific geographical area. The compromise text further refers to the obligation to “examine” whether an IPA exists, rather than an obligation to “determine” that an IPA exists when certain conditions are met. The latter leaves ambiguity as to whether a State is always required to examine and apply the internal protection alternative or whether, following examination, it may choose not to use it. Being a Regulation, its provisions would be directly applicable provided they are sufficiently precise as per the CJEU doctrine. However, if adopted, the provision on internal protection alternative would no doubt be a strong candidate for interpretation by the Court of Justice through a preliminary reference procedure.
Also, the punitive approach to secondary movements of asylum seekers envisaged by the Commission has been further diluted in trilogue negotiations on the recast Reception Conditions Directive. According to the provisional agreement, exclusion from general material reception conditions as a sanction can only be imposed as of the moment of notification of a transfer decision, and not for the entire duration of the Dublin procedure. While Parliament had proposed the deletion of this provision, it was the Council’s negotiating position which was finally adopted in the provisional compromise. At the same time, whereas the European Parliament had proposed an absolute ban on detention of children, the compromise in the provisional agreement further strengthened the exceptional nature of detention of children by allowing detention only as a measure of last resort and after detention is assessed to be in their best interest.
Whereas co-legislators have downplayed some of the controversial aspects of the Commission proposals in trilogue negotiations so far, others have been maintained such as additional grounds for detention and freedom of movement restrictions, or the prohibition of any proportionality assessment linked to the “fear of persecution’ in the application of exclusion clauses, based on an erroneous reading of the ruling of the CJEU in the case of B&D. Moreover, trilogue discussions on the two other core instruments of the asylum package, the Dublin IV proposal and the proposal for an Asylum Procedures Regulation, including most contentious provisions relating to safe country concepts, border procedures and solidarity, have not even started yet.
The composition of the new European Parliament may considerably change the dynamics in the negotiations with the Council and trigger additional concessions by the new European Parliament departing from core protection-oriented positions taken by its predecessor. Seen from that perspective, even in case the above mentioned request for continuation of the legislative process on the basis of the positions taken by the previous Parliament is honoured by the Conference of Presidents under rule 229, this does not necessarily imply that such positions would be defended with the same rigour, nor would it..
The EU legislature has agreed about a substantial reform of the Visa Code which still needs to be adopted officially. It will tie the cost, processing time of visas and availability of multiple entry visas to the success rate of Member States’ return efforts to the relevant country. The idea, proposed by the Commission and accepted both by the Council and the Parliament, is that nationals of countries on the EU’s visa black list, should be punished for the inability of EU Member States to return people (both nationals of the state and where permitted non-nationals who travelled through) to their state.
This collective punishment would take the form of a doubling of the cost of a visa (from the proposed € 80 to € 120 or even € 160) or exclusion from:
(a) simplified visa application procedures,
(b) waivers for holders of diplomatic and service passports,
(c) the 15 day visa processing time and
(d) access to multiple entry visas. Further, the penalties would not necessarily apply to all nationals of the ‘offending’ country but possibly only to some categories (unspecified) of them.
This principle could constitute discrimination on the basis of nationality within the class of States which are on the EU’s visa black list and discrimination on the basis of economic, social and/or educational status in the delivery and cost of EU visas for nationals of the same state. While the international community is becoming increasingly intolerant of discrimination on the basis of nationality in immigration procedures (see for instance No 2018/36 Qatar v UAE ICJ Provisional Measures decision of 23 July 2018 where the ICJ gave provisional relief to Qatar regarding the threatened collective expulsion of its nationals from UAE in reliance on the Convention on the Elimination of All Forms of Racial Discrimination), the EU appears to be embracing it in a particularly arbitrary form. The vast majority of the people the proposal would punish have no control over or influence on the rates of expulsion of their fellow countrymen and women from EU states, yet would be the objects of this discrimination.
The original proposal of the Commission issued in 14 March 2018 (COM(2018)252) was accompanied by an impact assessment explaining the merits of linking visa issuance with return (SWD(2018)77). Highlighting the ‘unprecedented’ levels of refugees and irregular migration into the EU, it tied the policy context to threats to internal security and terror attacks, a flimsy linkage if there even was one (see: SWD(2018)77, p. 8). Relying on the flagging of visa policy as a possible element to ‘improve’ return policies set out in the 2015 Action Plan on Return (COM(2015)453) which was then picked up loosely the following year in the Partnership Framework Communication (COM(2016)385), the proposal introduces a fully blown linking option where visa policy and return rates are tied together. This rather absurd proposal has now made it to the last stages of the legislature apparently with no serious consideration of the legality let alone the effectiveness of such a proposal.
The proposal’s Staff Working Document refers to the European Travel Information and Authorisation System (ETIAS) the EU’s equivalent of the US ESTA system of pre-travel check on travelers which was adopted on 12 September 2018. The ETIAS system of prior authorisation for visa exempt travelers is completely electronic but the objective is broadly similar as that for visas – to screen foreign nationals seeking to enter the EU before they arrive. It would have been most helpful if the proposal could have clarified why there is a need for two systems – visas and ETIAS authorizations. It might well be asked why a single system applicable to all third-country nationals would not be preferable to avoid differential costs and discrimination on the basis of (third-country) nationality in the authorization system. In the Staff Working Document much attention is devoted to the cost of the visa system and the justification of the general increase of visa fees from € 60 to € 80 for visa nationals on the basis of the cost of administering the system (minors under six go free but six to twelve year olds pay a € 40 visa fee). In case of sanction against one country the visa fee can raised up to €160.
But the visa fee is not the only unavoidable cost in obtaining a Schengen visa. Within the system there is an increasing important place for external service providers (private companies which carry out an increasingly important part of the visa process). Under the proposal Member States can make it mandatory to cooperate with these external service providers. The proposal increases their fees allowing then them to charge for their services up to € 80 or exceptionally € 120 on top of the visa fee. So at its most expensive a Schengen visa can cost an individual € 280 under this proposal. An ETIAS authorization will cost € 7 and be free for minors under 18 years old (see here). The discrimination on the basis of nationality among third country-nationals in fees payable for the pleasure of visiting the EU is obvious and growing. It also falls disproportionately on nationals of countries with lower GDP and average incomes. A quick look at the EU’s visa black list shows that it includes very few high GDP countries so their nationals will only pay the € 7 for an ETIAS authorization. Nationals of lower GDP countries which are mainly on the EU visa black list will have to pay between € 80 and € 160 for the visa plus the external service provider fees an additional amount between € 80 and € 120 (to a possible total between € 160 and 280 € per person).
The Visa Code, through the provisions on external service providers and their fees, has created the conditions for a cottage industry to develop around EU consulates consisting of companies which provide these services. It has also encouraged many consulates to rely ever more heavily on their services in processing visa applications. Allowing these companies to charge even more for their services will only entrench the economic interests of these companies in ensuring the continuation of the visa system.
The proposal sets out the criteria on the basis of which the Commission is to assess cooperation of third countries with regard to readmission for the purpose of punishing their nationals through visa procedures. These are:
The number of return decisions issued to persons irregularly staying on the territory of a Member State from the third country;
The number of actual forced returns of such persons (including other persons on the basis of bilateral agreements);
The number of readmission requests per Member State accepted by a third country as a percentage of the number of such applications submitted;
The level of practical cooperation in the area of return cooperation in the different stages of the return procedure such as:
Assistance provided in identification;
Acceptance of EU travel documents or laissez-passer;
Acceptance of readmission of persons who should be legally returned to their country;
Acceptance of return flights and operations.
These are the criteria on the basis of which the most egregious discrimination in the proposal (the possibility to increase the visa fee and service provider fee and otherwise punish nationals of countries to which EU Member States are having difficulties returning people) in the visa process is made possible. The justification contained in the Staff Working Document is far from satisfactory. It states that “in the case of 80 third countries the return rate (share of persons return to their country out of those ordered to leave the EU territory) fall below the EU average of 46.4%.” (para 2.2.3). Having regard to the Frontex Annual Risk Analysis 2019, the top three nationalities of persons against whom Member States took return decisions in 2018 were Ukraine, Morocco and Afghanistan. Iraq and Syria are in the top ten. Syria, Iraq and Afghanistan are among the top countries of origin of asylum seekers in the EU and have high recognition rates. The issue of return decisions against nationals of countries which are ravaged by civil war or otherwise highly unstable is a dubious practice. That these decisions are difficult if not impossible to execute comes as no surprise. It should not be used as an excuse to enmesh discrimination in the visa process.
Further, the Commission relies on the number of return decisions issued by Member States before the number of forced returns. In almost all Member States return decisions are challenged which may result in positive decisions in favour of the individual. In 2018, around 50% of individuals issued with a return decision choose to leave. Thus the relevant figure is that of forced returns – which according to Frontex numbered a total of 75,241 in 2018. This is hardly a figure on the basis of which to adopt punitive legislation. In any event, the top five nationalities of people subjected to forced return from the EU, according to Frontex, were Albanian, Moroccan, Algerian, Tunisian and Ukrainian (in that order). Two of these nationalities – Albanian and Ukrainian – are not subject to visa obligation so the policy would be irrelevant to them. The next five top nationalities of expulsion were Serbian, Georgian, Brazilian, Kosovar, and Russian. Of these, only Russians are visa nationals so once again the policy is irrelevant.
The argument can be made that other countries might be among the top ten if their states provided more cooperation to EU Member States in return. But this is not explained or justified. Indeed, the European Parliament’s own impact assessment study on the proposals for amendment to the Return Directive shows that alternatives to coercion are likely to be more efficient both in the short and long term. Good law making requires a direct relationship between the measure adopted and the objective to be achieved. Here there is only the weakest of links and no evidence that the measures proposed would have any effect in achieving the objective. This is a prime example of incoherent policy making.
A lack of practical cooperation of state authorities as a criteria for punishment of their citizens by the EU is hard to justify. There is little if anything an individual citizen can do to encourage his or her state to cooperate with the EU in return practices. Such an approach to international relations is likely to exacerbate already tense relations between the EU and many third countries around the issue of migration. The contention of the Commission in its Staff Working Document that “visa policy can play an important role to obtain better leverage vis-à-vis third countries on readmission of irregular migrants” (para 5.2.3) is unsupported by any evidence at all.
Targeting the citizens of third countries through collective punishment in the visa process for their state’s degree of cooperation with EU Member States in return is likely to be highly counter-productive in terms of the EU’s international relations with third countries. Such a policy, based on the principle of discrimination on the basis of nationality among third country nationals, is of dubious legality in international law. A policy founded on the collective punishment of people for actions beyond their control raises profound questions about the quality of the shared values of the EU.
The decision of the UN Committee on the Rights of the Child in the case of D.D. v. Spain sets a precedent that strongly condemns the policy and practice of push-backs, called “rechazos” by the Spanish authorities. As the first case on summary expulsions in front of the Committee, D.D. v. Spain plays a crucial role for future jurisprudence on individual communications. The Committee’s decision clearly affirms the rights of unaccompanied minors at Europe’s borders and beyond.
Push-backs are a longstanding policy and practice at the Spanish-Moroccan border with fences around the enclaves of Ceuta and Melilla, the only land border between the European and the African continent. Spanish authorities immediately apprehend those who manage to cross the fences and return them to Moroccan forces, including the return of unaccompanied minors without access to any procedure to safeguard their fundamental rights. In March 2015, the Spanish Parliament created a legal basis for indiscriminate automatic deportations from Ceuta and Melilla. The Organic Law 4/2015 on the “Protection of Public Safety” introduced a special legal regime for Spain’s terrestrial borders with Morocco, stipulating that “foreigners detected on the territorial border line of Ceuta and Melilla while trying to cross irregularly the border may be rejected in order to prevent their illegal entry into Spain”. These amendments do not contain procedures to protect unaccompanied minors from immediate expulsions to Morocco.
The Spanish-Moroccan border regime has served as a laboratory for push-backs, with other Member States such as Hungary, Croatia, Greece, Bulgaria and Poland replicating this practice of indiscriminate automatic returns. More recently, border controls and push-backs have also surged at Europe’s internal borders, such as between France and Italy. Some States root these practices in national law, some issue semi-formalised paperwork in fast-track procedures and some take informal measures without any record. The underlying common denominator is the denial of access to any effective possibility to explain personal circumstances and to present arguments against the measures taken. At Europe’s multiple borders, unaccompanied minors encounter institutionalized violence instead of protection.
Case D.D. v Spain in front of the UN Committee on the Rights of the Child
D.D. was deported from Melilla to Morocco as an unaccompanied minor in December 2014. He had climbed the Spanish border fence structure and remained on top for hours without receiving any kind of assistance from the Spanish authorities. When he eventually climbed down and stepped on Spanish ground, he was immediately apprehended by the Spanish Guardia Civil, handcuffed and returned to Morocco through doors in the fence structure. He was returned without any possibility to explain his personal circumstances and protection needs, without being asked a single question about his age or name, and without being able to come into contact with any lawyer, translator or social worker. Instead, he was returned to the very place where he had previously suffered repeated ill-treatment at the hands of Moroccan forces.
The complaint brought by D.D. constitutes the first individual submission against push-backs to the UN Committee on the Rights of the Child under the Optional Protocol on a communication procedure. D.D. submitted that his summary return to Morocco without any type of identification procedure or evaluation of his situation was in violation of three core principles of the Convention on the Rights of the Child (CRC): a) Contrary to Article 20, the Spanish State failed to provide special protection and assistance to him as an unaccompanied minor; b) contrary to Article 37, he was exposed to the risk of violence and inhuman or degrading treatment in Morocco; c) contrary to Article 3, the Spanish State failed to take into consideration the best interest of the child.
The Procedural and Substantive Rights of Unaccompanied Minors
The decision by the UN Committee on the Rights of the Child issued in February 2019 agreed with the complainant’s claims on all three grounds. Thereby, this precedent case confirms core principles and specifies the procedural and substantive rights of unaccompanied minors at Europe’s borders.
States exercise jurisdiction during border control operations and are therefore bound to respect the safeguards enshrined in the Convention. The Committee confirmed that D.D. was under the effective control of the Spanish authorities when he was apprehended, handcuffed and returned (§13.4 CRC). Therefore, he was under Spanish jurisdiction, irrespective of whether he is considered to have entered Spanish territory or not (ibidem). Spain had argued that the complainant never entered Spanish territory de jure, and therefore did not come under Spanish responsibility (§4.1, §12.4). Thus, according to the Spanish State, the actions of the Guardia Civil constitute a lawful rejection at the border (“rechazo en frontera”), not an unlawful return (§4.1). The Committee’s rejection of these arguments serves as a reminder that States exercise jurisdiction at the border, no matter whether the action is called “rejection”, “return”, expulsion”, “deportation” or otherwise. The Convention on the Rights of the Child applies to all children irrespective of their legal status and cannot be curtailed at the border. This is clearly stipulated in General Comment No 6 on the treatment of unaccompanied and separated children outside their country of origin (Article 12): “State obligations cannot be arbitrarily and unilaterally curtailed either by excluding zones or areas from a State’s territory or by defining particular zones or areas as not, or only partly, under the jurisdiction of the State. Moreover, State obligations under the Convention apply within the borders of a State, including with respect to those children who come under the State’s jurisdiction while attempting to enter the country’s territory.”
The Committee’s decision provides crucial clarifications on questions of admissibility. As to admissibility rationae personae, the Spanish State first claimed that D.D. was not a minor (§4.3) and that there was no definite evidence proving that he was in fact the person being summarily returned (§6.1 CRC), thus prima facie questioning his credibility. As to admissibility rationae materiae, Spain tried to reframe the claimant’s case around the right to asylum, which is not covered by the CRC (§4.4). Spain also argued that the claimant failed to exhaust domestic remedies, because he did not apply for a visa or asylum before trying to cross the fences (§4.5). The Committee rejected all these arguments. In assessing the evidence, the Committee not only asserted that D.D. had provided a consistent account of events, but crucially also highlighted the imbalance between an undocumented minor summarily returned and the State authorities, concluding that the burden of proof could not fall entirely on the applicant (§13.3). Further, the decision confirmed that the case was about rights under the Convention, not asylum (§13.5). On the requirement to exhaust domestic remedies before filing an individual communication, the Committee aligned its jurisprudence to that of other UN treaty bodies and the European Court of Human Rights, confirming that in cases of imminent returns with no procedure, there were no domestic remedies available that could be exhausted (§13.6). The Committee also underlined that visa or asylum applications in transit countries or the Spanish border posts concern alleged possibilities before the return and therefore do not constitute an effective remedy against the latter (ibidem). Moreover, the office for international protection at the Beni Enzar border post are factually inaccessible for individuals from Sub-Saharan Africa.
Regarding the merits, the Committee found that the lack of any procedure to identify D.D. as an unaccompanied minor and the lack of any possibility to present arguments against his imminent return constituted treatment prohibited by Articles 3 and 20 of the CRC (§14.7 CRC). The Committee underlined that respect for the best interest of the child and the special protection of unaccompanied minors includes positive obligations and “extends to requiring States to take all necessary measures to identify children as being unaccompanied or separated at the earliest possible stage, including at the border”, as stipulated in General Comment No 6 (Art 13). States are obliged to conduct an initial evaluation procedure prior to any return (§14.3). This is the first and most crucial step. Denying access to a procedure also implies the denial of any further rights, such as protection as an unaccompanied minor.
The Committee asserted that the lack of any risk assessment prior to the minor’s return violated his rights under Articles 3 and 37 CRC (§14.6). In light of Article 37 and the principle of non-refoulement, States are obliged to evaluate the risk of irreparable damage and grave human rights violations in the country of return (§14.4). In light of Article 3 and the best interest of the child, such an evaluation includes the grave consequences of insufficient alimentary and sanitary services for minors (ibidem). The Committee reached its conclusion taking into consideration D.D.’s past experiences of ill-treatment by Moroccan forces as well as the general violence in Morocco’s border zones (§14.6) – a violence that EU institutions and States seek to outsource and make invisible. Notably, the Committee confirmed the position expressed in its new General Comment No 23 on the human rights of children in the context of international migration, asserting that best interest assessments require “access to the territory, regardless of the documentation they have or lack” (Article 17, quoted at §14.4), as such access constitutes a prerequisite to the initial assessment process.
Last but not least, the Committee emphasized that the way the return operation itself was conducted violated Article 37 CRC: Although D.D. was an unaccompanied minor deprived of his family environment in the context of international migration, he was nevertheless handcuffed and returned without being heard, without legal assistance, without translation services and without taking into consideration his needs (§14.8).
Based on these findings of triple violation, the Committee demanded adequate reparation to D.D. for the harm he suffered. Moreover, the Committee ordered the Spanish State to amend Law 4/2015 and its ‘special regime for Ceuta and Melilla’ in order to avoid a repetition of the same violations and end the unlawful practice of indiscriminate automatic expulsions at the border (§15).
Consequences for Rights at EU Borders
The Spanish policy of summary returns denies those affected access to any procedure that would enable them to present arguments against their imminent return. By arguing that those apprehended at the fences do not come under Spanish jurisdiction, the authorities attempt to create a zone of de facto rightlessness at the border, where safeguards under national, European and international human rights law are suspended. The Committee’s decision constitutes a clear rejection of this rationale, because it clarifies that and how the rights of unaccompanied minors are to be respected at European borders and beyond.
The Committee’s decision further defines the rights of unaccompanied minors at Schengen borders both procedurally and substantively. In the proceedings, the Spanish State referred to Article 13 of the Schengen Borders Code to argue that the Guardia Civil’s actions constitute an exercise of their “right to refuse illegal entry” (§12.1) rather than an unlawful return. Other States at Europe’s borders likewise refer to Article 13 to justify push-backs. However, the Code does not allow summary returns and limits border control to actions in compliance with the State’s obligations under international law. As stated in Article 14 itself, States must issue a substantiated decision with the specific reasons for the refusal of entry. Moreover, as stipulated in Article 4 on Fundamental Rights, Member States shall act in full compliance with relevant international law and consider individual vulnerabilities. This includes the Convention on the Rights of the Child.
The Committee’s decision has further implications for the territorial application of the EU Charter on Fundamental Rights and the EU Return Directive. As highlighted in the Third Party Intervention to this case, the provisions of the Charter are applicable to rejections at the border. The Charter contains an explicit reference to the principle of the best interest of the child (Article 24). Moreover, it prohibits collective expulsions and torture or inhuman or degrading treatment (Article 19), stipulates the right to be heard (Article 41) and requires an effective remedy (Article 47). With regards to the Return Directive, although Member States may decide not to apply the Directive to third-country nationals subjected to a refusal of entry order (article 2, §2, a), this shall be without prejudice to the obligation to respect fundamental guarantees and to consider the needs of vulnerable persons (Article 4). The Return Directive explicitly includes unaccompanied minors in the list of vulnerable persons (Article 3) and refers repeatedly to the principle of the best interest of the child (Preamble 22, Articles 5, 10 and 17).
Although the Committee’s decision concerns unaccompanied minors, the general principles emphasised in it are relevant beyond this case. Access to a procedure to be able to explain their individual circumstances is crucial for unaccompanied minors in need of protection, but also for others who wish to express fear of ill-treatment upon return, submit an asylum application or indicate particular vulnerabilities. Immediate summary returns at the Spanish-Moroccan border deny access to any kind of possibility to present arguments against it. Several cases against such treatment are pending at the European Court of Human Rights, namely the complaints of N.D. and N.T. v. Spain, as well as Doumbe Nnabuchi v. Spain and Balde and Abel v. Spain. The arguments presented by the Spanish State in front of the Court are reminiscent of those presented to the Committee, in particular concerning jurisdiction, evidentiary questions and the reframing of cases as being about asylum applications rather than access to a procedure before a return.
Conclusion and Outlook
The significance of the decision on D.D. v Spain goes not only beyond the individual case, but also beyond the Spanish-Moroccan border. It constitutes a clear condemnation of States’ practice to create zones of exception at the border where basic rights are suspended. Yet, the response of European States including but not limited to Spain has been to adjust push-back practices to avoid legal consequences, not only by outsourcing border control, but also by preventing the collection of evidence on push-backs or by ordering individuals to sign untranslated documents waiving their rights. Courts or quasi-judicial bodies should also scrutinise such practices in order to counteract the de facto denial of access to rights at the borders of Europe.
Acknowledgments: This blog post reflects a collective work process and discussions in the ECCHR Migration Programme. Furthermore, I would like to thank Hanaa Hakiki and Carsten Gericke for their comments.
A recent case challenged successfully part of the UK’s ‘hostile environment’, the self-titled immigration policy created by Theresa May when she was Home Secretary to deter irregular migration to the UK. R (Joint Council for the Welfare of Immigrants) v Secretary of State for the Home Department was handed down by the High Court of England and Wales on 1 March 2019. In dispute was the so-called ‘right to rent’ scheme (‘the Scheme’) which stipulated that certain ‘disqualified persons’ (those who need but do not have leave to enter or remain in the UK) were prohibited from renting or occupying private residential accommodation (section 21(2) Immigration Act 2014). Notably the Scheme deputised landlords by prohibiting them from authorising any occupation of their privately owned properties by disqualified persons and obliging them to conduct ‘reasonable enquiries (i.e. documentary checks) into the immigration status of tenants (section 22). Failure to do so incurred a civil penalty of up to £3,000 (section 23) and any landlord who knew or had reasonable cause to believe that the premises were being occupied by a disqualified person committed a criminal offence (section 33A) punishable by up to five years in prison or an unlimited fine (section 33C).
The Joint Council for the Welfare of Immigrants (JCWI) asked the court to find that the measures were incompatible with the Article 14 ECHR prohibition of discrimination in conjunction with the Article 8 ECHR right to respect for private and family life, home and correspondence. The remedies sought were a declaration of incompatibility with Articles 14 and 8, and either the stopping of the roll out of the Scheme to Scotland, Wales and Northern Ireland, or delaying such roll out until proper research was done into the Scheme’s effects. JCWI’s specific allegation was as follows: the Scheme causes landlords to discriminate on the basis of nationality and or/race ‘against those who are perfectly entitled to rent with the result that they are less able to find homes than (white) British citizens’ [§6]. JCWI argued that landlords fearful of breaking the law would avoid renting to those with a complicated immigration history, those who did not have a passport and were not white, or whose status was not immediately clear. JCWI backed this up with empirical evidence including ‘mystery shopper’ data [§27-34] and surveys of landlords, of whom 42% said they were less willing to let to people without a British passport because of the Scheme and 27% said they were reluctant to rent to people who ‘appeared foreign’ [§20].
The court was satisfied by the Claimant’s empirical evidence of the discriminatory effect suffered by Black and Minority Ethnicity persons (‘BME’) and non-British/EU citizens and that the Scheme did not merely provide the occasion or opportunity for landlords to discriminate but it caused them to do so. Arguments of the Government that a racist landlord constituted a third-party intervenor and thus broke the chain of causation were not accepted [§106].
In relation to justification, the court acknowledged that immigration policy was an area with a particularly wide margin of appreciation. It accepted the Government’s submission that ‘how best to maintain a workable and fair immigration system is very much a matter for the executive’ and added that ‘The European Court of Human Rights is loath to interfere with the right of a State to control immigration where there is no consensus across the Council of Europe as to what is or is not acceptable as a means of controlling immigration’ [§121]. However, the court held this was counter-balanced by the particular abhorrence with which racial discrimination is regarded, and thus agreed with the Claimant that ‘weighty reasons’ were required for justification [§109 adopted by Court a §123]. Particularly persuasive for the court was the evidence that the Scheme was ineffective: of the total number of persons who had come to the Home Office’s attention as a result of the Scheme, only 5% had been deported. Furthermore, the Court was swayed by the evidence that the Home Office was neither monitoring the discriminatory effect, the cost-effectiveness, nor the impact of the Scheme on voluntary returns of irregular migrants [§111, adopted by the court in §123]. In conclusion, the Defendant has failed to justify the Scheme, indeed it has not come close to doing so.’[§123] A declaration of incompatibility with Articles 8 and 14 ECHR was made.
The national political context
The significance of the political moment at which the case was brought cannot be overstated. The 2018 Windrush scandal had involved the so-called Windrush generation, those from Caribbean countries with right to remain under the British Nationality Act 1948 whose families had been encouraged to migrate to the UK in the 1950s, 60s and 70s. In 2018 they were mainly retirement-age citizens who had lived and paid taxes in the UK for decades, but who were detained, made homeless, sacked or denied benefits and NHS treatment because they struggled to prove they were British (some example stories here). Public outrage was redoubled when it was revealed that the Home Office had destroyed landing card documents which were some people’s only record of lawful status. The outrage and injustice felt by many was specifically for those who were lawfully in the country, but being treated as if they were not.
JCWI ‘s litigation was solely mounted on the discrimination point; there was no direct challenge to the way in which the Scheme makes irregular migrants homeless. This is not because they are unaware: the hostile environment is one which, as they commented on the day of the judgment, makes life ‘unbearable’ for those at whom the hostility is directed. The JCWI did not bring — a challenge on one or more ECHR Articles on behalf of irregular migrants themselves, rather than on behalf of lawful migrants.
The irregular migrants’ case which was not brought
Firstly, it needs to be stressed that there is an arguable case against the UK Government directly challenging the Scheme itself and its violation of irregular migrants’ human rights, and not only its discriminatory effect on non-irregular migrants. That the court declined to find that Article 8 right to one’s home is directly engaged does admittedly feel like a closed door: Spencer J states ‘the Scheme does not engage Article 8 directly by reason of interference with the rights protected by that Article. If it did, then this would be tantamount to acknowledging that Article 8 gives a person the right to a home’ [§61], citing the case of Chapman v UK. However, the court opened a window with its conclusion that, although Article 8 does not give anyone the right to a home, ‘it gives everyone the right to seek to obtain a home for themselves and their family even if they are eventually unsuccessful, and the playing field should be even for everyone in the market for housing, irrespective of their race and nationality.’ Spencer J thus concludes that, ‘[w]here the State interferes with the process of seeking to obtain a home, in my judgment it must do so without causing discrimination and this either engages Article 8 or comes within its ambit.’ [§68] Therefore, a combined Article 8 and 14 ECHR challenge on behalf of irregular migrants, the Article 14 ground being on the basis of their migration status, is not excluded.
Furthermore, a stronger Article 8 ground alone is in relation to the eviction powers which were added to the Scheme. These gave landlords the power to evict ‘disqualified persons’ without court order, and indeed required an attempt at eviction to comply with their civil and criminal obligations. This judgment mentions evictions in passing [§11] but there is no analysis since they are not the focus (the focus is the lawful residents’ access to housing, and the lawful resident cannot be evicted). The necessity and proportionality within a democratic society of coercing landlords to evict would be analysed. The fact that such eviction would render those tenants likely homeless since the Scheme prevented them from renting further accommodation would have to be considered.
When it comes to this proportionality analysis, the perennial problem is the wide margin of appreciation in migration matters. Spencer J stressed this deference in the judgment, highlighting that:
iii) a fair and workable immigration system will involve many different parts or strands which will often, or usually, together form a coherent whole, intended to complement each other and work together: thus, for the court to interfere with one aspect potentially causes havoc to an overall strategy devised by the Government in accordance with its democratic mandate;
iv) The European Court of Human Rights is loath to interfere with the right of a State to control immigration where there is no consensus across the Council of Europe as to what is or is not acceptable as a means of controlling immigration;
v) Control of immigration must be recognised as a political issue which features near the top of highly charged political issues which are of concern to voters whether voting in a general election, by-election or a referendum. [§121]
Thus, any found interference with Article 8, or with a combination of 8 and 14, would have to pit enforced homelessness against the state’s sovereign ability to restrict social and economic rights as a form of migration control. Envisioning this analysis — under the heading of ‘proportionality’ in an Article 8 challenge and ‘justification’ under an Article 8 and 14 challenge — one can anticipate stronger arguments on both sides than in the present case. For example, JCWI argued that there was a disadvantage on the housing market for those lawful residents discriminated against, but their access to private rental housing was only reduced as opposed to the complete eradication of irregular migrants’ access. However, the counter-argument is also stronger in its ability to rely on the wide margin of appreciation in migration control since the human rights interference is not an unintended consequence of the Scheme — as with lawful residents — but it is the purpose of the Scheme itself. Within this analysis, therefore, not the purpose of the Scheme but perhaps its efficacy would be the most effective argument, asking: is a high level of homelessness or detention proportional to the paltry 5% deportation rate of the current Scheme?
Article 8 is not the only option, however. The Scheme was criticised in 2017 by Hannah Lewis, Louise Waite and Stuart Hodkinson for leaving irregular migrants with ‘no real or acceptable alternative but to turn to less-detectable informal lodgings’. As a result, their analysis argued a consequential marginalisation of other aspects of their social interactions, such as ‘transactional labouring experiences that act as a driver for entry into exploitative and forced labour conditions’. Furthermore, the causal link between the state and homelessness could realistically lead to an Article 3 challenge, to the extent that forced homelessness breaches the absolute right to be protected from inhuman and degrading treatment. Distinguishing the case from MSS v Belgium and Greece, which reiterated that ‘Article 3 cannot be interpreted as obliging the High Contracting Parties to provide everyone within their jurisdiction with a home … Nor does Article 3 entail any general obligation to give refugees financial assistance to enable them to maintain a certain standard of living’ [§249], the Scheme is not in any way removing financial assistance, or obstructing the State providing a home. Instead it prevents access to what would otherwise be a self-funded, independent contractual arrangement with a third party.
The Permanent People’s Tribunal considered the Scheme in November 2018, highlighting the full gamut of international human rights commitments which should prevent the UK from continuing to enforce the Scheme: The ‘right to rent’ measures breach the right to adequate housing without discrimination, which is recognised in the Universal Declaration on Human Rights (UDHR) Art 25 (as an integral part of the right to an adequate standard of living), and in Art 11 of the International Covenant on Economic, Social and Cultural Rights (ICESCR), further amplified by General Comment No 4 on Adequate Housing by the UN Committee on Economic, Social and Cultural Rights (1991). Insofar as homelessness affects human dignity and physical and mental integrity, the measures also breach Art 1 UDHR, Arts 3 and 8 of the European Convention on Human Rights (ECHR) and Arts 1 and 3 of the EU Charter of Fundamental Rights (EUCFR).
Furthermore, the matter is not only being highlighted retrospectively. In 2010, well before Theresa May announced the ‘hostile environment’ policy, the UN Special Rapporteur on adequate housing as a component of the right to an adequate standard of living reported specifically on the rights of migrants, concluding that:
80. Regardless of their immigration status, all migrants must be afforded essential levels of protection with regard to their access to adequate housing and living conditions.
81. The provision of housing should not be denied to undocumented migrants.
However, it is unsurprising that this report did nothing to deter the UK Government. The willingness of the UK to proceed in denying the right to rent to those with irregular status is another micro case of the macro trend: the presumption of state sovereign control over migration regulation, to the detriment of human rights. Linda Bosniak’s observation in 1991, of ‘the enormous almost talismanic power that assertions of state sovereignty have had often in the area of human rights for aliens’ is just as relevant today as it was then.
In the name of lawful residents: the case which was brought
JCWI’s excellent strategic litigation, then, reflects a wider truism within migration law: that of the questionable feasibility of a successful human rights challenge based on the rights of irregular migrants alone. The margin of appreciation remains wide when the targets of an immigration scheme are those whose human rights it hurts. However, as we see in the judgment, when the innocent bystander — the lawfully resident bystander — is also hurt, even when their access to housing is only partially impaired as opposed to the irregular migrants total ban — the margin narrows and ‘weighty reasons’ rhetoric enters the judgment. Thus, in focusing on the injustice of lawful residents being treated as if they were irregular migrants, JCWI were choosing the horse that they knew would run the course, in that particular Windrush moment and within the dynamics of the court’s deference to the state in the treatment of aliens. The cynical reading perhaps is that, in choosing not to directly challenge the Scheme, the legitimacy of its aim — migration control through enforced homelessness — is tacitly accepted, even condoned. The policy in theory is left untouched and only its de facto discriminatory effect on lawful residents needs to change.
However the non-cynical reading is one of indivisibility: in highlighting the discrimination, they do not allow the Scheme to be limited to irregular migrants’ rights alone. This indivisibility is underlined by de facto result of the judgment: the housing situation of both lawful and unlawful residents is improved: the Scheme has been declared incompatible with the ECHR rights. Furthermore the strategy here not only secured a de facto relief for irregular migrants, but through this tactic of indivisibility, there is a clever and effective seizing of an opportunity to protect irregular migrants by relying on a population who are more recognisably rights holders. And the message is: a hostile environment for them is a hostile environment for us. I repeat the earlier quotation from JCWI on the day of the judgment: ‘the aim of making life so unbearable for people in the country without leave cannot be enforced without creating a hostile environment for all of us — especially those who look or sound ‘foreign’.
Exploring the formation of asylum at the national level in Hungary, Slovenia and Croatia, it may be noted that not long ago these EU Member States had nothing close to a developed asylum system in place, this is why their EU membership had to be followed by the absorption of the relevant EU legislation in their legal orders. Unlike many of the older Member States who had the chance to participate in the negotiations and adoption of the main asylum instruments, these states accepted a pre-existing legal and policy reality which was not necessarily meant to accommodate the asylum demands and migration issues specific to their region. What is addressed in the contributions is the degree of autonomy enjoyed by the domestic legislator and the practical consequences stemming from this discretion, especially in relation to three areas: access to territory, access to asylum procedures, and application of the safe-third country rules. In reviewing the legislative changes in Hungary, Slovenia and Croatia since 2014, it can be noted that these are heavily modelled on the restrictive policy measures undertaken by other EU Members. As a result, although firmly embedded in international and European law, asylum law remains insufficiently implemented in the domestic legal systems in question.
THE NEIGHBOURING TRANSIT COUNTRIES
The EU’s neighboring countries, forming the so called Balkan route – Turkey, Macedonia, Serbia and Albania – have played a significant role during the crisis. As with new EU asylum countries, these countries lack experience in dealing with the reception and treatment of migrants and refugees, as well as in the adoption and proper implementation of integration policies. The movement of migrants and refugees towards the EU through the Balkan Route during the crisis exposed the unpreparedness of both the Union and the countries located in the broader region in coping with such events through meaningful cooperation. In addition, they suggest that the EU efforts to foster capacity building in these countries are likely to have unexpected implications in terms of refugee rights, as these efforts themselves are underpinned by contradictory assumptions and objectives (migration control rather than protection). As a result, the few positive outcomes of cooperative arrangements such as the financial aid provided by the EU to support refugees within the EU-Turkey deal seem to come at the expense of core refugee law principles.
THE QUESTION OF SOLIDARITY IN THE REGION
In an effort to understand the distributive effects of the European asylum mechanisms in combination with the reality of refugee movements, evidence suggests that national governments, both within and beyond the EU, are likely to continue informing the interpretation and implementation of asylum law and policy in the EU. Within this context, any future reform of the Common European Asylum System should take all interests into account, and aim at long-term rather than ad-hoc responses to the uneven distribution of asylum responsibilities amongst states.
THREE POLICY SHIFTS
The Common European Asylum System, for all its flaws, became a driving force of legal developments in asylum both on the ground and at the national and sub-regional levels. Nevertheless, the 2015/2016 crisis exposed the incoherence in the circulation and the transplantation of asylum norms across Europe, on the one hand, and the disconnection between the law and its application, on the other. In particular, the crisis appears to have triggered three national and sub-regional policy shifts:
1. Faced with the possibility of becoming the main destination for asylum seekers and refugees after Germany, Sweden shifted from actively welcoming asylum seekers to implementing strict asylum policies. Similarly, faced with the possibility of becoming a transit country, Albania altered its rather neutral position to one that returned ‘undocumented’ migrants to Greece. Furthermore, Hungary moved from being a Member State which in the course of adopting the EU acquis was more on the liberal side, to a country showing total denial of the asylum acquis and disapproval of decision-making in the EU. This demonstrates the great influence that states opposing cooperation exert on the application of asylum norms. From a fundamental rights perspective, the EU and non-EU countries in question appear to have operated under the assumption that asylum seekers and refugees are ‘a burden’ to be avoided at all costs. In other words, a refugee is not the person to be protected but a person to be fenced out. This ‘preservation of the system over protection’ rationale is reflected, among others, in the humanitarian visas X X case where the CJEU clarified that any state willing to provide refugees with alternatives to accessing asylum would have to deal with this individually as a matter of national policy. This seems to support the argument that the responsibility to protect applicants from refoulement is placed on each Member State individually while there is a collective responsibility to control and deter migrants and refugees from entering the EU.
2. Some countries both within and outside the EU, such as Slovenia and Serbia respectively, shifted from ‘transit’ to ‘destination’ countries. This resulted in a redirection of their policy from ensuring adequate treatment to migrants passing through their territory (in the form of shelter and humanitarian supplies) to fundamentally improving their asylum procedure and reception conditions. However, it seems there was no adequate effort from the EU nor the CJEU, as the guarantor of fundamental rights within the EU, to avoid another ‘MSS’. The CJEU had the opportunity in both the A.S. and Jafari cases to declare the first entry criterion in Dublin Regulation not suitable within a context of a crisis, chiefly as opposing the principle of solidarity between states. The court could have opted for a circumstance-specific interpretation of the Dublin Regulation based on relevant EU and international norms and principles which would have alleviated some of the disproportionate pressure put on the countries in question and, most importantly, make it more likely for asylum seekers to receive proper treatment (see AG’s Opinion). The Court also had the opportunity to clarify the limitations of readmission agreements such as the EU-Turkey statement from a human rights perspective and to ‘save’ the EU’s legitimacy. It did neither, but instead preserved the integrity of the system despite there being no chance of its effectiveness.
3. Countries that were once on the sidelines of asylum in Europe have started to play a significant role in shaping the asylum system in the EU. The case of Macedonia, for instance, highlights the significance of migration and asylum in the European integration process and explains why the EU needs to take these countries seriously in the future.
The radical rethink of the Common European Asylum System should take these three shifts as a starting point. Any reform that does not take the above into consideration is likely to contribute to a less effective, less solidary and less human-rights oriented system, eventually transforming the crisis into the new norm.
This is the second post in a two-part series on the internal protection alternative (IPA) based on Jessica Schultz ’s new book on the topic.
The surrogate role of refugee law: a principle or preoccupation?
As described in the previous post, the ‘internal protection alternative’ (IPA) is a limit on refugee status used to exclude claimants with access to adequate protection somewhere within their countries of origin. For example, an Afghan who fled Taliban recruitment in Helmand province may be returned to Kabul if conditions are sufficiently safe there. In most jurisdictions, return must also be reasonable for the person concerned.
Although there is no mention of the IPA in the 1951 Refugee Convention or its 1967 Protocol, this implied limited spread in state practice during the mid-1980s as a response to several factors, including the prominence of non-state persecution as a basis for refugee claims. Today, the concept has firmly taken root in many jurisdictions despite continued debate about the exact criteria for IPA application.
The IPA is also sometimes applied to justify decisions relating to the cessation of refugee status when some area of the country of origin is deemed secure enough to accommodate returning refugees. The migration of the IPA concept from the inclusion to the cessation analysis is not broadly endorsed (see UNHCR’s position and UK jurisprudence), but there is little doubt that states are increasingly interested in identifying some safe space – anywhere – to which refugees or refugee claimants may return within their countries of origin.
My book analyzes various ways that states and scholars have framed the IPA limit in refugee law. This post examines a common element of these efforts, the observation that international protection is ‘surrogate’, or subsidiary, to national protection provided by countries to their own citizens. As Zimmermann and Mahler explain, ‘the evolution of the [IPA] concept draws heavily on the notion of surrogacy as a basic principle of refugee law, according to which international protection only comes into play when national protection within the country of origin is not available.’ If national protection exists somewhere, the back-up remedy of protection abroad is not (or no longer) needed.
The surrogate role of refugee law: ‘thin’ versus ‘thick’ perspectives
As Goodwin-Gill and McAdam have written, the notion of ‘surrogacy’ usefully describes the overall role of refugee law. The 1951 Convention obliges state parties to provide replacement protection, when a person can demonstrate a well-founded fear of persecution within their country of ‘origin’ (citizenship or previous residence in the case of stateless persons) (Article 1A(2)). Beyond the inclusion provision, Article 1 also contains criteria for exclusion and cessation of refugee status. These too confirm the subsidiary, or surrogate, character of Convention protection. For example, Articles 1A(2) para 2, 1C(5), 1C(6), and 1E all indicate that when there is no well-founded fear of persecution within the country of origin or another country where the claimant has a national connection, refugee status need not be recognized. However, neither the text nor other sources (including the extensive drafting history, as discussed in the book) suggest a similar limit for persons for whom protection is available in only part of a country.
It may be helpful, then, to distinguish between a ‘thin’ surrogacy perspective and the ‘thick’ one that is sometimes referred to as a ‘principle’ of refugee law. According to a ‘thin’ surrogacy perspective, the lack of protection by the state of origin is relevant insofar as it negates the well-founded nature of a claimant’s fear. This absence of protection is a characteristic of a person who meets the Convention criteria; it is not a condition of refugee status. A fear of persecution is well-founded because the state has failed to ensure a minimum level of security to suppress a risk of serious harm. The IPA is not compelled by this interpretation of refugee law, although there may be cases in which the claimant’s unwillingness to avail him or herself of home state protection, despite a well-founded fear, cannot be reasonably justified.
According to a ‘thick’ surrogacy perspective, meanwhile, state protection comes into play at two separate stages in the assessment of refugee status. First, serious harms committed by non-state actors qualify as persecution for a Convention reason when the state is unable or unwilling to protect the claimant. Second, the possibility of protection elsewhere in the country of origin must be considered to establish the necessity of protection abroad. In other words, protection has a systemic aspect (related to the state’s ability and willingness to protect from the original harm) as well as a territorial one (related to the possibility of protection elsewhere).
There is nothing in the structure of the Convention to suggest that the presence of protection somewhere in a state’s territory can defeat a claim to refugee status. Article 1A(2) of the Refugee Convention defines a refugee as someone who ‘owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable, or owing to such fear, is unwilling to avail himself of the protection of that country….’
The criteria for refugee status are satisfied, in other words, when the claimant 1) is outside his or her country of origin; 2) has a legitimate fear of persecution for a Convention ground; and 3) is unable or unwilling, owing to the well-founded fear, to avail him or herself of that country’s protection. The ‘thick surrogacy’ perspective shifts focus from the claimant’s ability or willingness to avail herself of home state protection to the ability or willingness of the state to provide it (within the country of origin).
Despite the fact that a ‘thin’ surrogacy perspective aligns better with the text of the treaty, a ‘thick’ surrogacy perspective has gained traction in the scholarly discourse as well as in national jurisprudence. Below are some of the factors – by no means exhaustive – that help explain this development.
Factor 1: The influence of human rights law and principles on interpretation of the refugee concept
One explanation for the persistent power of ‘surrogacy’ as a justification for limiting the scope of refugee status relates to the misuse—in my view—of human rights law and principles for the purpose of interpreting the Refugee Convention. In terms of protection criteria, non-refoulement cases decided by human rights courts have influenced states’ interpretation of the 1951 Convention’s refugee concept. These do not frame the IPA as a limit on the scope of refugee status but rather as an extension of the basic risk analysis covering conditions on return.
Procedural concepts from the human rights field have also infiltrated substantive interpretations of refugee law. Some sources refer to the duty to ‘exhaust domestic remedies’ in support of IPA practice. The ‘exhaustion of local remedies’ rule applies in human rights law to preserve the subsidiary nature of supranational decision-making bodies vis-à-vis a more democratic local organ. Despite its authoritative ring, it makes no sense in the context of refugee claims. First, the rule is usually concerned with systemic procedures (has the case been heard by the state’s highest court?) and not with the absence of protection by local state agents in certain areas. It also involves a backward-looking analysis (what has been done?) in contrast to the prospective inquiry required to assess the need for protection.
Factor 2: The influence of other disciplines in the interpretation of the refugee concept
A second explanation for the prominence of the surrogacy discourse relates to broader developments in the field of refugee studies. In recent decades, important contributions to the refugee concept from social scientists and philosophers have filtered into legal debates. With some exceptions, these scholars (see here and here) argue that international protection should extend beyond the confines of the Geneva Convention to include persons forced to flee their countries of origin for reasons that may or may not have a Convention nexus.
For example, in his influential article ‘Who is a Refugee?’ (1985), Shacknove posits the following definition:
[R]efugees are, in essence, persons whose basic needs are unprotected by their country of origin, who have no remaining recourse other than to seek international restitution of their needs, and who are so situated that international assistance is possible.
Despite the fact that Shacknove’s contribution to the field of forced migration was never meant as a restatement of refugee law, legal scholars like Hathaway and Storey have cited him in support of their view that the possibility of protection somewhere in the country of origin may, under certain conditions, defeat a claim for refugee status. Scholarly contributions, even those motivated by the desire to expand the refugee concept, have reinforced the view that a deserving claimant is one whose compelling needs arise from the absence of adomestic alternative.
Factor 3: The self-referential nature of norm development in refugee law
Without a treaty monitoring body to oversee the Convention, the situation arises in which certain ideas take on a life of their own through the interplay of jurisprudence, scholarship and policy. The concept of surrogacy has undoubtedly gained momentum through the inter-jurisdictional ‘dialogue’ in the field of refugee law and the influence of leading scholars. Its expression, however, has also been marred by circular reasoning. For example, the Canadian Supreme Court in Ward refers to the first edition (1991) of The Law of Refugee Status of Hathaway to explain the surrogate role of refugee law. Although the legal question was about whether non-state persecution was covered by the Convention’s refugee concept (and not about a potential IPA), this case appears regularly in IPA literature. Indeed, authors of the second edition of The Law of Refugee Status cite Ward as evidence of positions proposed in the previous book.
Factor 4: Harmonization impulses versus the Refugee Convention as ‘lex specialis’
As most states operate with multiple regimes of international protection, the ‘surrogacy principle’ can be leveraged to justify the IPA as an unwritten exception to refugee status no matter what its treaty basis may be. For example, Article 8 of the EU Qualification Directive (2011) provides in general terms that internal protection may be considered ‘(a)s part of the assessment of the application for international protection’. By framing the IPA in this way, it is easier for states to justify a common IPA test in claims to Convention refugee status as well as those that relate to complementary forms of protection. As discussed elsewhere, the consequence has been that human rights courts like the ECHR are increasingly setting the standards for interpretation of the Refugee Convention – also for IPA practice – rather than the other way around.
Implications of a ‘thin’ surrogacy perspective for IPA practice
While the Refugee Convention’s purpose is to provide substitute protection, recognition as a refugee does not depend on the absence of a domestic alternative. Nonetheless, an IPA limit may apply under certain conditions: when a claimant can relocate within the country of origin with minimal negative impact (from either an objective human rights perspective or a more subjective, humanitarian one). In these cases, extending international protection would erode the treaty’s effectiveness. In addition to individual factors, sending states must consider structural ones: whether return would reinforce policies or practices of ethnic cleansing, or stress already fragile areas. Both dimensions require a displacement-sensitive analysis.
The scope for IPA practice when revoking refugee status is even narrower. As UNHCR has rightly maintained, where a risk of persecution persists in one region of a country, it is unlikely that the changed circumstances are ‘profound’ and ‘enduring’ as required under the cessation analysis. There are also situations in which circumstances have objectively changed but return is still unreasonable due to past experience of persecution.
The IPA question goes to the heart of what it means to be a refugee. Is a well-founded fear of persecution within the country of origin enough to establish a refugee claim? While persecution – especially by non-state actors – may not always justify the claimant’s recourse to refuge abroad, the presumption that it could should not be diluted. The surrogacy concept in many jurisdictions has done just that, by shifting the focus of refugee status determination from the risk of persecution to the possibility of return in line with minimal human rights standards. This not only downplays the relevance of refugee experience, but it also renders the unique protection issues attached to internal displacement – a consequence of IPA practice – largely invisible.
During the summer of 2018, a fierce debate on how to counter “secondary movements” (to Germany) of potential asylum seekers took place inside the German government. The discussions nearly put an end to the parliamentary group between the CDU and the (Bavarian) CSU, which had existed with a very short interruption in 1976 since 1949. The CSU, in particular the Federal Minister of the Interior (MoI), Horst Seehofer, opted for full controls at the German-Austrian border and the refusal of entry for all persons that are not in the possession of valid documents for an entry to Germany. The MoI claimed that the German constitution as well as immigration and asylum law would require/allow for a refusal of entry and push-backs of asylum seekers due to their entry from a safe third country (Austria). In this regard, German law should be applicable and so no Dublin procedure would be necessary. The CDU and in particular Chancellor Angela Merkel preferred a “European solution” in cooperation with the other Member States based on agreements under Article 36 Dublin III Regulation. This provision allows for Administrative Arrangements only for the purpose of “the exchange of liaison officers” and the “simplification of the procedures and shortening of the time limits relating to transmission and the examination of requests to take charge of or take back applicants.”
As a compromise it was decided to extend the “fiction of non-entry” (Fiktion der Nichteinreise) that applies to the (non-Schengen) transit zones of international airports to this border area. This fiction of non-entry should allow for controls on German soil and the refusal of entry would (legally speaking) push the person that had already reached German soil outside the territory into a perceived no man’s land between the controlling officer and the Austrian border. This effect might be called a “border spell” as the persons – unlike in the non-Schengen zone of an international airport under Annex VI of the Schengen Borders Code – have already legally and physically crossed the border between the two states and had previously already entered the Schengen territory (as Austria is also a Schengen Member State). According to its construction the zone that lies between the controlling officer and the physical Austrian border would be transformed by the refusal of entry into a non-Schengen territory and an EU Law free zone, where German national law and not the Schengen Border Code (SBC) or the Dublin III Regulation would be applicable.
Administrative Arrangements with Greece and Spain
Germany subsequently concluded Administrative Arrangements with Portugal (under Article 36 Dublin III Regulation), Greece and Spain. The latter two Administrative Arrangements contain the agreed procedure in the event that a person is refused entry on the occasion of a check at the German-Austrian border, knowing however that they are not applicable to unaccompanied minors. The conditions for the applicability comprise the refusal of entry, the submission of an application for international protection and a Eurodac Category 1 hit (for an “asylum application”) from Greece/Spain (No. 1 of both Administrative Arrangements – i.e. its first provision). If such a case occurs, the Greek/Spanish authorities will be notified, stating the Eurodac number, photograph, itinerary data and place of apprehension. For this notification the Annex to the Arrangements contains a form (No. 2). The “implementation of the refusal of entry” (i.e. the actual transfer of the person) shall be initiated within 48 hours “after the person has been apprehended at the internal border” (No. 3).
In the German-Greek agreement it is furthermore stated that the Greek side retains a possibility to object the transfer within six hours if it demonstrates why the conditions set out in No. 1 are not met. If the Greek side shows within seven days that the person has been wrongly transferred, Germany undertakes to readmit the person already transferred (No. 4). There is no such time limit in the German-Spanish agreement. No. 5 designates the National Coordination Unit for Border Controls, Migration and Asylum in Athens/the General Commissariat for Aliens and Border (Spain) and the Federal Police Headquarters in Munich are responsible for the implementation of the Arrangements. Germany undertakes to bear all costs associated with the transfers (No. 6). No. 15 of the German-Greek agreement contains the beginning and possibilities for unilateral termination of the Arrangement. It will end at the latest when the reform of the Common European Asylum System (CEAS) enters into force (No. 15 in fine). No. 7 of the German-Spanish agreement does also contain the start date and the scenarios for amending or ending the Arrangement. There is no reference to the CEAS reform in this document.
The first parts of the Arrangements between Germany and Greece/Spain allow for a procedure that is different from the procedure foreseen by the Dublin III Regulation. The agreed approach goes way beyond the simplification of procedures and the shortening of time limits and is therefore not compatible with the scope and purpose of Arrangements under Article 36 Dublin III Regulation. Therefore, the first parts of the Arrangements cannot stand as agreements under this provision. If EU Law was applicable, both agreements would alter the Dublin rules and violate EU law.
The agreement with Greece contains additional parts (Part 2 (Nos. 7-12) and Part 3 (Nos. 13-15)) that are not relevant as such to the topic of this blog. Part 2 was setting out a solution for pending Dublin family reunification procedures where the persons would travel from Greece to Germany. Timelines and procedures were clarified with a two-fold purpose: 1) Clearing the backlog by the end of 2018 and 2) (re-)starting the regular cooperation within the Dublin procedure for new family reunion cases. The final provisions (Part 3) describe the specific administrative cooperation, including the possibility of exchanging liaison officers (No. 13) and the establishment of a Dispute Settlement Committee (No. 14).
The provisions contained in the two parts on family reunion (Part 2) and general cooperation (Part 3) would generally be in line with the scope and purpose of Article 36 Dublin III Regulation. If the consultation and notification procedure with the EU Commission is initiated as foreseen in Article 36(3)-(5) Dublin III Regulation is followed, these parts may serve as a model for future cooperation between Greece and Germany. Such agreements exist inter alia between Germany and Austria and – concluded in September 2018 – between Germany and Portugal.
Outside the scope of EU Law?
In a letter of 20 November 2018 to the President of the German Bundestag (Wolfgang Schäuble) MoI Seehofer argued that the latter two Arrangements are purely bilateral and therefore not falling within the scope of Article 23(2) Basic Law, which provides for full and as early as possible information of the Parliament in “matters concerning the European Union.” “Looking at all circumstances” the Arrangements would have “no close relation to the law of the European Union.” The construction of this argument turns around two assumptions:
The refusal of entry is conducted under national law
The border zone at the internal border is in the context of these controls an EU law free zone
This view presupposes a “non-Schengen zone” created by the border spell within the Schengen area, where EU law is not applicable. Consequently, such controls would be conducted under national law and the Arrangements could apply without being in conformity with EU law. This means that there would need to be an EU law free zone in which neither the Schengen Borders Code , nor the Common European Asylum System (including the Dublin III Regulation) nor the Returns Directive are applicable. This view, which was also put forward in a blogpost on this matter, lacks a legal basis as EU law is applicable to all matters concerning asylum seekers that have reached a border or a transit zone. Consequently, the secondary legislative acts adopted under this Chapter are fully applicable (ratione temporis, personarum, materiae and loci) to interceptions at the Austrian-German border.
National rules have therefore to follow and comply with EU law. Looking at the principle of precedence of EU law it could not (to put it in the words of the CJEU in Costa/ENEL) “be overridden by domestic legal provisions, however framed”. Provisions and agreements not complying with EU law are inapplicable and need – even if they have been adopted subsequently – to be set aside in order to guarantee that they have no legal effect (see e.g. CJEU, Simmenthal II, para. 14pp).
Four examples related to the implementation of the Arrangements further demonstrate the legal non-sustainability or even absurdity of the „border spell“ and thereby show that EU Law is applicable in these situations:
If the refusal of entry at the German-Austrian border is solely subject to national law, why are returns to Austria governed by agreements under Article 36 of the Dublin III Regulation?
Article 14 (1) SBC states: “[Refusal of entry] […] shall be without prejudice to the application of special provisions concerning the right of asylum and to international protection or the issue of long-stay visas.” Why would this provision not apply at the internal border, especially as Germany has exactly argued that Article 14 SBC is applicable in the context of internal border controls in the CJEU case Arib? If Article 14 SBC does not apply, there would be no legal provision allowing for the refusal of entry at internal borders.
Article 3 (1) Dublin III Regulation states that the Dublin system applies to any asylum application lodged “on the territory of any [Member State], including at the border or in the transit zones.” How would this be different in border zones inside the Schengen area and lead to the application of national provisions instead of EU law?
How could an analogy to the airport procedure lead to a non-application of the Dublin procedure if Article 15 Dublin III Regulation explicitly refers to application on airport transit zones as one of the criteria?
As the procedure foreseen in the two Arrangements deviates from the above mentioned rules, the agreements are violating EU law. Consequently, the Arrangements need to be set aside and there is an obligation for authorities and courts to secure that there is not “any legal effect” of the Arrangements that is not in line with the Dublin rules.
Transforming the Arrangements into conformity with EU Law?
The Arrangements could therefore only take effect if it was possible to apply or interpret them in accordance with EU law. The only possibility not to conduct a Dublin procedure is to apply the so-called sovereignty clause (Article 17 Dublin III Regulation) and to conduct a national asylum procedure (see CJEU, Fathi of October 2018). In this national asylum procedure, a transfer to a safe third country remains possible (see CJEU, Mirza), but a transfer to another Member State violates EU law if no Dublin procedure is conducted (see mutatis mutandis CJEU, Hassan). The binding time limits (see CJEU, Mengesteab and Shiri) may be shortened but not altered to a by default acceptance on notification by Arrangements under Article 36 Dublin III Regulation. On a general level, the Dublin procedure does not leave room for exemptions from the general procedural rules (see e.g. CJEU, Hassan) and appeals procedures (see e.g. CJEU, M.A. and others). The ECtHR has made it clear that the lack of such remedy may constitute a violation of Article 13 in conjunction with Article 3 ECHR (see M.S.S. v. Belgium and Greece). The full assessment of the responsibility of another Member State has to be conducted before a request is lodged to such a Member State (see CJEU, Ghezelbash and Karim) that may be followed by a decision and a transfer. Moreover, before a transfer is carried out, a full and individual assessment of potential non-refoulement obligations has to be conducted (see e.g. CJEU, C.K. and others).
The same is true for border controls and returns within the Schengen area as the CJEU has confirmed inter alia in the cases Affum and A. This applicability of European Law does not change during periods where under exceptional circumstances internal border controls have been reinstated. Already the fact that the CJEU accepted the preliminary reference in the case Arib and others, as well as the Opinion of the AG Szpunar in this case, confirm this applicability. Moreover, as these border controls have been conducted for more than three years now, they lack – on a more general level – a legal basis as such, as the maximum period for such reinstated internal border controls is two years (see Articles 25pp SBC).
In essence, the two Arrangements violate EU law and may therefore not be applied due to the principle of precedence of EU law. Neither Arrangement leaves room for effective legal remedies. This is of concern especially regarding Greece as expedited transfers to this Member State seem like an attempt to circumvent non-refoulement obligations by implementing a procedure without an effective legal remedy in Germany before the transfer is carried out.
The Arrangements are unlawful as they foresee procedures that are incompatible with the Dublin rules. Therefore, they cannot be applied. Moreover, using the “border spell” as a trick to suspend the practical applicability of procedural rights and to circumvent non-refoulement obligations is not a new idea in the European asylum policy (see e.g. Frei/Hruschka on this blog). However, the application of such constructions within the Schengen area may actually do more than just suspending/circumventing the Dublin rules: the enacting of such procedures presupposes the existence of (quasi) permanent border controls. Such controls are not only violating the main principle of the Schengen acquis – the free movement within the Schengen area – but are also, as underlined by the Civil Liberties Committee of the European Parliament in the context of the current reform, “illegal” and a “misuse” of the system. The easiest way to avoid further confusion would be to use the unilateral possibility to discontinue the application foreseen in both agreements. In the meantime, authorities and courts are called upon to prevent “any legal effect” of these Arrangements. Any other solution would be incompatible with EU Law.
On se souvient de la vive polémique survenue en Belgique fin 2016, à propos de visas dits “humanitaires” sollicités par une famille syrienne en vue de rejoindre le territoire belge de façon sûre et légale pour y demander l’asile. La Cour de Justice de l’Union européenne (CJUE) était intervenue par son arrêt X et X c. Etat belge du 7 mars 2017 , décidant que de telles demandes de visas relèvent du seul droit national et non du droit de l’UE, de sorte que la Charte des droits fondamentaux de l’UE (CDFUE) ne trouve pas à s’appliquer. Bien que cet arrêt fut accueilli avec soulagement par le gouvernement belge et les 13 Etats intervenants, loin de clore la polémique, il a suscité de nombreuses réactions virulentes, certains allant jusqu’à dénoncer l’opportunisme politique de la Cour. Aujourd’hui, la saga continue devant la Cour européenne des droits de l’homme, la Grande Chambre ayant été saisie d’une affaire similaire, M.N. et autres c. Belgique, le 26 novembre 2018. La Cour de Strasbourg prendra-t-elle le même chemin prudent de la CJUE, ou bousculera-t-elle le système européen d’asile ?
1. La question de l’ouverture de voies d’accès légales et sûres
La question de l’octroi de visas humanitaires permettant aux demandeurs d’asile d’accéder légalement et en toute sécurité au territoire belge doit être replacée dans le contexte plus large de la « crise des réfugiés » qui a secoué l’UE en 2015-16. Ce contexte a mis en lumière l’un des grands paradoxes du droit international et européen de l’asile : autant le droit de fuir la persécution et de chercher l’asile est garanti, autant le droit d’atteindre le territoire d’un pays tiers en vue de solliciter l’asile ne l’est pas. Paradigme implicite du droit des réfugiés, les Etats ne se considèrent obligés vis-à-vis des réfugiés qu’une fois qu’ils sont parvenus sur leur territoire ou se trouvent sous leur contrôle. D’une part, et la « crise des réfugiés » l’a démontré à suffisance, l’accès spontané au territoire européen, déjà périlleux dès lors qu’il s’agit de traverser une mer, est rendu en pratique excessivement difficile par les Etats membres qui multiplient les obstacles frontaliers dans le cadre de la lutte contre l’immigration irrégulière devenue l’objectif essentiel de la politique européenne. Ces réactions contraignent les demandeurs d’asile à emprunter des voies toujours plus dangereuses, seuls moyens d’accéder à “l’Eldorado européen”, résultant en plus de 5200 morts en Méditerranée en 2016, 3100 en 2017 et plus de 3000 en 2018. D’autre part, s’il existe certaines voies d’accès légales et sûres, elles sont largement insuffisantes de sorte que la grande majorité des personnes qui se voient reconnaître le statut de réfugié ou le bénéfice de la protection subsidiaire, sont entrées irrégulièrement (c’est-à-dire sans remplir les conditions requises pour avoir accès au territoire européen, en particulier l’obligation de visa imposée à certains ressortissants de pays tiers par le règlement 539/2001.
Dans un tel contexte, le développement de voies légales et sûres d’accès au territoire européen n’a de cesse d’être revendiqué par de nombreux experts internationaux et acteurs de la société civile. Parmi ces voies d’accès légales, on trouve lles dispositifs de réinstallation et les visas dits “humanitaires”. Par rapport à la dénomination de cette dernière voie d’accès, il est permis de s’interroger sur son inadéquation puisqu’il s’agit plus précisément d’un visa “asile”. A ces deux voies s’ajoutent les mécanismes de migration régulière tels que la réunification familiale, les permis de travail et d’étude qui peuvent être accordés préférentiellement à des réfugiés. Toutefois, force est de constater que ces voies d’accès offrent un nombre de places tout à fait insuffisant par rapport au nombre de demandeurs d’asile , qu’elles font appel à des processus de sélection parfois douteux (la récente polémique de la vente des visas humanitaires en Belgique en fournit un parfait exemple) et, surtout, qu’elles semblent être basées sur la discrétion des Etats membres. C’est précisément en raison de cette dernière caractéristique que le débat sur les visas humanitaires présente toute son importance, comme nous le montrent les affaires X et X c. Etat belge et M.N. et autres c. Belgique.
2. Retour sur les deux affaires X et X c. Etat belge et M.N. et autres c. Belgique
L’affaire M.N et autres commence par un jeu de ping-pong incessant entre l’Office des étrangers compétent pour l’examen des demandes de visa, et la juridiction belge compétente en matière d’asile et d’immigration, le Conseil du Contentieux des Etrangers (CCE). Une famille syrienne avait déposé des demandes de visas à validité territoriale limitée auprès de l’ambassade belge à Beyrouth, le 22 août 2016, sur base de l’article 25 §1 a) du code des visas, le but étant de parvenir légalement et en toute sécurité sur le territoire belge en vue d’y demander l’asile. Ces demandes ont été rejetées par l’OE le 13 septembre 2016. Le 7 octobre 2016, le CEE a suspendu en extrême urgence ces décisions de refus, considérant qu’eu égard à la situation sécuritaire en Syrie, le risque de violation de l’article 3 de la CEDH était sérieux. Le CCE a alors ordonné à l’OE de prendre de nouvelles décisions dans les 48h. Toutefois, l’OE a rendu deux nouvelles décisions de refus, les 10 et 17 octobre 2016, motivées de la même façon que la première décision. Elles furent à nouveau suspendues en extrême urgence et, dans son troisième et dernier arrêt, le CCE a finalement ordonné à l’administration de délivrer dans les 48h un laissez-passer ou un visa aux requérants. L’administration belge a toutefois continué à refuser d’exécuter les arrêts du CCE de sorte que le Tribunal de première instance a ordonné à l’Etat belge de délivrer un visa ou un laissez-passer dans les 48h sous peine d’astreinte. Après avoir été confirmée, l’exécution de ces astreintes a été suspendue par la Cour d’Appel (pour commentaire, voy. ici). Les requérants ont finalement décidé d’introduire un recours devant la CEDH le 10 janvier 2018.
C’est pour sortir de ce cercle vicieux nourrissant une polémique médiatisée que le CCE a saisi la CJUE de questions préjudicielles, dans une affaire similaire qui a donné lieu à l’arrêt X et X c. Etat belge. Pareillement, une famille syrienne avait déposé des demandes de visas sur base de l’article 25, §1, a) du code visa, auprès de l’ambassade belge à Beyrouth, le 12 octobre 2016, afin de demander l’asile en Belgique. Le 18 octobre 2016, l’OE a rejeté ces demandes et les requérants ont introduit un recours en suspension en extrême urgence devant le CCE. Le 8 décembre 2016, le CCE a saisi la CJUE de deux questions préjudicielles. La première se présente comme suit : (1) « ‘Les obligations internationales’ visés à l’article 25, paragraphe 1, sous a), du code des visas, visent-elles l’ensemble des droits garantis par la Charte, dont, en particulier, ceux garantis par l’article 4 et 18, et recouvrent-elles également les obligations auxquelles sont tenus les Etats membres, au regard de la CEDH, et de l’article 33 de la convention de Genève ? » Plus précisément, il s’agissait de savoir si l’article 25 du code des visas doit être interprété en ce sens que l’Etat membre, saisi d’une demande de visa à validité territoriale limitée, est tenu de délivrer le visa s’il y a un risque sérieux de violation des articles 4 ou 18 de la Charte, 3 de la CEDH, ou 33 de la Convention de Genève.
Ces deux affaires remettent ainsi en cause le caractère discrétionnaire des visas asile. Considérer que le respect de leurs obligations internationales, telles que l’interdiction des traitements inhumains et dégradants (articles 4 de la Charte et 3 de la CEDH), impose aux Etats membres de délivrer des visas en cas de motifs sérieux et avérés permettant de penser que le refus exposerait les demandeurs à pareils traitements, fournirait une base légale précise d’accès au territoire européen. Le respect des droits fondamentaux imposerait aux Etats membres une obligation positive de délivrer un visa humanitaire afin de prévenir leur violation, et ce, en excluant toute appréciation discrétionnaire des Etats. C’est ce qu’avait avancé l’avocat général P. Mengozzi dans ses conclusions sur l’affaire X et X. Selon ce dernier, les demandes de visas relèvent bien de l’article 25, §1 du code des visas, peu importe que l’intention soit de demander l’asile par la suite. Par conséquent, en mettant ainsi en œuvre le droit de l’UE, la Belgique est tenue de respecter l’article 4 de la Charte qui lui impose de délivrer les visas dès lors qu’un refus exposerait les requérants à des traitements inhumains ou dégradants (voyez V. Moreno-Lax en ce sens).
3. Le refus de la CJUE de se prononcer sur le fond
La CJUE n’a pas suivi le raisonnement de son avocat général. Elle a décidé que la demande de visa humanitaire introduite auprès d’une ambassade d’un Etat Membre, dans un pays tiers, en vue de demander l’asile une fois arrivé sur le territoire de l’Etat membre ne relève pas du droit de l’UE. La Cour note que le code des visas concerne les visas pour les séjours d’une durée maximale de 90 jours, alors que les demandes des requérants ont été formées dans le but de demander l’asile en Belgique, et donc de se voir délivrer un permis de séjour supérieur à 90 jours. Les demandes ne relèvent donc pas du champ d’application du code des visas. La Cour relève ensuite qu’aucun acte législatif n’a été adopté par l’Union en ce qui concerne les conditions de délivrance, par les Etats membres, de visas ou de titres de séjour de longue durée pour motifs humanitaires. La situation ne relève donc pas du droit de l’Union mais du seul droit national de sorte que les dispositions de la CDFUE ne s’appliquent pas.
Ce faisant, la Cour adopte une approche prudente et formaliste, lui permettant de ne pas prendre avoir à trancher le fond de la question. En décidant que la celle-ci ne relève pas du droit de l’Union, la Cour balaye du revers de la main les arguments tirés de la Charte et manque l’occasion de se prononcer sur son application extra-territoriale. Si certains ont espéré que la CJUE allait saisir l’opportunité de mettre fin aux périls de la traversée de la Méditerranée par des réfugiés, celle-ci en a déçu plus d’un. S’il est vrai que l’affaire portée devant elle comportait de nombreuses implications politiques et auraient pu changer fondamentalement le système européen d’asile, le résultat est maigre d’un point de vue humain. La seule porte entrouverte par la Cour vise une évolution du droit de l’Union quant à l’adoption de règles communes sur la délivrance de visas humanitaires de courte ou de longue durée.
Cette porte pourrait s’ouvrir davantage tant que les discussions relatives aux visas humanitaires se poursuivent au sein du Parlement européen et de la Commission. Le 11 décembre 2018, le Parlement a en effet voté en faveur d’un rapport d’initiative législative appelant la Commission à examiner l’opportunité de déposer, pour mars 2019, une proposition de législation permettant aux personnes en quête d’une protection internationale de demander un visa humanitaire dans une ambassade ou consulat européen. Si elle était suivie, une telle proposition ferait donc entrer les visas humanitaires dans le champ du droit de l’Union, de sorte qu’il faudrait alors trancher la question de savoir si le refus d’un visa équivaut à une violation de l’article 4 de la CDFUE ou 3 de la CEDH. L’affaire M.N. et autres c. Belgique pendante devant la Grande Chambre de la CEDH viendra alimenter ce débat. Reste à savoir quelle sera la contribution de la Cour de Strasbourg.
4. Qu’attendre de l’affaire M. N. et autres c. Belgique?
Le cadre juridique devant la CEDH est évidemment différent de celui devant la CJUE en ce que la question préalable du champ d’application de la CDFUE et de la CEDH se pose de manière distincte. Tandis que la CJUE était appelée à se positionner quant à l’interprétation du code des visas combiné à la Charte, la CEDH doit décider s’il y a eu violation de la CEDH tout en devant se prononcer au préalable sur son champ d’application. Mais la question de fond sous-jacente est la même : les Etats ont-ils l’obligation de délivrer des visas afin de permettre aux requérants d’atteindre leur territoire en toute sécurité pour y demander l’asile, lorsque leur refus risque de soumettre les requérants à des traitements inhumains ou dégradants ? En d’autres termes, est-ce que le respect de l’article 3 de la CEDH impose aux États des obligations positives de délivrer des visas permettant aux demandeurs d’asile d’accéder légalement et en toute sécurité à leur territoire? Dans le cadre de l’affaire M.N. et autres c. Belgique, les requérants se plaignent en effet du refus des autorités belges d’exécuter les mesures ordonnées par le CCE (i.e. délivrer le visa ou le laissez-passer), alléguant avoir été maintenus dans une situation contraire à l’article 3 de la CEDH. En outre, les requérants invoquent la violation des articles 6, §1 (droit à un procès équitable) et 13 (droit à un recours effectif) en ce qu’il leur était impossible de poursuivre l’exécution par la voie judiciaire des mesures ordonnées par le CCE.
La question préalable qui devra être examinée par la Cour est de savoir si la CEDH est applicable aux demandes de visas introduites dans un consulat européen établi dans un pays tiers, aux fins de demander l’asile. Partant, la Cour devra décider si, en statuant sur des demandes de visas aux fins de solliciter l’asile, la Belgique exerçait sa juridiction sur les demandeurs de visa au sens de l’article 1er de la Convention. Cette question présente est similaire à celle qui était posée devant la CJUE en ce qu’il s’agit de déterminer si la CEDH trouve à s’appliquer, tout comme il s’agissait de déterminer si la CDFUE était d’application. Mais les critères d’application sont différents: celui de la “juridiction” d’un côté, et celui de la “mise en oeuvre du droit de l’Union” de l’autre.
Il ressort d’une jurisprudence constante de la CEDH que, si la juridiction est principalement territoriale, elle peut également s’exercer de façon extra-territoriale dans des circonstances exceptionnelles. En bref, ces situations exceptionnelles visent des situations de contrôle effectif d’un Etat sur une zone située en dehors de son territoire (par exemple, en cas d’occupation armée), ainsi que des hypothèses d’autorité ou de contrôle exercé(s) par un agent de l’Etat. On relèvera avec intérêt que la Cour a indiqué dans l’affaire Al-Skeini que la juridiction de l’Etat peut naître des actes des agents diplomatiques ou consulaires présents en territoire étranger, conformément aux règles du droit international, dès lors que ces agents exercent une autorité ou un contrôle sur autrui. Peut-on dire que les autorités consulaires belges exerçent un contrôle ou une forme d’autorité sur les demandeurs de visa?
Pour tenter de répondre à cette question, on peut tenter de comparer l’affaire M.N. et autres avec l’intéressante affaire M. c. Danemark dans laquelle un homme s’était réfugié au sein de l’ambassade danoise de Berlin-Est dans sa tentative de fuir l’Allemagne de l’Est et avait été livré à la police est-allemande par les agents consulaires. La Commission européenne des droits de l’homme avait décidé que M. se trouvait bien sous la juridiction du Danemark. Il est vrai que la situation dans M.N. et autres est différente puisque les requérants sont retournés en Syrie et n’étaient plus physiquement présents dans l’ambassade belge. Est-ce que le fait que les agents consulaires avaient le moyen de protéger les requérants (délivrer le visa ou garder le requérant à l’intérieur du consulat) mais refusaient de le mettre en œuvre est suffisant pour dire qu’il y a juridiction sous la forme d’une autorité/contrôle exercé sur les demandeurs de visa?, S’il est vrai qu’en fait comme en droit un Etat exerce une certaine autorité sur ses locaux consulaires considérés en droit international comme inviolables, est-ce suffisant pour conclure à l’exercice d’une juridiction surtout que, à la différence de l’affaire M. c Danemark, les requérants qui demandent un visa sont, non seulement entrés dans l’ambassade volontairement, mais en sont également sortis de leur plein gré sans y avoir été contraints par les agents consulaires et sans que ceux-ci aient alerté les autorités libanaises (Etat de transit) ou syriennes (Etat d’origine). Encore la liberté des requérant peut-elle être remise en cause dans la mesure où il peut, à juste titre, être avancé que ce sont les circonstances du conflit, l’absence de voies d’accès légales et sûres à la protection internationale et la multiplication des obstacles érigés par les Etats pour combattre l’immigration irrégulière (dont, en premier lieu, l’obligation de disposer d’un visa pour se rendre dans l’UE, ce qui rend impossible l’embarquement dans un moyen de transport en raison des obligations faites aux transporteurs de vérifier, sous peine de sanction, que les personnes disposent bien des documents requis pour entrer dans l’Etat de destination) qui ont contraint les requérants à pénétrer dans l’ambassade pour y demander des visas humanitaires.
A considérer que cette première étape est franchie, et que la Cour estime donc que la CEDH s’applique dans cette affaire, encore devra-t-elle déterminer si le refus de délivrer des visas humanitaires a emporté violation de l’article 3. On voit que si la question de l’applicabilité des instruments de protection des droits humains se pose différemment à Luxembourg et à Strasbourg, la question de fond est quant à elle identique. Il est bien connu que la CEDH a déduit dans la célèbre affaire Soering de l’interdiction mentionnée à l’article 3 un principe de non-refoulement. Ce principe impose d’abord des obligations négatives aux Etats parties de ne pas renvoyer un individu vers un autre Etat où il y a des motifs sérieux de croire qu’il risque d’être soumis à des traitements inhumains ou dégradants. On peut dès lors se demander si ce principe impose aussi aux Etats l’obligation positive de mettre en place des systèmes effectifs permettant de protéger les individus contre le risque de violation de l’article.
En effet, la CEDH répète que l’article 1er, combiné à l’article 3, implique que les Etats doivent adopter des mesures appropriées pour empêcher que les individus se trouvant sous leur juridiction ne soient soumis à des traitements inhumains ou dégradants, même par des acteurs privés (affaire Opuz c. Turkey). La Grande Chambre de la Cour a même réaffirmé, dans l’arrêt El Masri c. L’Ex-République Yougoslave de Macédoine, que la responsabilité de l’Etat peut se trouver engagée lorsque les autorités n’ont pas pris..