EU Immigration and Asylum Law and Policy Droit et Politique de l'Immigration et de l'Asile de l'UE

The ruling of the Court of Justice in Kinsa: A first step towards the decriminalisation of the facilitation of unauthorised entry? 

By Prof. Valsamis Mitsilegas, University of Liverpool, UK

One of the most eagerly awaited judgments by the Court of Justice in recent months has involved the litigation in the case of Kinsa. Its significance lies in the fact that this was the Court’s first opportunity to interpret the scope of EU criminal law on the facilitation of unauthorised entry, as well as the compatibility of the current overcriminalisation of facilitation with fundamental rights. The ruling is also a landmark because of its potential influence on shaping future EU criminal law in the field, particularly in view of the current negotiations on a new facilitation Directive to replace the facilitators’ package (see Mitsilegas, 2024). This commentary offers a rapid reaction to the Court’s ruling on Kinsa, which was delivered on 3 June 2025. The note will begin with outlining the background to the litigation and proceed with analysing the Opinion of the Advocate General and the ruling of the Court of Justice.

The Kinsa Litigation: background 

A game changer in the reform of EU criminal law on facilitation has appeared in the form of a reference for a preliminary ruling to the Court of Justice by the Tribunale di Bologna in Italy. In the Kinshasa reference, (eventually renamed as Kinsa), lodged on 21 July 2023, the referring Court has asked the CJEU whether the criminalisation of the facilitation of unauthorised entry in EU law and in national law is compatible with the Charter, even when the conduct is non-profit in nature, and given that the legal framework does not oblige Member States to exempt humanitarian assistance from criminal liability. The referring Court focused on the principle of proportionality referred to in Article 52(1), read in conjunction with the right to personal liberty and the right to property referred to in Articles 6 and 17, as well as the rights to life and physical integrity referred to in Articles 2 and 3, the right to asylum referred to in Article 18 and respect for family life referred to in Article 7 of the Charter. The reference is welcome in stressing the potential adverse effect of the overcriminalisation of facilitation on a wide range of fundamental rights.

The facts in Kinsa lay bare the shaky normative foundations and adverse effects of overcriminalisation of facilitation of unauthorised entry in EU and Italian law (Mitsilegas, 2024). They involve the prosecution of a Congolese woman arriving at the air border of Bologna for the facilitation of the unauthorised entry of her minor daughter and niece (paras. 1-5). The referring court queried the compatibility of the national legislation, and the underlying EU law, with the Charter. It stated that, under Italian legislation, the offence of facilitating unauthorised entry is by its nature an offence of danger, penalising the conduct solely on the basis of intent to procure unauthorised entry to a foreign national, regardless of motive and without requiring any profit-making intent (para. 8).

The Court added that the offence is “free-form”, in the sense that the offence may be committed in any way by the perpetrator, using any means (para. 9). The criminal penalty also applies to those who have facilitated the unauthorised entry of a foreign national for humanitarian assistance purposes and even if the foreign national is in need (para. 11). The referring court noted that the Italian legislation complied with the facilitators’ package (para. 12) and that in the present cases it was clear that the conduct of the accused objectively corresponded to conduct punishable for the offence provided for in domestic law (para. 22). Yet the referring court questioned the reasonableness of such criminalisation and its compatibility with fundamental rights enshrined in the Charter, noting that the protection of those rights must be taken into account in any balancing exercise which must form the basis of the common immigration policy. The court further noted that in both the EU regulatory framework and the Italian legislation, there is a lack of proportionality in favour of the protection of the interest in controlling migration flows, which also results in an unnecessary sacrifice of fundamental rights (para. 17).

The AG Opinion

The Opinion of AG De La Tour was delivered on 7 November 2024.  The Advocate General did not go as far as challenging the current EU legal framework on criminalisation, but placed national authorities under the duty to scrutinise criminalisation from the prism of the proportionality principle as enshrined in Article 49(3) of the Charter. According to the Advocate General, Article 1(1)(a) of Council Directive 2002/90/EC of 28 November 2002 defining the facilitation of unauthorised entry, transit and residence, must be interpreted as meaning that the act by which a mother, a third-country national, intentionally contributes to the unauthorised entry into the territory of a Member State of her daughter and niece, minors over whom she has custody, by using false identity documents, constitutes an offence. In assessing the question on overcriminalisation, therefore, the Advocate General has not questioned the validity of Article 1(1)(a) of Directive 2002/90 in the light of the principles of the legality and proportionality of criminal offences and penalties, enshrined in Article 49(1) and (3) of the Charter.

However, according to the Advocate General, the principle of the proportionality of criminal offences and penalties, enshrined in Article 49(3) of the Charter, precludes a system which would not allow the national court to balance the interests at stake and to differentiate between the criminalisation of a person acting out of humanity or necessity, in the sole interest of the minors, and that of a person motivated solely by the intent to commit the criminal offence for financial gain. It is for the referring court to carry out a specific examination of the proportionality of national legislation, which imposes on anyone facilitating unauthorised entry into national territory, of a custodial sentence of between two and six years and a financial penalty of EUR 15 000 per person concerned, having regard, in particular, to the possibility of exonerating from criminal liability persons whom are shown to have acted disinterestedly, out of altruism, compassion or solidarity, for humanitarian reasons or because of family ties, or of adapting the system of penalties applicable to them.

The Opinion of the Advocate General was a welcome step in placing national courts under a duty to assess the criminalisation of the facilitation of unauthorised entry in conformity with the principle of proportionality as enshrined in Article 49(3) of the Charter. It was also welcome in that it did not limit the proportionality test to the specific circumstances of the present case, but extended the test to lead to the exoneration from criminal liability of any person who is shown to have acted disinterestedly, out of altruism, compassion or solidarity, for humanitarian reasons or because of family ties. On the other hand, the Advocate General’s approach would perpetuate overcriminalisation and legal uncertainty, by maintaining the validity of the very broadly and vaguely defined offences in the EU facilitators’ package, and subsequently the validity of overbroad national criminal law on facilitation. National courts are only to intervene ex post to exonerate from criminal liability, which already exists in national law. In some Member States, assigning this task to national courts could be seen as a challenge in view of the separation of powers between the legislator and the judiciary. A similar issue has arisen in the Taricco litigation, also involving Italy, by the CJEU requirement for national courts to disapply national law on the statute of limitations (Mitsilegas, 2018). The state of legal uncertainty regarding the criminalisation of the facilitation of unauthorised entry is being perpetuated by the Council in negotiations on the new facilitators’ Directive: the Council’s General Approach adds the requirement for financial gain for criminalisation (Art. 3), but on the other hand leaves Member States free to adopt broader criminalisation in their national law (Preamble, recitals 6 and 6(a)).

The Court’s ruling

The Court of Justice delivered its ruling in Kinsa on 3 June 2025. The Court held, sitting in Grand Chamber, that Article 1(1)(a) of Directive 2002/90, read in the light of Articles 7 and 24 and Article 52(1) of the Charter, must be interpreted as meaning that the conduct of a person who irregularly brings into the territory of a Member State minors who are third-country nationals accompanying him or her and under his or her custody, does not fall within the scope of the general offence of facilitation of unauthorised entry. Additionally, the Court held that those articles preclude national legislation criminalising such conduct (para. 73). 

To reach this conclusion, the Court departed from the approach of both the referring court and the Advocate General. The Court reformulated the questions of the referring court, which addressed the examination of the compatibility of the criminalisation of humanitarian assistance with EU law, to a much narrower scope reflecting the facts of the specific case. The Court focused specifically on whether Article 1(1)(a) of Directive 2002/90, read in the light of Articles 7, 18 and 24 of the Charter, must be interpreted as meaning that the conduct of a person who irregularly brings into the territory of a Member State third-country national minors, over whom he or she exercises actual care, does not fall within the scope of the general offence of facilitation of unauthorised entry and whether those Charter provisions must be interpreted as precluding national legislation criminalising such conduct (para. 39).

Formulating the questions in this manner enabled the Court to avoid examining the validity of Article 1 of Directive 2002/90 or to interpret Article 1(2) of that Directive, which relates to exemption from criminal liability for conduct aimed at providing humanitarian assistance to the person concerned (para. 68). Unlike the Advocate General, who primarily focused on the interpretation of national law, the Court did proceed to re-interpret the scope of elements of EU law within the narrower confines resulting from the reformulation of the questions asked. On the other hand, also unlike the Advocate General, the Court refrained from extending the analysis to humanitarian assistance more broadly, again defining the scope of the questions in a narrower way. 

The Court justified the exclusion of the conduct under examination in this case from the scope of EU criminal law on facilitation of unauthorised entry on the basis of the requirement to respect Article 7 and 24 of the Charter, on the right to private life and best interests of the child respectively (paras. 46-51). The Court held that a contrary interpretation of that provision would entail a particularly serious interference with the right to respect for family life and the rights of the child, enshrined, respectively, in Articles 7 and 24 of the Charter, to such an extent that it would undermine the essence of those fundamental rights, within the meaning of Article 52(1) of the Charter (paras. 52-53). The Court held that a person such as the mother subject to prosecution in the present case, in principle merely assumes an obligation inherent to his or her personal responsibility, based on his or her family relationship with those minors, to secure their protection, well-being and development. The Court concluded that the conduct under examination is, above all, the concrete expression of his or her general responsibility towards those minors (para. 54). In addition to the fundamental rights justification, the Court held that criminalising such conduct is contrary to the objectives of Directive 2002/90 as such conduct does not constitute facilitation of illegal immigration, which that directive seeks to combat, but reflects the individual’s personal responsibility arising from the care they provide to the minors involved (para. 45). 

Importantly, the Court of Justice did not stop in the examination of the compatibility of the criminalisation of facilitation on the basis of the facts of the case with Articles 7 and 24 of the Charter, but proceeded to hold that the interpretation adopted by the Court is also necessary in the light of the right to asylum enshrined in Article 18 of the Charter (paras. 57-58). The Court further cited a number of protective provisions of secondary law in this context, including Article 6 of the facilitation Framework Decision (Geneva Convention – para. 59)  and Article 3(b) of the Schengen Borders Code (Schengen Borders Code – para. 60). Importantly, the Court reiterates its earlier case-law to hold that the right of any third-country national or stateless person to apply for international protection on the territory of a Member State, including at its borders or in its transit zones, must be recognised, even if he or she is staying illegally and irrespective of the prospects of success of such a claim. When such an application is made, an applicant cannot, in principle, be regarded as staying illegally on the territory of the Member State concerned, so long as no decision has been given on that application at first instance, in order to preserve the effectiveness of the right to asylum, as guaranteed by Article 18 of the Charter (Case C-808/18 para. 102; C-821/19 paras. 136 and 137 – para. 61). The Court also emphasises that measures which, without reasonable justification, result in discouraging a third-country national from submitting his or her application for international protection, are liable to undermine the effectiveness of the right to asylum, as guaranteed by Article 18 of the Charter (Case C-808/18 paras. 102, 103, 118 and 119 C-823/21 paras. 47 to 51- para. 62).

In terms of the impact of this interpretation of EU law on national criminal law, the Court held that when transposing Article 1(1)(a) of Directive 2002/90, Member States may not establish, in national law, rules that would go beyond the scope of the general offence of facilitating unauthorised entry, as defined by that provision, by including conduct not covered by it, in breach of Articles 7 and 24 and Article 52(1) of the Charter (para 71). The Court further clarified that Articles 7 and 24 of the Charter are self-sufficient and do not require further specification in EU or national law to grant individuals enforceable rights. Consequently, if the referring court were to find that it is not possible to interpret its national law in conformity with EU law, it would be required to ensure the judicial protection of the rights conferred on individuals by the Charter and to guarantee their full effectiveness by disapplying, if need be, Article 12 of the Consolidated Law on Immigration (para. 72). 

For those who view the glass as half empty, the Court’s ruling in Kinsa is a missed opportunity to examine the compatibility of the criminalisation of humanitarian assistance with fundamental rights. The narrow focus of the Court to the facts of the case perpetuates legal uncertainty regarding the position of those providing humanitarian assistance to migrants, including civil society organisations subject to a sustained establishment of a hostile environment. This means that the impact of the ruling on the current negotiations of a new Directive on facilitation may be more limited than hoped for.

On the other hand, those who view the glass as half full, will hail Kinsa as a significant step forward: the Court relied expressly on an array of fundamental rights and Article 52(1) of the Charter to interpret EU criminal law on facilitation, leading to a degree of decriminalisation at EU and at national level. In addition to family life and the rights of children, the Court examined specifically the impact of criminalisation on the right to asylum, in a significant finding that irregular entry leading to the submission of an asylum application, and until the issuance of a first instance decision, is not to be treated as a criminal offence. This finding, along with the Court’s express reliance on international law not only in terms of refugee law, but also in terms of the Palermo Convention (para. 66), can open the door towards the decriminalisation of humanitarian assistance, if the latter is linked to enabling the exercise of the right to asylum (see Mitsilegas, 2024).

Polish law restricting the right to asylum at borders

By Barbara Mikołajczyk, University of Silesia

Since July 2021, in response to EU sanctions imposed on Belarus following the rigged 2020 presidential election, the Belarusian authorities have launched a campaign to destabilise Poland by engaging in instrumentalisation of migrants at the border with Poland. The Polish Border Guard provides evidence that Belarusian officers are organising mass border crossings by migrants at the border, and that they are encouraging or facilitating the migrants’ use of violence against Polish officers. The Polish Border Guard staff are coming under attack from improvised weapons (sharpened sticks, broken glass, stones, slingshots and sometimes knives). The Belarusian forces have also been found to have engaged in an operation to provide migrants with equipment to destroy parts of a dam that was built in 2022. In May 2024, a soldier was stabbed in the chest and killed. Due to the growing tension at the border, in  June 2024, a buffer zone was introduced on the Podlasie section of the Polish-Belarusian border to limit access to the border. Prolonged for the third time in March 2025, this zone now covers additional areas and aims to limit the presence of people in the border strip. In turn, zones up to four kilometres wide have been introduced in some sections to prevent illegal crossings.

From July 2021 to November 2024, the Polish Border Guard recorded 110.595 attempts to cross the Polish-Belarusian border illegally according to the rationale for the act introducing the possibility of suspending the right to asylum. All these events are taking place in the context of Russia’s invasion of Ukraine and constant threats from Russia towards Poland. The need to take decisive steps to protect the border is therefore indisputable, especially since a visa-free travel agreement between Russia and Belarus came into force in January 2025. 

This  situation is a blatant example of the instrumentalisation of migration, which is a challenge not only for Poland but also for the Baltic States, Finland, and also recently Greece. Iris Goldner Lang has discussed on this blog (see here) the changes in the EU that led to the regulation on Crisis and Force Majeure Regulation 2024/1359  which broadly defines the situation of instrumentalisation as a type of crisis situation “in which a third country or hostile non-state actor encourages or facilitates the movement of third-country nationals to the external borders or into a Member State, with the aim of destabilising the Union or its Member States, where such activities could jeopardise the fundamental functions of a Member State, including the maintenance of law and order or the safeguarding of its national security” (Article 1(4)(b)). 

It is well known that migrants arriving from Belarus have faced push-back practices by Poland that began long before the summer of 2021 as documented by the reports and views of human rights bodies (UN Human Rights Committee [2016], [2023] and the UN Committee against Torture [2019]), as well as judgments of the European Court of Human Rights, including in the cases of M.K and others v. Poland and  D.A. and others v. Poland.  The situation at the EU–Belorussian border should also be considered a humanitarian crisis and human drama (see here). According to the Helsinki Foundation for Human Rights analysis devoted to mapping deaths at the border, in most cases migrants are forced to return to Belarus without their identity being verified and without the legally prescribed administrative procedures being carried out by Polish border guards. They are returned through service gates in the dam as well as at swamps, and border rivers. As a result, many people require urgent medical attention from fractures and injuries, dehydration, hypothermia and exhaustion as well as the exacerbation of symptoms of existing chronic diseases. This analysis provides that, from 2021 to the end of March 2024, 116 people have been documented as dying on the European Union’s eastern border in four countries (Belarus, Latvia, Lithuania and Poland).

The most well-known situation is that of a group (initially 60 people, later around 30) of Middle Eastern migrants (Afghans and Iraqis) who were stucked on the Belarusian side of the border from August to the end of October 2021.  The case R.A. and others v. Poland ended up before the ECtHR. Currently, it has come before the Great Chamber. The circumstances of this case are different from those in M.K. and D.A mentioned above concerning the refusal to accept applications at border crossings, because it concerns applications in the context of an attempt to cross the border outside border crossing points. It is problematic whether in the R.A. case the complainants had a real chance to apply at the border crossing or another “authorised” location. Still, formally, the border crossings were not closed at that time and the right to apply for international protection was not suspended. The judgment in this pending case and similar cases –  H.M.M. and others v. Latvia and C.O.C.G. and others v. Lithuania – will determine the standard of protection for migrants in a situation of instrumentalisation of migration. 

The situation at the Polish-Belorussian border prompted several legal changes in Poland, including a highly contentious amendment to the Act on Granting Protection to Foreigners. On 14 October 2021, despite widespread criticism from human rights institutions such as UNHCR, OHDIR and various NGOs, the act was revised. The amendment allowed the Head of the Border Guard Office to dismiss applications for international protection submitted by foreigners apprehended immediately after unlawfully crossing the EU external border. However, exceptions were made for individuals who arrived directly from territories where their life or freedom was at risk due to the threat of persecution or serious harm, provided they presented credible reasons for entering the Republic of Poland illegally and applied for international protection immediately after crossing the border. However, it should be noted that in Polish, “immediately” is a relatively vague concept and may not denote instant action. Consequently, this interpretation allows for scenarios where a foreigner might be apprehended after spending several days in the forest and covering significant distances from the border.

It was expected that, after the change of power following the Polish parliamentary elections of October 2023, these regulations would be repealed by the new governement as they were assessed as incompatible with the principle of non-refoulement, other international human rights obligations and Article 56 of the Constitution of the Republic of Poland which proclaims that “foreigners who, in the Republic of Poland, seek protection from persecution, may be granted the status of a refugee in accordance with international agreements to which the Republic of Poland is a party.” This happened, though not necessarily in the way that the human rights institutions and NGOs might have expected. The new law of 21 February 2025 (discussed below), introducing, among other things, the legal basis for temporarily restricting the right to apply for international protection, admittedly repeals Article 33.a [1]. Additionally, in the context of the instrumentalisation of migration, the adopted reforms seem to legalise pushbacks, which have already been or are still considered by the ECtHR.

The new Polish law

The new law of 21 February 2025 amending the Act on Granting Protection to Foreigners establishes a legal framework for temporarily restricting the right to apply for international protection. New provisions were adopted despite protests from the UNHCR, the Polish Bar Council, the Helsinki Foundation for Human Rights, the Ombudsman and the Center for Research, Studies and Legislation of the National Council of Legal Advisers. As a result of the consultation, only the provision preventing the submission of asylum applications on behalf of spouses and children were removed from the final version of the act. Even after the bill was passed in Parliament, a consortium of NGOs appealed to the Polish President to veto the new law, but he signed the amendment on 26 March 2025, and it came into force on 27 March 2025.

In this context of tense situation at the border with Belarus, introducing a legal basis for temporarily restricting the right to apply for international protection is the focus. The Council of Ministers can impose such restrictions for a period not exceeding 60 days. If the reasons for the restrictions have not ceased, they may be extended for another fixed period of up to 60 days if the Sejm (lower chamber of Parliament) consents to the extension. It is possible to lift the temporary suspension of the right to asylum earlier, either on the entire border or part of it if the situation improves.

First of all, the amendment introduces the notion of “instrumentalisation”, which includes actions by a country bordering Poland, or another entity, enabling foreigners to cross the external border in violation of the law. This applies in particular to the use of violence against service officers and soldiers of the Polish Armed Forces protecting the border, as well as the destruction of border infrastructure which may result in the destabilisation of the internal situation in Poland.

Crucial is a new Article 33a.[1] stipulating that “The right to apply for international protection may be temporarily restricted where:

1)     instrumentalisation takes place, and

2)     instrumentality activities pose a serious and real threat to the security of the state or society and

3)     the introduction of this restriction is necessary to eliminate the threat referred to in point 2, and other measures are not sufficient to eliminate it”. 

Exceptions to the temporary restriction on the right to apply for international protection are stipulated  in Article 33b, providing that “an application for international protection is accepted by the Border Guard authority from a foreigner who is:

1) an unaccompanied minor;

2) a pregnant woman;

3) a person who may require special treatment, in particular due to their age or state of health;

4) a person who, in the opinion of the Border Guard authority, is clearly at risk of serious harm in the country from which they have directly arrived in Poland”;

5) a citizen of the instrumentalising state from whose territory foreigners enter Poland.

However, the application will not be accepted if foreigners belonging to the above categories have used direct coercive measures and weapons or other armaments immediately after they crossed or attempted to cross the border by force and in cooperation with others.

More questions than answers

The day after the president signed the new law,  the Council of Ministers, in its regulation of 27 March 2025, implemented restrictions for 60 days along the whole border with Belarus, justifying the immediate implementation by the sharp increase in the number of attempts of illegal crossings. The Ministry of Interior and Administration reported that from 1 January 2025 to 20 March 2025, 2.227 such attempts had been recorded with 1.383 cases in March alone. This corresponds to the upward trends in spring 2023-2024, so a further significant increase is likely in the coming months. 

Border control and non-refoulement: are they contradictory?

Given the dangers emanating from Belarus and Russia, there is no doubt that the state has the right to strengthen border controls, especially in a situation of instrumentalisation of migration, aggressive behaviour on the part of migrants and influxes in large groups and attempts to cross the border outside of border crossings. 

In a communication dated 11 December 2024, the European Commission confirmed that Member States with external borders with Russia and Belarus should be able to act in a firm and decisive manner to control their borders effectively. They are responsible for maintaining law and order, safeguarding internal security in their national territories, and effectively controlling and managing the external borders as required by EU law. 

At the same time, the Commission continues to recognise that the right to asylum, respect for human dignity and the principle of non-refoulment are important foundations. However, it is also assumed that international law and EU law allow Member States to resolutely resolve situations where it is necessary to maintain security and public order while respecting the rights of all persons affected by these measures.  The Commission points to the significance of the Court of Justice in assessing the legitimacy and proportionality of the measures taken. Future case law in this area will reveal how flexible EU law is. As Daniel Thym notes on this blog, EU law, including Article 72 TFEU and Article 18 of the Charter of Fundamental Rights, may leave a certain margin for interpretation.    

The European Court of Human Rights also provides some guidance on the measures that can be taken at borders in crises, especially in the cases of N.D. and N.T. v. Spain (see here a comment on this blog) and A.A. and others v. North Macedonia. In the latter judgment of 5 April 2022, the Court ruled that the European Convention on Human Rights does not prevent states from requiring that requests for such protection be submitted at existing border crossings in fulfilling their obligation to control borders. Consequently, they may refuse entry to foreign nationals, including potential applicants for protection who without justified reasons  have not complied with these provisions by attempting to cross the border at another location, in particular by taking advantage of their large number (paragraph 115 of the judgment). Thus, the ECtHR confirms the state’s right to shape the way the border is protected, but on the assumption that there are real chances to submit an application for protection at a designated place, in particular when it comes to respecting Article 3 of the Convention. The exercise of necessary and effective border control cannot therefore be equated with a derogation from the prohibition of refoulement.

Which discretion for border guards?

The introduced restrictions mean that border guards on the Polish-Belarusian border will not accept applications for international protection unless they come from someone who belongs to one of the groups mentioned above. It doesn’t matter if the foreigner tries to apply at the border crossing or outside it. Thus, the solutions adopted not only suspend the right to seek asylum, the inherent principle of non-refoulement and the prohibition of collective expulsion of foreigners, but may also undermine non-derogable rights like freedom from torture, the right to life, the prohibition of slavery and of inhumane and degrading treatment. Any limitations on other rights (e.g. freedom of movement) should be proportional and necessary to achieve the goal (here preventing destabilisation). They therefore require strict precision. However, the adopted regulations here are vague and laconic and leave border guards with a great deal of discretion.

Even the term used in the Council of Ministers regulation “on the border with Belarus” is unclear. Does it refer to the borderline or the buffer zone, which was introduced in 2024 and extends up to four kilometres, or perhaps it is another undefined area where a foreigner was apprehended but had previously managed to cross the border illegally? The latter approach would correspond to the solution adopted in the repealed provision introduced in 2021.

The introduced provisions do not indicate what “restriction to apply for international protection” means formally. Does it mean that applications are not accepted as if migrants were invisible? Do they receive a decision of refusal to accept the application? Can they appeal, and to whom? The new regulations exclude the activity of the Head of the Office for Foreigners in this regard, which would normally consider applications for international protection and issue decisions granting or denying it.

The scope of individuals listed in Article 33 b from whom border guards are required to accept applications for protection, raises also many questions. After all, the explicitly mentioned “unaccompanied minor” and “pregnant woman” fall under the premise of “a person who may require special treatment, in particular, due to their age or state of health,” so why are these individuals explicitly highlighted and not others? Why, for example, are parents with young children not designated either? What about other vulnerable migrants? How will early pregnancy be verified? Do border services have the ability to verify a woman’s declaration of pregnancy under dignified conditions? In practice, will a certificate, whether genuine or falsified, of pregnancy be worth more at the border than a passport or evidence of persecution in the country of origin? Another question can be posed in the context of a minor who, due to travel disruptions, has arrived unaccompanied while their parents/caregivers or other guardians show up at the border a bit later. Will the current regulations lead to the separation of the family? Such a situation would violate Article 3 (the principle of the best interests of the child), Article 9 (a child and parents cannot be separated against their will), and Article 10 (the right to family reunification) of the Convention on the Rights of the Child.

Reading into Article 33b 3) and 4) of the law of 21 February 2025, it may be assumed that they do not prevent border guards from accepting applications from many categories of persons with special needs, including victims of torture or those at risk of torture from the Belarusian state and non-state actors involved in instrumentalisation. Thus, foreigners arriving from Belarus are “at risk of serious harm in the country from which they have directly arrived in Poland.” Here, it is worth adding that Polish courts in their judgments on the situation on the border with Belarus have recognised that this state does not guarantee respect for fundamental human rights, both for foreigners and its citizens, and there is evidence of brutality by Belarusian services and treatment of migrants in an inhumane or degrading manner, which may in some cases be considered torture (e.g. Supreme Court [2023] and Voivodship Administrative Court in Warsaw [2022a] and [2022b]). Thus, it seems that nothing stands in the way of an interpretation obliging the acceptance of an application not only from a Belarus citizen fleeing persecution by the authorities of their “instrumentalising state”, but also from citizens of other countries.    

Finally, also designated in Article 33 b, vulnerable individuals may not be allowed to apply if it is necessary to use force against them when they cross or attempt to cross the border. However, there are no provisions to verify the expediency and proportionality of the measures used by officers. Such provisionssubject to interpretation and the lack of procedures enshrined in the law may also appear challenging for services operating at the border.

In such a situation, however, procedures and minimum guarantees are needed, at least those very modest ones provided for in the Crisis and Force Majeure Regulation 2024/1359 (Article 11 §§6 to 10). Incidentally, it is worth noting that “instrumentalisation”, as defined in Polish law, differs from  the quoted above definition in Regulation 2024/1359. The Polish definition also includes the aggressive behaviour of migrants which is a response to the current situation at the border. Still, it may cause interpretation difficulties when the EU regulation takes effect on 1 July 2026.

The new Polish  law have been compared with the Finnish Act on Temporary Measures to Combat Instrumentalised Migration of 17 July 2024. This law stipulates in §5 that an application for international protection shall be accepted if, in the opinion of a border guard trained and instructed for the task, it is necessary to protect the rights of a child, a disabled person or another particularly vulnerable person.  Moreover, an application for international protection will also be accepted if the person has presented or there are circumstances which, according to a case-by-case assessment by a border guard who has been trained and instructed for the task, make it obvious that the person is in real danger of being subjected to death penalty, torture or other that violates human dignity, primarily in the country from which they have arrived in Finland. The case-by-case assessment can be waived only in exceptional situations where there is an entry using force or a large numbers of people. 

Compared to the Finnish model, the Polish solution is much stricter as, even when there is no use of force or a large group of migrants, it does not guarantee a case-by-case assessment by specially trained officers. The Polish model could be deemed the opposite of the Finnish as it only involves case-by-case examinations in exceptional cases. The Polish solution loses track of people who have probably been deceived by Belarusian authorities, are not aggressive and do not arrive at the border crossing point in large groups.  

How temporary is the temporary restriction?

Without procedures and at least minimal guarantees, a “restriction” or “suspension” may, in fact, become an “exclusion” of the possibility of applying for asylum for an indefinite period. After all, even the duration of the restriction although clearly set at 60 days, is not certain, since there is always the possibility of it being extended with the approval of the Parliament, and it is not specified how many times an extension can be requested. 

On 21 May 2025, the Sejm, by a majority vote of the ruling party and the opposition, agreed with the Council of Ministers that the reasons for introducing restrictions on the right to apply for international protection had not ceased to exist as there was still instrumentalisation posing a serious and real threat to the security of the state or society. Thus, maintaining restrictions for the next 60 days is necessary to prevent this threat, as the measures taken so far have proven insufficient to eliminate it.

In a letter to the Speaker of the Sejm, the Helsinki Foundation for Human Rights points to the groundlessness of extending the restrictions, stating that migration trends on the Polish-Belarusian border are constantly changing and that monthly statistics on attempts to cross the border vary, regardless of the practices of the Border Guard or the laws in force in Poland.

Conclusion 

The situation at the border is becoming increasingly tense. Belarusian officers have even engaged in direct provocations and aggressive actions close to the border (see here). Poland, like other countries, has the right to strengthen border protection. However, the nature of the measures introduced, particularly preventing the vast majority of foreigners from applying for asylum at the border crossing, means that the measures must be considered excessive and disproportionate to the objective pursued. They therefore violate human rights.

The adopted solution contradicts a wide array of international and EU laws, including Article 3 ECHR, Article 4 of Protocol No.4 ECHR, Article 18 and 19  of the EU Charter of Fundamental Rights, secondary EU law including the Crisis and Force Majeure Regulation 2024/1359 (Article 11 §10), Article 3 of the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, and Article 33 of the Geneva Convention Relating to Refugee Status. The new legislation also contradicts Article 56 [2]  of the Constitution of the Republic of Poland, which is provides a guarantee of the right to apply for international protection as well as Article 31 [3] which states that limitations upon the exercise of constitutional freedoms and rights cannot violate the essence of freedoms and rights.

Poland’s response to the instrumentalisation of migration seems to be an example of an extreme solution illustrating the conflict between national security and the right to asylum. Even after the restrictions will be withdrawn one day, the right to asylum will emerge more damaged from this conflict. As Daniel Thym pointed out in this blog and in his article, “the time of dynamic interpretation in favour of migrants may have come to an end.”

Towards Temporary Prolonged Protection?  

Dr Meltem Ineli Ciger, Associate Professor of International Law, Suleyman Demirel University 

The EU’s activation of the Temporary Protection Directive (TPD) in March 2022 marked a historic response to the mass displacement caused by Russia’s invasion of Ukraine. The TPD was formally activated by Council Implementing Decision (EU) 2022/382 on 4 March 2022. It first applied for an initial period of one year, until 4 March 2023, and was then automatically extended for one additional year until 4 March 2024. Council Implementing Decision (EU) 2023/2409 extended temporary protection until 4 March 2025. Finally, Council Implementing Decision (EU) 2024/1836 extended temporary protection for a further period of 1 year, until 4 March 2026. As of March 2025, around 4.2 million people fleeing Ukraine continue to benefit from temporary protection across the EU (Eurostat 2025). Although the Directive was conceived as a short- to medium-term emergency mechanism, the temporary protection status is currently set to expire in March 2026.

In anticipation of this deadline, the European Commission adopted a package of three interlinked documents to shape the EU’s future approach on 4 June 2025. First, it proposed a Council Implementing Decision to extend temporary protection for one additional year, until 4 March 2027. Second, it issued a Proposal for a Council Recommendation outlining a coordinated strategy to phase out temporary protection, with a focus on long-term solutions such as transitions to residence status and voluntary returns. Third, a Communication titled “A Predictable and Common European Way Forward for Ukrainians in the EU” set out the strategic framework underpinning the two proposals. On 13 June 2025, Member States unanimously endorsed the Commission’s proposal to extend temporary protection until 4 March 2027.

This post analyses the European Commission’s June 2025 proposals on the future of the Temporary Protection Directive, focusing on their legal basis, policy rationale, and transition strategy. It questions the compatibility of the proposed extension with the Directive’s time limits, critiques the Commission’s reading of recent CJEU case law, and highlights the lack of a harmonised and inclusive pathway beyond temporary protection.

Is 5 years still temporary? 

The application of the TPD has been effective in delivering swift protection and alleviating pressure on national asylum systems. It also affirmed displaced persons’ agency by exempting them from the Dublin rules and allowing freedom of choice regarding the host Member State. However, as its name suggests, temporary protection must remain temporary.

Article 4 of the Directive sets an apparently clear temporal limit: temporary protection must not exceed three years. The Commission’s proposal to extend the regime to March 2027 would exceed this maximum, raising serious concerns about the legal soundness and credibility of the Temporary Protection Directive as an emergency instrument. 

According to the Commission, the proposal to extend the protection until March 2027 “is fully consistent with the EU asylum acquis, as the Temporary Protection Directive forms an integral part of the Common European Asylum System.”  The Commission justifies the extension on four grounds: ongoing volatility in Ukraine; the need to safeguard those still requiring protection; the risk of overwhelming asylum systems if temporary protection ends abruptly; and the persistence of the original conditions that triggered the Directive’s activation in 2022. While these reasons are compelling from a policy perspective, they do not resolve the legal inconsistency with Article 4. 

As I previously argued, the Temporary Protection Directive including Article 4 and, in particular, Article 4(2), which the Commission continues to invoke as the basis for extending temporary protection to five years, does not provide a legal basis for prolonging protection beyond the three-year limit without a formal amendment to the Directive.  A further concern is the absence of clarity on whether March 2027 will represent the final expiration of the regime. The Proposal for a Council Recommendation relies on vague and open-ended language, noting that “at some future point in time,” when conditions permit, temporary protection “will need to come to an end.” This formulation provides no certainty or concrete benchmarks and leaves the door open to indefinite extensions, contrary to both the spirit and letter of the Directive.

While Steve Peers wryly observes that “with a bit of legal alchemy, the EU has cruised past” the generally accepted three-year time limit of the Temporary Protection Directive, I take issue with the EU’s legal alchemy. In my view, permitting the continuation of temporary protection without amending the Directive risks setting a dangerous precedent. Repeatedly extending the Directive on an annual basis, in the absence of clear legal or temporal constraints, opens the door to its indefinite application. This not only stretches the original intent of the Temporary Protection Directive but also undermines the fundamental principles of proportionality and necessity enshrined in both EU and international law. A protection regime initially designed as an exceptional, time-bound response must not become a de facto permanent fixture without any appropriate legislative reform and legal justification. Moreover, when five years marks the point at which third-country nationals can obtain long-term residence in the EU, how can a regime lasting just as long or longer still be credibly labelled ‘temporary’?

A restrictive interpretation of Krasiliva judgment (C-753/23)

The Commission also appears to adopt a narrow and arguably restrictive interpretation of the Court of Justice’s recent Krasiliva judgment (C-753/23). In both the explanatory memorandum and paragraph 6 of the Recommendation, the Commission advises that Member States should refrain from granting residence permits or social assistance to individuals already enjoying temporary protection in another EU country to avoid overlapping entitlements.

While this recommendation is framed as a safeguard for the system’s integrity, it arguably overstates the scope of this ruling. The Court did not prohibit a second Member State from issuing a residence permit to an individual who was holding temporary protection status in a first Member State. Instead, it addressed a narrow factual scenario in which an applicant had applied but not yet been granted protection in the first Member State. The CJEU ruled that Article 8(1) of the TPD does not allow national legislation to deny a permit solely because an application was made in another Member State (para 33). The Court also ruled that a Member State must allow appeals against such rejection decisions. However, the judgment does not clarify whether a person already granted temporary protection in one Member State can obtain the same status in another. Thus, the judgment does not mandate uniform refusal. The Commission’s reading fills a legal lacuna left by the Court but does so by advancing a restrictive approach that may limit Member State discretion more than the ruling itself requires. 

Ambiguous path beyond temporary protection

The Commission’s Proposal for a Council Recommendation sets out a coordinated transition strategy that encourages early movement towards national residence statuses. These include permits for employment, education, research, and family reasons, as well as statuses available under EU legal migration instruments such as the Blue Card, Single Permit, and Students and Researchers Directives. Member States are expected to inform displaced persons of the legal implications of such transitions and clarify that these statuses cannot be held concurrently with temporary protection.

Simultaneously, the Recommendation envisions a pathway for return built around voluntary return programmes once temporary protection ends, exploratory visits to Ukraine, and reintegration assistance coordinated with Ukrainian authorities. Specific provisions are foreseen for vulnerable individuals and families with school-aged children, ensuring continued residence under Article 23 of the TPD where needed. The proposed Unity Hubs, multi-purpose information centres supported by the Asylum, Migration and Integration Fund (AMIF), are tasked with providing displaced persons with guidance on legal options and return possibilities. 

A notable element of the Recommendation, in my view, is its implicit discouragement of asylum applications by persons currently benefiting from temporary protection. The Commission recommends that Member States emphasise the advantages of transitioning to national residence statuses “in comparison to temporary protection and international protection” (para 4). While this may serve to prevent asylum systems from being overwhelmed, it raises questions about access to international protection for individuals beyond a temporary period.

What should be the right exit strategy? Some reflections on the Commission Proposals 

First, given the EU institutions’ prolonged silence on post-temporary protection planning, the Commission’s initiative is a welcome step toward providing some direction on what should follow the eventual termination of the regime. Having said that, none of the reviewed proposals set out a unified way for the future of temporary protection beneficiaries. As Peers rightly notes, “the most obvious route to a common policy would have been to propose some form of harmonised EU-wide post-temporary protection status.” Instead, the Commission outlines two main trajectories: (1) integration via national pathways, including residence permits available under domestic law or through existing EU legal migration instruments, and (2) assisted voluntary return for those wishing to go back to Ukraine. 

Second, from the outset, I have argued that the Long-Term Residents (LTR) Directive could serve as a structured and rights-based pathway to permanent residence for those granted temporary protection. The LTR Directive allows third-country nationals to apply for long-term resident status after five years of legal residence. However, it remains unclear whether time spent under temporary protection counts towards this threshold. Moreover, temporary protection beneficiaries are currently excluded from applying. This exclusion constitutes a significant legal and policy gap. 

Unless the LTR Directive is amended, many individuals who may have satisfied core integration criteria may be left with fewer durable residence options after 2027. Amending the LTR Directive should include: (a) counting time under temporary protection toward the five-year threshold; (b) extending eligibility to temporary protection beneficiaries; (c) adjusting income and insurance requirements to acquire the long-term residence status to reflect socio-economic realities; and (d) incorporating gender-sensitive indicators such as caregiving roles and community participation. It is also important to recognise that while transitioning to LTR  status may offer a durable solution for some, it will not be accessible to all. Women with caregiving responsibilities, individuals engaged in informal or precarious employment, older persons, and survivors of trauma often struggle to meet the standard eligibility criteria, such as stable income, health insurance, and continuous legal residence, thus risking exclusion from this pathway. Though it is possible to amend the Directive, to incorporate inclusive and gender-sensitive criteria, recognising informal care work, community participation, and non-standard employment as valid indicators of integration. Notably, the Commission’s proposals remain silent on the Long-Term Residents Directive. This omission may reflect the stalled status of its recast proposal, which would have allowed time spent under temporary protection to count towards the five-year threshold for long-term residence. (cf. LTR Recast Proposal)

Third, while the Commission encourages Member States to transition temporary protection beneficiaries to national migration statuses, this approach risks deepening legal fragmentation across the Union. In the absence of a harmonised exit strategy, Member States have been adopting divergent pathways (cf. ICMPD Paper). As Wagner and Grama observe, several states have introduced regularisation mechanisms: for example, Poland’s 2025 legislation allows those with one uninterrupted year of temporary protection to apply for a three-year residence permit. Italy permits transitions to standard work-related permits, while Estonia and Latvia allow in-country applications for alternative statuses without requiring exit and re-entry. While these national initiatives may appear pragmatic, they undermine the original purpose of harmonised protection standards, increase the likelihood of unequal treatment, and heighten the risk of secondary movements and protection gaps. Without stronger coordination or guidance from EU institutions, this may lead to “transition shopping,” (as one of the speakers in a recent webinar aptly framed it), where temporary protection beneficiaries seek out Member States offering the most favourable legal frameworks, potentially undermining the principles of solidarity and leading to secondary movements. 

Fourth, the Commission’s focus on the transition to residence permits under national law or the EU acquis overlooks the situation of vulnerable individuals who may be unable to meet the required eligibility criteria, such as stable employment or self-sufficiency. The proposal offers no substantive guidance on how these individuals will continue to receive protection and assistance within the Union, risking their exclusion from future residence frameworks.

Fifth, for individuals who do not meet the eligibility criteria for residence permits under national laws or EU instruments, including those in particularly vulnerable situations, international protection remains one of the few viable and rights-based avenues for continued protection within the Union. However, when read in their entirety, the Commission’s proposals appear to encourage Member States to dissuade temporary protection beneficiaries from entering the asylum system. This is a matter of serious concern. The Council’s approach, unfortunately, mirrors the Commission’s approach on this: Council of Europe’s press release dated 13 June 2025 notes:

“The temporary protection system alleviates the pressure on national asylum systems because persons falling under this type of protection do not need to apply for asylum on an individual basis. The extension until March 2027 is a signal to displaced people from Ukraine that it is not necessary to apply for asylum.”

The right to asylum is a fundamental right guaranteed under EU law, notably enshrined in Article 18 of the Charter of Fundamental Rights of the European Union. Furthermore, it must be recalled that Article 17 of the Temporary Protection Directive explicitly provides that individuals enjoying temporary protection must be able to lodge an asylum application at any time. Any approach that discourages access to the asylum procedure risks violating EU law and the broader principles of international refugee protection.

Sixth, while the Commission rightly highlights the need to prevent national asylum systems from becoming overwhelmed, there are lawful and pragmatic ways to achieve this without restricting or discouraging access to asylum. One such approach is the adoption of group-based or prima facie recognition of international protection status. Under this model, Member States could grant refugee or subsidiary protection to categories of displaced persons based on common characteristics or shared circumstances, thereby bypassing the need for full individual status determination in every case. Although the forthcoming Qualification Regulation—set to apply in 2026—retains the requirement for individual assessments, it allows for flexibility in implementation. Member States retain discretion to use accelerated procedures for applicants whose claims are likely to be well-founded (see Tan & Ineli-Ciger, 2024). 

In the context of ongoing armed conflict in Ukraine and the risks posed by a sudden termination of temporary protection, group-based recognition could significantly ease pressure on asylum administrations while preserving procedural fairness and fundamental rights. This approach would be particularly relevant for countries hosting large Ukrainian populations, such as Germany and Poland, as it would facilitate a more efficient and rights-based transition to durable protection statuses while preserving the possibility of applying cessation clauses when protection needs no longer exist.

Eligibility criteria could preferably be formulated to include all individuals currently benefiting from temporary protection, recognising them as eligible for international protection. If such a broad approach proves unfeasible, the criteria could instead be narrowly tailored to reflect specific vulnerabilities such as region of origin (particularly for those from non-safe oblasts), special protection needs, and the absence of meaningful internal protection alternatives. For example, displaced persons from areas experiencing active hostilities or those facing specific risks could be granted international protection through streamlined procedures that maintain essential safeguards. 

Conclusion

The Commission’s June 2025 proposals attempt to balance continued protection for displaced Ukrainians with the need to plan a structured transition beyond the Temporary Protection Directive. However, they fall short of providing legal certainty, coordinated long-term solutions, or adequate safeguards for the most vulnerable. The extension of temporary protection to March 2027 exceeds the Directive’s maximum time limit without offering either a sound legal basis or an unequivocal commitment that this will be the final prolongation. Furthermore, the overreliance on national migration statuses coupled with the underutilisation of international protection as a viable pathway, the absence of amendments to the Long-Term Residents Directive, and the lack of a harmonised post-temporary protection status risk entrenching fragmentation and perpetuating legal uncertainty across the Union.

The Council’s decision to extend temporary protection for five years or maybe more will also establish a precedent for how the EU responds to protracted displacement and operationalises its temporary protection frameworks in future crises. It is clear that the EU needs a better exit strategy than what the Commission proposed. The Temporary Protection Directive was designed to unify Member States in times of crisis. Its final chapter must reflect that original spirit by fostering solidarity, coherence, and legal certainty.

Une meilleure protection de groupe pour les femmes-réfugiées

Par Jean-Yves Carlier et Eleonora Frasca, Université catholique de Louvain (UCLouvain), membres de l’équipe droits et migrations (EDEM).

 

Ce billet est une version révisée d’un extrait de la chronique de jurisprudence annuelle « Droit
européen des migrations » publiée dans le Journal de droit européen, n° 3, mars 2025.

En 2024, par trois arrêts, dont deux de grande chambre, la Cour de justice de l’Union européenne a tranché diverses questions relatives à la protection des femmes en qualité de réfugiées : les craintes de persécutions résultant de violences domestiques (W.S., C-621/21) ou de l’identification aux valeurs de l’égalité des sexes dans le pays d’accueil (K.L., C-646/21) ainsi que le constat d’appartenance de l’ensemble des femmes d’Afghanistan à un groupe social craignant avec raison des persécutions (aff. jointes AH et FN, C‑608/22 et C‑609/22). En confirmant et en élargissant la protection des réfugiées risquant des persécutions en raison de leur seule appartenance au groupe social des femmes, la Cour tient compte des réalités tout en consolidant l’importance des références aux droits fondamentaux et au droit international dans l’interprétation du droit européen de l’asile. 

1.Risques de persécutions liées à l’appartenance à un groupe social 

Au regard de l’actualité en Afghanistan, mais aussi de l’interprétation de la notion même de réfugié, en particulier en raison de risques de persécutions liées à l’appartenance à un certain groupe social, l’arrêt AH et FN, Bundesamt für Fremdenwesen und Asyl e. a. (Femmes afghanes), qui n’a pas été prononcé en grande chambre et qui est le dernier dans l’ordre chronologique, mérite néanmoins de servir de point de départ pour la présentation globale de ces jurisprudences. Le nombre de commentaires dont l’arrêt a déjà fait l’objet témoigne également de son importance (voy., ex multis, Nicolosi et Ertuna Lagrand ; Gupta ; Esmailian ; Golesorkhi ; Maheshe Musole ; Acierno). Dans cette affaire, les questions préjudicielles concernaient deux Afghanes séjournant en Autriche depuis plus de cinq ans. Elles y bénéficiaient de la protection subsidiaire, car elles « seraient exposées à des difficultés de nature économique et sociale si elles retournaient » en Afghanistan (point 21), mais le statut de réfugiées leur avait été refusé au motif notamment qu’« elles n’avaient pas adopté un mode de vie “occidental” qui serait devenu une composante de leur identité à ce point essentielle qu’il leur serait impossible d’y renoncer afin d’échapper à des menaces de persécution dans leur pays d’origine » (point 23).  

La décision autrichienne était une forme d’écho déformé de la décision antérieure de la Cour, prononcée en grande chambre, dans l’affaire K.L. Staatssecretaris van Justitie en Veiligheid (Femmes s’identifiant à la valeur de l’égalité entre les sexes). Cet arrêt a été, lui aussi, déjà amplement commenté (voy. Nicolosi et Ertuna Lagrand ; Raimondo et De Coninck ; Terlouw). Dans l’espèce, il s’agissait d’Irakiennes dont la jurisprudence du Conseil d’État des Pays-Bas considérait que le seul fait de s’être occidentalisées aux Pays-Bas ne suffisait à leur reconnaître la qualité de réfugiées car « ces “femmes occidentalisées” constituent un groupe trop hétéroclite pour qu’elles puissent être considérées comme appartenant à un “certain groupe social” » au sens de la convention de Genève relative aux réfugiés ou de la directive qualification (point 27). Écartant cette interprétation restrictive, la Cour avait dit pour droit que « en fonction des conditions qui prévalent dans le pays d’origine, peuvent être considérées comme appartenant à “un certain groupe social”, en tant que “motif de persécution” susceptible de conduire à la reconnaissance du statut de réfugié, les femmes ressortissantes de ce pays, y compris mineures, qui partagent comme caractéristique commune leur identification effective à la valeur fondamentale de l’égalité entre les femmes et les hommes, intervenue au cours de leur séjour dans un État membre » (dispositif).  

En sa motivation, la Cour, reprenant la définition du groupe social donnée par la jurisprudence depuis l’arrêt Ward de la Cour suprême du Canada en 1993 et par la directive qualification, la Cour précisait que « l’identification effective d’une ressortissante d’un pays tiers à la valeur fondamentale de l’égalité entre les femmes et les hommes peut être considérée comme “une caractéristique ou une croyance à ce point essentielle pour l’identité ou la conscience qu’il ne devrait pas être exigé d’une personne qu’elle y renonce” » (point 44, italique ajoutés) 

C’est en faisant une interprétation conditionnelle de cette précision inscrite dans les textes et  reproduite ci-dessus que certains États, dont l’Autriche, ont refusé, par la suite, la reconnaissance de la qualité de réfugié aux femmes n’établissant pas qu’elles auraient « adopté un mode de vie “occidental » qui serait devenu une composante de leur identité à ce point essentielle qu’il leur serait impossible d’y renoncer afin d’échapper à des menaces de persécution dans leur pays d’origine » (point 23 de l’arrêt Femmes afghanes). En conséquence, deux questions se posaient dans l’affaire Femmes afghanes. Comme le synthétisait fort bien l’avocat général Richard de la Tour en introduction de ses conclusions qui seront suivies par la Cour, « face à cette situation, les autorités des États membres hésitent entre accorder le statut de réfugié à ces femmes en raison uniquement de leur sexe et rechercher individuellement l’existence d’un risque de persécutions » (point 1 de ses conclusions).  

2.Qu’est-ce qu’une persécution fondée sur la discrimination de genre ?

S’agissant de la notion de persécution, l’article 9 de la directive qualification en donne une définition alternative reposant soit sur un critère qualitatif, soit sur un critère quantitatif. Le critère qualitatif vise les actes graves qui constituent une violation grave des droits fondamentaux de l’homme, en particulier des droits indérogeables. Le critère quantitatif vise une accumulation de mesures moins graves, mais suffisamment graves du fait de cette accumulation pour affecter un individu de manière comparable. La Cour constate que, bien qu’alternatifs, ces deux critères sont remplis s’agissant des discriminations subies par les femmes en Afghanistan. Ainsi, au regard du critère qualitatif, la Cour note que le  « mariage forcé […] est assimilable à une forme d’esclavage, prohibé par l’article 4 de la CEDH, et [que] l’absence de protection contre les violences fondées sur le sexe et les violences domestiques […] constituent des formes de traitements inhumains et dégradants, interdits par l’article 3 de la CEDH » (point 43). Ce faisant, la Cour de justice renforce la « condamnation » de persécutions de type privé qu’elle avait déjà affirmée dans l’arrêt Intervyuirasht organ na DAB pri MS (Femmes victimes de violences domestiques) rendu en grande chambre (pour des commentaires, voy. Flamand ; Fleury Graff ; De Vido et Möschel ; Loxa ; Kübek et Bornemann ; Ertuna Lagrand ; Kompatscher et al. ; Steininger ; Minardi ; Stamme ; Di Pascale).  

Dans cette affaire Femmes victimes de violences domestiques, relative à des femmes kurdes de Turquie victimes de violences domestiques, la Cour a aussi souligné le caractère alternatif du lien qui doit exister entre la crainte avec raison de persécutions et l’un des motifs de la convention de Genève (race, religion, nationalité, appartenance à un certain groupe social ou opinion politique). Ce lien peut tantôt être le motif direct de la persécution, tantôt le motif indirect en étant le motif de l’absence de protection contre la persécution « privée » (cette approche alternative du lien de rattachement avec l’un des motifs de persécution est devenue classique aussi depuis la jurisprudence Ward). En outre, dans l’arrêt Femmes afghanes, s’agissant du critère qualitatif résultant de l’atteinte à un droit fondamental indérogeable, la Cour n’hésite pas à assimiler le mariage forcé à l’esclavage ce qui rappelle la tragédie grecque des Suppliantes, dans laquelle Eschyle met en scène, au Ve siècle av. J.-C., des femmes, les Danaïdes, qui demandent et obtiennent l’asile car elles fuient un mariage forcé avec les fils d’Égyptos. Ceci est également en ligne avec la jurisprudence de la Cour pénale internationale qui a hissé le mariage forcé au niveau du crime contre l’humanité (CPI, aff. Dominique Ongwen ; CPI, aff. Al Hassan, qualifiant le mariage forcé de crime contre l’humanité même si, en l’espèce, Al Hassan a été acquitté de ce chef, voy. point 40 du résumé d’audience).  

Toujours dans l’arrêt Femmes afghanes, au regard cette fois du critère quantitatif de la notion de persécution, résultant du cumul d’atteintes aux droits fondamentaux, la Cour note que  « à supposer que, prises isolément, les mesures discriminatoires à l’égard des femmes qui restreignent l’accès aux soins de santé, à la vie politique et à l’éducation ainsi que l’exercice d’une activité professionnelle ou sportive, ou qui entravent la liberté d’aller et de venir ou encore portent atteinte à la liberté de se vêtir ne constituent pas une violation suffisamment grave d’un droit fondamental, […] ces mesures, prises dans leur ensemble, affectent les femmes d’une manière telle qu’elles atteignent le niveau de gravité requis pour constituer des actes de persécution […] (en effet), en raison de leur effet cumulé et de leur application délibérée et systématique, ces mesures aboutissent à dénier, de manière flagrante et avec acharnement, aux femmes afghanes, en raison de leur sexe, les droits fondamentaux liés à la dignité humaine. Lesdites mesures témoignent de l’établissement d’une organisation sociale fondée sur un régime de ségrégation et d’oppression dans lequel les femmes sont exclues de la société civile et privées du droit de mener une vie quotidienne digne dans leur pays d’origine » (point 43).  

 3.Une approche collective plutôt qu’individuelle est-elle possible ?

 La réponse à la deuxième question se déduit alors de ce constat : d’un point de vue procédural, « s’agissant des demandes de protection internationale déposées par des femmes ressortissantes d’Afghanistan, les autorités nationales compétentes peuvent considérer qu’il n’est actuellement pas nécessaire d’établir, lors de l’examen individuel de la situation d’une demanderesse de protection internationale, que celle-ci risque effectivement et spécifiquement de faire l’objet d’actes de persécution en cas de retour dans son pays d’origine, dès lors que les éléments relatifs à son statut individuel, tels que sa nationalité ou son sexe, sont établis » (point 57 et dispositif). En d’autres termes l’examen individuel peut se limiter à l’identité (en l’occurrence le sexe et la nationalité), le risque de persécution pouvant faire l’objet d’une « présomption de reconnaissance du statut de réfugié à l’égard des femmes et des jeunes filles afghanes, compte tenu des actes de persécution perpétrés contre elles par les talibans, uniquement en raison de leur genre » (point 56).  Ce qui conduit une partie de la doctrine centrant l’analyse sur une approche de genre, à voir dans cette décision une reconnaissance d’un « apartheid de genre », ce que confirme l’expression « régime de ségrégation » utilisée par la Cour (voy. Esmailian et, surtout pour son analyse détaillée des liens avec le droit pénal international, Golesorkhi). 

Dans une perspective plus large, il est aussi possible d’étendre ce raisonnement à d’autres groupes sociaux, constatant que « ce raisonnement peut conduire à une protection de groupe très présente dans le système interafricain des réfugiés », comme le souligne pertinemment Trésor Maheshe Musole dans son commentaire. Pour justifier cette ouverture vers une approche collective de la protection des réfugiées victimes de persécution en raison de leur appartenance à un groupe social, la Cour dit que « l’exigence même de procéder à une évaluation individuelle de la demande de protection internationale suppose que les autorités nationales compétentes adaptent les modalités d’appréciation des éléments de fait et de preuve en fonction des circonstances et des caractéristiques propres à chaque demande » (point 54). Elle ajoute par ailleurs que l’article 3 de la directive permet aux États membres d’adopter des normes plus favorables.  

Reste une interrogation pour le futur. Dans le nouveau Pacte, la directive qualification est, comme les autres textes, remplacée par un règlement. Si ce règlement ne modifie pas la définition des actes de persécution, il ne comporte plus la clause d’ouverture de l’article 3, permettant aux États membres d’accroître le niveau de protection. En conséquence, pour résister à l’évolution législative, cette jurisprudence de la Cour devra nécessairement se fonder sur l’adaptation réaliste de l’examen individuel à la situation de violation massive de droits fondamentaux dans le pays d’origine et non sur une possible protection étatique renforcée. Sur ce point les conclusions de l’avocat général étaient peut-être plus fermes en constatant que « les mesures discriminatoires […] relèvent d’un régime de ségrégation et d’oppression qui est mis en œuvre à leur égard du seul fait de leur présence sur le territoire, sans considération de leur identité ou de leur situation personnelle » (point 73, conclusions de l’avocat général Richard de la Tour) en manière telle que l’approche reste individuelle tout en étant modalisée par le niveau d’atteinte aux droits fondamentaux. Du reste, c’est ce que la Cour a déjà fait dans sa jurisprudence Elgafaji (C-465/07). Il s’agissait d’un autre texte et contexte relatif à la protection subsidiaire résultant de « menaces graves et individuelles contre la vie ou la personne d’un civil en raison d’une violence aveugle en cas de conflit armé interne ou international » au sens de l’article 15, lettre c), de la directive qualification. La Cour avait précisé cette approche en 2023 à propos de la situation en Libye dans l’arrêt Staatssecretaris van Justitie en Veiligheid (Notion d’atteintes graves) (C-125/22). Selon ce raisonnement il existe une sorte de proportionnalité interne à la gravité de l’atteinte aux droits fondamentaux qui pourrait s’appliquer à une persécution au sens de la Convention de Genève : plus celle-ci est élevée et généralisée, moins il convient d’établir qu’elle vise individuellement la personne en recherche de protection. 

Notons encore que, dans ces trois arrêts, la Cour fait de nombreuses références à l’élimination des discriminations à l’égard des femmes pour y adosser son interprétation large des persécutions à leur égard. Elle cite non seulement, de façon classique, la convention de Genève comme « pierre angulaire » de la protection internationale des réfugiés et les recommandations d’interprétation du Haut-commissariat aux réfugiés des Nations unies (HCR), mais aussi la convention sur l’élimination de toutes les formes de discrimination à l’égard des femmes (CEDEF), soulignant que tous les États membres de l’Union y sont parties, et la convention d’Istanbul du Conseil de l’Europe sur la prévention et la lutte contre la violence à l’égard des femmes et la violence domestique, soulignant que l’Union y est elle-même partie et qu’elle y est entrée en vigueur le 1er octobre 2023 (voy. Porcheron, 2024, p. 49). 

Conclusion  

Il est indéniable que, par ces trois jurisprudences, la Cour renforce la protection des femmes en qualité de réfugiées. En outre, par sa jurisprudence relative aux femmes afghanes, elle ouvre la voie vers une approche plus collective de certains bénéficiaires de la protection internationale en raison de l’ampleur des discriminations ou des violations massives de droits fondamentaux qui atteignent le niveau de persécutions subies par ce groupe social. À ce jour, une telle approche collective, introduisant une présomption juris tantum de besoin de protection, ne se faisait qu’à l’égard de catégories de personnes qui bénéficiaient de dispositions particulières, comme les Palestiniens contraints de fuir la zone de protection de l’UNRWA ou les Ukrainiens bénéficiaires de la protection temporaire. La Cour vient d’ajouter les femmes d’Afghanistan à ces deux catégories.  

A Better Group Protection for Refugee Women 

By Jean-Yves Carlier and Eleonora Frasca,  

Université catholique de Louvain (UCLouvain), members of the Equipe droits et migrations (EDEM)

 

This is a revised version of the commentary in “Droit européen des migrations”, the yearly case law column published in French in the Journal de droit européen, no. 3, March 2025.

In 2024, the Court of Justice of the European Union (CJEU) ruled on various issues relating to the protection of women as refugees in three decisions, including two Grand Chamber judgments: the fear of persecution resulting from domestic violence (W.S., C-621/21), the identification with the values of gender equality in the host country (K.L., C-646/21), as well as the finding that all women in Afghanistan belong to a social group with a well-founded fear of persecution (joined cases AH C-608/22 and FN C-609/22). By confirming and extending the protection of refugee women at risk of persecution solely because they belong to the social group of women, the CJEU is taking into account women’s realities while consolidating the importance of references to fundamental rights and international law in the interpretation of EU asylum law. 

1.Risk of Persecution Linked to Membership of a Particular Social Group

Not delivered by the CJEU’s Grand Chamber, the AH and FN judgment, also known as Bundesamt für Fremdenwesen und Asyl and Others (Afghan Women) is the last in chronological order. Nevertheless, it deserves to be taken as the starting point for the overall presentation of these cases, not only in light of current events in Afghanistan but also for the interpretation of the very concept of refugee, especially because of the risk of persecution linked to membership of a particular social group. The high number of case notes testify to its importance (see, ex multis, Nicolosi and Ertuna Lagrand; Gupta; Esmailian; Golesorkhi; Maheshe Musole; Acierno). 

The questions referred for a preliminary ruling in this case concerned two Afghan women who had been living in Austria for more than five years. They had been granted subsidiary protection because “they would face economic and social difficulties” if returned to Afghanistan (para 21), but had been refused refugee status on the grounds that “they had not adopted a ‘Western’ lifestyle that had become such an essential part of their identity that it was impossible for them to renounce it in order to escape the threat of persecution in their country of origin” (para 23). The Austrian decision was a distorted echo of the CJEU’s earlier Grand Chamber decision in K.L., Staatssecretaris van Justitie en Veiligheid (Women identifying with the value of gender equality). This judgment too has been widely commented (see Nicolosi and Ertuna Lagrand; Raimondo and De Coninck; Terlouw). The case concerned Iraqi women who – according to the Dutch Council of State’s case law – could not be recognised as “refugees” simply because “Westernised women constitute too diverse a group to be regarded as members of a particular social group” within the meaning of the Geneva Refugee Convention or the Qualification Directive (para 27).  

Dismissing this restrictive interpretation, the Court ruled that “depending on the circumstances in the country of origin, women, including minors, who share as a common characteristic the fact that they genuinely come to identify with the fundamental value of equality between women and men during their stay in a Member State may be regarded as belonging to a particular social group, constituting a reason for persecution capable of leading to the recognition of refugee status” (operative part). In its reasoning, reiterating the definition of a social group given in case law since Ward by the Supreme Court of Canada in 1993 and in the Qualification Directive, the Court stated that “the fact that a woman genuinely comes to identify with the fundamental value of equality between women and men may be considered a characteristic or belief that is so fundamental to identity or conscience that a person should not be forced to renounce it” (para 44, emphasis added). 

By making it a condition, certain States, including Austria, subsequently refused the recognition of refugee status to women who did not establish that they had “adopted a Western lifestyle that had become such an essential part of their identity that it was impossible for them to renounce it in order to escape the threat of persecution in their country of origin” (para 23 of the Afghan Women judgment). Consequently, two questions arose in the Afghan Women case. Advocate General Richard de la Tour summed up very well the stakes of this case in the introduction to his Opinion which was followed by the Court: “faced with this situation, Member State authorities are uncertain whether to grant refugee status to those women and girls solely on account of their gender, or to identify whether an individual risk of persecution exists” (para 1 of his Opinion).

2. What is Persecution based on Gender Discrimination?

Article 9 of the Qualification Directive gives an alternative definition of persecution based on either qualitative or quantitative criteria. The qualitative criterion refers to serious acts that constitute a severe violation of basic human rights, in particular non-derogable fundamental rights. The quantitative criterion refers to an accumulation of various measures that are less severe, but sufficiently severe by reason of that accumulation to affect an individual in a comparable manner. The Court finds that, although they are alternative, these two criteria are met in relation to the discrimination suffered by women in Afghanistan. Thus, with regard to the qualitative criterion, the Court notes that “forced marriage […] is comparable to a form of slavery prohibited under Article 4 ECHR, and the lack of protection against gender-based violence and domestic violence […] constitute forms of inhuman and degrading treatment prohibited by Article 3 ECHR” (para 43). In so doing, it reinforced the “condemnation” of private persecution that it had already affirmed in the Grand Chamber judgment in Intervyuirasht organ na DAB pri MS (Women victims of domestic violence) (see Flamand; Fleury Graff; De Vido and Möschel; Loxa; Kübek and Bornemann; Ertuna Lagrand; Kompatscher et al.; Steininger; Minardi; Stamme; Di Pascale).  

The case Women victims of domestic violence concerned Kurdish women from Turkey who were victims of domestic violence. The Court also emphasised the alternative nature of the link that must exist between a well-founded fear of persecution and one of the grounds in the Geneva Convention (race, religion, nationality, membership of a particular social group or political opinion). This link may sometimes be the direct reason for the persecution and sometimes the indirect reason for the lack of protection against “private” persecution (this alternative approach to the link with one of the reasons for persecution has become classic also since Ward). 

Furthermore, in the Afghan Women judgment, with regard to the qualitative criterion resulting from the infringement of a non-derogable fundamental right, the Court did not hesitate to equate forced marriage with slavery. This should be emphasised. It reminds the Greek tragedy The Suppliants in which Aeschylus, in the fifth century BC, portrays the Danaids, women who seek and obtain asylum because they are fleeing a forced marriage to the sons of Aegyptus. This is also in line with the case law of the International Criminal Court which has raised forced marriage to the level of a crime against humanity (ICC, Dominique Ongwen; ICC, Al Hassan case, even though Al Hassan was acquitted of this charge, see point 40 of the summary of the hearing). 

Also with regard to the quantitative criterion of the concept of persecution resulting from the cumulative infringement of fundamental rights, the Court noted in the Afghan Women judgment that “even if, taken separately, the discriminatory measures against women that restrict access to healthcare, political life and education and the exercise of a professional or sporting activity, restrict freedom of movement or infringe the freedom to choose one’s clothing do not constitute a sufficiently serious breach of a fundamental right […], those measures, taken as a whole, affect women to an extent that they attain the level of severity required to constitute acts of persecution […]. […] [G]iven that those measures have a cumulative effect and are applied deliberately and systematically, they blatantly and relentlessly deny Afghan women fundamental rights related to human dignity on account of their gender. Those measures reflect the establishment of a social structure based on a regime of segregation and oppression in which women are excluded from civil society and deprived of the right to lead a dignified daily life in their country of origin” (para 44).

3. Is a Collective Rather than Individual Approach Possible?

The answer to the second question can be deduced from this: from a procedural point of view, “regarding applications for international protection lodged by women who are Afghan nationals, the competent national authorities are entitled to consider that it is currently unnecessary to establish, in the individual assessment of the situation of an application for international protection, that there is a risk that she will actually and specifically be subject to acts of persecution if she returns to her country of origin, where the factors relating to her individual status, such as her nationality or gender, are established” (para 57 and operative part). In other words, the individual examination may be limited to identity (in this case gender and nationality), while the risk of persecution may be the subject of “a presumption of recognition of refugee status for Afghan women and girls, given the acts of persecution carried out against them by the Taliban solely on account of their gender” (para 56). This has led some scholars to focus their analysis on a gender approach (see Esmailian and especially for her detailed analysis of the links with international criminal law, Golesorkhi). They see in this decision the recognition of a “gender apartheid”, confirmed by the expression “regime of segregation” used by the Court (para 44). 

From a broader perspective, it is also possible to extend this to other social groups, noting that “this reasoning can lead to a group protection that is very present in the inter-African refugee system”, as Maheshe Musole rightly points out in his commentary (translation is our own). To justify this move towards a collective approach to the protection of refugees who are victims of persecution because they belong to a social group, the Court states that “the requirement to carry out an individual assessment of the application for international protection presupposes that the competent national authorities adapt the methods of assessing the facts and evidence according to the specific circumstances and characteristics of each application” (para 54). It adds that Article 3 of the Qualification Directive allows Member States to introduce more favourable standards. We endorse this realistic and progressive interpretation of protection against persecution on the grounds of membership of a particular social group. 

One question remains for the future. In the new Pact, the Qualification Directive is replaced by a regulation. While this regulation does not change the definition of acts of persecution, it no longer includes the opening clause in Article 3, allowing Member States to increase the level of protection. Consequently, in order to withstand legislative change, the Court’s case law will necessarily have to be based on a realistic adaptation of the individual examination to the situation of massive violation of fundamental rights in the country of origin, and not on a possible increase in the protection offered by a Member State. On this point, Advocate General Richard de la Tour’s Opinion was perhaps stronger. He noted that “the discriminatory measures […] are part of a regime of segregation and oppression conducted solely on account of their presence on the territory, regardless of their identity or personal circumstances” (para 73, Opinion of Advocate General) 

This interpretation remains an individual approach, modulated by the level of infringement of the fundamental right. Moreover, this is what the Court did in its Elgafaji case (C-465/07). Of course it is another text and context about the subsidiary protection resulting from “serious and individual threats to the life or person of a civilian by reason of indiscriminate violence in the event of internal or international armed conflict” within the meaning of Article 15(c) of the Qualification Directive. The Court clarified this approach in 2023 in relation to the situation in Libya in Staatssecretaris van Justitie en Veiligheid (Notion of serious harm) (C-125/22). But, according to this reasoning, there is a kind of internal proportionality to the seriousness of the infringement of fundamental rights that could be applied to persecution in the sense of the Geneva Convention: the more serious and widespread the infringement, the less it is necessary to establish that it is aimed individually at the person seeking protection. 

It should also be noted that the Court makes in these three judgments numerous references to international law on the elimination of all forms of discrimination against women to support its broad interpretation of the persecution of women. In addition to the standard references to the Geneva Convention as the “cornerstone” of international refugee protection and the Guidelines on International Protection of the United Nations High Commissioner for Refugees (UNHCR), the Court also cited the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), emphasising that all EU Member States are party to it, and the Istanbul Convention of the Council of Europe on the Elimination of All Forms of Discrimination against Women, pointing out that the Union is itself a party to it and that it came into force in the EU on 1 October 2023 (see Porcheron, 2024, p. 49). 

Conclusion 

It is beyond any doubt that, with these three judgments, the Court is strengthening the protection of women as refugees. Furthermore, the Court is paving the way for a more collective approach to certain beneficiaries of international protection, given the scale of the discrimination or massive violations of fundamental rights that reach the level of persecution suffered by this social group. To date, such a collective approach, introducing a presumption juris tantum of the need for protection, has only been applied to categories of people who have benefited from special provisions, such as Palestinians forced to flee the UNRWA protection zone or Ukrainians benefiting from temporary protection. The Court has now added Afghan women to these categories. 

Réfugié d’un Etat membre à l’autre: vers la reconnaissance mutuelle de plein droit ?

Par Jean-Yves Carlier et Eleonora Frasca, Université catholique de Louvain (UCLouvain), membres de l’équipe droits et migrations (EDEM)

 

Ceci est une version révisée de la chronique de jurisprudence annuelle publiée dans le Journal de droit européen (n° 3, mars 2025)

La Cour de justice de l’Union européenne apporte des précisions sur la question des effets à attribuer dans un État membre à une décision relative au statut de réfugié prise dans un autre État membre dans deux arrêts rendus en grande chambre le 18 juin 2024 (QY, C-753/22 et A., C-352/22). Dans ces deux affaires, les requérants cherchent à faire valoir l’effet obligatoire de la décision de reconnaissance du statut de réfugié, obtenue dans un État membre, à l’égard d’une procédure qui se déroule dans un autre État membre. Dans QY Bundesrepublik Deutschland (Effet d’une décision d’octroi du statut de réfugié), il s’agit d’une procédure d’asile dans laquelle une Syrienne demande que le statut de réfugiée obtenu en Grèce soit reconnu en Allemagne qui, à son tour, ne lui a accordé que la protection subsidiaire. Dans A. Generalstaatsanwaltschaft Hamm (Demande d’extradition d’un réfugié vers la Turquie), il s’agit d’une procédure d’extradition d’un Turc, d’origine kurde, reconnu réfugié en Italie, qui fait l’objet en Allemagne d’une demande d’extradition de la part de la Turquie, son pays d’origine. Ces deux affaires ont déjà fait l’objet de nombreux commentaires (Michel; Boucheron; Pahladsingh; Fratea; Simon and Rigaux; Peers). Ici, nous nous centrons sur la question des effets contraignants, dans un autre État membre de l’Union, d’une décision relative au statut de réfugié. Nous raisonnons en trois étapes : premièrement, le passé qui révèle l’aporie du droit européen en matière de (non-)reconnaissance mutuelle des décisions relatives au statut de réfugié (1) ; deuxièmement, le présent des deux jurisprudences commentées, avec la reconnaissance de quelques effets limités de décisions d’octroi du statut de réfugié (2) et, troisièmement, le futur avec, une évolution possible, notamment sous l’influence paradoxale du Pacte, vers la reconnaissance mutuelle des décisions d’octroi du statut de réfugié (3).

1. Passé : L’aporie de la reconnaissance des décisions positives d’un autre État membre

L’aporie est évidente. Les décisions négatives de refus du statut de réfugié, prises dans un État membre, sortent des effets contraignants dans un autre État membre. A l’inverse, les décisions positives attribuant le statut de réfugié ne sont pas reconnues dans un autre État membre. En voici quelques exemples. En 2021, dans l’affaire M e.a. (Transfert vers un État membre) (C‑673/19), portant sur le placement en rétention de réfugiés en séjour irrégulier en vue de leur transfert forcé vers l’État membre de protection, la Cour avait affirmé que ni les décisions de placement en rétention ni les décisions de transfert vers l’État de protection ne sont régies par les normes et les procédures communes établies par la directive retour. Ces décisions relèvent de la seule compétence de l’État membre car les règles et les procédures communes établies par cette directive ne portent que sur l’adoption et sur l’exécution de décisions de retour. 

Cette affaire, tout comme celles que nous analysons infra (C-753/22 et C-352/22), illustrent parfaitement l’aporie du droit européen de l’asile en matière de reconnaissance des décisions prises dans un autre État membre. En bref, une décision négative refusant le statut, prise dans un État membre, sera « reconnue » dans les autres États membres, en ce que, au mieux, elle conduira à un rejet de toute nouvelle demande sur la base d’une procédure accélérée. Comme le souligne l’avocate générale Laila Medina, dans ses conclusions sur la première affaire (C-753/22), le mécanisme de Dublin rend « le principe de reconnaissance mutuelle pleinement effectif et opérationnel, mais uniquement en ce qui concerne les décisions défavorables » (point 56 de ses conclusions). Même une décision positive de reconnaissance du statut de réfugié obtenue dans un État membre aura principalement un effet négatif dans un autre État membre. Ainsi, dans Ibrahim e.a. (C-297/17 e.a.) et dans Commissaire général aux réfugiés et aux apatrides (Unité familiale – Protection déjà accordée) (C-483/20), la Cour avait déjà souligné cet effet négatif: un autre État membre peut déclarer une nouvelle demande irrecevable au motif qu’« une protection internationale a été accordée par un autre État membre » (art. 33 (2) a) de la directive procédures), sauf circonstances exceptionnelles telles que des défaillances au niveau de la procédure d’asile et de l’accueil dans l’État membre qui a octroyé la protection, qui exposeraient l’intéressé à un risque réel d’y être soumis à des traitements inhumains ou dégradants au sens de l’article 4 de la Charte (sur ce sujet, voy. Frasca et Rizzuto).

2. Présent : Des effets positifs mais limités ?

Cela signifie-t-il qu’une décision de reconnaissance du statut de réfugié n’aura aucun effet positif dans un autre État membre ? Non. Les arrêts QY (C-753/22) et A. (C-352/22) permettent à la Cour de mieux préciser les effets positifs du fait de la protection des personnes concernées, d’un statut de réfugié obtenu dans un autre État membre. L’effet positif sur une nouvelle procédure d’asile éventuelle, comme dans QY, est limité à l’obligation d’échange d’informations entre les États membres. 

En revanche, les effets dégagés par la Cour dans A., en matière d’extradition, sont plus contraignants pour les États. La Cour considère que l’article 21 de la directive qualification, titré « protection contre le refoulement » doit être lu à la lumière des articles 18 et 19, paragraphe 2, de la Charte des droits fondamentaux, avec pour conséquence que l’extradition d’un réfugié vers son pays d’origine « aurait pour effet de le priver de la jouissance effective du droit que lui offre l’article 18 de la Charte » car cet article « s’oppose à son extradition vers le pays tiers qu’il a fui et dans lequel il risque d’être persécuté », tant qu’il « remplit les conditions pour jouir de la qualité de réfugié » (point 58). Allant au-delà des conclusions de l’avocat général Richard de la Tour, la Cour en déduit une condition préalable à toute extradition : la révocation du statut de réfugié par les autorités du premier État membre. La Cour explique que « pour autant que ce statut n’ait pas été révoqué ou retiré par l’État membre l’ayant octroyé », l’extradition doit être refusée (point 64). Ici aussi, un « échange d’informations » s’impose, en vertu du principe de coopération loyale (art. 4 (3) TUE). Cet échange est un processus à double sens. Il va permettre, d’une part, à l’autorité compétente pour l’extradition de procéder aux vérifications nécessaires en vertu des articles 18 et 19, paragraphe 2, de la Charte, et, d’autre part, à l’autorité de l’autre État membre d’éventuellement révoquer ou retirer le statut de réfugié. 

Deux expressions frappent dans les extraits qui précèdent. D’une part, l’expression « priver (un réfugié) de la jouissance effective du droit » offert par la Charte (point 58). Cela nous rappelle bien sûr le mantra qui permet désormais à la Cour d’entrer en matière dans les situations purement internes, depuis sa jurisprudence Ruiz Zambrano (C-34/09). Dans ces jurisprudences, la Cour vérifie si le citoyen européen risque d’être privé de la « jouissance effective de l’essentiel de [ses] droits ». Ici, la Cour vise l’examen de la jouissance effective « des droits », sans le centrer, comme dans le mantra Zambrano sur « l’essentiel des droits ». Elle  a fait de même en 2024 dans une affaire concernant un citoyen de l’UE dans une situation purement interne (aff. jointes NW et PQ, C-420/22 et 528/22). Cela montre à quel point le mantra devient un concept variable, au contenu évolutif, qui peut s’appliquer aux citoyens de l’Union comme aux ressortissants de pays tiers, mais avec des nuances (sur ce sujet, pour approfondir, voy. Carlier). La Cour explicite progressivement le contenu des droits protégés par référence à certaines dispositions de la Charte. Il y a là une référence de principe commune au cœur des droits fondamentaux. D’autre part, les mots « jouir de la qualité de réfugié » font à nouveau référence à la distinction entre qualité et statut de réfugié (point 58). La Cour avait, déjà en 2019, introduit cette distinction entre « qualité » et « statut » à propos des conséquences de la révocation du statut de réfugié (voy. C-159/21, C-391/16 et C-663/21). L’octroi du statut de réfugié peut être révoqué pour des raisons qui ne figurent pas dans la convention de Genève, mais la qualité de réfugié protège toujours les droits fondamentaux des réfugiés, comme le principe de non-refoulement. La Cour semble suivre la même approche dans ce cas.

Si on peut approuver la Cour de veiller à sauvegarder des garanties minimales pour la personne ayant été reconnue réfugiée, il reste que les zones d’ombre et les limbes juridiques dans lesquels ces distancions plongent le fugitif mériteraient d’être levés. Ainsi, rien n’indique que le retrait du « statut » de réfugié est exigé par la Cour comme condition pour permettre au deuxième État d’envisager l’extradition. Rien n’indique que cela permettrait effectivement à ce second État d’extrader vers le pays d’origine. En effet, selon la jurisprudence de la Cour elle-même, si les motifs de retrait ne sont pas conformes à la convention de Genève, l’intéressé perdrait certes le « statut », mais il devrait toujours garder la « qualité » de réfugié. En conséquence, il ne pourrait faire l’objet d’un refoulement vers son pays d’origine contraire à l’article 33 de la convention de Genève et donc ne pourrait pas être extradé. On mesure toute la complexité de ces solutions qui entendent respecter le droit international sans remettre en question les textes du droit de l’Union. Sera-t-il possible, dans le futur, de clarifier ces zones obscures ? 

3. Futur : vers une reconnaissance mutuelle de plein droit avec le pacte ?

La question de la reconnaissance mutuelle des décisions relatives au statut de réfugié reste ouverte au regard du Pacte sur la migration et l’asile. En l’état du droit de l’Union au moment de sa décision, c’est-à-dire avant le Pacte, la Cour a rappelé que les États membres demeuraient libres de prévoir des dispositions plus favorables, y compris sur la reconnaissance automatique de décisions d’octroi d’un statut de protection adoptées par un autre État membre (QY, point 69 au sens de l’art. 3 de la directive qualification et de l’art. 5 de la directive procédures). 

Déjà peu réaliste aujourd’hui, au regard de la concurrence entre États, chacun voulant éviter d’être plus accueillant que l’autre, cette possibilité de transposer le droit de l’Union en y ajoutant des dispositions plus favorables disparaitrait demain, compte tenu de la nouvelle forme de droit dérivé introduite par les textes du Pacte. Il s’agit de règlements directement applicables, non plus de directives à transposer. Seul le nouveau règlement procédures (2024/1348) prévoit en son considérant 6 la possibilité de dispositions nationales plus favorables « lorsque le présent règlement le prévoit ». Mais on chercherait en vain, dans les articles du règlement, ces hypothèses de dispositions plus favorables qui permettraient la reconnaissance mutuelle de décisions relatives au statut de réfugié. Comme l’article 61, paragraphe 2, qui permet, à propos du concept de pays d’origine sûrs, d’introduire des exceptions pour des parties de territoire ou catégories de personnes, il s’agit plutôt d’articles qui permettent des exceptions aux dispositions strictes qui ne font pas l’unanimité que d’articles prévoyant des dispositions plus favorables. En outre, les nouveaux règlements « qualification » ou « procédures » ne comportent aucune disposition substantielle relative à la reconnaissance mutuelle comme cela avait été évoqué de lege ferenda. Ce n’est pas dire que ces reconnaissances mutuelles ne seraient pas facilitées par l’entrée en application du Pacte ? En effet, on pourrait considérer que, par la voie de règlements, le Pacte a renforcé l’uniformisation du droit européen de l’asile (voy Barbou des Places). En conséquence, on devrait s’interroger, en pratique et en théorie, sur les conséquences de ce renforcement de l’intégration au sein d’un espace commun.

Des arguments pratiques aux arguments de principe

Les jurisprudences ici commentées confirment donc que, pour l’essentiel, une décision positive attribuant le statut de réfugié, prise dans un État membre, ne sera pas reconnue dans un autre État membre. En pratique, on comprend l’objectif qui est d’éviter les mouvements secondaires de réfugiés de certains États vers d’autres. Rien n’indique toutefois, statistiquement, que cela représente encore une réelle difficulté. A la différence des mouvements secondaires des candidats réfugiés, qui sont, pour partie, limités par le mécanisme de Dublin, les mouvements secondaires de réfugiés reconnus ne paraissent pas représenter un défi beaucoup plus important que celui des citoyens européens ou des ressortissants d’États tiers résidents de longue durée. En effet, le nombre de réfugiés reconnus demeure limité, en manière telle que ces mouvements secondaires, qui pourraient en partie se croiser, ne paraissent pas insurmontables. On se souviendra également que les personnes fuyant l’ Ukraine ont non seulement le libre choix de l’État membre dans lequel elles demandent le statut de protection temporaire, mais, en plus, une fois ce statut obtenu, elles peuvent le conserver et se déplacer dans l’Union, en le transportant comme un sac à dos, pour s’installer dans un autre État membre où ce statut doit être reconnu de plein droit. Au-delà des arguments pratiques qui ne résistent guère au réel, l’absence de reconnaissance mutuelle des décisions relatives au statut de réfugié est surtout contestable du point de vue des principes, tant en droit international privé qu’en droit de l’Union. 

Le droit international privé

En droit international privé, le principe est, de plus en plus, la reconnaissance de plein droit des actes et décisions relatives aux matières civiles et commerciales, que celles-ci émanent d’un autre État membre ou même d’un État tiers. Au-delà des règles de conflit de juridictions et de conflits de lois, la technique dominante du droit international privé devient la reconnaissance des situations (voy. Azoulai e.a. ; Lagarde ; Baratta ; d’Avout ; Pfeiff ; Fulli-Lemaire). Il est vrai que, en droit international privé européen, le règlement Bruxelles I bis (1215/2012) exclut de son champ d’application matériel l’état et la capacité des personnes physiques (art. 1, § 2, sous a). Toutefois, sur la base du principe de confiance mutuelle, les actes de l’état civil – les actes de naissance et d’établissement de nationalité établis dans un État membre – sont reconnus de plein droit dans un autre État membre. Cela est confirmé par de nombreuses jurisprudences relatives à la reconnaissance d’une nationalité attribuée par un autre État membre depuis l’arrêt Micheletti (C-369/90). 

Or, plus encore que la reconnaissance de la qualité de réfugié, fondée sur des règles de droit international et de droit européen communes, l’attribution de la nationalité est l’exercice de la souveraineté nationale. Pourquoi la nationalité attribuée par un autre État membre devrait-elle être reconnue, sous la seule réserve d’une atteinte disproportionnée au droit de l’Union, et pas la qualité de réfugié attribuée par ce même État membre et qui remplace précisément la nationalité de l’État d’origine qui ne protège plus cette personne ? 

En outre, au-delà de la nationalité, la jurisprudence impose également la reconnaissance mutuelle d’autres actes de l’état civil, alors même que l’ordre public de l’État requis pourrait être en cause. Il peut s’agir d’un mariage entre personnes de même sexe (Coman, C-673/16), de l’acte de naissance d’un enfant de couple lesbien (Stolichna obshtina, rayon « Pancharevo », C-490/20), ou encore, durant cette année 2024, de l’obligation pour l’État membre d’origine (Roumanie) de reconnaître et d’inscrire dans l’acte de naissance un changement de prénom et d’identité de genre d’une femme devenue homme, changement établi dans un autre État membre dont l’intéressé a également la nationalité (Royaume-Uni) (Mirin, C-4/23). La reconnaissance mutuelle des actes de l’état civil est également facilitée par le règlement visant à favoriser la libre circulation des citoyens en simplifiant les conditions de présentation de certains documents publics dans l’Union européenne (2016/1191). Certes, ce règlement porte uniquement sur la reconnaissance de l’instrumentum, non sur le contenu de l’acte, mais il participe à la libre circulation des statuts. Certes aussi, ces textes et jurisprudences concernent des citoyens européens et sont liés à la libre circulation des personnes, mais rien n’indique que ces raisonnements de reconnaissance ne pourraient s’étendre aux ressortissants d’États tiers, en particulier lorsqu’ils sont réfugiés reconnus. Ce qui est en cause, ce n’est pas la reconnaissance de la nationalité de la personne, mais la « nationalité » de l’acte, c’est-à-dire l’État d’origine de l’acte examiné dans l’État requis, tous deux États membres de l’Union européenne. Or, en qualité de réfugiés, tous les documents d’identité de ces personnes sont nécessairement délivrés par les autorités du pays d’accueil, conformément au prescrit de la convention de Genève (articles 25 et 27). Du reste, comme indiqué, le droit international privé, dans son ensemble, évolue vers cette « reconnaissance des situations », en ce compris à la suite d’actes ou décisions établis dans des États tiers, ce qui n’est même pas le cas pour notre sujet limité aux décisions sur le statut de réfugié prises dans un État membre.

Le droit de l’Union européenne

Au-delà du droit international privé, l’aporie de l’absence de reconnaissance mutuelle des décisions relatives au statut de réfugié pose également question du point de vue du droit de l’Union européenne. La contradiction avec d’autres aspects du droit de l’Union en général et de l’Espace de liberté, de sécurité et de justice en particulier, est tout aussi évidente. La libre circulation et la reconnaissance mutuelle sont au cœur de la construction de cet espace, que la Cour a été amenée à appeler « le territoire de l’Union pris dans son ensemble » (point 66 de l’arrêt Dereci, C-256/11,). Au-delà de la libre circulation des personnes, ce sont les principes mêmes d’un espace commun, d’un territoire commun qui sont mis en question. 

Bien que déplaisant, un rapprochement avec les marchandises est éclairant. Non seulement, selon la jurisprudence classique de la Cour depuis Cassis de Dijon (1979, 120/78), toute entrave disproportionnée à la libre circulation des marchandises produites dans un autre État membre, quelle qu’en soit la forme, est proscrite, mais même les marchandises importées d’États tiers bénéficient, après contrôle dans un État membre, de la libre circulation appelée « en libre pratique ». Ainsi, une voiture, produite dans un État tiers (Japon), entrée légalement dans un État membre (Allemagne), en respectant les critères de qualité et les droits de douane requis, sera considérée comme « en libre pratique ». Cela signifie que, ayant subi les contrôles de qualité requis dans un État membre, cette voiture pourra circuler, se trouver, être utilisée ou vendue dans un autre État membre. Un réfugié, officiellement reconnu tel, selon des critères communs aux États membres inscrits dans la directive qualification, mérite-t-il moins d’être « en libre pratique » et de pouvoir circuler et séjourner librement sur l’ensemble du territoire de l’Union ? 

Certes, cette liberté de circulation sera possible en qualité de résident de longue durée car le champ personnel de la directive résidents de longue durée a été étendu aux réfugiés (art. 3, § 3, sous c). Mais son application requiert, pour tous, y compris pour les réfugiés, une résidence régulière préalable de cinq ans dans le premier État membre. Cette condition de durée pourrait être supprimée, ou à tout le moins réduite, par exemple à deux ans, comme le font certains États en matière d’accès à la nationalité, appliquant en cela l’invitation expresse de la convention de Genève à faciliter la naturalisation des réfugiés (art. 34). Dans sa communication du 23 septembre 2020 « un nouveau Pacte sur la migration et l’asile », la Commission avait proposé de « modifier la directive sur les résidents de longue durée de sorte que les bénéficiaires d’une protection internationale soient incités à rester dans l’État membre leur ayant accordé cette protection, avec la perspective d’obtenir le statut de résident de longue durée au bout de trois ans de résidence légale et ininterrompue dans cet État membre » (p. 7). Cette idée n’a pas été retenue dans la proposition de refonte de la directive résidents de longue durée, actuellement sous négociation. 

Pour autant, il n’est pas certain que le refus de principe de la reconnaissance mutuelle des décisions relatives au statut de réfugié puisse, dans le futur, résister à l’entrée en vigueur du Pacte sur la migration et l’asile. Dans ses conclusions dans la première affaire (C-753/22), l’avocate générale Medina, après avoir constaté l’aporie de la seule reconnaissance des décisions négatives en matière de statut de réfugié et l’absence d’application du principe de reconnaissance mutuelle de toute décision sur le statut de réfugié, ajoutait que, « bien que des auteurs aient avancé des arguments convaincants, de lege ferenda, pour justifier la nécessité de consacrer un tel principe, le législateur de l’Union, malgré les efforts de la Commission, n’a pas pris de mesures définitives en ce sens ». Suivant les conclusions de l’avocate générale, la Cour affirme que, tant sur la base du droit primaire, en ce compris la Charte, que sur la base du droit dérivé, « le législateur de l’Union n’a pas, à ce stade, posé de principe selon lequel les États membres seraient tenus de reconnaître de manière automatique les décisions d’octroi du statut de réfugié adoptées par un autre État membre, ni précisé les modalités de mise en œuvre d’un tel principe » (QY, point 68 de l’arrêt). La Cour pourra-t-elle maintenir cette affirmation de principe après la mise en application du Pacte ? Peut-elle encore se limiter au constat que « le législateur de l’Union n’a, en l’état actuel du régime d’asile européen commun, pas encore concrétisé complètement l’objectif vers lequel tend l’article 78, paragraphe 2, sous a), TFUE, à savoir un statut uniforme d’asile en faveur de ressortissants de pays tiers, valable dans toute l’Union » (point 68) ? Ou, à l’inverse, la Cour devra-t-elle constater l’existence d’un statut « uniforme » d’asile et, en conséquence, admettre la reconnaissance automatique de toute décision d’octroi du statut de réfugié qui pourrait se déduire de plano du principe de reconnaissance mutuelle ?

Avec le Pacte, n’avons-nous pas atteint le niveau d’intégration de la politique européenne d’asile attendu en 2018 par l’ancien avocat général Melchior Wathelet dans ses conclusions sur l’affaire Jawo (C-163/17) ? Il écrivait :

« Certes, seule l’adoption d’une véritable politique sur la protection internationale au sein de l’Union, avec son propre budget, qui garantirait des conditions de vie minimales et uniformes aux bénéficiaires de cette protection, permettrait de réduire, voire d’éliminer, la survenance d’affaires (dans lesquelles la confiance mutuelle doit être mise en question au regard des droits fondamentaux), en assurant que le principe de solidarité et de partage équitable des responsabilités entre les États membres consacré par l’article 80 TFUE est une réalité au bénéfice non seulement des États membres, mais surtout des êtres humains en cause. Toutefois, dans cette attente (probablement longue !), il appartient aux États membres, y compris les juridictions nationales, d’assurer la pleine efficacité des normes actuellement en vigueur » (point 145). 

Cette réflexion, suscitée à l’époque par les dérives du mécanisme de Dublin, nous paraît transposable à la question de la reconnaissance des décisions d’octroi du statut de réfugié, ainsi du reste qu’à l’ensemble de la politique migratoire et d’asile de l’Union, à la suite de l’adoption du Pacte. De même que l’article 72 TFUE permet, pour motif de sécurité interne, selon la Commission, certaines mesures nationales dérogatoires comme les « pushbacks » de migrants instrumentalisés, de même le Pacte pourrait permettre la reconnaissance mutuelle des décisions. On pourrait certes considérer que le contenu du Pacte n’atteint pas encore le niveau d’une véritable politique commune au vu des analyses faites par la doctrine (voy. la série dans ce blog sur le Pacte sur la migration et l’asile). Toutefois, sur la forme du droit dérivé par des règlements et sur les objectifs inscrits expressément dans les textes, on pourrait aussi considérer que les conditions suffisantes pour la reconnaissance mutuelle des décision relatives au statut de réfugié sont remplies. L’objectif n’est-il pas de « renforcer la confiance mutuelle entre les États membres » par « une approche globale en matière de gestion de l’asile et de la migration » (nouveau règlement relatif à la gestion de l’asile et de la migration 2024/1351, considérant 2, abrogeant le reglement Dublin) ? On peut bien sûr admettre avec Daniel Thym qu’en avalisant les dérogations pour motif de sécurité nationale sur la base de l’article 72 TFUE, la Commission a franchi le Rubicon et qu’il y a un renversement de la dynamique d’interprétation en ce que « the time of dynamic interpretation to the benefit of migrants may have come to an end » (voy. Thym dans ce blog). Mais c’est à la Cour de justice de l’Union, conjointement avec la Cour européenne des droits de l’homme et les juridictions nationales, que reviendra le rôle final d’interprétation des textes. 

En guise de conclusion 

Fin janvier 2025, à l’occasion de la séance d’engagement solennel de la présidente et des membres de la Commission européenne devant la Cour de justice de l’Union européenne, dans son allocution, le Président Koen Lenaerts rappelait que 

« [c]es dernières années ont aussi été marquées par des atteintes portées, de l’intérieur ou de l’extérieur de l’Europe, aux valeurs de l’État de droit ainsi qu’aux libertés et aux droits fondamentaux. Face à ces atteintes, qui menacent nos systèmes et nos institutions démocratiques, et qui sont souvent exacerbées par des campagnes de désinformation et de déstabilisation véhiculées à travers certains réseaux sociaux, il est du devoir collectif des institutions européennes de continuer à œuvrer quotidiennement à la préservation du socle de valeurs qui fonde notre ordre juridique commun ». 

Ceci s’applique également au rôle de la Cour de justice. Tout en veillant à maintenir un équilibre entre, d’un côté le réalisme d’une politique migratoire, respectueuse des souverainetés et, de l’autre côté, la protection des droits fondamentaux des personnes concernées, la Cour doit maintenir une cohérence dans l’interprétation du droit de l’Union, en ce compris de l’Espace de liberté de sécurité et de justice en ses différentes composantes. Un des principes de base de cette cohérence est précisément la reconnaissance mutuelle.