Prof. Francesco Maiani, University of Lausanne

“Solidarity” and “responsibility” were dominant themes in the discussions that led to the new Pact. During the crisis, responsibility foundered amidst falling standards and wave-through practices, while solidarity was shown to be tenuous indeed. After the crisis, failure to find an agreed balance between responsibility and solidarity sealed the fate of the 2016 Commission proposals.

In 2020, the Commission went so far as to state that “[t]here is currently no effective solidarity mechanism in place, and no efficient rule on responsibility”, and promised a “fresh start” entailing the “aboli[tion of] the Dublin Regulation” and the institution of a new solidarity mechanism.

The new Asylum and Migration Management Regulation (EU) 2024/1351 (hereafter “the Regulation” or “AMMR”) keeps both promises, formally speaking: on July 1, 2026, it will abolish the Dublin III Regulation (EU) No 604/2013 (hereafter “DRIII”) and introduce new rules on responsibility as well as a new governance framework and permanent solidarity mechanism (see Art. 1 and 83ff AMMR).

Will this have a “transformative impact” and “[set] the stage for fair, efficient, and sustainable migration management over the long term” as claimed? While the proof of the proverbial pudding will be in the eating, I will attempt to gauge the main innovations – or lack thereof – introduced by the Regulation. In this post, I will address the new Dublin rules. A forthcoming post will examine the solidarity mechanism and the governance framework intimately linked to it.


Path-dependent policymaking: “We can Dublin again”

It would be misleading to discuss the innovations of the Regulation without pointing out its path-dependence. The EU legislator has again confirmed the same system that was established by the 1990 Dublin Convention, and that has by common consensus – including per the Commission’s own appraisal – performed poorly since 1995:

  • The system is still based on “objective” criteria independent from the subjective wishes of applicants (see recital 36 AMMR). Indeed, the idea that applicants have no right to choose their Member States of destination is repeatedly expressed in the legislative package (see e.g. recital 26, Art. 19(1)(c) and Art. 67(3) AMMR; recital 14 of the Reception Conditions Directive (EU) 2024/1346).
  • The criteria themselves, and the discretionary clauses, remain by and large the same.
  • The outline of Dublin procedures will also remain largely unaltered.

Instead of suggesting that the Dublin III Regulation would be “abolished”, EU Institutions might perhaps have borrowed the slogan used by Dublin City Council to invite the citizenry back to the streets after the pandemic: “We can Dublin again”. This is, indeed, the fundamental policy choice made in respect of responsibility-sharing.

The scope of the new Dublin rules

Assuming that Ireland opts in, and that Denmark and the four EFTA States elect to remain Dublin associates, the territorial scope of the Dublin system will remain unaltered (see recitals 90-94 AMMR).

All applications made in the territory of the Member States will trigger the application of the system, including those made “at the border or in the transit zones”. The fiction of non-entry introduced by Article 6 of the Screening  Regulation (EU) 2024/1356 will change nothing in this regard except delaying the application of Dublin rules. Indeed, while under screening, applicants will be entitled to make an application and their data will be stored in EURODAC. At the end of the screening process, they will be able to register the application and trigger the application of Dublin rules (see Art. 9 and 18(1) and (2) Screening Regulation; Art. 38(1) AMMR).

The personal scope of the Dublin system will be significantly extended. Under current rules, applicants and persons whose application has been withdrawn or rejected are subject to Dublin transfers, but beneficiaries of international protection are not. By contrast, under the AMMR, beneficiaries of protection will be subject to take back procedures if they move without authorization to another Member State. Art. 36(1)(c) AMMR makes this explicit for persons admitted under EU or national resettlement schemes (see also Art. 18 and 22 of the new EURODAC Regulation 2024/1358). The applicability of take back procedures to the other beneficiaries of protection results implicitly from Art. 36(1)(b), which refers “an applicant or a third-country national or a stateless person in relation to whom [a] Member State has been indicated as responsible” (emphasis added). This wording removes the textual elements of Art. 18 DRIII that made take back procedures inapplicable to beneficiaries of protection according to the Court of justice. Furthermore, recital 43 AMMR explains that “responsibility” is to be “determined only once, unless one of the cessation grounds […] applies”. Thus, once a Member State will be designated as responsible for an applicant, such responsibility will endure as status evolves until one of the cessation causes exhaustively enumerated in Art. 37 AMMR intervenes (e.g. another Member State grants a residence permit).

Annex I, paragraph 1(e) of the Qualification Regulation 2024/1347 – referring explicitly to the “take back” of beneficiaries – confirms this interpretation, which is also consistent with EURODAC rules whereby the data concerning an applicant are “marked” and stored when protection is granted (see Art. 31 and recital 66 of the new EURODAC Regulation 2024/1358).

Subjecting beneficiaries of protection to Dublin transfers is but part of a broader strategy to fight against their secondary movements (see Art. 27 and 40 the Qualification Regulation), as further detailed in a forthcoming post in this Series.

New duties and the “consequences” of non-compliance

The intention of limiting movement into the Dublin area also inspires a group of other amendments. Under Art. 17(1) and (2) AMMR, applicants will be newly obliged to make and register their application in the Member State of first entry. Under Art. 17(4) they will have to remain there during the take charge procedure, then be present in the responsible State, where the application must be lodged (see also Art. 9(1) and (2)(h), 10(1) and 28(2) of the Procedures Regulation (EU) 2024/1348). Non-compliance will entail, per recital 58 AMMR, “appropriate and proportionate procedural competences”. Such “consequences” must be reconstructed from a puzzle of legal provisions scattered across several acts.

According to Article 18(1) AMMR, from the moment that a transfer decision is notified, reception conditions are withdrawn as long as the applicant is not present in the “right” Member State. There are several caveats: Member States must in any case guarantee a “standard of living in accordance with Union law, including the Charter, and international obligations” (Art. 18(1)); victims of trafficking must be exempted (Art. 18(3)); any sanction must take into account individual circumstances and be proportionate (Art. 18(4)).

Bizarrely, Art. 21 of the new Reception Conditions Directive – which applies when the applicant is not present in the responsible State, i.e. in a subset of the cases covered by Art. 18(1) AMMR – foresees the same sanction but without the second and third caveats. Is it a badly drafted reference back to Art. 18 AMMR, or did the legislator wish, for unfathomable reasons, to punish some secondary movements more harshly?

Be that as it may, three comments may be made at this stage. First, reducing the benefits of an applicant based on a transfer decision that is not yet final, and may yet be struck down as illegal, is highly questionable. Second, in light of the caveats above, the potential for administrative complexity, litigation, and inconsistent decision-making is considerable. Third, until the EU ensures dignified living conditions throughout the Dublin area, it is far from self-evident that these provisions will dissuade their addressees from engaging in secondary movements: faced with destitution in the “right” Member State, applicants may find the prospect of a standard of living in accordance with the Charter in the “wrong” Member State quite attractive.

The “hostile environment policy” embedded in Art. 18(1) AMMR is part of a broader array of measures targeting persons engaging in secondary movements. On the one hand, Art. 36(3) AMMR no longer includes the language of Art. 18(2) DRIII to the effect that, on take back, files closed as implicitly withdrawn must be reopened and first instance negative decisions must be amenable to judicial review. On the other hand, Art. 41 of the Procedures Regulation stipulates that Member States must consider as implicitly withdrawn, and thus inadmissible, applications “lodged” in a Member State other than those referred to by Art. 17(1) and (2) AMMR, where the applicant does not remain in that State pending the take charge procedure or transfer. The wording of the provision lacks precision, since Art. 17 dictates where applicants should “make” and “register” an application, not where they should “lodge” one (see Art. 28(2) Procedures Regulation). Assuming that the situations envisioned are indeed those where an applicant does not “make” an application in the “right” State and/or moves away during Dublin procedures, it is unclear why this type of secondary movements should attract such a draconian sanction, and others – e.g. moving away from the responsible State pending the examination of the application – not.

Be that as it may, these provisions weaken considerably the guarantee that Member States “shall examine” the applications made to any of them, and therefore entail a serious risk of indirect refoulement in the operation of Dublin rules. They embed in the fabric of CEAS law the very same “discontinuation” practices that the Commission itself once condemned as incompatible with effective access to status determination.

Touch-ups to the criteria, security cases, and obstacles to transfer

As noted at the outset, the general outline of the hierarchy of criteria remains unaltered.

A few changes are introduced to expand responsibility-allocation based on “meaningful links” (see recital 55 AMMR) – a notion hitherto unknown to EU asylum legislation and reminiscent of UNHCR ExCom conclusions No 15:

  • The definition of family members will also include family ties formed in transit countries en route to the Dublin area, as opposed to only those formed in the country of origin (Art. 1(8) and recital 52 AMMR; see also e.g. Art. 3(9) Qualification Regulation).
  • While Art. 9 DRIII foresees the reunification of applicants with family members who are present in a Member State as beneficiaries of protection, Art. 26 AMMR will also cover family members who were beneficiaries of protection, and have been naturalised, as well as family members who have long-term resident status.
  • Immediately after the criteria based on possession of a residence document or visa, slightly expanded, Art. 30 introduces a new criterion based on having obtained a diploma in a Member State less than six years before the registration of the application. Art. 1(15) AMMR provides the relevant definition of diploma.

These innovations are welcome, but it is difficult to anticipate their practical impact. In particular, while the possibility to be reunited with long-term resident family members might constitute a relevant innovation, that status is currently “under-used” and “too difficult” to acquire according to the Commission itself.

Recital 54 AMMR addresses the paramount issue of proving family ties. But while it expresses the laudable intention of “allowing for swifter family reunification”, the AMMR includes no binding rules preventing Member States from imposing excessive evidentiary requirements. There are recommendations in the recital itself, an obligation to give reasons (Art. 40(6) AMMR), plus an unclear duty to “prioritise” the examination family-based requests. None of this will matter unless the Commission uses its powers under Art. 40(4) to facilitate the recognition of family ties.

Meanwhile, it is worth noting that the legislator has maintained the exclusion of marital relations from the family definition of the dependency clause of Art. 34 AMMR. It therefore remains the will of the legislator, with the apparent blessing of the European Court of Justice, that a severely traumatized or ill applicant be “normally kept or brought together” with her sister, but be kept apart or transferred away from her husband.

Coming to entry-based criteria, two changes deserve mention:

  • Under Art. 25(5) AMMR, absent a relative or family member, unaccompanied minors will be attributed to the Member State where they first applied, subject to an individualised best interest determination (see recital 53). This new provision, whose aim is to discourage secondary movements, overrules an old judgment of the ECJ which de facto exempted UMAs from unvoluntary take charge transfers precisely to preserve their best interest. In this connection, it is worth mentioning that under the new EURODAC Regulation children above six – instead of fourteen – will be fingerprinted, and that while “no form of force” shall be used, a “proportionate degree of coercion” may be applied (Art. 14 EURODAC Regulation).
  • Again to discourage secondary movements, the irregular entry criterion is maintained and expanded (Art. 33 AMMR). The new provision clarifies that the criterion applies to persons disembarked in the context of SAR operations and, save in these last situations, makes the criterion applicable for twenty months after entry instead of twelve currently.

Two further changes will have a bearing on the application of the criteria:

  • Under Article 16(4) AMMR, every applicant will have to undergo a security check, either under Art. 15 of the Screening Regulation, or right after the registration of the first application, or both. Upon finding “reasonable grounds to consider that the applicant poses a threat to internal security”, the Member State carrying out the check will become responsible. No other criterion will be considered, and no take charge procedure will be initiated. It is worth pointing out that this does not exclude the applicant from the scope of the Dublin system. Indeed, responsibility will have been determined, the responsible State will have to examine the application, and the applicant will be subject to take back transfers in case of unauthorized movement.
  • Like Art. 3(2) DRIII, Article 16(3) AMMR will determine the course of action when it is “impossible” to transfer an applicant to the responsible State on account of the conditions prevailing there. The responsibility rule remains the same, but the new provision does away with any reference to “systemic flaws” and instead states that transfers are “impossible” whenever they entail a real risk of inhuman or degrading treatment. This is welcome, given the recent tendency of the Court to lose sight of this core human rights standard, but it is not a foregone conclusion that the notion of “systemic flaws” will wholly disappear from Dublin case-law.

Procedural “simplifications” and cutbacks in due process rights

“Streamlining” Dublin procedures in a bid for greater efficiency (recital 64 AMMR) is an important Leitmotiv of the AMMR.

First, deadlines are shortened across the board. For instance, the ordinary time limit for submitting a take charge request is two months instead of three currently (Art. 38(1) AMMR). Two amendments counter this trend:

  • In order to discourage evasion of the Regulation, the time limit to carry out a transfer in cases where the transferee absconds is prolonged – from 18 months to three years – and any form of resistance is equated to absconding (Art. 46(2) AMMR);
  • Furthermore, to avoid putting a premium on secondary movements, Member States missing the deadline for requesting a take-back will no longer become responsible. This is a return to the past, since the Dublin II Regulation foresaw no deadline in the matter and introducing one was deemed necessary to make the procedure “more efficient and rapid”.

Much is being made of the fact that the current take back procedure “will be replaced by a simple take back notification”. Still, the AMMR foresees the possibility for the “notified” State to raise objections based on the same grounds that can today justify the rejection of a take back request (see Art. 41(3) and 46(2) AMMR; see C-228/21, para. 96f). The only change of some substance, in this regard, is the fact that cessation rules are stricter under the new Regulation (see Art. 37(4) AMMR).

Far more important than these technicalities, the AMMR heavily modifies procedural rights of persons subjected to the Dublin system. The right to information and to an interview are expanded, as are special guarantees for minors, and a qualified right to free legal aid at first instance is introduced (see Art. 19 to 22 AMMR). Recitals 38-39 elaborate on the benefits of such bountiful “human rights frontloading”.

At the same time, Art. 43 AMMR drastically reduces the appeal rights against transfer decisions: it retains the least protective of the options currently envisaged by Art. 27 DRIII for the suspensive effect of appeals, introduces a strict deadline to submit an appeal, which Member States may set as low as one week, and reduces the scope of the remedy to:

  • Risks of inhuman or degrading treatment;
  • “Circumstances subsequent to the transfer decision that are decisive for the correct application” of the Regulation;
  • Incorrect application of the family criteria.

The one-week time-limit may, or may not, fall foul of the right to an effective remedy depending on how one reads the case-law of the Court of Justice (see Steve Peers’ excellent and concise analysis).

The rules on the scope of the remedy do prima facie violate that right. True, the Court itself had similarly reduced the scope of remedies against transfers in Abdullahi, and when it backtracked in Ghezelbash, it did so based on changed legislative intent, not on the Charter. Given that legislative intent has once again changed with the new Regulation, the Court might revert to Abdullahi in one form or another.  This would, however, place EU law on a collision course with the ECHR.

Under Art. 13 ECHR, anyone having an arguable claim that her Convention rights have been violated has the right to an effective remedy. Under Art. 43 AMMR transferees may only invoke Art. 3 ECHR. They are not even entitled to invoke the right to family life, which Dublin transfers routinely interfere with and which, bizarrely, recital 62 AMMR expressly mentions. In this last regard, it may be useful to recall that the question of whether a take charge transfer violates the family criteria of the Regulation (Art. 43(1)(c) AMMR) is not coextensive with the question of whether the same transfer violates Art. 8 ECHR.

If Art. 43 AMMR indeed violates Art. 13 ECHR then, recital 62 notwithstanding, it also contravenes Art. 47 of the Charter, which incorporates and exceeds the minimum standard of that provision  (see e.g. para. 44 ff of the Berlioz judgment).

Even if Art. 43 AMMR could somehow be reconciled with superior law, its introduction will mark a stark reduction in applicants’ rights. Save for the grounds enumerated in Art. 43(1), transferees will be deprived of the right to a full judicial review of decisions that affect them profoundly. No matter how serious the errors in the application of the criteria, no matter how grievous the violations of their procedural rights at first instance, they will have no remedy. Seen in this light, the expansion of the so-called procedural “rights” of applicants in first-instance Dublin procedures seems like a cruel jest, to say nothing of the reference to “the rule of law” found in recital 5.


As far as responsibility-determination in the CEAS is concerned, the adoption of the Pact constitutes everything but a “complete overhaul”. Part III of the Asylum and Migration Management Regulation is the evil twin of the current Dublin Regulation.

While the 2013 recast had expanded the rights of the persons subjected to the Regulation, the 2024 recast is starkly regressive. True, there is a timid expansion of criteria based on “meaningful links” and the introduction of legal assistance at first instance. But this is little more than sugar-coating for the bitter concoction of punitive measures aimed at “secondary movers” and for the drastic cutback of transferees’ right to a judge. Overall, the new rules multiply the risks of human rights violations and endanger the raison d’être of Dublin rules, i.e. to guarantee access to an asylum procedure to every person seeking protection in the EU+. Every effort will have to be expended to promote a human-rights compliant implementation of this flawed body of rules. To this end, even the torrent of self-absolutory statements found in the preamble may be of some use.

Will the many “challenges” of the Dublin system be addressed at least, as recital 40 posits? It is far too early to say, but there is every reason to doubt it. As noted, the system is fundamentally unchanged. It will be as complex to implement as its predecessor, and it has little chances of winning the hearts and minds of applicants. Nor is it safe to assume that the “consequences” it threatens to secondary movers will daunt them into obedience, so long as the factors driving secondary movements remain unaddressed.

The new Dublin rules might even aggravate the burden-concentrating effects inherent in the system. Indeed, this might well be the combined effect of new duties to apply in the first State of entry, shortened take charge deadlines, take back procedures expanded to beneficiaries of protection and restricted cessation clauses. This lends an even greater salience to the question of whether the new solidarity mechanism, which will be examined in an upcoming post, has the potential to promote a fair sharing of responsibility among the Member States.

Responsibility-determination under the new Asylum and Migration Management Regulation: plus ça change…

Welcome to the Practice Source & House of Butter’s Global Law Blogs directory.

At and The House of Butter Blog we have been writing about lawyers, legal publishing and legal information on a daily basis for over 20 years.

We have decided to compile what we think, are, the best law / legal blogs written across language by lawyers, barristers & law firm (teams) from around the world.