Disapplication Unbound

Asylum seekers who come from a safe country of origin (SCO) may be detained under the so-called border procedure, a special asylum procedure provided for by Article 43 of Directive 2013/32/EU (Asylum Procedures Directive – APD). Article 36 APD establishes a protective scheme which permits setting aside the presumption of safety of the country of origin in individual cases. For the CJEU, this level of protection is insufficient. With its ruling of 1 August 2025, in the joined cases of Alace and Canpelli, it added another protective layer based on Article 37 APD. According to the ruling, national courts are required to review whether a SCO designation meets the criteria of Article 37, and, where it does not, disapply it. Neither the discretionary nature of the designation nor its inclusion in a legislative act is of relevance: since EU law is supreme, first-instance courts are bound to disapply conflicting provisions directly, without recourse to constitutional courts.

Legal scholars welcomed this ruling as a “total victory” for liberals supporting human rights and the independence of the judiciary. But the ruling fails to acknowledge that Article 37 APD is not unconditional: its direct effect is, at best, dubious. Second, in Member States like Italy, where the judiciary makes extensive use of disapplication in asylum matters, the laissez-faire approach of the CJEU paves the way for legal uncertainty and exposes judges to populist attacks.

Taking the direct effect for granted

According to CJEU’s settled case law, the duty to disapply is based not only on the primacy of EU law but also on its direct effect (Poplawski II [2019], §§ 59-69, is the leading case). This implies that EU directives can be invoked before national courts only when they are sufficiently precise, i.e. setting out an obligation in unequivocal terms, unconditional, i.e. not requiring any national implementing measure (e.g. RTS [2021], § 46), and not allowing any discretionary power (Link Logistic [2018], § 48).

Article 37 APD permits Member States to designate a country as safe in accordance with the criteria set out in Annex I to the APD. The national designating authorities need to provide an up-to-date assessment that takes into account, inter alia, the “legal situation, the application of the law in a democratic system and the general political circumstances” of the third country, as well as the “general and consistent” absence of persecution. It is evident that these are very broad criteria, which leave Member States a significant margin of discretion, and that Article 37 involves national implementing measures. Is it possible to consider Article 37 APD directly applicable?

In the Alace ruling, the CJEU delivers an affirmative but reticent response. According to § 63, Member States’ «discretion does not affect the obligation of every national court to give full effect to the provisions of Directive 2013/32, by disapplying as required, of its own motion, any provision of national legislation (…) which is contrary to the provisions of that directive with direct effect, without it having to request or await the prior setting aside of the provision of national legislation at issue by legislative or other constitutional means». This peremptory formula is taken directly from Simmenthal (1978). Usually, in the judgments of the last decade, this formula appears in the context of a broader reasoning explaining why an EU rule can be considered to have direct effect, according to the criteria developed in the subsequent case law. Not in this case, however. This omission is problematic. By ignoring the discretion of Member States, the supremacy of EU law risks undermining both that discretion and the domestic separation of powers.

As a rule, where a matter is governed exclusively by domestic law, the existence of legislative or administrative discretion means that judicial review is limited: the competent courts may examine the legality of the decision but must refrain from intruding into the sphere of discretion, the exercise of which requires democratic legitimacy. When that matter is also governed by EU law, that rule – meant to preserve the separation of powers – begins to falter. All too often, national courts intensify their judicial review: they interpret the direct effect of EU law as authorising them to substitute their own assessment for that of the other branches. For this reason, the Italian Court of Cassation – in a ruling of 30 December 2024, addressed in vain to the CJEU – proposed limiting judicial review to “extreme cases” of manifestly unreasonable designations. The CJEU itself is aware of the problem. In a rare case where a directive was denied direct effect, it admitted: if disapplication is based on an EU provision that is not unconditional, this “would in practice lead to the suppression of the discretionary power conferred exclusively on national legislators” (Link Logistic, § 54).

Yet, the suppression of that discretionary power is precisely what is conveyed by Alace. After failing to clarify why Article 37 APD would have direct effect, the CJEU adds that, to make judicial review effective, national courts “may verify whether that designation complies with the material conditions laid down in Annex I to that directive, by taking into account other information which it may itself have gathered” (§ 86). The Court thus clearly refers to substitutive judicial review, which would in principle only be possible when the decision does not imply any margin of discretion. As a result, Alace allows national courts to assume the role of governments and legislators in conducting the safety assessment underlying the designation. Is the CJEU ready to apply the same standard of review vis-à-vis the EU list of SCOs when it will be approved?

Fuelling counter-majoritarian activism

This form of heightened judicial review has become a problem of its own in Italy, from which the case originates and on which it will have significant effects. The Italian judiciary has organised itself along political lines and migration is a politically salient area. In 2017, the Gentiloni Government established specialised sections of first-instance courts to deal with immigration issues. Most of those sections swiftly aligned with an “universalist” vision of rights: human rights considerations must take absolute precedence in disapplying formal laws, especially when the law is very restrictive, as in the field of asylum. Eurostat confirmed that in 2024, Italian courts upheld 84.7% of asylum appeals, by far the highest rate in the EU, compared with 28.7% in Germany, 21.6% in France, 4.5% in Spain and an EU rate of 27.1%.

One of the favourite targets of Italian courts is the border procedure, both when applied in Italy and, even more so, to Albania. And one of the court’s favourite “weapons of resistance” to the Italian government is disapplication. This process is more effective in offering migrants immediate protection than a referral to the CJEU would be; and it is more “certain” in its outcome than a referral to the Constitutional Court, which is perceived as too deferential in these matters.

Accordingly, since September 2023, when the right-wing Meloni Government introduced cogent rules on the border procedure, Italian courts have systematically disapplied different aspects of that regime. Initially, they paralyzed the border procedure by targeting the government’s rules on the financial guarantee as alternative to detentions: a measure that Article 8(4) of Directive 2013/33/EU allows but courts set aside on the basis of a dubious interpretation of the MSF ruling. Then, in autumn 2024, when the implementation of border procedures following the Italy-Albania Protocol began, the focus shifted to SCO rules: courts disapplied them on the basis of a broad interpretation of the CV judgment. As these examples illustrate, Italian courts have no timidity in delivering pro-migrant and anti-government interpretation of EU law whenever they find it convenient, even when the conditions of acte clair or act éclairé (i.e. unequivocal or already settled EU law) are not met (for these conditions, see Consorzio Italian Management).

Alace condones all this: it embraces the expansive reading of the CV judgment followed by Italian courts (with its “all or nothing” approach to exceptions that, in my understanding, contradicts the objective of SCO rules) and affirms an unlimited form of disapplication. Italian courts will continue to disregard SCO designations in the name of an EU provision (Article 37 APD) that is far from unconditional. And they will continue to do so even without a “substantive examination” of the request for protection (§ 84) – a bold concession by the CJEU, which departs from the procedural “principle of demand” and, therefore, creates tensions with the procedural autonomy of Member States. As a result, the case-by-case process of disapplication continues. This, in turn, inevitably feeds legal uncertainty, administrative chaos and the de facto paralysis of the border procedure. Ironically, all this occurs in the same Member State that, from June 2026, will have to carry out the highest number of annual border procedures in the EU.

What kind of liberalism?

The Alace ruling is thus legally weak. It is not only elusive on the most crucial issue – whether Article 37 APD is directly applicable – but also disregards Article 36 APD, which fully protects genuine applicants from inaccurate SCO designations. Instead, it focuses on Article 37 APD to endorse the invasive control that only Italian courts (so far) carry out on designations. By disconnecting that legality review from the “substantive examination” of the asylum claim, i.e. from the applicant’s entitlement to protection, this approach falls into a contradiction: insofar as it goes beyond Article 36, Article 37 can only serve to extend the protection to non-genuine applicants. This clearly undermines the objective of the Directive, which is not only to speed up the examination of applications, as the CJEU held (§ 101), but also to prevent non-genuine applicants from engaging in secondary movements and absconding to avoid return.

As I have argued elsewhere, judicial review of SCO designation is necessary to serve a super-individual purpose: namely, to prevent national groups of asylum seekers from being directed towards an incorrect asylum procedure. If the problem is not the protection of individual rights but of national groups, case-by-case disapplication is structurally inadequate. The appropriate remedy is to annul the incorrect designation with the erga omnes powers of constitutional courts. The alliance with those courts is also necessary for the CJEU to promote consistent and uniform implementation of EU law at national level.

On the contrary, in the Alace case, the CJEU preferred to empower its long-standing allies, national lower courts. That choice is not only based on legally questionable assumptions (see above). It is also problematic in the current context of the crisis of liberal democracies. If there is a close correlation between «the rise of the unelected and the wave of populism» (e.g. Kosaf et al.), then the Alace ruling is a gift to populism. It conveys the image of an almost omnipotent judiciary which, while demanding respect for its independence, infringes upon the prerogatives of the other branches in the name of fundamental rights and EU law. This law, those rights and that judiciary are thus at risk of being perceived as biased and further delegitimised. The tensions with domestic constitutional systems and the separation of powers are at risk of being exacerbated and further exposed.

To counter the populist narrative, European citizens need good politics to explain why certain migration policies are wrong, counterproductive or both. But they also need our legal culture to be more than just militant liberalism, exclusively focused on the ambition to expand the reach of individual rights and the role of courts. What is required is a liberal vision that is genuinely and robustly democratic.