Introduction
This essay provides a legal advisory for Maya Antonescu, a client facing a 75% reduction in her housing benefit due to an automated decision by the National Welfare Authority (NWA) in Member State G. The analysis draws on EU law principles, particularly in relation to the fictional Directive 2024/77/EU (the Algorithmic Transparency Directive or ATD), which aims to ensure procedural protections for individuals affected by automated administrative decisions. Following the group roles assigned, the essay is structured into three key sections: indirect effect, supremacy, and the preliminary ruling procedure. These sections collectively advise on Maya’s options to enforce her rights under Articles 3 and 4 of the ATD, considering the incomplete transposition via the Digital Administration and Public Services Act 2025 (DAPSA) and the constitutional challenges under Article 72 of Member State G’s Constitution. The discussion is grounded in established EU law doctrines, highlighting how Maya might challenge the generic notification and the lack of suspension pending review. By examining these aspects, the essay outlines practical steps for Maya, while acknowledging the fictional nature of the scenario and relying on real EU legal precedents for analytical depth.
Indirect Effect
In advising Maya on enforcing her rights under the ATD, it is essential first to explore indirect effect, a principle established in EU law that requires national courts to interpret domestic legislation in conformity with EU directives where possible (Von Colson and Kamann v Land Nordrhein-Westfalen, 1984). This approach prioritises harmonious interpretation over direct enforcement, potentially allowing Maya to rely on the DAPSA as a vehicle for ATD compliance without immediately invoking the Directive itself.
Section 12 of the DAPSA mandates “appropriate notification” for automated decisions, but it lacks specificity on content, such as individualised explanations required by Article 3 of the ATD. A court could interpret “appropriate notification” broadly to include the ATD’s demands for an explanation of principal factors and their impact, in language intelligible to non-specialists. This conforms with the EU principle of consistent interpretation, which national judges must apply to the fullest extent possible, even if the national law was not intended to implement the directive (Marleasing SA v La Comercial Internacional de Alimentacion SA, 1990). For Maya, this means arguing before a national court that the NWA’s generic letter fails this interpreted standard, potentially entitling her to a more detailed explanation without proving direct effect.
However, indirect effect has limits, particularly when national law explicitly contradicts the directive or when criminal liability is involved. The DAPSA preamble expressly excludes suspension of adverse effects pending review, citing constitutional incompatibility under Article 72, which emphasises administrative finality. Here, consistent interpretation reaches its boundary, as forcing suspension would contradict the preamble’s clear intent, and EU law prohibits interpretations that impose unforeseen criminal penalties (Criminal proceedings against Kolpinghuis Nijmegen BV, 1987). Since Maya’s case involves welfare entitlements rather than criminal matters, this limitation is less directly applicable, but it underscores that indirect effect cannot override explicit national exclusions.
If indirect effect fails, direct effect becomes relevant. The CJEU’s fictional ruling in Fischer confirms Article 3’s direct effect, being clear, precise, and unconditional (Van Gend en Loos v Nederlandse Administratie der Belastingen, 1963). Maya can thus enforce Article 3 directly against the NWA, a public body, claiming her right to an individualised explanation. Article 4, however, lacks such confirmation; the Supreme Administrative Court deemed it not directly effective due to discretion in “suitably qualified official.” Arguably, this phrase introduces some margin, but core obligations like suspension pending review could still be precise enough for direct effect, drawing parallels with cases like Marshall v Southampton and South West Hampshire Area Health Authority, 1986.
Should direct enforcement not apply to Article 4, Maya has alternatives: state liability for non-transposition post-deadline (Francovich v Italy, 1991), potentially claiming damages for eviction threats caused by the failure to suspend effects. Additionally, she could seek judicial review in national courts, invoking EU loyalty principles under Article 4(3) TEU to pressure compliance. In summary, indirect effect offers a primary route for Article 3, with direct effect as a fallback, while Article 4 may require pursuing damages or further litigation.
Supremacy
The principle of supremacy in EU law, as articulated in Costa v ENEL (1964), requires that EU provisions prevail over conflicting national law, including constitutional norms. For Maya, this means advising whether Member State G’s courts must disapply the DAPSA, particularly its preamble’s constitutional limitation under Article 72, in favour of the ATD. Supremacy imposes a duty on national courts to set aside incompatible domestic rules without awaiting legislative repeal (Simmenthal SpA v Italian Minister for Finance, 1978), ensuring EU law’s uniform application.
Practically, this duty to disapply means that in proceedings before the NWA or national courts, judges must ignore the DAPSA preamble’s exclusion of suspension if it conflicts with Article 4 ATD. For Maya, a court could thus order suspension of her benefit reduction pending human review, treating the ATD as overriding. This extends to administrative bodies like the NWA, which must apply EU law directly where possible, though ultimate enforcement often requires judicial intervention.
Engaging with the constitutional argument, Article 72 is framed as a “constitutional identity” clause, interpreted by Member State G’s Constitutional Court as protecting national competence in welfare procedures, citing limited EU powers under Article 153 TFEU. This echoes real debates in cases like Landtová (2011), where national identity under Article 4(2) TEU can limit EU intrusion into core constitutional elements. However, the argument likely fails here. Supremacy is absolute; national constitutions cannot unilaterally exempt areas from EU law once competence is exercised (Melloni v Ministerio Fiscal, 2013). The ATD, as a directive under EU internal market or data protection competences (potentially Articles 16 or 114 TFEU), overrides supporting competences in social policy. The CJEU has rejected similar claims when they undermine EU objectives, as in Commission v Hungary (2020) on higher education, emphasising that identity clauses cannot create blanket derogations.
Furthermore, Article 72’s focus on “administrative certainty” appears more as a sovereignty claim than a true competence limitation. EU law permits procedural harmonisation where necessary for rights protection, as in algorithmic decisions affecting welfare (GDPR recitals on automated processing). Thus, courts must disapply the preamble, enabling Maya to enforce ATD rights. If unsuccessful domestically, infringement proceedings by the Commission could follow, though this offers indirect relief. Overall, supremacy strengthens Maya’s position, rendering the constitutional barrier ineffective.
Preliminary Ruling Procedure
The preliminary ruling procedure under Article 267 TFEU obliges national courts to refer questions on EU law interpretation to the CJEU, ensuring uniformity. Advising Maya, we assess whether the Supreme Administrative Court of Member State G acted lawfully in its handling of the ATD, and the implications.
For Article 3, the court declined a reference, citing acte clair under CILFIT (1982), which allows non-referral if the provision’s meaning is obvious considering prior CJEU rulings. The litigation history complicates this: the Westham court’s contradiction of the Fischer ruling suggests inconsistency, yet the Court of Appeal’s alignment and the Supreme Court’s view post-Fischer indicate the matter may indeed be clear. However, CILFIT requires the point to be equally obvious to courts of other Member States, and the Bauer contradiction highlights potential doubt. Arguably, the Supreme Court erred by not referring, as acte clair should not be invoked amid domestic divergence (Foto-Frost v Hauptzollamt Lübeck-Ost, 1987). This could undermine the ruling’s validity for Maya, allowing her to challenge it on procedural grounds.
Regarding Article 4, the Supreme Court resolved it without reference, holding no direct effect and upholding constitutional limits. As a court of last instance, it was obliged to refer unless acte clair or acte éclairé applies (CILFIT criteria). No prior CJEU ruling exists on Article 4, and terms like “suitably qualified official” introduce ambiguity, making non-referral unlawful. The lack of explanation for non-referral breaches transparency duties, potentially violating EU law effectiveness.
Consequences for Maya include weakened enforceability; she could seek a fresh reference in her own case from a lower court, or pursue infringement actions via the Commission. Remedies include state liability for judicial breaches (Köbler v Austria, 2003), claiming damages for prolonged benefit cuts. Practically, I recommend Maya initiates judicial review, requesting a reference to the CJEU on Article 4, leveraging the procedural flaws to strengthen her claim for suspension and explanation.
Conclusion
In conclusion, Maya has viable avenues under EU law to challenge the NWA’s decision. Indirect effect offers a pathway to interpret the DAPSA conformably with Article 3, with direct effect as a robust alternative; supremacy mandates disapplying constitutional barriers for both articles; and procedural lapses in preliminary rulings provide grounds for further challenges. These elements collectively empower Maya to seek an individualised explanation, suspension pending review, and potential damages, highlighting EU law’s protective role in automated decisions. However, success depends on national judicial willingness, underscoring ongoing tensions between EU integration and national identity. Implications include reinforcing the need for full directive transposition to avoid such disputes, ultimately benefiting individuals like Maya in an increasingly digital administrative landscape.
References
- Craig, P. and de Búrca, G. (2020) EU Law: Text, Cases, and Materials. 7th edn. Oxford University Press.
- European Court of Justice (1963) Van Gend en Loos v Nederlandse Administratie der Belastingen, Case 26/62, ECLI:EU:C:1963:1.
- European Court of Justice (1964) Costa v ENEL, Case 6/64, ECLI:EU:C:1964:66.
- European Court of Justice (1982) CILFIT v Ministry of Health, Case 283/81, ECLI:EU:C:1982:335.
- European Court of Justice (1984) Von Colson and Kamann v Land Nordrhein-Westfalen, Case 14/83, ECLI:EU:C:1984:153.
- European Court of Justice (1986) Marshall v Southampton and South West Hampshire Area Health Authority, Case 152/84, ECLI:EU:C:1986:84.
- European Court of Justice (1987) Criminal proceedings against Kolpinghuis Nijmegen BV, Case 80/86, ECLI:EU:C:1987:431.
- European Court of Justice (1987) Foto-Frost v Hauptzollamt Lübeck-Ost, Case 314/85, ECLI:EU:C:1987:452.
- European Court of Justice (1990) Marleasing SA v La Comercial Internacional de Alimentacion SA, Case C-106/89, ECLI:EU:C:1990:395.
- European Court of Justice (1991) Francovich v Italy, Joined Cases C-6/90 and C-9/90, ECLI:EU:C:1991:428.
- European Court of Justice (2003) Köbler v Austria, Case C-224/01, ECLI:EU:C:2003:513.
- European Court of Justice (2011) Landtová, Case C-399/09, ECLI:EU:C:2011:415.
- European Court of Justice (2013) Melloni v Ministerio Fiscal, Case C-399/11, ECLI:EU:C:2013:107.
- European Court of Justice (2020) Commission v Hungary, Case C-66/18, ECLI:EU:C:2020:792.
