When Does Credit Stop Being Finance? Debt Collection before the CJEU

 

Cătălin Gabriel Stănescu, Reader in Law at Newcastle University and member of the European Commission’s Financial Services Users Group.

Photo credit: BNP Paribas HQ, by Boubloub, via Wikimedia Commons

A pending reference from the Greek Council of State brings an old question back before the Court of Justice: when credit is transferred to a purchaser and its administration and recovery are entrusted to a separate company, how should EU financial law classify the activities of that company? In Trapeza tis Ellados II, C-210/26, the Court is asked whether an undertaking engaged exclusively in credit servicing, without assuming credit risk, qualifies as a financial institution under the Regulation (EU) No 575/2013 on prudential requirements for credit institutions (Capital Requirements Regulation, CRR).

In a 2021 article,  I asked whether informal debt collection qualified as a financial service under EU law and what that qualification would mean from a legal perspective (pp 535-538). The answer was less straightforward than the question suggested. Treating collection as a financial service might activate additional protection for those subjected to it, but it might also make that protection depend on the origin of the debt. The difficulty itself was older: my 2015 analysis of MKG-Kraftfahrzeuge-Factoring had already shown how factoring, debt purchasing and collection could overlap without becoming identical.

Two recent CJEU judgments make the pending reference particularly interesting. In Kosmiro, the Court describes debt collection, for VAT purposes, as financial transactions designed to obtain payment (para 50). In IFIS NPL INVESTING, it excludes the purchase or management of non-performing loans from the economic activities covered by a particular definition of financial institution (para 37). The reference in Trapeza II concerns a servicer whose activities include administering and collecting debts, negotiating with borrowers and restructuring repayment obligations. The question is not simply whether these judgments can be reconciled, but how far a classification reached under one EU regulatory regime can be carried over into another. I argue that regulatory differentiation does not require conceptual separation. Lending, assignment, servicing and collection may attract different legal classifications while remaining connected stages in the life of the same credit relationship.

1.     The question in Trapeza II

The reference in Trapeza II was lodged by the Greek Council of State on 13 March 2026 and is currently pending. It arises from sanctions imposed by the Bank of Greece on the former managing director of a credit-servicing company (summary of the request, paras 6-8). The company was authorised and supervised by the Bank but operated exclusively as a servicer. Its activities included legal and accounting monitoring, recovery, negotiations with debtors, settlements and debt restructuring (paras 14-15). It assumed no credit risk (para 29).

The question is whether such a company is a “financial institution” under Article 4(1)(26) of CRR. The definition applicable at the material time covered an undertaking, other than an institution or pure industrial holding company, whose principal activity was acquiring holdings or pursuing one or more activities listed in points 2-12 and 15 of Annex I to Directive 2013/36/EU on access to the activity of credit institutions and the prudential supervision of credit institutions (Capital Requirements Directive, CRD). The dispute turns on two possible Annex I classifications: point 2, ‘lending’, which expressly includes factoring with or without recourse, and point 11, ‘portfolio management and advice’.

The Bank of Greece characterises the work as administrative factoring and, where securitised loans are concerned, portfolio management (summary of the request, para 10). The referring court provisionally rejects both propositions. That conclusion forms part of a broader line of reasoning. Drawing on Lineas – Concessões de Transportes and Others, joined Cases C-207/22, C-267/22 and C-290/22, judgment of 26 October 2023, the referring court starts from the proposition that an undertaking which does not carry out activities in the financial sector cannot qualify as a financial institution (para 23). It then reads the Annex I activities as, in principle, activities linked to the banking or financial sector. It invokes Directive (EU) 2021/2167 on credit servicers and credit purchasers as confirmation, in its view, that credit servicers are not financial institutions within the meaning of the CRR and do not simultaneously carry out activities in the financial sector (para 27). It ultimately concludes that companies performing only the specified servicing activities, without assuming credit risk, ‘do not carry out activities in the banking or financial services sectors’ (para 29). The classification is of immediate importance because it determines whether the sanctions applicable to financial institutions could be used against the applicant. The reference nevertheless raises a broader issue: how EU financial law classifies an undertaking that performs functions arising from a credit relationship while neither originating the credit nor assuming the associated credit risk.

There is also a temporal limit. Regulation (EU) 2024/1623 (CRR III) amended that definition with effect from 1 January 2025. The current Article 4(1)(26) also expressly includes an “ancillary services undertaking” as an alternative basis for classification as a financial institution. Article 4(1)(18), as amended, defines such an undertaking by reference to a principal activity consisting, inter alia, of a direct extension of banking or other activities ancillary to banking. That express route to financial-institution status was absent from the version of Article 4(1)(26) applicable to the dispute. The judgment will therefore interpret a superseded definition. Its reasoning on factoring, portfolio management and credit servicing may nevertheless remain relevant under current law, even though the amended definition now provides an additional route to financial-institution status.

2.     Factoring as debt collection

Kosmiro, C-232/24, judgment of 23 October 2025, concerned two factoring models. Under trade factoring, the factor purchased invoices and assumed the debtors’ default risk. Under invoice factoring, the debts remained with the client, which retained the default risk while obtaining finance secured by those debts. In both arrangements, the factor administered and collected the debts (paras 11-13). The VAT question was whether the relevant fees constituted consideration for exempt credit or taxable debt collection under Article 135(1)(b) and (d) of Directive 2006/112/EC on the common system of value added tax (VAT Directive).

The Court began from the debt-collection exception to the exemption for transactions concerning debts. Exemptions are interpreted strictly, while the exception, which restores taxation, is interpreted broadly (paras 46, 49). Debt collection refers to “financial transactions designed to obtain payment of a pecuniary debt” (para 50). That concept encompasses both factoring models (paras 51-53).

That conclusion builds on an established line of VAT-related activities case law. In MKG-Kraftfahrzeuge-Factoring, C-305/01, judgment of 26 June 2003, the Court held that factoring is merely a variant of the more general concept of debt collection, whatever the manner in which it is carried out (paras 76-77). In AXA UK, C-175/09, judgment of 28 October 2010, it added that debt collection need not concern an overdue debt or involve coercion (paras 31-35). The resulting VAT concept is therefore not confined to coercive recovery of debts already in default.

In Kosmiro, the Court held that, in principle, factoring constitutes a single economic transaction whose main purpose is to transfer responsibility for recovery and collection from the client to the factor and splitting that transaction would be artificial (para 60). In trade factoring, the funds paid for the purchased debts were not a loan to the client and there was consequently no credit relationship between factor and client (para 61). In invoice factoring, by contrast, financing was provided, but not independently of the debt-collection service, of which it was the corollary (paras 62-63). The Court distinguished arrangements whose essential purpose was independently the provision of financing (paras 64-66). Importantly, it rejected the argument that all financing arrangements must receive the same VAT treatment: the legislative distinction between debt collection and the granting of credit permits “financing arrangements subject to VAT and financing arrangements exempt from VAT” to coexist (para 67). The relevant invoice-factoring fees therefore constituted consideration for a single and indivisible debt-collection service subject to VAT (paras 68-69). 

There are firm limits. The Court classifies the transactions for VAT purposes but does not define the institutional perimeter of prudential supervision. It establishes that factoring falls within debt collection, not that every collector performs factoring. Nor does the debtor purchase the service examined in Kosmiro: the factor supplies it to the creditor-client. Finally, purchasing debts need not supply a remunerated service at all. In GFKL Financial Services, C-93/10, judgment of 27 October 2011, a price reflecting the actual economic value of defaulted claims was not consideration for a service to the seller (paras 22-26).

Kosmiro consequently provides no trans-substantive definition of financial services. Its narrower significance is nevertheless important. EU VAT law describes debt collection as consisting of financial transactions designed to obtain payment of a pecuniary debt, and Kosmiro demonstrates that financing and collection can form part of the same economic arrangement without becoming identical legal services. What it undermines is therefore any assumption that the functions of financing and collection must always be legally or economically separate.

3.     The classification puzzle

The difficulty is that the Court has recently used a different regulatory definition to reach a conclusion that appears to leave little room for treating the purchase or management of non-performing loans as a financial activity. In IFIS NPL INVESTING, C-65/25, judgment of 11 June 2026, the Court stated that the economic activities covered by Directive (EU) 2015/849 on the prevention of the use of the financial system for the purposes of money laundering or terrorist financing (Fourth Anti-Money Laundering Directive, AMLD) do not include the purchase or management of non-performing loans (para 37). The conclusion is relevant for Trapeza II because the Greek undertaking did not purchase the claims it serviced, but its activities included their administration, collection, negotiation and restructuring. The question is whether the exclusion of NPL “management” in IFIS also encompasses those distinct servicing functions.

IFIS arose from Italian legislation governing bulk assignments of non-performing loans during the period before the expiry of the deadline for transposing Directive 2021/2167. The referring court asked, among other things, whether entities carrying out such assignments had to be subject to anti-money-laundering supervision.

In this instance, the definition before the Court was Article 3(2)(a) of AMLD, not Article 4(1)(26) CRR. It covered an undertaking other than a credit institution which carried out one or more activities listed in points 2-12, 14 and 15 of CRD Annex I. That statutory route nevertheless matters for Trapeza II because it incorporates the two Annex I activities at issue there: point 2, which includes factoring, and point 11, portfolio management and advice. Yet the Court does not analyse either Annex I activity separately. Its unqualified exclusion of NPL purchase and management consequently raises the question whether either activity can bring an undertaking within the corresponding CRR definition.

That passage is difficult to confine to its facts. Unlike the historical CRR definition, the AMLD did not require an Annex I activity to constitute the undertaking’s principal activity. If the purchase or management of NPLs does not fall within the relevant Annex I activities at all, making that activity the undertaking’s principal business cannot by itself bring it within the CRR definition. IFIS therefore points strongly towards exclusion if the Greek servicer’s work is characterised simply as the management of NPLs.

IFIS bears directly on Trapeza II, but does not resolve it. The dispute in IFIS arose from bulk assignments of NPL portfolios, whereas the Greek undertaking in Trapeza II neither acquired the claims it serviced nor assumed their credit risk. Paragraph 37 of the IFIS judgment extends the Court’s conclusion beyond the purchase of NPLs to their “management”, but does not examine what that activity encompasses. In particular, the Court did not consider whether the administration, collection, negotiation and restructuring performed by a credit servicer could constitute factoring under point 2 of Annex I to the CRD, or whether the management of a portfolio of credit claims could fall within “portfolio management and advice” under point 11 of that Annex. Instead, the reasoning moves directly from the scope of the AMLD to the exclusion of NPL purchase and management. The temporal analysis in IFIS does not qualify that conclusion. Paragraphs 31–36 of the judgment explain why Directive 2021/2167 did not apply to the transfers at issue and could not be applied retroactively; paragraph 37 states a separate conclusion concerning the scope of the AMLD. That conclusion must therefore be confronted on its own terms.

The two definitions also perform different functions. The AMLD identifies obliged entities for the purposes of anti-money-laundering supervision, whereas the CRR definition operates within a system of prudential regulation and financial-sector supervision. In Lineas, the Court interpreted the historical CRR definition in that regulatory context. It concluded that an undertaking whose sole activity consisted in acquiring holdings in companies which did not carry out activities in the financial sector was not a “financial institution”. Its reasoning emphasised the distinction drawn by the CRR between holdings in financial-sector entities and holdings outside that sector, and the inconsistency that would result from subjecting the latter to rules designed for the former (paras 69-81).

Those different regulatory purposes matter, but they do not by themselves resolve Trapeza II. Both definitions incorporate points 2 and 11 of the same CRD Annex I. If those common terms produce different classifications in the AMLD and CRR contexts, the difference must be explained through the wording, structure and purpose of the respective regimes. Regulatory purpose alone cannot determine what counts as factoring or portfolio management.

Trapeza II thus offers the Court an opportunity to provide the analysis compressed into paragraph 37 of IFIS. It could explain whether servicing without credit risk can constitute factoring under point 2; what “portfolio management and advice” means in point 11 when the portfolio consists of credit claims; and whether the servicer’s combination of collection, negotiation and restructuring affects either classification. If the result remains exclusion, the reasoning should identify why those activities fall outside the particular definition rather than turn that exclusion into a general proposition that servicing and collection are external to finance.

Consumer law reveals the relevance of that distinction. In Gelvora, C-357/16, judgment of 20 July 2017, the Court held that the concept of commercial practices directly connected with the sale of a product extends beyond the conclusion of the contract to its performance, to include measures taken to obtain payment (paras 19-21). The claims at issue originated in consumer credit, and their recovery could fall within the Unfair Commercial Practices Directive even after assignment to a debt collection agency (paras 22-25). Excluding credit-repayment transactions following assignment could undermine consumer protection by encouraging traders to separate the recovery phase in order to escape the Directive (para 28). The Directive could also apply where unilateral collection measures were pursued in parallel with formal enforcement (paras 29-31).

Gelvora does not classify debt collection as a financial service. Paragraph 25 appears to distinguish the two activities: the collection agency does not provide consumer credit “as such”. Its significance lies instead in the continuity recognised for the particular purposes of consumer protection. Assignment changes the creditor and recovery changes the activity, but neither necessarily breaks the legal connection between collection and the credit from which the payment obligation arose.

INKO, C-127/15, judgment of 8 December 2016, provides a complementary, activity-specific example. Under Directive 2008/48/EC on credit agreements for consumers, a debt collection agency acting on behalf of a lender to conclude a rescheduling agreement for unpaid credit must be categorised as a “credit intermediary” within the meaning of Article 3(f) (paras 42-44). The classification follows not from debt collection alone but from the additional function performed by the agency within the existing credit relationship: concluding the rescheduling agreement on the creditor’s behalf. The Court further held that acting as a credit intermediary only in an ancillary capacity may remove particular pre-contractual information obligations without altering that classification (paras 47-53).

4.     Continuity without regulatory identity

The case law does not produce a single answer to whether collection is a “financial service”. And, perhaps, it should not. Lending creates credit; purchasing transfers a claim; servicing manages rights and obligations arising from it; collection seeks payment; and enforcement invokes legally compulsory procedures. Those functions may overlap, but they are not identical and may justify different authorisation, conduct, prudential and tax rules.

The difficulty begins when a conclusion designed for one of those purposes is detached from it. Under Directive 2021/2167, a credit purchaser does not require authorisation to operate because it acquires existing non-performing credit rather than creating new credit and raises negligible systemic risk (Recital 40). IFIS excludes the purchase or management of NPLs from the economic activities relevant to the AMLD, while Trapeza II asks whether particular servicing functions fall within the historical CRR definition. A particular acquisition of debts may not constitute a remunerated service for VAT purposes. None of those situations establishes, without another legal step, that collection has become external to the financial relationship from which the payment obligation arose.

Directive 2021/2167 embodies this distinction. A credit servicer manages and enforces rights and obligations related to a creditor’s rights under a non-performing credit agreement, or to the agreement itself, on behalf of a credit purchaser. Credit-servicing activities include collecting or recovering payments, specified renegotiation, complaints administration and informing borrowers of changes in interest rates or charges or of payments due (Article 3(8) and (9)(a)-(d)). Rather than simply designating servicers as credit institutions or financial institutions, the Directive subjects them to a bespoke authorisation and supervisory regime (Articles 4-14 and 21-23).

At the same time, the regulatory object remains connected to the underlying credit relationship. Article 10 of Directive 2021/2167 regulates the conduct of purchasers and servicers towards borrowers, requiring good faith, fairness and professionalism and prohibiting harassment, coercion and undue influence. It also requires specified information to be communicated after transfer and always before the first debt collection. The limited prudential concerns associated with credit purchasers do not, however, displace the conduct obligations applicable to them and their servicers. Yet Union and national consumer-protection rules continue to apply and borrowers retain the rights arising from the initial credit agreement. Regulatory identity changes; legal continuity remains.

This also separates different forms of risk. Whether an undertaking assumes the economic loss resulting from non-payment may be highly relevant to prudential classification. It says much less about the conduct risks created by controlling communications with borrowers, repayment negotiations and collection. A servicer’s absence of credit risk may therefore matter to one regulatory classification without establishing that its activities are unrelated to the underlying financial relationship. Undertakings might pose little or no systemic risk, but that does not mean no risk to consumer-debtors. Directive 2021/2167 responds to such conduct risks through authorisation and conduct obligations tailored to servicing rather than lending.

It should be emphasized that the Directive is selective evidence, not a complete solution. Its framework concerns non-performing credit agreements issued by credit institutions established in the Union and contains significant exclusions (Articles 1-2). It does not create a common regime for the collection of rent, utility debts or other non-credit claims. That limitation leaves intact a problem I identified in my earlier work: if protection against abusive collection depends on the regulatory identity of the debt or collector, comparable collection practices may attract different legal responses for reasons having little to do with the risks those practices create.

The answer is not to declare every collector a financial institution, but to resist allowing institutional definitions to determine questions they were not designed to answer. Prudential classification should follow prudential purposes. Consumer-facing collection should remain subject to effective conduct controls because of the activity and its connection to the underlying obligation. Where gaps remain outside the sectoral regime, they require direct regulation rather than conceptual disappearance.

Credit does not stop being finance at a single identifiable moment. Default, assignment and outsourcing redistribute its functions and risks. They may change the applicable service, institution and regulatory regime without erasing the relationship between the claim being collected and the transaction that created it. IFIS requires purpose-specific exclusions to be taken seriously. Kosmiro, Gelvora and INKO require equal caution before turning such an exclusion into a general separation between finance and collection. Trapeza II is an opportunity to articulate both propositions and to clarify a boundary that EU law has so far defined only imperfectly.

A candle in the dark: the CJEU rules against criminalising parents for smuggling their children

 

Professor Steve Peers, Royal Holloway University of London

Photo credit: Bologna, taken by Ввласенко, via Wikimedia Commons

Introduction

The world is increasingly bleak for those concerned about civil liberties, poverty or climate change – including for those concerned about the right to asylum and the rights of migrants more generally in Europe. Constraints on those rights have reached the point where a law supposedly intended to combat wicked gangs of people smugglers ended up potentially being applied to criminalise a mother who applied for asylum for assisting her little girl to enter a Member State illegally with her.

In today’s judgment in Kinsa, the Court of Justice ruled that this interpretation of EU law went too far. But there’s a twist: there’s a proposal to amend the legislation, tabled a few months after the Kinsa case was referred to the CJEU, and Member States tried to foresee the Court’s judgment, taking the opportunity to build themselves an escape route to get around it. Yet there’s a further twist here: today’s judgment implicitly appears to be aware of the Member States’ plans, and subtly attempts to block that escape route. It’s the latest move in an unstated game of cat-and-mouse between the EU legislature and judiciary. But above and beyond these games, the Court’s judgment also has broader implications for the highly controversial relationship between criminal law and immigration and asylum law.

 

Background

Current law

The current EU law on smuggling of migrants – or, more precisely, the facilitation of illegal entry or stay – is set out in two parallel linked laws adopted in 2002: a Directive setting out the prohibition on facilitating illegal entry or stay, and a Framework Decision providing for criminal offences based on the prohibition. According to the 2002 Directive, Member States must impose ‘appropriate sanctions’ on anyone who ‘intentionally assists a person who is not a national of a Member State to enter, or transit across, the territory of a Member State in breach of the laws of the State concerned on the entry or transit of aliens’ (illegal entry), and anyone ‘who, for financial gain, intentionally assists a person who is not a national of a Member State to reside within the territory of a Member State in breach of the laws of the State concerned on the residence of aliens’ (illegal stay; italics added).

While the prohibition on assisting illegal stay is subject to the requirement that such assistance must be ‘for financial gain’, the prohibition on facilitating illegal entry is not. However, the Directive goes on to say that any Member State ‘may decide not to impose sanctions’ in ‘illegal entry’ cases, ‘by applying its national law and practice for cases where the aim of the behaviour is to provide humanitarian assistance to the person concerned’.

The Framework Decision then requires Member States to penalise the facilitation of illegal entry or illegal stay (as defined in the Directive) via the  criminal law, although the level of that penalty is only set out for the most serious cases (where the offence was committed as part of an organised crime group, or ‘while endangering the lives of the persons who are the subject of the offence’). However, it includes a safeguard for asylum:

This framework Decision shall apply without prejudice to the protection afforded refugees and asylum seekers in accordance with international law on refugees or other international instruments relating to human rights, in particular Member States' compliance with their international obligations pursuant to Articles 31 and 33 of the 1951 Convention relating to the status of refugees, as amended by the Protocol of New York of 1967.

Proposed law

The Commission’s proposal to replace the two 2002 measures with a single Directive would make a number of changes to that law, but I’ll focus on the scope of criminalisation of assistance (for a full discussion of the proposal and the Member States’ response to it, see my analysis here).

On this point, the obligation to criminalise either illegal entry or illegal stay would be limited to ‘financial gain’ cases, and the preamble clarifies that these elements of the offence:

…will usually not be fulfilled when it comes to assistance among family members or the provision of humanitarian assistance or the support of basic human needs. Third-country nationals should not become criminally liable for having been the subject to such criminal offences. Moreover, it is not the purpose of this Directive to criminalise, on the one hand, assistance provided to family members and, on the other hand, humanitarian assistance or the support of basic human needs provided to third-country nationals in compliance with legal obligations.

A version of the current asylum safeguard would also appear in the preamble:

Member States should apply this Directive in accordance and in full compliance with the 1951 Convention Relating to the Status of Refugees as amended by the Protocol of New York of 1967, obligations related to access to international protection, in particular the principle of non-refoulement, and fundamental rights.

However, this is only the Commission’s proposal. EU legislation must also be agreed by the Council (ie Member States’ representatives) and the European Parliament. According to the Council’s version, agreed last December (after the Advocate-General’s opinion in the Kinsa case, but before today’s judgment) the preamble expressly reserves the possibility for Member States to criminalise support not motivated by financial benefits (new text underlined):

Assistance of unauthorised entry, transit or stay in the Union should constitute a criminal offence at least when there is a link with an actual or a promised financial or material benefit. This is without prejudice to the way in which Member States deal in their national law with facilitation conducts for which an actual or a promised financial or material benefit is not a constituent element of the offence.

A wholly new recital in the preamble makes the point even more clearly (italics are mine):

This Directive is an instrument of minimum harmonisation. As a consequence, Member States are free to adopt or maintain legislation providing for a broader incrimination than what is set out in this Directive, in the interests of enhancing the effectiveness of the fight against migrant smugglers. Member States should ensure that intentionally assisting a third-country national to enter, or transit across, or stay within the territory of any Member State in breach of relevant Union law or the laws of the Member State concerned on the entry, transit and stay of third-country nationals constitutes a criminal offence at least where the person who carries out the conduct requests, receives or accepts, directly or indirectly, a financial or material benefit, or a promise thereof, or carries out the conduct in order to obtain such a benefit. However, since this Directive is an instrument of minimum harmonisation, Member States are free to criminalise such conduct when no financial or another material benefit has been provided.

And the Commission’s proposed recital stating that family members should not themselves be criminalised, is replaced by a recital giving Member States more leeway to do just that:

Nothing in this Directive should be understood as requiring the criminalisation, on the one hand, of assistance provided to close family members and, on the other hand, of humanitarian assistance or the support of basic human needs provided to third-country nationals in compliance with the applicable national and international legal framework.

For its part, to date the European Parliament has not yet agreed its negotiation position on the issue, although there is a draft report on the proposal that exempts humanitarian assistance from criminal liability, and is critical of criminalising family members in the preamble:

(7a) Where an individual or organisation offers assistance to migrants through actions which are selfless and from which no financial or other material benefit is derived, or which are based on the principles of humanity, neutrality, impartiality and independence, then those actions should not be criminalised. Such actions may include the carrying out of search and rescue operations in accordance with international law, assistance to family members, assistance to third-country nationals in situations of danger or distress, the provision of legal assistance, medical assistance, healthcare, housing, shelter, clothing, food and water to third-country nationals or supporting third-country nationals in asserting their rights.

However, it remains to be seen whether the majority of MEPs support the draft report.

 

The Kinsa case

The mother being prosecuted in this case came originally from the Congo, flying to Bologna, Italy from Casablanca, with an 8-year-old and a 13-year-old girl (one was her daughter; the other was her niece, who she was looking after since the girl’s own mother died). (Some of the facts are omitted from the Court’s judgment; there’s more detail in the summary of the national court reference) Although she and the children did not enter via an unsafe vessel across the Mediterranean, they all travelled on false passports, and the Italian authorities decided that she should be prosecuted for breach of the EU Directive and Framework Decision, as implemented in national law, because she assisted the girls’ entry. She subsequently applied for asylum (the fate of that application is unknown).

In the criminal case, the national court had doubts that the EU legislation or its national implementation was compatible with the EU Charter of Fundamental Rights, particularly given that there was no mandatory exclusion for humanitarian assistance. It therefore asked the CJEU if the EU legislation was invalid, or alternatively if EU law precluded Italian law’s implementation of it on this point.

According to the Advocate-General’s opinion, the EU Directive was valid, and obliged Member States to criminalise a mother in these circumstances, although Member States must impose lower penalties in such cases, due to the Charter rule on proportionality of criminal penalties.

The CJEU took a radically different view.

It began by noting that EU legislation should be interpreted in conformity with the Charter as far as possible, in effect redrafting the national court’s questions to focus on Articles 7, 18 and (added by the Court) 24 of the Charter (respectively the rights to family life and asylum and rights of the child).

Interpreting the wording of the Directive, in the Court’s view, the prohibition of assistance to illegal entry was broad; it neither expressly included or excluded its application to those who assisted the illegal entry of children they had effective custody over. But the Court rejected the application of the prohibition of assistance to such cases, for several reasons.

First of all, it was inconsistent with the objectives of the EU law, which concerned assistance to illegal immigration, not the exercise of parental responsibility.

Secondly, this would contradict the right to family life (in particular the family unity of parents and children) and the rights of the child set out in the Charter (the right to protection and care, the best interests of the child, and the right of contact with parents); the latter right is also relevant to decisions addressed to adults which have major consequences for children. As the explanations to the Charter state, the UN Convention on the Rights of the Child (CRC) is relevant too; and that Convention refers to parents or guardians having the primary responsibility for children. It followed that the prohibition in EU law could not apply in these circumstances, even if the mother herself entered unlawfully.

Moreover, the Court ruled that applying the prohibition in such cases would undermine the ‘essence’ of these fundamental rights (see Article 52 of the Charter), given that a parent or guardian is simply exercising their responsibility to look after the children.

Thirdly, applying the prohibition to such case would breach the right to asylum in the Charter too; this right was relevant given the mother’s subsequent asylum application. Both the Charter and (as we have seen) the EU legislation concerned require compliance with the Refugee Convention. In particular, the Court referred to Article 31 of that Convention (also referred to explicitly in the EU legislation), which states that:

1. The Contracting States shall not impose penalties, on account of their illegal entry or presence, on refugees who, coming directly from a territory where their life or freedom was threatened in the sense of article 1 [ie the definition of ‘refugee’], enter or are present in their territory without authorization, provided they present themselves without delay to the authorities and show good cause for their illegal entry or presence.

Similarly, the Court referred to a number of related points: the obligation set out in the Schengen Borders Code to apply that Code in accordance with human rights, the Refugee Convention, and non-refoulement; the right to submit an asylum application despite any illegal stay, after which the stay could not be considered illegal until a first-instance decision on the application; the prohibition on Member States dissuading an applicant for applying for asylum, without reasonable justification (on the last two points, referring to its judgments in Commission v Hungary (Reception of applicants for international protection), Commission v Hungary (Criminalisation of assistance to asylum seekers) and Commission v Hungary (Declaration of intention prior to an application for asylum); it’s fair to assume that the Hungarian government’s observations in the Kinsa case were unpersuasive); the EU reception conditions Directive and qualification directive refer to the best interests of the child and family unity; and the Final Act to the Refugee Convention refers to family unity of refugees too.

So, in the Court’s view, as an asylum-seeker the mother could not be prosecuted either for her own illegal entry or for accompanying her daughter and niece.  

Fourthly, the Court noted that – as the Commission, and indeed many critics of EU law in this field point out – a narrower reading of the obligation to criminalise for assisting illegal entry is confirmed by the protocol on smuggling of migrants, attached to the UN Convention on organised crime, which the EU and Member States have ratified, and which the EU legislation was intended to give effect to. The Court noted that the intention of the protocol is to criminalise the smuggling of migrants, while protecting the rights of those migrants. However, it also points out that other conduct, such as trafficking in people or sexual exploitation of children, could be criminalised.  

Finally, the Court addressed remedies. Its interpretation of the EU legislation had addressed the argument that this legislation might be invalid for breach of human rights. As for the national law, Member States could not breach fundamental rights when implementing EU law; so when transposing the 2002 Directive, they could not include conduct not covered by it, in breach of Articles 7 and 24 the Charter. In any event, the Court ruled that those provisions of the Charter were, in effect, directly effective; if a national court could not interpret national law consistently with the Charter, it would have to disapply that conflicting national law, ie the principle of supremacy/primacy (referring to Egenberger on enforcing Charter equality rights; on that judgment see here). The judgment did not comment on whether or not Article 18 of the Charter could be used in the same way.

 

Comments

There’s…a lot to unpack here. I’ll discuss in turn: the family life and rights of the child aspects of the judgment; its broader implications for asylum and immigration law; and the even broader legal and political context of the Court’s judgment.

Family life and rights of the child

The Court’s first point (distinguishing between control of irregular migration and parental responsibility) is solely connected to the parent/child relationship, but its second point (the Charter rights to family life and rights of the child, plus the CRC) is potentially broader – as there are family relationships other than the parent/child link. Although the Court’s analysis of the issues is understandably focussed on the facts of this case, there may be other circumstances where it might be relevant by analogy: assistance for other family members, or assistance from a parent already on the territory. The Court’s strong emphasis on the rights of the child aspects (covering all three paragraphs of Article 24 of the Charter, plus the CRC) could be relevant to other issues besides irregular entry; and the very strong remedy – disapplying national law in breach of Articles 7 and 24 of the Charter – could be applicable to many EU law issues inside and outside the scope of immigration or asylum law. And the Court’s ruling that this issue goes to the ‘essence’ of the rights in question makes it impossible to justify a limitation on those rights. (The Court doesn’t discuss whether or not the restriction goes to the essence of the right to asylum or not)

Having said that, the issue of humanitarian assistance goes beyond family relationships: the impulse to rescue strangers is distinct from a mother’s (or father’s) love for their child. In the context of revising the EU legislation on this issue, could the EU legislature simply leave the possible criminalisation of family support out of the law altogether – leaving Member States free to decide whether to criminalise it, untroubled by the constraints of EU law, because they would not be implementing it? Equally, could they evade those constraints as regards humanitarian assistance generally – as the Council’s negotiation position on the Commission’s proposal to amend the law seeks to do – by simply leaving the issue outside the scope of that law? With one bound, could the Member States be free?

Not so fast. While the CJEU could perhaps have confined its analysis to the family life and rights of the child points, it did not – choosing also to answer the national court’s questions about the right to asylum, and doing so in a particularly broad way.

Right to asylum

The mother’s subsequent application for asylum is crucial to the impact of this case – giving the national court the opportunity to ask about the Charter right, thus allowing the CJEU to wax lyrical about a wide swathe of EU asylum and border control law, plus the Refugee Convention.

Starting with the Convention, the Court’s willingness to refer to and apply Article 31 of the Convention – ruling out the prosecution of the mother for illegal entry even though the national court did not ask about that (it’s not even clear if she was charged for it) – is striking. The Court has perhaps forgotten that back in 2014, in the case of Qurbani (discussed here) it claimed that it had no jurisdiction to interpret Article 31 of the Convention. This was very unconvincing, given the express or implied references to that Article in EU law (in the main text of the smuggling Framework Decision and implicitly in the human rights safeguard in the borders code, both referred to in this judgment; Article 31 is also mentioned in the preambles to other EU asylum law), and its obvious relevance to criminalisation of immigration.

Having finally worked up the enthusiasm to discuss Article 31, the Court applies it without much analysis – possibly because it was not even asked about criminal penalties for the mother’s own illegal entry, or perhaps because of the other legal factors it mentioned before reaching its conclusion that she could not be prosecuted for her own illegal entry or for assisting that of the children.

But let’s look at the Article 31 issues in this case. The Court obviously (and surely correctly) assumes that criminalisation of illegal entry constitutes a ‘penalty’ for the purposes of Article 31; the definition of ‘penalty’ is an issue as regards other restrictions following illegal entry, though. The Court also assumes that asylum-seekers – not only recognised refugees – are covered by Article 31.  Another key issue in the interpretation of Article 31 is the meaning of ‘coming directly’: here, the summary of the facts indicates that the family originated in Congo but flew to Italy from Morocco, and the details of their travel to Morocco are not mentioned. Arguably the Court has implicitly accepted that an asylum-seeker must still be regarded as ‘coming directly’ even though they went through an intermediary country, but does this only apply if direct flights were unavailable? (We don’t know if they were). Or if the asylum-seeker only transited there, or the stay was otherwise brief? Perhaps the story of refugees trapped in Casablanca, desperately trying to obtain documentation to travel further abroad, would be a good idea for a film.

Whether the mother’s life or freedom was threatened on Convention grounds is the issue to be decided in her substantive asylum application (she’s fleeing domestic violence, which was the subject of an important CJEU judgment last year). She’s certainly present without authorisation, and appears to have presented herself to the authorities without delay (although this may be more of an issue in other cases). Arguably, seeking asylum is in itself a ‘good cause’ for illegal entry or presence, although national interior ministries might well contest that. And the Court does not even mention Article 31(2) of the Convention, on restrictions on movement or possible movement to other countries – although these issues might arise in relation to detention cases (the mother was only detained for one day), or ‘safe third country’ arguments.

This detailed analysis of Article 31 may be begging the question, though, because, as mentioned already, the Court also mentioned a series of other legal factors in its assessment of possible criminal penalties. So presumably Article 31 is not the only issue to be taken into account in such cases, and the finer points of interpreting it sit alongside other issues to be interpreted (the borders code human rights safeguard; the legality of an asylum-seeker’s stay until the first instance decision, which is explicitly set out in EU asylum procedures law; and the prohibition of dissuading asylum applications). And while the Court does mention family unity and the rights of the child in this context, most of the points it makes here are relevant to any asylum application, and also to other types of assistance to asylum-seekers (cf the Court’s reference to the previous judgment against Hungary for criminalising assistance to asylum-seekers). The Court seems to have taken the opportunity arising from this specific case to reach some very general findings about criminalisation and border control in the context of asylum.

All of this encourages national courts to send the EU court further questions to clarify these points, in an environment where asylum-seekers’ access to the territory and the asylum procedure is often very difficult on the ground. But what about the EU legislature?

As noted already, the Member States’ negotiation position on the proposed new Directive on smuggling of persons takes the issue of family assistance in particular, and humanitarian assistance in general, outside the scope of the smuggling law, leaving it up to Member States to determine whether to criminalise such forms of assistance or not. But on the face of it, the Court’s judgment has thwarted this cunning plan – at least in cases where the assistance is offered to, or by, an asylum seeker. For the Court has linked its interpretation also to the Schengen Borders Code (where the human rights safeguard was untouched by amendments last year), (implicitly) the asylum procedures law, the reception conditions directive, and the qualification law – along with the Refugee Convention. So even if Member States win the battle on this point as regards the smuggling of migrants law, the Court has opened four more fronts – and its rulings on the point not only interpret secondary law, but also largely link that interpretation to EU primary law (the Charter, plus the quasi-primary law of the Refugee Convention – linked to the Charter and the EU Treaties). 

Nor could the Member States argue that the future Directive sets minimum standards only, leaving them free to criminalise what they want above that level, because the Kinsa judgment insists that the EU Charter constrains Member States' implementation of the law regardless.

The legislative mouse seems to have been caught by the judicial cat.

Finally, the Court’s interpretation of the UN Protocol is unhelpful for the border control industrial complex too. The Court is correct to point out that the protocol only requires the criminalisation of smuggling in the context of organised crime, although the judgment exaggerates the extent to which the protocol protects migrants’ rights. Again, these are general points not limited to family assistance, or even humanitarian assistance more generally; they could equally apply to criminalising the migrants as such.   

The broader context

Trying to fire a legal weapon devised to combat organised crime at the desperate mother and aunt of two young girls offers us a moment of rare moral clarity. The facts in this case are so sympathetic that the ruling in her favour may not be as controversial as if the facts had been somewhat different. Nevertheless, the CJEU eschewed the opportunity to limit its judgment to such compassionate cases, instead providing fuel for much more general legal challenges to border controls on human rights grounds, even apparently undercutting an attempt to legislate an escape from EU law constraints.

And the Court is surely well aware of the political environment in which it delivered its judgment. Even setting aside the assault on immigration and asylum rights across the Atlantic, the judgment comes shortly after a group of Member States’ leaders delivered a broadside against the case law of the European Court of Human Rights in immigration cases. The CJEU is a different court, but it’s unlikely that these leaders are too thrilled with its asylum and immigration case law either. Yet its response to this broadside was not to surrender and prepare to be boarded. It was to fire back.  

It’s not certain what the full scope of this judgment is; it’s unlikely that every CJEU migration judgment will be as liberal; and it’s possible that Member States may find a way to circumvent this judgment anyway – perhaps by simply ignoring it, as they do with pushbacks. But the judgment is at least a sign that all is not yet lost. In an increasingly dystopian world, it is at least a flicker of hope.

Further reading - for a comparative law angle, the Supreme Court of Canada's R v Appulonappa judgment 

Does the Court of Justice of the European Union Respect the Limits of EU Competence?

 

 

Vilija Vėlyvytė, British Academy Post-Doctoral Fellow, Somerville College, University of Oxford

 

Photo credit: civarmy, by wikipedia

 

My new book, ‘Judicial Authority in EU Internal Market Law: Implications for the Balance of Competences and Powers’ (Hart Publishing 2022), examines how the Court of Justice of the European Union expands EU competences through the interpretation of EU internal market law.

 

Much has been written about the EU’s so-called ‘competence creep’, describing the extension of EU competence (or power) to regulate the internal market into the areas in which the EU has not been granted explicit competence to act, such as public health, social security, and others. That discussion criticizes the EU legislature – Commission, Council and Parliament acting together – for failure to respect the limits of EU competence and the Court for failure to police those limits. Yet it largely overlooks a related, and equally important, question: does the Court itself observe the limits of EU competence in the interpretation of the rules of the internal market laid down in the Treaties? This question lies at the heart of my new book. The inquiry conducted in the book exposes the scale of the problem of the EU’s competence creep and demonstrates that the exercise by the Court of its interpretative authority is its major, and often dominant, cause.

 

The limits of EU competence are governed by the principles of conferral, subsidiarity and proportionality. Whilst the principle of conferral defines the scope of EU competence, the principles of subsidiarity and proportionality limit its exercise. The book examines to what extent the Court observes these principles in the interpretation of EU free movement rules. It argues that the Court’s observance of the three principles has been inconsistent, thereby creating constitutional tensions in the EU’s relationship with the Member States and upsetting the institutional balance of powers between the EU legislature and judiciary.

 

Observance of Conferral in Free Movement Case Law

 

According to the principle of conferral (Art 5(2) TEU), the EU can act only within the scope of the competences that Member States have conferred upon it in the Treaties. The Treaties contain a list of EU competences, dividing them into exclusive, shared and supplementary (see here). Notably, the EU does not have any real powers in the areas belonging to its supplementary competences, such as, for example, healthcare, education and, for the most part, social policy. Its action in these areas is limited to the adoption of measures of a recommendatory character.

 

The phrasing of the principle of conferral in the Treaties suggests that there is a clear line as to what the EU can and cannot do. But in reality this is far from true. In the way conferral is interpreted by the Court, it allows the EU to interfere into the areas that in principle fall outside the scope of its competences to the extent that such interference stems from the exercise of a competence that has been conferred upon the EU. For instance, the EU has not been granted competence to legislate in the area of public health. Yet it can enact such legislation as long as it contributes to the functioning of the internal market (see Tobacco Advertising II judgment, concerning the validity of EU directive on the advertising and sponsorship of tobacco products).

 

EU legislation aside, Member States must comply with the rules of the internal market laid down in the Treaties in all areas, regardless of whether these belong to EU or Member State scope of competence. So, for example, Member States retain the competence to regulate the exercise of collective labour rights, namely the right to collective bargaining and right to strike. However, they must regulate these rights in a way that does not violate the EU’s economic freedoms – freedom of establishment and of provision of services in particular (see Viking and Laval judgments).

 

The book argues that if conferral is to have a meaningful normative function in EU law, there needs to be a limit to how far EU internal market law can encroach into the areas of competence that should in principle be exempted from any binding interference on behalf of the EU. In the context of the judicial interpretation of free movement law, that limit lies in the conditions that trigger the application of free movement law. These are the conditions that have to be satisfied for the matter to fall within the scope of free movement. For instance, in order to benefit from the EU’s economic freedoms, the activities in question have to be of an economic, as opposed to social, nature and the challenged national law or practice has to constitute a restriction on those activities.

 

When interpreting the meaning of these conditions in a given case, the Court should adopt a systemic approach, contextualising them in the requirements of the principle of conferral. This presupposes a judicial interpretation of the concepts of ‘economic activity’ and ‘restriction’ that balances the objectives of the internal market inherent in these concepts with the concerns relating to the protection of national competence and regulatory autonomy. Such balancing exercise should be performed whenever free movement law collides with policy areas that are constitutionally sensitive from the EU law point of view. As a consequence, activities which have some economic features but are otherwise heavily regulated at the national level in the interests of social policy and solidarity, would fall outside the scope of the economic freedoms. A prominent example of such activities is the provision of public healthcare and education services.

 

Having conducted an inquiry into free movement case law involving three sensitive areas of national policy – healthcare, education and collective labour law, the book finds that the Court has systematically failed to engage in any meaningful reflection of competence concerns in the interpretation of the scope of free movement rules. This practice places the Court in tension with the principle of conferral.

 

The case law has also diminished the effectiveness of the legislative dimension of the principle of conferral. Specifically, it has opened up the possibility for matters that had been reserved to the regulatory prerogative of the Member States to be addressed at the EU’s legislative level through the functionally broad provisions of the Treaties (legal bases) dedicated to the regulation of the internal market. For instance, by holding that the provision of public healthcare is an economic activity, the Court sent a signal to the EU legislature that healthcare, as a service, could be legitimately legislated on under Art 59 or Art 114 TFEU. This enabled the adoption of EU Directive on Patients’ Rights, which governs access to cross-border healthcare, notwithstanding that EU-wide harmonisation in the area of public health is excluded by Art 168 TFEU.

 

Observance of Subsidiarity and Proportionality in Free Movement Case Law

 

The principles of subsidiarity (Art 5(3) TEU) and proportionality (Art 5(4) TEU) govern the exercise of EU competences. Subsidiarity applies to the areas of competence that are shared between the EU and Member States. It sets out the conditions under which the competence concerned should be exercised by the EU rather than Member States. Essentially, it stipulates that policy decisions should be taken at the level of the Member States unless action at the EU level could be more effective. Proportionality, on the other hand, controls the intensity of EU action. It limits intervention by the EU to what is necessary in light of the objectives pursued.

 

The book demonstrates that both principles are designed to protect national regulatory autonomy in politically sensitive areas of EU action, but they do so in slightly different ways. If subsidiarity is primarily concerned with the protection of the scope of national regulatory autonomy, proportionality ensures that EU regulatory intervention does not intrude excessively into the various values and interests protected via the exercise of national regulatory autonomy.

 

How do the principles of subsidiarity and proportionality manifest themselves in the judicial interpretation of the Treaties? They do at the stage of the judicial inquiry into the justification of national measures restrictive of free movement. The degree to which restrictive measures can be justified depends on the standard of judicial review employed by the Court. Accordingly, the standard of judicial review determines the space left for domestic regulatory action in the situation concerned. It thus affects both the scope of national regulatory autonomy (subsidiarity) and the interests protected within its exercise (proportionality). So, for example, when free movement law is construed to require that trade unions employ the ‘least restrictive’ means in negotiations with foreign businesses, this inhibits their autonomy to regulate the terms and conditions of employment as well as their ability to protect workers’ interests. This implicates both subsidiarity and proportionality.

 

The Court has been inconsistent in its observance of the principles of subsidiarity and proportionality across the policy areas studied in the book.

 

In three of the four areas – healthcare, education and collective labour law, the Court has employed a high standard of review of national measures, focused on the test of necessity. This approach is well illustrated by the Bressol judgment. The case concerned Belgian measures restricting foreign students’ access to some of the oversubscribed medical programmes. For context, in some of these programmes foreign students (mainly French) would make up more than 80% of the class. This situation raised concerns about the potential shortage of health specialists in Belgium and, more broadly, about the financial sustainability of the country’s higher education system, which was based on free and open access. The Court held that in order to be justified, government measures had to be appropriate and proportionate. Notably, the assessment of proportionality had to take into account the possibility that the government might have been able to achieve its regulatory goals with less restrictive means. Finally, the government was required to provide ‘an objective, detailed analysis, supported by figures, … capable of demonstrating, with solid and consistent data, that there [were] genuine risks to public health’ (para 71).

 

Essentially, the Court’s reasoning made clear that the Belgian higher education policy could only be pursued by employing measures that did not restrict the free movement of students in the EU.

 

This approach can be contrasted with the one employed by the Court in the gambling case law. Governments often limit access to their gambling markets for foreign operators, often by monopolising the operation of gambling. (This keeps the revenues derived from gambling within national borders.) The Court’s review of these types of measures is highly deferential. It is grounded in respect for national regulatory autonomy in the gambling sector. In that connection, the Court holds that Member States are entitled to a margin of appreciation, or discretion, when regulating gambling activities. Accordingly, ‘it is for those authorities to consider whether, in the context of the aim pursued, it is necessary to prohibit activities of that kind, totally or partially, or only to restrict them and to lay down more or less rigorous procedures for controlling them’ (see Zenatti judgment, para 33).

 

As a result, governments are allowed to bar foreign operators from entering their gambling markets as long as such restrictions are consistent with the regulatory objectives pursued.

 

The book ultimately argues that discrepancies in the standard of judicial review employed in relation to the areas under study – healthcare, education and collective labour law, on the one hand, and gambling, on the other – cannot be justified considering their comparable political and social sensitivity. Rather, these discrepancies indicate that the Court has been unfairly selective in its observance of the principles of subsidiarity and proportionality in the free movement case law.  

 

Judicial Reasoning Reform

 

In light of its findings, the book advances a proposition aimed at restoring the balance of competences within the EU and thus easing the tensions created by the case law. Specifically, it proposes that the principles of subsidiarity and proportionality should be reinforced into EU-level judicial practice concerning the areas of national policy which are in principle reserved to Member State competence. In practice, this would require the Court to employ, in relation to those areas, a set of techniques of judicial interpretation – already to be found in the Court’s wider case law – that are aimed at lowering the standard of judicial review. These techniques are the margin of appreciation, the ‘fair balance’ approach and exemption from the scope of EU law.

 

The proposed reform would ensure that the exercise by the Court of EU judicial authority through the interpretation of internal market law respects the limits of EU competences.