The EU (non) co-ordination of minimum subsistence benefits: What went wrong and what ways forward?

This contribution deals with the co-ordination of minimum subsistence benefits in EU law. It is argued that the distinction between social assistance schemes and non-contributory benefits in EU social security law is becoming increasingly redundant. This is recognised in the case law of the CJEU, although paradoxically not in a way that strengthens the rights of mobile citizens, but in an adverse manner that undermines the co-ordination efforts of non-contributory benefits under Regulation 883/2004. In order to overcome this problem, it is argued that social assistance should be included in the material scope of application of Regulation 883/2004. This regulation should abandon the concept of special non-contributory benefits and introduce a new category of minimum subsistence benefits, which would also include social assistance schemes. Such a change could be accompanied by a single, coherent principle to govern the relationship between the right to benefits (Regulation 883/2004) and residence rights (Directive 2004/38), if necessary supported by a cost sharing mechanism for minimum substance benefits under Regulation 883/2004 and, preferably, by a recognition of minimum protection standards for economically non-active EU citizens without a legal right of residence.


Introduction
The place of minimum subsistence benefits in EU law is shrouded in controversy. Member States are reluctant to grant tax financed social advantages to mobile EU citizens at home, let alone to those moving to other parts of the EU. Yet, the principles of EU citizenship and the freedom of movement of persons impose certain obligations in this regard. Thus, the state of EU law on the right to minimum subsistence benefits for mobile persons reflects a compromise between the interests of Member States and these principles, as enshrined in the EU treaties. The state of law is volatile, not only because of the dynamics of EU case law, but also because of changes in the role, nature and composition of minimum substance benefits in the Member States. The public debate in the Member States stirs up the fire even more.
This article analyses the state of co-ordination of minimum subsistence benefits in EU law. In it, I refer to these benefits as an umbrella concept to denote both 'social assistance' within the meaning of Directive 2004/38 and 'special non-contributory benefits' within the meaning of Regulation 883/2004. As such, their co-ordination is well recorded. There are descriptive chapters in in the handbooks on social security law, 1 as well as the separate contributions of colleagues. 2 Also, many commentaries have been written in the slipstream of the latest wave of regressive judgements on social assistance and the Residence Directive 2004/38 À Brey (C-140/12), Dano (C-333/13), Alimanovic (C-67/14) and Garcia (C-299/14) and the judgement in the infringement procedure Commission v. UK (C-308/14) dealing with the lawful residence requirement under Regulation 883/2003. 3 What these commentaries have in common is that they look at developments in case law primarily through the lens of EU Citizenship and the freedom of movement of persons. While these lenses will be used in this paper, I also intend to look at the EU co-ordination of minimum subsistence benefits from the perspective of the changes in the architecture of the social security systems of the Member States. Before continuing, this concept must be explained.
Present EU co-ordination law is based upon two marked distinctions between social security schemes, which have major consequences for the quality of the co-ordination rules. One is between general social security benefits and special non-contributory benefits, and the other one between special non-contributory benefits and social assistance benefits.
The central question I wish to address in this contribution is the extent to which these distinctions are adequately supported by social security developments in the Member States. And if not, what should the consequences be for the future of EU co-ordination law for minimum subsistence benefits, and to what extent are these consequences taken on board as part of current revision process of Regulation 883/2004? 4 I will argue that the distinction between social assistance schemes and non-contributory benefit schemes is becoming increasingly redundant. This is implicitly recognised in the case law of the CJEU, although paradoxically not in a way that strengthens the rights of mobile citizens, but in an adverse manner that undermines the co-ordination efforts of non-contributory benefits under Regulation 883/2004. In order to overcome this problem, I argue that social assistance should be included in the material scope of application of Regulation 883/2004 and excluded from the ambit of the Residence Directive 2004/38. More specifically, Regulation 883/2004 should abandon the concept of special non-contributory benefits and introduce a new category of minimum subsistence benefits, which would also include social assistance schemes. Such a change could be 1. Fuchs and Cornelissen (2015) and Pennings (2015). 2. Thus, for example, Herwig Verschueren has written extensively the controversies surrounding the regime of the of special non-contributory benefits adopted in the then new Regulation 883/2004 and how these were finally settled by the CJEU. See Verschueren (2009)  accompanied by a single, coherent principle to govern the relationship between the right to benefits (Regulation 883/2004) and residence rights (Directive 2004/38), if necessary supported by a cost sharing mechanism for minimum substance benefits under Regulation 883/2004, and preferably by a recognition of minimum protection standards for economically non-active EU citizens without a legal right of residence.
This contribution is structured as follows. It begins by giving a brief account of the present state of (non) co-ordination of social assistance and non-contributory benefits (Section 2). It then goes on to look at the justifications for a separate regime for both non-contributory benefits under Regulation 883/2004 (Section 3) and for the special place of social assistance in the corpus of free movement regulations (Section 4). This contribution then turns to the question of where the CJEU took a wrong turn in recent social assistance case law, which in my opinion, happened in the case of Brey (C-140/ 12) (Section 5). Finally, the possibilities of an alternative co-ordination regime for minimum subsistence benefits are discussed (Section 6), and the article ends with a brief afterthought (Section 7).

The co-ordination of non-contributory benefits and the noncoordination of social assistance
The background to the exclusion of the material scope of application of Regulation 883/2004 is a historical one. Access to social assistance for migrants was always more problematic than access to social insurance. The explanation goes back to the fact that the origins of social assistance schemes are based upon the notion of a unilateral charitable obligation, rather than a reciprocal insurance relation between the insured person and the social insurance institutions. On the international level, social assistance has been the subject of a different type of instrument; not primarily dealing with safeguarding social security rights, but with residence rights, or more specifically, the right to send poor people back to their country of origin (although this may be accompanied by mutual obligations to continue to pay social assistance as long as the repatriation has not materialised). 5 This historical situation is still visible in EU law, where social assistance remains the exclusive domain of the Residence Directive 2004/38. It is not proper to speak of a form of co-ordination of social assistance under Directive 2004/38. At most, it is a negative form of co-ordination. The purpose is primarily to exclude non-active citizens with temporary EU residence status from the right to social assistance in the host state. Admittedly, in many instances, this is formulated in a positive manner by stating in which situations there is a right of residence and corresponding equal access to social assistance benefits. Nevertheless, these provisions would not be necessary if, down the line, there are also situations in which such residence rights no longer exist should individuals be without sufficient resources of their own. In such cases, equal access to social assistance would be denied. From this perspective, it is fair to speak of a situation of non-co-ordination of social assistance, which is even less far reaching than the 1952 Convention on social and medical assistance, and which, at least, guarantees the payment of social assistance until a deportation order has been issued, provided there is no stay of execution. 6 5. C.f. The European Convention on social and medical assistance, 1953. Only by exception international social assistance agreements uniquely focus on benefit rights, rather than residence rights. An example is the Abkomen zwischen der Republik Ö sterreich und der Bundesrepublik Deutschland ü ber Fü rsorge und Jugendwohlfahrtspflege, 17 January 1966. 6. Art. 1 and Art.11 of the European Convention on social and medical assistance, 1953.
Conversely, the co-ordination of special non-contributory benefits under Regulation 883/2004 constitutes a true form of co-ordination because its sole purpose is not only to exclude exportability but also to guarantee access to these benefits in the host state. The guarantee included in Article 70, paragraph 4 Regulation 883/2004, is adamant about that. The exclusive obligation for the residence member state to provide mobile citizens with special non-contributory benefits at its own expense was part of the great bargain between the Commission and the Member States in the 1992 amendment of the then Regulation 1408/71. 7 Also, Implementation Regulation 987/2009 contains interesting procedural rules in Article 11 covering the situation in which residence is contested between two states, taking into account a number of criteria, such as the duration of the stay, family ties, the intention of the person, etc. Such rules are a typical by-product of a collective coordination effort.

Justification for a separate regime for special non-contributory benefits?
It is possible to question the validity of a separate co-ordination regime for special noncontributory benefits in the system of co-ordination regulations from the point of view of the architecture of social security in the Member States. At present, social insurance has so many characteristics that deviate from genuine insurance principles: minimum income guarantees, a lack of equivalence between contributions and the level of benefit, partial means testing, etc. On top of these characteristics, there are more fundamental questions: is there really such an enormous difference between financing by general taxation or by earmarked contributions, and, furthermore, is it not the case that many insurance schemes partly financed by both sources?
Yet, overall, the distinction between contributory social insurance and other types of benefit is proving to be much more persistent than could have been expected on the basis of such trends and arguments. Social insurance is still there as a separate category in the legislation of all EU countries, with its own legal principles of membership, obligations and personal rights and separate institutions responsible for administration. And while a considerable body of non-contributory benefits has come into being to address the shortcomings of the social insurance system, this does not seem to have affected the continuation or the legitimacy of social insurance itself. This holds true not only for continental countries like Austria, Germany and France, but even for the UK where the social security project has meandered in different directions.
Perhaps the 'stubborn persistence' of the social insurance approach to social security can be explained by the fact that people and politicians are attached to the myth of personal rights following from the insurance principle. It was Beveridge who famously said: 'Benefit in return for contributions, rather than free allowances from the State, is what the people of Britain desire.' 8 It is posited that this subjective preference is even stronger in Europe, where social insurance often involves a form of co-ownership by employers and employee organisations. 9 Social insurance is rooted in the European social model. This is much less the case with regard to non-contributory benefits, which are a state affair and outside the realm of the social partners.
In view of this, there still seems to be a case for justifying a different status for non-contributory benefits under Regulation 883/2004, with social insurance rights presented as personal rights following from membership of a particular scheme (suitable to be exported to other countries) and special non-contributory benefits that are more closely linked to nationhood and territory. A corollary is that the promise of protection provided by a minimum subsistence level is limited to the circle of national solidarity. EU citizenship is expressed by including mobile persons from other Member States in this solidarity circle.

Justifications for a separate regime for social assistance benefits
It is more difficult to find justifications for the distinction between social assistance, on the one hand, and special non-contributory benefits, on the other. Many years ago, the background to the exclusion of social assistance from the co-ordination regulations was well-researched by Advocate General Mayras in the case of Frilli (1/72), which dealt with the Belgian guaranteed income for the elderly (a non-contributory moderately means tested benefit scheme). Social assistance can be traced back to the discretion of the poor laws and the denial of a subjective right. But when legislation grants recipients a clear legal position, according to the Advocate General, such benefits become part of the wider concept of social security as a whole. In this sense, the Belgian guaranteed income for the elderly could no longer be classified as social assistance as it provided recipients with clearly defined subjective rights, analogous to those in social security. The Court accepted this but later added the additional requirement that, for benefits to be classified as social security they must be linked to the enumeration of the risks included in today's Art. 3 Regulation 883/2004. 10 That seemed to be a good additional standard because in those days, schemes that were nationally classified as social assistance were indeed of a general nature: Sozialhilfe in Germany, Revenu minimum d'insertion in France, Supplementary Benefits in the UK, Sociale Bijstand in the Netherlands, the Bestaansminimum in Belgium, etc.
Nevertheless, since then, in many states the formerly general social assistance schemes have become more categorical. One of the reasons for this is that, as part of activation policies, the 'able bodied unemployed' (to re-use this classical notion once more) have been separated from the general social assistance scheme and included in separate benefit arrangements, with a separate regime of work duties, sanctions and re-integration services, etc. 11 Thus, in the UK, the noncontributory Jobseekers Allowance was taken out of the Income Support regime. Equally in Germany, Arbeitslosengeld II now constitutes a different scheme alongside the residual Sozialhilfe safety net for the elderly and for those who are not capable of earning an income through work. 12 But also, within minimum subsistence schemes that have not been split up along these categorical lines, clear distinctions in legal regimes between different groups of beneficiaries may appear. Thus, in the Netherlands, with the introduction of the Participatiewet in 2015, handicapped and elderly beneficiaries enjoy a distinct legal position compared to other social assistance recipients. The same is the case for the Universal Credit Scheme in the UK which, as part of effort to integrate a number of means tested benefits, is now being rolled out. This regime is incorporated in a layered 10. C.f. the cases of Hoeckx (C-249/83) and Scrivner (C-122/84). 11. For an extensive comparative analysis of contemporary minimum subsistence schemes, see Bahle, Hubl, and Pfeifer (2011). 12. Leistungen der Grundsicherung im Alter und bei Erwerbsminderung nach dem Vierten Kapitel des Zwö lften Buches Sozialgesetzbuch.
system consisting of four levels. 13 The first three levels represent lighter forms of conditionality and sanctions and their personal scope of application is narrowly defined: they cover inter alia claimants with limited capacity for work, (lone) parents of children under the age of five and foster parents in a comparable situation. The lighter degrees of conditionality aim at preparing claimants for a transition into paid employment (work-focused interviews and work preparation requirements) and the sanctions employed are relatively moderate: in most cases, benefit payments are suspended until the claimant has complied with the duty in question. 14 The final degree of full conditionality applies as a default mode to all claimants who do not fall into one of the narrowly defined categories of individuals mentioned above. 15 Under this state of affairs, the distinction between general social assistance and special noncontributory benefits is becoming very narrow. Still, social assistance schemes for the unemployed or for separate groups, such as the handicapped or the elderly, are fully coordinated by Regulation 883/2004 (respectively as unemployment benefits or special non-contributory benefits), while general social assistance schemes, which do not make formal distinctions between these categories, are characterised by a de facto differentiation of legal positions and fall under the principles of non-co-ordination of the Residence Directive 2004/38. This apparent lack of justification for the difference in treatment between these two categories is not only questionable in view of the function and nature of minimum subsistence benefits in the Member States but may also lead to all sorts of confusions, misunderstandings and aberrations in the application of EU law itself. For example, the Germans have adopted their residual Sozialhilfe scheme (SGB XII) in Annex X Regulation 883/2004, while Income Support in the UK, which has the same function and character as the German equivalent, is not mentioned in this Annex. This state of affairs is illogical. 16 This state of affairs becomes even more murky when we consider the fact that, in recent decades, many states have introduced and/or extended social cash transfers programmes in order to combat general household poverty. These schemes may come under taxation law in the form of tax exemptions, but they may equally be part of the social security system where they take the form of allowances, credits, income supplements, etc. A central characteristic of these schemes is that they are tax financed and means tested. Often, such measures are targeted towards parents with children, in particular lone parents. 17 Hence, they are likely to be co-ordinated by Title III, Chapter 8 of Regulation 883/2004 as family benefits, whereas in fact, these cash transfers have much in common with general social assistance regulated by universal social assistance schemes.

Case law of the CJEU: Where did it go wrong?
In my opinion, it is only a matter of time before this differentiation in co-ordination regimes of comparable benefits systems will collapse under the pressure of its own lack of inner logic. As a matter of fact, the structure has probably already collapsed, but unfortunately not with positive consequences for EU mobile citizens. Thus, the question must be asked: where did it go wrong? 13. For more details, see Dwyer and Wright (2014). 14. See e.g. Sections 20 and 21 Welfare Reform Act 2012; Regulations 104 and 105 Universal Credit Regulations 2013. 15. Section 22 Welfare Reform Act 2012. 16. The judgement of the CJEU in Swaddling (C-90/97) probably has something to do with it, but that can hardly be seen as justification for the difference in approach to Annex X by Germany and the UK. 17. Marchal and Marx (2015).
For this to be answered, we have to go to Austria, where by a decision of 2 March 2011, the Pensionsversicherungsanstalt refused the Ausgleichszulage application of a German national of Russian origins on the grounds that he did not have sufficient resources to establish his lawful residence in Austria. The individual was Mr. Brey, and his case was brought before the CJEU (C-140/12).
In this case, the Court did three things which, in my view, have been fundamental for subsequent developments in case law. First of all, the CJEU qualified the Ausgleichszulage, which is undeniably a non-contributory benefit within the meaning of Regulation 883/2004 (adopted in Annex X) as social assistance within the meaning of the Residence Directive 2004/38. Secondly, the CJEU gave exclusive preference of the latter Directive over the former Regulation. Thirdly, the court employed a very broad definition of social assistance, 18 which has constantly been repeated in subsequent case law so broadly that it can easily engulf the entire landscape of non-contributory benefits as defined in Regulation 883/2004.
In this manner, the CJEU has indeed refused to uphold a dividing line between social assistance and special non-contributory benefits, but with what consequences? The Residence Directive 2004/38 has become the default standard, and hence the restrictive conditions of this directive have become the weakest links in the protection of mobile EU citizens.
Paradoxically, the potential damage of the Brey case went largely unnoticed because the CJEU got rather stuck on how to apply the Residence Directive 2004/38 in this particular situation. But when, in the later cases of Dano (C-333/13), Alimanovic (C-67/14) Garcia (C-299/14) and the Commission v UK (C-308/14), the CJEU showed its true colours with respect to the right to social assistance for economically non-active citizens, the damage was already done. Potentially, the new trend in the case law can undo all the good work done under the co-ordination regulation on mixed benefit systems since the case of Frilli (C-1/72) in 1972. 19 I have chosen the qualification 'regressive trend' in the introduction to this article with care.
The recent trend in the case law has further sharpened the schism in quality of protection in Residence Directive 2004/38 between non-economically active citizens and those who are economically active. Recent case law, which involved the right to social assistance for workers, has confirmed this. 20 However, from the point of view of the logic of minimum subsistence benefits, which mostly do not recognise any differentiation on the basis of work status, such variation in protection is certainly not logical.
Finally, the recent trend in case law has painfully exposed the limited value of fundamental social rights in EU law. In this respect, it should be pointed out that Dano was handed down in the same week that the complaint authority of the European Social Charter, the European Social Rights Committee (ESRC), took two decisions in collective complaints procedures against the Netherlands concerning its refusal to grant shelter and emergency assistance to foreign nationals without residence status (Council of European Churches v. the Netherlands;No. 90/2013), and about the practice of local councils applying a local connection test before giving shelter to the homeless (FEANTSA v. the Netherlands: No. 86/2013)). According to the ESRC, the Netherlands violated several articles of the European Social Charter. This is, ironically, the current state of affairs under European law when it comes to the right to protection for non-nationals offered by social assistance schemes. EU citizens cannot invoke their EU Charter rights before the CJEU but must rely on NGOs using the complaint procedure adopted by the Council of Europe under the European Social Charter. 21

Ways forward
What can be done to bend the regressive trend back to a more friendly, integrationist regime for granting access to minimum income protection for economically non-active citizens? In my view, the first way is to create new conceptual guidance as to the meaning of social assistance and special non-contributory benefits. These two categories should be merged into the single category of 'minimum subsistence benefits'. For the meaning of these benefits, it is possible to refer to the first part of the definition employed by the CJEU in Brey (C-140/12): i.e. benefits 'introduced by the public authorities, whether at national, regional or local level, that can be claimed by an individual who does not have resources sufficient to meet his own basic needs and the needs of his family'. So, it is the public character, the individual right and the (household) means test that are the three constituents of this definition.
All benefits that can thus be defined as minimum subsistence benefits should fall exclusively under Regulation 883/2004, either in a general capacity or in a specific, categorical capacity. This also means that any references to social assistance in the Residence Directive 2004/38 can be scrapped. The same is the case for the exclusion of social assistance from the material scope of application of Regulation 883/2004. As long as the definition of minimum subsistence benefits contains a reference to a statutorily defined right to a means tested minimum benefit, this is no longer necessary. The technique for the co-ordination of minimum income benefits can remain the same as today, i.e. as guaranteed access in the host state and no exportability.
The second thing that must be done is to somehow establish a single, coherent principle that governs the relationship between benefit rights (Regulation 883/2004), on the one hand, and residence rights (Directive 2004/38), on the other. In my view, the key to this is a legal residence test, such as the one that is already included in the proposal for a revised Regulation 883/2004, 22 which in Article 4(2) contains an exclusion for non-active persons who no longer have legal residence within the meaning of the Residence directive. By offering Member States this proviso in return, the sacrifice made in Commission versus the UK 23 would not have been in vain. But, there are two important additions to be met in order to make this pathway palatable. 21. C.f. My blog on https://europeanlawblog.eu/2014/11/25/eu-freedom-of-movement-no-protection-for-the-strandedpoor accessed on 15 January 2020. However, the ECtHR adopted a less generous stance in Hunde v. the Netherlands, 28 First of all, it is necessary to include a clause similar to article 11(b) of the European Convention on Social and Medical Assistance of 1952; i.e. that residence shall only be deemed unlawful from the date of any deportation order made out against the person concerned, unless a stay of execution is granted. The consequence is that there can no longer be an automatic exclusion of minimum subsistence solely on the ground that an economically non-active person is not deemed to have sufficient resources of himself. First, the legality of the residence will have to be officially established and given effect to by the immigration authorities in order to end the right to benefits. In this sense, the consequence of Dano and subsequent cases needs to be repaired. If necessary, this solution can be accompanied by the introduction of a cost sharing mechanism for benefits granted to economically non-active citizens. There are precedents for such mechanisms in pre-war European history 24 and, of course, in Regulation 883/2004 itself (for medical costs and unemployment benefits 25 ).
Secondly, the EU should not fully abandon those stranded EU citizens who no longer hold lawful EU residence status, even when they are excluded from rights under Regulation 883/2004. At least some minimum care obligation should be established, comparable that which is accepted for asylum seekers and persons applying for international protection. A suitable point of reference is Directive 2013/33/EU, which sets out standards for the reception of applicants for international protection. Article 17(2) of this Directive provides a credible description of the protective standard involved: 'Member States shall ensure that material reception conditions provide an adequate standard of living for applicants, which guarantees their subsistence and protects their physical and mental health'. In this way, the Directive would stipulate further rules as to what is to be understood by this protective standard (Article 17-Article 19) as well as additional guarantees such as the right of the families to stay together (Article 12) and access to housing for minor children (Article 14). By assuming responsibilities for stranded EU citizens, EU law would respect its human rights obligations, 26 not only for those seeking international protection, but also to its own citizens.

Afterthought
My generation of European lawyers was raised on the idea that European integration is a civilizing force, resulting from the co-operation between nations and peoples that arose after the experiences of the past. For this reason, I had always assumed that, one day or another, the rights to income support for poor mobile citizens in the EU would be fully recognised as an automatic by-product of the forces of progress built into the European project.
But times, they are a changing. Shadows of nationalism are once more dimming the lights over the old continent. Benefit rights for the poorest citizens are the first thing to be sacrificed. This theme is always picked out as the emblem by those who want to make political gain by stirring the pool of discontent. EU law is not immune to this pressure. It is in these circumstances that we realise that nothing in this life can be taken for granted and that everything valuable has to be constantly defended and argued for. This puts some responsibility on the shoulders of the next generation of European lawyers.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.