Abstract
The effects of crossing borders can be advantageous or disadvantageous for the persons concerned; these are all part of the game and cannot be challenged on the basis of EU law. After all, the Treaty on the Functioning of the European Union (TFEU) does not provide powers for harmonisation, but only for coordination. However, the coordination rules themselves may make a person worse off when he or she makes use of the right to free movement. More precisely, such an effect may occur in combination with differences between national systems to which coordination rules are applied. One example is that the coordination rules provide that a person is subject to unemployment benefits in the country of residence and, as a result, if that person becomes ill, also to sickness benefit in that country. If the duration of sickness benefit in the country of residence is 52 weeks, but the waiting period for disability benefit (supposing, for instance, that this is (mainly) due from the country of employment) is 104 weeks, there is a gap of 52 weeks in protection. The relevance of such gaps is not to solve particular cases as such; after all, these are closely linked to particular national systems. The relevance lies in the more general approach that is now being developed by the Court of Justice to address such gaps. This will be useful in cases other than those discussed here and may be further developed in order to be codified in the Coordination Regulation.
Keywords
1. Introduction
When people consider moving to another country, one of things they may worry about is the impact of their move on their social security position. Will they be sufficiently protected if they lose my job, become ill or disabled? Are their pension rights well protected? Questions involving the position of family members also arise. Such questions were already foreseen at the time when the EEC Treaty was drafted. Therefore Regulation 3 (one of the first Regulations) was adopted in 1958 to deal with these issues. Regulation 3 was replaced, first by Regulation 1408/71, and then in 2010 by Regulation 883/2004 (henceforth: the Coordination Regulation). 1 These Regulations have created, in combination with the case law of the Court of Justice of the European Union (henceforth: the Court of Justice), a coherent and comprehensive coordination system for migrants.
The objective of the Coordination Regulation is not to harmonise the social security systems of the Member States, but to provide a ‘bridge’ for migrants to assist them to gain access to the social security system determined to be applicable to them. Examples are aggregation rules for having recourse to periods of insurance or work completed in another country for satisfying the conditions in the country where benefit is claimed. Coordination rules do not change the benefit conditions (e.g. on their employment history) of national benefit schemes, since that would amount to harmonisation.
Some of the problems that migrants are confronted with are inherently linked with the lack of harmonisation of social security systems. One example is the difference in benefit levels. If a person first works in a country where unemployment benefit is 80 per cent of the last earned wage and she moves to work and live in a country where the level is 50 per cent, she is paid the latter after becoming unemployed. As a result, she has a lower benefit than she would have had if she stayed in her country of origin. Such differences also appear in other circumstances, such as entitlement conditions, duration of benefit, criteria for disability and sanctions. 2 As previously stated, this is an unavoidable effect of the lack of harmonisation.
The harmonisation of national social security systems would make mobility much easier, but it is one of the largest taboos in the EU. Member States fear, inter alia, higher costs and a reduced ability to respond to social problems and to problems relating to the impact on other institutions (such as the labour market) in their own countries if they had to follow common standards. Therefore, currently the social security protection of migrants can only be addressed by the coordination rules.
It would go far beyond the limits of this contribution to describe the coordination system; neither is it necessary to do, since there are already several discussions of this issue. 3 Nor will this contribution discuss the advantages and disadvantages of making use of the right to free movement. 4 Instead, by describing some situations derived from practice, 5 it will discuss the gaps that remain in the coordination system and for which case law has been developed. For this purpose, it will focus on those cases where, as the result of the application of coordination rules, there is reduced protection, or no protection at all, for migrants. The question arises in such situations as to which country has to fill such gaps. Which principles can be applied in such cases and which instruments can be used to find solutions?
In Section 3, the effects of the differences in waiting periods and between pension ages are elaborated. Section 4 discusses the position of so-called ‘mini jobs’, and Section 5 deals with the position of a person who is or was insured in the wrong country.
At this moment a proposal for revision of the Coordination Regulation is pending; the European Parliament has adopted some amendments that address the gaps highlighted in this article. We will discuss these in Section 6 and will make some alternative proposals that fit with the analysis made in the preceding sections.
2. A short overview of the coordination system
In this section, some of the main rules that are needed to understand the analysis of our cases are discussed.
The Coordination Regulation determines which legislation is applicable for determining the conflict rules. The State whose legislation has been determined in this way is called the competent State. The conflict rules have exclusive effect: a person to whom the Regulation applies is subject to the legislation of a single Member State. As a result, if the legislation of the State of residence provides that a person is insured in that State, even if she is working in another State (which is possible in residence schemes), is ‘withdrawn’ from the State of residence’s system if the legislation of the country of employment is applicable (which is the main rule). This exclusive effect of the conflict rules prevents double insurance and, as a result, the obligation to pay two sets of contributions. Further conflict rules are that when a worker or self-employed person works in two countries, in principle the system of the State of residence is applicable. However, if she works in one country as a civil servant, and in another as an employee, the legislation of the former State applies.
These provisions can sometimes have undesirable results, for instance, in the latter case when an employee accepts a job as civil servant for a short time only. In this case the shift in legal position has many administrative consequences. The Coordination Regulation has a safety net for solving such problems. Article 16 of Regulation 883/2004 provides that two or more Member States, the competent authorities of these Member States or the bodies designated by these authorities may by common agreement provide for exceptions to Articles 11 to 15 of the Regulation (the conflict rules) in the interest of certain persons or categories of persons.
The chapter on unemployment benefits has special rules for frontier workers. They receive, in the case of full unemployment, unemployment benefit from the State of residence (Article 65(2)). The argument for this is that they are supposed to have a better chance on the labour market of the home country. If the State of residence is the competent State as a result of this rule, the person concerned is subject to the full social security legislation of that State (as otherwise things would become very complicated). As a result, the person is insured in the latter State for sickness benefit, old-age benefit, etc.
Coordination rules for short-term benefits (unemployment, sickness etc.) differ from those for long-term benefits (old age, disability, etc.). Benefits in the first category are paid only by the competent State. 6 Benefits in the second category are, in principle, paid on a pro rata basis, i.e. every State in which a person has been insured pays part of the benefit, depending on how long the person was insured in that State. In other words, the principle that legislation of only one State is applicable does not prevent benefits from being paid by more than one State on the basis of insured periods in the past.
3. Overview of the situations to be discussed
As discussed in the Introduction, this contribution focuses on those cases where, as the result of the application of coordination rules, there is reduced protection or no protection at all for migrants. For this purpose, it will first discuss a situation in which there is a difference in waiting periods or a difference in pension age between two systems. It will then discuss the situation in which a person is not insured in the State of residence as a result of the conflict rules, in spite of the fact that the national system provides coverage and the person is not insured in the State of employment. Then a situation is discussed in which a person has a very small job as civil servant which makes her subject to the system of the State where this work is performed, although she carries out her main activities in another Member State.
At first sight the cases to be discussed seem to deal with quite separate and casuistic situations. However, they have more in common than might appear. In each of these situations, it can be said – although this is a simplification - that the problems result from the application of the coordination rules. This is the common element, although below we will also show the differences between the situations are also a result of differences in approaches and solutions.
Note that the cases presented below are only intended to be illustrative; the interaction of coordination rules is what is relevant for the analysis. In other words, these problems do not occur only in Belgian-Netherlands cross-border situations.
4. Differences in waiting periods or pension age
4.1 The case of John
John worked throughout his life in the Netherlands and has lived in Belgium since 2008. When he lost his job, on 31 December 2018, he was 65. In 2019, the entry age for Dutch old-age benefit was 65 years and 4 months. For this reason, John did not yet qualify for this benefit. As a frontier worker, he would normally be entitled to Belgian unemployment benefits, but persons over the age of 65 are not entitled to Belgian unemployment benefit. Instead, they can claim Belgian old-age benefit from age 65, but John had not acquired entitlement to such pension. As a result of the coordination rules, he was excluded from Dutch unemployment benefit (see Section 2 supra). Since he was not entitled to any benefit at all before reaching the Dutch pension age, the new year did not start happily for John.
Belgium has introduced a solution that partially fills this gap by providing, in a special national regulation, that frontier workers receive Belgian unemployment benefit until they are entitled to a foreign statutory foreign old-age pension. 7 A condition for this generous solution is, however, that a worker belonging to this category has worked for 15 years as a frontier worker in Belgium. Since John has lived for only ten years in Belgium, this special scheme did not provide a solution for him.
A member of the Dutch Parliament asked questions to the Dutch Minister of Social Affairs about persons in a position such as John’s. 8 The Minister replied, without giving specific arguments, that he liked the Belgian solution and that he would ask his Belgian colleague why the scheme required 15 years of work in Belgium. In other words, he preferred Belgium to provide a solution for all frontier workers in John’s position. But is this really a solution for all cases and on what grounds? This question is addressed in the next section.
4.2 Analysis on the basis of EU law
Questions about the issue of different waiting periods for benefits have been referred to the Court of Justice in a couple of cases. The first was the Leyman judgment. 9 Ms Leyman was first insured in Belgium and then in Luxembourg. When she became disabled, because of her high level of disability and in accordance with the applicable Luxembourgish rules, she was immediately (i.e. without a waiting period) awarded, a (pro rata) Luxembourgish invalidity benefit. Since she had worked for only a short period in Luxembourg, this benefit was small. However, she was also entitled to a Belgian disability benefit, which could be paid only after a waiting period of one year. Persons insured in Belgium receive sickness benefit during this waiting period for Belgian disability benefit, but since Ms Leyman was subject to the Luxembourgish system, she was only paid the small Luxembourgish disability pension during this period.
The Court of Justice, while acknowledging that Article 48 TFEU leaves the differences between the Member States’ social security systems intact, considered that the aim of Article 45 TFEU would not be met if, through exercising their right to freedom of movement, migrant workers were to lose social security benefits guaranteed to them by the laws of a Member State. The law concerned causes a disadvantage for workers such as Ms Leyman during the first year after claiming, compared with workers who are also definitively or permanently incapable of work, but who have not exercised their right to freedom of movement. The Court concluded that Articles 45 and 48 TFEU preclude application by a Member State of national legislation which makes acquisition of the right to disability benefits subject to the condition that a period of illness of one year has elapsed, where such application has the result that a migrant worker who has paid contributions into the social security scheme of that Member State for which there is no return and is at a disadvantage compared with a non-migrant worker.
So, there are two conditions: the national legislation places the worker at a disadvantage compared with those who have pursued all their activities in the Member State where the legislation applies since this results in the payment of social security contributions for which there is no return. Although this situation was contrary to the aforementioned Articles, the Regulation did not provide a solution for this. Therefore, the Court of Justice referred to the principle of cooperation in good faith, laid down in Article 4(3) TFEU, which requires the competent authorities in the Member States to use all the means at their disposal to achieve the aim of Article 45 TFEU.
Since the problem of different waiting periods can also occur in situations other than that of Ms Leyman (for instance in the case of John), there has been some uncertainty as to how to act in general in such cases. For instance, in the Dutch system, the waiting period for disability benefit is 24 months, and in Germany it is 78 weeks. A claimant having worked in both the Netherlands and Germany, who becomes ill in Germany, receives a pro rata German disability benefit after 78 weeks of sickness. This person has to wait another 26 weeks before he or she becomes entitled to a pro rata Dutch disability benefit as well. 10 The problem becomes even greater when the German insurance period for disability is relatively shorter than the Dutch one, since a relatively shorter German period means a lower German disability benefit.
Although the Dutch Minister of Social Affairs was asked in Parliament how to deal with these situations, he did not provide a general solution for solving the problems. Instead, patchwork solutions have been made by the benefit administration. One of the solutions is that an income supplement is paid, that supplements the benefit up to the applicable social minimum. However, for those who have worked in higher paid jobs, this does not compensate the full missing income. Another solution is that a Dutch unemployment benefit is paid (from which the German disability benefit is deducted). The decision as to which solution is chosen in each case is not very transparent and no general rules have been published.
In the Vester judgment 11 the Court had to deal with the waiting period for the Dutch disability benefit mentioned in the preceding paragraphs. After having become unemployed from a job in the Netherlands, as a frontier worker Ms Vester received unemployment benefit in Belgium from the Belgian benefit administration. Then she became ill. Since she received unemployment benefit in Belgium, she was subject to the Belgian social security legislation and became entitled to Belgian sickness benefit. When the sickness benefit period expired after 52 weeks, she did not qualify for Belgian disability benefit, due to her very short insurance period in Belgium. She was not paid Dutch disability benefit either, until the Dutch waiting period for this benefit of 104 weeks had been fulfilled (of which she had only completed 52 weeks, receiving Belgian sickness benefit). Therefore, she had to wait 52 weeks without benefit.
The Court of Justice, which was asked whether this gap in protection was compatible with Articles 45 and 48 TFEU, repeated the considerations it had made in the Leyman judgment mentioned above. It noted that Ms Vester had to wait 12 months before she could receive Dutch disability benefits, during which period she received no benefits at all. It also noted that workers who, unlike Ms Vester, have not exercised their right to free movement and had completed the entire period of incapacity to work under Dutch law, received benefits for the full duration of sickness. It then made a key remark: where national law, in breach of EU law, provides that groups of persons are to be treated differently, the members of the group placed at a disadvantage must be treated in the same way and made subject to the same arrangements as the other persons concerned. The arrangements applicable to members of the group placed at an advantage remain, for want of the correct application of EU law, the only valid point of reference, the Court ruled.
So, here, the relevant framework was mentioned by the Court: the arrangements applicable to members of the group placed at an advantage. The Court continued by saying that it is for the competent national authorities of the Member States concerned to determine what are the most appropriate means for achieving equal treatment for migrant and non-migrant workers. However, it added that this objective may, a priori, also be achieved by granting migrant workers sickness benefits during the second year of incapacity to work required of them under Dutch law before they are entitled to disability benefit.
The Court did not completely make clear which benefit had to be paid to Ms Vester. However, that was not be expected from the Court, since it is up to the Member States concerned – the Netherlands and Belgium – to find a solution. However, it is clear from the framework mentioned by the Court that the Dutch situation provides the relevant framework and Ms Vester has to be compared with persons who have worked solely in the Netherlands. From this it can be concluded that she should receive Dutch sickness benefit. This is the way Van der Mei and Melin interpret this judgment. 12
It is also possible to argue that Belgium, which is the competent State for sickness benefit, has to pay the benefit for the additional year. This would, however, be an odd outcome: Ms Vester had not paid contributions in Belgium and furthermore the Court had made clear that the Dutch system provided the relevant framework.
The judgment can also be read as making it clear that, if no sickness benefit is paid in the second year, the waiting period for the Dutch disability benefit has to be shortened. This would be in line with Leyman and would fit better in the system of the Coordination Regulation and the Dutch legislation.
4.3 Application of the Vester case law to John’s case
Let us now come back to the case of John, which was also concerned with the effect of differences between systems when one becomes eligible for benefit. John had paid contributions for Dutch old-age benefit but had not yet reached pension age. He had also paid contributions for Dutch unemployment benefit, but he was not entitled to this benefit either, since, as a frontier worker, he could only claim this from the Belgian institutions. However, since claimants are eligible for Belgian old-age benefit at 65, there is no right to Belgian unemployment benefit after this age. As a result, there was no benefit covering the period before John became entitled to the Dutch old-age pension.
Since John did not receive unemployment benefit, he was worse off compared with workers in the Netherlands who had not made use of the right to free movement. In line with the Vester case law, John should have been paid unemployment benefit by the Netherlands in order to bring him into a situation comparable to that of workers in the Netherlands who had not made use of the right to free movement. In this approach, the Dutch system provides the point of reference for John, meaning that his right to unemployment benefit did not end when he reached the age of 65. Unlike the Vester case law, however, one cannot claim that persons in a situation such as John’s are always entitled to unemployment benefit until they reach pension age. After all, unemployment benefit is not by definition paid until pension age, but only for as long as benefit rights acquired on the basis of the individual’s employment history (though not beyond pension age).
Since the Court of Justice did not prescribe a fixed method, the Netherlands could solve the problem by lowering the access age for old-age benefit for a person who is confronted by the termination of unemployment benefit in Belgium because he or she has reached pension age.
The risk of this solution is, we have to acknowledge, that a ‘Belgian route’ is created for persons close to Belgian pension age who seek work for a short period in Belgium only in order to access the Dutch pension age at 65. Of course, it can be argued that this solution is possible only for those who have worked for a certain period as a frontier worker, but then we could have new ‘John’ cases, i.e. persons who fall between the new rules. Where do we draw the line?
In view of the specific character of unemployment benefit, it may be advisable that, in this situation, giving access to Dutch unemployment benefit is a better solution (provided that all the qualifying conditions are fulfilled, and neglecting the coordination rule on unemployment benefit for frontier workers). This makes it possible to take account of the reason why a person has become unemployed (e.g. voluntary unemployment is a ground for exclusion) and of the duration of benefit rights that are acquired. Since John was insured for unemployment benefit in the Netherlands, John could be given access to Dutch unemployment benefit on the basis of Dutch law.
In the proposal for the revision of Regulation 883/2004, it is proposed to give persons who have been insured for a certain period (the exact minimum is still subject to debate) in the State of employment (but do not live there) access to unemployment benefit in that country (see also the contribution on unemployment benefits in this issue). This proposed new rule would have prevented cases as John and Vester. However, not all problems are solved by this new rule. Suppose that John works in Belgium and lives in the Netherlands and becomes unemployed (the reverse situation of the John case) and has acquired rights to Dutch old-age benefit. Under the new rule he receives Belgian unemployment benefit, but this stops on his 65th birthday. In this case he will also have to bridge a period without benefit until he becomes entitled to the Dutch old-age benefit. In such case, the Vester case law can still be used to find a solution.
5. Persons excluded from coverage in the country of employment
In this section, the situation of persons who would be covered in a residence scheme had the conflict rules not determined another system as applicable will be discussed. However, because of their low number of working hours, these persons are not insured in most of the schemes of the country of employment. Consequently, they fall into a gap between the systems.
5.1 The case of Ria
Ria received disability benefit in the Netherlands, her State of residence. The Netherlands has a residence scheme for old age, and, on the basis of this, she would have acquired old-age benefit rights during her period of employment in Germany, since she was still residing in the Netherlands. She was given permission by the Dutch benefit administration to accept work in Germany, which happened to be a so-called ‘mini job’. A ‘mini job’ is subsidiary employment (work for which less than EUR 450 per month is paid) that is not covered by a large part of German social security, such as old age insurance. 13 As a result she was no longer covered by social security. The coordination rules thus ‘deprived’ her of her right to Dutch insurance.
5.1.1 Analysis according to EU law: The Ten Holder and Bosmann judgments
Those familiar with coordination law will think immediately of the Franzen, Giesen and Van den Berg judgment 14 when reading Ria’s case. This judgment concerned the joined cases of three individuals, who were deprived of insurance or benefits under the Dutch residence schemes because they worked in Germany. However, they were excluded from German insurance because they worked in ‘mini jobs’.
This judgment was preceded by the Ten Holder judgment, 15 delivered in the 1980s, in which the Court of Justice ruled that, as a result of the conflict rules of the then Coordination Regulation, only the legislation of the competent State was applicable to the (former) worker. In other words, there was a strict application of the exclusive effect of the rules for determining the legislation applicable that meant that the non-competent State was not allowed to apply its legislation and thus to provide benefit. This could have painful effects, such as in the case of Ms Ten Holder. She had lived in the Netherlands and would, without the Regulation, be entitled to Dutch disability benefit, since this was a residence scheme. She was not entitled to this benefit, however, because Germany was the competent State, but in Germany, there was no benefit scheme from which she could derive benefit rights.
In the Bosmann judgment 16 of 2008, the Court of Justice mitigated this effect. The Court ruled that the Member State of residence cannot be deprived of the right to grant child benefit to those who are resident within its territory. This is because the purpose of the Coordination Regulation is not to prevent the Member State of residence from granting, pursuant to its legislation, child benefit to that person.
5.2 The Hudziński judgment
A principle question arising from Bosmann was: when can it be said that the right to benefit is, what the Court terms, ‘pursuant to its legislation’? This question was central in the Franzen, Giesen and Van den Berg case. 17 The crucial scheme here was the Dutch old-age scheme, which is a residence insurance scheme, meaning that all residents of the Netherlands are insured. 18 Persons who are insured have to pay contributions, but those with a very low or no income, including school pupils, students and housewives, do not have to pay contributions or pay contributions at a reduced level (2 per cent of the full amount per year) although they still acquire benefit rights. Persons who are obliged to pay contributions, but fail to do so, remain insured, but the amount of their old-age benefit can be reduced by 2 per cent if the non-payment is due to ‘culpable negligence’. In this system, the non-payment of contributions does not automatically result in non-insurance or the reduction of benefit rights. Thus, the question was: is this a residence scheme from which benefits flow or is it an insurance scheme, which requires additional elements, such as the paying of contributions, before benefits can flow from this? An answer to this question is essential to determine whether a person such as Ria acquires the right to Dutch old-age benefits when the conflict rules of the Regulation do not make it clear that the Netherlands as the competent State.
The uncertainty as to whether the right of benefit flows from national legislation was reinforced by the circumstance that whereas the Court started by saying that Member States are free to pay benefits (if they do not impose contributions), the actual issue was part of a dispute in which the benefit administration did not want to pay the benefit. In this case, the national court had doubts whether the national legislation provided for the obligation to pay benefit. The uncertainty is reinforced where the Court of Justice gives an interpretation of the national legislation in view of EU law: this means that Member States are not completely free to apply their own rules. An example of this is the Hudziński judgment. 19 In this case, it was relevant that Article 62 of the German Federal Law on Income Tax provides that a person is entitled to child benefit under this Law if he has his permanent or habitual residence within the national territory, or, if he is subject to unlimited income tax liability in Germany. The latter was the case for Polish workers who were working in Germany, and, although they were covered by the Polish social security system (due to posting or because they worked simultaneously in two countries), were subject to German income tax.
The referring German court asked whether the Bosmann judgment was applicable in this situation, since unlike Ms Bosmann, these workers had not suffered a legal disadvantage by reason of the fact that they had exercised their right of free movement. The Court of Justice answered that the fact that the workers did not suffer a disadvantage was not relevant. It considered that being subject to unlimited income tax liability is ‘a connecting factor’ for the right to child benefit under German legislation, as a result of which it can be said that a right to benefit flows from German legislation. However, as the German court noted that Paragraph 65 of the German Income Tax Act excludes entitlement to child benefits in the case where a comparable benefit must be paid in another State. Does it follow that the right to family benefits does not flow from German law?
The Court of Justice answered that if the legislation of a non-competent Member State confers additional social protection on the migrant worker by virtue of the fact that he was subject to unlimited income tax liability in that State, any rules against overlapping laid down by that legislation cannot be applied, if their application is found to be contrary to EU law. The application of a national anti-overlapping rule which does not limit a reduction in the amount of the benefit but excludes that benefit completely, constitutes a substantial disadvantage, affecting a greater number of migrant workers than settled workers who have worked exclusively in the Member State concerned.
So, here, the Court restricted the freedom of national legislators to design their own systems. It added that the aim of Articles 45 TFEU and 48 TFEU would not be achieved if, as a result of exercising their freedom of movement, migrant workers were to lose social security advantages guaranteed to them by the legislation of one Member State. However, the Court added, a reduction in the amount of the benefit corresponding to the amount of a comparable benefit received in another State would be compatible with the Treaty. So, Polish family benefits would have to be supplemented to the German level.
This approach is quite complicated. The Court of Justice made clear that the non-competent State is not obliged to take away the disadvantages of exercising free movement. However, the Regulation does not preclude the non-competent State from granting benefit if it flows from its legislation but, if there is a connecting factor to granting benefit, an overlapping rule in national legislation that excludes entitlement is contrary to Article 45 TFEU.
5.3 The Franzen, Giesen and Van den Berg judgment
This is confusing. Above we discussed the characteristics of the Dutch old-age benefit scheme. This indeed has a connecting factor: residence. Do the other characteristics mentioned (obligation to pay contributions, although this is not a condition for acquiring rights) mean that benefit does not flow from this scheme, or can these conditions not preclude a person from acquiring benefit under this scheme?
In the Franzen, Giesen and Van den Berg judgment the Dutch benefit administration mentioned an additional aspect of Dutch legislation as an argument for preventing a benefit flowing from this legislation, specifically that the Dutch Old age Pension Act (Article 6a) provides that persons shall not be considered to be insured persons if, by virtue of a treaty, the legislation of another State applies to them. Is this provision sufficient to exclude persons from coverage?
The Court of Justice answered that ‘leaving aside the exclusion provided for in Article 6a, which aims to transpose the single State principle into national legislation, the mere fact of residence in the Netherlands is sufficient to establish entitlement to child benefits’ (point 62). It also referred to the Dutch decree extending and restricting the personal scope of, inter alia, the Old-Age Benefits Act, which contains a hardship clause, and which empowers the benefit administration to derogate, in certain cases, from the other provisions of that decree in order to remedy an unacceptable degree of unfairness that might arise from the insurance obligation or the exclusion from it. The Court considered that it was for the referring court to disregard the exclusion clause and to apply the hardship clause. In this way it did not answer the principle question of the relevance of the ground for exclusion but referred to the hardship clause. The referring national court subsequently applied this hardship clause. 20
5.4 The Van den Berg judgment
The dispute did not end here since the Dutch benefit administration appealed to the Hoge Raad (Supreme Court). It is quite a complicated case and needs, to some extent, to be simplified. The Supreme Court considered that the hardship clause was relevant to the Decree on the extension or restriction of the personal scope but could not remove the exclusion provided for in Article 6a. It decided that the Dutch legislator had deliberately made this exclusion. This could be removed only if Article 6a is removed. The question was, according to the Supreme Court, whether Article 45 TFEU precludes a national rule such as Article 6a.
This led to a new preliminary procedure before the Court of Justice, set out in the Van den Berg case. 21 In response to the question on Article 45, the Court of Justice replied that Articles 45 and 48 TFEU may not be interpreted as obliging the State of residence to grant benefits to a migrant worker who is not entitled to such benefits under the legislation of the competent Member State.
The Dutch Supreme Court also asked a question about the situation before Article 6a came into in force (which was the case before 1989) since, in that situation, Article 6a did not exclude persons working abroad from the right to benefit or insurance. For that situation, the question amounted to the issue (already mentioned above) of whether persons acquire old-age benefit rights solely on the basis that they reside in the Netherlands. In other words, do the other conditions described above, i.e. that the insured persons have to pay contributions, mean that residence is not sufficient for a right to flow from the Dutch legislation?
The Court of Justice replied that, under the principle of single applicable legislation, the Member State in which the migrant worker resides cannot require that worker to be insured without undermining the system of coordination under Article 48 TFEU. Such an insurance obligation entails the payment of contributions to a non-competent Member State and that could result in the migrant worker being obliged to contribute to social security in two different Member States.
The Court added to this that the fact that the non-competent State may not make the right to benefit conditional on insurance must not be understood as precluding any affiliation of a migrant worker in that Member State. The Member State of residence may, on the basis of a connecting criterion other than employment or insurance conditions, grant benefits, in particular an old-age pension, to a person residing in its territory, if the possibility of granting such benefits arises, in actual fact, from its legislation.
The Court of Justice left it to the referring court to determine whether the person concerned was entitled to an old-age pension independently of any obligation to pay contributions. This made it complicated since the criteria for this were not provided.
Finally, the Court added that Articles 45 and 48 TFEU and the Coordination Regulation are intended in particular to prevent the situation in which a worker who has exercised his right of free movement is treated, without any objective justification, less favourably than one who has completed his entire employment career in only one Member State (Hudziński). That would be the case if the national law placed the migrant worker at a disadvantage in relation to those who have only worked in the Member State where that law applies and that law obliges the worker to pay social contributions on which there is no return, which is something for the referring court to determine.
The Dutch Supreme Court had not asked the Court of Justice whether its interpretation that, as a result of Article 6a, no right to benefit flows from residence schemes was correct. Therefore, this issue was not addressed by the Court of Justice.
The Dutch Supreme Court completed the procedure by deciding that the Dutch old-age pension depends on the obligation to be insured and that the obligation to pay contributions is linked to this. This does not change with the circumstances that insured persons with a low or with no income do not have to pay contributions. 22 Whether this is actually in line with the judgment of the Court of Justice remains unclear, but the Court had left it to the national judge to decide, so this is the outcome.
5.5 The application to the case of Ria and appreciation of the case law
From the previous section it appears that the Bosmann judgment does not provide a solution for Ria. Leaving aside the criticisms of the reasoning of the Court of Justice and the national court that could be made, it has to be acknowledged that whether the Bosmann approach is a good way to solve the problems of persons in ‘mini jobs’ can be disputed. The most basic interpretation of Bosmann is that it does not preclude a Member State from paying benefit when it is not the competent State. It is quite harsh to say – as was the case in Ten Holder - that this is not allowed: why prohibit a State from paying a benefit if is willing to do so?
The issues brought before the Court are, however, those in which a benefit administration is unwilling to pay benefit and a national court has doubts whether, on the basis of the national legislation, benefit is due or not. Thus, these cases are not examples of generosity that is not allowed, but controversial issues. In such controversial cases, arguments based on national law are used (such as that the scheme is an insurance scheme or that an overlapping rule excludes paying national benefit when only benefit from the competent State is received) and also on EU arguments (such as that one should not be treated less favourably than those who have not made use of their right to free movement). If the Court of Justice does not want to take a decision itself, considerable legal certainty will remain.
A second remark is that application of the Bosmann case law would mean that no contributions have to be paid by applicants whereas residents who have not made use of their right to free movement have to pay such contributions. Although this problem is limited since persons with a low income do not have to pay contributions (and those concerned have a low income by definition), there can still be circumstances in which migrants get insurance benefits ‘for free’.
A third remark is that in the meantime, Germany has adopted legislation to better protect persons in ‘mini jobs’. This is a better approach, since after all Germany is the competent State and has to be held responsible for the protection of those falling under its legislation.
Still, this does not solve the problems faced by persons such as Ria. A more principled solution is therefore needed, something that is considered in Section 6.
6. Employees who paid contributions in the ‘wrong’ country
6.1 The case of Mark
Mark lives in the Netherlands and works full-time as an employee for a small enterprise in Belgium, for which he and his employer pay employee contributions to Belgium. In addition, he works for the Ministry of Defence (army) in the Netherlands, from which he receives a small remuneration (about EUR 1,600 a year); his activities being considered as those of a civil servant. Mark paid contributions on the income from work earned in each country to the respective collecting bodies, since he and his employers did not appear to be aware of the conflict rules of Regulation 883/2004.
At a certain point, the Dutch tax administration (correctly) concluded that Mark was subject to Dutch social security only, as persons who are civil servants (as he was) are subject, according to the Coordination Regulation, and as discussed in Section 2 above, to the legislation of the State where the person is employed as a civil servant. This means that the Dutch contribution rules on his income as an employee in Belgium were applicable from the moment he started to work as a civil servant.
Consequently, the contribution rules were previously applied wrongly and contributions on his income from work in Belgium were paid in the wrong country. As a result, the Netherlands could still claim contributions on the income from work in Belgium made in the past. Mark could ask Belgium to reimburse the incorrectly paid contributions, but if benefits have been paid, there could be large complications.
In Mark’s case, the problem was solved on the basis of an Article 16 Agreement (mentioned in Section 2 above): the previous division of social security responsibilities between the two countries was ‘legalised’ and nothing had to be reimbursed. However, Article 16 agreements are made, according to the policy of most Member States, for a limited period only (in principle a maximum of five years). This meant that the solution found for Mark could not be used in future.
Mark and his Belgian employer considered the situation to be too complicated to continue after the five-year period expired. Not only was a lot of paperwork involved, but foreign rules, such as the obligation to continue to pay wages during 104 weeks in case of sickness, had to be applied. Therefore, Mark decided to terminate his employment in the Netherlands.
Of course, the effects Mark and his employer complained about could also occur if a person works in two countries as an employee or a self-employed worker. However, what is particular to the case of Mark is that the rules of the country where the employee is a civil servant apply even if the activities undertaken are marginal.
7. Analysis and recommendations
7.1 The common problems in our cases
The cases discussed in the previous sections have more in common than appears at first sight.
First of all, the problems described follow from a combination of differences between national schemes and the application of the Coordination Regulation. The persons concerned would be ‘better off’ without the Regulation. This does not contradict the positive meaning of the Regulation in other situations, but here the analysis has focused
The second common element is that the Regulation does not offer (direct) solutions if problems are a result from differences in waiting periods or pension ages, or from an incorrect application of the coordination rules.
The third element is that there is no obligation for Member States to solve problems by, for example, harmonising pension ages or waiting periods. The reference to the loyal cooperation of Member States ‘only’ provides a basis for a solution.
The fourth element is that a solution is possible on the basis of Article 16, but Member States are free to decide whether or not to conclude such an agreement, so individuals are fully dependent on the benevolence of the Member States involved.
There are also differences between the cases. In the case of the differences in waiting periods there is an obligation to fill the gap. The Court of Justice has even mentioned a reference framework that is applicable, although it does not define exactly which Member State has which obligation. In the case of ‘mini jobs’ there is no obligation for the non-competent (rather than the competent) State to seek a solution.
7.2 Should the State of Residence be responsible for a solution?
At first sight it seems attractive to point to the State of residence to find a solution. Sometimes, indeed, the State of residence feels responsible: Belgium solved the income gap of foreign unemployment benefit recipients, as seen in the case of John. Another example is an amendment accepted by the European Parliament to the text of the proposal for the revision of the Coordination Regulation (amendment 23),
23
which provides as follows: ‘(6 g) Where, owing to a mismatch between social security systems, a group of persons working in a Member State other than their Member State of residence are, as a result of the provisions of Articles 45 to 48 TFEU, placed at a disadvantage in comparison with those who have not availed themselves of freedom of movement for workers, in so far as they are, for a certain period, given a significantly lower level of protection than citizens of the Member State of residence, and where the matter cannot be resolved under the coordination rules, the Member State of residence of those citizens and their families should, in agreement with the Member State(s) concerned, find a way of remedying those disadvantages’.
It is argued here that different solution would be preferable. Since the gaps in protection that have been identified are not the same, two different proposals for filling them are proposed.
7.2.1 First proposal, concerning the case where a person is not insured in the State of employment or where conflict rules result in an ‘undesirable’ outcome
As seen in Section 4 above, making the State of residence responsible in a case such as that of Ria 24 is not consistent with the Regulation and is very complicated. As long as there is no chance of requiring the competent State, on the basis of the Regulation or Treaty, to find a solution for such case, an agreement on the basis of Article 16 seems the best instrument to use. Moreover, such an agreement also makes it possible to conclude an arrangement on the contributions to be paid.
One problem, however, is that, when a worker wants an Article 16 agreement to be made, he or she is completely dependent on the willingness of the benefit administrations to conclude such an agreement. Moreover, by definition two Member States are involved: a court decision in one country deciding that an agreement must be made cannot ensure that the other country is also bound by it. This creates a legal gap in protection.
A second problem is that it is uncertain whether such an agreement can provide for a sufficiently fine-tuned solution. For example, making those in a ‘mini job’ fully subject to the social security of the State of residence could result in higher contributions having to be paid, certainly compared with the income from the ‘mini job’ he or she is in. These workers may not receive all (tax) compensation for low income that are paid to persons in the State of residence, so for those working abroad it may be more expensive to be insured than for those who are not. Moreover, mini jobbers living in Germany may benefit from the insurance of their spouse without extra cost, whereas frontier workers may not enjoy such protection. Finally, if a mini jobber is made fully subject to the social security legislation of the State of residence, it will also affect their employer, who will most likely want to get rid of them.
Therefore, a patchwork solution is required. We propose to add to Article 16 of the Regulation: ‘If a worker or self-employed person makes an application to conclude an agreement as defined in the previous sentence, the competent body of the State of residence takes the initiative to conclude an agreement with the other Member State(s) that fits as well as *possible with the circumstances of the person concerned. This agreement has a duration, or is renewed, as long as there is a need for this. If one or more of the Member States involved have objections to the proposed agreement the competent body of the State of residence will inform the applicant of these in writing. The applicant can have these arguments tested by the Administrative Commission
25
or a body assigned by the Commission, which can give a decision binding both States’.
7.2.2 Second proposal, concerning differences in admission conditions for benefits (waiting periods and pension ages)
In respect of the differences between waiting periods and pension ages, such as those in the case of John, the reference framework is, following the Vester judgment, that of the country that would be applicable if the person concerned had not made use of the right to free movement. An example is that of a State from which a long-term benefit is due (such as old-age or disability benefit) but a waiting period following from its own legislation excludes the claimant temporarily from this benefit (and there is no other source of income for the claimant).
We take again the amendment of the European Parliament mentioned above and propose to change this as follows.
‘If a person who works or worked in a Member State other than their Member State of residence is, owing to a mismatch between social security systems, placed at a disadvantage in comparison with those who have not availed themselves of the freedom of movement for workers, to the extent that they are deprived from coverage or do not have any coverage at all, whereas he or she paid contributions for the benefit concerned and where the matter cannot be resolved under the coordination rules, the Member State the State of residence takes the initiative to conclude an agreement with the other Member State concerned to bring the situation of the person concerned in as far as possible to the state it would be if no use was made of the right of free movement. The other State concerned is bound to cooperate with finding a solution by means of the agreement. In case no agreement with the other State can be reached, the State of residence pays an advance that corresponds with the solution sought and this State can ask the Administrative Commission to provide a binding solution. The State of residence can ask the Administrative Commission to impose the obligation on the other Member State to compensate the costs made for finding this solution’.
This provision could be included in Title I of Regulation 883/2004 (after Article 10 on overlapping of benefits) or in Title III. 26
8. Conclusions
In this contribution, the cases of three mobile workers, all frontier workers, who found themselves in the situations of no protection have been examined. The Court of Justice could not provide a (final) solution to these problems. Also, the Regulation could not do this and Member States have not always appeared to take a responsibility for it.
Gaps in the protection of mobile workers are created by differences between national systems. The Regulation cannot take these away since there is no legal basis for harmonisation in the Treaty. Principles that can be used to decide which country has the responsibility for providing a solution have therefore been sought.
With respect to differences in the waiting periods for accession ages, the Vester judgment was investigated. From this it follows that the system of the country that deprives those who have made use of the right to free movement, and have paid contributions for these benefits, provides the reference framework for the solution.
For other cases, an adjustment of Article 16 on the agreement that can be made between two Member States can be of help. For this purpose, proposals to provide the individual mobile worker with more possibilities of requiring Member States to conclude such an agreement and ensuring that this agreement is adjusted to the individual situation, have been made.
Footnotes
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.