1. Introduction

Free movement of persons is a fundamental freedom of the EU. It lies at the foundation of the European Economic Community (EEC) and is an important condition for realising the internal market. Free movement can help to reallocate labour where it is most needed and thus it can contribute to economic growth. In other words, the market is the (original) source of this fundamental right. Furthermore, free movement and equal treatment of workers have been closely intertwined from the very moment establishing the EEC. If workers have the right to work in another country, but can be denied a job on grounds of nationality, if lower wages are paid compared to national workers or access to social security is restricted, they will have to return to the State of origin and thus the right of free movement becomes illusory. This is still true nowadays and for this reason equal treatment is essential to the right of free movement.

Yet, free movement and equal treatment of workers have come under pressure in the past decades. One factor was the creation in 1992 of a Treaty right to move and reside granted to all EU citizens (now Article 21 of the Treaty of the Functioning of the European Union (TFEU)) explicitly detaching free movement from its economic foundation and the interest of the market.1 A second factor was the accession of new Member States in 2004. This led to increased disparities in wealth and social protection across the borders. The disparities between the Member States encourage movement of citizens (both economically active and inactive) from the poorer to the richer Member States. It also meant that entrepreneurs in the poorer countries have made use of the right of free movement, this time of services, by accepting work in the richer Member States. They were and are often able to compete with lower wage rates for their workers. Both this social development (awarding the right to free movement to EU citizens) and market development (persons and enterprises seeking work in Western States) in a context of important differences between Member States have impacted negatively on the principle of equal treatment underpinning the free movement of persons and the social dimension of the market. Indeed, social protection of EU migrants is mainly at the cost of the host State, which may threaten, in the view of the latter, its welfare system. Furthermore, service providers can pay lower wages than the rates applicable in the host country, which can threaten social cohesion in the host State and lead to loss of jobs in some sectors. It is thus frequently argued that free movement of persons can lead to welfare tourism threatening the welfare State and to social dumping when EU workers are working at lower wage rates. However, in general the available data put the actual occurrence of welfare tourism and social dumping into perspective, inter alia because the figures on persons exercising their right of free movement are still relatively low.2 Yet, heated discussions and related fears do exist and have had important influence on Member States’ policies and on the case law of inter alia the Court of Justice of the European Union (CJEU). Tensions linked to free movement of persons have been further exacerbated following the refugee crisis of 2015. Migration, mobility and free movement are thus considerably more contested and negatively depicted than before. This negative sentiment reached a provisional peak in the pre-Brexit referendum discussions (see also Barnard and De Vries in this Special Issue).

In order to contribute to an analysis of this situation and propose some recommendations, we will highlight and discuss the social and economic conditions of the free movement of EU citizens. We will discuss their relevance in the post Lisbon context of a social market economy. These analyses may help to understand why particular distinctions are made and how to assess current developments.

This contribution proposes to investigate three forms of mobility within the EU, that of workers (Section 2),3 of EU citizens, i.e. economically inactive persons (Section 3) and of posted workers (Section 4). In respect of all three categories, there is a tension between social and market interests, which appears most clearly, when we compare the right of equal treatment of these respective categories with workers or persons in the host State. For each of the distinguished groups we will discuss which equal treatment perspective is chosen by the EU legislature, the underlying rationale and how this can be maintained in a social market economy. In so doing, we examine the legal developments against the three essential elements underpinning the economic model praised by the EU: the social, the economic and the market. Yet, these concepts are hard to define and depend to a certain extent on the eye of the beholder. For the purpose of this contribution, by the term social we will essentially refer to the level of protection of migrant EU citizens in the host State compared to nationals. The ‘economic’ element relates, in this contribution, essentially to budgetary considerations of the host State, where EU citizens can be considered, depending on the situation, somewhere between an asset and a burden. The market concept is closely tied to the economic considerations just mentioned, but is more specifically linked to the core concepts of free trade and competition between undertakings. This means that in order to realise an internal market, barriers to offer services across borders have to be removed as well as restrictions on competition. In the context of a social market economy, the freedom of undertakings and competition are restricted, if necessary, by social values. This balancing between market and social values is at the core of this research. Thus, when discussing legal developments within the context of the free movement of people, we will turn to these concepts and assess how they are weighted against or linked to each other and how they have evolved over time.

We thus expect that the social market economy will, on the one hand, explain the ‘why’ in discussing the right of equal treatment of free movers and the limitations to this right. On the other hand, the social market economy has also a normative dimension, which enables us to discuss perspectives on how equal treatment can develop. We argue that the social component in the market economy requires a revision of what is seen as fair competition in the direction of outlawing social dumping. We also find that the persistence of the market elements pleads for allowing a free movement of persons, which does not substantially disturb the economy of the host Member State. In our discussions and conclusion, we propose some avenues to address the tensions linked to the free movement of persons in the perspective of a social market economy.

2. Free movement of workers: a fundamental freedom under pressure

The right of free movement of workers has been the cornerstone of the EU since the creation of the EEC in 1957. It has a firm economic basis but also a social dimension as demonstrated, for example, by the broad interpretation of the concept of worker. Yet, such social dimension has recently been contested at national level. We will conclude this section by proposing a few avenues to address the tension between the economic and the social.

2.1. The case law of the CJEU: a broad interpretation of ‘workers’

This freedom has, like the other freedoms, a firm economic basis: it aims at ensuring that labour is allocated where it is needed most, stimulating movement of workers from areas with high unemployment to places with shortages of labour. Such reallocation certainly benefits the economies of the Member States and is necessary for an internal market. Furthermore, workers from other Member States have to be treated equally compared to national workers, as is stipulated by Article 45 of the TFEU and Regulation 492/2011 (and their predecessors).4 Equal treatment is seen as necessary, as otherwise the right of free movement may be hampered; if EU workers are not put on the same footing as the nationals of the host State, they may not make use of this freedom. Moreover, if foreign workers are cheaper than the national workers and are recruited on a large scale, the social system of the host State would be undermined. This would put pressure on wages and on the social contributions that are paid to the social funds in the host State.

This argument combines both economic and social rationales. In its case law, the CJEU does not ‘weigh’ the relevance of the economic factors in each situation to grant free movement and equal treatment. It is sufficient that a person performs an economic activity, however small. Workers are thus protected because of their status as workers, regardless of the amount of the contribution to the host State as long as the threshold for being a worker is met. They are assumed – as a category – to contribute to the economy of the host State where they pay taxes and social security contributions.5 Equal treatment extends to all social advantages6 and workers must not be discriminated on the basis of nationality as regards all welfare benefits, including social assistance,7 in the country of work (with the exception of war compensation benefits).8

The question remains of who qualifies as a ‘worker’ and thus deserves equal treatment. According to the CJEU, e.g. in the Blum judgment,9 the term ‘worker’ in Article 45 of the TFEU and Regulation 492/2011 has a Community meaning and, since it defines the scope of a fundamental freedom it must not be interpreted narrowly. The essential feature of workers is that they perform work for a certain period of time for and under the direction of another person or organisation for which they receive remuneration. Only EU citizens, who perform activities on such a small scale that they are purely marginal and ancillary are not considered to be workers.10 Such broad interpretation was further justified on the social ground of raising the standard of living of EU citizens, one of the objectives of the Community. As the CJEU explained in Levin, ‘(…) part-time employment, although it may provide an income lower than what is considered to be the minimum required for subsistence, constitutes for a large number of persons an effective means of improving their living conditions’.11 Sometimes it is argued that workers with a low wage or a very short work history in the host State may be a burden for the welfare State if they have to be treated equally from the first moment. Such workers can have access to study grants, even if they work in a small part-time job or they can claim an income support benefit if they have a low income (if such benefits exist in the host country). The CJEU has, however, not changed its interpretation of the term worker so far. On the contrary, it has maintained its broad interpretation, including persons who have worked for a very short period and/or in a job with minor income. The Court recently confirmed its inclusive approach in Tarola, where it ruled that two weeks in employment might be sufficient to be regarded as a worker and maintain this status upon redundancy for at least six months.12 Indeed, workers who lose their job retain their status as workers and the protection thereof under certain circumstances. The CJEU was the first to pin down this legal principle, which was subsequently endorsed by the EU legislature (see also Article 7(3) of Directive 2004/38 (Citizenship Directive)).13

There have been some limited exceptions to the principle of equal treatment concerning workers who are not living in the country where the social benefit are claimed, such as the judgments Hartmann14 and Geven.15 In the latter case, the link with the host State – where the worker was not living and where she worked in only minor employment – was considered too weak to grant equal treatment. Also in respect of children of frontier workers,16 the CJEU considered the sole link of working by the parents as insufficient and required additional links.17 Yet, this does not mean that the broad concept of ‘worker’ has been abandoned by the Court. In contrast, in its recent ruling in the Tarola case mentioned above, the Court made clear that while it is up to the Member States to exclude persons with short employment from social benefits, they cannot discriminate on grounds of nationality in accessing the benefit.

2.2. Tensions at national level

The CJEU’s broad approach to the concept of worker has not always been followed by national authorities as some interpretations are more restrictive. In the UK, for instance, a person performing work must have an income above the minimum earnings threshold, i.e. GBP 155 (approx. EUR 207) a week, in order to be considered a worker, which amounts to 23 hours of work a week for a person earning the minimum wage.18 In Vatsouras, the Court considered that a wage of EUR 169 a month below the minimum subsistence level was not necessarily too small for Mr. Vatsouras to qualify as a worker.19 Furthermore, if we reason in terms of working hours, 5.5 hours work a week in the Genc case, could depending on the circumstances, be seen as genuine and effective work pursuant to Article 45 of the TFEU.20 In the Netherlands, EU citizens are considered as workers if they earn more than 50% of the applicable public assistance rate, i.e. more than 50% of EUR 1,524 per month for a couple, or more than 50% of EUR 1,060 for a single person in 2016, or if they work at least 16 hours per week.21 This requirement seems higher than that discussed in the Genc judgment. There is no general overview of the conditions in other Member States, but there seem to be different approaches, often more based on discretion by civil servants than legislation. It is obvious that this creates serious legal uncertainty.22

These national interpretations stand in contrast to the case law of the CJEU, in which the fundamental rights of migrant workers to free movement and equal treatment have been given a broad meaning. If national authorities do not recognise these rights for EU citizens in small jobs, this would lead to unequal treatment in accessing social rights and in denying them social citizenship. Some Member States claim that admitting marginal foreign workers with a short working history results in a burden for the State. Economic grounds thus restrict the social rights of these workers. For example, some Member States have attempted to curtail rights of workers to some social benefit on an equal treatment basis. The Settlement Agreement between the UK and the EU prior to the British referendum on Brexit is an example of this.23 Also, Denmark adopted an Act subjecting the right to family benefits to a condition of two years’ residence for payment of full benefits. Following pressure from the EU Commission, practice has changed but the law was not revised, as no majority could be found to amend it.24 The UK has also adopted a policy to restrict access of migrants to social benefits in order to shield the alleged magnetic pull of its welfare system.25 Adding conditions for being a worker under Article 45 of the TFEU, granting discretion to the administrative services or allowing a difference between law and practice all create legal uncertainty. This may seriously undermine the right of free movement.

2.3. Possible avenues for reconciliation of the social and the economic

How can the concerns of the Member States possibly be reconciled with the free movement of workers? Member States can make their systems less vulnerable to access by foreign migrants in marginal jobs. They can, for instance, take measures to encourage larger and better paid jobs; introduce fiscal advantages for employers for giving larger jobs, give better education for workers, introduce (higher) minimum wages etc. States could also reduce the number of income support schemes and other non-contributory welfare schemes. This measure is more controversial as it impacts negatively on their ‘own’ citizens and might backfire against the EU. However, there are alternatives that could realise a similar protection while reducing the burden problem. Take a country with a national health system free at the point of access. It could be difficult to shield its access from ‘outsiders’, who might easily be targeted as a burden even if they pay taxes. An alternative is a health system in which only those who are insured have access. Instead of selecting those who are not entitled to health care by difficult criteria, such as habitual residence, such system obliges residents to be insured, contribute and enable them to better ‘track’ the money. Persons with low income could be covered by an income subsidy scheme, so that the social objectives are not undermined. Under the first system, foreigners are more seen as a threat, in the second more as a source of income.26

Solutions can also be found at the EU level. For example, some academics have proposed to amend the concept of worker, in particular in Regulation 492/2011, which could further specify who is a worker.27 For example, the threshold for being a worker could be set at a minimum of 40 per cent of the minimum wage, thus making clear that stricter definitions are prohibited. Such proposition would reduce concerns of benefit tourism and/or EU nationals being a burden on national schemes for student grants and loans. After all, the difference between a person working in a very low productivity job and a person not working at all is not always easy to justify. This proposal enhances legal certainty but might be challenged before the CJEU. The danger of revising the term ‘worker’ itself is that it has an impact on all aspects of Article 45 of the TFEU (thus not only on the access to social advantages). Such revision would allow employers to pay lower wages to EU citizens than national workers, exclude them from a particular job on ground of nationality or affect their right to seek work in another Member State. A more acceptable variation of the same proposal is that the thresholds depend on the type of benefit and only apply to specific benefits. Income support topping up low wages and study grants are examples of such benefits. This can be achieved through amending Article 24 of Directive 2004/38 and Article 7 of Regulation 492/2011, adding a few limited exceptions to the principle of equal treatment. This is a more subtle approach. If it is desirable that workers must, say for the first five years, have no access to a particular income support benefit if they work less than 20 hours a week, it is better to state this in the equal treatment rule than to change the concept of worker as a whole. Again, there might be a risk that the CJEU would find such derogations in conflict with the Treaty, this time especially with the principle of equal treatment enshrined in Articles 45 and 18 of the TFEU. Yet, the Court has traditionally granted a great leeway to the EU legislator, which is much broader than that given to national legislators and authorities restricting free movement. Of course, this proposal requires that a majority for the revision of the Regulation can be reached within the Council and the European Parliament.

A last proposal is to give Member States the power to define the term worker and thus legitimating the practice actually taking place in some Member States. The idea is that in order to save the fundamental principle of equal treatment of workers, Member States should be allowed to rely on their national definitions of the concept of worker, provided that they apply to their own citizens.28 In this way, equal treatment will be strengthened by extending its application to the very concept of worker. This proposition has the advantage of firmly anchoring free movement in the principle of equal treatment. It thus avoids the danger of allowing exceptions to equal treatment, which might be a slippery slope. In addition, it might be more politically acceptable as it rests on national definitions and competences, and has thus greater chances to be adopted. Yet, this would inevitably lead to increasing national divergences in protection and legal uncertainty. Furthermore, leaving the issue to national authorities increases the risk of more escape routes in order not to apply equal treatment. For example, a country could define a worker in a particular sector, where many foreigners are working, as a person working at least 30 hours a week, while having a different definition in other sectors. Another problem is that national systems often, if not always, have different concepts of worker depending on the regulation it is part of. For example, for employment relationships, for social security or for tax law there may be different definitions and there are good reasons for this, so it is not easy to change. Which one should then be used for residence purposes?

We can conclude that there are several proposals for addressing the problems both at national and EU level. We tend to prefer the second proposal – that the thresholds depend on the type of benefit – since it fits best with maintaining the right of free movement of workers. In the next section, we will discuss the issue of economically inactive EU citizens, who have a more limited right of free movement and protection against unequal treatment.

3. Free movement of economically inactive EU citizens

Before the Maastricht Treaty introduced the concept of EU citizenship and attached rights, in 1990 the EU legislature adopted three Directives extending the right of free movement and residence to EU citizens who are not economically active, such as students, pensioners and a rest group.29 The rationale for this extension of free movement rights was linked to the realisation of the internal market with the abolition of internal frontiers securing free movement for goods, persons, services and capital. In addition, the preamble of the Directives also referred to the activities of the Community, which includes the abolition of obstacles to the free movement of persons. Furthermore, the economic rationale underpinning the Directives is also evident from their focus, precluding inactive persons becoming an unreasonable burden on the State finances. Yet, the legal basis was not found in the internal market provisions (now Articles 94 and 95 of the TFEU) enabling the EU to remove obstacles to free movement and competition. Instead, they were all initially based on the residual provision (now Article 308 of the TFEU) granting competence to the EU in areas where the EU does not have the necessary power to legislate and requiring unanimity within the Council.30 The Directives thus transcend to a certain extent the internal market logic by its choice of legal basis and its focus on the inactive.

3.1. The introduction of citizenship and the CJEU’s initial approach to equal treatment

The Treaty of Maastricht, adopted in 1992, introduced a new concept: EU citizenship in (what are currently) Articles 20 and 21 of the TFEU. EU citizens were thus granted a Treaty-based right of free movement regardless of economic activity. Yet, this right is, as the articles read, ‘subject to the limitations and conditions laid down in this Treaty and by the measures adopted to give it effect’. It thus follows that the right of free movement of citizens is subject to more conditions than that of workers. Workers are, of course, also citizens, but Article 45 of the TFEU has priority over Article 21 of the TFEU. Hence, for the purpose of this article, the expression free movement and equal treatment of EU citizens refers to economically inactive EU citizens. In the TFEU, the right to equal treatment is not connected to the provision on free movement of citizens itself, but the CJEU explicitly linked the two provisions together. It follows from Article 20 of the TFEU that EU citizens enjoy the rights ensured by the Treaties, and this includes protection against discrimination on ground of nationality enshrined in what is now Article 18 of the TFEU (Martínez Sala judgment).31 In the Grzelczyk judgment,32 this approach was confirmed. Since Belgian students could receive social assistance, the exclusion of Mr. Grzelczyk solely on the ground of nationality was not allowed. Before this judgment only workers and self-employed persons enjoyed protection against discrimination in accessing social benefits, so the new provisions entailed an important extension of rights. As the Advocate-General remarked in the Föster case,33 the concept of EU citizenship, as developed by the case law of the CJEU, marks a process of emancipation of Community rights from their economic paradigm. Community law rights – in particular the right not to be subject to unjustified discrimination – are no longer bestowed upon citizens solely when they make use of the economic freedoms and assume a corresponding status (worker, provider of services etc.), but directly by virtue of their status as a citizen of the EU.

However, since then, the approach of the CJEU has importantly changed. In the early stage of EU citizenship case law, equal treatment of citizens with national citizens was central and the question of objective justification to exclude persons from a right on the basis of nationality was not raised. Most likely Member States had never thought that their social schemes would be involved in a discrimination dispute and thus that it had to produce justifications. The Bidar judgment was the first in which the CJEU accepted that exceptions to equal treatment are possible.34Although Member States are required to show a certain degree of financial solidarity with nationals of other Member States, it is permissible to ensure that students from other Member States do not become an unreasonable burden, which could have consequences for the overall level of assistance, which may be granted by that State. For this purpose, ‘a certain degree of integration’ can be required. However, the conditions to ascertain such a degree of integration must be appropriate and proportionate. Member States were not always successful in giving justifications. For instance, in D’Hoop, the Belgian scheme on a maintenance allowance for school-leavers required claimants to have attended secondary school in Belgium.35 The CJEU considered that Belgium could have required a link with the labour market for this unemployment benefit scheme, but not with the country where school was attended. In contrast, in Förster – issued after the implementation period of the Citizenship Directive 2004/38, but the facts of the case were before that – the Court accepted that a condition of five years of prior residence for study grants was appropriate for the purpose of guaranteeing that the applicant is integrated into the society of the host State and proportionate.36

We can conclude from this brief overview, that the Treaty provisions and the case law extend the right of free movement and equal treatment to economically inactive persons. However, Member States can legitimately require a link to their territory in respect of certain claims before economically inactive EU citizens can claim equal treatment. This approach fits well with the concept of social citizenship requiring equal treatment for those belonging to a community. The link approach is useful to assess when one belongs to a particular community. Article 18 in conjunction with Article 21 of the TFEU do not remove all nationality discrimination, and thus follow a different approach than Article 45 of the TFEU that only accepts justifications for direct discriminations on grounds of public health, order and security. Unlike workers, economically inactive persons cannot take the right of equal treatment for granted, but have to earn it. This is an issue discussed in the following section.

3.2. A different positioning of workers and economically inactive persons

In Commission v the Netherlands, the CJEU had to explain the difference in treatment of workers and economically inactive EU citizens.37 The dispute concerned the policy of the Netherlands that had, following the Förster judgment, excluded children of frontier workers from exporting study grants. Yet, the Court had consistently ruled that, in principle, the equal treatment rules of Regulation 492/2011 also apply to family members.38 The question was therefore whether, after the Förster judgment, the limitations to equal treatment of economically inactive persons could extend to workers. The Court decided that this was not the case; it considered that although budgetary considerations may underlie a Member State’s choice of social policy, they do not in themselves constitute an aim pursued by that policy and cannot therefore justify discrimination against migrant workers. The Court addressed the issue of the difference between economically inactive persons and workers. It considered that participation by migrant workers and frontier workers in the employment market of a Member State establishes, in principle, a sufficient link of integration with the society of that Member State. This allows them to benefit from access to social advantages on the same footing as national workers. Migrant and frontier workers are thus per se integrated in the host country. This rests, inter alia, on the circumstances that they pay taxes there and thus contribute to the financing of its social policies.39

It is the first time that the CJEU argued for equal treatment of workers on the basis of the economic link with the host State making an explicit reference to the payment of taxes. Yet, the Court still followed a macro perspective and argued that workers generally pay taxes and contribute to the economy of the host State. For this reason, they have a link with the host State. The Court did not examine whether this is true in the individual case. For economically inactive persons, there is no direct economic link with the host State. Therefore, they are not automatically integrated in that country, but have to demonstrate such integration link before equal treatment can be claimed.40

It thus follows from the CJEU’s case law that there is a priority order. Article 21 of the TFEU finds specific expression in Article 45 of the TFEU in relation to freedom of movement for workers. This means that when Article 45 of the TFEU is applicable, the restrictions connected with Article 21 do not apply.

3.3. The impact of Directive 2004/38 on the position of EU citizens

Directive 2004/38 specifies the right to reside, which is granted to all for up to three months and for longer to those who are workers, self-employed or have sufficient resources for themselves and their family members and thus do not become a burden on the social assistance system of the host Member State (e.g. with Article 7). The Directive also contains some exceptions to the principle of equal treatment. as enshrined in its Article 24. Thus, the host Member State is not obliged to confer entitlement to social assistance during the first three months of residence or for a longer period in respect of first-time jobseekers. Furthermore, a Member State is not obliged, prior to the acquisition of the right of permanent residence (i.e. after five years), to grant maintenance aid (i.e. study grants or loans for studies) to persons other than workers, self-employed persons (and persons who have retained such a status) and members of their families. This rule deviates from the Bidar judgment (discussed supra), but fits well with the Förster judgment, where the CJEU referred to the Directive to support its conclusion. After five years of legitimate residence, EU citizens are, according to Directive 2004/38, entitled to the right to permanent residence and no further conditions for the residence right or equal treatment can be imposed.

After the entry into force of the Citizenship Directive, the non-discrimination rules were restricted considerably in the CJEU’s case law. Its restrictive line began in Brey and was reinforced in especially Dano and Alimanovic. The Brey judgment41 concerned a claim for a special non-contributory benefit and the question was whether it could have the same effect as claiming social assistance. Special non-contributory benefits are benefits at a subsistence level meant for specific risks, e.g. that of old aged or disability. For Regulation 883/2004, the coordination regulation for social security, these benefits are not considered as social assistance. Consequently, it could be expected that they are not seen as social assistance under the Citizenship Directive either. Yet, the CJEU considered that, although special non-contributory benefits are indeed not social assistance for Regulation 883/2004, they are to be considered as such for Directive 2004/38, since the objectives pursued by Regulation 883/2004 are different from those of the Directive. Regulation 883/2004 intends to remove the possible negative effects of the exercise of free movement on social security benefits. While Directive 2004/38 is also meant to facilitate and strengthen the exercise of the right of free movement, it also requires that inactive EU citizens need to have sufficient resources in order to allow Member States to protect their public finances. Yet, the CJEU emphasised that claiming social assistance does not automatically mean that the person does not have sufficient resources to avoid becoming an unreasonable burden on the social assistance system of the host Member State and therefore loses the right to residence. The national authorities must first carry out an overall assessment of the specific burden granting that benefit would place on the national social assistance system as a whole, by reference to the personal circumstances characterising the individual situation of the person concerned.

This tailor-made approach was abandoned in the Dano judgment,42 where the CJEU followed a radically different approach. Ms. Dano was refused a German unemployment benefit, a special non-contributory benefit, since foreign nationals whose right of residence arises solely out of the search for employment were excluded from it. The Court first found that Ms. Dano had neither worked in the host State, Germany, nor in the country of origin, Romania, and there was nothing to indicate that she had been looking for a job. The Court then made a firm statement: EU citizens can claim equal treatment with nationals of the host Member State only if their residence in the territory of the host Member State complies with the conditions of Directive 2004/38. This means that they have a right of residence in the host State only as long as they do not become an unreasonable burden on the social assistance system of the host Member State. According to the referring court, the applicant did not have sufficient resources and the CJEU ruled that she could not claim a right of residence in the host Member State under Directive 2004/38. Consequently, she could not invoke the principle of non-discrimination in Article 24(1) of the Directive and could be disqualified from the benefit.

The fear of the impact on the public funds appears to have been decisive for the CJEU to leave its case-by-case assessment of the Brey judgment and the link approach followed in the Bidar judgment– in any case for persons in the situation of Ms. Dano. It seemed to have read the newspapers and provided a reassuring and clear message to those fearing benefit tourism. Thus, economic arguments have gained weight in the balance against social rights. Yet, it could still be argued that the situation of Ms. Dano was special since she had (according to the German court) made clear that she was not seeking work, and does therefore not exclude more nuanced approaches like the one adopted in the Brey judgment.43 It is remarkable that the persons concerned can still have the right to reside according to national law (Ms. Dano for example), so they are not expelled from the country, but the right to equal treatment is disconnected from this. The Brey and Dano case law was extended to a broader range of benefits than that of social assistance and special non-contributory benefit. Indeed, the Commission v United Kingdom judgment concerned non-contributory social security benefits, such as child benefits, which were conditioned to the right to reside in the UK in accordance with Directive 2004/38.44 The CJEU ruled that such condition of lawful residence in respect of economically inactive EU citizens was not contrary to the social security regulation. Interestingly enough, this ruling was given just a few days before the British referendum on Brexit.

It is noteworthy that Directive 2004/38 also had an impact on a particular group of persons making use of the freedom to move under Article 45 of the TFEU, i.e. jobseekers. Although they have never been considered workers in the sense of Article 45, since they do not satisfy the criteria of worker mentioned in Section 2.1, they can derive some rights from this provision including those mentioned in Regulation 492/2011. For instance, Article 5 provides that a national of a Member State who seeks employment in the territory of another Member State shall receive the same assistance there as that afforded by the employment offices to their own nationals seeking employment. However, in Directive 2004/38 (first-time) jobseekers are in a way treated as economically inactive persons. Consequently they can be excluded from social assistance by the Member States during three months and as long as they are seeking work (Articles 24(2) and 14(4)(b) of Directive 2004/38). This provision had an impact on their access to so-called jobseekers allowances. The CJEU ruled in Collins that in view of the establishment of citizenship of the EU, it was no longer possible to exclude from the scope of Article 45 of the TFEU a benefit of a financial nature intended to facilitate access to employment in the labour market of a Member State.45 Yet, access to such benefits could be subject to a condition of sufficient connection (real link doctrine) to the national labour market of the host State. In Alimanovic, the Court departed from this approach and ruled that when a jobseeker allowance both facilitates access to the labour market and provides minimum means of subsistence, it constitutes social assistance, from which jobseekers can be disqualified (Article 24(2) of the Citizenship Directive).46

3.4. Assessment of the position of EU citizens and possible solutions for improvement

The right of free movement and equal treatment of EU citizens has been subject to important developments. They are evidence of an important step in favour of social citizenship, although they are not as unconditional as those of workers. Yet, the fear of benefit tourism seriously influenced the interpretation of the provisions of the Citizenship Directive by the CJEU and led to important restrictions of their rights and legal uncertainties. Although the existence of broad political support for denying social assistance to persons in an early stage of their stay in a host State is difficult to deny, this cannot be the (sole) basis for defining the legal position of EU citizens. Indeed, what is the meaning of the right to free movement of economically inactive citizen where because of disparities in economic wealth between Member States, EU citizens have insufficient income or capital in their State of origin to spend some time abroad?47 It can thus be argued that for the time being such general right of free movement becomes illusory for a great number of EU citizens.48

Secondly, it can be said that the legal position of economically inactive persons has become quite unsure. A claim for a subsistence benefit can mean that one has no (longer) a legal right to stay; this can also mean that these persons do not acquire a right to permanent stay after five years, since they cannot prove that they fulfilled the condition of lawful residence in this first-five year period. It is not clear yet whether the nuance in the Grzelczyk judgment (a person having provided for his own needs during three years encountering temporary financial problems could not be denied social assistance) still applies. After all, the judgments discussed above, i.e. Brey and Dano, concern persons in an early stage of their stay or who seemed to be permanently unable to earn their own living and are thus in a different position than Mr. Grzelczyk. However, the CJEU uses quite absolute words in denying the right of equal treatment in Dano and Commission v the UK, so a general exclusion from the right of equal treatment until a permanent right to stay is reached is very well possible.49

Finally, the issue is not so much whether Ms. Dano has been correctly denied the right to social assistance, but more in general, the legal uncertainty for EU citizens. It is therefore time to rethink their right of equal treatment. This is not an easy task because of the considerable differences between the Member States. Although one might be tempted to make the State of origin more responsible for the subsistence income of their citizens moving abroad, this raises huge problems since several countries have no such system or have only very low benefits.50 It is therefore preferable to look for a solution in the host country. After all, also for persons who worked in the host State and then become unemployed the responsibility lies with that country. Yet, in view of the large differences between social systems of Member States, equal treatment of economically inactive persons upon entry into the host country creates financial and social cohesion problems. A complete denial from equal treatment for five years if a person has no subsistence means is, on the other hand, a very radical approach and a barrier to their free movement rights. Therefore, the real link approach developed by the CJEU in its earlier case law, is a good alternative: it allows for nuanced outcomes and fits well in a policy to prevent benefit tourism. We therefore recommend to ‘include’ this approach in Directive 2004/38.51 In this way, it will allow for a certain flexibility while not being too intrusive on the principle of legal certainty: the longer EU citizens have stayed in a country, the more rights they acquire. The Grzelczyk case law was, for instance, elaborated in Dutch policy rules so that EU citizens and their families claiming social assistance (benefit or lodging) are less easily evicted the longer they have resided in the Netherlands. Making use of night shelters was not considered an unreasonable burden if used less than eight nights in the first two years of residence, less than 16 nights in the third year, less than 32 nights in the fourth year and less than 64 nights in the fifth year. Generally, unless personal circumstances oppose so, a request for social assistance payment in the first two years of residence means that one does no longer meet the conditions under which the permission to stay in the Netherlands was granted. From the third year of residence onwards, it is possible to apply for social assistance without being expelled. No expulsion measures are taken in case of the following claims: in the third year during less than two months for less than 50 per cent of the subsistence benefits norm or during less than three months for a more supplementary assistance (for an amount which is up to 50 per cent of the subsistence benefits norm), in the fourth year less than four months respective six months and in the fifth year less than six months respective nine months. Social assistance during less than 15 months in a period of three years can be received without being expelled.52 This is only an example of how the link approach can be translated into policy rules, but it shows how legal certainty can be increased within the framework of the link approach and accommodate the tensions between the economic and social.

4. Posted workers

The third form of movement is that of persons who are sent by their employers to work for them temporarily in another Member State. This situation is called ‘posting’ of workers. The CJEU ruled that posted workers cannot rely on Article 45 of the TFEU as they do not form part of the working community of the host State. They thus do not benefit from the principle of equal treatment. Posting takes place, in this interpretation, within the framework of providing services in other Member States (Article 56 of the TFEU). As we will see below, this makes competition on the basis of wages possible potentially leading to social dumping. Indeed, employers can rely on the rules of employment and social protection of the sending State and are not bound by the equal treatment principle of Article 45.

Posting of workers is a complex phenomenon, which has many faces. Yet, for the purpose of this article, we have to keep in mind that the problems linked with posting take place especially in specific sectors of the economy. These are sectors where work is mobile in nature, such as the construction and transport sectors. The construction sector has the highest number of posted workers in the EU (39% of the total number).53 Secondly, although posting can take place from any country to another, the discussion and controversies centre on posting from low wage countries to high wage countries.54 The number of postings in the EU has significantly risen with a total increase of 58% between 2011-2016.55 In principle, the general economy or labour market of the host country will not seriously feel the effects of such posting. Yet, in specific sectors and areas posted workers have in practice substituted considerable numbers of national workers.56 Finally, in practice even the minimum EU rules of protection of posted workers are often not followed on the ground, leading to abusive practices. In this section, we will first look at the legal development in the initial period, before the Posted Workers Directive of 1996 was adopted (Section 4.1) and then at how this directive was interpreted (Section 4.2). We will assess whether the recent revision of the Directive makes a change in the balance between social and economic interests in view of equal treatment (Section 4.3) and assess the fate of the Directive (Section 4.4).

4.1. The CJEU’s initial case law: an uncertain balancing between economic and social interests

The debate on which rules should be applicable to posted workers started in the 1980s.57 In Rush Portuguesa,58 the CJEU ruled that posted workers could not be considered as migrant workers under (what is now) Article 45 of the TFEU as they do not at any time gain access to the labour market of the host country. It must be kept in mind that this case concerned the issue of a fine being imposed on the employer established in Portugal for not having a labour permit for employees posted to France. The question of equal treatment of the workers was not at issue here. Yet, it follows from this judgment that EU law does not require posted workers to be treated equally with local workers. In terms of social citizenship, they do not form part of the community where they temporarily work. Thus, like in the citizenship cases discussed in the previous section, there is no sufficient link to the host country to trigger the application of equal treatment. In contrast, the undertakings posting the workers were protected against unjustified restrictions on their freedom to provide cross-border services. Thus, the criterion of competition between undertakings of the Member States seems to be decisive to regulate the legal position of the workers. However, in the Rush Portuguesa judgment, the CJEU also considered in an obiter dictum that the host Member State could impose its legislation or collective labour agreements on posted workers and enforce those rules (§ 18). Although this did not impose equal treatment in general, it gave Member States the freedom to treat posted workers in the same way as national workers and thus outlaw competition on the basis of labour law standards. Yet, in later case law, the Court gave preference to the possibility to compete and ruled that imposing national labour law can restrict the freedom to provide services and thus needed to be justified. The objective of social protection of workers could, in principle, legitimately justify the application of the legislation of the host State. However, there was uncertainty as to the extent of this protection, as some host States applied all their labour law rules to posted workers. The question then arose when competition should prevail and when labour law rules legitimately protected workers.

4.2. Downgrading protection: Directive 96/71 and the CJEU

In order to find a compromise on this issue, in 1996 the EC legislature adopted – after long deliberations – a Directive regulating the issue of posting of workers.59 The Directive’s objective was to enhance legal certainty by providing a clear set of rules to be followed. Thus, the Directive specified that the posted workers are subject to a nucleus of mandatory rules for minimum protection of the host State. A crucial part of this is Article 3(1)(c) that required and allowed imposing the minimum rates of pay of the State where the work is performed. When asked to interpret Article 3 in the Laval case, the CJEU considered that the EU is more than a market economy as it also has a social purpose. It followed that the economic freedoms such as the free provision of services shall be balanced against the social objectives of the EU including a high level of employment and adequate social protection.60 Thus, the ruling which was issued a few days after the signature of the Lisbon Treaty implicitly referred to the novel concept of a ‘social market economy’. Yet, the CJEU ruled that the provisions of the Directive on Posting of Workers should not only work as a floor for social protection of workers, but also as a ceiling. In other words, the minimum must be imposed, but not more than the minimum. A point of discussion was Article 3(7) that allows Member States to impose more favourable terms and conditions than those mentioned in Article 3(1). Could this mean that higher rates of pay than the minimum could be required? The Court decided in favour of the service provider that the host Member State can only impose the obligation to pay minimum wages to posted workers.61 According to the Court, Article 3(7) only refers to more favourable terms of the sending State (or of the chosen law).

This interpretation of the Directive and other aspects of the ruling led to massive criticism of the CJEU for favouring economic interests to the detriment of the social protection of posted workers and the workers of the host State. It already followed from the Posted Workers Directive that no equal treatment was required, but the case law of the Court added that nothing more than the minimum could be imposed. The way to equal treatment seemed completely barred. Equal treatment with the workers and protection against unfair competition on the basis of wages had to give way to the economic dimension, i.e. the free movement of services. Thus, the balance tipped in favour of the economic interests.

To some extent, in the CJEU’s rulings in Elektrobudowa62 and RegioPost63 slightly different tones can be heard. They lift the ceiling of social protection flowing from Laval by allowing a broader interpretation of the concept of minimum pay. In Elektrobudowa, it was permitted to include all elements of the wage structure pursuant to the collective agreement, provided that they were clear and accessible to the service provider. In RegioPost the Court allowed the local contracting authority to impose the wage provided for in the law of the German land.64 Nevertheless, the Court put much effort in distinguishing this case from its more restrictive approach followed in Rüffert.65 It emphasised that the regional public procurement rule in question laid down a minimum rate of pay, and that no other minimum pay existed for the specific sector.66 These two rulings are rather an accommodation of previous case law than a fundamental shift from it as they are still based on minimum protection, but now more broadly interpreted.67 While the Court has softened its approach to social dumping, it has not fundamentally altered it as it is bound by the framework of the Directive and its minimum rights logic.

4.3. From wage competition to equal treatment?

As we saw in the preceding section, the market paradigm based on wage competition prevailed to a great extent in the law on posting of workers. In 2018 there seemed to be an important change, when the Directive on Posting of Workers was revised to introduce the principle of equal remuneration.68 In this section, we will discuss whether this constitutes a genuine departure from the previous approach. We will assess its further implications in terms of shift of values underpinning the internal market. Is the revised directive a step towards a more social market economy, where social protection of workers prevails over competition between undertakings?

The revision of the Directive was preceded by pressure of, in particular, the French, British and Dutch Governments and the President of the EU Commission. They intended to put the wage competition, which followed from the Directive on the agenda. They argued that wage competition undermines sectors of the economies of countries with higher wages and creates the risk of social dumping, i.e. a deterioration of wages and social protection in these sectors. In fact, this effect was the result of the differences between the old and the new Member States, or between the well-developed economies and the economies still in development. This tension can be clearly seen when the ‘new’ Member States, with (the somewhat remarkable) support of Denmark, drew the so-called yellow card to argue that the proposal of the Directive was contrary to the subsidiarity principle. The substantial objections to the Directive seemed, however, to have been the real reason for this procedure.69

The revised Directive has the Treaty provision on free movement of services as a legal basis, although it also refers to social provisions such as Article 3 of the TFEU with its focus on a social market economy and Article 9 of the TFEU referring inter alia to a high level of employment and social protection. The revised Directive no longer mentions minimum rates of pay, but remuneration (so no longer minimum remuneration) as the core of labour law of the host State to be imposed. For the purpose of the Directive, remuneration means all the elements of remuneration rendered mandatory by national law, regulation or administrative provision, collective agreements or arbitration awards, which have been declared universally applicable. In the absence of the latter systems, other collective agreements or arbitration awards can be used. Such mechanisms were already mentioned in the previous Directive (Article 3(8) second subparagraph).70 These provisions are a remarkable development, since social interests, in any case those of the workers in the host countries have been given more weight. Posted workers shall receive the same wage as national workers in so far as this is laid down in the instruments mentioned above.71

This is a very interesting development. One might wonder whether it is an effect of the concept of the social market economy, that requires a more equal balancing between potentially conflicting values such as the economic and the social interests. Or is it only because economic (and political) interests in the richer countries became so heavy that they switched the balance to the social? Both may be true. Firstly, the free movement of services is now more firmly connected to the principle of equal treatment. Equal treatment with local workers in respect of wages protects not only the posted workers, but also the local labour market from unfair competition. In contrast, the market element prohibiting restrictions, which ‘render less attractive the right to provide services’, has now been given less room. This is a new development; so far the Directive and the Court’s interpretation of it relied heavily on the competition argument to assess whether there was a barrier to free movement.72 The newly adopted Directive is therefore a paradigm shift, though to some extent only. This paradigm shift has been challenged by Hungary and Poland, which have brought the Directive before the CJEU for infringing the provisions on free movement of services, which form the legal basis of the Directive.73 In the context of a social market economy, it could be argued that competition between providers now has to take place with larger barriers than before (i.e. equal remuneration where so required), thus giving more room for social aspects. Indeed, according to the new directive in ‘a truly integrated and competitive internal market, undertakings compete on the basis of factors such as productivity, efficiency, and the education and skill level of the labour force (…)’.74 In other words, not on wages.

Secondly, the adoption of the revised Directive shows that where there is a political will, there is a way. Status quo, which risks being the favoured position in a Europe of 28 Member States, can thus be overcome. It has indeed been argued that ‘negative integration favours the negotiation position of Member States with a “Liberal Market Economy” who have no incentive to legislate upward to change the liberalised status quo’.75 Yet, the newly adopted Directive is evidence that a majority can be found for better protection of workers despite the opposition of countries that prefer competition on the basis of wages.

4.4. West against East?

‘Economic’ and ‘social’ are not unambiguous and neutral terms, and their meaning depends to a large extent on the eye of the beholder. For posted workers and their employers, for instance, it may be ‘social’ that they can escape unemployment in their State of origin and contribute to its economy; if they have to be paid the same wage as national workers, it may deprive them of this advantage. Therefore, the issue of social dumping in connection to posting is a complex and sensitive political issue. Likewise, the sending employers may claim that they are not in the same position as employers in the host State as they incur extra cost such as those linked to accommodation and travel and therefore should not be subject to the same rules.76 As mentioned above, Hungary and Poland have challenged the legality of the Directive before the CJEU arguing that it amounts to protectionism and breaches Article 56 of the TFEU. It will be interesting to see how the CJEU will solve the issue and argue for its findings. We expect that the Court will follow the adopted Directive. It has traditionally granted a broad discretion to the EU legislature, and has not been as strict as it is in respect of national measures that restrict the Treaty freedoms. Furthermore, the reference in consideration 3 of the Preamble to Article 3 of the TEU requiring the promotion of a highly competitive social market economy aiming at full employment and social progress and a high level of protection is a powerful counterargument for a rebalancing of the social and economic interests in favour of workers. According to recital 10 of the Directive, ‘ensuring greater protection for workers is necessary to safeguard the freedom to provide (…) services on a fair basis’ (our emphasis). Finally, reference to the horizontal clause of Article 9 of the TFEU requiring that the EU shall take into account requirements linked to the promotion of a high level of employment and the guarantee of adequate social protection when defining and implementing its policies and activities is another strong argument restricting possibilities of competition on wages and requiring equal pay for equal work. The latter principle constitutes one of the social pillars endorsed by the EU to redirect the EMU and the EU in a more social direction.77 To conclude on this issue, the introduction of the concept of market economy pleads for less competition on the basis of wages and for more social protection of all workers including those of the host State.

5. The normative power of the social market economy

What can we conclude from these analyses of the free movement of workers, persons and posted workers in terms of balance between the economic and social?

Free movement of workers has remained a pillar of the EU. It has a very important economic dimension, and exactly because of this dimension it has been included in the Treaty since 1957. Article 45 of the TFEU itself requires equal treatment of migrant workers with national workers in order to make free movement attractive, but also to avoid wage competition and social dumping. In this way, the economic dimension has also social effects. The economic dimension of this provision was an important reason for the CJEU to give the term worker a broad dimension. Yet, there was also a strong social underpinning, as a broad interpretation of the term worker meant that Member States were not allowed to restrict free movement and equal treatment by excluding groups of persons performing work in a dependent position (the Levin judgment).78 Probably, since the economic and social interests were closely connected, equal treatment of migrant workers has not been substantially restricted in the case law. Yet, we saw that some tensions exist at national level threatening the right to equal treatment of workers. It is therefore recommended that the EU legislature takes its responsibility and reinforces the right of equal treatment by defining where and for whom it exists.

Free movement of EU citizens is a more recent phenomenon and is subject to more restrictions than the free movement of workers. This has been the case from the beginning, as one important requirement has always been that the EU citizen should not become an unreasonable burden for the host State. The CJEU’s approach was initially case-by-case based and it considered that a public fund would be strong enough to bear the claim of an individual. In addition, it developed the interesting approach that citizens can be required to demonstrate a link with the host society for accessing some social rights. Such approach constitutes a good middle way between exclusion on basis of nationality and having to admit everyone. This approach requires, however, an extensive investigation by the public administration of each case and raises concerns of legal uncertainty. The CJEU’s more recent rulings in Dano, Alimanovic and Commission v the United Kingdom indicate a shift towards a more restrictive approach. Yet, it is still uncertain whether this case law is already giving the full picture, or that it will be further refined. Can the very claim of social assistance in all cases (so also after a couple of years without claims) imply that one has become an unreasonable burden to the system and therefore no longer has a right to stay pursuant to the Directive? And what is the position of those who do not make a claim, but cannot show that they have sufficient resources? In any event, this case law leaves it unclear to what extent equal treatment still exists in respect of economically inactive EU citizens. The adoption of Directive 2004/38 partly explains this case law, since the Directive contains a few limitations and derogations to the principle of equal treatment. However, the CJEU added an important restriction: it assimilated non-contributory benefits with social assistance benefits, and thus introduced different interpretations of this type of benefit, depending on the piece of legislation involved.

As mentioned above, the earlier requirement of showing a certain degree of integration in the host State was attractive since it found a way between simply excluding all EU citizens on ground of nationality and admitting every citizen to the welfare system from the first day of stay. In contrast, the Dano approach is an absolute one of excluding citizens who claim social assistance and is not further explained. The same outcome could have been reached by means of the link approach or by simply saying that a person not seeking work is not entitled to the non-contributory unemployment benefit that was claimed by Ms. Dano. Dano seems therefore to be more inspired by the fear of welfare tourism expressed by some Member States than by consistent criteria. Equal treatment of economically inactive persons has thus radically shrunk, in any case during the first five years of their stay in the host State. When EU citizens become workers, however, they are treated in the same way as nationals. This is an essential difference. It is clear that there are no economic interests to ensure equal treatment of burdensome inactive EU citizens and that makes their position vulnerable when economic arguments such as the possible costs of their stay are put forward. Still, also free movement of citizens deserves protection and we have made suggestions for the legislature to elaborate the link approach in order to ensure, in any case, equal treatment gradually.

Conflicting economic and social interests could be seen in particular in the area of posting, where the social protection of workers in the host State and that of the posted workers had to give in to the economic rights of the services providers. Yet, unequal treatment in the host State can have a negative effect on public confidence in the EU and on social cohesion in the host State. In fact, the discussion around posting clearly shows the importance of the connection of equal treatment and the free movement of workers in Article 45 of the TFEU in terms of legitimacy of the EU’s action. Although the development towards equal treatment in the revised Posted Workers Directive can be seen as pushing the balance in favour of the social side, it must be recalled that the motives for this shift are also of an economic nature. They are not only economic, since the promotors of this shift also had to do with pressures from workers, unrest and lack of trust in governance. However, these arguments go hand in hand with economic interests, such as losses in competition in the high wage countries and extra costs for social funds, and that explains the pressure to revise the Directive. A remaining problem is that of the service providers and their workers. They may lose their markets in the more affluent countries. This problem cannot be solved in the latter countries, but this has to be addressed in the countries of establishment. EU funds could help to improve the labour on offer and to educate workers, so that competition no longer has to be on wage costs, but can be on particular expertise. This contributes more to a coherent social market economy in the EU than the previous wage competition.

Finally, there are parallels between the issues of benefit tourism and social dumping. Combatting social dumping through posting is combatting unfair competition in order to protect jobs for domestic workers. Combatting welfare tourism is avoiding unreasonable expenses for the State budget, again for the more affluent countries. Both are meant to protect the economy and maintain social cohesion, in any case of the more developed countries, so that jobs and social benefits are not threatened by outsiders. Such actions are largely influenced by the disparities between the Member States and this should not be the end of the development. We have suggested some ways to respond to the various problems; some of these are based on the principle of equal treatment. Yet, as long as the differences between countries are still as important as they currently are, it is more sensible to make adjustments to the present rules as suggested in the preceding sections than radically require equal treatment in all situations. Still, equal treatment should be the final objective and make it possible to come to a better reconciliation of the economic and social.