Introduction
The question of what is to be understood as labour and what is not is closely linked to the emergence of the European Economic Community (EEC) in the 1950s. The Treaties of Rome of 1957, in which the free movement of labour was enshrined as a so-called principle, was followed by not only corresponding debates but also legal disputes before the European Court of Justice (CJEU) in Luxembourg. Initially, most of these involved cross-border migration, as this was where Community law was evidently pertinent. In the long run, however, this narrow understanding would give way to a considerably broader application. While claims to social security benefits by Italian miners employed outside their homeland dominated in the first years following the Treaties taking effect, over the course of the decades a vast variety of welfare and social policy issues would come to be associated with the free movement of labour. As time went on, the trajectory pointed to a broader, both more complex and more flexible understanding of what constituted labour, including an ever-increasing range of activities previously not regarded as work. Students and sex workers, unemployed and sick persons were demanding national benefits through European channels, transcending the boundaries of national welfare state systems and helping re-define labour in the process. The concept of ‘European citizenship’ evolved gradually, eventually becoming reality in the 1992 Maastricht Treaty.Footnote 1
A few legal scholars, but mostly sociologists, have traced this development in morphological terms, depicting the evolution from European workers to European consumers and finally to European citizens.Footnote 2 This approach has been developed further by ascribing a social citizenship status to European Union (EU) citizens, reflecting the Union’s transformation from a European legal area into a socio-political space of benefits.Footnote 3 The call for a re-conceptualisation of the Europeanisation of social policy that would redirect the traditional focus on the Community’s and Union’s elites to citizens and their representatives, if from a largely synchronic perspective,Footnote 4 chimes with the legal–historical approach taken in this article, in recent historical research on EU social policy,Footnote 5 and, specifically, in the contribution of lawyers to European social policy at the CJEU.Footnote 6
The present article does not aspire to a comprehensive account of all cases related to labour migration, or even of all so-called landmark cases,Footnote 7 that have ever been brought before the CJEU. Neither does it offer a history of legal dogma, even though dogmatic problems figure in how events unfolded. Instead, it takes a problem identified by sociologists and social historians,Footnote 8 namely the transformation of ‘labour’ as a political and economic category, and asks how it was rendered into a legal problem and co-shaped by different legal, especially judicial, actors such as the CJEU. The ensuing renegotiation of ‘work’ vis-à-vis ‘non-work’ will be analysed by looking at a blend of cases that all pointed in the same direction, the expansion of an originally clearly circumscribed term, labour, into an encompassing category that covers a wide range of activities and economic statuses.
Based on both the published (CJEU judgements and its advocate generals’ opinions) and unpublished documents of both the Court and the Legal Service of the European Commission, which include the written statements by involved actors to the proceedings, this article will chart this development by studying a sample of cases that arose in Belgium from 1972, when the first case from Belgium arrived at the CJEU, to 1988, when the last case of this selection was decided at the CJEU. Although by no means alone in their effort, Belgian lawyers were particularly active in bringing cases to the CJEU during the 1970s and 1980s, reflecting the country’s manifold and intense intra-European migration networks.Footnote 9 With the (frequent) backing of the advocate generals as well as the Legal Service of the Commission, these lawyers managed to convince the judges in Luxembourg (who were anyhow on an active course to interpret the free movement principle broadly) of rejecting member states’ regular claims for a restrictive understanding of work and, by implication, a wide construction of non-work. The article sketches firstly the evolution of the principle of free movement of workers from the nineteenth century up to the Treaties of Rome. Secondly, the article analyses chronologically as well as systematically how social transformations in terms of pensions, unemployment, disability, sex work and education that gave rise to the legal claims were translated into the language of European law and were at times endorsed, at others rejected by the CJEU. In that process – to take up a point made about law’s prescriptive ambition by Andrei Marmor – the arguments invoked by judges, advocates generals, lawyers and the Commission did not merely reflect changing social realities but also contributed to the change they purported to describe.Footnote 10
The Evolution of a Principle
Long before individuals could bring their legal claims to the CJEU, the complex field that was social policy had developed in a twofold, yet intertwined process. It had taken shape as a central policy field with the emergence and consolidation of European nation states in the nineteenth century, and it had become a both international and transnational, increasingly institutionalised discursive field in the first half of the twentieth century. Because of mass poverty and increased labour mobility that resulted from accelerated industrialisation, social security systems evolved to placate workers’ demands and stabilise societies across Europe, of both democratic and autocratic stripe. Early welfare state legislation became a key component of governmental legitimacy and state authority while marking the dividing line between citizens and immigrants.Footnote 11
National social insurance systems and their bureaucratic institutions placed huge administrative hurdles in the way of and caused very real losses to labour migrants. Victims of work-related accidents often did not receive any pension benefits from social insurances they had already contributed to. Neither could retirees transfer old-age pension rights – if they existed in the first place – to their new or old places of residence. Workers who continuously moved or who commuted to another country for work usually lacked political and legal support to make their claims heard wherever they were.Footnote 12 As a response to these problems, bilateral and multilateral agreements between European countries sought to establish common standards and offer guidelines of how to account for cross-border labour migration in social policy terms. At the same time, the turn of the century coincided with an internationalist turn,Footnote 13 notably the foundation of several international organisations that were devoted to matters of social policy. They dealt with a wide spectrum of labour- and welfare-related concerns from social insurance to labour legislation and policies to alleviate unemployment.Footnote 14
These debates and institutions would feed into the decision to establish the International Labour Organisation (ILO) after the First World War, when the double pressure to reconstruct European economies and provide for the subcontinent’s population in dire times aligned with a general trend towards societal modernisation, activist states and corporatism.Footnote 15 The ILO offered a framework of policy recommendations that member states were called on to implement by means of bilateral and multilateral conventions.Footnote 16 It was within this context that the notion of the free movement of workers took shape, laying the foundation for one of the EEC’s central legal principles. Also, already in these early stages, it became clear that social policy solutions to migration-related problems would oscillate between harmonisation on the one hand and co-ordination on the other.Footnote 17
More than half a century later, the same dual mode of organising social security would enter the Treaties of Rome. While its article 48 laid down the principle of free movement of workers, article 7 prohibited discrimination on the grounds of nationality and article 51 stipulated that social security was to be coordinated by Community institutions.Footnote 18 In order to elaborate the social provisions of the Treaties of Rome and to manage co-ordination, two crucial regulations of the European Council came into force together with the Treaty the same year. Regulations nos. 3 and 4 effectively stipulated five fundamental principles that had already been established by international agreements but would prove salient to the new community’s integrative framework: first, the place of employment determined the responsibility of the respective pension insurance institution. Second, the regulations guaranteed foreign workers treatment on equal terms with national workers within the Community. Third, in case of several benefits claims, the member states were obliged to transfer social security benefits within the Community (transferability). Fourth, employees who had worked in several member states could from now on claim social security in all the countries they had worked; their total benefits titles were to be aggregated (aggregation). Fifth, the pension of a beneficiary who had completed insurance periods in two or more member states was calculated in accordance with the regulation (pro rata temporis).Footnote 19
Five years later, yet another regulation (No. 1612/68) was issued by the European Council that strengthened the legal status of migrant workers and their family members. Its far-reaching implications would only become apparent when courts began applying it. The regulation introduced a non-discrimination clause that guaranteed the same ‘social advantages’ – as it said in the regulation – to both national and foreign workers.Footnote 20 In 1971, Regulations nos. 3 and 4 would eventually be replaced by their remote descendant, Regulation No. 1408/71, in order to adjust the legislative ground to the problems of co-ordination of European social provisions at a time when both geographical expansion and economic crisis were imminent.Footnote 21 None of the various articles and regulations, however, did much to define what was actually at stake; neither was it clarified what kind of work came under the protection of Community law nor which migratory patterns were meant. The underlying assumptions of steady, waged work were unmistakable, yet never made legally explicit.
It therefore fell to the judiciary to fill the gaps, first and foremost the CJEU. After 1957, the CJEU’s case law rendered the applicability of European law more concrete,Footnote 22 especially through the preliminary ruling procedure that integrated national and supranational jurisdiction. Frequently causing dissatisfaction among member state governments concerned about judicial sovereignty, the preliminary ruling’s intermediate procedure made them an integral part of national proceedings that began and ended before the judges of national courts who decided on the merits of concrete cases but left the generally applicable interpretations to the CJEU.Footnote 23 While lower-level courts could decide whether or not to make references to Luxembourg, national courts of last instance were obliged to ask for the CJEU’s ruling on questions that were clearly related to Community law (once again, ‘clearly’ was far from an unequivocal category itself, leaving it up to judges to decide whether or not to involve the CJEU).Footnote 24 Once submitted, preliminary rulings were binding on national courts.
The authority that the CJEU had been given, and its willingness to wield it, became first apparent in two cases in the early 1960s, the famous Van Gend en Loos (1963) and Costa v ENEL decisions (1964). Both appeared highly technical and unlikely to stir controversy, yet they would set into motion ‘revolutionising’Footnote 25 changes in legal discourse and jurisprudential practice. In effect, the CJEU judges ruled that Community law had direct effect, imposing rights and obligations on not only individuals but also national courts.Footnote 26 In Costa v ENEL, the Court doubled down on its previous ruling and stipulated an unambiguous hierarchy: Community law had priority over national law.Footnote 27 These doctrines entrenched the Court’s constitutional rule – and they provided scores of lawyers with food for thought regarding what else could be henceforth achieved by drawing on Community law. That the CJEU’s Hoekstra decision, handed down the same year as Costa v ENEL, concluded that it could very well define ‘workers’ (and by implication ‘work’) without taking recourse to member states’ own classifications showed the way.Footnote 28
After Work: Various Pensions and Unemployment
Pensions conclude careers, but they initiated the CJEU’s role in European social policy and (re)defined ‘work’ along the way. First and foremost, cases of Italian workers in Belgium played a key role in shaping the Court’s jurisprudence. Due to an increased demand for coal to revive the shattered European economies after the Second World War, Achille van Acker, Belgian prime minister and member of the Belgian Socialist Party, called out the ‘bataille du charbon’ not only to increase energy production but also to find the means to fund the Belgian welfare state.Footnote 29 Thus, the Belgium and Italian governments made an agreement in 1946: Italy promised to send 50,000 miners and Belgium would send up to 3 million tonnes of coal on an annual basis in return to Italy. Over the following decade roughly 150,000 workers from Italy found work in coal mines located in the French-speaking Wallonia. This did not include family members of those workers who came to stay in Belgium.Footnote 30 In 1970, roughly 9.5 million people lived in Belgium, of whom 721,000 were foreigners, and with around 40 per cent, Italians made up the largest group of all immigrants.Footnote 31
These entangled relations between Belgium and Italy provided a complex set of challenges for national welfare states vis-à-vis a newly European labour market: both permanence, that is, Italians staying for good, and remigration, that is, workers leaving Belgium again, posed difficulties for social security systems that had never been devised to accommodate transnational lives. European integration forced member states to tackle these challenges and often in ways that were not intuitive to national authorities. As the largest and earliest country to send hundreds of thousands of citizens to other countries, Italy’s emigrants became the pioneers of the case constellations European jurists pondered over. As they frequently paid for the hard labour they did with massive occupational health hazards and subsequent medical problems, these translated into the legal issues of, for instance, invalidity pensions, co-ordination of social benefits and practical cross-border transfers.Footnote 32
Italian miners received support from their compatriot Daniele Rossini and the Christian Associations of Italian Workers (Associazioni Cristiane Lavoratori Italiani; ACLI). Between 1972 until 2002, Rossini brought fifty-three cases to the CJEU to oppose Belgian social security institutions that rejected a broad interpretation of the Treaties of Rome and pertinent regulations. While Rossini and his clients often found support among the Commission, the advocates general and ultimately the CJEU judges, practical help was notoriously late, as clients were seriously ill or passed away before the announcement of the judgement.Footnote 33
There was little evidence that the authors of Community legislation had given the issue of retirement much thought; the point had been to enable the free flow of active labour, not the transfer of legal titles and funds to those who were dropping out of the labour market. Yet against the backdrop of expanding welfare states where benefit schemes were a blend of collective bargaining and state-collected contributions deducted from monthly wage bills, the CJEU treated pensions as evident remuneration for labour provided.Footnote 34 In other words, retirement became an integral part of a lifecycle centred on waged work.
With its underlying notion of literally having been earned, old-age benefits rhymed with wage labour. The same was not true when it came to unemployment, narrowly understood as able workers who were out of paid work. Here, legal views very much depended on normative assumptions of what the dole essentially was: an act of charity; a temporary support to bridge over to the next employment contract and keep workers in the national labour pool; or a legal title that had, just like pensions, been earned through taxed income. Of the three, only the last conception was amenable to migration. At the CJEU, such cases occurred especially throughout the 1980s from the Flemish part of Belgium (and thus was not represented by Rossini). They also played an important part in developing the CJEU’s work-related jurisprudence. The pertinent cases revolved around residence, that is, whether workers could literally take their unemployment benefits and move to another EEC country.Footnote 35
These issues of the entitlement to social security benefits for labour migrants need to be contextualised in the broader evolution of European social policy and migration policy. First, during the 1970s, Belgian lawyer Éliane Vogel-Polsky brought three cases to the CJEU that all concerned the principle of equal pay laid down in the Treaties of Rome and the subsequent Equal Treatment Directive. The three proceedings were based on the unequal treatment of female flight attendants, who were sent into retirement much earlier than their male colleagues. Consequently, their old-age pensions were lower.Footnote 36 As a ‘pathfinder’Footnote 37 in the fight for equal treatment between women and men, Vogel-Polsky inspired lawyers from other EEC member states to bring similar cases to the CJEU in the following decades.Footnote 38 Although some plaintiffs remained unsuccessful in their claims with regard to equal pay and the directive, the CJEU contributed to the application of Community law and the (sometimes slow) amendments of national legislation by member state governments.Footnote 39
Second, the European Council had initiated the so-called Social Dialogue to enhance European social policy, such as intra-community migration in 1972. Belgian Prime Minister Leo Tindemans took this up in his 1975 report proposing that ‘a citizen’s Europe’Footnote 40 could materialise by abolishing controls between member state borders and guaranteeing fundamental rights.Footnote 41 Third, the member states established the European Social Fund to fight unemployment by funding projects supporting vocational training in the mid-1970s.Footnote 42
Furthermore, when the cases of Italian workers arrived at the CJEU, it was a time of tension: in the early 1970s, Belgian and other Western European countries, such as West Germany, stopped taking on labour migrants due to economic decline.Footnote 43 In most cases at the CJEU, besides the plaintiffs, the Commission and the Italian government were in favour of transferability as stipulated in Regulations nos. 3 and 4, while the Belgian government – which had to pay for the respective social security benefits – would interpret the principle of free movement and further regulations narrowly.
Never Working (Again): Disability
From old age, infirmity and unemployment, it was not a big step to ask where member state citizens with neither the physical nor mental ability to do waged work figured in the emerging picture of social benefits within the EEC. Echoing larger contemporary debates, disabilityFootnote 44 became a recurring legal concern of and at the CJEU between 1973 and 1975. In turn, the Court’s extension of the principle of free movement of workers to claimants whose labour-market integration was not even alleged would create legal precedents for a discourse that reconceptualised civic rights in social terms.Footnote 45
The CJEU developed its principles in a mere five cases in the 1970s and 1980s, fourFootnote 46 of which came once again from Belgian courts and concerned Italian immigrants, with the fifth being submitted by a French magistrate. The Belgian cases were a direct result of Belgian reform policy from the previous decade, aiming at giving people with disabilities greater autonomy and enabling them to participate actively in society. Breaking with (Europe-wide) traditions of institutionalising people with disabilities to separate them from mainstream society,Footnote 47 social reformers in the late 1960s aimed at destigmatising disability. Social change translated into legal reform with a 1969 law that explicitly justified financial benefits with the objective of enhancing individual independence.Footnote 48
Among those to whom the new law catered was Odette Callemeyn. Born in France in 1934, she had been residing in Belgium since 1957 and was married to a Belgian citizen. Callemeyn had been employed in her earlier life but received incapacity benefits from the Belgian social security insurance due to a 70 per cent incapacity to work. In 1972, she applied for disability grants with the Belgian Ministry of Social Security (Ministère de la prévoyance sociale) in addition to her impairment benefits. Her application was rejected on account of her French nationality, with the Ministry of Social Security arguing that the new law entitled only Belgian nationals, not foreigners.Footnote 49 Represented by her trade union, the General Labour Federation of Belgium (Fédération générale du travail de Belgique; FGTB), Callemeyn filed a complaint at the labour court (Tribunal du Travail) in Tournai. Her lawsuit was based on the premise that Regulation No. 1408/71, with its extension of a variety of social benefits entitlements to family members of migrant workers, had been breached.Footnote 50 However, the regulation did not explicitly list disability benefits and its scope would thus stand at the heart of the case.
During the proceedings at the CJEU, the Belgian government sought to distinguish strictly between incapacity and disability benefits by ascribing to them different rationales. In their reading, the regulation covered only incapacity (article 4), which was a substitute for lost income. For that reason, the Belgian delegation further elaborated, a specific assessment system existed that evaluated the degree of impairment in relation to a specific occupation, that is, work typically done in the respective profession. The disability grant, on the contrary, was not intended to compensate reduced earning capacity, and thus there was in no relation to prior employment. Instead – and here evidence from the Belgian parliamentary debates of the late 1960s was invoked – the law entitled people who had never been able to work in their lives to receive benefits that guaranteed a minimum subsistence level. In short, Regulation No. 1408/71 was not pertinent, and Callemeyn had missed her mark.Footnote 51
The Belgian government’s distinction was essentially between contributory and non-contributory benefit schemes. Yet the crucial question was about the dogmatic distinction between social assistance and social security, as only the latter qualified under Regulation No. 1408/71. This fundamental legal question – which would also inform the three other cases – was addressed by the Italian government, which, mindful of the role mass emigration played in the domestic economy, seconded Callemeyn’s case. The Italian lawyers recalled the difficulties of definition, as both concepts were of a ‘fluid and evolving nature’.Footnote 52 Older ideas of welfare had been justified as social security, that is, as an individual, legally enforceable right, shedding traditional notions of charity along the way. The Belgian law on disability benefits, the Italian government argued, combined facets of social assistance (no requirement of contributions to social security funds or gainful employment) with enforceability before Belgian courts. This blend, the Italian government concluded, was sufficient to qualify as social security.Footnote 53
The same stand was taken by the Legal Service. It interpreted the Belgian provisions as ‘systems mixtes’Footnote 54 that were enough to invoke Regulation No. 1408/71 and its prohibition of discrimination on national grounds.Footnote 55 In this context, the Commission reminded of an earlier judgement, in which the CJEU had declared that social security and social assistance were in some circumstances difficult to distinguish from each other.Footnote 56 The Legal Service’s stance towards an extension of worker’s rights towards citizenship rights reflected broader European politics. In the 1960s and 1970s, the European Parliament (EP) was committed to the principle of free movement and its broader implication for workers. In its debates, it went ‘beyond Treaty provisions’Footnote 57 by extending it to the people behind the work and also taking into account the conditions of living of women and people with disabilities, to name a few.Footnote 58
Gerhard Reischl, the advocate general in Callemeyn’s case, not only reiterated the reasoning of the Italian government and the Legal Service but also reconceptualised the Belgian law’s meaning. Effectively claiming the primacy of supranational over national interpretation, Reischl ascribed to the disability law a dual function: to guarantee a minimum income for disabled people outside the social security system and to provide an allowance to recipients of social security where it is inadequate; the latter opened the door for applying Community law.Footnote 59 Reischl was rather aware of his interpretative leap. By pointing to past judgements, he claimed that the CJEU had been in favour of dynamic interpretation, as it had ‘always striven not to allow the protection of migrant workers to be defeated by the organisational peculiarities of national systems, but to take account of the emergence of new forms of social protection that cannot be classified in time-hallowed categories’.Footnote 60 The judges agreed and found in favour of Callemeyn.Footnote 61
While Callemeyn’s case ventured further into the realm of protecting people who were not gainfully employed through the Treaties of Rome’s freedom of labour, it was still tied to her previous employment. The constellation in the following two cases of Luciana Costa and Renzo Fracas would loosen the tie to waged work even further. Costa, another Italian who had moved to Belgium in 1956 and was married to a local, had never been gainfully employed when she applied for the disability grant;Footnote 62 that fourteen-year-old, 100 per cent disabled Renzo would never be able to work in his life was self-evident when his parents applied for an ordinary disability grant on behalf of their son.Footnote 63 Both applications were rejected in the early 1970s by the Belgian Ministry of Social Security because of the applicants’ Italian nationality.Footnote 64 Thus, the questions that drove these two cases was whether the wife of a domestic worker and the child of a foreign worker were entitled to receive disability benefits respectively. What was at stake was not employees’ rights, but those of their dependents.
On behalf of Renzo, Daniele Rossini argued that Regulation No. 1408/71 explicitly included family members to improve the living conditions of migrant workers. He referred to the regulation that provided for equal treatment of Community foreigners and nationals (article 3) as well as the principle of free movement (article 48) of the Treaties of Rome, which also ruled out any discrimination based on nationality. Furthermore, Rossini reminded the court that the state’s income also indirectly consisted of contributions from migrant workers.Footnote 65 The Legal Service shared Rossini’s argument and pointed out that the child of a migrant worker who received benefits due to the employee status of the parent was entitled to benefits even after the child came of age.Footnote 66
In contrast to this, the Belgian government argued that the child of a worker could only fall within the personal scope of Regulation No. 1408/71 if it covered the relevant national legislation according to its material scope. This was not the case with the Belgian legislation. The Belgian legislation only provided for a personal entitlement regardless of whether they were employees or family members.Footnote 67
Given that the available statutory law had little to say on the matter, and precedent was missing, the written opinion submitted by Advocate General Alberto Trabucchi appealed to the judges’ sense of justice, rephrasing the argument about the aspirations of European integration:
If we want Community law to be more than a mere mechanical system of economics and to constitute instead a system commensurate with the society which it has to govern, if we wish it to be a legal system corresponding to the concept of social justice and to the requirements of European integration, not only of the economy but of the people, we cannot disappoint the Belgian court’s expectations [that Community law was the solution to Renzo’s problem], which are more than those of legal form.Footnote 68
For Trabucchi, ensuring the welfare of ‘the people’ was the ultimate objective of Community law. Hence, workers’ rights had to be conceptualised comprehensively, including their roles as spouses, parents and families. Workers had a life, and obligations, outside of and beyond work. In June 1975, the CJEU followed Trabucchi’s argument and decided Regulation No. 1408/71 included national systems of entitlements to allowances for disabled persons and that Renzo, as a child of a migrant worker, was as equally entitled as Belgian children. While Trabucchi had argued that in cases such as Renzo’s, children would keep being entitled if the working parent stayed in the country that offered these rights, the CJEU went even further by stating that Renzo, who would never be able to work in his life, should be entitled to benefits even if he came of age.Footnote 69
After a fifth, related case involving a disability grant application by a son of Italian nationals in France had also been decided in favour of the applicant in 1977,Footnote 70 the representative of the Legal Service, Marie-José Jonczy, drew some conclusions for her colleagues at the Commission. The CJEU, Jonczy pointed out, had developed, possibly even replaced, traditional understandings of social security as derivative of paid employment with a new, broader concept that based every claim of social security rights on the question of indigence. Ultimately, rights did not have to be earned through paid work but were an expression of solidarity in and between member states.Footnote 71
Thanks to Rossini, people with disabilities were now included under the principle of free movement at the CJEU. However, this did not lead to their immediate integration into the European labour markets. Throughout the 1980s, the Commission would initiate action programmes to include people with disabilities in education and the Council would give out recommendations to member states to work on employment opportunities for people with disabilities, but it would take until the mid-1990s before people with disabilities were to receive equal opportunities (on paper).Footnote 72
(In)Decent Work:Footnote 73 Sex Work
If workers’ family members had hardly been on the mind of those drafting and signing the Treaties of Rome, terrorists and weapons traffickers, prostitutes and drug users, scientologists, convicted rapists and persons without valid passports certainly had not been either. Yet such labels obscure that these were often people who had been in gainful employment at one point before engaging in shady or outright criminal activities; indeed, some considered themselves workers who happened to hold beliefs, or to do work, considered as immoral in the public eye. And it was in that capacity that a member state prerogative such as extradition or refusal of entry on criminal grounds would none the less come under the EEC’s governing principle. Indeed, national authority in intra-Community migration had been waning for some years. In 1961, a European Council regulation abolished general quotas for work permits for EEC citizens; after five years of legal employment, these persons had the right to a permanent work permit. However, member states were entitled to reject workers from within the EEC when the economy was under pressure. This changed three years later when Directive 64/221 prohibited the practice of expelling labour migrants and their families in times of economic crisis; henceforth, only internal security–related reasons, some of the broad, interpretable categories such as threats to public safety and order, justified extradition. Previous criminal convictions were no longer a sufficient reason for expulsion in their own right, nor were expired passports.Footnote 74
The improved legal status of Community migrants prompted a range of court cases, the first of which was that of a Dutch citizen who had been hired as a secretary by the British branch of the Church of Scientology. In 1973, she was refused entry at Gatwick because Scientology, though not banned, was considered a threat to public policy in the United Kingdom (UK). Her case went from the High Court to the CJEU. The secretary’s lawyers (who were paid by Scientology) argued that the directive allowed rejection of entry only because of individual transgressions, not due to collective suspicions – but lost in Court. Over the course of the following years, various residents in, for instance, West Germany, the United Kingdom and Belgium would make claims of a similar nature to the CJEU to challenge the concept of public policy and its relation to labour migration and personal conduct.Footnote 75
Far from being specific to individual member states or idiosyncratic in their constellations, these early cases proved rather typical of the organisational and legal path-dependencies that national immigration systems had laid out. This was driven home by three cases that came to the CJEU from Belgian courts during the late 1970s and early 1980s. These cases concerned French nationals Josette Pecastaing, Rezguia Adoui and Dominique Cornuaille, all of whom Belgian police suspected to be female sex workers. As elsewhere, sex work was certainly common in Belgium,Footnote 76 but the governing 1948 law stood in an abolitionist tradition: it did not criminalise prostitution as such but rather its commercial exploitation. It also gave local authorities the power to intervene if prostitution was considered to jeopardise public policy.Footnote 77 Behind this stood a notion of unrespectable and abusive work that did not merit professional protection, which was in line with international trends. Sex work had received negative attention in fora of international organisations for decades, and the ILO had never legally recognised sex work employment, as historians Eileen Boris and Magaly Rodríguez García have shown.Footnote 78
Against this backdrop, the three women had come for work to Liège Province, on the French-Walloon border in 1977, 1978 and 1980 respectively. They had entered Belgium lawfully, taking up employment as bar and café waitresses, and had soon after applied for residence permits with the Aliens Office of the Public Security Administration (Administration de la Sûreté Publique, Office des Étrangers). However, their formal employment was considered a mere cover for prostitution by the Belgian authorities, who did not want to make immigration of additional sex workers particularly easy and therefore rejected the applications and decided to deport the three French women.Footnote 79
In an echo of the Scientology case, the decision did not hinge on illegality but on vague concepts such as ‘public policy’ and ‘personal conduct’, as Pecastaing, Adoui and Cornuaille learned when they applied to the Consultative Committee for Aliens (Commission consultative des étrangers) about the reasons for their extraditions.Footnote 80 The Committee, which supervised rejections of residence permits of EEC citizens,Footnote 81 stated in all three cases that the women’s employment constituted a danger to public policy, since they were working in bars that were considered morally repugnant. Aggravatingly in Pecastaing’s case, the police found out that she had been a prostitute in Germany and France previously,Footnote 82 while all of them were reported to have displayed themselves in shop windows.Footnote 83 The Committee therefore issued orders for rapid relocation.Footnote 84 Unfazed, the three women refused to comply with the expulsion orders and instead sought legal counsel from a local law firm in Jacques Levaux and Luc Misson before the court of first instance (Tribunal de Premièr Instance) in Liège.Footnote 85
Misson and Levaux had experience with similar cases,Footnote 86 but the early stages of his work proved inauspicious. On the technical issue of whether, under Community law, Pecastaing could be deported before her case was decided by the Belgian court, the CJEU came out in support of the authorities: member states were not obliged to let foreign nationals stay for the duration of the proceedings, and neither had national courts ‘additional powers’ to suspend ‘measures referred to by the directive or to empower them to review the urgency of an expulsion order’.Footnote 87 Similarly, the attempt in the cases of Adoui and Cornuaille to point to unequal procedural safeguards for foreigners as opposed to Belgian nationals met with stiff resistance from not only the Belgian government but also the governments of the United Kingdom, France and Italy. None of the national bureaucrats was eager to see their civil servants’ independence and impartiality put in doubt; nor did they want to set a precedent that judicial review could be vested in not only national courts but also the CJEU.Footnote 88
Misson and Levaux's arguments found broader support when they turned to substantial points. In the cases of Adoui and Cornuaille, the first issue to be discussed was the interpretation of the term public policy. Given its vagueness, the lawyers suggested, it had to be interpreted ‘strictly and restrictively’Footnote 89 and should in no way lead to discrimination. In dogmatic terms, only behaviour that was covered by the criminal law of the respective country should qualify as a disruption of public policy, and this had come in the form of a clearly defined catalogue of criminal law. Otherwise, migrant workers would not develop a sense of legal security and ‘confidence in the right of residence and establishment’.Footnote 90 Thus turning the tables on the Belgian state that took recourse to imprecise language and arbitrary implementation in rejecting a worker’s right to residence on moral taste rather than criminal grounds, the two lawyers doubled down in Adoui’s case. Invoking the dire consequences such decisions had on individual lives, they asked what a finding against their client would mean if applied to long-term immigrants with family responsibilities (as many Italian migrants in Belgium had), if found in a similar situation. In their reasoning, any such extradition would breach several articles of the European Convention on Human Rights, the Universal Declaration on Human Rights and the International Agreement on Civil and Political Rights. To avoid this, Misson and Levaux brought up the idea of ‘double proportionality’,Footnote 91 which would weigh the public policy against the individual situation of the working migrant.
While the Liège-based lawyer received support from the ranks of the Legal Service,Footnote 92 the Dutch, Belgian, French, Italian and UK governments were unwilling to commit to a clear definition of public policy, as it was precisely its variation over time and space that reflected the ‘progrès de chaque societé’.Footnote 93 From a somewhat less idealistic perspective, national bureaucrats rather preferred remaining in charge of whom to expel and whom to grant the right to stay, and the interpretability of the clause served this goal just fine. That states had no choice but to treat their nationals differently from foreigners could not be helped since the former could not be expelled.Footnote 94 While that argument rather disingenuously missed the very point of Community prohibition of national discrimination, the Belgian government’s retort to Misson’s call for a more generous treatment of individual life decisions betrayed the underlying moral stance. It hardly mattered, the Belgian team suggested, as Adoui was not a long-term migrant with family responsibilities but was young and just living ‘by her charms’.Footnote 95
The second issue at stake before the CJEU was the personal ‘conduct’ of the women, a wording that both highlighted and hid their job’s particularities. In the proceedings, all three women stated that they had tried to find jobs other than as waitresses, readily admitted that their actual work included being on display in windows but denied ‘engaging in prostitution’.Footnote 96 The difference mattered, as being on display was considered only a rather ‘small danger to public policy’Footnote 97 and was indeed permitted in many other places in Belgium. Where it was banned, offences were punished with simple fines. Against that background, the inequity of extradition was rather glaring, especially, as Misson and Levaux added, as the Belgian authorities made no secret of the fact that they systematically expelled French waitresses in bars because of their alleged ties to the criminal scene in France. That, the two lawyers asserted, amounted to universal suspicion rather than the individual examination criminal cases demanded.Footnote 98
The link to organised crimes was not gratuitous but illustrated an awareness in the Belgian government that qualifying sex work as a danger to public policy was hardly consensus. Indeed, their Italian colleagues, in deliberately guarded language, went no further than to acknowledge that prostitution could jeopardise public policy, but that any such assessment was subject to change over time and space, requiring a case-by-case decision.Footnote 99 To counter such relativising arguments, the Belgian authorities justified their actions by painting a picture in which prostitution furthered the ‘milieu international’Footnote 100 more generally – yet notably failed to specify how French sex workers played a more significant role in that than their Belgian colleagues.Footnote 101 This weakness was instantly spotted by the lawyers of Pecastaing, Adoui and Cornuaille; herein lay the violation of European anti-discrimination law by means of enforcing immigration control, as it was not the nature of the work that made the difference, but who did it.Footnote 102
The national governments’ negative stance towards an inclusive interpretation of the principle of free movement translated into legal terms against the background of the directive. To ‘strengthen the enforcement of European law’,Footnote 103 the doctrine of direct effect, meaning the direct applicability of directives, had been pushed forward by the Legal Service as well as the CJEU since the late 1950s and the mid-1960s. In the context of the Van Duyn case, when the CJEU had decided that individuals could in certain constellations claim Directive 64/221 directly at national courts, the British parliament had rejected the CJEU’s decision because it would contribute to blurring the lines between regulations (that were directly applicable) and directives (that needed implementation by national legislators) and would thus create legal uncertainties.Footnote 104 Over the course of the following decade, the Legal Service and the CJEU judges would continue working towards the direct applicability of directives but received backlash from member state institutions, including the French Conseil d’Etat.Footnote 105 In the late 1970s, this rejection made the Legal Service as well as the CJEU become more cautious and restrictive to interpret directives as directly applicable due to its concerns of a resistance of the member states as well as national courts.Footnote 106
The Legal Services’ more cautious approach became also apparent in the cases of Pecastaing, Adoui and Cornuaille, when it chose a middle way. The Commission’s lawyers supported accounting for differences in moral and sexual conceptions throughout the EEC but emphasised that the connection between the individual and the criminal milieu needed to be investigated thoroughly rather than being generally assumed.Footnote 107 This stance reflected the tension the Commission had to handle between member states’ interests as well as the enforcement of the principle of free movement. At the Paris summit of 1974, the national governments had decided to establish a passport union to accomplish a Europe for citizens. While the Commission had proposed to harmonise the different immigration laws of the member states, it stopped working on the matter in the subsequent years. In 1981, in parallel with the proceedings at the CJEU, the European passport had been introduced, and three years later, the Commission proposed to gradually relax internal border controls, receiving support from the EP.Footnote 108 However, the European Council disagreed on the practical terms, fearing to lose control of migration.Footnote 109 Only in the mid-1980s, under the aegis of the governments of France, West Germany, Belgium, the Netherlands and Luxembourg, was the Schengen agreement signed to gradually abolish borders. To stay in control of migration, the UK government did not sign the agreement and would thus not join the Schengen Area that came into effect in 1995.Footnote 110
Misson’s argument struck a chord with Advocate General Francesco Capotorti, a lawyer with a strong interest in human rights matters.Footnote 111 In contrast to the national governments, the advocate general agreed with the plaintiffs that procedural rights had indeed been violated. Capotorti doubted that the administrative division of power between executive and advisory bodies in this case met the standards that informed Directive 64/221. And he was not convinced either that the procedures truly were no less favourable to foreigner than to Belgium’s own nationals. He also suggested that labour migrants from within the Community should have the right to reapply for a permanent residence permit. On substance, Capotorti sought to walk the tightrope between respecting member states’ own prerogatives for guaranteeing public policy on the one hand and upholding the principle of freedom of movement on the other. If the one was not to come at the expense of the other, proportionality was key, and this meant that the personal consequences of deportation on the migrant’s life must be considered. While not disputing national governments’ claim that behaviour that was not punishable by law could also endanger public policy, Capotorti discounted universal suspicion as insufficient; only the individual case counted – and in the case of Adoui and Cornuaille, a concrete link to the crime had yet to be proven.Footnote 112
Thus, backed up by the advocate general, Misson and Levaux's overall argument managed to convince the judges that national authorities were no longer at liberty in dealing with EEC citizens. Procedurally, the Court shared the view that the measures taken against Pecastaing, Adoui and Cornuaille were based on a general rather than an individual assessment. Substantially, it stipulated that what national authorities thought about the nature of their profession had little to with their rights under Community law. Member states’ interest to preserve a specific definition of ‘public policy’ lay within their own discretion, but they could not be used to discriminate arbitrarily. And while the Court did suggest that member states were entitled to consider prostitution socially harmful (for what that category was worth), they had to be consistent in how they defined and handled it. It could not be one thing – dubious, but ultimately tolerable work – for their own citizens, and another – criminal behaviour – for EEC migrant workers.Footnote 113 This was not the same as putting sex work on an equal footing with other types of employment; that was left to the mostly conservatively inclined member state governments. But the CJEU reminded that sex workers from other member states had to be treated equally both with an eye to the substance and to procedural safeguards. In effect, what the CJEU said was that, in terms of intra-Community labour migration, sex work was more about work than it was about sex.
Before Work: Education
As with social policy, education fell almost completely into member states’ responsibility. The Treaties of Rome had laid down that the Commission and the European Council would establish working groups on education and vocational training, and cooperation would gain momentum in the 1970s, particularly with an eye to the mutual recognition of diplomas and the mobility of students.Footnote 114 Yet there were few, if any, firm commitments to integrating fellow member states’ citizens into what continued to be educational systems that were essentially conceived of in national terms. In Smithian terms, education was (and is) about improving human capital through the allocation of public funding, that is, taxpayers’ money.Footnote 115 Two conclusions could be drawn from that: one was to prioritise domestic apprentices and students, the other was to improve qualification and employment through better education. Against the backdrop of European labour market integration, these two conclusions were potentially conflicting.Footnote 116
This tension came to the fore in a number of cases that were brought to the CJEU once more by ACLI’s lawyers as well as Luc Misson. The plaintiffs and their lawyers all directed their claims against national legislation that excluded European Community citizens from, or limited their access to, domestic education and vocational training. Prominent among these proceedings were a couple of cases that concerned children of migrant workers whose applications for education grants had been rejected on grounds of their nationality. In West Germany, these grants – according to the 1971 Bundesausbildungsförderungsgesetz (Bafög) – could be sought by students who attended secondary schools and universities and had limited family resources to draw on.Footnote 117
One of the first cases that linked the principle of free movement and education was the case of Donato Casagrande, an Italian national who had applied for an educational grant for finishing his secondary education in Munich after his father’s death in 1971, only to be rejected. Before German courts and, eventually, the CJEU, ACLI’s lawyer invoked article 12 of Regulation No. 1612/68, which stated that ‘the children of a national of a member state who is or has been employed in the territory of another member state shall be admitted to that State’s general educational, apprenticeship and vocational training courses under the same conditions as the nationals of that State’.Footnote 118 Excluding non-Germans constituted a significant disadvantage for children of migrant parents. It increased their risk of dropping out of school and getting a qualified job.Footnote 119 The CJEU’s judges sided with ACLI’s understanding of education: more than simply implying equal formal admission to labour markets, any financial support that was crucial to entering these markets on an equal footing came under the purview of the discrimination prohibition.Footnote 120
The European headaches did not stop there, and education-related lawsuits continued to arrive at the CJEU also from Belgium. While the ACLI pursued a course of advocating the extension of rights for family members of resident workers, Misson worked continuously on widening the scope of the Treaties of Rome’s prohibition for students. This needs to be contextualised within two major, interlinked trends. First, Belgium experienced an economic decline in the second half of the 1970s, which the ruling Flemish Christian Democratic Party countered with austerity measures in the 1980s.Footnote 121 Second, since the early 1960s, Belgium had been federalising gradually into four language areas that would significantly shift competences from the national government to the regional ones in the second state reform in 1980.Footnote 122 General state-funded education had been introduced in 1959. Primary and secondary education was free of charge, and public institutions of higher education were only entitled to charge low registration fees to cover the administrative costs they incurred. In 1976, however, a new law of budget on education authorised the Minister of Education to introduce an enrolment fee for pupils and students with non-Belgium citizenship if their parents had resided outside of Belgium since 1976. The new rule did not hit children of labour migrants in Belgium but targeted educational cross-border migration, notably from France where numerus clausus legislation restricted access to university studies significantly.Footnote 123 In a circular of the 1985 education budget law, the Minister of Education announced that foreign students in the arts had to pay an enrolment fee.Footnote 124 In the course of the 1980s, Misson would represent several French students’ claims against Belgian educational institutions in an attempt to end the enrolment fee practice for foreign students and to recover previous payments.Footnote 125
In the proceedings of the first such case at the CJEU, that of Francoise Gravier, an aspiring comic artist, the Belgian government (and its constituent body, the Communauté Française) justified the imposition of tuition fees with an increase in foreign students in Belgium. This had put pressure on the education budget, which was not offset by tax revenue, as foreign citizens did not pay taxes in Belgium. Turning Misson’s discrimination argument on its head, the authorities suggested that the introduction of tuition fees therefore counteracted the discrimination against Belgian nationals.Footnote 126 In that, the Belgian government received support from its Danish and British peers who accepted that tuition fees would be unjustified in the case of children of migrant workers but drew a line at the ‘too general’Footnote 127 exemption of all EEC students.Footnote 128 Significantly, the Belgian authorities’ effort to disqualify cartoon drawing as vocational training failed to convince even the local judgeFootnote 129 – maybe not all that surprising in a country that took great pride in the prestige of its bande dessinée industry.Footnote 130
In contrast to to the national governments and in line with Misson, the Legal Service found that by paying tuition fees, non-Belgian students were discriminated against on the grounds of the Treaties of Rome (article 7). Furthermore, the principle of free movement encompassed vocational training.Footnote 131 This reflected Community education policy in the early 1980s. While up until then, education had been part of ‘culture’ and thus not part of the Treaties of Rome, it was only in 1981 that the Commission’s Directorate-General V for Employment and Social Affairs became responsible for education.Footnote 132 This change recognised the economic dimension of education, at a time when youth unemployment was high throughout the Community. This trend also reflected the shift from manufacturing to services in the European labour markets. Thus, higher education became more important to national economies as well as social policies.Footnote 133
Gordon Slynn, the advocate general assigned to the Gravier case, was receptive to Misson’s and the Legal Service’s arguments, notably in light of the attention to education. Community institutions had been paying their effort for years to enhance mobility within the Community, to improve life of employees and, ultimately, to contribute to economic growth. According to Slynn, the education policy in the Community evolved gradually over time. In article 128 of the Treaties of Rome, the European Council had laid down the principle to harmonise the different education policies among member states. In 1963, the European Council had laid down a guarantee that everyone should be able to choose education and profession freely. Then, in 1968, the regulation had stipulated that labour migrants and their children should have the same access to education as nationals of the respective host country. During the 1970s, the European Council had repeatedly called for cooperation between member states regarding education; indeed, the education ministers had agreed on common measures to improve young people’s chances on the labour market, among them measures towards vocational education policy adopted in 1983.Footnote 134 Combining these jigsaw puzzle pieces, Slynn arrived at the following definition:
Vocational training is such form of education as prepares for and leads directly to a qualification for a particular profession, trade or employment or which provides the necessary training and skills for such profession, trade or employment where no formal qualification is available, and which is over and above that given by general education.Footnote 135
In effect, learning to draw cartoon strips was a perfectly valid professional qualification, a conclusion with which the judges would agree. They also ruled the Belgian law was discriminating against students of EEC countries compared to Belgian students because the former had to pay fees.Footnote 136
Keeping the argumentative link to potential employment that was needed to invoke the free movement of labour while decoupling it from any concrete professional training, Slynn’s definition would become the basis for several related proceedings before the CJEU the years to come. It cut to the heart of the three next cases arriving from Belgium, which was the question what kind of education or vocational training was affected by Community law. The constellations were roughly the same as in (and indeed inspired by) that of Gravier: the complaints formulated by Misson on behalf of Bruno Barra and Vincent Blaizot (each representing a larger number of students) also concerned French students who had enrolled in Belgian institutions of higher education. Barra had taken up training in the armoury department of the municipal institute for training in precision mechanics, armoury and watchmaking (Institut communal d’enseignement technique de la fine mécanique, de l’armurerie et de l’horlogerie) in Liège, essentially a vocational school.Footnote 137 While the latter fact was undisputed, Slynn made it a point to remind the court and the parties to the case that the level of institution was irrelevant: it was about the type of education that was on offer. He then went on to denounce the Belgian government’s insistence that Barra’s school did not offer vocational training but secondary school education; vocational training was about gaining qualifications for a future job – even if they were taught in secondary school.Footnote 138
In the third case, that of the French veterinary student Vincent Blaizot, education was once more up for definition. In Belgium, the study programme of veterinary medicine was divided into two three-year stages; the first consisted of basic studies leading to an intermediate degree; then followed by postgraduate studies that were concluded with a diploma.Footnote 139 While the Belgian authorities argued that only the postgraduate study programme formed part of vocational training, because it taught practical skills as opposed to the general knowledge offered in the undergraduate studies, Slynn again begged to differ: general education was needed to proceed with the more specific training that was part of the postgraduate studies. Thus, the whole six years had to be seen holistically.Footnote 140
In its eventual decisions of 1988, the CJEU followed the advocate general’s guidance and found that the tuition fee discriminated against foreign students from within the Community. However, the judges also proved to be astute tacticians; appreciating the resistance Gravier’s retroactivity had met with in Belgium, Blaizot stipulated that tuition fees for EEC citizens were obsolete henceforth – but there would be no general reimbursement.Footnote 141 After all, it was better to compromise than to be ignored.
Conclusions
The tuition fee cases have been called ‘judicial creations of great boldness’Footnote 142 for their determination to interpret the Treaties of Rome in a way that was both ‘Community-friendly’ and intent on enhancing working Europeans’ rights. In that they stand pars pro toto for a period in which the expansive dynamics of the CJEU’s jurisprudence on social policy led to significant advances in the scale and scope of social security and other welfare state provisions to countless numbers of Community citizens, both nationals of the states and foreign workers whose lives had taken them to more than one place. Starting out from a fairly simple and static understanding of the freedom of movement of workers, which focused on employed, industrial wage labour, the Court’s jurisdiction contributed to an expansion that covered ever more groups initially not considered to come under this rule: pensioners, people with disabilities, (alleged) criminals and students.
Likewise, equal pay and equal treatment requirements were, in a long series of complaints (not all of which were successful), further developed to encompass a growing number of working and non-working situations, stretching from access to the labour market to old age. As working biographies and family patterns changed, the ‘worker’ became a broad term that accounted for work lifecycles, covering everything from education to retirement.
Crucial to this development towards greater integration was a constellation that resulted from the confluence of a specific historical situation with highly contingent individual factors: on the one hand the massive growth of European welfare states, the establishment of an intra-European labour migration regime, the ageing of a first post-war generation of migrant workers as well as the coming-of-age of second-generation migrants and the expansion of educational opportunities; on the other, a dynamic Commission at Brussels, organisational networks through associations and trade unions, an emerging cohort of specialised lawyers such as Misson and responsive lawyers on CJEU’s bench as well as advocate generals, such as Slynn.
If the conditions were benign for social policy integration, its evolution was incremental. National sovereignty had to be curtailed, a thicket of member states’ financial reservations had to be navigated, and national traditions of defining welfare and social benefits had to be overcome. In this process, re-defining the division line between what work was, and what it was not, proved to be key. The narrow concepts that had informed the original treaty-making, intending to supply labour markets with usually male, industrial, waged workers, proved increasingly inept and were challenged by the realities of people’s mobility in Western Europe. The cases brought to the CJEU exploded these confines in both scale and scope, affecting a greater number of people while widening the notion of work itself.
Lawyers rarely make for revolutionaries, yet the trajectory of the CJEU’s decisions had revolutionary implications that would contribute manifestly to the growing idea of European Community and Union citizenship. In fleshing out the Treaties of Rome’s modest provisions, the lawyers involved effectively laid bare that the capitalist notion of labour was a fiction: it ignored that workers were citizens and as such complex social actors, with beliefs and biographical twist and turns, with families and friends, in short: with lives. In that sense, the CJEU merely validated in legal terms what Swiss playwright Max Frisch had observed as early as 1965: European states had called for workers, just to find that it was people who came.Footnote 143
But whereas Frisch had indeed had Italian ‘guest workers’ in mind, immigrants from Yugoslavia, Turkey, Algeria and other non-member states were also to arrive to take up work in the EEC. And to them the trend towards broader, more liberal understandings of ‘work’ would meet its limitations. This was not because claims to schooling and pensions were deemed non-pertinent when applicants who did not hold EU citizenship asked for it, but because such claims remained, at the end of the day, derivative of the right to access the inner-European labour market granted by national laws only.Footnote 144 From that angle, European citizens’ inclusion appeared as a privilege (though not an irreversible one, as the 2016 Brexit referendum with its explicit goal of bringing down EU immigration to the United Kingdom showed) and the flip side of other migrants’ second-rate status or, indeed, their wholesale exclusion from European principles.
Mala Loth is a postdoctoral fellow at ARENA Centre for European Studies at the University of Oslo. Her research focuses on the history of European law.
Acknowledgements
The author would like to thank the three anonymous reviewers for their constructive comments, as well as Annelie Ramsbrock and Kim Christian Priemel for their thorough reading of earlier drafts of this article.
Open access
